Supersaturation as a driving force for drug absorption ...€¦ · mucin interaction on drug absorption is likely difficult to predict and may be expected to vary for different drug

i

Supersaturation as a driving force for drug absorption from

colloidal lipid species in the intestine

A thesis submitted for the degree of

DOCTOR OF PHILOSOPHY

from

Monash Institute of Pharmaceutical Sciences

Monash University (Parkville Campus)

by

YAN YAN YEAP

BPharm (Hons) Monash University

March 2013

Drug Delivery, Disposition and Dynamics

Monash Institute of Pharmaceutical Sciences

Monash University (Parkville Campus)

381 Royal Parade, Parkville,

Victoria 3052, Australia

Saturday, 7 September 2013 To: Prof Shinji Yamashita, Setsunan University Re: PhD thesis by Yan Yan Yeap Dear Prof Yamashita, Please find below the response to questions raised for the thesis entitled: Supersaturation as a driving force for drug absorption from colloidal lipid species in the intestine

by Yan Yan Yeap

The comments and feedback by Prof Yamashita are highly appreciated. I am honored that the Professor took his time to read the thesis thoroughly, and was interested enough to ask questions. It is my pleasure to provide the answers to the questions here. 1. In Chapter-3, author clearly demonstrated that receptor mediated lipid uptake does not play important roles to promote the absorption of model PWSD, cinnarizine (CIN). However, in Chapter 5, it was concluded that lipid absorption stimulates drug supersaturation and enhance the absorption. It might be possible to consider that the receptor mediated lipid uptake also promotes drug supersaturation by reducing the solibilisation capacity of LBF. Please make some comments on this possibility. The Professor raises an excellent point where receptor-mediated lipid absorption is potentially an inherent mechanism where drug supersaturation and absorption is promoted. I very much agree with the point raised, and add that the phenomenon will most likely apply for lipids that contribute significantly to colloidal drug solubilization capacity when incorporated into mixed micelles e.g. long-chain fatty acid (LCFA) more so than cholesterol. The experiments in Chapter 3 were early experiments that were carried out before the possibility of lipid absorption-induced drug supersaturation was hypothesized and tested. Therefore, in those experiments, we only tracked the systemic availability of lipids whose absorption was most likely facilitated by the receptors investigated (i.e. cholesterol for SR-BI and NPC1L1, oleic acid for CD36). For cholesterol, while its presence in vesicles is thought to stabilize the bilayer, it generally is not thought to contribute significantly to the drug solubilization capacity of colloids. Thus, cholesterol absorption is not expected to be a significant trigger for drug supersaturation and absorption. This is reflected in Fig 3.2A & Fig 3.2B of the thesis where cholesterol absorption is almost completely abolished by ezetimibe co-administration, but drug

absorption remains unchanged. On the other hand, we now know that the presence of oleic acid in bile micelles contributes significantly to the colloidal solubilization capacity for a range of lipophilic drugs. Thus, receptor-mediated absorption of oleic acid will likely stimulate drug supersaturation and absorption too. Some evidence that support this notion may be found in Fig 3.2A & Fig 3.2C of the thesis, where the absorption of oleic acid and cinnarizine appeared to be concomitantly reduced (albeit insignificantly) by CD36 inhibitor co-administration. At the time of the experiments, we did not think that the difference was significant enough to warrant further investigation. After the findings in Chapter 5 of the thesis however (i.e. LCFA absorption is a significant trigger for drug supersaturation and absorption), it seems entirely plausible that receptor-mediated LCFA uptake could also enhance drug absorption via the stimulation of drug supersaturation. At present, the absorption of LCFA is thought to occur mainly via facilitated diffusion at low concentrations, and passive diffusion at high concentrations [1]. Thus, at high LCFA load (as is typically found in the small intestine after lipid ingestion), the physiological significance of active LCFA absorption is questionable. In addition, it is still unclear if the receptors indicated to be involved in LCFA absorption (e.g. CD36 [2], FATP [3], SR-BI [4]) are indeed transporters which facilitate uptake across the apical membrane, or proteins that facilitate intracellular lipid trafficking/modify signalling processes that mediate lipid absorption [5-6]. Given these uncertainties and complexities, we were not able to adequately probe the role of receptor-mediated lipid absorption in drug supersaturation/absorption. However, in light of the comment made by the professor, it may be interesting to note the data in Fig 5.2A of the thesis, where the attenuation of the UWL acidic microclimate did not completely abolish the absorption of oleic acid, and significant amounts of oleic acid was still absorbed. This could be due to the receptor-mediated absorption component for oleic acid, and it remains highly possible that if this component of lipid absorption was also inhibited, drug absorption may be reduced further. 2. In Fig. 3.7, CIN supersaturation terminated at 1-20min (in the case of Vesicle) or >20 min (in the case of Micelles) after bile addition, however in both cases, decreased concentration of CIN becomes plateau at significantly higher level than equilibrium solubility. Generally, this phenomenon is understood as drug precipitated as amorphous or unstable polymorph. However, using polarized light microscopy, author clearly showed that the CIN precipitated in the crystalline form (Fig. 3.11). Could author explain the reason of the higher plateau concentration of CIN than the equilibrium solubility? On several occasions, the remaining mixture at the end of the precipitation kinetics experiments (colloids + bile + precipitated drug) was collected, and daily samples (after centrifugation) were taken to track the rate and extent of drug precipitation. We found that while the majority of drug precipitation occurred within the first day of mixing (as shown in Fig 3.7 of thesis), drug concentration in the supernatant took a few days to decrease to equilibrium solubility. At present I do not have an explanation for this phenomenon (where non-instantaneous equilibration to the equilibrium solubility is observed), but speculate that intestinal colloids such as mixed micelles and vesicles may have supersaturation-stabilizing properties. This would be consistent with recent studies from our group that showed that during simulated lipid digestion experiments, stabilization of drug supersaturation was typically possible at and below a supersaturation ratio (SS ratio) of 3, while rapid precipitation was typically observed at SS ratios > 3 [7]. In the experiments depicted in Fig 3.7 of thesis, it is likely that the high degree of drug supersaturation induced immediately after the addition of bile (SS ratio of 6 and 12 for micelles and vesicles,

respectively) was not sustained by the colloidal phases, leading to rapid and significant precipitation. The lower degree of drug supersaturation (SS ratio < 2 for both micelles and vesicles) that resulted after the majority of the precipitation occured, however, was able to be stabilized for longer periods of time by the colloidal mixture, giving rise to the observation where drug concentration appears to ‘plateau’ at a concentration higher than the equilibrium solubility. 3. Please speculate the possibility of LCFA colloids to interact with the components of mucin layer. I wonder if the LCFA colloids interact with mucin, which might prevent the diffusion of colloids in UWL and affect the absorption of drugs included. The possibility of LCFA colloids interacting with mucus components is an interesting one and is an area that might be usefully explored in detail. The ability of mucus glycoprotein to bind with a variety of lipids does suggest the possibility that the diffusion of LCFA colloids may be impeded by mucus-LCFA interaction. It follows that the diffusion of colloid-solubilized drugs may be slower in the mucin layer as well, although this will be in contradiction to previous modeling efforts that report enhanced solubilizate diffusion across the UWL when solubilized in bile micelles [8-10] (the UWL is often thought to be indistinguishable from the mucus layer). In the interest of discussion however, if we assume that LCFA colloid-mucin interactions do lead to slower diffusion of solubilized drug, whether or not that is beneficial or detrimental to drug absorption remains unknown. On one hand, if the diffusion of poorly water-soluble drugs through the aqueous mucin layer is indeed the rate-limiting step to drug absorption, mucin-LCFA interaction may then be expected to reduce drug absorption. However mechanistically we understand very little what happens at the mucus/UWL boundary and so it is too early to make predictions. In the presence of mucin-LCFA interaction, it is plausible that enhanced retention time in close proximity to the absorptive membrane could serve to enhance the absorption of somewhat permeability-limited drugs, or, increase the likelihood of acidic microclimate-triggered lipid absorption (which, as shown in Chapter 5 of thesis, is an important driver for drug supersaturation and absorption). In addition, analogous to lipid absorption-induced drug supersaturation, it is also possible that LCFA-mucin interaction serves to deplete LCFA from mixed micelles, and stimulate drug supersaturation as a result. The net effect of LCFA colloid-mucin interaction on drug absorption is likely difficult to predict and may be expected to vary for different drug types. Nonetheless, available literature and practical experience seem to suggest that co-administration of a lipophilic drug with precursors of LCFA often lead to enhanced drug absorption. I am therefore inclined to think that LCFA colloid-mucin interactions, if present, are in fact beneficial to the absorption of lipophilic drugs, although this judgment is made with reservations due to the lack of available scientific data. 4. Since LBFs performance in the GI tract to produce supersaturation and enhance the absorption of PWSD profoundly depends on the endogenous lipid processing events, conversely, it might become a factor to cause inter- and/or intra individual differences in drug absorption. Please make comments on the possibility of such risks of LBF system. I agree that it is possible the reliance on endogenous events to produce the necessary enhancement in drug thermodynamic activity may be a contributing factor to intra/inter-individual variability in drug absorption. However, it is well-known that the body is extremely efficient and robust in its ability to process, digest and absorb dietary lipids. Therefore, it is debatable that the

dependence of LBF on endogenous lipid processing events to promote drug absorption is any more variable than other processes that may underpin drug absorption (e.g. dose form disintegration, drug dissolution etc), or any more variable than other formulation approaches. It is worth remembering that first and foremost, LBF increases drug absorption by circumventing the need for drug dissolution in the GI tract, by increasing the mass of drug that may be present in solution in the small intestine, and by increasing the efficiency of drug presentation to the absorptive membrane (via micellar solubilization). These processes likely vary little between healthy individuals. However, in the interest of discussion, if LBF’s dependence on endogenous lipid processing events to produce drug supersaturation was indeed a significant cause for intra/inter-individual variability, the susceptibility to variability is likely different for different drug types, since different drug types may have different propensity for supersaturation to be induced by endogenous lipid processing events (one of the conclusions in Chapter 4 is that endogenously triggered supersaturation is more apparent for weak bases than neutral compounds). Therefore, for compounds where high variability in absorption may be expected, LBF may be more usefully directed to increase the bioavailability of drugs with broad therapeutic windows. References

1. Chow, S. L.; Hollander, D. A dual, concentration-dependent absorption mechanism of linoleic acid by rat jejunum in vitro. Journal of Lipid Research 1979, 20, (3), 349-56.

2. Nassir, F.; Wilson, B.; Han, X.; Gross, R. W.; Abumrad, N. A. CD36 is important for fatty acid and cholesterol uptake by the proximal but not distal intestine. Journal of Biological Chemistry 2007, 282, (27), 19493-19501.

3. Stahl, A.; Hirsch, D. J.; Gimeno, R. E.; Punreddy, S.; Ge, P.; Watson, N.; Patel, S.;

Kotler, M.; Raimondi, A.; Tartaglia, L. A.; Lodish, H. F. Identification of the major intestinal fatty acid transport protein. Molecular Cell 1999, 4, (3), 299-308.

4. Bietrix, F.; Yan, D.; Nauze, M.; Rolland, C.; Bertrand-Michel, J.; Coméra, C.; Schaak, S.;

Barbaras, R.; Groen, A. K.; Perret, B.; Tercé, F.; Collet, X. Accelerated lipid absorption in mice overexpressing intestinal SR-BI. Journal of Biological Chemistry 2006, 281, (11), 7214-7219.

5. Nauli, A. M.; Nassir, F.; Zheng, S.; Yang, Q.; Lo, C. M.; Von Lehmden, S. B.; Lee, D.; Jandacek, R. J.; Abumrad, N. A.; Tso, P. CD36 Is important for chylomicron formation and secretion and may mediate cholesterol uptake in the proximal intestine. Gastroenterology 2006, 131, (4), 1197-1207.

6. Tran, T. T. T.; Poirier, H.; Clément, L.; Nassir, F.; Pelsers, M. M. A. L.; Petit, V.; Degrace, P.; Monnot, M.-C.; Glatz, J. F. C.; Abumrad, N. A.; Besnard, P.; Niot, I. Luminal lipid regulates CD36 levels and downstream signaling to stimulate chylomicron synthesis. Journal of Biological Chemistry 2011, 286, (28), 25201-25210.

7. Williams, H. D.; Sassene, P.; Kleberg, K.; Calderone, M.; Igonin, A.; Jule, E.;

Vertommen, J.; Blundell, R.; Benameur, H.; Müllertz, A.; Pouton, C. W.; Porter, C. J. H.

Toward the establishment of standardized in vitro tests for lipid-based formulations, Part 3: Understanding supersaturation versus precipitation potential during the in vitro digestion of Type I, II, IIIA, IIIB and IV lipid-based formulations. Pharmaceutical Research 2013, DOI 10.1007/s11095-013-1038-z.

8. Westergaard, H.; Dietschy, J. M. The mechanism whereby bile acid micelles increase the rate of fatty acid and cholesterol uptake into the intestinal mucosal cell. Journal of Clinical Investigation 1976, 58, (1), 97-108.

9. Miller, J. M.; Beig, A.; Krieg, B. J.; Carr, R. A.; Borchardt, T. B.; Amidon, G. E.; Amidon,

G. L.; Dahan, A. The solubility–permeability interplay: Mechanistic modeling and predictive application of the impact of micellar solubilization on intestinal permeation. Molecular Pharmaceutics 2011, 8, (5), 1848-1856.

10. Sugano, K. Estimation of effective intestinal membrane permeability considering bile

micelle solubilisation. International Journal of Pharmaceutics 2009, 368, (1–2), 116-122.

I hope my response to the questions have been satisfactory, but please do not hesitate to contact me if I can provide further clarification. Sincerely,

Yan Yan Yeap, BPharm (Hons)

Current address Department of Chemical Engineering Northeastern University, Boston, Massachusetts, USA Previous address Drug Delivery Disposition and Dynamics Monash Institute of Pharmaceutical Sciences Monash University (Parkville campus), Melbourne, Victoria, Australia

ii

Notice 1

Under the Copyright Act 1968, this thesis must be used only under the normal conditions of

scholarly fair dealing. In particular no results or conclusions should be extracted from it, nor

should it be copied or closely paraphrased in whole or in part without the written consent of the

author. Proper written acknowledgement should be made for any assistance obtained from this

thesis.

Notice 2

I certify that I have made all reasonable efforts to secure copyright permissions for third-party

content included in this thesis and have note knowingly added copyright content to my work

without the owner’s permission.

iii

昨夜西风凋碧树。独上高楼，望尽天涯路。

衣带渐宽终不悔，为伊消得人憔悴。

众里寻他千百度，蓦然回首，那人却在灯火阑珊处。

- 王国维《人间词话》“三种境界”

iv

TABLE OF CONTENTS

ABSTRACT .............................................................................................................................................. ix

GENERAL DECLARATION ................................................................................................................... xi

ACKNOWLEDGEMENTS ...................................................................................................................... xii

PUBLICATIONS .....................................................................................................................................xiv

COMMUNICATIONS .............................................................................................................................. xv

LIST OF ABBREVIATIONS ...................................................................................................................xvi

CHAPTER 1 : GENERAL INTRODUCTION ................................................................... 20

1.1 STATEMENT OF THE PROBLEM ............................................................................................... 21

1.2 LIPIDS FOR THE ENHANCEMENT OF ORAL DRUG ABSORPTION .................................... 21

1.2.1 Poorly water-soluble drugs and the Biopharmaceutics Classification System ................... 22

1.2.2 Lipid-based formulations used in oral drug delivery .......................................................... 23

1.3 OVERVIEW OF LIPID DIGESTION AND ABSORPTION ......................................................... 30

1.3.1 Lipid digestion in the oral cavity and stomach ................................................................... 30

1.3.2 Lipid digestion and solubilisation in the small intestine ..................................................... 31

1.3.3 Lipid absorption across the enterocyte apical membrane ................................................... 37

1.3.4 Lipid transport into the systemic circulation ...................................................................... 38

1.4 MECHANISMS BY WHICH LIPIDS ENHANCE THE ORAL BIOAVAILABILITY OF

DRUGS ........................................................................................................................................... 40

1.4.1 Enhancement of drug dissolution in gastrointestinal fluids ................................................ 42

1.4.2 Enhancement of apparent drug solubility in intestinal fluids ............................................. 43

1.4.3 Enhancement of intestinal permeability and inhibition of enterocyte-based efflux

transporters and metabolism ............................................................................................... 44

1.4.4 Promotion of lymphatic drug transport ............................................................................... 47

1.5 PROPOSED MECHANISMS OF DRUG ABSORPTION FROM INTESTINAL MIXED

MICELLES AND VESICLES ........................................................................................................ 48

1.5.1 Current model of drug absorption and its limitations ......................................................... 48

1.5.2 Collisional uptake of solubilised drug ................................................................................ 53

1.5.3 Supersaturation-enhanced drug absorption ......................................................................... 57

1.6 STRUCTURE OF THIS THESIS ................................................................................................... 59

CHAPTER 2 : GENERAL METHODS .............................................................................. 60

2.1 INTRODUCTION ........................................................................................................................... 61

v

2.2 AUTOPERFUSED RAT JEJUNUM .............................................................................................. 61

2.2.1 Materials ............................................................................................................................. 61

2.2.2 Methods .............................................................................................................................. 62

2.2.3 Data analysis ....................................................................................................................... 68

2.2.4 Validation of the autoperfused rat jejunum ........................................................................ 70

2.2.5 Adsorption of cinnarizine onto the recirculating perfusion apparatus ................................ 78

2.3 IN VIVO BIOAVAILABILITY STUDIES .................................................................................... 81

2.3.1 Materials ............................................................................................................................. 82

2.3.2 Methods .............................................................................................................................. 82

2.4 VALIDATION OF ANALYTICAL METHODS ........................................................................... 85

2.4.1 HPLC assays for quantification of cinnarizine, halofantrine, fenofibrate, danazol and

meclofenamic acid in intestinal lipid colloidal phases ....................................................... 85

2.4.2 HPLC assay for quantification of cinnarizine in rat plasma ............................................... 89

2.4.3 LC-MS assay for quantification of danazol in rat plasma .................................................. 90

2.4.4 Enzymatic colorimetric assay for quantification of total bile salt in whole rat bile ........... 92

CHAPTER 3 : INTESTINAL BILE SECRETION PROMOTES DRUG ABSORPTION

FROM LIPID COLLOIDAL PHASES VIA INDUCTION OF SUPERSATURATION96

3.1 ABSTRACT .................................................................................................................................... 97

3.2 INTRODUCTION ........................................................................................................................... 98

3.3 METHODS .................................................................................................................................... 102

3.3.1 Materials ........................................................................................................................... 102

3.3.2 Experimental outline ......................................................................................................... 103

3.3.3 Formulation preparation ................................................................................................... 105

3.3.4 Particle sizing ................................................................................................................... 108

3.3.5 Equilibrium solubility of cinnarizine in the model micelles and vesicles ........................ 109

3.3.6 Kinetics of cinnarizine precipitation ................................................................................. 109

3.3.7 Solid-state analysis of the cinnarizine precipitate ............................................................. 109

3.3.8 Animals ............................................................................................................................. 110

3.3.9 Surgical procedures .......................................................................................................... 110

3.3.10 Cinnarizine bioavailability studies ................................................................................... 111

3.3.11 In situ single-pass rat jejunum perfusion .......................................................................... 113

3.3.12 Analytical procedures ....................................................................................................... 114

3.3.13 Statistical analysis ............................................................................................................. 117

3.4 RESULTS ...................................................................................................................................... 117

vi

3.4.1 SR-BI, CD36, NPC1L1 and endocytosis have little impact on drug absorption from

intestinal colloidal phases ................................................................................................. 117

3.4.2 Drug absorption from micelles and vesicles is determined by Cfree and not colloidal

structure ............................................................................................................................ 122

3.4.3 Bile-mediated dilution of cinnarizine-loaded micelles and vesicles generates drug

supersaturation .................................................................................................................. 126

3.4.4 Bile-induced drug supersaturation increases jejunal absorptive flux for micelles but not

vesicles .............................................................................................................................. 129

3.4.5 Bile-induced drug supersaturation increases in vivo cinnarizine exposure after

intraduodenal infusion ...................................................................................................... 132

3.5 DISCUSSION ................................................................................................................................ 132

3.6 CONCLUSION ............................................................................................................................. 141

CHAPTER 4 : THE POTENTIAL FOR DRUG SUPERSATURATION DURING

INTESTINAL PROCESSING OF LIPID-BASED FORMULATIONS IS ENHANCED FOR

BASIC DRUGS .................................................................................................................... 145

4.1 ABSTRACT .................................................................................................................................. 146

4.2 INTRODUCTION ......................................................................................................................... 147

4.3 METHODS .................................................................................................................................... 149

4.3.1 Materials ........................................................................................................................... 149


4.3.3 Preparation of model intestinal colloidal phases containing long-chain lipids ................. 151

4.3.4 Equilibrium solubility of drugs in oleic acid and soybean oil .......................................... 153

4.3.5 Equilibrium solubility of drugs in long-chain colloids ..................................................... 155

4.3.6 Kinetics of cinnarizine precipitation ................................................................................. 155

4.3.7 Solid-state analysis of the cinnarizine precipitate ............................................................. 156

4.3.8 Animals ............................................................................................................................. 156



4.3.11 Cinnarizine bioavailability after intraduodenal infusion .................................................. 159



4.4 RESULTS ...................................................................................................................................... 162

4.4.1 Trends in drug solubility in long chain lipid-based colloids as a function of bile and lipid

concentration are different for basic, neutral and acidic drugs ......................................... 162

vii

4.4.2 The addition of rat bile to drug-loaded model colloids reduces cinnarizine solubility and

promotes supersaturation, but increases danazol solubilisation ....................................... 168

4.4.3 Co-perfusion of bile with model colloids increases cinnarizine absorption to a greater

extent than danazol ........................................................................................................... 170

4.4.4 Supersaturation increases in vivo cinnarizine exposure after intraduodenal infusion of

drug-loaded model colloids .............................................................................................. 175

4.5 DISCUSSION ................................................................................................................................ 177

4.6 CONCLUSION ............................................................................................................................. 186

CHAPTER 5 : LIPID ABSORPTION TRIGGERS DRUG SUPERSATURATION AT THE

INTESTINAL UNSTIRRED WATER LAYER AND PROMOTES DRUG ABSORPTION

FROM MIXED MICELLES .............................................................................................. 190

5.1 ABSTRACT .................................................................................................................................. 191

5.2 INTRODUCTION ......................................................................................................................... 192

5.3 METHODS .................................................................................................................................... 197

5.3.1 Materials ........................................................................................................................... 197


5.3.3 Preparation of LCFA-containing intestinal colloids ......................................................... 199

5.3.4 Preparation of Brij 97 colloids .......................................................................................... 200

5.3.5 Equilibrium solubility studies of cinnarizine in colloids .................................................. 200

5.3.6 Animals ............................................................................................................................. 201





5.4 RESULTS ...................................................................................................................................... 206

5.4.1 Attenuation of the acidic microclimate using amiloride reduces oleic acid and cinnarizine

absorption from model LCFA colloids; but has no effect on the absorption of cinnarizine

from fatty acid-free Brij 97 colloids ................................................................................. 206

5.4.2 Exposure of model LCFA colloids to the acidic microclimate, and absorption of lipid

components, leads to cinnarizine supersaturation and enhanced thermodynamic activity

.......................................................................................................................................... 211

5.4.3 Attenuation of the acidic microclimate using amiloride abolishes bile-induced,

supersaturation-enhanced, cinnarizine absorption from model LCFA colloids ............... 217

5.5 DISCUSSION ................................................................................................................................ 221

viii

5.6 CONCLUSION ............................................................................................................................. 229

CHAPTER 6 : SUMMARY AND PERSPECTIVES ....................................................... 231

REFERENCES ........................................................................................................................................ 241

APPENDIX 1 ........................................................................................................................................... 261

ix

ABSTRACT

This thesis seeks to elucidate the mechanism of drug absorption from the colloidal micellar and

vesicular species that form in the gastrointestinal (GI) tract during lipid digestion. In addition to

absorption from the free fraction of drug molecules that exist in equilibrium with drug solubilised

in the colloidal reservoir, two alternative models of drug absorption were explored: (i) collisional

drug absorption where lipid colloidal phases interact directly with the absorptive membrane, and

(ii) supersaturation-enhanced absorption where transient changes to colloid structure and content

in situ lead to drug supersaturation, thereby boosting drug thermodynamic activity and rendering

solubilised drug more available for absorption via the free fraction. Collisional drug absorption

was explored by comparing the intestinal absorptive flux of cinnarizine (CIN) from two distinctly

different colloids (micelles vs. vesicles) that were matched for CIN solubilisation capacity and

drug loading (and therefore thermodynamic activity). In these systems the number of micellar

particles was substantially higher than that of vesicles, and as such collisional absorption of CIN

was expected to be higher from micelles. The absorption of CIN from micelles and vesicles,

however, was not statistically different, suggesting little collisional involvement in drug

absorption. Receptor-mediated collisional absorption was examined by assessing CIN

bioavailability from a lipid emulsion in the absence and presence of inhibitors of common lipid

uptake transporters (e.g. SR-BI, CD36, NPC1L1). CIN bioavailability was unchanged by inhibitor

co-administration. Collectively, the data suggest that collision-mediated uptake is not a significant

driver for drug absorption from intestinal lipid colloidal phases, and that drug absorption occurs

largely from the free fraction. Subsequently, attention turned to the possibility that drug

supersaturation might be stimulated during endogenous processing of intestinal lipid colloidal

phases. Two mechanisms were investigated: (i) interaction of lipid colloidal phases with bile

secretions, where bile-induced changes to colloid microstructure may lead to reductions in drug

solubilisation capacity and (ii) lipid absorption from intestinal colloids, where reductions in

colloidal lipid content may reduce drug solubilisation capacity at the intestinal unstirred water

layer (UWL). The addition of donor rat bile to CIN-loaded colloids (CIN was loaded at sub-

x

saturated concentrations) resulted in a decrease in CIN solubilisation and the generation of CIN

supersaturation. Bile-induced supersaturation was subsequently shown to increase the intestinal

absorptive flux and systemic exposure of CIN from both medium-chain and long-chain lipid

containing colloids. To assess the potential for lipid absorption to induce drug supersaturation, the

intestinal absorptive flux of CIN from oleic acid-containing colloids was assessed under

conditions of normal lipid absorption vs. inhibited lipid absorption (oleic acid absorption was

inhibited by co-administration of amiloride, an inhibitor of UWL acidity). When oleic acid

absorption was suppressed, the absorption of CIN was dramatically attenuated. Assessment of

CIN solubilisation behaviour under conditions that simulate lipid absorption at the UWL

subsequently indicated that supersaturation was likely to be responsible for the enhanced CIN

absorption observed during normal lipid absorption. In summary, supersaturation appears to be an

important driving force for drug absorption from lipid-based intestinal colloids. Two novel

mechanisms have been identified by which drug supersaturation may be naturally triggered in the

small intestine (bile dilution and lipid absorption). The findings enhance mechanistic

understanding of the effects of lipids in food or formulations on drug absorption and are expected

to inform the development of more rational design criteria for LBF.

xi

Monash University Declaration for thesis based or partially based on conjointly published or unpublished work

GENERAL DECLARATION

In accordance with Monash University Doctorate Regulation 17 Doctor of Philosophy and Research Master’s regulations the following declarations are made: I hereby declare that this thesis contains no material which has been accepted for the award of any other degree or diploma at any university or equivalent institution and that, to the best of my knowledge and belief, this thesis contains no material previously published or written by another person, except where due reference is made in the text of the thesis. This thesis includes one original paper published in peer reviewed journals and two submitted publications. The core theme of the thesis is an understanding of the mechanism of drug absorption from intestinal colloidal species following oral administration of lipids. The ideas, development and writing up of all the papers in the thesis were the principal responsibility of myself, the candidate, working within the Drug Delivery, Disposition and Dynamics Theme of the Monash Institute of Pharmaceutical Sciences under the supervision of Prof Christopher J. H. Porter and Dr Natalie L. Trevaskis. The inclusion of co-authors reflects the fact that the work came from active collaboration between researchers and acknowledges input into team-based research. In the case of Chapter 3 my contribution to the work included the majority of the experimental work, all data analysis and interpretation, the concept and design of all studies, the preparation of initial drafts of all manuscripts and the subsequent revision and formulation of conclusions and hypotheses resulting from the relevant studies. In the case of Chapters 4 and 5 my contribution to the work included all the experimental work, data analysis and interpretation, the concept and design of all studies, the preparation of initial drafts of all manuscripts and the subsequent revision and formulation of conclusions and hypotheses resulting from the relevant studies. Thesis chapter

Publication title Publication status* Nature and extent of candidate’s contribution

3 Intestinal bile secretion promotes drug absorption from lipid colloidal phases via induction of supersaturation

In press

Planning and execution of experimental work, data evaluation, drafting and revision of manuscript

4 The potential for drug supersaturation during intestinal processing of lipid-based formulations is enhanced for basic drugs

In submission


5 Lipid absorption triggers drug supersaturation at the intestinal unstirred water layer and promotes drug absorption from mixed micelles

In submission


I have renumbered sections of submitted or published papers in order to generate a consistent presentation within the thesis. Signed: ………………………………………… Date: ………………………………

xii

ACKNOWLEDGEMENTS

I’d first like to thank my supervisors Prof Chris Porter and Dr Natalie Trevaskis for their guidance

and support. They are outstanding scientists who still believe in the personal development of their

students, for that I could not have asked for two better people to see me through my PhD. Thank

you for consistently being there, for the times when you have shown great insight, care,

enthusiasm and encouragement. Thank you also for having the foresight and courage to ask an

excellent research question that led to the conception of this project – I have had a ball, and these

years spent under your supervision will be some of the most profound and memorable experience

in my life.

To my parents, I feel truly lucky to be your daughter. You are such resilient individuals, and have

strived hard to give us a good life, and importantly, good and free minds. Thank you for always

listening and entertaining my thoughts and ideas, even if I was and still am, a little dreamy. I think

it has given me the confidence to just be myself. Thank you also for your unconditional love. I

didn’t realise the importance of having such unwavering support behind me, but I know now that

it’s the reason I do not think twice before undertaking a difficult endeavour. Thank you especially

to my mum who, in retrospect, really made the effort to ensure that our upbringing was a well-

considered one. To my siblings YXY, YYH (and Alicia and baby Xuan Xuan) and YYY, thank

you for always looking out for me and showing an interest in what I do, I hope I have outgrown

little sister status over the years.

To Hywel, in the past two years or so I have found it difficult in my mind to disentangle you from

the wonderful turn in PhD journey I have had, maybe good things do come in pairs. You have

been an absolute inspiration to me, and I am still astounded by the depth of support you continue

to provide. Thank you for so many things, from making me dinner to keeping me warm on top of

a mountain to showing me how best to import powerpoint figures to always being enthusiastic and

insightful when discussing ideas and concepts. I look forward to our future together.

To Shalini and Sifei, how will I complete my PhD a sane person without you two? Shalini you

have been a great PhD buddy. Throughout the years you have provided me with lots of laughs,

endless conversation, and some serious emotional support through the personally challenging

times. I will treasure the friendship and good times we’ve had. Sifei your wise words are worthy

of compilation, and your convincing me that breakthrough is near at my lowest point is one of the

most important things anyone has said to me. Thank you for always pointing me in the right

direction. I am grateful to have known such kind-hearted individuals full of integrity as you two.

xiii

The financial support by Australian Postgraduate Award is also gratefully acknowledged.

Lastly, I would like to thank everyone who cared and shared good times with me. Tri and Kathy

for their ‘special’ way of caring, LJ for her friendship and wisdom especially in the last few

months, Lisa for showing me how ‘easy’ it is to write a thesis/paper (in a week!), Jenny for giving

me a job with a handsome pay during my PhD years, Joe and Ian and Danielle and Khay for

lunchtime friendship during my write-up, Rachelle for yummy Thursday night Chinese takeaways,

my sister Ying, Ginny, Kary and Tim for their hospitality in this last month when I don’t have a

regular place to call home, and a special mention to Alice and Tom for being so interested in my

work. These are the things I will remember.

xiv

PUBLICATIONS

This thesis is a compilation of the following manuscripts:

Chapter 3:

Yeap YY, Trevaskis NL, Quach T, Tso P, Charman WN, Porter CJH. Intestinal bile secretion

promotes drug absorption from lipid colloidal phases via induction of supersaturation. Molecular

Pharmaceutics, 2013, in press. (Proofs appended at the end of this thesis)

Chapter 4:

Yeap YY, Trevaskis NL, Porter CJH. The potential for drug supersaturation during intestinal

processing of lipid-based formulations is enhanced for basic drugs. Manuscript in submission.

Chapter 5:

Yeap YY, Trevaskis NL, Porter CJH. Lipid absorption triggers drug supersaturation at the

intestinal unstirred water layer and promotes drug absorption from mixed micelles. Manuscript in

submission.

xv

COMMUNICATIONS

Yeap YY. Rethinking drug absorption from post-digestion lipid colloidal phases: A

supersaturation-based model of absorption. Podium presentation, AAPS Annual Meeting and

Exposition 2012, Chicago, IL, USA. (Award winning)

Yeap YY, Trevaskis NL and Porter CJH. The acidic microclimate of the intestinal unstirred water

layer can promote drug absorption from long-chain mixed micelles via induction of

supersaturation at the absorptive site. Poster presentation, AAPS Annual Meeting and Exposition

2012, Chicago, IL, USA.

Yeap YY, Trevaskis NL and Porter CJH. Bile secretion increases drug absorption from intestinal

mixed micelles via induction of supersaturation. Poster presentation, AAPS Annual Meeting and

Exposition 2011, Washington, DC, USA.

Yeap YY, Trevaskis NL and Porter CJH. The role of small intestine lipid uptake transporters in

the oral absorption of lipids and poorly water soluble drugs. Poster presentation, APSA Annual

Conference 2009, Hobart, Tasmania, Australia.

xvi

LIST OF ABBREVIATIONS

% CV % coefficient of variation

AUC area under the curve

BA bioavailability

BBMV brush border membrane vesicles

BCRP breast cancer resistant protein

BCS Biopharmaceutics Classification System

BLT-1 Block Lipid Transport-1

BS bile salts

CD36 Cluster of Differentiation 36

Ch cholesterol

CIN cinnarizine

cm centimetre

CMC critical micellar concentration

DAN danazol

DG diglyceride

DMSO dimethyl sulfoxide

FA fatty acids

FABP fatty acid binding protein

FABPpm membrane-associated fatty acid binding protein

FAT Fatty Acid Translocase

FATP fatty acid transport protein

FF fenofibrate

g gram

GI gastrointestinal

h hour

xvii

HDL high density lipoprotein

HF halofantrine

HLB Hydrophilic-Lipophilic Balance

HPLC high performance liquid chromatography

kg kilogram

L litre

LBF lipid-based formulations

LC long-chain

LCFA long-chain fatty acids

LC-MS liquid chromatography-mass spectrometry

LCT long-chain triglycerides

LDL low density lipoprotein

LFCS Lipid Formulation Classification System

Log D logarithm of the octanol-water distribution coefficient

LPC lysophosphatidylcholine

LPL lysophospholipids

m metre

mg milligram

mL millilitre

mm millimetre

mM millimolar

min minute

M molar

MC medium-chain

MCT medium-chain triglycerides

MFA meclofenamic acid

MG monoglyceride

xviii

µCi microCurie

µg microgram

µL microlitre

µm micrometre

µM micromolar

MRP multidrug resistance-associated protein

nm nanometre

ng nanogram

NMR nuclear magnetic resonance

NPC1L1 Niemann-Pick C1 Like 1

OA oleic acid

o/w oil in water

Papp apparent permeability coefficient

PC phosphatidylcholine

PEG polyethylene glycol

P-gp P-glycoprotein

PL phospholipids

PWSD poorly water-soluble drugs

rpm revolutions per minute

SAXS small-angle X-ray scattering

SDS sodium dodecyl sulphate

sec second

SEDDS self-emulsifying drug delivery systems

SEIF simulated endogenous intestinal fluid

SEM standard error of the mean

SI small intestine

SMEDDS self-microemulsifying drug delivery systems

xix

SR-BI Scavenger Receptor Class B Type 1

S-SEDDS supersaturable self-emulsifying drug delivery systems

SSO sulfo-N-succinimidyl oleate

SS ratio supersaturation ratio

TBME tert-butyl methyl ether

TG triglyceride

TRL triglyceride-rich lipoproteins

UV ultraviolet

UWL unstirred water layer

VLDL very low density lipoprotein

v/v volume in volume

w/o water in oil

w/v weight in volume

w/w weight in weight

x g relative centrifugal force

20

CHAPTER 1 : GENERAL INTRODUCTION

Chapter 1: General Introduction

21

1.1 STATEMENT OF THE PROBLEM

Lipid-based formulations (LBF) are commonly employed to increase the oral bioavailability of

poorly water-soluble drugs (PWSD). Confident and widespread application of LBF, however, is

limited by the lack of a holistic understanding of the mechanisms by which lipids enhance the

absorption of PWSD. Currently, the design and evaluation of LBF is largely predicated on the

ability of formulations to maintain drug solubilisation during in vitro simulations of formulation

dispersion and digestion in the gastrointestinal (GI) tract. This approach, while generally useful, is

largely empirical and does not always adequately predict formulation performance in vivo. In

contrast, considerably less attention has been directed towards an understanding of the

mechanism(s) of drug absorption from the colloidal phases (e.g. micellar and vesicular species)

that form in the small intestine (SI) following lipid ingestion. Improved understanding of the

fundamental processes that govern drug absorption from intestinal lipid colloidal phases has the

potential to provide enhanced design parameters for LBF, and is the main aim of this PhD thesis.

1.2 LIPIDS FOR THE ENHANCEMENT OF ORAL DRUG ABSORPTION

Growing application of high throughput activity screens and the use of complex chemical

scaffolds in drug discovery has led to increasingly frequent identification of poorly water-soluble

compounds as prospective drug candidates. Indeed, recent estimates suggest that at least 40% of

new chemical entities in development are classified as ‘poorly water-soluble’1. While these

compounds may possess high intrinsic potency at target sites, the development and clinical utility

of poorly water-soluble drug candidates are often limited by low systemic exposure after oral

administration. This reflects the need for drugs to be molecularly dispersed (i.e. in solution) in

order to pass across the absorptive cells that line the GI tract, and for poorly water-soluble


22

compounds, this prerequisite is limited by slow dissolution rates and low solubility in the aqueous

GI milieu.

Several approaches have been employed to improve the dissolution and solubilisation

characteristics of poorly-water soluble compounds in the GI tract. These include the isolation of

drug candidates in alternate salt forms, polymorphs or co-crystals with improved dissolution and

solubility profiles; particle size reduction strategies such as milling or nanomilling; and

formulation strategies such as the use of solid dispersions, cyclodextrins, and surfactant or lipid-

based drug delivery systems. The common strategies to enhance the dissolution and solubility of

poorly water-soluble compounds has recently been reviewed in detail by Williams et al.2. The

focus of the studies in this thesis, however, is the use of LBF to enhance the absorption of PWSD.

1.2.1 Poorly water-soluble drugs and the Biopharmaceutics Classification

System

The Biopharmaceutics Classification System (BCS) categorises drugs into four classes according

to their water solubility and membrane permeability properties (Table 1.1), and allows for the

broad prediction of the rate-limiting steps in drug absorption following oral administration3.

Table 1.1: The biopharmaceutical classification system (BCS) for drugs according to their

solubility and permeability properties.

Class Solubility Permeability Examples

I High High ketoprofen, propranolol, midazolam, paracetamol

II Low High cinnarizine, danazol, phenytoin, griseofulvin

III High Low metformin, digoxin, cefotaxamine, cimetidine

IV Low Low cyclosporin A, lovostatin, amphotericin, paclitaxel


23

PWSD are contained within Class II and IV of the BCS. In general, the absorption of Class II

drugs is limited by solubility alone, whereas the absorption of Class IV drugs is limited by both

solubility and permeability. PWSD include examples that have been likened to ‘grease balls’ that

are both hydrophobic (i.e. solubility is limited by hydration and interaction with aqueous solvents)

and lipophilic (i.e. where lipid solubility is high and melting point is typically low); and

compounds that have been likened to ‘brick dust’ that are hydrophobic but not lipophilic (i.e.

where solubility is limited by strong intermolecular bonds within the crystal lattice thereby

restricting the solubility of compounds in both water and lipids). ‘Grease ball’ like compounds

(typically found in Class II) are well-suited to formulation in LBF since the higher lipophilicity of

these compounds dictates that they have higher solubility in the formulation when compared to

‘brick dust’ like molecules. As new molecular entities become increasingly lipophilic and poorly

water-soluble1, LBF are becoming an increasingly effective and popular strategy to improve oral

absorption. However, a thorough understanding of the mechanisms by which lipids exert their

absorption-enhancing effects must first be obtained before LBF technologies can be confidently

applied.

1.2.2 Lipid-based formulations used in oral drug delivery

‘Lipid-based formulation’ is an umbrella term that encompasses a diverse group of formulations

including lipid solutions, suspensions, emulsions, microemulsions, nanoemulsions, lipid

nanoparticles, self-emulsifying drug delivery systems (SEDDS), liquid crystalline materials such

as lamellar, hexagonal, and cubic liquid crystals, and lipid complexes such as cochleates. LBF

may be employed for oral or parenteral administration. In oral drug delivery, LBF comprising

simple isotropic liquids that may be filled into gelatin capsules are the most widely used4. LBF

typically comprise a mixture of excipients including triglycerides, mixed mono- and diglycerides,

non-ionic surfactants, and cosolvents5. The choice of excipients is usually based on the need to


24

solubilise the drug dose in the formulation, the ability to promote rapid initial emulsification of the

formulation in the GI fluids, the capacity to maintain drug solubilisation during dispersion and

digestion of the formulation in the GI tract, and the potential to alter metabolic processing

pathways such as the stimulation of lymphatic transport or the inhibition of enterocyte-based

efflux transporters and metabolism. Some examples of LBF that have been shown to enhance the

oral bioavailability of PWSD are listed in Table 1.2.


25

Table 1.2: Examples of LBF used to enhance the oral bioavailability (BA) of PWSD.

Compound Formulation Study Observation Ref.

Cinnarizine Oleic acid solution Oral BA (beagle dogs)

4 fold increase in relative BA c.f. tablets

6

Penclomedine Trioctanoin, triolein, soybean oil, mineral oil and tributyrin o/w emulsion

Intraduodenal dosing (rats)

3-7 fold increase in absolute BA when dosed in digestible and non-digestible lipids c.f. aqueous suspension

7

Danazol Monoolein o/w emulsion Oral BA (humans)

4 fold increase in relative BA c.f. capsules

8

Griseofulvin

Dexamethasone

Peanut oil, Captex 355 and triacetin oil suspensions

Oral BA (rats) Griseofulvin: 2-3 fold increase in absolute BA c.f. aqueous suspension

Dexamethasone: 8-9 fold increase in absolute BA c.f. aqueous suspension

9

Griseofulvin Corn oil o/w emulsion Oral BA (humans)

2 fold increase in relative BA c.f. aqueous suspension

10

Ontazolast Soybean oil o/w emulsion

Oral BA (rats) 15 fold increase in absolute BA c.f. aqueous suspension

11

Halofantrine Peanut oil, Captex 355 and tributyrin solutions

Oral BA (rats) 2-3 fold increase in absolute BA c.f. aqueous suspension

12

Carvedilol SEDDS (MCT) Oral BA (beagle dogs)

4 fold increase in relative BA c.f. tablets

13

Cyclosporin Sandimmune® (SEDDS), Sandimmune Neoral® (SMEDDS)

Oral BA (humans)

Up to 2.4 fold increase in relative BA from SMEDDS c.f. SEDDS

14

Itraconazole SMEDDS (PEG) Oral BA (humans)

1.9 and 1.5 fold increase in BA in fasted and fed states, respectively

15

Silymarin SMEDDS (ethyl linoleate)

Oral BA (rabbits)

49 fold increase in relative BA c.f. suspension

16

Paclitaxel S-SEDDS Oral BA (rats) 5-fold increase in relative BA c.f. intravenous solution (Taxol®)

17

o/w = oil in water; MCT = medium-chain triglycerides; LCT = long-chain triglycerides; SEDDS = self-emulsifying

drug delivery systems; SMEDDS = self-microemulsifying drug delivery systems; S-SEDDS = supersaturable self-

emulsifying drug delivery systems


26

To assist in the description and comparison of oral LBF, Pouton proposed a classification scheme

that originally grouped LBF into three categories18 with a fourth recently added19. The Lipid

Formulation Classification System (LFCS) allows formulations to be described according to their

composition and general behaviour during in vitro dispersion and digestion tests (Table 1.3).

Table 1.3: Typical composition of LFCS types I, II, IIIA, IIIB, and IV LBF and their

solubilisation capacity for drug before and after dispersion and digestion in the GI tract.

Type I Type II Type IIIA Type IIIB Type IV

Hydrophobic composition high

low

Lipid: TG, DG, MG (% w/w) 100 40-80 40-80 <20 -

Water-insoluble surfactants HLB < 12 (% w/w)

- 20-60 - - 0-20

Water-soluble surfactants HLB > 12 (% w/w)

- - 20-40 20-50 30-80

Hydrophilic cosolvents (% w/w) - - 0-40 20-50 0-50

Solubilisation capacity prior to dispersion

low

high

Particle size on dispersion (nm) coarse 250-2000 100-250 50-100 very fine

Loss of drug solubilisation capacity upon dispersion

limited

high

Importance of digestion crucial

not required

Loss of drug solubilisation capacity upon digestion

Possible Possible Possible Possible Possible

TG, DG, MG = triglycerides, diglycerides, monoglycerides; HLB = Hydrophilic-Lipophilic Balance

Type I formulations are comprised of a single lipid vehicle or a blend of lipids typically consisting

common plant oils or fractionated glycerides20. Type I formulations disperse poorly upon contact

with aqueous media, but typically maintain drug solubilisation during dilution by GI fluids.


27

Digestion processes are required to increase the amphiphilicity of formulation lipids, thereby

improving dispersion properties and promoting the formation of micellar and vesicular species

that are required to maintain PWSD solubilisation and facilitate transfer to the absorptive

membrane (see Section 1.3.2 and Section 1.4.2). The requirement for digestion however, may not

limit absorption as many studies suggest that formulations containing digestible lipids have

superior performance in vivo when compared to formulations that contain poorly or non-digestible

lipids. For example, Yoshiya et al. showed that the systemic exposure of the lipophilic drug SL-

512 in rats was significantly higher after oral administration of a 2 mg/kg dose in MCT (readily

digested) when compared to administration in N-α-methylbenzyllinoleamide (poorly digested)21.

Myers et al. also noted higher bioavailability of penclomedine when administered intraduodenally

in soybean oil, triolein and trioctanoin, when compared to mineral oil7. The major drawback of

Type I formulations is their low solvent capacity for many drugs (except for drugs that possess

very high lipophilicity).

Type II formulations differ from Type I formulations in that they have improved emulsification

properties on dispersion due to the inclusion of lipophilic surfactants (HLB < 12). Type II LBF

exhibit typical ‘self-emulsifying’ drug delivery system (SEDDS) behaviour, and self-emulsify on

contact with aqueous media to form emulsions with particle sizes of 250-2000 nm. Compared to

Type III and IV systems (discussed below), Type II LBF contain limited quantities of hydrophilic

excipients and as such drug precipitation out of the formulation is unlikely during formulation

dispersion (since most formulation components are water immiscible). The use of Type II LBF,

however, has been largely superseded by Type III and IV formulations that have higher drug

solubilisation capacities and result in smaller particle sizes on initial dispersion.


28

Type III formulations include lipids, water-soluble surfactants (HLB > 12) and cosolvents. They

form fine emulsions (particle size of 50-250 nm) on contact with aqueous media, and are often

described as self-microemulsifying drug delivery systems (SMEDDS). Type III formulations are

arbitrarily divided into IIIA and IIIB to distinguish between formulations that contain a significant

proportion of oils (Type IIIA) and those that contain greater quantities of hydrophilic components

(Type IIIB). Type III formulations are expected to have widely varying performance

characteristics due to the diverse group of formulations that fall under this category. Many

marketed LBF are Type III formulations.

Type IV formulations were most recently added to the LFCS19. They contain no classical lipids,

and are comprised predominantly of hydrophilic surfactants and cosolvents. The impact of

digestion on the performance of Type IV systems is therefore expected to be less pronounced than

those containing digestible lipids, although surfactant digestion is still possible in some cases. The

advantage of Type IV systems lies in the high solvent capacity of the formulations, which allows

for higher drug loading when compared to other types of LBF. However, since the formulations

consist of mainly water-miscible excipients, drug precipitation is likely on dilution with GI fluids.

Some toxicity concerns have also been raised regarding the chronic use of Type IV formulations

containing high quantities of surfactant and cosolvent, however there are many examples of

marketed products (e.g. the HIV protease inhibitors) where daily excipient ‘doses’ are high, but

seemingly well tolerated.

The selection of excipients for LBF is currently guided primarily by the capacity for the excipients

(or combinations of excipients) to solubilise the required drug dose, and to maintain drug

solubilisation during in vitro dispersion and digestion tests. Other factors that may influence

excipient selection include mutual miscibility, toxicity, irritancy, capsule compatibility, purity,


29

chemical stability and cost. A summary of the common excipients used in LBF is listed in Table

1.4.

Table 1.4: Excipients commonly used in oral LBF. The table is adapted with permission from

Williams et al.2.

Excipient Class Examples

Medium-chain triglycerides Coconut or palm seed oil, Miglyol, Captex

Long-chain triglycerides Soybean, sesame, safflower, sunflower, corn, cottonseed,

olive, palm, peanut (arachis), rapeseed oils; hydrogenated

vegetable and soybean oils

Medium- and/or long-chain monoglycerides,

diglycerides, or mixed glycerides

Medium-chain: Capmul, Imwitor, Labrafac, Capmul GMO

Long-chain: Maisine 35-1, Peceol, Geleol

Low HLB surfactants (HLB < 8) Lipophilic sorbitan fatty acid esters (e.g. Span 85, Span 80);

esters of propylene glycol and fatty alcohols (Lauroglycol,

Capryol)

Water-insoluble, lipophilic surfactants (HLB 8-12) Predominantly oleate esters and include polysorbate 85

(Tween 85) and Tagat TO

Water-soluble surfactants (HLB > 12) Include products synthesized by reacting alcohols with

ethylene oxide to produce alkyl ether ethoxylates (e.g., Brij,

cetomacrogol); sorbitan esters with ethylene oxide to form

sorbitan ester ethoxylates (e.g. Tween); ethylene oxide with

castor oil to produce castor oil ethoxylates (e.g. Cremophor

EL, RH40); vegetable oils with polyethylene glycol (PEG)

to produce mixtures of monoglycerides, diglycerides,

triglycerides and PEG esters of fatty acid (Labrasol,

Labrafil, Gelucire)

Cosolvents PEG 400, propylene glycol, propylene carbonate, ethanol

HLB = Hydrophilic-Lipophilic Balance


30

1.3 OVERVIEW OF LIPID DIGESTION AND ABSORPTION

Ingested dietary or formulation-derived lipids are subjected to a series of digestion and

solubilisation processes in the GI tract. These facilitate presentation of poorly water-soluble lipids

to the absorptive membrane of the SI at high concentration and in a molecularly dispersed,

solubilised form. The main stages of lipid processing from ingestion to uptake into the systemic

circulation are described below:

1.3.1 Lipid digestion in the oral cavity and stomach

The digestion of lipids is initiated in the oral cavity by lingual lipase, prior to the passage of food

into the stomach via the oesophagus. The presence of lipids in the GI tract leads to the secretion of

gastric lipase from gastric mucosa chief cells22, 23. Gastric lipase is acid-resistant, and partially

hydrolyses triglycerides (TG) to diglycerides (DG) and fatty acids (FA)24, 25 with specific

preference for digestion at the sn-3 position on the glycerol backbone26, 27. Gastric lipid digestion

is thought to constitute a small portion of total lipid digestion. Indeed, FA concentrations (as an

indicator of the extent of lipid digestion) are up to 7-fold higher in the duodenum than in the

stomach of humans28, and 600-fold higher in pigs29. Nonetheless, the amphiphilic nature of lipid

digestion products (DG, FA), coupled with the shear force exerted on lipid droplets by gastric

propulsion, grinding and retropulsion, aid the emulsification of lipids in the stomach and increases

the surface area available for digestion. As a result, lipids enter the duodenum via the pyloric

sphincter as crude emulsion droplets with diameters less than 0.5 µm30. The rate of gastric

emptying is thought to be regulated (slowed) by a feedback mechanism stimulated by the presence

of FA in the lower SI31, 32, which further facilitates efficient lipid digestion in the duodenum.


31

1.3.2 Lipid digestion and solubilisation in the small intestine

The presence of lipids in the duodenum stimulates the secretion of bile from the gall bladder and

pancreatic fluid from the pancreas33, 34. Bile contains bile salts (BS), phospholipid (PL),

cholesterol (Ch), bile pigments, organic wastes, and bicarbonate ions whereas pancreatic fluid

contains enzymes that catalyse the hydrolysis of lipids (lipase and co-lipase), proteins and peptides

(trypsin, chymotrypsin and carboxypeptidase) and carbohydrates (amylases). pH change from the

stomach environment to the small intestinal environment leads to ionisation of FA, which imparts

further amphiphilicity and results in improved emulsification of lipid droplets35.

Pancreatic lipase acts on the oil/water interface of lipid droplets to hydrolyse TG and DG to 2-

monoglyceride (2-MG) and FA as the final digestion products while amphiphilic bile constituents

such as BS and PL further aid the emulsification process to produce smaller lipid droplets,

increasing the effective surface area for lipid digestion. Co-lipase is required to anchor lipase to

the oil/water interface of oil droplets in the presence of bile salts36. In addition to the digestion of

TG and DG, PL (predominantly phosphatidylcholine (PC)) is hydrolysed to lysophospholipids

(LPL) (predominantly lysophosphatidylcholine (LPC)) and FA by pancreatic phospholipase A237;

and cholesterol esters (which comprises 10-15% of dietary cholesterol) are hydrolysed to free

cholesterol by pancreatic cholesterol esterase38.

Although the aqueous solubility of lipid digestion products is inherently low, the apparent

solubility of these molecules in the aqueous environment of the SI may be increased many fold via

solubilisation within bile-derived micellar species. In effect, the amphiphilic nature of the biliary

lipids (BS, PL, LPL) and the lipid digestion products (MG, FA) enable self-association of

endogenous and exogenous sources of lipids to form a series of colloidal structures that maintain

lipid digestion products in a solubilised state.


32

Digesting lipids in the SI display complex phase behaviour depending on the types and quantities

of lipid digestion products and biliary components present. Several studies have examined the

changes in phase behaviour that occur as lipids are diluted and digested in intestinal fluids and

provide insights into the expected phase changes that occur in vivo as lipid digestion products are

formed on the surface of a lipid droplet and are progressively diluted (i.e. solubilised) by intestinal

fluid39-41. The physical changes to dietary lipids that occur during lipid processing in the SI are

summarised in Figure 1.1.


33

Figure 1.1: Schematic diagram showing the physical changes to dietary lipids during lipid

processing in the SI. Crudely emulsified oil droplets enter the duodenum from the stomach and

stimulate the secretion of biliary and pancreatic fluids. BS, PL and Ch further emulsify the lipid

droplets and increase the surface area available for digestion. Pancreatic lipase and co-lipase

catalyse the hydrolysis of one TG molecule to 2 FA molecules and 1 MG molecule. Lipid

digestion products accumulate on the surface of digesting oil droplets, where they become

increasingly hydrated and eventually slough off as liquid crystalline lamellar, cubic or hexagonal

phases. In the presence of raised BS, PL, Ch concentrations in the SI lumen, lipid digestion

products are incorporated into bile micelles to form a solubilised system often referred to as the

intestinal mixed micellar phase. Continued dispersion of lipid colloidal phases by secreted bile

results in the generation of progressively smaller species. Under conditions of low lipid dispersion

(i.e. high lipid, low bile), larger multilamellar vesicles are thought to predominate; whereas

smaller unilamellar vesicles and mixed micelles co-exist under conditions of high lipid dispersion

(i.e. low lipid, high bile). Absorption of MG and FA is thought to occur from mixed micelles and

unilamellar vesicles, species with smaller particle sizes (< 500 nm) that are capable of permeating

the intestinal unstirred water layer and accessing the absorptive membrane.


34

During lipid digestion, more polar (but still poorly-water soluble) digestion products accumulate

at the surface of digesting oil droplets, where they are thought to ultimately swell (due to

hydration) and slough off as liquid crystalline phases ranging from lamellar to cubic and

hexagonal phases39-41. The presence of lamellar, inverse hexagonal and micellar/bicontinuous

cubic phases during lipid digestion has recently been confirmed by the use of a coupled in vitro

lipid digestion-synchrotron SAXS (small-angle X-ray scattering) model to study real time

evolution of liquid crystalline structures40-42. Currently, there is increasing research effort to

understand the impact that these structures have on drug solubilisation and absorption, and the

potential for the provision of modified release characteristics43, 44. In the SI lumen, the liquid

crystalline phases are diluted by (or solubilised by) intestinal fluids and biliary secretions, leading

to the generation of smaller and less lipid-rich colloidal vesicular and micellar phases. Under

conditions of low dilution, that is, when the concentration ratio of biliary components:lipid

digestion products is low, larger multilamellar vesicles are thought to predominate (an exception

to this is seen in the presence of lipids with lower hydrocarbon chain lengths (C8) where micelles

are the only species observed, presumably due to the ability of C8 FA to self-aggregate and form

micelles at high concentration39). Further dilution of the phases (i.e. increasing the concentration

ratio of biliary components:lipid digestion products) leads to colloidal properties that reflect co-

existence of smaller mixed micelles and unilamellar vesicles39. Importantly, these colloids (i.e.

mixed micelles and unilamellar vesicles) that persist under conditions of high dilution likely

represent the phases that are responsible for the presentation of lipid digestion products to the

absorptive membrane. Despite the co-existence of vesicular structures with mixed micelles in the

aqueous phase of the SI lumen during lipid digestion, the solubilised phase is frequently referred

to as the ‘mixed micellar phase’. In reality the intestinal fluids may consist of a range of highly

dispersed lipid colloidal species. The properties of different types of intestinal colloidal aggregates

are summarised in Table 1.5.


35

Solubilisation of lipid digestion products in small colloidal species in the mixed micellar phase is

also important in the promotion of lipid transport across the intestinal unstirred water layer (UWL).

The UWL is an aqueous diffusion barrier for poorly water-soluble compounds that separates bulk

intestinal fluid from the absorptive surface of enterocytes. The UWL is estimated to be 500-800

µm wide45, 46, has a lower (acidic) pH compared with the bulk intestinal fluid, and is

indistinguishable from a viscous mucus layer consisting water (~ 95%), glycoproteins, lipids,

mineral salts and free proteins45, 47, 48. Since nano-scale particles are able to permeate intestinal

mucus layers49, 50, small colloidal phases that persist under conditions of high dilution (such as

mixed micelles and unilamellar vesicles) are expected to diffuse more readily across the UWL.

Although micellar and vesicular species are larger and slower to diffuse when compared to lipid

monomers, colloidal solubilisation enhances the net diffusion rate of lipids across the UWL, due

to their ability to shuttle more lipid molecules across the UWL in a given time, when compared to

the diffusion of lipid molecules as monomers33, 51.


36

Table 1.5: Properties of lipid colloidal phases found in the SI during lipid digestion.

Colloidal species Properties

Simple micelles Simple micelles are small aggregates of 2-4 bile salt molecules that form spontaneously at concentrations above the critical micellar concentration of bile salts52. The bile salt molecules orientate in such a way that the aromatic hydrocarbon moieties form the core of the micelle, with the polar head groups facing the aqueous environment. Lipid digestion products are incorporated into the hydrophobic core of micelles, thus are ‘solubilised’ in the aqueous SI environment.

Mixed micelles Mixed micelles are micelles that are composed of more than one lipidic species. Simple micelles containing solubilised lipids are mixed micelles by definition. The hydrodynamic radii of mixed micelles are ~ 3–10 nm52. Some properties of simple micelles may change upon solubilisation of lipids. For example, bile salt-lecithin and bile salt-monoglyceride micelles are swollen (larger) and have greater solubilisation capacity for cholesterol and fatty acids respectively, when compared to simple bile salt micelles52.

Vesicles Vesicles are composed of a single lipid bilayer (unilamellar), or alternating lipid bilayers and layers of water (multilamellar). Simple vesicles are typically formed by phospholipids (phospholipids are capable of forming lamellar liquid crystals in water); while the inclusion of cholesterol in vesicles has been suggested to stabilise the bilayer53. Phospholipids and cholesterol in bile are thought to be secreted into the duodenum partly as unilamellar vesicles54. The hydrocarbon chains of the lipid bilayer form the hydrophobic sites in which hydrophobic regions of lipid digestion products can be solubilised in, while the hydrophilic polar groups of the lipid bilayer orientate towards the aqueous layers of water.

Two types of vesicle may be found in the SI during lipid digestion: larger multilamellar vesicles (hydrodynamic radii ~ 50–250 nm54) under conditions of low bile salt and smaller unilamellar vesicles (hydrodynamic radii ~ 40–60 nm52) at higher bile salt concentrations. Vesicles are typically able to solubilise more lipid molecules per particle when compared to micelles due to their larger size.

Liquid crystal phases

PL, FA and MG are insoluble swelling amphiphiles that may form liquid crystal phases when dispersed in aqueous media52. Liquid crystals exhibit both the long range order of crystalline materials and the disorder of liquid systems, and may include lamellar, cubic and hexagonal phases. The relationship between lipid structure and liquid crystal phase formation is defined by the critical packing parameter (p), which is calculated from the effective hydrocarbon volume (v), the fully extended hydrocarbon chain length (l) and the area of the hydrophilic headgroup (a), where p = v/al55. Factors that may affect the packing parameter include hydration, pH and temperature. For example, as water is added to anhydrous amphiphilic lipid (thus decreasing the packing parameter), structural changes from inverse cubic and hexagonal phases (w/o systems, v/al > 1) to lamellar phase (v/al = 1) to normal cubic and hexagonal phases (o/w systems, v/al < 1) may be observed55. During lipid digestion, FA and MG are thought to accumulate at the surface of digesting oil droplets, where they are increasingly hydrated and ultimately sloughed off as liquid crystal phases.


37

1.3.3 Lipid absorption across the enterocyte apical membrane

Realising that micellar structures act as carriers of lipids across the UWL, Westergaard and

Dietschy suggested a model of absorption where micelle-solubilised lipids served as a reservoir to

replenish the passively absorbed free fraction via an equilibrium relationship51. Later, Shiau and

colleagues postulated that absorption of FA (specifically LCFA) was further facilitated by the

acidic microclimate of the intestinal UWL. They suggested that exposure of LCFA to pH values

below their pKa promoted LCFA protonation and led to an attenuation in LCFA amphiphilicity.

This in turn was suggested to enhance LCFA absorption via (i) enhanced LCFA partitioning

across the lipophilic absorptive membrane, and (ii) enhanced LCFA thermodynamic activity (due

to reduced LCFA solubility in bile micelles)56, 57. In both cases, however, uptake of FA across

enterocyte brush border membrane was thought to occur via passive diffusion.

In contrast, in a seminal paper by Chow and Hollander, linoleate (a LCFA) uptake across the

absorptive membrane was shown to be concentration dependent, and facilitated diffusion was

suggested to be the main mechanism of absorption at low lipid concentrations, whereas simple

passive diffusion was thought to dominate at high concentrations58. In subsequent years, findings

from several authors have supported the concept of dual, concentration-dependent mechanism of

absorption for LCFA59, 60, and efforts have been made to identify the plasma membrane lipid

transporters responsible for facilitated LCFA absorption. Transporters that have been identified

include CD3661, FATP462, SR-BI63 and FABPpm64, 65

, although conclusive evidence to support the

role of these receptors in the intestinal absorption of LCFA remains elusive. A number of

transporters including CD3661, SR-BI66, 67 and NPC1L168 have also been implicated in cholesterol

uptake across intestinal brush border membrane. The mechanisms of lipid absorption into

enterocytes are summarised in Figure 1.2.


38

1.3.4 Lipid transport into the systemic circulation

The GI tract is supplied by an extensive network of blood and lymphatic vessels. Absorbed FA

and MG may be transported to the systemic circulation via either the portal blood, or the intestinal

lymph (Figure 1.2 (iv)). Lipids that are destined for transport into the portal vein diffuse directly

across enterocytes and gain access to the blood capillaries via the lamina propria. Lipids that are

destined for transport into the lymphatic vessels are trafficked to the endoplasmic reticulum where

they are re-synthesised to TG via either the 2-monoglyceride pathway (which predominates in the

fed state)69, 70 or glycerol-3-phosphate pathway (which predominates in the fasted state)71, 72. Re-

synthesised TG, together with cholesterol esters, constitutes the primary core lipids of intestinal

lipoproteins. Intestinal lipoproteins are large colloidal particles with a hydrophobic core and a

hydrophilic surface (primarily consisting phospholipid, free cholesterol and apolipoproteins). In

the fasted state, very-low-density lipoproteins (VLDL) are preferentially produced by the

enterocytes whereas after the ingestion of lipids, both VLDL and chylomicrons (larger but less

dense lipoproteins) are produced. Intestinal lipoproteins that are assembled in the endoplasmic

reticulum of enterocytes are subsequently exocytosed into the lamina propria where they

preferentially access the lymphatic vessels rather than the blood capillaries. Preferential lymphatic

access occurs because the diffusion barrier of the vascular endothelium to large lipoprotein is

significant (resulting from the presence of tight junctions between adjacent endothelial cells, and

an underlying basement membrane), whereas the lymphatic endothelia barrier is considerably

more permeable (due to the presence of wider intercellular spaces between lymphatic endothelial

cells (Figure 1.2 (iv)). In general, FA and MG with hydrocarbon chain lengths of 14 and above are

primarily transported into the intestinal lymph whereas the more water-soluble short-chain and

medium-chain FA and MG are primarily transported by the portal blood73-76. Absorbed LPC may

be re-synthesised to PC77-79 (and incorporated into the hydrophilic surface of lipoproteins) or

hydrolysed to glycerol-3-phosphorylcholine that is transported via the portal blood80, 81. In the fed


39

state, absorbed cholesterol is esterified to cholesterol esters82-84, prior to incorporation into the

hydrophobic core of lipoproteins.

Figure 1.2: Schematic diagram outlining the mechanisms of lipid absorption into enterocytes (i, ii,

iii) and lipid transport into the systemic circulation (iv). (i) Solubilisation of lipid digestion

products within mixed micelles or unilamellar vesicles enhances the net diffusion rate of lipids

across the UWL and maximises the concentration of lipids adjacent to the absorptive membrane.

Lipid absorption occurs via the free fraction, which is in equilibrium with the solubilised fraction.

(ii) Diffusion of colloids across the acidic UWL leads to protonation of LCFA (depicted as a loss

of head groups on lipid molecules), which reduces LCFA solubility in bile micelles (thus swelling

micelles), and enhances LCFA absorption via enhanced partitioning across the absorptive

membrane and enhanced thermodynamic activity. (iii) LCFA absorption at low concentration is

thought to be mediated by a saturable, facilitated process. (iv) Absorbed lipids are transported to

the systemic circulation either via the portal blood, or the intestinal lymph. The intestinal lymph

drains via the thoracic lymph directly into the systemic circulation at the junction of the left

subclavian vein and left jugular vein, therefore avoiding first-pass through the liver.

Blood vessel

Lymph vessel

Unstirred water layerBulk lumen

lipoprotein

carrier-mediated uptake

Lipid digestion products

BS, LPC, Ch(i)

(ii)

(iv)

(iii)

Enterocyte

lamina propria


40

1.4 MECHANISMS BY WHICH LIPIDS ENHANCE THE ORAL

BIOAVAILABILITY OF DRUGS

In the absence of carrier-mediated uptake, flux per unit area (F) across the intestinal absorptive

membrane can be represented by the following equation:

F P. C Equation 1.1

where P is the permeability coefficient of drug across the apical membrane; and C is the

concentration of drug on the abluminal side of the apical membrane.

For solubilised drug, and assuming that drug absorption is driven only by the free concentration in

the GI lumen, the maximum flux (per unit area) that can be obtained across the apical membrane

is a function of the permeability coefficient and the maximum free concentration. In the absence

of supersaturation, the maximum free concentration that can be obtained in equilibrium with a

solubilised reservoir is the equilibrium solubility (Cs) of a drug in the inter-micellar phase of the

SI content (a value that approximates the solubility of drug in water buffered to intestinal pH). The

rate of drug absorption is therefore limited by permeability across the absorptive membrane and/or

its solubility in the SI lumen. Furthermore, although drug solubilisation increases the total mass of

drug in solution, the free drug concentration is not expected to be raised above that of drug in

simple solution. Based on Equation 1.1, therefore, drug solubilisation would not be expected to

increase drug absorption. In contrast, a large number of studies detail the utility of solubilising

formulations in enhancing the oral bioavailability of PWSD85-87. This apparent anomaly may be

explained by the impact of drug solubilisation on dissolution (as described in Section 1.4.1 below)

and the potential for formulations to promote supersaturation, where the free concentration can be


41

raised above the equilibrium aqueous solubility (see Section 1.5.3). The latter is the major topic of

this thesis.

After oral administration, the dissolution rate of a drug in the GI fluids (i.e. the rate at which drug

passes into solution in the GI tract) may also be an important determinant of drug absorption.

Since transit time along the GI tract is finite, the rate of drug dissolution must be sufficiently rapid

to allow drug to pass into solution, and be available for absorption prior to transit past the

absorptive site. The dissolution rate of a drug is related to its solubility according to the Noyes-

Whitney equation88:

. C C Equation 1.2

where dc/dt is the dissolution rate of the drug; D is the diffusion rate of the drug in water; A is the

surface area of contact of the solid with the dissolution fluid; h is the width of the diffusion layer;

Cs is maximum drug concentration at the surface of the dissolving fluid i.e. drug solubility; and C

is the concentration of drug in the well-stirred bulk.

The low aqueous solubility of PWSD (i.e. low Cs) dictates that the dissolution rate is also low and

that the absorption of many PWSD is limited by low solubility and/or slow dissolution in the GI

fluids. In general, co-administration with lipids provides PWSD with the opportunity to intercalate

into endogenous lipid digestion and absorption pathways (stimulated by the ingestion of lipids),

ultimately leading to enhancements in drug dissolution, solubilisation, absorption and systemic

exposure as described below.


42

1.4.1 Enhancement of drug dissolution in gastrointestinal fluids

Dosing PWSD in LBF offers significant dissolution advantages when compared to a conventional

solid dose form. LBF typically comprise lipid solutions or emulsion-preconcentrates that are filled

into soft or hard gelatin capsules prior to administration. In this cases, drug is pre-solubilised

within the formulation, thus obviating the need for traditional solid-liquid drug dissolution76.

Ideally, as these lipidic formulations are dispersed, digested and solubilised in the intestinal

content, drug remains solubilised in a monomolecular form and in equilibrium with drug free in

solution. In effect therefore, dissolution is replaced with a (rapid) partitioning process between

solubilised drug and drug in free solution. Bearing in mind the intrinsically slow rate of

dissolution of PWSD from solid dose forms, drug solubilisation in LBF therefore provides

significant dissolution rate advantages.

However, drug precipitation frequently occurs during formulation dispersion and digestion. Slow

drug re-dissolution from the precipitated solids into GI fluids may therefore limit drug absorption

even though the dose is initially solubilised in the formulation. Nonetheless, the co-administration

of lipids may also enhance the dissolution rate of solid drug in the SI via improvements in wetting,

and enhancements in drug solubilisation capacity. Specifically, raised concentrations of

amphiphilic biliary lipids (BS, PL) in the SI after lipid ingestion reduces the contact angle between

solid drug and intestinal fluid, thereby increasing the surface area available for dissolution. This

mechanism has been documented to increase the dissolution rate of phenylbutazone and

hydrocortisone89, 90. Solubilisation of drugs within fed state intestinal mixed micelles (which may

consist BS, PL, FA, MG) also significantly increases the apparent solubility of many PWSD in

intestinal fluids when compared to the fasted state39, 91, thereby increasing dissolution via

increases in solubility according to the Noyes-Whitney equation (Equation 1.2)92. Increases in

dissolution afforded by enhanced solubilisation however, may be off-set by the larger sizes of


43

micellar species which have slower diffusion rates (hence lower D value in Equation 1.2), which

explains the less-than-proportional increase in dissolution rate (in relation to solubility) observed

in many studies93, 94.

1.4.2 Enhancement of apparent drug solubility in intestinal fluids

After the ingestion of lipids, biliary secretions combine with lipid digestion products to generate a

series of intestinal lipid colloidal phases including emulsion droplets, vesicles and mixed micelles

that have markedly increased solubilisation capacities for PWSD. Although increases in the

concentrations of BS, PL, Ch in the fed state GI lumen alone results in significantly increased

solubilisation capacity of a wide range of PWSD95-98, incorporation of digested lipids into bile salt

micelles has been shown to afford further enhancements in drug solubilisation capacity39, 91,

presumably due to the ability of lipid digestion products to swell endogenous bile micelles and to

generate a range of lipid colloidal phases with increased solubilisation capacity (discussed in

Section 1.3.2). The solubilisation enhancement afforded by the incorporation of exogenous lipids

is dependent on the nature of the digestion products and the structures of the colloidal species

formed. For example, medium-chain FA and MG are more amphiphilic than long-chain FA and

MG and more readily combine with endogenous BS, PL and Ch. Thus, at the same lipid load,

medium-chain lipids typically form smaller, more optically clear micellar dispersions while long-

chain lipids form slightly turbid systems containing both vesicular and micellar species. Drug

solubilisation enhancement afforded by the two systems at similar lipid loads are therefore

significantly different, with a less than 3-fold enhancement (relative to drug solubility in

endogenous BS, PL, Ch species) observed in the case of medium-chain lipids and up to 20-fold

enhancement observed in the case of long-chain lipids39, 99. Thus, the mass of drug that may be

solubilised in fed state GI fluids is markedly higher than that in the absence of lipid co-

administration.


44

Analogous to the mechanism by which micellar solubilisation increases FA transport across the

intestinal UWL (outlined in Section 1.3.2), micellar solubilisation may also increase drug

transport across the UWL, thereby increasing the mass of solubilised drug that is presented to the

absorptive membrane. This concept has been applied to good effect in mathematical models

simulating the absorption of progesterone100 and danazol101 from bile micelles. Thus, current

understanding suggests that drug absorption across the apical membrane of enterocytes occurs via

the free drug fraction that is in equilibrium with the solubilised fraction; and that upon drug

absorption from the free fraction, drug rapidly partitions out of the solubilised reservoir to

replenish the free concentration and maintain solubilisation equilibrium2. In turn, the presence of

drug-rich micelles at the absorptive membrane is likely maintained by the continuous diffusion of

micelles from the lumen where lipid digestion, dispersion and drug solubilisation is on-going.

Therefore, although the concentration of drug in the free fraction (the fraction available for

absorption) is not raised by micellar solubilisation (the maximum attainable inter-micellar free

concentration remains at the solubility limit of drug in water), micellar solubilisation overcomes

the solubility limitations of PWSD on the abluminal side of the apical membrane via the provision

of a solubilised reservoir which replenishes free drug concentration during on-going absorption.

1.4.3 Enhancement of intestinal permeability and inhibition of enterocyte-

based efflux transporters and metabolism

Many LBF components (pre- or post-digestion) have been suggested to enhance passive

paracellular (i.e. between cells) and transcellular (i.e. across cells) membrane diffusion102, 103.

Excipients that enhance transcellular permeability are of relevance for PWSD, as PWSD are likely

to be absorbed via the transcellular route rather than the paracellular route due to their

hydrophobicity. Medium-chain FA and glycerides104, 105, long-chain FA and MG103, and a variety

of surfactants106, 107 have been shown to increase the transcellular permeability of the intestinal


45

epithelial membrane in vitro as well as in Caco-2 cells. The ability of excipients to increase

membrane permeability has been attributed to their amphiphilic/surfactant properties, which

disrupt the organisation of the lipid bilayer. Generally, surfactants that have large, charged or

hydrophilic head groups and a relatively small hydrophobic tail are most likely to increase

transcellular permeability as they are taken up rapidly by the outer membrane leaflet but are slow

to translocate to the inner membrane leaflet2. The increased residence time of surfactant molecules

between the bilayers, coupled with the inability of the hydrophobic tail to fill the void between

adjacent membrane lipids, leads to monolayer/bilayer curvature strain, and the generation of a

more permeable membrane. In extreme cases, membrane solubilisation is possible where the

curvature strain is sufficiently high and membrane lipids are expelled. Bile salts are strong

membrane solubilisers103, however, the presence of phospholipid in endogenous bile is expected to

reduce the permeability enhancement effects of bile salts by decreasing bile salt thermodynamic

activity in mixed bile salt-phospholipid micelles108. Although many drugs that are formulated in

LBF have high membrane permeability (e.g. typical BCS Class II compounds), the permeability-

enhancing effects of lipid excipients may be of relevance to BCS Class IV drugs where absorption

is likely limited by both poor solubility and membrane permeability.

An increasing number of LBF excipients have recently been described to enhance drug

permeability via inhibition of intestinal efflux transporters such as P-glycoprotein (P-gp), breast

cancer resistant protein (BCRP) and multidrug resistance-associated protein (MRP)2, 109-112.

Intestinal efflux transporters are present on the apical membrane of enterocytes, and actively pump

absorbed drugs back to the intestinal lumen, thus limiting the absorption of their substrates. Well-

known substrates of efflux transporters include digoxin113, doxorubicin114, vincristine114,

cyclosporine A115, verapamil116 and the HIV protease inhibitors117. Examples of common LBF

excipients that inhibit P-gp, BCRP and MRP have been compiled by Williams et al..2 The


46

mechanisms of efflux inhibition are thought to include competitive or allosteric inhibition of drug

binding sites111, 118-120, changes to membrane fluidity leading to protein destabilisation121-123, and

reductions in transporter expression124-126.

The CYP3A family of metabolic enzymes (also present in enterocytes) share similar substrate

specificity to P-gp127, 128, and have been suggested to be functionally linked with P-gp to limit the

systemic exposure of their substrates129, 130. Although the mechanism of interaction between P-gp

and CYP3A is not well-understood, it is postulated to be due to the function of the efflux

transporter in increasing substrate exposure to the metabolising enzyme in a given time (via

repeated cycles of drug efflux and reabsorption131, 132 which may also modulate the degree of

enzyme saturation), or increased affinity of CYP3A-generated metabolites for P-gp when

compared to the parent compound131. Thus, inhibition of intestinal efflux might also be expected

to reduce enterocyte-based first-pass metabolism, and increase the oral bioavailability of drugs.

Dietary lipids and LBF components may also affect pre-systemic metabolism via direct effects on

enzyme activity and indirect effects on enzyme expression levels133. Examples of lipids or

solubilising excipients that increase drug exposure by altering pre-systemic metabolism may be

found in reviews by Patel et al.133 and Buggins et al.134, respectively. Systemic metabolism may

be altered by lipid co-administration via the alteration of systemic lipoprotein levels (therefore

changing the extent of drug binding in the blood which may decrease or increase uptake into

metabolising compartments)133, 135, 136, although this mechanism is less likely for formulation

lipids due to the small lipid doses contained.

Although the use of LBF excipients as permeability enhancers and metabolic and efflux inhibitors

is receiving increasing interest, relatively few studies, to this point, have shown significant in vivo


47

effects for transcellularly transported drugs. As such the major effects on bioavailability

enhancement appear to be driven by differences in solubility and dissolution.

1.4.4 Promotion of lymphatic drug transport

As described in Section 1.3.4, ingestion of lipids promotes the assembly of triglyceride-rich

lipoproteins (TRL) in the enterocyte and these are transported from the enterocyte to the systemic

circulation via the intestinal lymph. Drug transport into the systemic circulation may occur via

either the portal blood, or the intestinal lymph. Since the blood flow of the portal vein is

approximately 500-fold higher than the lymph flow in the mesenteric lymph duct137, the majority

of absorbed drugs are transported to the systemic circulation via the portal blood76. However, for

highly lipophilic (log P > 5) and highly lipid soluble (LCT solubility > 50 mg/g) drugs, the

potential for lymphatic transport is increased due to the increased likelihood of drug association

with developing TRL in the enterocyte138. Since intestinal lymph drains directly into the systemic

circulation (by-passing the liver), co-administration with lipids may increase the bioavailability of

lymphatically transported drugs via the avoidance of first-pass metabolism in the liver.

Enhancements in lymphatic drug transport have been observed for many drugs after co-

administration with dietary or formulation lipids137. Significantly, studies with halofantrine have

shown that even small amounts of long-chain lipid (e.g. quantities available within a single

capsule) are capable of supporting substantial lymphatic drug transport in the fasted state. In this

way, LBF have the potential to increase oral drug bioavailability for compounds that are highly

first-pass metabolised and that are lymphatically transported139. In general, lymphatic drug

transport increases with administered lipid load138-140, and is more effectively promoted by long-

chain and monounsaturated glycerides than their medium-chain/short-chain and saturated

counterparts12, 138, 139.


48

1.5 PROPOSED MECHANISMS OF DRUG ABSORPTION FROM

INTESTINAL MIXED MICELLES AND VESICLES

1.5.1 Current model of drug absorption and its limitations

In contrast to the wealth of data describing lipid absorption from intestinal mixed micelles

(discussed in Section 1.3.3), the mechanism of absorption of PWSD from mixed micelles remains

poorly understood. Currently, the absorption of PWSD is believed to occur via passive diffusion

of drug monomers across the enterocyte apical membrane (see Figure 1.3(A), solid black arrow),

and drug in the solubilised fraction serves as a reservoir to replenish absorbed drug via re-

establishment of the solubilisation equilibrium:

Ccolloid ↔ Cfree

where Cfree and Ccolloid represent the free concentration and the concentration solubilised in

micelles, respectively. The total solubilised drug concentration (Ctotal) is the sum of Cfree and

Ccolloid:

Ctotal = Cfree + Ccolloid Equation 1.3

In addition to passive diffusion, some drugs may be actively absorbed by specific transporters

present on the apical membrane (see Figure 1.3 (A), dotted black arrow), usually as a result of

structural similarities of the drug to endogenous substrates of the transporter141, 142. Some

examples of active transporters in the intestine and their drug substrates are listed in Table 1.6.


49

Table 1.6: Examples of active transporters in the intestine with specific reference to drug

substrates141.

Transporter Family and Example Substrates

amino acid – large amino acid transporter L-dopa, baclofen, gabapentin

oligopeptide – PepT1 lisinopril, cefadroxil, bestatin, cephradine

Na+ dependent phosphate transporter foscarnet, fosfomycin

monocarboxylic acid – MCT1 pravastatin, salicylic acid, carindacillin

reduced folate transporter (RFT) methotrexate

monosaccharide transporters – SGLT1 glucose, p-nitrophenyl-β-D-glucopyranoside

apical Na+ dependent bile acid transporter (ASBT) S3744 (investigational compound)

nicotinic acid transporter valproic acid, penicillins

Importantly, regardless of active or passive transport, the maximum driving concentration for drug

absorption (i.e. maximum Cfree) is not typically increased by micellar solubilisation and is defined

by drug solubility in the inter-micellar fluid (essentially the aqueous solubility). From the

equilibrium shown above, the relationship between the free fraction (f) and the solubilised fraction

(s) may be described by the following equations143:

s = (Cm – Cs)/Cm Equation 1.4

f = 1 - s Equation 1.5

where Cm is the solubility of drug in the solution with micelles and Cs is the solubility of drug in

the same medium without micelles (essentially the aqueous solubility). It is assumed that s is


50

constant and independent of the drug concentration for a particular concentration of micelles, and

that only one type of micelles exist. Thus,

Ctotal = f.Ctotal + s.Ctotal Equation 1.6

Since the solubilised and free drug concentration are in equilibrium, the degree of drug saturation

in the free fraction and solubilised fraction are expected to be equal. For example, when a micellar

system is loaded with drug at its full solubilisation capacity, Cfree will be the saturated solubility of

the drug in the colloid-free medium (i.e. Cs), while Cmicelle will be the saturated solubility of the

drug in the colloids (i.e. Cm – Cs).

Therefore, even if micelles are efficient in promoting drug transport across the intestinal UWL and

replenishing the absorbed free fraction, the inter-micellar drug concentration at the absorptive

membrane is at best maintained at the solubility limit of free drug in water, and is very low. Since

solubilisation does not increase (and may reduce) Cfree, and drug flux across an absorptive

membrane is the product of the free drug concentration and the drug permeability across the

membrane, lipid co-administration may not be expected to enhance the absorption of drugs that

are solubility- (but not dissolution rate) limited. Indeed, many authors have shown that increasing

the total concentration of solubilised drug within bile salt or synthetic micelles does not

necessarily lead to proportional enhancements in intestinal absorptive flux100, 144, 145. These

observations recently led Miller et al. and Dahan et al. to suggest that formulations that enhance

apparent drug solubility via solubilisation may do so at the expense of overall drug absorption due

to decreases in the free fraction100, 146, 147.


51

Co-administration with lipids, however, remains an effective means to enhance the oral absorption

of PWSD, as exemplified by the myriad drugs that demonstrate positive food effect and

bioavailability enhancement when administered in a LBF. In addition, mathematical models of

intestinal drug absorption that utilise Cfree as the driving concentration frequently underestimate

the absorption of BCS Class II compounds101, 148, 149. Indeed, the use of total solubilised

concentrations (i.e. Cfree plus Cmicelle) often led to better predictions101, 148. These observations

suggest that the absorption of lipophilic, poorly water-soluble compounds from lipid-based

colloids may not be solely dictated by membrane permeability and free solubility, and that instead

additional mechanisms may contribute to overall flux.

Therefore, the principle hypothesis in this thesis is that after lipid co-administration, the

solubilised fraction (i.e. the fraction where the majority of drug in solution resides) is not simply

an inert reservoir that replenishes Cfree, but may contribute to drug flux across absorptive

membrane via enhancements in apparent drug permeability or apparent drug solubility. The

proposed mechanisms where the solubilised fraction may contribute to drug absorption are

discussed below:


52

Figure 1.3: Schematic diagram outlining the mechanisms of drug absorption from intestinal mixed

micelles or vesicles. (A) Current understanding of drug absorption where uptake across the enterocyte

apical membrane is driven by the free fraction (black solid arrow) while the solubilised reservoir

replenishes absorbed free drug via an equilibrium relationship. Additionally, the absorption of drugs that

possess structural similarities to endogenous substrates may be facilitated by specific transporters present

on the apical membrane of enterocytes (black dotted arrow). (B) Proposed mechanism of drug absorption

where in addition to passive diffusion of free drug, absorption may also occur via collisional uptake (red

dotted arrows), where micelles or vesicles interact directly with the apical membrane to mediate selective

transfer of solubilised drug or endocytosis of colloid-receptor complex. Collisional uptake may or may not

be receptor-mediated. (C) Proposed mechanism of drug absorption where the stimulation of drug

supersaturation in a previously stable drug-containing colloid leads to transient increases in the

thermodynamic activity (free concentration) of drug, thereby increasing the number of free drug molecules

that are available for passive diffusion across the apical membrane (red dotted arrows). These alternative

mechanisms of drug absorption from colloidal species are examined in detail in this thesis.

D DSSMixed micelle / vesicle Drug Supersaturated drug

Drug transporters Lipid uptake transporters

DDD

DDD

(A)

DD

DD DD

(B)

DDSS

DSSD DSSDSS

(C)

DD

DDD

Supersaturation stimulus


53

1.5.2 Collisional uptake of solubilised drug

The transfer of FA between model cell membranes, proteins (receptors) and intestinal lipid

colloidal phases via collisional transfer has been extensively investigated by Storch and

colleagues150-153. Collisional transfer refers to the transfer of solubilised content from donor to

acceptor compartments via direct interactions (collisions). For example, Narayanan et al. showed

that FA transfer between mixed micelles occurred as a consequence of micelle collisions rather

than a diffusion process where FA first dissociate from donor micelles into the aqueous phase

prior to association with acceptor micelles152. Similar findings have also been demonstrated for the

transfer of LCFA from phospholipid vesicles to proteins such as brain-FABP151.

In the SI, since lipid digestion products and PWSD are also compounds with limited free solubility

in water, it is possible that the absorption of these compounds could occur via direct collisional

transfer from mixed micelles or vesicles to the absorptive membrane (see Figure 1.3(B), red dotted

arrows). Collisional transfer may or may not be receptor-mediated, however recent interest in the

role of intestinal lipid uptake transporters such as CD3661, 154, 155, SR-BI63, 67, 155, 156, NPC1L1157,

caveolin-1158-160, FATP462 and FABPpm64, 65 in lipid absorption, and reports of direct interactions

of colloidal or macromolecular species in the blood (e.g. albumin, HDL) with one or more of these

receptors have led to the hypothesis that absorption of solubilised lipids and drugs may also occur

via a similar mechanism. Thus colloid-receptor interactions between mixed micelles or vesicles

and lipid uptake transporters in the brush border membrane of the SI may facilitate lipid and/or

drug absorption (Figure 1.3(B), red dotted arrows). In this model, after initial collision,

monomeric lipids and PWSD may be selectively transferred into cells, or the entire colloid-

receptor complex may be endocytosed. A brief description of the relevant receptors and their role

in lipid uptake is given below:


54

1.5.2.1 SR-BI

SR-BI (Scavenger Receptor Class B Type 1) is an 83 kDa glycoprotein that can bind to a variety

of ligands such as anionic phospholipids, apoptotic cells and modified/native lipoproteins in

vitro161. In vivo, SR-BI is a physiologically relevant receptor for HDL and plays a key role in

cholesterol transport into and out of cells. SR-BI acts as a docking receptor for plasma HDL (high

density lipoprotein), and mediates the selective uptake of cholesterol esters into hepatocytes and

steroidogenic tissues161. In the GI tract, SR-BI is expressed on the apical surface of intestinal villi

in the proximal small intestine67. However, the physiological ligand(s) and mechanisms of action

of SR-BI at this site remain unclear. Whilst a study by Hauser et al. reported inhibition of sterol

collisional transfer into brush border membrane vesicles (BBMV) and Caco-2 cells by anti-human

SR-BI IgG67, and there has been evidence of SR-BI receptor endocytosis in pig enterocytes156

during fat absorption, the role of SR-BI in lipid absorption is yet to be fully elucidated.

1.5.2.2 CD36

CD36 (Cluster of Differentiation 36), also known as Fatty Acid Translocase (FAT), is also a class

B scavenger receptor (88 kDa) that shares many ligands (such as native and modified lipoproteins,

anionic phospholipids and serum HDL) with SR-BI154, 155, 161. The distribution of CD36, however,

favours tissues with high metabolic needs for FA e.g. adipose tissue, heart muscle, skeletal muscle

and the intestine162, consistent with its established role as a FA transporter. The role of CD36 in

native LCFA uptake into cells has been well-studied especially in muscle162 and adipose tissues163.

For example, CD36 expression and FA transport are increased during chronic muscle

stimulation162, and CD36 knockout mice show deficits in cellular FA uptake i.e. elevated plasma

FA levels and reduced FA accumulation in muscle and adipose tissue163. Even though the role of

CD36 as a FA transporter is well-known, the exact mechanism of transport under normal

physiological conditions i.e. where FA are mostly bound to albumin, is less clear. Indeed, whether


55

CD36 interacts with free FA or the entire FA-albumin complex in the plasma, and whether it acts

as a docking receptor or an endocytotic receptor to facilitate cellular uptake is not yet known. In

the SI, CD36 is most highly expressed on the apical side of intestinal villi of the duodenum and

jejunum163, and is believed to facilitate FA and cholesterol absorption. For example, FA and/or

cholesterol absorption is lower in vitro and in vivo in CD36 knockout mice when compared to

wild type mice61, and one study which investigated the inhibitory effect of CD36 antibodies on

radiolabelled sterol uptake into human BBMV concluded that CD36 is important in free

cholesterol uptake in the small intestine155. In contrast, in an investigation of FA uptake into the SI

of CD36 knockout mice, CD36 was found to be unimportant in FA uptake but instead played a

role in chylomicron formation and secretion in the enterocytes154. A potential role for CD36 in

intestinal lipid absorption is therefore well established. However, the mechanism whereby CD36

facilitates lipid uptake into enterocytes is not yet known.

1.5.2.3 NPC1L1

Ezetimibe, a cholesterol absorption inhibitor, is widely used in practice and has been shown to

have modest efficacy in lowering plasma cholesterol164. Although the exact mechanism of action

of ezetimibe remains unknown165, pursuit of potential sites of ezetimibe activity helped to reveal

the role of Niemann-Pick C1 Like 1 (NPC1L1) protein in cholesterol absorption. NPC1L1 is a 145

kDa165 membrane protein that is highly expressed in the SI, with detectable levels in tissues such

as the liver, gallbladder, testis, and stomach165. In the SI, NPC1L1 is predominantly expressed on

the apical side of the proximal small intestine, with the highest levels found in the jejunum164.

Unlike other lipid transporters, little is known about NPC1L1, especially its role outside the

intestine, in large part reflecting the relatively recent (2000) identification of NPC1L1 as a

potential lipid transporter166. What is known however, is the seemingly important role of NPC1L1

in cholesterol absorption and involvement in the ezetimibe-sensitive cholesterol absorption


56

pathway. Thus, cholesterol absorption is reduced by more than 70% in NPC1L1 knockout mice

when compared to wild type mice167, and ezetimibe treatment does not reduce cholesterol

absorption further in the same knockout animals167. The available data therefore lend strong

support to a potential role of NPC1L1 in cholesterol absorption, however, the mechanism by

which NPC1L1 facilitates cholesterol absorption is still not established165. Data suggesting that

NPC1L1 aids cholesterol absorption by facilitating intracellular cholesterol transport167 (rather

than absorption directly) further complicate the issue of whether NPC1L1 is an authentic

cholesterol transporter, or a protein that facilitates intracellular trafficking, or both.

1.5.2.3 Endocytosis

Endocytosis can be mediated by several pathways including clathrin coated pits, caveolae-

mediated pathways, and non-clathrin, non-caveolae mediated endocytosis168. Briefly, clathrin-

mediated endocytosis occurs ubiquitously and involves the deformation of the plasma membrane

into a coated pit, a process mediated by surface proteins such as clathrin and adaptor protein-2.

Lipid rafts are domains of the plasma membrane that have a higher degree of order due to

enrichment by cholesterol, sphingolipids and phospholipids. Due to their less fluid (more rigid)

structure, lipid rafts are able to spatially segregate membrane lipids and proteins such that

membrane processes are also functionally separated169. Lipid rafts have received increasing recent

attention as plasma membrane platforms with involvement in the binding and uptake of LCFA160.

Caveolae are subsets of lipid rafts and are characterised by flask-like invaginations in the plasma

membrane that resemble vesicles that are poised to bud170. The expression of the scaffolding

protein, caveolin, is essential to the formation of caveolae170. Evidence also exists to suggest a role

for endocytotic pathways mediated by clathrin coated pits and caveolae in lipid uptake. For

example, the uptake of LCFA by human microvascular endothelial cells has been shown to

involve endocytosis mediated by clathrin-coated vesicles and caveolae171, and the lipid uptake


57

transporters CD36 and SR-BI are found preferentially located in caveolae on the apical membrane

of enterocytes161, 172. Caveolin-1, the scaffolding protein of caveolae, binds FA with strong

affinity169, and may be a potential lipid uptake transporter itself. These observations suggest that

an endocytotic component to lipid uptake in the SI is possible.

1.5.3 Supersaturation-enhanced drug absorption

Supersaturation is a transient solution state that precedes precipitation, where the concentration of

solute in a system is higher than the equilibrium solubility of the solute in the same system.

Supersaturation results in an increase in thermodynamic activity, and therefore a potential means

by which drug absorption can be enhanced.

The potential for supersaturation to enhance the oral bioavailability of PWSD has received

increasing recent interest173, 174, and delivery systems that aim to generate supersaturated drug

concentrations in the GI lumen, and to stabilise the metastable supersaturated state (i.e. delaying

drug precipitation) are increasingly common. The attainment of intraluminal drug concentrations

in excessive of its equilibrium solubility may be achieved by delivering drugs in a solubilised

form that lose solubilisation capacity in situ (e.g. cosolvent systems, LBF), and the delivery of

high-energy crystal forms or amorphous materials that provide accelerated dissolution and/or

higher solubility than their most stable form173, and where precipitation does not occur

immediately.

In the context of LBF, Gao and Morozowich first reported on “supersaturable SEDDS” (S-

SEDDS), where dispersion of the formulation in the GI fluids led to drug precipitation, but where

addition of a precipitation inhibitor resulted in transient stabilisation of a supersaturated state17.


58

They subsequently showed that the oral bioavailability of a number of drugs was significantly

improved after dosing of S-SEDDS when compared to a conventional suspension or SEDDS17, 175.

Recent data, however, suggests that many LBF, whether they are labelled as ‘supersaturable’ or

not, frequently lose drug solubilisation capacity during GI processing and have the potential to

trigger drug supersaturation. Solubilisation capacity can be lost via initiation of the digestion of

glyceride lipids91, 176 and/or surfactants177, and the dilution of co-solvents and surfactants5, 178. LBF

that lose solubilisation capacity in vivo may therefore benefit from supersaturation-enhanced drug

absorption, provided that precipitation is sufficiently delayed. In this regard, a range of polymers

have been explored for their utility in supersaturation stabilisation during GI processing of LBF,

with cellulose-based polymers appearing to be the most useful to date179.

In contrast, the possibility that supersaturation-enhanced absorption may play a role in the

absorption of solubilised drug from intestinal micellar and vesicular species has not been explored

in detail. Importantly, for a solubilised system, supersaturation is expected to increase Cfree above

drug aqueous solubility. This may provide a means to reverse the reduction in drug

thermodynamic activity resulting from solubilisation, and explain the paradox that is the ability of

solubilising LBF to increase the absorption of many PWSD despite apparent reductions in drug

thermodynamic activity (see Figure 1.3(C), red dotted arrows). In essence, the stimulation of

supersaturation in stable colloidal systems may provide a means to mobilise the solubilised drug

reservoir such that solubilised drug is more readily available for absorption. Since lipid-based

micelles and vesicles are dynamic structures that are in constant interaction with lipid dispersion,

digestion and absorption processes in the GI environment, changes to solubilisation capacity (and

therefore supersaturation opportunities) are likely to exist within endogenous lipid processing

pathways. This thesis has examined these possibilities in detail.


59

1.6 STRUCTURE OF THIS THESIS

Chapters 3, 4 and 5 of this thesis have been directly reproduced from published or submitted

manuscripts. The constraints of journal publication dictate that these chapters focus on the results,

discussions, and conclusions of the experiments, but include relatively limited detail of the

methods employed and validation studies. A general methods chapter (Chapter 2) has therefore

been added to further describe the method development and validation studies conducted. The

thesis concludes with a final summary and perspectives chapter (Chapter 6).

60

CHAPTER 2 : GENERAL METHODS

Chapter 2: General Methods

61

2.1 INTRODUCTION

This general methods chapter contains further information on the in situ autoperfused rat jejunum

preparation, brief outline of the surgical procedures required for the conduct of in vivo

bioavailability studies, and a detailed description and validation of the analytical methods used

throughout this thesis. Other experimental methods have been included in the experimental

chapters (3 through 5) which are reproduced from published or submitted manuscripts.

2.2 AUTOPERFUSED RAT JEJUNUM

2.2.1 Materials

Sodium chloride (NaCl), D-mannitol, antipyrine, D-mannitol, [1-14C] (46.6 mCi/mmol),

antipyrine, [3-14C] (5.4 mCi/mmol) (Sigma-Aldrich, Australia), sodium hydroxide pellets (NaOH)

and ethanol 96% v/v (Merck, Australia), disodium hydrogen orthophosphate (Na2HPO4) and

sodium dihydrogen orthophosphate (NaH2PO4.2H2O) (Ajax Finechem, Australia), 0.9% Sodium

Chloride Intravenous Infusion BP (Baxter, Australia), polyethylene glycol, [1,2-3H] (1.50 mCi/g),

Irga-Safe PlusTM, SOLVABLETM (Perkin Elmer Life Sciences, MA, USA), oleic acid, [9,10-

3H(N)] (60 Ci/mmol) (American Radiolabeled Chemicals, MO, USA), heparin sodium injection

BP (1000 I.U./mL, Hospira, Australia), xylazine (100 mg/mL, Troy Laboratories, Australia),

acepromazine (10 mg/mL, Ceva Delvet, Australia), ketamine (100 mg/mL, Provet, Australia) and

pentobarbitone sodium (325 mg/mL, Virbac, Australia) were obtained from listed suppliers. Water

was obtained from a Millipore milliQ Gradient A10 water purification system (Millipore, MA,

USA).


62

2.2.2 Methods

2.2.2.1 Surgical equipment and accessories

The surgical instruments required for the setup of the autoperfused rat jejunum include a scalpel, a

pair of sharp straight scissors, a pair of iridectomy scissors, two curved anatomical forceps and

one fine tip curved forceps. A portable, multi-head, flexible cold fibre-optic illuminating system

(Microlight 150®, Fibre optic light guides, Australia) was used to obtain adequate lighting for

surgery. A heated surgical board (Ratek Instruments, Australia) was used to maintain body

temperature of the animals throughout surgery and experiments.

2.2.2.2 Animal care and anaesthesia

All rat studies were approved by the institutional animal ethics committee, and were conducted in

accordance with the guidelines of the Australian and New Zealand Council for the Care of

Animals in Research and Teaching. Male Sprague-Dawley rats (280–330 g) were used in all

experiments, and were allowed to acclimatize in the institutional animal housing facility for at

least 7 days with free access to standard chow and water. All animals were fasted overnight (12–

18 h) prior to surgery. Anaesthesia was induced in rats by subcutaneous injection of 1.0 mL/kg of

‘Cocktail I’ (37.3 mg/mL ketamine, 9.8 mg/mL xylazine, 0.4 mg/mL acepromazine in saline), and

maintained throughout the study with subcutaneous doses of 0.44 mL/kg of ‘Cocktail II’ (90.9

mg/mL ketamine, 0.9 mg/mL acepromazine) when required. To prepare for surgery, rats were

shaved and cleaned aseptically with 70% v/v ethanol on both the abdomen and the skin overlaying

the trachea, and placed in dorsal recumbency. Throughout surgery and experiment, the body

temperature of rats was maintained at 37 ºC.


63

2.2.2.3 Surgical technique

The autoperfused rat jejunum model involves the perfusion of an isolated jejunal segment and

simultaneous blood collection from the corresponding mesenteric vein branch (see Figure 2.1).

The surgical procedures for the perfusion studies are similar to those described by Singhal et al.

with slight modifications180. Firstly, the right jugular vein was cannulated to enable infusion of

donor rat blood (the surgical procedures for jugular vein cannulation is detailed in Section 2.3.2.5).

The small intestine was then exposed by a longitudinal midline incision to the abdomen, and a

piece of jejunal segment (~ 10 cm) carefully isolated and externalised. Access to the lumen was

made at the proximal and distal ends by electrocautery, and jejunal contents were flushed out with

warm perfusion buffer (150 mM Na+, 18 mM H2PO4-, 12 mM HPO42-, 108 mM Cl-, adjusted to

pH 6.30 ± 0.01 with NaOH solution). The segment was then cannulated at the proximal and distal

ends with sections of Teflon tubing (0.03-inch i.d. proximal/inlet, Upchurch Scientific, Oak

Harbor, WA, Australia; 0.0625-inch i.d. distal/outlet, Shimadzu, Kyoto, Japan) which were

secured with surgical sutures. The mesenteric vein that drains the cannulated jejunal segment was

then identified, and prepared for catheterisation by careful blunt dissection of the surrounding

connective tissue under a dissecting microscope. The rat was then heparinised (90 I.U./kg) via the

jugular vein using an appropriate volume of a 100 I.U./mL heparinised saline solution, and the

mesenteric vein immediately catheterised under the dissecting microscope with a 24G intravenous

catheter (AngiocathTM, BD, Australia), which had the top 1.5 cm tip cut so that the tip remained in

the mesenteric vein after the guide needle was withdrawn. A drop of instant cyanoacrylate glue

was placed over the site of catheterisation, the guide needle removed, and a 30 cm piece of

silicone tubing (0.025-inch i.d., Helix Medical, CA, USA) attached to the catheter tip for the

collection of venous blood. Immediately following catheterisation of the mesenteric vein and for

the remainder of the experiment, the jugular vein cannula was connected to a peristaltic pump

(Adelab Scientific, SA, Australia), and rats were infused with heparinised donor rat blood (5


64

I.U./mL) at a rate of 0.3 mL/min (this rate was selected to match the outflow of the mesenteric

blood). Donor rat blood was collected from donor rats via cardiac puncture under isoflurane

anaesthesia (IsoFlo®, Abbott Laboratories, IL, USA) immediately prior to surgery. The

temperature of the animal and preparation were maintained by a heated surgical board and lamp,

and the exposed jejunal segment was covered by saline-soaked gauze.

Figure 2.1: Autoperfused rat jejunum preparation depicting mesenteric venous blood collection

from the perfused jejunal segment (~ 10 cm). The model allows for simultaneous perfusion of the

jejunal segment and collection of mesenteric venous blood, which enables analysis of compound

absorption in a closed system.

2.2.2.4 Variations in the setup of the autoperfused rat jejunum

Various perfusion techniques may be used to conduct the autoperfused rat jejunum experiments,

including single-pass, recirculating, oscillating and closed loop perfusion181. Among the

techniques mentioned, single-pass and recirculating perfusion are most commonly employed182, 183,

and were established to determine their suitability as model systems to collect the data required for

this thesis. During single-pass perfusion, perfusate from a reservoir is continuously flowed


65

through the jejunal segment and outflowing perfusate continuously collected (Figure 2.2A). Thus,

the jejunal segment is exposed to fresh perfusate containing a constant concentration of compound

at all times (although the concentration of compound decreases during intestinal transit due to

absorption). On the other hand, during recirculating perfusion, outflowing perfusate is recirculated

back to the perfusate reservoir, and re-perfused into the jejunal segment (Figure 2.2B). Thus,

assuming that the compound of interest is absorbed, the intestine is exposed to a gradually

declining concentration of compound. As a result, the mathematical modelling for single-pass and

recirculating techniques are slightly different. Both techniques, however, have been satisfactorily

employed to study drug absorption kinetics in the small intestine182.

2.2.2.5 Experimental protocol for single-pass jejunal perfusion

The animals were equilibrated for 30 min after surgery. During this time, blood that was collected

from the cannulated mesenteric vein (~ 0.3 mL/min) was mixed with donor rat blood and re-

infused via the jugular vein. The inlet jejunal cannula was connected to a second peristaltic pump

and perfusion buffer was pumped through the jejunal segment at a rate of 0.1 mL/min. Outflowing

buffer during the equilibration period was discarded to waste. After the equilibration period and

immediately before the experiment was commenced (i.e. immediately before the solution of

interest was perfused), air was pumped through the jejunal segment for 10 sec at 0.1 mL/min, to

create an air pocket in the outflowing perfusate which differentiated outflowing perfusion buffer

and outflowing solution of interest during the first perfusate sampling interval. Outflowing

perfusate was continuously collected into 1.5 mL polypropylene tubes at 10-min intervals whilst

venous blood draining the perfused jejunal segment was collected into pre-weighed 1.5 mL

polypropylene tubes at 5-min intervals and weighed. The setup of the single-pass autoperfused rat

jejunum preparation is shown in Figure 2.3.


66

At the end of the experiment, the animal was euthanised via an intravenous injection of 100 mg

sodium pentobarbitone into the jugular vein. The jejunal segment was flushed with 10 mL warm

perfusion buffer followed by 10 mL air, excised, and dissected longitudinally for accurate

determination of surface area (calculated by multiplying the diameter by the length of the perfused

intestinal segment). The jejunal segment was then weighed and stored at -20 ⁰C for future analysis

of compound content.

2.2.2.6 Experimental protocol for recirculating jejunal perfusion

The experimental protocol for recirculating jejunal perfusion was similar to single-pass perfusion

(described in Section 2.2.2.5), except that perfusate was passed through the jejunal segment at a

rate of 0.6 mL/min, and outflowing perfusate was recirculated back to a reservoir (typically 10 mL)

and re-perfused into the jejunal segment (see Figure 2.2B). During the 30-min equilibration period,

perfusion buffer was recirculated through the jejunal segment at a rate of 0.6 mL/min.

Immediately before the experiment was commenced (i.e. immediately before the solution of

interest was perfused), air was pumped through the jejunal segment for 5 sec at 0.6 mL/min. The

perfusate reservoir was sampled at 10-min intervals (typically 100 µL aliquots) whilst venous

blood draining the perfused jejunal segment was collected into pre-weighed 1.5 mL polypropylene

tubes at 5-min intervals and weighed.


67

(A)

(B)

Figure 2.2: Schematic of autoperfused rat jejunum with (A) single-pass perfusion, and (B)

recirculating perfusion. In single-pass perfusion, perfusate is flowed through the isolated jejunal

segment once and collected. In recirculating perfusion, outflowing perfusate is redirected back to

the perfusate reservoir and re-perfused into the isolated jejunal segment. Diagrams are adapted

from Cummins et al.184.


68

Figure 2.3: Setup of the single-pass autoperfused rat jejunum preparation.

2.2.3 Data analysis

2.2.3.1 Permeability calculations for single-pass perfusion

In the single-pass perfusion model, permeability coefficients were calculated after attainment of

steady state drug absorption, and using steady state compound concentrations in perfusate and

plasma. Two permeability coefficients were calculated to describe compound absorption from the

jejunal lumen into the mesenteric venous blood as described previously184 and as follows:

′Disappearance′P . ln Equation 2.1

′Appearance′P∆

∆.

Equation 2.2


69

where ‘Disappearance’ Papp is the apparent permeability coefficient calculated from compound

loss from the perfusate (cm/sec); ‘Appearance’ Papp is the apparent permeability coefficient

calculated from compound appearance in the mesenteric blood (cm/sec); Q is the perfusate flow

rate (mL/sec); A is the surface area of the perfused jejunal segment (cm2), which is calculated by

multiplying the diameter and the length of the perfused intestinal segment; C1 is the average

steady state compound concentration exiting the perfused jejunal segment (ng/mL); C0 is the

compound concentration entering the jejunal segment (ng/mL); ∆MB/∆t is the average rate of

compound mass appearance in mesenteric venous blood at steady state (ng/sec); and <C> is the

logarithmic mean compound concentration in the lumen (ng/mL), where <C> = (C1 – C0) / (ln C1

– ln C0).

2.2.3.2 Permeability calculations for recirculating perfusion

Similar to single-pass perfusion, two permeability coefficients were calculated in the recirculating

perfusion model to describe compound absorption from the jejunal lumen into the mesenteric

venous blood. The disappearance permeability coefficient was assessed from the rate of

compound loss from the perfusate reservoir185; while the appearance permeability coefficient was

calculated from cumulative compound appearance data in the mesenteric venous blood:

′Disappearance′P . k Equation 2.3

′Appearance′P .. Equation 2.4

where ‘Disappearance’ Papp is the apparent permeability coefficient calculated from compound

loss from the perfusate (cm/sec); ‘Appearance’ Papp is the apparent permeability coefficient

calculated from compound appearance in the mesenteric venous blood (cm/sec); V is the effective


70

volume of recirculating perfusate (mL); A is the surface area of the perfused jejunal segment

(cm2); ku is the first order disappearance rate constant (sec-1), calculated from the slope of a loge

perfusate concentration vs. time plot; Co is the initial perfusate concentration (ng/mL); and dM/dt

(ng/sec) is the appearance flux of compound in the mesenteric venous blood, calculated from the

slope of a cumulative mass transported vs. time plot.

2.2.4 Validation of the autoperfused rat jejunum

2.2.4.1 Passive permeability markers

The autoperfused rat jejunum was validated via comparison of calculated disappearance and

appearance permeability coefficients for the model compounds antipyrine and mannitol to

published data using the recirculating perfusion technique. Antipyrine is passively absorbed via

the transcellular route (i.e. diffusion across the apical and basolateral membranes of enterocytes)

whereas mannitol is passively absorbed via the paracellular route (i.e. diffusion through the tight

and gap junctions between enterocytes). Antipyrine and mannitol thus act as markers for passive

permeability via the transcellular and paracellular route. Since paracellular transport comprises ~

0.01% of the total surface area available for absorption in the small intestine186, 187, the intestinal

permeability of mannitol is expected to be significantly lower than antipyrine. Comparison of the

disappearance and appearance permeability coefficients of antipyrine and mannitol to literature

values enables assessment of the reproducibility of the autoperfused rat jejunum model and the

integrity of the intestinal segments during perfusion experiments180, 182, 188.

In the validation experiments, the perfusate reservoir contained 10 mL of 1 mM mannitol in

perfusion buffer spiked with 1 µCi 14C-mannitol or 10 mL of 1 mM antipyrine in perfusion buffer

spiked with 1 µCi 14C-antipyrine. The experiments were conducted over 60 min, during which


71

time 100 µL aliquots of perfusate were sampled from the reservoir at t = 0 and at 10 min intervals

whilst blood draining the jejunal segment was collected every 5 min and plasma separated by

centrifugation (10,000 xg, 5 min). At the end of the experiment, the perfused jejunal segment was

removed and longitudinally dissected for accurate determination of surface area and quantification

of radiolabel. For quantification of antipyrine or mannitol in perfusate and plasma, 100 µL of

perfusate samples and 500 µL of plasma samples were placed in liquid scintillation (LSC) vials

containing 2 mL Irga-safe Plus scintillation fluid, and vortexed for 10 sec. For quantification of

antipyrine or mannitol in the perfused jejunal segment, the segment was placed in a 20 mL glass

vial, 2 mL of SOLVABLETM added, and the vial was incubated at 40 ºC for 48 h. The dissolved

tissue solution was divided into two equal portions where 12 mL of Irga-safe Plus was added to

each vial followed by a 1-min vortex. Radioactivity of the perfusate, plasma, and intestine samples

were quantified via scintillation counting on a Packard Tri-Carb 2000CA liquid scintillation

analyser (Packard, Meriden, CT, USA). Blood concentrations were calculated by multiplying the

plasma concentrations of antipyrine and mannitol and their respective blood:plasma concentration

ratios, which are determined in Section 2.2.4.2.

The perfusate disappearance and blood appearance profiles for antipyrine and mannitol are shown

in Figure 2.4, and calculated permeability coefficients are in Table 2.1. The apparent permeability

coefficient of antipyrine based on disappearance from the perfusate (~ 45 x 10-6 cm/sec) and the

apparent permeability coefficient of mannitol based on appearance in the mesenteric venous blood

(~ 1.9 x 10-6 cm/sec) were comparable to reported values in the literature (47-73 x 10-6 cm/sec182,

183 and 1 x 10-6 cm/sec180, respectively). The permeability coefficient of antipyrine based on

appearance in the mesenteric blood is not available in the literature to date, and the permeability

coefficient of mannitol based on disappearance from the perfusate was not calculated due to its

low disappearance from the perfusate (see Figure 2.4A) which precluded accurate assessment of


72

the permeability coefficient. The intestinal permeability of antipyrine was therefore significantly

higher than mannitol, consistent with their respective mode of absorption as discussed earlier. The

total recoveries (i.e. the sum of recovery percentages in the perfusate, blood, and the perfused

jejunal segment) of antipyrine and mannitol in the autoperfused rat jejunum preparation were

93.37% ± 8.25% and 97.02% ± 0.20% (mean ± SEM, n =3), respectively. A summary of the mass

balance for mannitol and antipyrine is shown in Table 2.2.

Thus, the permeability data obtained using mannitol and antipyrine validate the reproducibility of

the autoperfused rat jejunum and the integrity of the intestinal segments during perfusion

experiments. Although the validation experiments were conducted using the recirculating

perfusion model, the data also apply for the single-pass perfusion model, since previous studies

that compared the permeability of antipyrine using both techniques reported statistically equal

apparent permeability coefficients3, and the apparent permeability coefficient of mannitol obtained

in this thesis (using the recirculating model) was comparable to literature value that employed the

single-pass model180.


73

(A)

(B)

Figure 2.4: (A) Disappearance profiles of antipyrine (filled circles) and mannitol (open circles)

from the perfusate and (B) Appearance profiles of antipyrine (filled circles) and mannitol (open

circles) in mesenteric venous blood during recirculating perfusion of an isolated ~ 10 cm segment

of rat jejunum. Data represent mean ± SEM of n = 3 experiments, linear regression lines utilised in

calculation of permeability coefficients are shown as solid (antipyrine) or dotted (mannitol) lines.

Time (min)

0 10 20 30 40 50 60

Per

cent

initi

al

perf

usat

e co

ncen

trat

ion

(%)

50

60

70

80

90

100

AntipyrineMannitol

Time (min)

0 10 20 30 40 50 60

Cu

mu

lativ

e tr

ansp

ort

into

mes

ente

ric b

lood

(%

dos

e/10

cm

2 )

0

5

10

15

20

AntipyrineMannitol


74

Table 2.1: Permeability coefficients of antipyrine and mannitol based on disappearance from the

perfusate and appearance in mesenteric venous blood during recirculating perfusion of an isolated

rat jejunal segment (~ 10 cm). Data represent mean ± SEM of n = 3 experiments.

Disappearance Papp

(x 106 cm/sec)

Appearance Papp

(x 106 cm/sec)

Antipyrine 45.49 ± 7.86 38.27 ± 5.56

Mannitol n/a a 1.94 ± 0.42

a The disappearance Papp for mannitol could not be determined accurately due to low disappearance from

the perfusate

Table 2.2: Mass balance for mannitol and antipyrine in the autoperfused rat jejunum preparation.

Total mass balance (as a %) is the sum of percentages recovered in the perfusate, mesenteric

venous blood and perfused jejunal segment at the end of the perfusion experiments. Data represent

mean ± SEM values for n = 3 experiments.

Antipyrine Mannitol

% Recovery in perfusate 75.24 ± 6.89 94.38 ± 0.43

% Recovery in blood 16.78 ± 2.32 1.13 ± 0.20

% Recovery in jejunal segment 1.35 ± 0.14 1.50 ± 0.16

Total Mass Balance 93.37 ± 8.25 97.02 ± 0.20


75

2.2.4.2 Blood:plasma ratio determination

Calculation of appearance permeability coefficients requires accurate quantitation of total

compound transport into mesenteric venous blood. Since compound concentrations were assayed

in plasma (and not whole blood), the partitioning behaviour of compounds between whole blood

and plasma was determined such that concentrations of compound in whole blood could be back

calculated. Blood:plasma concentration ratios were therefore determined for mannitol and

antipyrine (for validation experiments), and for cinnarizine, danazol and oleic acid (for

experiments described in Chapters 3, 4 and 5), to enable conversion of plasma concentrations into

blood concentrations.

Briefly, in triplicate, 0.5 mL of blank blood was spiked with known amounts of compound to

achieve low, medium, and high concentrations (the concentrations were chosen based on the likely

amount of compound transported into mesenteric venous blood during permeability experiments)

and gently mixed. The final blood concentrations were 1, 10, 20 µM for mannitol and antipyrine;

20, 100, 500 ng/mL for cinnarizine and danazol; and 1, 5, 10 µg/mL for oleic acid. Traces of

radiolabelled mannitol, antipyrine and oleic acid were also spiked into blank blood to facilitate

quantitation. Blood samples were centrifuged at 10,000 xg for 5 min, and plasma concentrations

of compound were assayed by scintillation counting (mannitol, antipyrine and oleic acid), HPLC

(cinnarizine), or LC-MS (danazol). The blood:plasma ratio was calculated from the ratio of known

concentration in spiked blood to the concentration measured in plasma separated from spiked

blood. The blood:plasma ratios for mannitol, antipyrine, cinnarizine, danazol and oleic acid were

0.48 ± 0.02, 1.23 ± 0.03, 0.68 ± 0.03, 0.65 ± 0.01, and 1.24 ± 0.00 (mean ± SEM, n = 3),

respectively, and were not concentration-dependent within the tested range.


76

2.2.4.3 Assessment of net water flux

Changes in perfusate volume (due to water absorption from or water secretion into the lumen of

the perfused jejunal segment) affect the concentration of compound in the perfusate, and therefore

impact accurate calculation of disappearance permeability coefficients. For example, water

secretion into the lumen increases perfusate volume and lowers the concentration of compound in

the perfusate. If correction factors were not applied to the measured concentrations, compound

loss from the perfusate (i.e. mass of compound assumed to be absorbed) would be overestimated,

leading to overestimation of disappearance permeability. Therefore, net water flux across the

perfused jejunal segment is often determined during individual experiments via the inclusion of a

non-absorbable marker, such as PEG 4000189, in the perfusate. The concentration of the non-

absorbable marker is only expected to change when there is net water flux into or out of the

perfused jejunal segment. Thus, changes in the concentration of a non-absorbable marker directly

reflect volume changes in the perfused jejunal segment, and correction factors may be

subsequently calculated to adjust the concentrations of compound of interest in the perfusate at

each sampling point.

In the experiments conducted in this thesis however, the routine inclusion of a radiolabelled

marker of water flux (e.g. 14C- or 3H-PEG 4000) in each experiment was impractical since other

radiolabelled compounds (e.g. 14C-cholesterol and 3H-oleic acid) were also often included in the

perfusate. Net water flux was therefore determined in separate validation experiments, where 14C-

PEG 4000 was perfused through isolated jejunal segments using the recirculating perfusion model

(a total of five experiments were conducted). The recirculating perfusate reservoir consisted of 10

mL perfusion buffer spiked with 0.25 µCi 14C-PEG 4000. The experiments were conducted over

60 min, and 100 µL aliquots of perfusate were sampled from the reservoir at t = 0 and at 10 min


77

intervals. The perfusate samples were analysed for radiolabelled PEG 4000 content as described in

Section 2.2.4.1.

A plot of mean calculated perfusate volume vs. time is shown in Figure 2.5. The plot shows that

significant volume gain mainly occurred in the first 10 min of the experiment, and the perfusate

volume was relatively constant for the remainder of the experiment (relative to the calculated

volume at 10 min, percent volume changes from 10-60 min were < 2%, a value below

experimental error limits). The volume gain in the first 10 min likely reflects the incorporation of

small amounts of perfusion buffer that remained in the jejunal segment after the equilibration

period. The PEG 4000 experiments therefore suggested that no significant water flux occurred

between 10 and 60 min. As such, rather than applying correction factors at each sampling point,

the analysis of perfusate compound disappearance was modified slightly such that calculations of

disappearance permeability coefficients were performed using compound concentrations obtained

between 10 and 60 min (instead of between 0 and 60 min) for recirculating perfusion experiments.

For single-pass perfusion, since the analysis of perfusate compound disappearance was performed

after drug loss from perfusate reached steady state (a process that typically takes > 30 min as

exemplified by perfusate disappearance profiles in the experimental chapters), no modifications to

the method of analysis was required.


78

Figure 2.5: Mean plot of calculated perfusate volume vs. sampling time points for five perfusion

experiments where 0.25 µCi of 14C-PEG 4000, a non-absorbable marker, was included in the

perfusate and recirculated through ~ 10 cm segments of rat jejunum. Broken line represents

theoretical volume of perfusate in the absence of net water flux. A trend of water gain was evident

at all sampling time points. Data represent mean ± SEM values.

2.2.5 Adsorption of cinnarizine onto the recirculating perfusion apparatus

Validation of the autoperfused rat jejunum preparation was performed using the recirculating

perfusion model (see Section 2.2.4) since the model provides better simulations of the absorption

process in vivo, where compound absorption is continuous along the length of the small intestine.

However, in subsequent experiments where drug-containing lipid colloidal phases were perfused,

the model poorly water-soluble drug (PWSD), cinnarizine, was found to adsorb extensively to the

recirculating perfusion apparatus. Specifically, in sham experiments where cinnarizine was

solubilised in intestinal lipid colloidal phases (composition shown in Figure 2.6) and pumped

through the recirculating perfusion apparatus only (i.e. through the perfusion tubing but not

through the intestinal segment of an animal) at a rate of 0.6 mL/min, cinnarizine concentration in

Time (min)

0 10 20 30 40 50 60

Cal

cula

ted

perf

usat

e vo

lum

e (m

L)

9.0

9.5

10.0

10.5

11.0


79

the perfusate declined appreciably, with > 60% loss over 2 h. Since the stability of cinnarizine in

the phases was validated for a period of 24 h (data not shown), the decrease in perfusate

concentration of cinnarizine was attributed to adsorption onto the recirculating apparatus,

presumably to the relatively lengthy silicone tubing that was required for the recirculating model

setup.

To overcome the problem of cinnarizine adsorption onto the recirculating apparatus, the duration

of the sham experiment was increased to 6 h in an attempt to saturate the tubing with cinnarizine,

so that a constant cinnarizine perfusate concentration could be achieved before perfusate was

passed through the rat jejunal segment. This approach, however, did not prove viable as the

concentration of cinnarizine decreased to ~ 10% of initial after 5 h (Figure 2.6A), and was too low

to be used in the experiments. Thus, the recirculating model was deemed unsuitable for the jejunal

perfusion of cinnarizine in intestinal lipid colloidal phases, and the single-pass model was trialled

as an alternate method instead. The single-pass perfusion setup required shorter lengths of

polyethylene tubing, and perfusate is only passed through the tubing once. Therefore, cinnarizine

adsorption was expected to occur to a lesser extent than with the recirculating perfusion setup.

Indeed, a repeat sham experiment using the single-pass perfusion setup where cinnarizine was

solubilised in the same lipid colloidal phase and pumped through the perfusion apparatus at a rate

of 0.1 mL/min revealed negligible cinnarizine adsorption, and the concentration of cinnarizine in

the perfusate was within ± 10% of initial throughout the experiment (Figure 2.6B). Similar

outcomes were later obtained for another model PWSD, danazol, where no adsorption was

observed during the sham perfusion of danazol containing-intestinal lipid colloidal phase

(composition shown in Figure 2.7) through the perfusion apparatus at a rate of 0.1 mL/min (Figure

2.7).


80

Therefore, while validation of the animal model was carried out using the recirculating perfusion

setup, the single-pass perfusion setup was the method employed to investigate the absorption

kinetics of cinnarizine and danazol from intestinal lipid colloidal phases.

(A)

Time (mins)

0 100 200 300

Per

cent

initi

al

cinn

ariz

ine

conc

entr

atio

n (%

)

0

20

40

60

80

100

(B)

Time (min)

0 20 40 60 80 100 120

Per

cent

initi

alci

nnar

izin

e co

ncen

trat

ion

(%)

50

60

70

80

90

100

110

Figure 2.6: Plot of percent initial cinnarizine concentration in perfusate vs. time during an

adsorption test experiment where an intestinal lipid colloidal phase* containing 10 µg/mL

cinnarizine was flowed through (A) the recirculating perfusion apparatus for 6 h, and (B) the

single-pass perfusion apparatus for 2 h. In the recirculating perfusion setup, samples were taken at

t = 0, at hourly intervals for the first 5 h, and at 10 min intervals thereafter to reflect the approach

of attempting to saturate the adsorption of cinnarizine onto the perfusion apparatus for 5 h prior to

commencement of a 1 h jejunal perfusion experiment. Even though steady state (i.e. constant)

concentrations appeared to have been reached after 5 h, the concentration (~ 10% of initial) was

deemed too low for the perfusion experiments. In the single-pass perfusion setup, samples were

taken at t = 0 and at 10 min intervals for 2 h, and negligible cinnarizine adsorption was observed

during the sham experiment.

* Intestinal lipid colloidal phase consisted of 0.025 %w/v oleic acid and 0.016 %w/v monoolein solubilised

in 4 mM sodium taurocholate, 1 mM lysophosphatidylcholine and 0.25 mM cholesterol at pH 6.50.


81

Time (min)

0 10 20 30 40 50 60 70

Per

cent

initi

alda

nazo

l con

cent

ratio

n (%

)50

60

70

80

90

100

110

Figure 2.7: Plot of percent initial danazol concentration in perfusate vs. time during an adsorption

test experiment where an intestinal lipid colloidal phase** containing 14.4 µg/mL danazol was

flowed through the single-pass perfusion apparatus for 70 min. Samples were taken at t = 0 and at

10 min intervals, and negligible danazol adsorption was observed during the sham experiment.

** Intestinal lipid colloidal phase consisted of 0.10 %w/v oleic acid and 0.064 %w/v monoolein solubilised

in 4 mM sodium taurocholate, 1 mM lysophosphatidylcholine and 0.25 mM cholesterol at pH 6.30.

2.3 IN VIVO BIOAVAILABILITY STUDIES

To determine whether absorption trends observed using the in situ autoperfused rat jejunum

preparation could be extended to the in vivo setting, the bioavailability of cinnarizine and danazol

was assessed in rats following intraduodenal administration in intestinal lipid colloidal phases. In

the case of danazol, intraduodenal administration did not lead to detectable systemic plasma

concentrations (limit of quantification for danazol plasma assay was 5 ng/mL) despite significant

absorption into the mesenteric venous blood during perfusion experiments (suggesting significant

hepatic first-pass metabolism). Thus, bioavailability data for danazol is not included in this thesis.

The experimental protocol for the bioavailability studies is described in detail in Chapters 3 and 4.

This section contains details of the surgical procedures required for the setup of the animal model.


82

2.3.1 Materials

Heparin, saline, ketamine xylazine, acepromazine were sourced from suppliers listed in Section

2.2.1.

2.3.2 Methods

The right carotid artery, right jugular vein, common bile duct and duodenum were cannulated for

the purpose of: blood sample withdrawal, intravenous administration of formulation or saline,

pancreatobiliary fluid diversion and intraduodenal administration of formulation, respectively.

2.3.2.1 Conscious state

Studies that assess the bioavailability of drugs in oral formulations are most often conducted using

conscious animals since the rate of gastric emptying, intestinal mixing and digestion are likely to

be reduced in anaesthetised animals. The formulations used in the experiments throughout this

thesis, however, were highly dispersed intestinal lipid colloidal phases assembled from fully

digested lipids, and as such, drug absorption from the formulations is less likely to be affected by

the conscious state of the animals. In addition, the protocol for the bioavailability studies

necessitated abdominal incisions to be made to the animals (for the diversion of pancreatobiliary

fluids as well as the intraduodenal administration of formulations - to avoid the effects of gastric

mixing since the formulations were designed to simulate colloidal phases found in the small

intestine). These surgical interventions were expected to significantly enhance the degree of post-

operative pain and stress in the animals. Therefore, bioavailability studies were conducted using

anaesthetised rats to reduce the logistical and ethical burden of the experiments, and to match the

conditions used in the in situ jejunal perfusion experiments.


83

2.3.2.2 Surgical equipment and accessories

The surgical instruments required for the cannulation of the carotid artery, jugular vein, common

bile duct and duodenum include a scalpel, a pair of sharp straight scissors, two curved anatomical

forceps, one fine tip curved forceps, one fine tip straight forceps, a pair of iridectomy scissors, one

fine tip haemostat and one small ‘alligator’ clamp. A portable, multi-head, flexible cold fibre-optic

illuminating system (Microlight 150®, Fibre optic light guides, Australia) was used to obtain

adequate lighting for surgery. A heated surgical board (Ratek Instruments, Australia) was used to

maintain body temperature of the animals during surgery.

2.3.2.3 Animal care and anaesthesia

The care and anesthesia of animals have been described in Section 2.2.2.2.

2.3.2.4 Cannulation of the right carotid artery

A longitudinal incision (1.5 cm) was made to the right of the central neck muscle, starting from

above the right breast bone towards the head. The carotid artery (visualised as a red, pulsating

vessel running parallel to the white vagus nerve ~ 1 cm below the skin surface) was isolated by

blunt dissection of surrounding connective tissue and muscle layers. Particular care was taken in

separating the carotid artery from the vagus nerve. Blood flow to the artery from the heart was

temporarily halted by the placement of a pair of fine tip straight forceps (bridged over a 10 mL

syringe to provide leverage) underneath the vessel. A small incision was made on the top surface

of the vessel using iridectomy scissors. A bevelled 30 cm piece of polyethylene tubing (0.50 mm

i.d., 0.80 mm o.d., Microtube Extrusions, Australia; attached to a 25G syringe containing 2

I.U/mL heparinised saline) was then inserted into the incision, advanced 2.5 cm into the vessel,

and secured in place with surgical sutures (with the aid of a haemostat if needed). To ensure


84

cannula patency, the cannula was flushed with 0.2 mL heparinised saline and sealed using a naked

flame. The neck incision was closed with surgical suture.

2.3.2.5 Cannulation of the right jugular vein

A longitudinal incision (1.5 cm) was made to the right of the central neck muscle, starting from

above the right breast bone towards the head. The jugular vein (visualised as a bluish-grey vessel

close to the skin surface) was isolated by blunt dissection of surrounding connective tissue and

muscle layers. A ligature was placed around the vessel (using surgical suture), anterior to the site

of cannulation, to halt blood return from the brain via the right jugular vein. A small incision was

made on the top surface of the vessel using iridectomy scissors. A bevelled 30 cm piece of

polyethylene tubing (0.50 mm i.d., 0.80 mm o.d., Microtube Extrusions, Australia; attached to a

25G syringe containing 2 I.U/mL heparinised saline) was then inserted into the incision, advanced

2 cm into the vessel, and secured in place with surgical suture. To ensure cannula patency, the

cannula was flushed with 0.2 mL heparinised saline and sealed using a naked flame. The neck

incision was closed with surgical suture.

2.3.2.6 Cannulation of the common bile duct

The abdominal muscle wall was opened by a latitudinal incision (~ 3 cm) 1-2 cm below the

ribcage, approximately 0.5 cm to the left side of the animal from the midline. The duodenal loop

was isolated and exteriorised to expose the common bile duct, which may be visualised as a small,

yellow vessel originating from the liver and draining into the duodenum. The duct was cleared of

surrounding connective tissue using forceps, and a small puncture hole made on the top surface,

close to the duodenum, with a 25G needle. A bevelled 15 cm piece of polyethylene tubing (0.28

mm i.d., 0.68 mm o.d., Microtube Extrusions, Australia) was then inserted into the duct and

secured in place with surgical sutures. The position of the cannula was selected to be below the


85

entry points of all exocrine branches from the liver and pancreas. The cannula was externalised

through the abdominal incision which was subsequently closed with surgical suture.

2.3.2.7 Cannulation of the duodenum

The abdomen muscle wall was opened by a latitudinal incision (~ 2 cm) 1-2 cm below the ribcage,

approximately 0.5 cm to the left side of the animal from the midline. The duodenum (identified as

the brighter pink section of the small intestine, which upon gentle downward pulling reveals the

stomach) was isolated and a small puncture hole made in the duodenum, 1 cm below the pylorus,

with a 23G needle. A bevelled, J-shaped (previously heat-moulded with a cigarette lighter) piece

of polyethylene tubing (0.58 mm i.d., 0.96 mm o.d., Microtube Extrusions, Australia) was then

inserted into the hole and glued in place using instant cyanoacrylate adhesive. The cannula was

externalised through the abdominal incision which was subsequently closed with surgical sutures.

2.4 VALIDATION OF ANALYTICAL METHODS

2.4.1 HPLC assays for quantification of cinnarizine, halofantrine,

fenofibrate, danazol and meclofenamic acid in intestinal lipid colloidal

phases

The concentration of cinnarizine, halofantrine, fenofibrate, danazol and meclofenamic acid in

intestinal lipid colloidal phases was determined by High Performance Liquid Chromatography

(HPLC). The HPLC system used for the quantification of cinnarizine consisted of a Waters 610

fluid unit, Waters 717 autosampler, Model 600 fluid controller (Waters, Milford, MA, USA) and a

RF-10A XL fluorescence detector supplied by Shimadzu (Shimadzu Corp., Kyoto, Japan). Data

was recorded and integrated using Empower 2 personal chromatography data software (Waters,


86

Milford, MA, USA). The HPLC system (Shimadzu Corp., Kyoto, Japan) used for the

quantification of halofantrine, fenofibrate, danazol and meclofenamic acid consisted of a LC‐

20AD binary pump, a SIL‐20A HT autosampler, and a temperature‐controlled column

compartment (CTO‐20A) coupled with a RF-10A XL fluorescence detector and a SPD-20A

UV/Vis detector. The column compartment was maintained at 40 °C. The chromatographic data

were recorded and integrated using LabSolutions software package (Shimadzu). The injection

volume for all samples was 50 µL. The HPLC assay conditions for the five compounds are

described in Table 2.3.

The assays were validated on three separate days by analysing drug standards (prepared at low,

medium and high concentrations) in quadruplicate and comparing against a linear standard curve.

The assays for all five compounds were found to be accurate (to within ± 10% of target

concentration except for the lower limit of quantitation which was ± 15%) and precise (% CV <

10 for all concentrations except the lower limit of quantification which was < 15) for the

concentration ranges reported in Table 2.3. The validation results are summarised in Table 2.4.


87

Table 2.3: HPLC assay conditions for cinnarizine (CIN), halofantrine (HF), fenofibrate (FF),

danazol (DAN) and meclofenamic acid (MFA) in intestinal lipid colloidal phases.

Drug Stationary phase

Mobile phase Flow rate (mL/min)

Detection Retention time (min)

Validated concentration

range

CIN Waters Symmetry®

C18, 5 µm, 3.9 x 150 mm

50% v/v Acetonitrile : 50%

v/v 20 mM NH4H2PO4

1.0

Fluorescence

λ = 249/311 nm

5.7 20–1000

ng/mL

HF Phenomenex Luna® C8(2), 5 µm, 4.6 x 250

mm


v/v H2O (with 0.2% w/v SDSa & 0.2% v/v

acetic acid)

1.5 UV

λ = 254 nm

4.2 0.5–25

µg/mL

FF Waters XbridgeTM C18, 5 µm, 4.6 x 150

mm


v/v H2O (total 0.01% v/v formic acid)

1.0 UV

λ = 286 nm

3.9 0.5–25

µg/mL

DAN Waters XbridgeTM C18, 5 µm, 4.6 x 150

mm

75% v/v Methanol : 25% v/v H2O

1.0 UV

λ = 286 nm

4.4

0.25–50

µg/mL

MFA Phenomenex Luna® C8(2), 5 µm, 4.6 x 250

mm

80% v/v Acetonitrile : 20% v/v 10 mM H3PO4

1.0 UV

λ = 285 nm

4.8 0.5–25

µg/mL

a SDS is sodium dodecyl sulphate


88

Table 2.4: Precision (% coefficient of variation) and accuracy (% of the target value) of the HPLC

assays for cinnarizine (CIN), halofantrine (HF), fenofibrate (FF), danazol (DAN) and

meclofenamic acid (MFA) in intestinal lipid colloidal phases.

Drug Target concentration

Day 1 Day 2 Day 3

Precision Accuracy Precision Accuracy Precision Accuracy

CIN 50 ng/mL 1.9 107.7 1.5 109.3 0.8 104.1

200 ng/mL 1.0 95.1 0.5 93.4 0.8 92.6

1000 ng/mL 0.2 101.8 0.2 104.5 1.2 106.1

HF 0.5 µg/mL 0.5 102.4 0.7 103.6 0.4 103.3

5 µg/mL 0.1 100.4 0.2 100.1 0.4 100.5

25 µg/mL 0.2 100.1 0.2 100.4 0.5 101.2

FF 0.5 µg/mL 0.2 94.4 0.4 86.1 0.5 86.2

5 µg/mL 0.2 100.0 0.0 97.5 0.0 97.8

25 µg/mL 0.2 100.4 0.0 99.5 0.1 99.7

DAN 0.25 µg/mL 0.3 110.0 0.2 108.6 0.3 92.3

5 µg/mL 0.2 99.4 0.2 98.9 0.1 97.8

50 µg/mL 0.1 100.0 0.1 100.1 0.8 100.4

MFA 0.5 µg/mL 0.5 99.7 0.3 96.8 0.3 113.1

5 µg/mL 0.6 98.1 1.6 95.8 1.0 97.8

25 µg/mL 0.6 101.0 0.1 99.7 0.1 100.2


89

2.4.2 HPLC assay for quantification of cinnarizine in rat plasma

Cinnarizine plasma samples were prepared for HPLC analysis using a validated extraction

procedure. Standards for the assay were prepared by spiking 100 μL aliquots of blank plasma (in 4

mL polypropylene tubes) with 20 μL of 50, 100, 200, 400, 800, 1600 ng/mL cinnarizine in mobile

phase solution, which provided spiked plasma concentrations in the range of 10‐320 ng/mL

cinnarizine. 20 µL of an internal standard solution (500 ng/mL flunarizine in mobile phase) was

also added, and tubes briefly vortexed. 20 µL of trichloroacetic acid was added to precipitate

plasma protein, and the tubes vortexed for 1 min. 2 mL of tert-butyl methyl ether (TBME) was

added to extract cinnarizine and flunarizine into the organic phase. The tubes were vortexed for 1

min, allowed to equilibrate for 10 min, and vortexed for a further minute before being centrifuged

at 10,000 xg for 15 min. 1.6 mL aliquots of the supernatant were then transferred into new

polypropylene tubes, and TBME evaporated using a nitrogen evaporator at 40 ºC. Cinnarizine and

flunarizine in the tubes were reconstituted with 150 µL mobile phase and assayed via HPLC.

The HPLC assay conditions for the quantification of cinnarizine in rat plasma were similar to that

for the quantification of cinnarizine in intestinal lipid colloidal phases described in Section 2.4.1,

with slight modifications. The mobile phase used in the plasma assay consisted 45% v/v

acetonitrile : 55% v/v 20 mM NH4H2PO4, i.e. the organic solvent content was lowered to enable

better separation of cinnarizine and flunarizine peaks. The injection volume was 100 µL, and the

retention time for cinnarizine and flunarizine were 8.9 min and 11.3 min, respectively. Unknown

concentrations were determined by comparing the unknown cinnarizine:flunarizine area under the

curve (AUC) ratio against a standard curve of cinnarizine:flunarizine AUC ratio vs. cinnarizine

concentration. The assay was validated on three separate days by analysing drug standards

(prepared at low, medium and high concentrations) in quadruplicates and comparing against a

linear standard curve. The assay was found to be accurate (to within ± 10 % of target


90

concentration except for the lower limit of quantitation which was ± 15%) and precise (% CV <

10 for all concentrations except the lower limit of quantification which was < 15) for cinnarizine

plasma concentrations between 10-320 ng/mL. The validation results are shown in Table 2.5.

Table 2.5: Precision (% coefficient of variation) and accuracy (% of the target value) of the

plasma cinnarizine HPLC assay.

[Cinnarizine]

(ng/mL)

Day 1 Day 2 Day 3


10 11.5 91.0 6.6 93.2 1.3 92.2

80 2.2 91.4 6.3 110.0 2.8 98.8

320 9.0 100.9 2.5 106.2 2.0 99.3

2.4.3 LC-MS assay for quantification of danazol in rat plasma

Danazol plasma samples were prepared for LC-MS analysis using a validated precipitation

procedure. Standards for the assay were prepared by spiking 125 μL aliquots of blank plasma with

5 μL of 125, 250, 1250, 2500 and 6250 ng/mL danazol in acetonitrile, which provided spiked

plasma concentrations in the range of 5‐250 ng/mL danazol. 5 µL of an internal standard solution

(2000 ng/mL progesterone in acetonitrile) was also added to each sample, and tubes vortexed for

30 sec. Following that, 62.5 μL of saturated ammonium sulphate solution was added, tubes

vortexed for 30 sec; 125 µL ACN was added, and the tubes vortexed again for 30 sec. The tubes

were left to stand for 20 min at room temperature prior to centrifugation at 35,000 xg for 5 min.

100 μL aliquots of the organic phase (top layer of supernatant) were then transferred into

autosampler vials and 10 μL injected onto the LC‐MS. Unknown concentrations were determined


91

by comparing the unknown danazol:progesterone peak height ratio against a standard curve of

danazol:progesterone peak height ratio vs. danazol concentration.

The LC-MS system (LCMS 2020, Shimadzu, Japan) consisted a LC‐20AD binary pump, a SiL‐

20AC Ht refrigerated autosampler, a mobile phase vacuum degassing unit (DGU‐20A3), and a

temperature‐controlled column compartment (CTO‐20A) coupled with a single‐quadrupole mass

spectrometric (MS) detector (Shimadzu LCMS 2020) equipped with an electrospray ionisation

source. The autosampler was maintained at 4 °C and the column at 40 °C. The stationary phase

was a Phenomenex Gemini® C6 phenyl column (50 × 2.0 mm, 3 µm). Samples were eluted using

a gradient at a total flow rate of 0.3 mL/min. The mobile phases consisted a mixture of solvent A

(95% v/v water : 5% v/v methanol ) and solvent B (5% v/v water : 95% v/v methanol) both

containing 1 mM ammonium formate and 0.1% v/v formic acid. The initial percentage of solvent

B was 60% and was linearly increased to 100% over 6.5 min and was held at 100% for 7 min.

After 14.25 min, the gradient was returned to 60% solvent B which was held until the end of the

run (17 min). The MS conditions were as follows: drying gas flow 20 L/min; nebulising gas flow

1.5 L/min; drying gas temperature 200 °C; interface voltage 3.5 kV; detector voltage 1.0 kV.

Selected‐ion monitoring was accomplished at m/z +338.2 for danazol [M+H]+ and m/z +314.9 for

progesterone [M+H]+. The chromatographic data were acquired and analysed using LabSolutions

software package (Shimadzu).

The assay was validated on three separate days by analysing drug standards (prepared at low,

medium and high concentrations) in quadruplicates and comparing against a linear standard curve.

The assay was found to be accurate (to within ± 10 % of target concentration except for the lower

limit of quantitation which was ± 15%) and precise (% CV < 10 for all concentrations except the


92

lower limit of quantification which was < 15) for cinnarizine plasma concentrations between 5-

250 ng/mL. The validation results are shown in Table 2.6.

Table 2.6: Precision (% coefficient of variation) and accuracy (% of the target value) of the

plasma danazol LC-MS assay.

[Danazol]

(ng/mL)

Day 1 Day 2 Day 3


5 8.2 96.0 7.8 103.5 5.3 87.6

50 4.5 103.2 1.1 108.5 2.0 104.4

250 6.4 97.6 4.4 106.7 2.0 93.0

2.4.4 Enzymatic colorimetric assay for quantification of total bile salt in

whole rat bile

Total bile salt concentration in whole rat bile was analysed using an enzymatic colorimetric assay

(Total Bile Acids kit # 431-15001; Wako Pure Chemical Industries, Osaka, Japan). Standards for

the assay were freshly prepared 50, 100, 200, 400, 500 µM sodium glycocholate in water. Rat bile

samples were diluted 1:100 with water. Two 20 µL aliquots of sample were added to two wells of

a 96-well plate. 50 µL of ‘enzyme reagent’ and 50 µL of ‘enzyme reagent for blank test’ were

added to the first and second sample, respectively. The plate was then incubated at 37 ⁰C for 10

min. At the end of the incubation period, 50 µL of ‘stopper solution’ was added to both samples.

Sample absorbance was then measured at a wavelength of 540 nm on plate reader (Fluostar

Optima plate reader, BMG Labtechnologies, Germany). The difference between the specific

absorption of the sample treated with ‘enzyme reagent’ and the sample treated with ‘enzyme

reagent for blank test’ was used in the determination of total bile salt concentration. The assay was


93

validated on three separate days by analysing three standards (prepared at low, medium and high

concentrations) in quadruplicates and comparing against the linear standard curve. The assay was

found to be accurate (to within ± 10 % of target concentration except for the lower limit of

quantitation which was ± 15%) and precise (% CV < 10 for all concentrations except the lower

limit of quantification which was < 15) for total bile salt concentrations between 50-500 µM. The

validation results are shown in Table 2.7.

Table 2.7: Precision (% coefficient of variation) and accuracy (% of the target value) of the total

bile salt enzymatic colorimetric assay.

[Glycocholate] (µM)

Day 1 Day 2 Day 3


50 13.2 85.9 14.6 93.1 11.6 87.3

200 4.5 100.6 1.2 102.8 4.8 107.3

500 4.0 101.9 2.8 101.8 0.7 103.6

94

Monash University

Declaration for Thesis Chapter 3 Declaration by candidate In the case of Chapter 3, the nature and extent of my contribution to the work was the following:

Nature of contribution Extent of contribution (%)

Concept and design of studies, planning and execution of experimental work, data analysis and interpretation, formulation of conclusions and hypotheses resulting from the relevant studies, drafting and revision of manuscript

60%

The following co-authors contributed to the work. Co-authors who are students at Monash University must also indicate the extent of their contribution in percentage terms:

Name Nature of contribution Extent of contribution (%)

C.J.H. Porter Project supervisor, data and manuscript review NA

N.L. Trevaskis Project co-supervisor, data and manuscript review NA

T. Quach Synthesis of sulfo-N-succinimidyl oleate NA

W.N. Charman Manuscript review NA

P. Tso Manuscript review NA

Candidate’s Signature

Declaration by co-authors The undersigned hereby certify that:

(1) the above declaration correctly reflects the nature and extent of the candidate’s contribution to this work, and the nature of the contribution of each of the co-authors.

(2) they meet the criteria for authorship in that they have participated in the conception, execution, or interpretation, of at least that part of the publication in their field of expertise;

(3) they take public responsibility for their part of the publication, except for the responsible author who accepts overall responsibility for the publication;

(4) there are no other authors of the publication according to these criteria; (5) potential conflicts of interest have been disclosed to (a) granting bodies, (b) the editor or

publisher of journals or other publications, and (c) the head of the responsible academic unit; and

(6) the original data are stored at the following location(s) and will be held for at least five years from the date indicated below:

Location(s) Monash Institute of Pharmaceutical Sciences

Department of Pathology and Laboratory Medicine, University of Cincinnati

95

Signature 1

Signature 2

Signature 3

Signature 4

Signature 5

96

CHAPTER 3 : INTESTINAL BILE

SECRETION PROMOTES DRUG

ABSORPTION FROM LIPID COLLOIDAL

PHASES VIA INDUCTION OF

SUPERSATURATION

Yan Yan Yeap1, Natalie L. Trevaskis1, Tim Quach2, Patrick Tso3, William N. Charman1,

Christopher J. H. Porter1

1 Drug Delivery, Disposition and Dynamics, Monash Institute of Pharmaceutical Sciences, Monash University, 381 Royal Parade, Parkville, Victoria, 3052, Australia

2 Medicinal Chemistry and Drug Action, Monash Institute of Pharmaceutical Sciences, Monash University, 381 Royal Parade, Parkville, Victoria, 3052, Australia

3 Department of Pathology and Laboratory Medicine, University of Cincinnati, Cincinnati, OH 45221, USA

Molecular Pharmaceutics (2013); In press.

Chapter 3: Bile-induced supersaturation

97

3.1 ABSTRACT

The oral bioavailability of poorly water-soluble drugs (PWSD) is often significantly enhanced by

co-administration with lipids in food or lipid-based oral formulations. Co-administration with

lipids promotes drug solubilisation in intestinal mixed micelles and vesicles, however, the

mechanism(s) by which PWSD are absorbed from these dispersed phases remain poorly

understood. Classically, drug absorption is believed to be a product of the drug concentration in

free solution and the apparent permeability across the absorptive membrane. Solubilisation in

colloidal phases such as mixed micelles increases dissolution rate and total solubilised drug

concentrations, but does not directly enhance (and may reduce) the free drug concentration. In the

absence of changes to cellular permeability (which is often high for lipophilic, PWSD), significant

changes to membrane flux are therefore unexpected. Realising that increases in effective

dissolution rate may be a significant driver of increases in drug absorption for PWSD, we explore

here two alternate mechanisms by which membrane flux might also be enhanced: (1) collisional

drug absorption where drug is directly transferred from lipid colloidal phases to the absorptive

membrane, and (2) supersaturation-enhanced drug absorption where bile mediated dilution of lipid

colloidal phases leads to a transient increase in supersaturation, thermodynamic activity and

absorption. In the current study, collisional uptake mechanisms did not play a significant role in

the absorption of a model PWSD, cinnarizine, from lipid colloidal phases. In contrast, bile-

mediated dilution of model intestinal mixed micelles and vesicles led to drug supersaturation. For

colloids that were principally micellar, supersaturation was maintained for a period sufficient to

promote absorption. In contrast, for primarily vesicular systems, supersaturation resulted in rapid

drug precipitation and no increase in drug absorption. This work suggests that on-going dilution

by bile in the gastrointestinal tract may invoke supersaturation in intestinal colloids and promote

absorption, and thus presents a new mechanism by which lipids may enhance the oral absorption

of PWSD.


98

3.2 INTRODUCTION

The potential for lipid-based formulations (LBF) to enhance the oral bioavailability of poorly

water-soluble drugs (PWSD) has been recognised for over 40 years2. Lipid co-administration is

thought to enhance the oral absorption of PWSD by providing mechanisms to overcome both slow

dissolution and low solubility of PWSD in the aqueous gastrointestinal (GI) milieu. Firstly, LBF

present drug to the GI tract in a molecularly dispersed form (i.e. in solution in the formulation),

thereby circumventing the need for dissolution from the solid to the liquid state. Subsequently, the

intercalation of formulation lipids into endogenous lipid digestion pathways results in the

generation of intestinal lipid colloidal phases (such as vesicular and micellar species) that enhance

the solubilisation capacity of the small intestine, promote drug solubilisation and reduce the risk of

drug precipitation.

Solubilisation within lipid colloidal phases therefore increases the apparent solubility of PWSD in

the intestinal fluids and promotes dissolution. However, in the absence of solid drug, solubilisation

also results in a reduction in thermodynamic activity143. In simple micellar systems this reduction

in thermodynamic activity is manifest in a decrease in the free concentration of drug. Where

solubilised drug exists in equilibrium between the free concentration (Cfree) and the concentration

solubilised in intestinal colloids (Ccolloid), the total solubilised drug concentration (Ctotal) is the sum

of Cfree and Ccolloid,


Under these circumstances, the solubility of drug in the inter-micellar phase (effectively the

aqueous solubility of the drug) provides the upper limit for Cfree and the presence of solubilising

colloidal species typically increase the total drug concentration but do not increase, and often


99

reduce, Cfree. Classical models of passive drug absorption suggest that drug flux across an

absorptive membrane is the product of the free drug concentration and the drug permeability

across the membrane. Therefore, where solubilisation reduces Cfree (but does not alter

permeability), absorption is expected to be reduced. Indeed, in solubilised systems, even at

saturation, Cfree does not exceed the equilibrium solubility of drug in (non-micellar) aqueous

solution. Solubilisation in intestinal colloidal phases therefore provides no practical advantage in

free drug concentration, and in the absence of changes to permeability, is unlikely to lead to

appreciable increases in membrane flux when compared to an aqueous solution containing drug at

close to saturated solubility. In support of this suggestion, many authors have shown that

increasing the total concentration of a PWSD by solubilisation does not necessarily result in

proportional increases in absorptive flux100, 143-146. Dahan et al. and Miller et al. recently proposed

a model to quantify this phenomena and referred to the existence of a ‘solubility-permeability

interplay’ where potential increases in membrane flux due to increases in solubilised drug

concentration were off-set by a reduction in the apparent permeability, the latter being, in large

part, a function of decreases in free fraction100, 146.

The dispersion and digestion of LBF therefore creates a solubilised reservoir that is in rapid

equilibrium with drug in free solution and provides significant advantage in the effective rate of

dissolution of a PWSD when compared to traditional dissolution from a solid dose form. In

contrast, the inherent solubility limitations to flux (rather than dissolution-rate limitations) are

seemingly unchanged when viewed from the perspective of the free concentration or may be made

worse. This appears at odds with the wealth of experimental and practical observations that

suggest the ability of lipids (either formulation- or dietary-derived) to enhance the oral absorption

of a range of PWSD2. A possible explanation for this anomaly is that the traditional view of drug

absorption from colloidal dispersions may not adequately describe the dynamic manner in which


100

LBF exert absorption-enhancing effects in the GI tract. Two alternative potential mechanisms of

drug absorption are therefore examined here (Figure 3.1).

Figure 3.1: Alternative mechanisms of drug absorption from intestinal lipid colloidal phases. In

collisional drug absorption (left panel), in addition to the diffusion of free drug molecules (a), lipid

colloidal phases collide with the absorptive membrane and facilitate direct transfer of solubilised

drug into absorptive cells (b). Collisional transfer may or may not be receptor-mediated. In

supersaturation-enhanced drug absorption (right panel), the interaction between secreted bile and

lipid colloidal phases leads to drug supersaturation (possibly via stimulation of phase changes to

less lipid-rich colloids with lowered drug solubilisation capacity). The increase in thermodynamic

activity manifests in increases in free drug concentration, and enhanced diffusional flux (a) across

the absorptive membrane.

The first mechanism evaluated was the potential for drug absorption to occur via direct collisional

transfer from lipid colloidal phases to the absorptive membrane, and thus to be mediated not only

Vesicles Mixed micelles Receptors

D Drug Dss Supersaturated drug

Bile components

DssD (a)

(a)Bile

secretion

Dss

DssD Dss

DDss

Diffusion Collision(a) (b)

Collisional drug absorption Supersaturation-enhanced drug absorption

D

D

(b)

(a)

(a)

(b)

D

DD

D

DD

D

DD

D

DD

DD

DD

(b)

(b)D

DD

D

DDDD

DssD


101

by Cfree but also by the solubilised fraction, Ccolloid. Previously, studies by Storch and colleagues

have shown that the transfer of poorly water-soluble fatty acids between model cell membranes

and proteins may occur via collisional transfer150, 151. More recently, the possibility of drug

absorption via collisional uptake has been suggested by Yano et al. and Gao et al.190, 191.

Collisional uptake may or may not be receptor-mediated192, however, lipid uptake receptors such

as CD3661, SR-BI67 and NPC1L168 have been suggested to facilitate the absorption of cholesterol

and fatty acids; poorly water-soluble molecules that are also solubilised in intestinal colloidal

phases. In the case of SR-BI and CD36, direct interaction of the receptor with colloidal structures

such as HDL (high-density lipoprotein), bile salt micelles and phospholipid vesicles66, 67, has also

been suggested, raising the possibility that lipid uptake receptors may interact directly with

intestinal lipid colloidal phases to facilitate collisional absorption of solubilised contents,

including PWSD.

The second mechanism evaluated was the potential for endogenous lipid processing pathways to

lead to drug supersaturation in lipid colloidal phases. Supersaturation increases the

thermodynamic activity of solubilised drug, and in the solubilisation model described by Equation

3.1, will increase Cfree above the equilibrium aqueous drug solubility. This in turn is expected to

enhance drug flux. The potential for supersaturation to enhance the oral bioavailability of PWSD

has received increasing recent interest173, 191. For LBF, supersaturation may be generated by the

loss of drug solubilisation capacity resulting from the digestion of triglycerides91, 176, 193 and/or

surfactants177, and the dilution of co-solvents5 during GI processing176. In contrast, the possibility

that supersaturation may result directly from interactions between lipid colloidal phases and

biliary fluids has been almost entirely ignored. Although traditional micellar solubilisation models

suggest that increases in bile salt concentrations increase drug solubilisation, previous studies have

also shown that dilution of lipid colloidal phases with model intestinal fluids (containing bile salts,


102

phospholipid and cholesterol) may lead to the generation of less lipid-rich colloidal phases with

lowered solubilisation capacities194. This provides a plausible mechanism for supersaturation

generation in the small intestine and has been examined in detail here.

The data suggest that under the conditions explored, collisional drug absorption has little impact

on drug absorption from intestinal colloidal species. In contrast, supersaturation-enhanced drug

absorption, mediated by the interaction between secreted bile and lipid colloidal phases, may

provide an endogenous mechanism to promote supersaturation and to facilitate drug absorption

from lipid colloidal phases.

3.3 METHODS

3.3.1 Materials

Cinnarizine, flunarizine dihydrochloride, monensin sodium, sodium taurocholate, sodium

taurodeoxycholate, sodium glycocholate, sodium glycochenodeoxycholate, cholesterol, L-α-

lysophosphatidylcholine (LPC, from egg yolk), L-α-phosphatidylcholine (PC, from dried egg

yolk), oleic acid, caprylic acid, monocaprylin, N-hydroxysulfosuccinimide sodium,

dicyclohexylcarbodiimide solution (60%w/v in xylenes), N,N-dimethylformamide, TWEEN® 80,

potassium dihydrogen phosphate (KH2PO4) and sodium chloride (NaCl) were obtained from

Sigma-Aldrich, Australia. Sodium taurochenodeoxycholate, sodium glycodeoxycholate, ortho-

phosphoric acid 85% (H3PO4), sodium hydroxide pellets (NaOH), tert-butyl methyl ether (TBME),

dimethyl sulfoxide (DMSO), glacial acetic acid and absolute ethanol were from Merck, Australia.

Disodium hydrogen orthophosphate (Na2HPO4), sodium dihydrogen orthophosphate

(NaH2PO4.2H2O) and ammonium dihydrogen orthophosphate (NH4H2PO4) (Ajax Finechem,


103

Australia), cholesterol, [4-14C]- (49.8 mCi/mmol) and Irga-Safe PlusTM (Perkin Elmer Life

Sciences, MA, USA), oleic acid, [9,10-3H(N)] (60 Ci/mmol) (American Radiolabelled Chemicals,

MO, USA), Block Lipid Transport-1 (BLT-1, i.e. 2-hexyl-1-cyclopentanone thiosemicarbazone)

(Chembridge, CA, USA), ezetimibe (Jai Radhe Sales, AMD, India), heparin sodium injection BP

(1000 I.U./mL, Hospira, Australia), xylazine (100 mg/mL, Troy Laboratories, Australia),

acepromazine (10 mg/mL, Ceva Delvet, Australia), ketamine (100 mg/mL, Provet, Australia) and

pentobarbitone sodium (325 mg/mL, Virbac, Australia) were obtained from listed suppliers.

Acetonitrile, methanol, and chloroform used were analytical reagent grade. Water was obtained

from a Millipore milliQ Gradient A10 water purification system (Millipore, MA, USA).

3.3.1.1 Sulfo-N-succinimidyl oleate (SSO) synthesis

SSO synthesis was adapted from the procedure of Harmon et al.195. Briefly,

dicyclohexylcarbodiimide (1.26 mmol) and N-hydroxysulfosuccinimide (sodium salt, 1.20

mmol) were added to a solution of oleic acid (1.20 mmol) dissolved in N,N-dimethylformamide (2

mL) and the reaction stirred at room temperature overnight. The precipitated dicyclohexylurea

was removed by filtration and ethyl acetate (2 mL) added to the filtrate, which was left to stand at

4° C overnight. SSO (precipitate) was then collected by filtration and dried under vacuum (1

mmHg). SSO identity was confirmed by NMR and mass spectrometry.

3.3.2 Experimental outline

To assess the role of receptor-mediated collisional drug absorption, cinnarizine bioavailability

was assessed after administration of a lipid emulsion formulation in the absence and presence of

BLT-1, SSO, and ezetimibe. BLT-1, SSO and ezetimibe are chemical inhibitors of SR-BI196,

CD36197, and NPC1L1198, respectively. The possibility of endocytosis-mediated uptake was also

investigated by the use of a general endocytosis inhibitor, monensin.


104

The role of collisional drug absorption was assessed more generically using an in situ rat jejunum

perfusion model to compare the absorptive flux of cinnarizine from two distinctly different lipid

colloidal phases (micelles vs. vesicles) with matched drug solubilisation capacities. Colloidal

systems with the same total solubilisation capacity, loaded with drug at the same concentration

have the same thermodynamic activity and therefore Cfree is the same in both cases. Under these

circumstances, comparison of the flux profiles obtained from two structurally different colloids,

but with identical Cfree, provides a means of determining whether the nature of the colloid, or Cfree,

is the principal determinant of absorption. Generation of identical flux profiles from both systems

would therefore confirm the dependence of flux on thermodynamic activity and free concentration,

whereas a significant difference in flux would indicate a role for factors beyond Cfree in

determining flux. These include the potential for collisional drug absorption mechanisms since

collision rates are a statistical function of particle number and are expected to be markedly higher

for micelles (where the smaller particle size results in higher particle numbers) when compared to

vesicles.

To assess the potential for intestinal fluids to enhance drug absorption from lipid colloidal phases

via the induction of drug supersaturation, whole bile was collected from fasted rats, and mixed

with model micelles and vesicles to simulate the process of interaction with bile in vivo. The

potential for bile to generate drug supersaturation was evaluated in vitro by assessing changes in

cinnarizine solubility, and by monitoring the kinetics of cinnarizine solubilisation and

precipitation, following bile addition to cinnarizine-loaded micelles and vesicles. Subsequently,

the impact of drug supersaturation on the intestinal absorptive flux of cinnarizine from micelles

and vesicles was assessed in an in situ rat jejunum perfusion model, with and without co-perfusion

of donor bile. Finally, the relevance of bile-induced drug supersaturation in vivo was assessed via


105

examination of changes to cinnarizine bioavailability after administration of drug-loaded micelles

and vesicles (with matched thermodynamic activity) in bile-intact vs. bile-diverted rats.

3.3.3 Formulation preparation

3.3.3.1 Lipid emulsion

The lipid emulsion (3 mL per dose) consisted of 1 mg cinnarizine and 49 mg oleic acid solubilised

in 8 mM sodium taurocholate, 2 mM phosphatidylcholine, 2 mM cholesterol and trace amounts of

14C-cholesterol (1 µCi/3 mL) and/or 3H-oleic acid (3 µCi/3 mL). The emulsion was prepared in

7.5 mL batches by weighing appropriate masses of cinnarizine in oleic acid stock solution (20

mg/g), phosphatidylcholine and cholesterol into a glass vial, and the mixture made up to volume

with a buffered sodium taurocholate solution (buffer consisted 18 mM Na2H2PO4.2H2O and 12

mM Na2HPO4). Appropriate volumes of 14C-cholesterol, 3H-oleic acid, 5 mg/mL BLT-1 in

ethanol (for SR-BI inhibition experiments only), 25 mg/mL SSO in DMSO (for CD36 inhibition

experiments only), and 10 mg/mL monensin in ethanol (for endocytosis inhibition experiments

only) were spiked into the vial, and vortexed for 1 min. The formulation was emulsified by

ultrasonification with a Misonix XL 2020 ultrasonic processor (Misonix, Farmingdale, NY, USA)

equipped with a 3.2-mm microprobe tip running at an amplitude of 240 µm and a frequency of 20

kHz for 1.5 min. The total solvent concentration in the emulsion was ≤ 2.5% v/v. The emulsion

was used within 4 h of preparation, and the concentration of drug and labelled cholesterol and/or

oleic acid assayed before dosing (in duplicate) to confirm compound content in the emulsion and

to allow for dose normalisation between rats.


106

3.3.3.2 Intravenous formulation

The formulation (1 mL per dose) used for intravenous administration of cinnarizine and 14C-

cholesterol comprised 0.5% w/v Tween® 80 in buffer (36 mM Na2HPO4 and 22 mM KH2PO4,

adjusted to pH 4 with acetic acid). Cinnarizine and cholesterol were added to the formulation by

spiking 5% v/v DMSO (containing 10 mg/mL cinnarizine) and 5% v/v ethanol (containing 5

mg/mL cholesterol and 40 µCi/mL 14C-cholesterol) into the micellar solution. The formulation

was mixed by vortexing, and the concentration of drug and labelled cholesterol assayed before

dosing to confirm compound content in the formulation and to allow for dose normalisation

between rats. The formulation was used within 1 h of preparation.

3.3.3.3 Model micelles and vesicles

The preparation of model micelles and vesicles was guided by the methods and phase diagram

published by Kossena et al.39. Medium-chain lipid containing colloids were chosen over long-

chain systems since the former have previously been shown to generate monophasic micellar and

vesicular systems39. Relatively high lipid concentrations were chosen to reflect the species that are

expected to initially form during the digestion of medium-chain triglycerides91. The model

colloids consisted of tricaprylin digestion products (caprylic acid and monocaprylin) solubilised in

simulated endogenous intestinal fluid (SEIF). SEIF comprised the six most prevalent bile salts in

human bile199, lysophosphatidylcholine (LPC), and cholesterol. The total bile salt:LPC:cholesterol

molar ratio was maintained at 16:4:1, reflecting known ratios within fasted human bile200, 201. The

combination of bile salts used here comprised 25 mol% sodium glycocholate, 17.5 mol% sodium

glycodeoxycholate, 25 mol% sodium glycochenodeoxycholate, 12.5 mol% sodium taurocholate,

7.5 mol% sodium taurodeoxycholate and 12.5 mol% sodium taurochenodeoxycholate. The

concentration ratios of the bile salts were chosen based on average concentrations of the six most

prevalent bile salts found in human bile199. The caprylic acid:monocaprylin molar ratio was kept at


107

2:1, reflecting the ratio of digestion products expected on digestion of 1 mole of triglyceride. The

concentration of micellar and vesicular components was varied by trial and error (but maintaining

the ratios described above) to identify systems with similar drug solubilisation capacities. It has

previously been shown that the thermodynamic activity (i.e. free concentration) of drug in a

solubilised system may be estimated via assessment of solubility behaviour, such that different

colloidal solutions containing drug at a fixed proportion of the saturated solubility, results in

matched free concentrations202. Thus, drug was loaded into either micellar or vesicular systems at

the same concentration (and the same proportion of saturated solubility) and was therefore present

at the same thermodynamic activity (i.e. Ctotal, Ccolloid and Cfree were the same in both micellar and

vesicular systems). The compositions of the identified micellar and vesicular systems are shown in

Table 3.1.

Table 3.1: Composition of model micelles and vesicles.^

Total bile salt# (mM)

LPC$ (mM)

Cholesterol (mM)

Caprylic acid (mM)

Monocaprylin (mM)

Micelles 8 2 0.5 69.3 34.7

Vesicles 2 0.5 0.125 52.0 26.0

^ Micelles and vesicles also consist of 18 mM NaH2PO4.2H2O and 12 mM Na2HPO4. Sodium strength was adjusted to 150 mM with NaCl. Final pH of phases was adjusted to 6.30 ± 0.01

# Total bile salt consist of 25 mol% sodium glycocholate, 17.5 mol% sodium glycodeoxycholate, 25 mol% sodium glycochenodeoxycholate, 12.5 mol% sodium taurocholate, 7.5 mol% sodium taurodeoxycholate, 12.5 mol% sodium taurochenodeoxycholate

$ LPC is lysophosphatidylcholine

SEIF (8 mM total bile salt:2 mM LPC:0.5 mM cholesterol) was prepared in 50 mL batches.

Briefly, LPC and cholesterol were dissolved in 1 mL chloroform in a round bottom flask, followed


108

by solvent evaporation under vacuum. The thin film formed by solvent evaporation was

reconstituted with buffered bile salts solution (8 mM total bile salt, 18 mM NaH2PO4.2H2O and 12

mM Na2HPO4, 100 mM NaCl), vortexed for 1 min, and allowed to equilibrate at room

temperature overnight. When vesicles were prepared, a similar procedure was adopted, but in this

case SEIF was diluted 4-fold with buffer (18 mM NaH2PO4.2H2O and 12 mM Na2HPO4, 108 mM

NaCl) to reduce the bile salt:lipid concentration ratio. Micelles and vesicles were prepared in 10

mL batches by adding caprylic acid and monocaprylin (quantities in Table 3.1) to SEIF, followed

by pH adjustment to 6.30 with solid NaOH and vortexing for 1 min. The phases were then

ultrasonicated as described earlier (30 sec continuous ultrasonication followed by pulsatile, 1 sec-

on/1 sec-off ultrasonication for 5 min). When included in the colloids, cinnarizine was pre-

dissolved in caprylic acid, and the drug/fatty acid solution allowed to equilibrate overnight prior to

micelle/vesicle preparation.

3.3.4 Particle sizing

The particle size of the model micelles and vesicles was determined by photon correlation

spectroscopy (Malvern Instruments Nano-ZS Zetasizer, Malvern, UK) at 37 °C. Light scattering

cells were cleaned by rinsing with ultrapure laboratory grade water (Milli-Q, Millipore, MA,

USA), and allowed to dry while inverted on lint-free wipes (Kimwipes®, Kimberly-Clark,

Australia). Samples were transferred into the cells using sterile 1 mL syringes, and analysed

without dilution to prevent structural changes to the phases. CONTIN analysis method was used to

analyse the autocorrelation functions and to provide a calculated polydispersity index. Micelles

had a mean particle size of 9 ± 1 nm [polydispersity index 0.128 ± 0.040], and vesicles had a mean

particle size of 443 ± 32 nm [polydispersity index 0.456 ± 0.038]. Data reported are mean ± SEM

of n = 3 determinations.


109

3.3.5 Equilibrium solubility of cinnarizine in the model micelles and vesicles

Excess solid cinnarizine was added to 2 mL micelles or vesicles in glass vials. Vials were briefly

vortexed, incubated at 37 °C, and samples taken every 24 h over a period of 120 h. During

sampling, vials were centrifuged (2,200 xg, 10 min, 37 °C), 50 µL of supernatant sampled, and

vials re-vortexed. Equilibrium solubility was defined when drug concentrations in consecutive

samples varied by ≤ 5%, and was determined on three separate occasions. Equilibrium solubility

of cinnarizine was also determined after 1:1 addition of bile/bile pH 6.30/buffer pH 6.30 to the

different colloidal phases. The pH of fresh bile was adjusted to 6.30 with H3PO4. Buffer pH 6.30

consisted of 18 mM NaH2PO4.2H2O, 12 mM Na2HPO4, and 108 mM NaCl.

3.3.6 Kinetics of cinnarizine precipitation

The kinetics of cinnarizine precipitation was monitored after addition of bile to the micellar and

vesicular phases, to determine whether a period of drug supersaturation existed prior to drug

precipitation. In a temperature and stirring rate-controlled vessel, 2.5 mL of bile was added to 2.5

mL micelles or vesicles containing 0.2 mg/mL cinnarizine (~ 80% saturated solubility). Samples

(100 µL) were taken before the addition of bile, and at 1, 10, 20, 30, 40, 50, 60, 80, 100, 120 min

after bile addition. Samples were immediately centrifuged (2,200 xg, 5 min, 37 °C) to separate

precipitated drug, and 50 µL of supernatant assayed for drug content. The proportion of the initial

solubilised cinnarizine concentration, that remained solubilised after bile addition, was assessed as

the percent of the drug mass remaining in solution (i.e. concentration in the supernatant multiplied

by the volume remaining in vessel) relative to the total drug mass in the vessel at each time point.

3.3.7 Solid-state analysis of the cinnarizine precipitate

Selected cinnarizine pellets from the precipitation kinetics experiments were analysed using a

Zeiss Axiolab microscope (Carl Zeiss, Oberkochen, Germany) equipped with crossed polarising


110

filters. At the end of the precipitation kinetics experiments, 1.5 mL of remaining bile + colloid

mixture were centrifuged (2,200 xg, 10 min, 37 °C), the supernatant was discarded, and a small

amount of pellet was carefully placed on a microscope slide. Samples were analysed under cross-

polarised light, and images were recorded using a Canon PowerShot A70 digital camera (Canon,

Tokyo, Japan).

3.3.8 Animals

Animal care and anaesthesia have been described in Section 2.2.2.2. All animals were fasted

overnight (12–18 h) prior to surgery. At the end of all experiments, rats were euthanized via an

intravenous or intracardiac injection of 100 mg sodium pentobarbitone.

3.3.9 Surgical procedures

3.3.9.1 Cinnarizine bioavailability studies following intraduodenal

administration

The surgical procedures for the conduct of bioavailability studies included cannulations of the

right carotid artery, right jugular vein, duodenum (1 cm below pylorus), and common bile duct

(only for bile-diverted rats). The surgical procedures for the cannulations have been described in

Section 2.3.2.

3.3.9.2 Fasted rat bile collection

The bile duct was cannulated near the hilum of the liver (where the duct is free of pancreatic

tissue) in order to facilitate the collection of bile fluid without contamination by exocrine

pancreatic secretions203. Rats were rehydrated via saline infusion (1.5 mL/h) into a cannula

inserted into the right jugular vein, and bile continuously collected for 5 h. The concentration of


111

total bile salt in collected bile was assayed using a validated enzymatic colorimetric assay (Total

Bile Acids kit #431-15001; Wako Pure Chemical Industries, Osaka, Japan) on a plate reader

(Fluostar Optima plate reader, BMG Labtechnologies, Germany) measuring absorbance at a

wavelength of 540 nm. In all subsequent experiments, bile was used within 24 h of collection.

3.3.9.3 Single-pass rat jejunum perfusion

The model employed to assess flux across rat jejunum involved in situ perfusion (single-pass) of

an isolated jejunal segment and simultaneous blood collection from the corresponding mesenteric

vein branch. The surgical procedures for the setup of the single-pass rat jejunum perfusion model

have been described in Section 2.2.2.3.

3.3.10 Cinnarizine bioavailability studies

A 30-min equilibration period was allowed between the end of surgery and drug dosing. To

examine the impact of lipid uptake receptors on drug absorption, studies were conducted in the

presence or absence of lipid uptake inhibitors. To examine the impact of bile-induced drug

supersaturation on drug absorption, studies were conducted in bile-intact or bile-diverted rats.

Colloidal systems (lipid emulsion, micelles, vesicles) containing cinnarizine were infused into the

duodenum of rats at a rate of 1.5 mL/h for 2 h. When the dose infusion was complete, saline was

infused at a rate of 1.5 mL/h for 10 min to flush any remaining formulation in the tubing into the

duodenum. Blood samples (0.3 mL) were collected via the carotid artery cannula up to 8 h after

infusion initiation into tubes containing 3 I.U. heparin. The sampling intervals were: t = 0, 1, 2, 2.5,

3, 4, 6, 8 h for the receptor inhibition studies; and t = 0, 1, 1.5, 2, 3, 4, 6, 8 h for the bile-induced

drug supersaturation studies. After each blood sample was taken, the cannula was flushed with 0.3

mL of 2 I.U./mL heparinised saline to ensure cannula patency, and to replace the volume of blood


112

removed. Plasma was separated by centrifugation (10,000 xg, 5 min) to enable analysis of drug

and labelled lipid content.

3.3.10.1 Administration of lipid uptake inhibitors (BLT-1, SSO, ezetimibe) and

endocytosis inhibitor (monensin)

Ezetimibe has previously been dosed at 0.3 mg/kg intravenously into rats, and has been shown to

inhibit cholesterol absorption (from an intraduodenally dosed lipid emulsion) without reports of

toxicity204. Therefore, in our study, intravenous administration of ezetimibe (0.3 mg/kg via the

jugular vein) was selected as the route to administer the inhibitor at the beginning of the 30-min

equilibration period. An appropriate volume of 5 mg/mL ezetimibe in ethanol was spiked into

blank rat plasma, and 0.8 mL of resultant plasma dosed into rats as an intravenous bolus. The total

ethanol concentration was less than 2.5% v/v. In the case of BLT-1 and SSO, the inhibitors had

not been previously administered intravenously, therefore local (i.e. intestinal) administration was

selected to limit the systemic effects of the inhibitors. Monensin was also administered directly

into the intestine to minimise systemic endocytosis inhibition. Here, 100 µM BLT-1, 1 mM SSO

and 100 µM monensin were pre-infused intraduodenally (in saline) at a rate of 1.5 mL/h during

the 30-min equilibration period, and subsequently co-infused at the same concentration as part of

the lipid emulsion. BLT-1 and SSO have previously been shown to inhibit lipid uptake in cell-

based studies at concentrations of 1-10 µM196 and 400 µM195, respectively. Monensin has

previously been shown to inhibit endocytosis in cultured cells at 10 µM205.

Intravenous administration studies were also conducted in control rats and monensin-treated rats

to assess the effect of endocytosis inhibition on the systemic distribution and clearance of

cinnarizine and cholesterol. In these studies, blank (i.e. not containing cinnarizine and 14C-

cholesterol) lipid emulsion (with or without 100 µM monensin) was infused intraduodenally as


113

described above, and intravenous infusion of the cinnarizine and 14C-cholesterol containing

intravenous formulation was commenced at the same time. The intravenous formulation (1 mL)

was infused into the right jugular vein at a rate of 0.05 mL/15 sec (total infusion period < 5 min).

Blood samples (0.3 mL) were collected via the carotid artery cannula at t = 5, 15, 30, 60, 120, 180,

240, 360, 480 min after infusion initiation into tubes containing 3 I.U. heparin. After each blood

sample was taken, the cannula was flushed with 0.3 mL of 2 I.U./mL heparinised saline. Plasma

was separated by centrifugation (10,000 xg, 5 min) to enable analysis of drug and labelled

cholesterol content.

3.3.11 In situ single-pass rat jejunum perfusion

After surgery, animals were equilibrated for 30-min, during which time heparinised donor rat

blood was infused via the jugular vein as described above. During re-equilibration, blood from the

cannulated mesenteric vein (~ 0.3 mL/min) was collected for re-infusion. Perfusion buffer was

pumped through the jejunal segment at a rate of 0.1 mL/min and outflowing buffer discarded to

waste. The exposed jejunal segment was kept moist by covering with saline-soaked gauze

throughout the experiment.

In all experiments, the concentration of cinnarizine in the perfusate was held at 0.1 mg/mL (~ 40%

saturated solubility). Therefore, in experiments where micelles or vesicles were perfused alone,

cinnarizine was loaded into the perfusate at 0.1 mg/mL. In experiments where micelles or vesicles

were co-perfused in a 1:1 v/v ratio with a secondary perfusate of bile/bile pH 6.30/buffer pH 6.30,

cinnarizine was loaded into the primary perfusate at 0.2 mg/mL, such that 1:1 v/v dilution led to a

final perfusate concentration of 0.1 mg/mL.


114

Perfusate flow was maintained at 0.1 mL/min in all experiments to minimise variations in the

thickness of the unstirred water layer206. For experiments where 1:1 v/v co-perfusion of the phases

with bile/bile pH 6.30/buffer pH 6.30 was required, micelles/vesicles and bile/buffer were pumped

at 0.05 mL/min, and mixed via a three-way “T” connector immediately prior to entry into the

jejunal segment, providing a total perfusate flow of 0.1 mL/min. Perfusate was sampled at t = 0 to

confirm lipid and drug concentrations. After this time, the outflowing perfusate was continuously

collected at 10-min intervals, and briefly vortexed before samples were taken for analysis of drug

and lipid content. For experiments where drug supersaturation was generated, perfusate samples

were taken before and after centrifugation (2,200 xg, 2 min), to assess the degree of drug

precipitation within the jejunal segment. Blood draining the perfused jejunal segment was

collected at 5-min intervals, plasma separated by centrifugation (10,000 xg, 5 min), and samples

taken for analysis of drug content by HPLC as described below.

3.3.12 Analytical procedures

3.3.12.1 Sample preparation and HPLC assay conditions for cinnarizine

Samples of lipid emulsion were prepared for HPLC assay by an initial 80-fold dilution with

chloroform:methanol (2:1 v/v), followed by a 10-fold dilution with mobile phase (50% v/v

acetonitrile:50% v/v 20 mM NH4H2PO4). Samples of IV formulation and micelles/vesicles were

prepared for HPLC assay by a 900-fold, and a 400-fold dilution with mobile phase (50% v/v

acetonitrile:50% v/v 20 mM NH4H2PO4), respectively. Plasma samples were prepared for HPLC

using a validated extraction procedure, with flunarizine as an internal standard, as reported

previously43.


115

Cinnarizine HPLC assay conditions were as described previously43, with slight modification to the

mobile phase employed for cinnarizine quantification in plasma, to 45% v/v acetonitrile:55% v/v

20 mM NH4H2PO4 in this study. Replicate analysis of n = 4 quality control samples revealed

acceptable accuracy and precision (± 10%, ± 15% at the limit of quantification) for concentrations

between 20–1000 ng/mL for IV formulation, emulsion, micelles and vesicles, and 10–320 ng/mL

for plasma.

3.3.12.2 Scintillation counting

Quantification of 14C-cholesterol and 3H-oleic acid in the plasma was performed via scintillation

counting on a Packard Tri-Carb 2000CA liquid scintillation analyser (Packard, Meriden,

Connecticut, USA). Plasma samples (50 µL) were added to 2 mL Irga-safe Plus scintillation fluid

followed by a 10-sec vortex. Samples were corrected for background radioactivity by the inclusion

of a blank plasma sample in each run.

3.3.12.3 Blood:plasma ratio determination for cinnarizine

The blood:plasma ratio for cinnarizine was determined by spiking 0.5 mL blank blood with known

amounts of cinnarizine to achieve low, medium, and high concentrations (in triplicate). Plasma

was separated by centrifugation (10,000 xg, 5 min) and plasma drug concentration assayed by

HPLC. The blood:plasma ratio was calculated from the ratio of known concentration in spiked

blood to the concentration measured in plasma separated from spiked blood. The mean

blood:plasma ratio was subsequently used to convert plasma concentrations to blood

concentrations in perfusion experiments, enabling quantification of total drug transport into

mesenteric blood.


116

3.3.12.4 Calculations

In the single-pass rat jejunum perfusion model, permeability coefficients were calculated using

steady state drug concentrations in perfusate and blood. Two apparent permeability coefficients

(Papp) were calculated as described previously184:

′Disappearance P . ln Equation 3.2


∆.

Equation 3.3

where ‘Disappearance’ Papp is the apparent permeability coefficient calculated from drug loss from

the perfusate (cm/s); ‘Appearance’ Papp is the apparent permeability coefficient calculated from

drug appearance in the mesenteric blood (cm/s); Q is the perfusate flow rate (mL/s); A is the

surface area of the perfused jejunal segment (cm2), which is calculated by multiplying the

diameter by the length of the perfused intestinal segment as described previously207; C1 is the

average steady state drug concentration exiting the perfused jejunal segment (ng/mL); C0 is the

drug concentration entering the jejunal segment (ng/mL); ∆MB/∆t is the average rate of drug mass

appearance in mesenteric blood at steady state (ng/s); and <C> is the logarithmic mean drug

concentration in the lumen (ng/mL), where <C> = (C1 – C0)/(ln C1 – ln C0).

3.3.12.5 Non-compartmental pharmacokinetic analysis

The maximum plasma concentration (Cmax), time to reach Cmax (Tmax), area under the plasma

concentration-time curve from time zero to the last measured concentration (AUC0-8 h), area under

the plasma concentration-time curve extrapolated to infinity (AUC0-inf), elimination rate constant


117

(Ke), volume of distribution (Vd), and clearance (Cl) were calculated using WINONLIN version

5.3 (Pharsight Inc., Apex, NC, USA).

3.3.13 Statistical analysis

Results were analysed using Student’s t test. A P value of < 0.05 was considered to be a

significant difference. Analyses were performed using SPSS v19 for Windows (SPSS Inc.,

Chicago, IL, USA).

3.4 RESULTS

3.4.1 SR-BI, CD36, NPC1L1 and endocytosis have little impact on drug

absorption from intestinal colloidal phases

Inhibition of the lipid uptake receptors SR-BI (by co-infusion of 100 µM BLT-1), CD36 (by co-

infusion of 1 mM SSO), and NPC1L1 (by intravenous administration of 0.3 mg/kg ezetimibe) did

not result in significant changes to the systemic plasma concentration-time profiles or systemic

exposure of cinnarizine (Figure 3.2A, Table 3.2). Inhibition of endocytosis (by co-infusion of 100

µM monensin) unexpectedly led to significant increases in cinnarizine systemic plasma

concentration at t = 2.5 and 4 h (Figure 3.2A), and increased the AUC0-8 h 1.5-fold (Table 3.2).

However, subsequent intravenous dosing studies revealed significantly lower Vd and Cl values in

the monensin-treated rats when compared to control (Table 3.3), suggesting that the increase in

cinnarizine systemic exposure was due to a decrease in cinnarizine systemic distribution and

clearance, rather than changes to intestinal absorption. The data suggest a limited role for SR-BI,

CD36, NPC1L1, and endocytosis generally, in the absorption of cinnarizine from micelles and

vesicles.


118

In contrast, inhibition of NPC1L1 did lead to significantly lower systemic plasma concentrations

of exogenously dosed cholesterol at t = 2, 2.5, 3, 4, 6, 8 h (Figure 3.2B). Plasma concentrations of

exogenously dosed oleic acid also appeared lower in rats following inhibition of CD36 although

differences were not significantly different (Figure 3.2C). While inhibition of SR-BI and CD36

did not lead to significant changes to the systemic plasma concentration-time profile of

exogenously dosed cholesterol, inhibition of endocytosis significantly increased the systemic

plasma concentrations of exogenously dosed cholesterol at t = 4, 8 h (Figure 3.2B). This increase

in exposure is likely explained by a decrease in systemic distribution of 14C-cholesterol in the

monensin-treated rats, as plasma 14C-cholesterol concentration at early sample time points (which

reflect the distribution phase) in intravenous studies were significantly higher in the monensin-

treated rats (Figure 3.3B) when compared to controls (Vd and Cl could not be calculated for

cholesterol as a typical elimination phase was not evident in the systemic plasma concentration vs.

time profiles of cholesterol - Figure 3.2B and Figure 3.3B).


119

Figure 3.2: Systemic plasma concentration-time profiles of (A) cinnarizine (CIN), (B) 14C-

labelled cholesterol (Ch), and (C) 3H-labelled oleic acid (OA) following intraduodenal infusion of

a 3 mL lipid emulsion consisting 1 mg cinnarizine emulsified in 59 mM oleic acid, 8 mM sodium

taurocholate, 2 mM phosphatidylcholine, 2 mM cholesterol, 1 µCi 14C-cholesterol and/or 3 µCi 3H-oleic acid. Experiments were performed in control rats (filled circle); rats treated with 100 µM

BLT-1 (open circle); rats treated with 1 mM SSO (open triangle), rats treated with 0.3 mg/kg

ezetimibe (open square), and rats treated with 100 µM monensin (open diamond), to inhibit the

lipid uptake receptors SR-BI, CD36, NPC1L1, and endocytosis, respectively. BLT-1, SSO, and

monensin were co-infused as part of the lipid emulsion; ezetimibe was administered intravenously.

Data represent mean ± SEM of n = 4 rats for (A) and (B); and n = 3 rats for (C). Statistically

significant difference with respect to control rats (p < 0.05) is denoted by the symbol *.

Time (h)

0 2 4 6 8

Pla

sma

CIN

con

c.

(ng

/mL)

0

100

200

300

400

500 ControlBLT-1SSO Ezetimibe Monensin

A

*

*

Time (h)

0 2 4 6 8

Pla

sma

14C

-Ch

con

c.(n

g/m

L)

0

20

40

60

ControlBLT-1SSOEzetimibeMonensin

B

* ** * * *

* *

Time (h)

0 2 4 6 8

Pla

sma

3H

-OA

co

nc.

(ng

/mL)

0.00

0.01

0.02

0.03

0.04

0.05

0.06ControlSSO

C


120

Table 3.2: Pharmacokinetic parameters for cinnarizine after intraduodenal administration of a 3

mL lipid emulsion in rats. Experiments were performed in control rats; rats treated with 100 µM

BLT-1; rats treated with 1 mM SSO, rats treated with 0.3 mg/kg ezetimibe, and rats treated with

100 µM monensin, to inhibit the lipid uptake receptors SR-BI, CD36, NPC1L1, and endocytosis,

respectively. Values represent mean ± SEM of n = 4 rats.

Experimental group Formulation type

CIN dose

(mg/kg)

AUC0-8 h

(ng h/mL)

Cmax

(ng/mL)

Tmax

(h)

Control Emulsion 3.33 978 ± 122 288.1 ± 26.6 2.4 ± 0.1

BLT-1 treated Emulsion 3.33 883 ± 21 219.0 ± 20.0 2.6 ± 0.5

SSO-treated Emulsion 3.33 924 ± 88 195.5 ± 33.6 2.4 ± 0.2

Ezetimibe-treated Emulsion 3.33 1048 ± 92 333.9 ± 13.6 2.3 ± 0.1

Monensin-treated Emulsion 3.33 1487 ± 103a 432.8 ± 45.9a 2.3 ± 0.1

a Significant difference when compared to control group


121

Time (min)

0 100 200 300 400 500

Pla

sma

CIN

co

nc.

(ng

/mL

)

10

100

1000ControlMonensin

** * *

(A)

Time (min)

0 100 200 300 400 500

Pla

sma

14

C-C

h c

on

c.

(ng

/mL

)

100

1000ControlMonensin*

**

(B)

Figure 3.3: Systemic plasma concentration-time profiles of (A) cinnarizine (CIN), and (B) 14C-

labelled cholesterol (Ch) following intravenous infusion over 5 min of 1 mL formulation

containing 0.5 mg cinnarizine and 1 µCi 14C-Ch. Experiments were performed in control rats

(filled circle) and rats treated with 100 µM monensin (open diamond). Intraduodenal infusion of

blank lipid emulsion was also commenced at t = 0 (to match experimental conditions in Figure

3.2). Monensin, an endocytosis inhibitor, was pre-infused intraduodenally 30 min prior to t = 0,

and subsequently co-infused as part of the lipid emulsion. Data represent mean ± SEM of n = 4

rats for (A); and n = 3 rats for (B). Statistically significant difference with respect to control rats (p

< 0.05) is denoted by the symbol *.

Table 3.3: Pharmacokinetic parameters for cinnarizine after intravenous administration in rats.

Values represent mean ± SEM of n = 3-4 rats

Experimental group

CIN dose (mg/kg)

AUC0-inf

(ng h/mL) Ke

(1/h) Vd

(L/kg) Cl

(L/h/kg)

Control 1.67 821 ± 24 0.25 ± 0.03 5.8 ± 0.2 2.03 ± 0.06

Monensin-treated 1.67 1077 ± 28a 0.28 ± 0.04 4.0 ± 0.4a 1.55 ± 0.04a

a Significant difference when compared to control group


122

3.4.2 Drug absorption from micelles and vesicles is determined by Cfree and

not colloidal structure

The intestinal perfusion of colloidal media with markedly different compositional profiles

(micelles and vesicles) but with comparable Cfree and thermodynamic activity did not result in

significant differences in steady-state absorptive flux, disappearance Papp or appearance Papp of

cinnarizine (Table 3.4, Figure 3.4A – filled symbols). Similarly, in spite of large differences in

particle size and composition, the intraduodenal infusion of the same micellar or vesicular systems

to bile-diverted rats did not result in significant differences in systemic plasma concentration-time

profiles (Figure 3.5 – filled symbols) and pharmacokinetic parameters of cinnarizine (Table 3.5 -

micelles vs. vesicles in bile-diverted rats). The results suggest that cinnarizine absorption from

lipid colloidal phases is relatively insensitive to the physical nature of the colloidal milieu, and is

instead controlled largely by Cfree.


123

Figure 3.4: (A) Absorptive flux of cinnarizine (CIN) into mesenteric blood (ng/5 min/10 cm2) and

(B) & (C) CIN disappearance from intestinal perfusate (% drug dose passing through jejunum)

when micelles/vesicles were perfused through an isolated rat jejunal segment (~ 10 cm2), with

(open symbols) and without (filled symbols) 1:1 v/v co-perfusion with rat bile. Co-perfusion of

micelles and vesicles with donor bile generates drug supersaturation in situ within the perfused

jejunal segment. SS denotes experiments where drug is supersaturated in the perfusate. The degree

of drug precipitation within the perfusate is illustrated in (B) and (C) as the difference in perfusate

concentration between pre- and post-centrifugation data. Precipitation was significant in the

vesicle, but not micellar groups. Experiments were performed using an in situ single-pass rat

jejunum perfusion model. In all experiments, the concentration of cinnarizine in perfusate and the

total perfusate flow rate were kept constant at 0.1 mg/mL and 0.1 mL/min, respectively. Data

represent mean ± SEM of n = 3-4 experiments.

Time (min)

0 10 20 30 40 50 60

CIN

flu

x in

to m

ese

nte

ric

blo

od

(n

g/5

min

/10

cm

2)

0200400600800

100012001400

Micelles

Micelles + Bile ss

Vesicles

Vesicles + Bile ss

A

Time (min)

0 20 30 40 50 60

% D

ose

pas

sin

g

thro

ug

h je

jun

um

0

20

40

60

80

100

MicellesMicelles + Bile SS (pre-centrifugation)Micelles + Bile SS (post-centrifugation)

B

Time (min)

0 20 30 40 50 60

% D

ose

pa

ssin

g

thro

ug

h je

jun

um

0

20

40

60

80

100

VesiclesVesicles + Bile SS (pre-centrifugation)Vesicles + Bile SS (post-centrifugation)

C


124

Table 3.4: Cinnarizine disappearance Papp (x 106 cm/s) from the intestinal perfusate, appearance

Papp (x 106 cm/s) in the mesenteric blood, and steady state absorptive flux into mesenteric blood

(ng/5 min/10 cm2) after 60 min of single-pass perfusion of ~ 10 cm2 segments of rat jejunum with

model micelles and vesicles, with and without 1:1 v/v co-perfusion with rat bile, rat bile pH 6.30

or buffer pH 6.30. Values calculated using data obtained after steady state attainment (t = 40-60

min). In all experiments, cinnarizine concentration in perfusate and total perfusate flow rate were

kept constant at 0.1 mg/mL and 0.1 mL/min, respectively. Data represent mean ± SEM of n = 3-4

experiments.

SS ratio #

Disappearance Papp

(x 106 cm/s)

Appearance Papp

(x 106 cm/s)

Flux into mesenteric blood (ng/5 min/10 cm2)

Micelles 0.4 19.4 ± 3.0 1.1 ± 0.2 310 ± 50

Micelles + Bile SS 6.1 29.3 ± 3.0 a 3.7 ± 0.1 a 979 ± 30 a

Micelles + Bile pH 6.30 SS 3.0 39.5 ± 5.6 a 4.2 ± 1.0 a 1099 ± 279 a

Micelles + Buffer pH 6.30 SS 2.1 38.9 ± 4.1 a 4.4 ± 0.4 a 1180 ± 100 a

Vesicles 0.4 22.0 ± 2.2 1.2 ± 0.1 340 ± 29

Vesicles + Bile SS 11.6 66.0 ± 18.3b 2.0 ± 0.7 499 ± 155

a Significant increase from micelles alone b Significant increase from vesicles alone SS denotes drug supersaturation in perfusate # SS ratio = Supersaturation ratio = (Supersaturated) concentration of drug in perfusate / Equilibrium

solubility of drug in perfusate


125

(A)

(B)

Time (h)

0 2 4 6 8

Pla

sma

CIN

con

c.

(ng/

mL)

020406080

100120140160

Micelles (Bile-diverted)Micelles (Bile-intact)

* * *

Time (h)

0 2 4 6 8

Pla

sma

CIN

con

c.

(ng/

mL)

020406080

100120140160

Vesicles (Bile-diverted)Vesicles (Bile-intact)

Figure 3.5: Systemic plasma concentration-time profiles of cinnarizine (CIN) following a 2 h

intraduodenal infusion of cinnarizine-loaded (0.2 mg/mL) (A) micelles and (B) vesicles to bile-

intact and bile-diverted rats. Consistent with observations in rat jejunum perfusion studies, bile-

induced supersaturation translated into increased in vivo exposure during the absorption phase in

the case of micelles but not vesicles. Data represent mean ± SEM of n = 4 rats. Statistical

significance (p < 0.05) is denoted by the symbol *.

Table 3.5: Pharmacokinetic parameters for cinnarizine after intraduodenal administration of

cinnarizine-loaded (0.2 mg/mL) micelles and vesicles to bile-intact and bile-diverted rats. Values

represent mean ± SEM of n = 4 rats.

Experimental group Formulation type

CIN dose (mg/kg)

AUC0-8 h (ng h/mL)

Cmax (ng/mL)

Tmax

(h)

Micelles (bile-intact) Micelles 2 502 ± 59a 138.8 ± 11.6a 1.6 ± 0.2

Micelles (bile-diverted) Micelles 2 362 ± 39 85.7 ± 14.3 2.0 ± 0.4

Vesicles (bile-intact) Vesicles 2 411 ± 74 109.4 ± 10.6 2.3 ± 0.3

Vesicles (bile-diverted) Vesicles 2 393 ± 82 97.6 ± 12.8 2.0 ± 0.4

a Significant increase when compared to micelles (bile-diverted) group


126

3.4.3 Bile-mediated dilution of cinnarizine-loaded micelles and vesicles

generates drug supersaturation

The equilibrium solubility and change in cinnarizine solubilisation capacity of model micelles and

vesicles before and after 1:1 v/v dilution with bile, bile pH 6.30 or buffer pH 6.30 are tabulated in

Table 3.6, and shown graphically in Figure 3.6.

Table 3.6: Equilibrium solubility (37 °C) and percent original solubilisation capacity values of

cinnarizine in model micelles and vesicles, before and after 1:1 v/v addition of rat bile, rat bile pH

6.30 or buffer pH 6.30. Data represent mean ± SEM of n = 3-4 determinations.

Equilibrium solubility (µg/mL)

Percent original solubilisation capacity (%) #

Micelles Micelles alone 236 ± 9.3 100

+ Bile (1:1) 16.5 ± 0.8 14.0 ± 0.7

+ Bile pH 6.30 (1:1) 33.1 ± 1.0 28.0 ± 0.9

+ Buffer pH 6.30 (1:1) 48.7 ± 0.8 41.2 ± 0.6

Vesicles Vesicles alone 232 ± 3.3 100

+ Bile (1:1) 8.60 ± 0.5 7.4 ± 0.4

+ Bile pH 6.30 (1:1) 17.4 ± 0.4 15.0 ± 0.3

+ Buffer pH 6.30 (1:1) 24.9 ± 0.3 21.4 ± 0.2

Percentoriginalsolubilisationcapacity# Solubility ∗ VolumeSolubility ∗ Volume

x100%


127

% o

rigin

al C

IN

solu

bilis

atio

n ca

paci

ty

0

20

40

60

80

100

120

140Colloid onlyColloid + Bile (1:1)Colloid + Bile pH 6.30 (1:1)Colloid + Buffer pH 6.30 (1:1)

Micelles Vesicles

**

*

* * *

Figure 3.6: Percent original cinnarizine (CIN) solubilisation capacity of model colloids (micelles

or vesicles), before and after a 1:1 v/v addition of rat bile, rat bile pH 6.30 or buffer pH 6.30. Data

represent mean ± SEM of n = 3-4 determinations. Statistical significant difference (p < 0.05) to

colloid only is denoted by the symbol *.

Dilution of the micellar or vesicular systems in a 1:1 v/v ratio with bile obtained from donor

animals led to significant decreases in cinnarizine solubilisation capacity. As a proportion of initial,

the solubilisation capacity of the micellar system dropped significantly from 100% to 14%, and

for vesicles the decrease was even greater from 100% to 7%. This was in spite of the fact that the

bile salt concentration in donor bile was higher than that in the micellar or vesicular system

(average total bile salt concentration in donor bile was 14.7 ± 0.9 mM; mean ± SEM, n = 13), and

therefore ‘dilution’ with bile did not reduce bile salt concentrations below the critical micellar

concentration (CMC) and instead increased the overall bile salt concentration.

The pH of donor bile was higher than intestinal pH (average pH of donor bile was 8.02 ± 0.02;

mean ± SEM, n = 5). As such additional studies were performed to examine whether the effects on

solubility reflected a pH effect. 1:1 v/v dilution of the phases with pH adjusted bile (pH of bile

adjusted to 6.30 to match the pH of the micelles and vesicles) also resulted in significant decreases


128

in solubilisation capacity (to 28% and 15% of initial for micelles and vesicles, respectively),

although the decrease was slightly attenuated when compared to non-pH adjusted bile. Finally, the

micellar and vesicular systems were diluted 1:1 v/v with buffer at pH 6.30 in an attempt to

uncouple simple dilution effects from pH effects and bile-mediated effects. Dilution with pH 6.30

buffer also decreased cinnarizine solubilisation capacity significantly, although to a lesser extent

(41% and 21% in micelles and vesicles, respectively).

Analysis of the kinetics of drug precipitation (Figure 3.7) demonstrated that bile addition to

cinnarizine-loaded micelles and vesicles (cinnarizine present in phases at 0.2 mg/mL, ~ 80%

saturated solubility) did not result in immediate drug precipitation, and was preceded by a period

of supersaturation. The time taken for cinnarizine precipitation to occur was variable, however

supersaturation was maintained for longer periods in micelles (> 20 min in all cases) when

compared to vesicles (1-20 min).

Time (min)

0 20 40 60 80 100 120

% o

rigin

al C

IN in

sol

utio

n

0

20

40

60

80

100

120

Micelles + BileVesicles + Bile

Equilibrium solubility: Micelles + Bile Equilibrium solubility: Vesicles + Bile

Add bile

Figure 3.7: Kinetics of cinnarizine (CIN) precipitation from model micelles (filled circles, n = 5)

and model vesicles (open circles, n = 4) upon addition of rat bile (in a 1:1 v/v ratio) at t = 0.

Addition of bile reduces the equilibrium cinnarizine solubilisation capacity of micelles and

vesicles to 14% and 7% of initial, respectively (see Table 3.6, and shown here as the lines denoted


129

equilibrium solubility). Cinnarizine supersaturation appeared to be maintained for longer in

micelles than in vesicles. Cinnarizine was loaded into micelles and vesicles at 80% saturation (~

0.2 mg/mL). Each line represents individual experiments.

3.4.4 Bile-induced drug supersaturation increases jejunal absorptive flux for

micelles but not vesicles

The absorptive flux vs. time profiles for cinnarizine under supersaturated conditions (i.e. when bile

was co-perfused with the phases) and under non-supersaturated conditions (when bile was not co-

perfused with the phases) are shown in Figure 3.4A. Steady state-absorptive flux, disappearance

Papp, and appearance Papp of cinnarizine in all perfusion experiments are reported in Table 3.4.

Co-perfusion of micelles with bile in a 1:1 v/v ratio increased the absorptive flux, disappearance

Papp, and appearance Papp of cinnarizine from micelles 3.2-fold, 1.5-fold, and 3.4-fold, respectively.

In contrast, 1:1 v/v co-perfusion of vesicles with bile did not lead to significant changes in

cinnarizine absorptive flux or appearance Papp. Disappearance Papp for cinnarizine did increase 3.0-

fold when vesicles were co-perfused with bile, however the drop in perfusate drug concentration

was largely a result of rapid drug precipitation in the perfusate (see below).

For the micellar preparation, bile-induced supersaturation was relatively stable throughout the

experimental period. Thus the cinnarizine concentration in the perfusate was essentially the same

before or after centrifugation (Figure 3.4B). In contrast, when vesicles were co-perfused with bile,

significant drug precipitation was observed during the time required for perfusate to transit the

jejunal segment (as indicated by the difference between the pre- and post-centrifugation data in

Figure 3.4C). This was consistent with the in vitro dilution profiles in Figure 3.7.

.


130

To distinguish between bile-induced increases in cinnarizine absorptive flux resulting from drug

supersaturation, pH increases (since an increase in pH might be expected to increase the

permeability of a weak base), and non-specific effects of bile components on membrane

permeability, micelles were also co-perfused with pH-adjusted bile (pH 6.30) and buffer (pH 6.30).

1:1 v/v co-perfusion of micelles with bile pH 6.30 or buffer pH 6.30 increased cinnarizine

absorptive flux 3.5-fold and 3.8-fold, respectively, when compared to the perfusion of micelles

alone (Figure 3.8, Table 3.4). Stable supersaturation was also generated within the perfused

jejunal segment in these experiments (Figure 3.9 - cinnarizine concentration in the perfusate was

essentially the same before and after centrifugation).

Time (min)

0 10 20 30 40 50 60

CIN

flu

x in

to m

esen

teric

blo

od (

ng/5

min

/10

cm2 )

0

500

1000

1500

2000MicellesMicelles + Bile SS

Micelles + Bile pH 6.30 SS

Micelles + Buffer pH 6.30 SS

Figure 3.8: Absorptive flux-time profiles of cinnarizine (CIN) when micelles were perfused

through an isolated rat jejunal segment (~ 10 cm2), with and without 1:1 v/v co-perfusion with rat

bile, rat bile pH 6.30 or buffer pH 6.30. SS denotes experiments where drug is supersaturated in

perfusate. Co-perfusion of rat bile, rat bile pH 6.30 or buffer pH 6.30 with micelles generates drug

supersaturation in situ within the perfused jejunal segment in all cases, and increased cinnarizine

absorptive flux by 3.2-fold, 3.5-fold and 3.8-fold, respectively. Experiments were performed using

an in situ single-pass rat jejunum perfusion model. In all experiments, the concentration of

cinnarizine in perfusate and the total perfusate flow rate were kept constant at 0.1 mg/mL and 0.1

mL/min, respectively. This series of experiments provides further support for the suggestion that


131

supersaturation was responsible for the increase in absorptive flux seen in Figure 3.4A. Data

represent mean ± SEM of n = 3-4 experiments.

Time (min)

0 20 30 40 50 60

% D

ose

pass

ing

thro

ugh

jeju

num

0

20

40

60

80

100

Micelles + Bile pH 6.30 (pre-centrifugation)Micelles + Bile pH 6.30 (post-centrifugation)

(A)

Time (min)

0 20 30 40 50 60

% D

ose

pass

ing

thro

ugh

jeju

num

0

20

40

60

80

100

Micelles + Buffer pH 6.30 (pre-centrifugation) Micelles + Buffer pH 6.30 (post-centrifugation)

(B)

Figure 3.9: Perfusate disappearance (% drug dose passing through jejunum) profiles of

cinnarizine (CIN) when micelles were perfused through an isolated rat jejunal segment (~ 10 cm2),

with and without 1:1 v/v co-perfusion with (A) rat bile pH 6.30 or (B) buffer pH 6.30. Co-

perfusion of micelles with rat bile pH 6.30 or buffer pH 6.30 generates drug supersaturation in situ

within the perfused jejunal segment. The degree of drug precipitation within the perfusate is

represented by the difference in perfusate concentration between pre- and post-centrifugation data.

Experiments were performed using an in situ single-pass rat jejunum perfusion model. In all

experiments, the concentration of cinnarizine in perfusate and the total perfusate flow rate were

kept constant at 0.1 mg/mL and 0.1 mL/min, respectively. Data represent mean ± SEM of n = 3-4

experiments.

Since 1:1 v/v co-perfusion of micelles with bile, bile pH 6.30 and buffer pH 6.30 are all expected

to generate cinnarizine supersaturation within the perfused jejunal segment (according to the

solubility data in Table 3.6, and the lack of drug precipitation in outflow perfusate in all cases), the

observation that absorptive flux enhancement was similar in all groups (Figure 3.8) suggest that

the enhancement was attributable to drug supersaturation, and not an increase in system pH (by

comparing micelles + bile group with micelles + bile pH 6.30 group), or non-specific effects of

bile on membrane permeability (by comparing micelles + bile pH 6.30 group with micelles +


132

buffer pH 6.30 group). However, the degree of flux enhancement did not correlate with the degree

of supersaturation, as flux enhancement was similar in all groups despite the significantly higher

supersaturation ratio generated in the micelles + bile group, when compared to the micelles + bile

pH 6.30 group and micelles + buffer pH 6.30 group (supersaturation ratio of 6 vs. 2-3) (Table 3.4).

Previous studies have shown that for lipophilic drugs such as cinnarizine, the drug fraction

extracted into octanol (and analogous to the permeable fraction) as a function of pH is shifted to

lower pHs than would be expected based on the unionised fraction208, 209. This left shift would

limit pH effects on permeability over the range of pH 6-8, consistent with the observations here.

3.4.5 Bile-induced drug supersaturation increases in vivo cinnarizine

exposure after intraduodenal infusion

The systemic plasma concentration-time profiles of cinnarizine following intraduodenal infusion

of cinnarizine-loaded micelles and vesicles (0.2 mg/mL; ~ 80% saturated solubility) to bile-intact

and bile-diverted rats are shown in Figure 3.5. The pharmacokinetic parameters of cinnarizine for

the bioavailability studies are reported in Table 3.5. Consistent with the results from the rat jejunal

permeability studies, intraduodenal administration of micelles to bile-intact rats resulted in

significantly higher systemic plasma concentrations of cinnarizine at t = 1, 1.5 and 2 h (AUC0-8 h

increased 1.4-fold) when compared to bile-diverted rats. In contrast, when vesicles were dosed to

bile-intact rats, the systemic plasma concentration-time profiles and AUC0-8 h of cinnarizine were

not different to that observed in bile-diverted rats.

3.5 DISCUSSION

After oral administration, the absorption of poorly water-soluble drugs (PWSD) is often limited by

slow dissolution and low solubility in the GI tract. LBF overcome many of the dissolution

limitations of PWSD (by providing a mechanism to circumvent dissolution from the solid to liquid


133

state), however, the solubility limitations of PWSD are seemingly unaddressed, since

solubilisation in the lipid-based colloidal phases that result from the digestion of LBF does not

typically enhance free drug concentrations101, 143. Nonetheless, co-administration of lipids (either

formulation lipids or via co-administration with lipid-rich foods) remains a highly effective means

to promote the absorption of PWSD. This suggests the potential for alternative mechanisms by

which LBF enhance drug absorption. In the current communication, two possible alternatives to

the traditional model of drug absorption from LBF have been explored. Firstly, that drug

absorption from lipid colloidal phases may involve a collisional uptake component (i.e. drug

absorption directly from the solubilised phase), and secondly that flux across the absorptive

membrane may be enhanced by a transient increase in the thermodynamic activity of drug in

intestinal colloidal phases due to supersaturation.

The data describing drug absorption from micellar and vesicular colloidal phases suggest that

direct interactions between colloids (or at least the systems examined here) and the absorptive

membrane do not play an important role in cinnarizine absorption. Thus, comparable absorptive

flux (Figure 3.4A – filled symbols) and systemic plasma concentration-time profiles (Figure 3.5 –

bile diverted rats) were observed following jejunal perfusion and intraduodenal infusion of

micelles and vesicles. The thermodynamic activity and Cfree of both systems were held constant,

but the large difference in hydrodynamic radius (9 nm of micelles vs. 443 nm of vesicles) and

higher bile salt, LPC and lipid concentrations in the micellar system (Table 3.1) suggest that the

number of administered micellar particles was substantially higher than that of vesicular particles.

Collision-mediated absorption is highly sensitive to increased particle number, since this increases

the statistical likelihood of collisions and collisional transfer150. As such collisional interactions

did not appear to dictate the degree of drug absorption from micelles and vesicles and instead,

absorption was seemingly controlled by thermodynamic activity (or Cfree). Differences in particle


134

size might also be expected to alter colloid diffusion across the unstirred water layer (UWL). The

similarity in absorption profiles from micelles and vesicles (Figure 3.4A – filled symbols)

therefore suggests that diffusion across the UWL was either non-limiting, or was not affected by

particle size in this experimental model. Similarly, the rates of replenishment of Cfree (i.e. the rate

of re-establishment of the equilibrium between solubilised and free drug) might be expected to be

different between the two different particle size colloids, however this was presumably

sufficiently fast in both cases to have little impact on drug absorption in the current model.

Data obtained from the lipid uptake receptor inhibition studies further support the notion that

collisional uptake mechanisms have a limited role in drug absorption from these systems. SR-BI,

CD36, and NPC1L1 were examined since they have previously been shown to mediate the cellular

uptake of fatty acids and/or cholesterol61, 66, 210; the absorption of which is also facilitated by

micellar solubilisation. Whether SR-BI, CD36, and NPC1L1 function as authentic transporters

that directly mediate lipid absorption61, 66-68, 210, or whether they act to facilitate intracellular lipid

trafficking or to modify signalling processes that mediate lipid absorption154, 211, 212, or both,

remains contentious but nonetheless all merit examination here. Recent reports also suggest the

involvement of SR-BI, CD36, and NPC1L1 in the absorption of fat-soluble nutrients (such as

carotenoids66, 213, vitamin D214, and vitamin E215), providing further support for a role in drug

absorption. In the current study, consistent with previous reports in vivo, inhibition of NPC1L1

reduced cholesterol absorption (Figure 3.2B)68, and inhibition of CD36 reduced (albeit non-

significantly) the absorption of oleic acid (Figure 3.2C)61. In contrast, inhibition of SR-BI and

CD36 had little impact on cholesterol absorption (Figure 3.2B). Inhibition of SR-BI and CD36

was expected to reduce cholesterol absorption based on previous in vitro studies61, 66, 67. However,

in vivo evidence of a role of SR-BI and CD36 in cholesterol absorption is less clear and, for

example, no significant differences in intestinal cholesterol absorption were reported in SR-BI


135

knockout vs. wild-type mice66, 157. The lack of effect of SR-BI and CD36 inhibition on cholesterol

absorption in the current study therefore runs contrary to previous in vitro studies but is in

agreement with some previous in vivo data. Inhibition of endocytosis pathways via administration

of monensin, also failed to reduce cholesterol and cinnarizine absorption. Rather, the volume of

distribution and clearance of cholesterol and cinnarizine appeared to be reduced, resulting in

increases in plasma exposure. Since exogenously dosed 14C-cholesterol and cinnarizine are likely

to be present within lipoproteins in the systemic circulation, the changes in systemic disposition of

14C-cholesterol and cinnarizine in the monensin-treated rats may reflect inhibition of receptor-

mediated endocytosis of LDL216.

Inhibition of SR-BI, CD36 and NPC1L1 did not result in significant changes to the systemic

exposure of cinnarizine following intraduodenal infusion of a lipid emulsion formulation in rats

(Figure 3.2A). The data suggest that while lipid colloidal phases may be capable of direct

interaction with lipid receptors (as suggested previously), the transfer of solubilised content into

absorptive cells is likely selective and more applicable to solubilised lipids and nutrients rather

than drugs. Together with the data describing cinnarizine absorption from micelles and vesicles,

the results indicate that cinnarizine absorption from lipid colloidal phases is largely independent of

the physical nature of the infused colloid (realising that in these first experiments cinnarizine was

present at concentration well below the solubilisation limit and therefore under conditions where

solubility/precipitation-mediated events were avoided), is not influenced significantly by common

lipid uptake receptors, and instead appears to be primarily dependent on the free drug

concentration in equilibrium with the solubilised reservoir.

Subsequent studies addressed the possibility that bile secretion may enhance drug absorption via

the induction of supersaturation during the intestinal processing of dietary or formulation-related


136

lipids. The addition of bile to cinnarizine-loaded micelles led to sustained drug supersaturation

that ultimately led to increased intestinal drug absorption and systemic drug exposure. In contrast,

although addition of bile to cinnarizine-loaded vesicles also led to supersaturation, the metastable

supersaturated state was less stable than that generated by bile addition to micelles, resulting in

more rapid precipitation of solubilised drug and therefore a lack of increase in drug absorption and

systemic drug exposure. The data indicate that bile-mediated dilution of lipid colloidal phases may

represent an endogenous mechanism of supersaturation generation during lipid processing in the

small intestine; and that the transient increase in thermodynamic activity may lead to enhanced

drug absorption. The observations also highlight both the potential for supersaturation to enhance

drug absorption, and the need to achieve an optimal balance between drug supersaturation and

drug precipitation.

Interaction of intestinal colloidal phases with bile leads to dilution, an increase in pH, and an

increase in the concentrations of bile components (bile salts, phosphatidylcholine and cholesterol)

associated with the colloidal species. For a solubilised system above the critical micellar

concentration (CMC), simple 1:1 v/v dilution is expected to reduce the solubilised concentration,

but to maintain total solubilisation capacity (i.e. for the drug concentration to drop by 50% but the

volume to double and therefore for solubilisation capacity to remain unchanged). However, the

data in Table 3.6 show that dilution of the micelles or vesicles with bile or buffer results in a drop

in total solubilisation capacity to only 7-41% of initial. Greater proportional decreases were

apparent for dilution of vesicles (in all cases when compared to micelles), and after dilution with

bile rather than buffer (Table 3.6). The loss of solubilisation capacity on dilution suggests the

likelihood of a phase transition to structures with reduced solubilisation capacity. Although a

complete explanation for these phase transition is not apparent at this time, for the micellar

systems, it may be related to the ability of ionised caprylic acid to self-associate and form fatty


137

acid micelles at high lipid concentration217. Thus, dilution of medium-chain colloidal phases may

reduce the concentration of caprylic acid below the CMC, leading to a loss of cinnarizine

solubilisation capacity in the micellar phase. For vesicles, previous studies have suggested that

increasing bile concentrations facilitate a vesicular to micellar transition218, 219. Since micelles are

expected to have lower solubilisation capacities for lipophilic drugs than vesicles (micelles are

smaller and less lipid-rich)92, 194, a reduction in cinnarizine solubilisation capacity might therefore

be anticipated when bile is added to model vesicles. In the case of the micelles, therefore, it seems

likely that the addition of bile disrupted the structure of swollen, mixed micelles leading to lower

colloidal lipid content and lower solubilisation capacity. Unfortunately, attempts to quantify

changes to particle size on bile addition were unsuccessful due to high polydispersity. However,

the broad trends observed were consistent with the suggestions above and decreases in particle

size were apparent for vesicles (consistent with initiation of a vesicle to micelle transition) and

increases in particle size for the mixed micelles (consistent with micellar destabilisation and

transformation to less dispersed structures) (data not shown). An additional complexity in these

dilution studies was the realisation that the pH of bile (pH 8.02) is higher than that normally

employed for simulated intestinal fluids39, 220, and therefore incubation of the colloidal phases

(prepared at pH 6.30) with bile increases system pH. Cinnarizine is a weak base with a pKa of

7.47221 and is therefore expected to be less ionised, less soluble and potentially more permeable at

higher pHs. The impact of pH in the current studies was therefore examined by dilution of

micelles and vesicles with pH-corrected bile at pH 6.30. Comparison to the data obtained with bile

at pH 8.02 suggests that the higher pH of bile provides an additional driver for drug

precipitation/supersaturation since the solubility drop was greater after incubation with bile at pH

8.02 vs. bile at pH 6.30.


138

While dilution of micelles and vesicles with bile resulted in decreases in drug solubilisation

capacity, cinnarizine did not precipitate immediately. A period of drug supersaturation was

evident for both micellar and vesicular systems, although drug precipitation from vesicles was

much more rapid than from micelles. This in turn translated into increases in drug absorption

(Figure 3.4A) and systemic exposure (Figure 3.5A) for the micellar systems. The difference in the

capacity of micelles and vesicles to maintain drug supersaturation (Figure 3.7) may be explained

by the difference in the degree of drug supersaturation induced by dilution with bile. The degree

of supersaturation is described by the supersaturation ratio, which is the ratio of the

(supersaturated) concentration of drug in solution relative to the equilibrium solubility of the drug

in the same system173. Crystallization theory suggests that the thermodynamic drivers of

precipitation from supersaturated solutions increase with increasing supersaturation ratios, as the

likelihood of nucleation and crystal growth increases with increases in (metastable) drug

concentration in solution173, 174. Here, 1:1 v/v addition of bile resulted in the attainment of

cinnarizine supersaturation ratios of 6 and 12, in micelles and vesicles, respectively. Therefore, the

faster rate of drug precipitation from vesicles (when compared to micelles) may be explained by

the higher supersaturation ratio induced by bile addition. It is also possible that micellar structures

are more effective in stabilising supersaturation when compared to vesicular structures, although

this has not been examined explicitly here. Notably, cinnarizine was found to precipitate in the

crystalline form in these experiments (Figure 3.10), precluding the possibility that the enhanced

cinnarizine absorption from micelles (when bile was co-perfused) observed in Figure 3.4A was

due to accelerated cinnarizine dissolution from precipitated amorphous forms.


139

(A)

(B)

Figure 3.10: Analysis of the cinnarizine precipitate using polarised light microscopy following

precipitation kinetics experiments where fasted rat bile was added in a 1:1 v/v ratio to cinnarizine-

loaded (A) micelles or (B) vesicles. Cinnarizine was loaded into micelles and vesicles at 80%

saturation (~ 0.2 mg/mL). Addition of bile reduced the equilibrium cinnarizine solubilisation

capacity of micelles and vesicles to 14% and 7% of initial, respectively, and triggered cinnarizine

supersaturation and precipitation (see Figure 3.7). The polarised light microscopy images show

that cinnarizine precipitated in the crystalline form after the addition of bile. The crystals obtained

from the vesicles + bile group were significantly smaller than those obtained from the micelles +

bile group, an observation that likely reflects the higher degree of supersaturation stimulated in the

vesicular system (SS ratio = 12) when compared to the micellar system (SS ratio = 6).

The role of bile in enhancing drug absorption from LBF has been reported previously222-224. In the

majority of cases, bile-mediated bioavailability enhancement has been suggested to stem from the

ability of bile to expand the solubilisation reservoir for PWSD in the GI tract. This is typically

assumed to occur via PWSD solubilisation in simple bile micelles, or via the ability of bile to

solubilise lipid digestion products and to generate more complex lipid colloidal phases with

enhanced solubilisation capacities. It seems likely that the ability to solubilise lipid digestion

products and to promote colloid formation remains an integral part of the role of bile in supporting

drug (and lipid) absorption. However, the data described here suggest that continued dilution of

lipid colloidal phases with bile in the small intestinal lumen may also lead to physical changes that


140

promote drug supersaturation, and ultimately promote drug absorption. In doing so,

supersaturation induction may be a means by which the decrease in thermodynamic activity

inherent in solubilisation is reversed, such that the free concentration of drug available for

absorption is maximised. Thus, a dual role of bile in facilitating drug absorption from LBF may be

conceived (see Figure 3.11). First, bile-mediated solubilisation of lipid digestion products at the

interface of a digesting lipid droplet results in the generation of lipid colloidal phases such as

vesicles and micelles that promote drug solubilisation during lipid digestion. Second, continued

bile-mediated dilution of existing lipid colloidal phases promotes drug supersaturation, and

enhances drug absorption by significantly increasing drug thermodynamic activity in colloidal

phases. The combination of these two highly kinetic events likely contributes to the effective drug

absorption often observed with lipid co-administration, as it affords a means to simultaneously

increase solubilisation capacity and promote thermodynamic activity of co-administered PWSD in

the small intestine.

Supersaturation induction via interaction with bile also provides a means of overcoming the

recently described solubility-permeability interplay observed in studies where PWSD are co-

administered with cosolvents, cyclodextrins or surfactant systems100, 146, 225. In these studies the

authors describe the reduction in thermodynamic activity common to most solubilisation

technologies and show that this off-sets the potential increases in membrane flux that might be

expected by an increase in solubilised drug concentration100, 146. More recent studies by the same

authors have shown that this solubility-permeability interplay can be addressed via the use of

amorphous solid dispersion formulations that stimulate supersaturation, but do not promote

solubilisation147, 226. Here we report that essentially similar outcomes are also possible with

solubilising formulations, when the solubilising formulations contain lipids and when the kinetic

changes that occur in the GI lumen in the presence of bile secretion promote supersaturation. The


141

current data therefore suggest that endogenous lipid processing pathways provide an exquisitely

sensitive and triggerable supersaturation mechanism that allows drug to remain in a solubilised

state during initial lipid digestion and at high lipid:bile concentration ratios, but that on-going bile

secretion subsequently provides a boost to thermodynamic activity and in doing so supports

enhanced absorption. This is in contrast to other common solubilisation strategies that may not

interact with the dynamic GI environment, or for which interaction with bile typically reduces

thermodynamic activity by increasing solubilisation capacity. Conversely, the ability of LBF to

promote drug solubilisation until drug supersaturation is triggered in the small intestine may

confer an advantage over formulation approaches that only utilise supersaturating strategies, as the

risk of drug precipitation in the GI tract may be reduced by solubilisation within lipid colloidal

phases.

3.6 CONCLUSION

Improved understanding of the mechanism of drug absorption from lipid colloidal phases such as

micelles and vesicles is required to provide a platform for more rational design of lipid-based

formulations. Using medium-chain lipids, we have demonstrated that the absorption of cinnarizine

(a lipophilic, poorly water-soluble drug) may be enhanced when drug supersaturation is generated

during bile-mediated dilution of lipid colloidal phases. Previous studies suggest that a similar

induction of supersaturation may occur as a result of initiation of digestion of some microemulsion

based LBFs21. These observations indicate that supersaturation, and its associated benefits in

enhancing drug absorption, may occur intrinsically during LBF incorporation into endogenous

lipid processing pathways in the small intestine. Future work will be directed toward assessing the

impact of bile dilution on supersaturation tendency, and thus absorption, for an extended range of

PWSD and in a series of different micellar and vesicular colloidal systems.


142

Figure 3.11: The dual role of bile during lipid digestion and dispersion. (i) Bile-mediated

solubilisation of lipid digestion products at the interface of a digesting oil droplet results in the

generation of lipid colloidal phases such as vesicles and micelles that maintain drug solubilisation

in the small intestine. (ii) The continuing interaction of secreted bile with existing lipid colloidal

phases in the lumen results in progressively less lipid-rich phases with lowered solubilisation

capacity. Thus, on-going bile-mediated dilution of lipid colloidal phases promotes drug

supersaturation and enhances drug absorption by increasing drug thermodynamic activity in

colloidal phases. The combination of the two highly kinetic, bile-mediated events affords a means

to simultaneously increase solubilisation capacity and promote thermodynamic activity of co-

administered drug in the small intestine, and may contribute to the increase in drug absorption

often observed with lipid co-administration. D represents the free concentration of drug available

for absorption. Dss is used to signify the increase in free concentration resulting from bile-

mediated supersaturation that drives increases in drug absorption.

Oildroplet

Common bile duct

DD

D D

D

DD

Multilamellarvesicles

Mixed micelles

Unilamellarvesicles

Simple micelles

(i) (ii)

D

DDD D

DD D

D

D

DDss DssDDss Dss

Absorption Absorption

Pancreatic lipase/co-lipase

Bile micelles


Drug

Supersaturateddrug

Small intestine

Increasing bile dispersion

(ii) Bile-mediated supersaturation(i) Bile-mediated solubilisation

143

Monash University




70%














144

Signature 1

Signature 2

145

CHAPTER 4 : THE POTENTIAL FOR

DRUG SUPERSATURATION DURING

INTESTINAL PROCESSING OF LIPID-

BASED FORMULATIONS IS ENHANCED

FOR BASIC DRUGS

Yan Yan Yeap, Natalie L. Trevaskis, Christopher J. H. Porter

Drug Delivery, Disposition and Dynamics, Monash Institute of Pharmaceutical Sciences, Monash University, 381 Royal Parade, Parkville, Victoria, 3052, Australia

Manuscript in submission.

Chapter 4: Supersaturation potential is enhanced for basic drugs

146

4.1 ABSTRACT

Co-administration of poorly water-soluble drugs (PWSD) with dietary or formulation lipids

stimulates the formation of lipid colloidal phases such as vesicular and micellar species, and

significantly expands the drug solubilisation capacity of the small intestine. The mechanism of

drug absorption from the solubilising phases, however, has not been fully elucidated. Recently, we

observed that drug supersaturation may be triggered during endogenous processing of lipid

colloidal phases containing medium-chain lipid digestion products, and that this may represent a

mechanism to enhance absorption and to reverse the reduction in thermodynamic activity inherent

in drug solubilisation. The current studies expand these preliminary findings and explore the

supersaturation tendency of five model PWSD during endogenous processing of intestinal

colloidal phases containing long-chain lipid digestion products. Bile:lipid concentration ratios

progressively increase during colloid transit through the gastrointestinal tract due to biliary

dispersion of lipid digestion products and lipid absorption. Supersaturation potential was therefore

evaluated under conditions of increasing bile and decreasing lipid concentrations and was found to

be greater for basic drugs (cinnarizine and halofantrine) than neutral (fenofibrate and danazol) or

acidic drugs (meclofenamic acid). Assessment of intestinal absorptive flux using rat jejunal

perfusion experiments subsequently showed that the absorption enhancement afforded by bile

dilution was greater for cinnarizine than danazol. The results confirm that bile may play a

significantly greater role in the absorption of weak bases such as cinnarizine from long-chain

intestinal colloids when compared to uncharged molecules such as danazol, and that this

difference reflects a greater propensity for supersaturation as intestinal colloids are dispersed and

diluted by bile. The data also suggest that co-administered digestible lipids may have a particular

utility in enhancing the absorption of poorly water-soluble weak bases.


147

4.2 INTRODUCTION

The oral bioavailability of many poorly water-soluble drugs (PWSD) is low and variable due to

slow dissolution and low solubility in the aqueous gastrointestinal (GI) contents. Lipid-based

formulations (LBF) have proven to be an effective approach in improving the oral bioavailability

of lipophilic PWSD. LBF overcome the dissolution and solubility limitations of PWSD by

presenting drug to the GI tract in a molecularly dispersed form (i.e. in solution in the formulation),

and by stimulating the generation of lipid colloidal phases that promote drug solubilisation and

reduce the risk of drug precipitation2, 76. Classically, therefore, absorption enhancement by LBF

has been considered to result from mechanisms that enhance drug dissolution and expand the

solubilisation capacity of the GI tract.

Recently, interest has increased in the potential for LBF to both promote solubilisation, and

simultaneously to stimulate supersaturation as a means of enhancing drug absorption175, 191. The

dispersion and digestion of LBF typically leads to a reduction in drug solubilisation capacity5, 91,

176, 177, 193, and where precipitation is not immediate, a period of supersaturation results where

solubilised drug concentrations are higher than the equilibrium drug solubility in an identical

colloidal milieu176, 227. Supersaturation has the potential to enhance drug absorption via transient

increases in thermodynamic activity. Conversely, however, supersaturation is also a driver of drug

nucleation and precipitation, processes that break the continuity of solubilisation and commonly

reduce drug absorption. Balancing supersaturation and precipitation is therefore a key design

element of LBF.

The digestion of triglycerides or partial glycerides in LBF, or dietary lipids, liberates amphiphilic

digestion products including fatty acids and monoglycerides. Digestion products so formed

combine with biliary components to generate a range of lipid colloidal phases in which co-


148

administered PWSD may be solubilised. At the surface of a digesting oil droplet, digestion

products accumulate and ultimately slough off to form lipid-rich dispersed liquid crystalline

phases such as lamellar, cubic and hexagonal phases with relatively large particle sizes and high

drug solubilisation capacities39-41. On dilution with intestinal fluids these colloidal structures

become smaller and less lipid-rich and typically have lower drug solubilisation capacities39. Drug

absorption is thought to occur from these smaller, more dilute lipid colloidal phases such as

unilamellar vesicles and mixed micelles, although the mechanisms involved in drug absorption

from these species have not been fully elucidated2, 76, 101.

Recently we described the ability of bile to promote drug absorption from lipid colloidal phases in

the small intestine. Contrary to traditional solubilisation paradigms, the addition of bile to

intestinal colloidal phases containing digestion products of medium-chain (MC) triglycerides

enhanced drug absorption in vivo by reducing the solubilisation capacity of the colloidal phases

for a model PWSD (cinnarizine), thereby generating a transiently supersaturated state (Chapter 3).

These initial findings were suggested to provide a means to reverse the reduction in drug

thermodynamic activity, and therefore absorptive drug flux that is inherent in solubilisation-

mediated formulation approaches100, 146.

The current study expands this preliminary evaluation of supersaturation events during the LBF

processing and has examined the impact of bile on the potential for drug supersaturation in

intestinal colloidal phases containing digestion products of long-chain (LC) triglycerides.

Importantly the work also delineates the relative sensitivities to supersaturation for weak bases,

acids and neutral compounds and shows that weak bases may be particularly suited to

supersaturation-mediated absorption enhancement from LBF.


149

4.3 METHODS

4.3.1 Materials

Cinnarizine, flunarizine dihydrochloride, fenofibrate, meclofenamic acid, progesterone,

ammonium sulphate, sodium taurocholate, sodium taurodeoxycholate, sodium glycocholate,

sodium glycochenodeoxycholate, cholesterol, L-α-lysophosphatidylcholine (LPC, from egg yolk),

oleic acid, monoolein, sodium chloride (NaCl), sodium dodecyl sulphate (SDS), formic acid and

ammonium formate were obtained from Sigma-Aldrich, Australia. Sodium

taurochenodeoxycholate, sodium glycodeoxycholate, ortho-phosphoric acid 85% (H3PO4), sodium

hydroxide pellets (NaOH), tert-butyl methyl ether (TBME), glacial acetic acid and absolute

ethanol were from Merck, Australia. Disodium hydrogen orthophosphate (Na2HPO4), sodium

dihydrogen orthophosphate (NaH2PO4.2H2O) and ammonium dihydrogen orthophosphate

(NH4H2PO4) (Ajax Finechem, Australia), oleic acid, [9,10-3H(N)] (60 Ci/mmol) (American

Radiolabeled Chemicals, MO, USA), danazol (Sterling Pharmaceuticals, Australia), halofantrine

base (GlaxoSmithKline, King of Prussia, PA, USA), heparin sodium injection BP (1000 I.U./mL,

Hospira, Australia), xylazine (100 mg/mL, Troy Laboratories, Australia), acepromazine (10

mg/mL, Ceva Delvet, Australia), ketamine (100 mg/mL, Provet, Australia) and pentobarbitone

sodium (325 mg/mL, Virbac, Australia) were obtained from listed suppliers. Acetonitrile,

methanol and chloroform used were analytical reagent grade. Water was obtained from a

Millipore milliQ Gradient A10 water purification system (Millipore, MA, USA).


Nine colloidal systems designed to mimic the colloidal phases that form in the small intestine

during the digestion of LC triglycerides were prepared. The systems consisted of triolein digestion

products (oleic acid and monoolein) solubilised in bile components (bile salts,


150

lysophosphatidylcholine and cholesterol). The concentrations of lipid and bile components in the

systems were varied systematically, as shown in Table 4.2. The equilibrium solubility of five

lipophilic PWSD (two weak bases and two neutral drugs with differing triglyceride solubilities,

and one weak acid) were determined in each of the colloidal systems, to assess drug solubility

trends as a function of colloidal bile salt concentration [bile] and colloidal lipid concentration

[lipid]. The melting points, octanol-water distribution coefficients, pKa, molecular weights,

soybean oil and oleic acid solubilities of the five drugs are shown in Table 4.3.

During bile-mediated dilution of lipid colloidal phases in the small intestine, increases in the

concentration of biliary components and decreases in lipid concentrations occur in tandem. Under

these circumstances, drug solubilisation capacity may increase (decreasing thermodynamic

activity at constant drug concentration) or decrease (invoking the potential for precipitation or

supersaturation where precipitation is not immediate). The net effect is dictated by the magnitude

of the change in solubility expected by increasing [bile] and decreasing [lipid]. Under conditions

of lipid absorption, lipids from intestinal colloids are also removed via absorption, further

decreasing lipid concentrations. In the second part of the current studies, two model drugs with

markedly different solubility relationships with [bile] and [lipid] were selected for further study.

For these compounds, absorption was assessed from a model long-chain colloidal system in situ

and in vivo, to determine the potential for endogenous lipid processing pathways to generate drug

supersaturation, and to enhance drug absorption, from intestinal colloidal phases containing LC

lipid digestion products.


151

4.3.3 Preparation of model intestinal colloidal phases containing long-chain

lipids

Colloidal systems containing triolein digestion products (oleic acid and monoolein) solubilised in

simulated endogenous intestinal fluid (SEIF) were prepared. SEIF comprised the six most

prevalent bile salts in human bile199, lysophosphatidylcholine (LPC), and cholesterol. The total

bile salt:LPC:cholesterol molar ratio was kept constant at 16:4:1, reflecting known ratios within

fasted human bile200, 201. The bile salts employed were glycocholate, glycodeoxycholate,

glycochenodeoxycholate, taurocholate, taurodeoxycholate, and taurochenodeoxycholate, at a

molar ratio of 4.0:2.8:4.0:2.0:1.2:2.0. The concentration ratios of the bile salts were chosen based

on average concentrations of the six most prevalent bile salts found in human bile199. The oleic

acid:monoolein molar ratio was kept at 2:1, reflecting the ratio of digestion products expected

from digestion of 1 mole of triolein. The concentrations of lipid and bile components in the

systems were varied systematically, according to the compositions shown in Table 4.2. Bile salt

concentrations of 4–16 mM were selected to span the typical concentration range seen in vivo76

and previously used in simulated intestinal fluids220; lipid concentrations were chosen to reflect

the solubilised concentrations that might be attained in the intestine following the digestion of LC

triglycerides91. Based on physical examination and the maintenance of consistent drug solubilising

capacities, colloids were stable for 5 days after preparation.

SEIF 16 (16 mM total bile salt:4 mM LPC:1 mM cholesterol) was prepared in 50 mL batches.

Briefly, LPC and cholesterol were dissolved in 2 mL chloroform in a round bottom flask, followed

by solvent evaporation under vacuum. The thin film formed was reconstituted with buffered bile

salt solution (consisting 16 mM total bile salt, 18 mM NaH2PO4.2H2O, 12 mM Na2HPO4, 92 mM

NaCl), vortexed for 1 min, and allowed to equilibrate at room temperature overnight. When SEIF

8 (8 mM total bile salt:2 mM LPC:0.5 mM cholesterol) and SEIF 4 (4 mM total bile salt:1 mM


152

LPC:0.25 mM cholesterol) were required, SEIF 16 was diluted 2-fold and 4-fold, respectively,

with buffer (18 mM NaH2PO4.2H2O and 12 mM Na2HPO4, 108 mM NaCl). The colloidal systems

were prepared in 10 mL batches by adding appropriate masses of oleic acid and monoolein into

SEIF, followed by a 1-min vortex. The mixture was then ultrasonicated (30 sec continuous

ultrasonication followed by pulsatile, 1 sec-on/1 sec-off ultrasonication for 5 min) with a Misonix

XL 2020 ultrasonic processor (Misonix, Farmingdale, NY, USA) equipped with a 3.2-mm

microprobe tip running at an amplitude of 240 µm and a frequency of 20 kHz. The colloids were

subsequently adjusted to pH 6.30 with H3PO4 solution and/or NaOH solution.

For the preparation of drug-loaded model colloids (for in situ jejunal perfusion and in vivo

bioavailability studies), cinnarizine was first dissolved in oleic acid at a concentration of 61 mg/g

and 115 mg/g prior to generating colloidal systems as described above, resulting in colloids

containing 65 µg/mL cinnarizine (representing ~ 40% of cinnarizine saturated solubility in the

colloidal system, solubility determined as described below) and 130 µg/mL (~ 80% saturated

solubility), respectively. Due to the lower solubility of danazol in oleic acid, colloids containing

danazol were drug-loaded by spiking 50 µL of 1.4 mg/mL or 2.8 mg/mL danazol in ethanol

solution into 10 mL blank model colloids, followed by vortexing for 1 min, and equilibration for 1

h to prepare colloids containing 7 µg/mL (~ 40% saturated solubility) and 14 µg/mL (~ 80%

saturated solubility) danazol, respectively. As a final step, trace amounts of 3H-oleic acid (0.25

µCi/mL in jejunal perfusion studies; 1 µCi/mL in bioavailability studies) were also added to the

drug-loaded colloids, followed by a 1-min vortex. The total ethanol concentration in the colloids

was ≤ 1% v/v. Drug and oleic acid concentrations in the colloidal systems were confirmed by

HPLC and scintillation counting immediately prior to use.


153

4.3.4 Equilibrium solubility of drugs in oleic acid and soybean oil

Excess solid drug (cinnarizine (CIN), halofantrine (HF), fenofibrate (FF), danazol (DAN),

meclofenamic acid (MFA)) was added to 2 mL glass vials containing approximately 0.5 g oleic

acid or soybean oil. Vials were briefly vortexed, incubated at 37 °C, and samples taken every 24 h

over a period of 120 h. During sampling, vials were centrifuged (2,200 xg, 10 min, 37 °C), and 1-2

drops of supernatant accurately weighed into 5 mL volumetric flasks using glass Pasteur pipettes.

After sampling, vials were re-vortexed such that undissolved drug particles were re-suspended in

the lipid slurry. The 5 mL volumetric flasks containing accurately weighed supernatant samples

were made up to volume with chloroform-methanol (2:1, v/v) and briefly vortexed. 50 µL aliquots

were then diluted > 10-fold with respective mobile phases (Table 4.1) and analysed for drug

content via HPLC (HPLC assay conditions for CIN, HF, FF, DAN and MFA in LC colloids are

reported in Table 4.1). Equilibrium solubility was defined when drug concentrations in

consecutive samples varied by ≤ 5%, and was determined on three separate occasions.


154

Table 4.1: HPLC assay conditions for cinnarizine (CIN), halofantrine (HF), fenofibrate (FF), danazol

(DAN) and meclofenamic acid (MFA) in long-chain colloidal systems.

Drug Stationary phase

Mobile phase Flow rate (mL/min)

Detection Retention time (min)

Validated concentration

range

CIN^ Waters Symmetry®

C18, 5 µm, 3.9 x 150 mm


v/v 20 mM NH4H2PO4

1.0

Fluorescence

λ = 249/311 nm

5.7 20–1000 ng/mL

HF Phenomenex Luna® C8(2), 5 µm, 4.6 x 250

mm

75% v/v Acetonitrile : 25% v/v H2O (0.2% w/v SDS & 0.2% v/v

acetic acid)

1.5 UV

λ = 254 nm

4.2 0.5–25 µg/mL

FF Waters XbridgeTM C18, 5 µm, 4.6 x 150

mm


v/v H2O (total 0.01% v/v formic acid)

1.0 UV

λ = 286 nm

3.9 0.5–25 µg/mL

DAN# Waters XbridgeTM C18, 5 µm, 4.6 x 150

mm

75% v/v Methanol : 25% v/v H2O

1.0 UV

λ = 286 nm

4.4

0.25–50 µg/mL

MFA Phenomenex Luna® C8(2), 5 µm, 4.6 x 250

mm

80% v/v Acetonitrile : 20% v/v 10 mM H3PO4

1.0 UV

λ = 285 nm

4.8 0.5–25 µg/mL

^ CIN in long-chain colloidal systems was assayed via HPLC as described above; while CIN in plasma was assayed via HPLC using a validated extraction procedure (with flunarizine as an internal standard) and slightly modified HPLC conditions as reported previously43.

# DAN in long-chain colloidal systems was assayed via HPLC as described above; while DAN in plasma was assayed via LC-MS using a validated precipitation procedure (with progesterone as an internal standard) and slightly modified HPLC conditions as reported previously228.


155

4.3.5 Equilibrium solubility of drugs in long-chain colloids

Excess solid drug (CIN, HF, FF, DAN, MFA) was added to 2 mL of long-chain colloids (prepared

as above) in glass vials. Vials were briefly vortexed, incubated at 37 °C, and samples taken every

24 h over a period of 120 h. During sampling, vials were centrifuged (2,200 xg, 10 min, 37 °C), 50

µL of supernatant removed, and vials re-vortexed. 50 µL aliquots were then diluted > 10-fold with

respective mobile phases (see Table 4.1) and analysed for drug content via HPLC (HPLC assay

conditions for CIN, HF, FF, DAN and MFA in LC colloids are reported in Table 4.1). Equilibrium

solubility was defined when drug concentrations in consecutive samples varied by ≤ 5%, and was

determined on three separate occasions. Equilibrium solubilities of CIN and DAN were also

determined after 1:1 v/v addition of fasted rat bile (obtained from donor animals as described

below) or buffer (18 mM NaH2PO4.2H2O, 12 mM Na2HPO4, 108 mM NaCl; adjusted to pH 6.30)

to the colloidal phases.

4.3.6 Kinetics of cinnarizine precipitation

The kinetics of cinnarizine precipitation was monitored after addition of fasted rat bile (obtained

from donor animals as described below) to model colloids, to determine whether a period of drug

supersaturation preceded drug precipitation. In a temperature (37 ºC) and stirring rate-controlled

vessel, 2.5 mL of bile was added to 2.5 mL model colloids containing 130 µg/mL cinnarizine (~

80% saturated solubility). A 1:1 v/v addition ratio of bile to colloids was used based on the known

rates of bile flow in rats in vivo (~ 1.2 ± 0.05 mL/h; mean ± SEM; n = 22) and the volume of a

typical oral dose in rats (~ 1 mL/h). Samples (100 µL) were taken before the addition of bile, and

at 1, 10, 20, 30, 40, 50, 60, 80, 100, 120 min after bile addition. Samples were immediately

centrifuged (2,200 xg, 5 min, 37 °C) to separate precipitated drug, and 50 µL of supernatant

assayed for drug content. The proportion of the initial solubilised cinnarizine concentration that

remained solubilised after bile addition was assessed as the percent of the drug mass remaining in


156

solution (i.e. the concentration in the supernatant multiplied by the volume remaining in vessel

after sampling) relative to the total drug mass in the vessel at each time point.

4.3.7 Solid-state analysis of the cinnarizine precipitate

Selected cinnarizine pellets from the precipitation kinetics experiments were analysed using a

Zeiss Axiolab microscope (Carl Zeiss, Oberkochen, Germany) equipped with crossed polarising

filters. At the end of the precipitation kinetics experiments, 1.5 mL of remaining bile + colloid

mixture were centrifuged (2,200 xg, 10 min, 37 °C), the supernatant was discarded, and a small

amount of pellet was carefully placed on a microscope slide. Samples were analysed under cross-

polarised light, and images were recorded using a Canon PowerShot A70 digital camera (Canon,

Tokyo, Japan).

4.3.8 Animals






The bile duct was cannulated near the hilum of the liver (where the duct is free of pancreatic tissue)

in order to facilitate the collection of bile fluid without contamination with exocrine pancreatic

secretions203. The bile duct and jugular vein were cannulated as described previously12, 229. Rats

were rehydrated via a continuous infusion of saline (1.5 mL/h) into a cannula inserted in the right

jugular vein, and bile continuously collected for 5 h. The concentration of total bile salt in


157

collected bile was assayed using a validated enzymatic colorimetric assay (Total Bile Acids kit

#431-15001; Wako Pure Chemical Industries, Osaka, Japan) on a plate reader (Fluostar Optima

plate reader, BMG Lab technologies, Germany) measuring absorbance at a wavelength of 540 nm.

In all subsequent experiments, bile was stored at 4°C (and allowed to warm up to ambient

temperature before use) and used within 24 h of collection.

4.3.9.2 Single-pass rat jejunal perfusion





4.3.9.3 Cinnarizine bioavailability studies following intraduodenal

administration

The surgical procedures for the conduct of bioavailability studies included cannulations of the

right carotid artery, right jugular vein, duodenum (1 cm below pylorus), and common bile duct

(only for bile-diverted rats). The surgical procedures for the cannulations have been described in

Section 2.3.2.


After surgery, animals were equilibrated for 30 min, during which time heparinised donor rat

blood was infused via the jugular vein as described previously12, 229. During the equilibration

period, blood from the cannulated mesenteric vein (~ 0.3 mL/min) was collected to enable re-

infusion via the jugular vein. Perfusion buffer was pumped through the jejunal segment at a rate of


158

0.1 mL/min and outflowing buffer discarded to waste. The exposed jejunal segment was kept

moist by covering with saline-soaked gauze throughout the experiment.

In all experiments to assess cinnarizine intestinal absorptive flux, the inflow concentration of

cinnarizine was constant at 65 µg/mL. Therefore, in experiments where model colloids were

perfused alone, cinnarizine was loaded into the perfusate at 65 µg/mL (~ 40% saturated solubility).

In experiments where model colloids were co-perfused in a 1:1 v/v ratio with a secondary

perfusate of bile, cinnarizine was loaded into the primary perfusate at 130 µg/mL (~ 80% saturated

solubility), such that 1:1 v/v dilution led to a final inflow perfusate concentration of 65 µg/mL. In

this way the concentration of cinnarizine on entry into the intestine was corrected for biliary

dilution effects, realising that the intent of the current studies was to evaluate whether bile addition

resulted in additional changes to drug solubilisation capacity and thermodynamic activity beyond

simple dilution.

Similarly, in all experiments that assessed danazol intestinal absorptive flux, the concentration of

danazol flowing into the perfused jejunum segment was held at 7.2 µg/mL. Therefore, in

experiments where model colloids were perfused alone, danazol was loaded into the perfusate at

7.2 µg/mL (~ 40% saturated solubility). In experiments where model colloids were co-perfused in

a 1:1 v/v ratio with a secondary perfusate of bile, danazol was loaded into the primary perfusate at

14.4 µg/mL (~ 80% saturated solubility), such that 1:1 v/v dilution led to a final inflow perfusate

concentration of 7.2 µg/mL.

Perfusate flow was maintained at 0.1 mL/min in all experiments to minimise variations in the

thickness of the unstirred water layer that may influence drug flux206. For experiments where 1:1

v/v co-perfusion was required, model colloids and bile were individually pumped at 0.05 mL/min,


159

and mixed via a three-way “T” connector immediately prior to entry into the jejunal segment,

providing a total perfusate flow of 0.1 mL/min. Perfusate was sampled at t = 0 to confirm drug and

3H-oleic acid concentrations. After this time, the outflowing perfusate was continuously collected

for 10 min intervals, and briefly vortexed before samples were taken for analysis of drug and 3H-

oleic acid content. For experiments where drug supersaturation was generated (i.e. cinnarizine in

model colloids + bile experiments), perfusate samples were briefly vortexed, and samples taken

before and after centrifugation (2,200 xg, 2 min) to obtain an indication of the degree of drug

precipitation within the jejunal segment. Blood draining the perfused jejunal segment was

collected at 5 min intervals, plasma separated by centrifugation (10,000 xg, 5 min), and samples

taken for analysis of drug and 3H-oleic acid concentrations.

4.3.11 Cinnarizine bioavailability after intraduodenal infusion

A 30 min-equilibration period was allowed between the end of surgery and drug dosing. To

examine the impact of bile-induced drug supersaturation on cinnarizine absorption, studies were

conducted in bile-intact or bile-diverted rats. Model colloids containing 130 µg/mL cinnarizine (~

80% saturated solubility) were infused into the duodenum of rats at a rate of 1.5 ml/h for 2 h.

Following infusion of the model colloids, saline was infused at a rate of 1.5 mL/h for 10 min to

flush remaining formulation in the tubing into the duodenum. Blood samples (0.3 mL) were

collected via the carotid artery cannula up to 8 h after infusion initiation into tubes containing 3

I.U. heparin. The sampling intervals were t = 0, 1, 1.5, 2, 3, 4, 6, 8 h. After each blood sample was

taken, the cannula was flushed with 0.3 mL of 2 I.U./mL heparinised saline to ensure cannula

patency, and to replace the volume of blood removed. Plasma was separated by centrifugation

(10,000 xg, 5 min) to enable analysis of plasma drug and 3H-oleic acid concentrations as below.


160


4.3.12.1 Sample preparation and HPLC assay conditions for CIN, HF, FF, DAN,

MFA

The HPLC assay conditions for CIN, HF, FF, DAN and MFA in LC colloids are reported in Table

4.1. LC colloid samples were prepared for HPLC assay by a minimum of 5-fold dilution with the

respective mobile phases (see Table 4.1). The injection volume for all samples was 50 µL.

Replicate analysis of n = 4 quality control samples revealed acceptable accuracy and precision (±

10%, ± 15% at the limit of quantification) for the reported concentration ranges.

CIN plasma samples were prepared for HPLC using a validated extraction procedure, with

flunarizine as an internal standard, as reported previously43. DAN plasma samples were prepared

for LC-MS using a validated precipitation procedure, with progesterone as an internal standard, as

reported previously228. The assay conditions for CIN and DAN in plasma are described

elsewhere227, 228. Replicate analysis of n = 4 quality control samples revealed acceptable accuracy

and precision (± 10%, ± 15% at the limit of quantification) for CIN plasma concentrations

between 10–320 ng/mL, and DAN plasma concentrations between 5–250 ng/mL.


Quantification of 3H-oleic acid in the perfusate and plasma was performed via scintillation

counting on a Packard Tri-Carb 2000CA liquid scintillation analyser (Packard, Meriden, CT,

USA). Perfusate samples (100 µL) and plasma samples (50 µL for in vivo studies, 200 µL for in

situ studies) were added to 2 mL Irga-safe Plus scintillation fluid followed by a 10 sec vortex.

Samples were corrected for background radioactivity by the inclusion of a blank sample in each

run.


161

4.3.12.3 Blood:plasma ratio determination for CIN, DAN, oleic acid

The blood:plasma ratio for cinnarizine, danazol, and oleic acid were determined by spiking 0.5

mL blank blood with known amounts of compound to achieve low, medium, and high

concentrations (in triplicate). Plasma was separated by centrifugation (10,000 xg, 5 min) and

plasma drug concentration assayed by HPLC (cinnarizine), LC-MS (danazol), or scintillation

counting (oleic acid). The blood:plasma ratio was calculated from the ratio of known

concentration in spiked blood to the concentration measured in plasma separated from spiked

blood. The mean blood:plasma ratio was subsequently used to convert plasma concentrations to

blood concentrations in perfusion experiments, enabling quantification of total transport into

mesenteric blood.


In the single-pass rat jejunum perfusion model, permeability coefficients were calculated from the

flux data obtained after attainment of steady state drug transport into mesenteric blood. Two

apparent permeability coefficients (Papp) were calculated as described previously184:



∆.

Equation 4.2

where ‘Disappearance’ Papp is the apparent permeability coefficient calculated from drug loss

from the perfusate (cm/sec); ‘Appearance’ Papp is the apparent permeability coefficient calculated

from drug appearance in the mesenteric blood (cm/sec); Q is the perfusate flow rate (mL/sec); A is

the surface area of the perfused jejunal segment (cm2), which is calculated by multiplying the


162

diameter by the length of the perfused intestinal segment as described previously207; C1 is the



appearance in mesenteric blood at steady state (ng/s); and <C> is the logarithmic mean drug

concentration in the lumen (ng/mL), where <C> = (C1 – C0) / (ln C1 – ln C0).

4.3.12.5 Pharmacokinetic analysis

The maximum plasma concentration (Cmax), the time to reach Cmax (Tmax) and the area under the

plasma concentration-time curve from time zero to the last measured concentration (AUC0-8 h)

were calculated using WinNonLin version 5.3 (Pharsight Inc., Apex, NC, USA).


Results were analysed using Student’s t test. A P value of < 0.05 was considered to be a

significant difference. Analyses were performed using SPSS v19 for Windows (SPSS Inc.,

Chicago, IL, USA).

4.4 RESULTS

4.4.1 Trends in drug solubility in long chain lipid-based colloids as a

function of bile and lipid concentration are different for basic, neutral and

acidic drugs

The solubilities of cinnarizine (CIN), halofantrine (HF), fenofibrate (FF), danazol (DAN) and

meclofenamic acid (MFA) in the lipid colloidal systems examined are presented in Table 4.2. The

solubility enhancement (or reduction) resulting from increasing bile concentrations (holding lipid


163

concentration constant), and solubility reduction caused by decreasing lipid concentration (holding

bile concentration constant), are shown in Figure 4.1 and Figure 4.2, respectively. Percent changes

in solubility were calculated relative to low bile concentration (i.e. 4 mM) and high lipid

concentration (i.e. 7.08 mM) in Figure 4.1 and Figure 4.2 respectively, to assess likely changes in

drug solubilisation during bile dilution (where bile concentration increases and lipid concentration

decreases in tandem) and lipid absorption (where lipid concentration decreases).

Table 4.2: Equilibrium solubility (37 °C) of cinnarizine (CIN), halofantrine (HF), fenofibrate

(FF), danazol (DAN) and meclofenamic acid (MFA) in long-chain colloidal systems.

[Lipid] #

(mM) ##

[Bile] ^

(mM) ^^

Saturated solubility (µg/mL)

CIN HF FF DAN MFA

1.77 4 49 330 13 12 85

8 19 267 15 18 90

16 13 250 21 29 130

3.54 * 4 * 163 704 37 18 62

8 84 610 32 24 86

16 35 471 37 35 157

7.08 4 357 1270 84 29 84

8 261 1276 71 36 107

16 144 1023 63 48 268

# Lipid refers to oleic acid and 1-monoolein in 2:1 molar ratio ## Concentration refers to concentration of oleic acid ^ Bile contained bile salts, lysophosphatidylcholine, cholesterol in 16:4:1 molar ratio. The bile salt mixture comprised 25 mol% sodium glycocholate, 17.5 mol% sodium glycodeoxycholate, 25 mol% sodium glycochenodeoxycholate, 12.5 mol% sodium tauroocholate, 7.5 mol% sodium taurodeoxycholate and 12.5 mol% sodium taurochenodeoxycholate

^^ Concentration refers to concentration of total bile salt

* Model long-chain colloidal system used in in situ and in vivo experiments is highlighted


164

CIN

HF

FF

DAN

MFA

Figure 4.1: Percent change in cinnarizine (CIN), halofantrine (HF), fenofibrate (FF), danazol

(DAN) and meclofenamic acid (MFA) solubility stimulated by increasing concentrations of bile at

varying lipid concentrations. Values calculated using raw data in Table 4.2. Solubility change was

calculated relative to the solubility in 4 mM bile at each lipid concentration.

The different drugs were found to have very different solubility behaviours with respect to bile

concentration in the prepared colloidal systems (Table 4.2, Figure 4.1). The solubilisation of the

weak bases (CIN and HF) was significantly reduced by increasing bile concentration (i.e. an

inverse relationship was apparent between solubility and bile concentration – as exemplified by

negative slopes in Figure 4.1); whilst the solubilisation of the neutral drugs (DAN and FF) and the

weak acid (MFA) increased in most cases in the presence of increasing bile concentrations (i.e. a

direct relationship between solubility and bile concentration - as exemplified by positive slopes in

Figure 4.1). The rank order by which bile ‘enhanced’ the solubilisation of drugs was CIN < HF <

[Bile] (mM)4 8 164 8 164 8 164 8 164 8 16

Sol

ubili

tyen

hanc

emen

t (%

)

020406080

100

4 8 164 8 164 8 164 8 164 8 160

20406080

100

4 8 164 8 164 8 164 8 164 8 160

100

200

300

4 8 164 8 164 8 164 8 164 8 160

100

200

300

4 8 164 8 164 8 164 8 164 8 160

100

200

300


165

FF < DAN < MFA, and appeared to be inversely correlated to drug solubility in oleic acid, such

that drugs that were most soluble in oleic acid (see Table 4.3) had their solubility reduced the most

by increases in bile concentration, whereas those with lower oleic acid solubility had their

solubilities increased by increases in bile concentration. The magnitude of the change was

seemingly greater for drugs with lower triglyceride (TG) solubility, for example, CIN

solubilisation was reduced to a greater extent when compared to HF; and the solubilisation of

DAN was increased to a greater extent when compared to FF.

In contrast, the solubility relationship between the drugs and colloidal lipid concentrations were

more predictable (Figure 4.2). With the exception of MFA at lower bile concentrations, the

solubility of all the drugs in the model colloids decreased with decreasing lipid concentration (i.e.

all drugs demonstrated a direct relationship between solubility and lipid concentration). The

proportional decrease in drug solubility varied significantly across the drug types, however, and

was greatest for CIN (which reduced to ~ 7% of the solubility value in colloids containing 7.08

mM lipid when the lipid content was dropped to 1.77 mM), followed by FF (which reduced to

15%), HF (21%), DAN (41%), and MFA (49%). The rank order by which decreasing lipid

concentration reduced the solubilisation of drugs was therefore CIN > HF ≈ FF > DAN > MFA,

and reflected drug solubility in oleic acid, such that drugs with the highest solubility in oleic acid

(see Table 4.3) had their solubility reduced the most by decreasing lipid (oleic acid and monoolein)

concentration.


166

CIN

HF

FF

DAN

MFA

Figure 4.2: Percent solubility reduction of cinnarizine (CIN), halofantrine (HF), fenofibrate (FF),

danazol (DAN) and meclofenamic acid (MFA) by decreasing lipid concentration, at various bile

concentrations. Values calculated using raw data in Table 4.2. Solubility reduction was relative to

the solubility in colloids containing 7.08 mM oleic acid at each bile concentration. Blue solid lines

(without symbols) represent theoretical changes in solubility upon simple dilution.

[Lipid] (mM)7.08 3.54 1.777.08 3.54 1.777.08 3.54 1.777.08 3.54 1.777.08 3.54 1.77

Sol

ubili

tyre

duct

ion

(%)

020406080

100

7.08 3.54 1.777.08 3.54 1.777.08 3.54 1.777.08 3.54 1.777.08 3.54 1.770

20406080

100

7.08 3.54 1.777.08 3.54 1.777.08 3.54 1.777.08 3.54 1.777.08 3.54 1.770

20406080

100

7.08 3.54 1.777.08 3.54 1.777.08 3.54 1.777.08 3.54 1.777.08 3.54 1.770

20406080

100

7.08 3.54 1.777.08 3.54 1.777.08 3.54 1.777.08 3.54 1.777.08 3.54 1.770

20406080

100


167

Table 4.3: Melting points, Log DpH7.5 (octanol-water distribution coefficients at pH 7.5), pKa,

molecular weights, and lipid solubilities of cinnarizine (CIN), halofantrine (HF), fenofibrate (FF),

danazol (DAN) and meclofenamic acid (MFA).

Drug Drug type Tm* (°C)

Log DpH7.5 pKa Molecular weight

Solubility in soybean oil (mmol/mol)

Solubility in oleic acid

(mmol/mol)

CIN Weak base 120$ 5.48** 1.95^^

, 7.47§§ 368.52 63.9** 264.0

HF Weak base 83# 8.85** 5.58*** 536.88 82.4** 111.9

FF Neutral 81^ 5.24$$ n/a 360.83 197.7 81.6

DAN Neutral 229$ 4.53** n/a 337.46 10.1** 4.8

MFA Weak acid 257-260§ 3.61##

3.76$$$ 296.15 19.4 7.8

n/a – not applicable

* Tm denotes melting point $ Data from Bergström et al.230 # Data from McIntosh et al.231 ^ Data from Van Speybroeck et al.232 § Data from Sanphui et al.233

** Data from Kaukonen et al.91 $$ Data quoted in Guichard et al.234 ## Calculated using Advanced Chemistry Development (ACD/Labs) Software Solaris V9.12 (1994–2006

ACD/Labs) ^^ Data quoted in Gu et al.235 §§ Data from Peeters236

*** Data from Khoo et al.237 $$$ Data quoted in Marriner et al.238


168

4.4.2 The addition of rat bile to drug-loaded model colloids reduces

cinnarizine solubility and promotes supersaturation, but increases danazol

solubilisation

Cinnarizine (CIN) and danazol (DAN) were selected as model drugs for more detailed assessment

of the impact of solubilisation properties on in situ and in vivo absorption. These compounds were

chosen since bile had opposing effects on their colloidal solubility properties (Figure 4.1,Table

4.2), and lipid concentration had a greater effect on the solubility of CIN when compared to DAN

(Figure 4.2,Table 4.2). To explore the impact of dilution with bile on CIN and DAN solubilisation

(rather than dilution with model intestinal fluids in Figure 4.1), donor bile was collected from rats,

and the change in CIN and DAN solubilisation capacity in model colloids assessed before and

after 1:1 v/v addition of rat bile (Figure 4.3A). 1:1 v/v addition of donor bile decreased CIN

solubility from 157.7 µg/mL to 9.2 µg/mL and therefore, as a proportion of initial, reduced CIN

solubilisation capacity from 100% to 12% (taking into account the 2-fold increase in volume after

the addition of bile). In contrast, for DAN, combination of the model colloids with bile in a 1:1 v/v

ratio increased DAN solubility from 17.6 µg/mL to 22.6 µg/mL, and therefore increased DAN

solubilisation capacity from 100% to 257% (again taking into account a doubling in volume on

bile addition). Dilution of the colloids with the same volume of buffer did not result in significant

changes in the solubilisation capacity of either DAN or CIN (Figure 4.3A). As such, the changes

in solubilisation capacity observed on addition of bile were not simply due to a volume/dilution

effect (i.e. decreasing colloidal lipid concentration), but instead appear to reflect increases in bile

concentration, since the average total bile salt concentration in donor bile was 15.0 ± 1.0 mM

(mean ± SEM, n = 4) whereas the bile concentration in the model colloidal system was 4 mM.

The solubility of CIN therefore decreased with increasing bile concentration, whereas the

solubility of DAN increased with increasing bile concentration. This trend with donor rat bile was


169

consistent with the in vitro data obtained using SEIF (Figure 4.1), and suggests that CIN may

supersaturate when drug-loaded model colloids interact with secreted bile in vivo.

To assess the kinetics of drug precipitation on interaction with bile, CIN-loaded model colloids

(CIN loaded at 130 µg/mL, ~ 80% saturated solubility) were incubated 1:1 v/v with donor rat bile

in a temperature (37 °C) and stirring rate-controlled vessel, and samples taken over time to

evaluate whether precipitation occurred. Figure 4.3B shows that bile addition did not result in

immediate drug precipitation, and instead was preceded by a period of supersaturation that lasted

for at least 80 min in all cases. Since DAN solubility increased rather than decreased on bile

addition, supersaturation/precipitation did not occur and therefore precipitation kinetics were not

evaluated.

(A)

% o

rigin

al d

rug

solu

bilis

atio

n ca

paci

ty

0

100

200

300

400

Model colloidModel colloid + Buffer (1:1)Model colloid + Bile (1:1)

Cinnarizine Danazol

100 100 100

12

257

94

*

*

(B)

Figure 4.3: (A) Percent original solubilisation capacity of cinnarizine (CIN) and danazol (DAN),

before and after a 1:1 v/v addition of buffer or rat bile to model colloids. After the addition of rat

bile, solubility of CIN decreased from 157.7 ± 3.0 µg/mL to 9.2 ± 0.7 µg/mL, while solubility of

DAN increased slightly from 17.6 ± 0.5 µg/mL to 22.6 ± 2.5 µg/mL. After the 1:1 v/v addition of

buffer, solubility of CIN and DAN was approximately halved, from 157.7 ± 3.0 µg/mL to 73.8 ±

1.0 µg/mL, and 17.6 ± 0.5 µg/mL to 8.8 ± 0.2 µg/mL, respectively. Percent original solubilisation

capacity is the solubilisation capacity of the colloids after bile addition (taking into account

Time (min)0 20 40 60 80 100 120%

orig

inal

CIN

in s

olut

ion

0

20

40

60

80

100

120

Add bile

Equilibrium solubility: Model colloid + Bile (1:1)


170

volume increases as a result of dilution) relative to the solubilisation capacity of the colloids prior

to bile addition. Data represent mean ± SEM of n = 3 determinations. Significant difference (p <

0.05) to colloid only is denoted by the symbol *. (B) Kinetics of CIN precipitation from model

colloids following 1:1 v/v addition of rat bile at t = 0. CIN was loaded into colloids at 80%

saturation (~130 µg/mL). Each line represents individual experiments. The addition of bile to

drug-loaded colloids resulted in CIN supersaturation that was maintained for at least 80 min in all

experiments.

4.4.3 Co-perfusion of bile with model colloids increases cinnarizine

absorption to a greater extent than danazol

The perfusate disappearance and mesenteric blood appearance profiles of cinnarizine (CIN),

danazol (DAN) and oleic acid after perfusion of drug in LC lipid-loaded model colloids through

an isolated jejunal segment (~ 10 cm2), with and without 1:1 v/v co-perfusion with donor rat bile,

are shown in Figure 4.4. Steady state-absorptive flux, disappearance Papp, and appearance Papp

from all perfusion experiments are reported in Table 4.4.

Co-perfusion of model colloids with bile increased the absorptive flux, disappearance Papp, and

appearance Papp of CIN 3.2-fold, 3.0-fold, and 4.2-fold, respectively. Throughout the perfusion

experiments, bile-induced supersaturation appeared to be stable, consistent with the in vitro

studies (Figure 4.3B), since the CIN perfusate concentration was unchanged before or after

centrifugation, therefore indicating a lack of CIN precipitation during co-perfusion with bile

(Figure 4.4AI).

Co-perfusion of model colloids with bile was expected to lead to an increase in DAN

solubilisation capacity within the perfused jejunal segment and therefore to not stimulate DAN

supersaturation or an increase in thermodynamic activity. In spite of this, however, co-perfusion of


171

model colloids with bile led to a moderate (1.7-fold) increase in both the absorptive flux and

appearance Papp of DAN (disappearance Papp was unchanged) (Figure 4.4II and Table 4.4). The

increase in absorptive flux in the presence of bile was, however, notably higher for CIN than DAN

(3.2-fold increase in CIN vs.1.7-fold increase in DAN).

In assessing the permeability of CIN and DAN, the concentration of the drug in the perfusate was

kept constant (65 µg/mL and 7 µg/mL, respectively), allowing direct comparisons of absorptive

flux in the absence and presence of bile. For oleic acid, however, the concentration in the

perfusate was reduced by 50% due to bile co-perfusion. Therefore, direct comparisons of absolute

oleic acid absorptive flux was not possible, and transport into the mesenteric blood is represented

instead as % dose transported (Figure 4.4BIII). Nonetheless, bile co-perfusion led to an increase in

the proportion of oleic acid absorbed (Figure 4.4BIII), and an increase in the appearance Papp of

oleic acid (Table 4.4).


172

(AI)

Time (min)

0 30 40 50 60 70

% C

IN d

ose

pass

ing

thro

ugh

jeju

num

40

60

80

100

Model colloids Model colloids + Bile (pre-centrifugation)SS

Model colloids + Bile (post-centrifugation)SS

(BI)

Time (min)

0 10 20 30 40 50 60 70

CIN

flu

x in

to m

esen

teric

bl

ood

(ng/

5 m

in/1

0 cm

2 )

0

100

200

300

400

500

600Model colloids (CIN 65 µg/mL)Model colloids + Bile (CIN 65 µg/mL)SS

(AII)

(BII)

Time (min)

0 10 20 30 40 50 60 70

DA

N f

lux

into

mes

ente

ricbl

ood

(ng/

5 m

in/1

0 cm

2 )

0

50

100

150

200

250

300Model colloids (DAN 7 µg/mL)Model colloids + Bile (DAN 7 µg/mL)

(AIII)

(BIII)

Time (min)

0 10 20 30 40 50 60 70% O

A d

ose

into

mes

ente

ric

bloo

d (%

/5 m

in/1

0 cm

2 )

0

1

2

3

4

5Model colloids (OA 1 µg/mL)Model colloids + Bile (OA 0.5 µg/mL)

Figure 4.4: (A) Perfusate disappearance (% dose passing through jejunum) and (B) mesenteric

blood appearance of (I) cinnarizine (CIN), (II) danazol (DAN), and (III) oleic acid (OA) when

model colloids were perfused through an isolated rat jejunal segment (~ 10 cm2), with (open

symbols) and without (closed symbols) 1:1 v/v co-perfusion with donor rat bile. Appearance

profiles are plotted as absorptive flux (ng/5 min/10 cm2) for CIN and DAN, and % dose passing

Time (min)

0 30 40 50 60 70

% D

AN

do

se p

ass

ing

thro

ug

h je

jun

um

40

60

80

100

Model colloids Model colloids + Bile

Time (min)

0 30 40 50 60 70

% 3

H-O

A d

ose

pa

ssin

gth

rou

gh

jeju

nu

m

40

60

80

100

Model colloidsModel colloids + Bile (pre-centrifugation)Model colloids + Bile (post-centrifugation)


173

through jejunum for OA. Co-perfusion of model colloids with bile generates supersaturation in

situ within the perfused jejunal segment for CIN but not DAN. It is not known if co-perfusion of

colloids with bile generates OA supersaturation within the perfused jejunal segment. SS denotes

experiments where drug is supersaturated in perfusate. Degree of CIN and OA (if any)

precipitation within the perfusate is illustrated in (AI) and (AIII) as the difference in perfusate

concentration pre- and post-centrifugation. Experiments were performed using an in situ single-

pass rat jejunum perfusion model. In all experiments, the total perfusate flow rate was kept

constant at 0.1 mL/min. Data represent mean ± SEM of n = 3-4 rats.


174

Table 4.4: Cinnarizine (CIN), danazol (DAN) and oleic acid (OA) disappearance Papp (x 106 cm/s)

from the intestinal perfusate, appearance Papp (x 106 cm/s) in the mesenteric blood, and steady

state absorptive flux into mesenteric blood (ng/5 min/10 cm2) after 70 min of single-pass perfusion

of ~ 10 cm2 segments of rat jejunum with model colloids (composition reported in Table 4.2), with

and without 1:1 v/v co-perfusion with donor rat bile. Values calculated using data obtained after

steady state attainment (t = 55-70 min). In all experiments, total perfusate flow rate was kept

constant at 0.1 mL/min. Data represent mean ± SEM of n = 3-4 rats.

Perfusate Conc. perfused (µg/mL)

SS ratio #^

Disappearance Papp (x

106 cm/s)

Appearance Papp

(x 106 cm/s) Flux into

mesenteric blood (ng/5 min/10 cm2)

CIN Model colloid only

65.0 0.4 12.2 ± 2.8 0.6 ± 0.2 138 ± 28

Model colloid + Bile SS

65.0 7.1 36.9 ± 5.5 a 2.5 ± 0.2 a 443 ± 34 a

DAN Model colloid only

7.2 0.4 27.7 ± 6.8 3.8 ± 0.1 75 ± 2

Model colloid + Bile

7.2 0.3 29.4 ± 3.0 6.6 ± 0.8 a 131 ± 15 a

OA Model colloid only

1.0 n/a 18.0 ± 5.9 2.0 ± 0.3 6 ± 1

Model colloid + Bile

0.5 n/a 26.1 ± 3.8 6.1 ± 0.8 a 9 ± 1

a Significant difference from model colloid only SS denotes drug supersaturation in perfusate # SS ratio = Supersaturation ratio = (Supersaturated) concentration of drug in perfusate / Equilibrium

solubility of drug in perfusate ^ SS ratio not reported for OA because equilibrium solubility of OA in perfusate was not determined


175

4.4.4 Supersaturation increases in vivo cinnarizine exposure after

intraduodenal infusion of drug-loaded model colloids

The systemic plasma concentration-time profiles for cinnarizine (CIN) and 3H-oleic acid

following intraduodenal infusion of CIN-loaded model colloids (130 µg/mL CIN; ~ 80% saturated

solubility) to bile-intact and bile-diverted rats are shown in Figure 4.5, and the pharmacokinetic

parameters reported in Table 4.5. Consistent with the results from the rat jejunal permeability

studies, intraduodenal administration of drug-loaded model colloids to bile-intact rats resulted in

significantly higher systemic plasma concentrations of CIN at t = 1.5, 2, 3 and 4 h, and

significantly higher AUC0-8 h (2.0-fold increase) when compared to bile-diverted rats. The

systemic plasma concentrations of 3H-oleic acid were also significantly higher in bile-intact rats at

all sample time points.

Bioavailability studies for danazol (DAN) are not reported here because the dose administered

(0.14 mg/kg – corresponding to 80% saturated solubility) did not lead to detectable systemic

plasma concentrations in both bile-intact and bile-diverted rats at all sample time points (limit of

quantification for DAN plasma assay was 5 ng/mL).


176

(A)

Time (h)

0 2 4 6 8

Pla

sma

CIN

con

c.

(ng/

mL)

020406080

100120140

Bile-divertedBile-intact*

*

*

*

(B)

Figure 4.5: Systemic plasma concentration-time profiles of (A) cinnarizine (CIN) and (B) oleic

acid (OA) following a 2-hour intraduodenal infusion of 3 mL model colloids containing 130

µg/mL CIN (~ 80% saturated solubility) and 1 µCi/mL OA to bile-intact and bile-diverted rats.

Consistent with observations in rat jejunum perfusion studies, bile-induced supersaturation

translated into increased in vivo CIN exposure. Data represent mean ± SEM of n = 3 rats.

Statistical significance (p < 0.05) is denoted by the symbol *.

Table 4.5: Pharmacokinetic parameters for cinnarizine (CIN) after intraduodenal administration

of model colloids (composition reported in Table 4.2) loaded with 1.3 mg/kg CIN (~ 80%

saturated solubility) to bile-intact and bile-diverted rats. The interaction between the colloids and

secreted bile in the duodenum is expected to generate CIN supersaturation in bile-intact animals,

whilst CIN is expected to be sub-saturated in bile-diverted animals. Values represent mean ± SEM

of n = 3 rats.

Compound Experimental group

AUC0-8 h

(ng h/mL) Cmax

(ng/mL) Tmax

(h)

Cinnarizine Bile-intact 403 ± 3 a 107.1 ± 10.8 a 1.8 ± 0.2

Bile-diverted 190 ± 17 53.0 ± 8.8 1.7 ± 0.3

a Significant difference when compared to bile-diverted group

Time (h)

0 2 4 6 8

Pla

sma

3H

-OA

con

c.(d

pm/m

L)

0

20000

40000

60000Bile-divertedBile-intact

**

**

** *


177

4.5 DISCUSSION

Previously, we described the ability of bile secretion to trigger drug supersaturation and to

enhance the absorption of a model PWSD (cinnarizine) from intestinal colloidal phases containing

digestion products of medium-chain triglycerides (Chapter 3). In the current study, we aimed to

extend these initial observations to include assessment of colloidal phases containing long-chain

lipids (to evaluate the dependence on lipid chain length) and PWSD with differing

physicochemical properties (to assess the importance of acid/base functionality, lipid solubility

and octanol-water distribution coefficient). The supersaturation tendency of the different PWSD

during endogenous processing of post-digestion lipid colloidal phases was predicted by assessing

drug solubility relationships with increasing bile concentration and decreasing lipid concentration;

and the absorption of drugs with different supersaturation tendencies was evaluated in situ and in

vivo. The data suggest that under the conditions explored, bile dilution of, and lipid

removal/absorption from, LC colloids may induce supersaturation and enhance the absorption of

PWSD, and that the potential for drug supersaturation during lipid processing is more pronounced

for basic drugs.

The solubility of lipophilic PWSD in lipid-containing intestinal colloids is a complex function of

the solid state properties of the drug (usually captured by melting point), and drug affinity for core

lipids, the colloid interface and water, and is not easily predicted de novo. For example, the lipid

solubility of FF far exceeds that of the similarly lipophilic (by log P) DAN (Table 4.3),

presumably due to differences in melting point. However, both drugs had comparable solubility in

the colloidal systems examined here (Table 4.2). In contrast, although CIN and HF have similar

solubility properties in lipids, the solubility of HF in the model colloids was up to an order of

magnitude higher than CIN, suggesting specific affinity of HF for the colloidal structures rather

than their component parts. MFA has low solubility in lipids but had moderate solubility in the


178

prepared colloids, presumably due to ionisation (and higher water solubility) at the pH of model

intestinal fluids.

Examination of the impact of bile concentration on the drug solubilisation capacities of the long-

chain colloidal systems revealed widely varying solubility trends for the drugs evaluated (Figure

4.1 and Table 4.2). Traditionally it has been assumed that bile secretion promotes drug

solubilisation (and in this way promotes drug absorption), and indeed increases in drug solubility

were evident with increasing bile concentration for the neutral (DAN, FF) and acidic (MFA) drugs.

In contrast, however, and in large part counter-intuitively, the solubility of the basic drugs (CIN,

HF) in the long-chain colloids decreased with increasing bile concentration. The different

relationships between the colloid solubility of the basic, neutral and acidic drugs and bile

concentration was not expected, but may reflect changes in the structure of the colloidal phases on

bile addition, as well as differences in the preferred drug solubilisation site. Under digesting

conditions in the small intestine, many amphiphilic molecules (e.g. bile salts, lyso-phospholipids,

fatty acids, monoglycerides) co-exist, and the colloidal species formed display complex phase

behaviour that is dependent on the type and concentration of the components present199, 239, 240. For

example, raising the concentration of endogenous biliary components (e.g. bile salts) relative to

the concentration of lipid digestion products has been shown to stimulate phase changes from

larger lipid rich species such as cubic, hexagonal and multilamellar liquid crystals to smaller, more

highly dispersed species such as unilamellar vesicles and mixed micelles39, 40. Continued phase

changes from unilamellar vesicles to mixed micelles on addition of higher concentrations of bile

salts are also well-documented218, 241, 242. In view of the dynamic nature of the structure of

intestinal lipid colloidal species, drug solubilisation capacity may also change with the addition or

removal of amphiphilic components, since different phases possess different solubilisation

capacities for PWSD194, 243.


179

Solubility changes as a function of microstructure have been described for many poorly water-

soluble compounds in larger, microemulsion-based colloidal systems244-246. The dependency of

compound solubility on microstructure is thought to be related to the preferred solubilisation locus

within the microstructure, and drugs that preferentially reside at an interface may have increased

solubility when the interfacial area of colloidal systems is increased247. It is thus possible that in

the colloidal systems examined here, differences in the solubility behaviour of the

basic/neutral/acidic drugs reflect differences in preferred solubilisation sites within the colloids.

CIN and HF have higher solubilities in oleic acid (presumably due to favourable electrostatic

interactions in lipid solution), and are therefore more likely to co-localise with oleic acid in the

hydrophobic core of a micelle or in the lipid bilayer of a vesicle when compared to DAN and

MFA, which in turn may be more highly solubilised at the amphiphilic interface. Under these

circumstances, the addition of bile constituents to the colloids (which disperses the lipids,

decreasing the proportional volume of the lipid core and increasing the colloid interfacial area)

may lead to increased DAN and MFA solubilisation, but reduced solubilisation of CIN and HF.

It is also possible that within the lipophilic sub-compartments of the dispersed colloidal structures,

molecules of oleic acid and the weak bases orientate in such a way that the hydrophobic aspects of

each molecule remain dissolved in the lipid core, whereas the charged species orientate towards

the interface where the cationic weak bases and the carboxylic acid head groups of oleic acid are

free to favourably interact. Under these circumstances increased concentrations of bile acids may

also disrupt intermolecular interactions between oleic acid and weak bases at the colloidal

interface, further lowering drug solubilisation capacity.

Reduced solubilisation of the weak bases with increasing bile concentration may also occur via the

formation of insoluble drug-bile salt complexes between the cationic weak bases and anionic bile


180

salts. To investigate the potential importance of the effects of bile on interactions between the

weak bases and oleic acid vs. drug interaction or complexation with bile salt, a series of colloidal

systems analogous to those in Table 4.2 (i.e. containing bile components and monoolein) but

without the inclusion of oleic acid were prepared, and CIN solubility assessed. As expected, the

solubility of CIN in the oleic acid-free colloidal systems (Table 4.6) was substantially lower than

the corresponding oleic acid-containing colloidal systems (Table 4.2), illustrating the high

dependency of the colloid solubility of the weak bases on the presence of oleic acid. More

significantly, in the oleic acid-free systems, CIN solubility increased (rather than decreased) with

increasing bile concentration (Table 4.6). The stark difference in CIN solubility behaviour in oleic

acid-free vs. oleic acid-containing colloidal systems suggests that the effect of bile in reducing

CIN solubilisation in bile salt mixed micelles is oleic acid dependent, and that the reduction in

CIN solubility was not due to bile salt-drug complexation, but rather bile-mediated disruption of

intermolecular interactions between oleic acid and the weak bases.


181

Table 4.6: Equilibrium solubility (37 °C) of cinnarizine (CIN) in colloidal systems consisting 1-

monoolein solubilised in bile components. Colloidal systems presented in this table correspond to

those in Table 4.2, but without the inclusion of oleic acid.

[1-Monoolein] (mM) [Bile]^ (mM)^^ CIN saturated solubility (µg/mL)

0.88 4 5

8 7

16 13

1.77 4 8

8 10

16 16

3.54 4 16

8 16

16 20

^ Bile refers to total bile salt, lysophosphatidylcholine, cholesterol in 16:4:1 molar ratio. Total bile salt consisted of 25 mol% sodium glycocholate, 17.5 mol% sodium glycodeoxycholate, 25 mol% sodium glycochenodeoxycholate, 12.5 mol% sodium tauroocholate, 7.5 mol% sodium taurodeoxycholate, 12.5 mol% sodium taurochenodeoxycholate

^^ Concentration refers to concentration of total bile salt

Endogenous lipid processing pathways that lead to decreases in the drug solubilisation capacity of

the intestinal colloidal phases may lead to drug supersaturation (provided that precipitation is not

instantaneous). Since the bile:lipid concentration ratio progressively increases on passage down

the GI tract, the tendency for a drug to supersaturate may be predicted by the relationship between

drug solubility and increasing bile/decreasing lipid concentrations. Thus, during bile dilution of

lipid colloidal phases, drugs that have an inverse solubility relationship with bile concentration are

expected to supersaturate, and the higher the proportional decrease in solubilisation (caused by

increasing bile concentration), the higher the supersaturation tendency. Although bile dilution also


182

leads to a decrease in colloidal lipid concentration, the percent solubility reduction vs. [lipid] plots

(Figure 4.2) suggest that drug solubility decreases no more than, and in most cases less than, the

dilution factor (solubility plots mostly overlap or are above the dilution line). As such, the overall

solubilisation capacity of the intestinal colloidal phases (with respect to a decrease in lipid

concentration) is likely to remain unchanged or even increase. Supersaturation tendency due

simply to bile dilution is therefore expected to be less dependent on changes in lipid concentration

and more dependent on changes in bile concentration. During lipid absorption, however, colloidal

structures are further depleted of lipid content (in the absence of volume changes). In this case

solubilisation capacity is expected to drop (and therefore supersaturation tendency to increase) and

this would be greatest for drugs with the highest solubility dependency on lipid concentration.

Taking into consideration the proportional changes in drug solubility caused by increasing bile

concentration and decreasing lipid concentration, the overall drug supersaturation tendency

resulting from both bile dilution of and lipid absorption from lipid colloidal phases follows the

rank order CIN > HF > FF > DAN > MFA, i.e. supersaturation tendency was greatest for basic

drugs followed by neutral drugs and acidic drugs.

Indeed, when CIN-loaded colloids were mixed with endogenous rat bile (to simulate the process

of bile dilution), CIN solubilisation capacity was reduced significantly, and the period of drug

supersaturation that preceded precipitation translated into enhanced drug absorption (Figure 4.4BI)

and enhanced systemic exposure (Figure 4.5A). Notably, cinnarizine was found to precipitate in

the crystalline form in these experiments (Figure 4.6), precluding the possibility that the enhanced

cinnarizine absorption from model colloids (when bile was co-perfused) observed in Figure 4.4BI

was due to accelerated cinnarizine dissolution from precipitated amorphous forms. In addition,

during bile co-perfusion, the proportion of oleic acid absorbed was increased (Figure 4.4BIII).

Since CIN solubilisation capacity is directly related to colloidal lipid content, the increased


183

absorption of oleic acid (when bile was co-perfused) was expected to magnify the decrease in drug

solubilisation capacity, and further increase CIN supersaturation potential and absorption.

Figure 4.6: Analysis of the cinnarizine precipitate using polarised light microscopy following

precipitation kinetics experiments where fasted rat bile was added in a 1:1 v/v ratio to cinnarizine-

loaded model colloids (cinnarizine was loaded into model colloids at 80% saturation (~ 130

µg/mL). Addition of bile reduced the equilibrium cinnarizine solubilisation capacity of model

colloids to 12% of initial, and triggered cinnarizine supersaturation and precipitation (see Figure

4.3). The polarised light microscopy image show that cinnarizine precipitated in the crystalline

form after the addition of bile.

When similar studies were repeated with DAN (which has significantly lower absolute solubility

in lipid digestion products and lower supersaturation potential based on the impact of increases in

bile concentration and decreases in lipid concentration on DAN solubility; Table 4.3, Figure 4.1,

Figure 4.2), absorptive flux also increased when model colloids were co-perfused with bile, but to

a much lower extent than CIN (1.7-fold, compared to 3.2-fold for CIN). The effect of bile in

enhancing DAN absorption most likely reflected differences in oleic acid absorption in the

presence of bile (Figure 4.4BIII). Thus, bile increased oleic acid absorption, and in turn decreased

DAN solubilisation capacity in the lipid depleted mixed micelles, resulting in an increase in DAN


184

thermodynamic activity at the absorptive membrane. This occurred despite a slight decrease in

DAN thermodynamic activity in the bulk lumen due to bile addition (via an increase in drug

solubilisation capacity due to bile). Other potential explanations for the change in DAN absorption

include an increase in flux resulting from an increase in permeability due to bile-induced

membrane damage248, 249 or bile-induced changes to pH (the average pH of donor bile used was

8.02 ± 0.02 (mean ± SEM, n = 5)). These factors were, however, excluded in previous studies

using similar experimental methods and colloidal systems (Chapter 3). The presence of

phospholipids in endogenous bile is also expected to reduce the toxicity (and permeability

enhancement effects) of bile salts by decreasing bile salt thermodynamic activity in mixed bile

salt-phospholipid micelles108. The permeability of DAN, a non-ionisable drug, is not expected to

be affected by changes to pH.

Differences in the extent to which bile enhanced the absorption of CIN and DAN may therefore be

attributed to differences in the supersaturation tendencies of the two drugs during bile dilution and

lipid absorption. In the case of CIN, increases in thermodynamic activity were triggered by both

bile dilution and lipid absorption; whereas in the case of DAN, the increase in thermodynamic

activity was triggered only during lipid absorption. These observations suggest that bile-induced

supersaturation may be a phenomenon that is specific to basic drugs; while lipid absorption-

induced supersaturation is expected to apply to all lipophilic drugs, and the degree of

supersaturation induced is likely to be greatest for basic drugs (which had higher solubility

dependency on the concentration of lipid digestion products within bile micelles).

Thus, the in vitro and in vivo observations in this study suggest that drugs with high solubility in

lipid digestion products (such as weak bases in fatty acid-containing colloids) may better harness

the supersaturation-generating potential of endogenous lipid processing pathways. This reflects

three separate, but sequential, processes (see Figure 4.7). First, strong affinity for solubilised lipid


185

digestion products may enable favourable drug partitioning from digesting triglyceride droplets

into post-digestion lipid colloidal phases250. Subsequently, drug is preferentially solubilised in the

core of lipid colloidal phases at high concentration and is therefore predisposed to bile-induced

supersaturation in the event of a drop in solubilisation capacity. Finally, lipid absorption (i.e. lipid

removal from colloids) exaggerates the loss of drug solubilisation capacity and therefore

encourages supersaturation if precipitation is not immediate. Drug affinity for lipid digestion

products may therefore be a useful indicator of supersaturation potential during LBF processing in

the GI tract.

In summary, we previously described a dual role for bile in facilitating drug absorption after co-

administration with medium-chain formulation lipids, where bile-mediated solubilisation of lipid

digestion products and bile-mediated dilution of existing lipid colloidal phases provides a means

to simultaneously increase solubilisation capacity and promote the thermodynamic activity of a

co-administered drug in the small intestine (Chapter 3). The data described here further establish

that endogenous processing of post-digestion lipid colloidal phases (via bile dilution and lipid

absorption) may generate drug supersaturation after co-administration of long-chain lipids; and

that the potential for supersaturation generation and ensuing absorption enhancement is greatest

for basic drugs. The unique manner by which lipids promote both drug solubilisation and drug

supersaturation likely contributes to effective drug absorption after lipid co-administration, and

has implication both for LBF design and better understanding the effects of lipids in food on drug

absorption. The sequential manner (see Figure 4.7) by which supersaturation is triggered during

endogenous lipid processing (e.g. during lipid dispersion and digestion176, 227, as a result of on-

going bile dilution, and in response to lipid absorption), combined with the presence of drug

solubilising lipid colloidal phases in the small intestine that ‘buffer’ against the generation of

excessive supersaturation, dictates that the rate and extent of supersaturation generated by LBF in


186

vivo is highly controlled, and has the potential to maximise absorption-enhancement and reduce

the risk of drug precipitation.

4.6 CONCLUSION

Supersaturation-triggered enhancements in thermodynamic activity may represent an important

mechanism by which solubilised drug is made available for absorption from the lipid colloidal

phases formed in the intestine following digestion of dietary or formulation lipids. In this study,

we investigated drug supersaturation during bile-mediated dilution of intestinal colloid phases

containing digestion products of long-chain triglycerides. The increase in colloidal concentration

of biliary solubilisers upon dilution with bile was a highly effective trigger for supersaturation for

basic drugs, whereas for acidic or neutral drugs, bile-mediated enhancements in lipid absorption

and therefore decreases in colloidal lipid content appeared to provide the required reduction in

solubilisation capacity to promote supersaturation. The data suggest that supersaturation tendency

during endogenous processing of intestinal colloids is greatest for basic when compared to

neutral/acidic drugs and likely reflects the high affinity of weak bases for colloidal systems

containing oppositely charged fatty acids. Studies to evaluate the importance of intestinal colloids

containing natural glyceride digestion products (i.e. fatty acids rather than systems based on

uncharged lipids or synthetic solubilisers) in maximising the absorption of co-administered drugs

are on-going.


187

Figure 4.7: Endogenous lipid processing pathways that facilitate both drug solubilisation and drug

supersaturation in the small intestine. (i) Bile-mediated solubilisation of lipid digestion products at the

interface of a digesting oil droplet results in the generation of lipid colloidal phases such as vesicles and

micelles that promote drug solubilisation (ii) Digestion and dispersion of lipid based formulations (LBF)

may reduce drug solubilisation capacity and trigger drug supersaturation176, 227 (iii) The continuing

interaction of secreted bile and existing lipid colloidal phases in the lumen may reduce the solubilisation

capacity of basic drugs and trigger drug supersaturation (iv) The removal of colloidal lipid components via

lipid absorption reduces drug solubilisation capacity and also triggers drug supersaturation at the absorption

site. Drugs that have high affinity for lipid digestion products (e.g. basic drugs such as cinnarizine and

halofantrine) may be more prone to supersaturation events as they preferentially partition into post-

digestion phases (instead of remaining in a digesting TG droplet), are predisposed to bile-induced

supersaturation, and show exaggerated losses in drug solubilisation capacity (i.e. greater supersaturation

potential) during lipid absorption. The sequential manner by which supersaturation is triggered during

endogenous LBF processing (ii, iii, iv), combined with the supersaturation ‘buffering’ properties of lipid

colloidal phases (via increases in intestinal solubilisation capacity), dictate that the rate and extent of

supersaturation generated by LBF in vivo is intrinsically controlled, thus maximising absorption-

enhancement potential of supersaturation and reducing the risk of drug precipitation. The unique ability of

LBF components to simultaneously increase solubilisation capacity and promote thermodynamic activity of

co-administered drug in the small intestine may contribute to effective drug absorption often observed with

LBF co-administration.

Oildroplet

Common bile duct

DD

D D

D

DD

Multilamellarvesicles

Mixed micelles

Unilamellarvesicles

DD

D

DD

AbsorptionAbsorption

Pancreatic lipase/co‐lipase

Bile micelles


Drugmolecule

Small intestine

(i) Bile‐mediated solubilisation(ii) LBF digestion and dispersion‐induced supersaturation(iii) Bile dilution‐induced supersaturation(iv) Lipid absorption‐induced supersaturation

DD

DSS DSS

Absorption

Enterocytes

Supersaturated drug

(ii)

(i) (iii)(iv)

DDSS DSS

(iii) (iii)

DDSS DSS DDSS DSS

DLipid absorption

Lipid absorption

DD

DD

DDD


188

Monash University




70%















189

Signature 1

Signature 2

190

CHAPTER 5 : LIPID ABSORPTION

TRIGGERS DRUG SUPERSATURATION

AT THE INTESTINAL UNSTIRRED WATER

LAYER AND PROMOTES DRUG

ABSORPTION FROM MIXED MICELLES

Yan Yan Yeap, Natalie L. Trevaskis, Christopher J. H. Porter

Drug Delivery, Disposition and Dynamics, Monash Institute of Pharmaceutical Sciences, Monash University, 381 Royal Parade, Parkville, Victoria, 3052, Australia

Manuscript in submission.

Chapter 5: Lipid absorption triggers drug supersaturation

191

5.1 ABSTRACT

Purpose. To evaluate the potential for the acidic intestinal unstirred water layer (UWL) to induce

drug supersaturation and enhance drug absorption from intestinal mixed micelles, via the

promotion of fatty acid absorption. Methods. Using a single-pass rat jejunal perfusion model, the

absorptive-flux of cinnarizine and 3H-oleic acid from oleic acid-containing intestinal mixed

micelles was assessed under normal acidic microclimate conditions and conditions where the

acidic microclimate was attenuated via the co-administration of amiloride. As a control, the

absorptive-flux of cinnarizine from micelles of Brij® 97 (a non-ionisable, non-absorbable

surfactant) was assessed in the absence and presence of amiloride. Cinnarizine solubility was

evaluated under conditions of decreasing pH and decreasing micellar lipid content to assess

changes in solubilisation and thermodynamic activity during micellar passage across the UWL.

Results. In the presence of amiloride, the absorptive-flux of cinnarizine and 3H-oleic acid from

mixed micelles decreased 6.5-fold and 3.0-fold, respectively. In contrast, the absorptive-flux of

cinnarizine from Brij® 97 micelles remained unchanged by amiloride, and was significantly lower

than from the long-chain micelles. Cinnarizine solubility in long-chain micelles decreased under

conditions where pH and micellar lipid content decreased simultaneously. Conclusion. The acidic

microclimate of the intestinal UWL promotes drug absorption from intestinal mixed micelles via

the promotion of fatty acid absorption which subsequently stimulates drug supersaturation. The

observations suggest that formulations containing absorbable lipids (or their pre-digestive

precursors) may outperform formulations that lack absorbable components due to the benefits

associated with lipid absorption-induced drug supersaturation.


192

5.2 INTRODUCTION

Co-administration of poorly water-soluble drugs (PWSD) with lipids often leads to a significant

enhancement in oral bioavailability2. In the small intestine, the digestion of formulation or dietary-

derived di/triglycerides liberates fatty acids and monoglycerides that are solubilised by biliary

components (bile salts, phosphatidylcholine, cholesterol) to generate a series of lipid colloidal

structures including vesicles and mixed micelles. These colloidal phases in turn provide dispersed

lipidic microenvironments for the solubilisation of co-administered PWSD, thereby increasing the

drug solubilisation capacity of the small intestine when compared to conditions in the fasted

state76.

Although solubilisation increases the apparent solubility of PWSD in the small intestine, the total

concentration of drug in solution (Ctotal) exists in equilibrium between the concentration

solubilised in the colloidal fraction (Ccolloid) and the concentration in the free fraction (Cfree):


In the absence of solid drug, solubilisation in colloidal structures such as micelles and vesicles is

expected to result in a reduction in drug thermodynamic activity143. In simple micellar systems a

reduction in thermodynamic activity manifests as a decrease in Cfree. Thus, solubilisation in

colloids does not increase (and may reduce) free drug concentrations. Whether increases in total

solubilisation capacity translate into enhancements in drug absorption is therefore difficult to

predict with certainty. Indeed, recent studies suggest that in the absence of an increase in free drug

concentrations, solubilisation may not result in enhanced drug absorption despite increases in total

solubilised drug concentrations100, 145, 146.


193

Recently, however, we have observed that formulations containing lipids may provide unique

absorption benefits for solubilising formulation, since drug supersaturation appears to be triggered

during lipid processing in the gastrointestinal (GI) tract4, 176, 227, 251. Under these circumstances, the

induction and maintenance of supersaturation has the potential to reverse (or at least attenuate) the

reduction in drug thermodynamic activity inherent in solubilisation, and may significantly enhance

free drug concentrations above the aqueous solubility. For lipid-based formulations (LBF), drug

supersaturation may be generated by several processes. Firstly, when the formulation loses

solubilisation capacity during the dilution of water miscible co-solvents or surfactants17, 178, 191,

secondly, as a result of the digestion of triglycerides and/or surfactants within the formulation4, 91,

176, 251, and thirdly as lipid-rich colloidal species are diluted by biliary secretions (Chapters 3 and

4).

The ability for drug supersaturation in lipid-based colloids to translate into enhanced drug

absorption has been shown using systems that model the species likely to form during the

digestion of glyceride lipids (i.e. micelles and vesicles containing bile salts, lysophospholipid,

cholesterol, fatty acid and monoglyceride) (Chapters 3 and 4). In these studies, interaction of

secreted bile with lipid colloidal phases reduced the solubilisation capacity of the colloids for

poorly water-soluble weak bases, but since drug precipitation was not immediate, supersaturation

was induced. The period of drug supersaturation that preceded drug precipitation coincided with

significant enhancements in the absorption of cinnarizine across rat jejunum (Chapters 3 and 4).

Interestingly, in the same studies, some increase in absorption in the presence of bile was also

apparent for danazol, even though interaction with bile did not reduce micellar solubilisation

capacity in vitro (and therefore did not stimulate supersaturation). Indeed, addition of bile

increased drug solubility in a fashion more consistent with traditional models for micellar


194

solubilisation (where the addition of solubilising species such as bile typically increases

solubilisation capacity). For danazol, the driver for increased drug absorption was suggested to be

the potential for lipid absorption (i.e. the removal of micellar lipid content) to reduce micellar drug

solubilisation capacity and to trigger drug supersaturation at the absorptive site (assuming lipid

absorption was faster than drug absorption). These preliminary data provide the background for

the current chapter that has examined in detail the role of lipid absorption from intestinal micellar

species as a driver of drug supersaturation, and therefore increased drug absorption.

The absorption of long-chain fatty acids (LCFA) is extremely efficient and facilitated by the acidic

microclimate (pH 5.3–6.245, 46, 57, 252) that is present within the unstirred water layer (UWL) at the

absorptive surface of the small intestine56, 57. The UWL (see Figure 5.1) separates bulk intestinal

fluid from the surface of intestinal absorptive cells, and is estimated to be 500-800 µm wide45, 46.

The UWL exists coincident with, and is indistinguishable from, a viscous mucus layer consisting

of water (~ 95%), glycoproteins, lipids, mineral salts and free proteins45, 47, 48. The acidic

microclimate of the UWL is maintained by the action of the Na+/H+ antiporter present at the brush

border membrane of enterocytes46, as well as the mucus coating which retards H+ diffusion into

bulk luminal fluid45, 46. Shiau and colleagues were the first to describe the facilitatory role of the

acidic microclimate in dietary LCFA absorption from intestinal mixed micelles56. These studies

showed that LCFA absorption was higher in the presence of the low pH microclimate of the UWL.

The authors postulated that the exposure of micelles to the UWL acidic microclimate led to the

protonation of ionised LCFA and an increase in lipid absorption via two mechanisms (depicted in

Figure 5.1(i)). Firstly, protonated LCFA were expected to preferentially partition into and across

the absorptive membrane in accord with classical pH-partition theory56, 57. Secondly, the

protonation of fatty acids was suggested to reduce LCFA amphiphilicity and thereby reduce

LCFA solubility in bile salt micelles. The decrease in micellar LCFA solubility was subsequently


195

suggested to stimulate micellar dissociation, resulting in increased LCFA thermodynamic activity

and increased LCFA absorption. A decrease in pH at the UWL is therefore expected to lead to a

reduction in the lipid content of intestinal mixed micelles via promotion of LCFA micellar

dissociation and absorption (Figure 5.1(i)).

Since the presence of lipid digestion products within mixed micelles contributes significantly to

drug solubilisation capacity39, 91, in the current chapter we have explored the hypothesis that in

promoting LCFA micellar dissociation and absorption, the acidic microclimate also promotes a

reduction in the drug solubilisation capacity of LCFA-containing intestinal mixed micelles. This

in turn is expected to facilitate drug absorption via the induction of drug supersaturation at the

UWL.

The importance of the presence of absorbable lipids in intestinal mixed micelles as a driver of

drug absorption has been studied in two ways. Firstly by examining drug absorption from LCFA-

containing intestinal mixed micelles in the absence and presence of amiloride. Amiloride is an

inhibitor of the Na+/H+ antiporter that is responsible for acidifying the UWL and is therefore an

inhibitor of fatty acid absorption. Secondly, comparison has been made between drug absorption

from LCFA-containing micelles and from micellar systems comprising non-ionic surfactant (Brij

97), where the micellar components are not expected to be absorbed and therefore where changes

to drug solubilisation capacity are not expected.

The data suggest that lipid absorption is a significant trigger for the induction of drug

supersaturation, and that the combination of fatty acid-containing solubilising species and the

acidic intestinal unstirred water layer may be a particularly powerful driver for drug

supersaturation and absorption. The results provide an improved mechanistic understanding of the


196

enhancements in drug absorption often observed from lipid-based systems containing digestible

lipids, and also serve to exemplify the beneficial effects of digestible lipids in food on drug

absorption.

Figure 5.1: Schematic of the proposed mechanisms by which the UWL acidic microclimate

facilitates the absorption of micellar solubilised (i) long-chain fatty acids (LCFA) and (ii) poorly

water-soluble drug (PWSD). (i) Exposure of mixed micelles to the acidic microclimate leads to

protonation of LCFA, attenuating their amphiphilic character and reduces LCFA solubility in

mixed micelles. Increased LCFA thermodynamic activity subsequently promotes LCFA

dissociation from mixed micelles and absorption across apical membrane. (ii) At the UWL,

removal of LCFA from mixed micelle via dissociation and absorption decreases the solubilisation

capacity for PWSD, therefore triggering drug supersaturation in close proximity to the absorptive

site, and enhances drug absorption via increases in thermodynamic potential. D represents drug.

Drug is either free and available for absorption or associated with micelles. Dss is used to signify

acidic microclimate-induced drug supersaturation that drives increases in drug absorption.

Unstirred water layer(acidic microclimate)

(mucus layer)

H+

H+

D

H+

Na+

Bile saltsFatty acid

Bulk lumen

(i) Lipid

(ii) Drug

--

--

----

-

-

-

-

-

-

--

----

-

-D DD DD

-

-

DD DD DD

-

-D

Dss

Dss

Dss

D D

H+

Small Intestine


197

5.3 METHODS

5.3.1 Materials

Cinnarizine, flunarizine dihydrochloride, amiloride hydrochloride hydrate, sodium taurocholate,

sodium taurodeoxycholate, sodium glycocholate, sodium glycochenodeoxycholate, cholesterol, L-

α-lysophosphatidylcholine (LPC, from egg yolk), oleic acid, sodium chloride (NaCl) and Brij® 97

were obtained from Sigma-Aldrich, Australia. Sodium taurochenodeoxycholate, sodium

glycodeoxycholate, ortho-phosphoric acid 85% (H3PO4), sodium hydroxide pellets (NaOH) and

tert-butyl methyl ether (TBME) were from Merck, Australia. Disodium hydrogen orthophosphate

(Na2HPO4), sodium dihydrogen orthophosphate (NaH2PO4.2H2O) and ammonium dihydrogen

orthophosphate (NH4H2PO4) (Ajax Finechem, Australia), Irga-Safe PlusTM (Perkin Elmer Life

Sciences, MA, USA), oleic acid, [9,10-3H(N)] (60 Ci/mmol) (American Radiolabelled Chemicals,

MO, USA), transcutol HP (Gattefossé, France), heparin sodium injection BP (1000 I.U./mL,

Hospira, Australia), xylazine (100 mg/mL, Troy Laboratories, Australia), acepromazine (10

mg/mL, Ceva Delvet, Australia), ketamine (100 mg/mL, Provet, Australia) and pentobarbitone

sodium (325 mg/mL, Virbac, Australia) were obtained from listed suppliers. Acetonitrile and

chloroform used were analytical reagent grade. Water was obtained from a Millipore milliQ

Gradient A10 water purification system (Millipore, MA, USA).


In situ rat jejunal perfusion experiments were conducted to assess the role of the acidic

microclimate in LCFA and drug absorption from LCFA-containing intestinal colloids. Specifically,

the intestinal absorptive flux of oleic acid and cinnarizine from a model LCFA-containing colloid

(“model LCFA colloids”) was assessed in the absence and presence of 2 mM amiloride, a

competitive inhibitor (with respect to Na+) of the plasma membrane Na+/H+ exchanger253 that has


198

previously been shown to attenuate the acidic microclimate on the cell surface of the rat jejunum46.

As a control, the absorption of cinnarizine from “model Brij 97 colloids” (Brij 97 is a non-

ionisable and non-absorbable surfactant) in the absence and presence of 2 mM amiloride was also

assessed. The total cinnarizine concentration (130 µg/mL) and cinnarizine thermodynamic activity

(~ 80% saturated solubility) were matched in both colloidal systems. Cinnarizine (a weak base)

was selected as a model PWSD in this proof-of-concept study, as the solubility of cinnarizine in

LCFA-containing intestinal colloids was previously found to be highly dependent on oleic acid

content (Chapter 4), and therefore may be more amenable to enhancement in drug thermodynamic

activity induced by fatty acid absorption. The model LCFA colloids used in this study were

representative of the diluted post-digestion lipid colloidal phases likely responsible for the

presentation of solubilised drug to the absorptive membrane39, 76.

In vitro solubility studies were conducted to evaluate expected changes in cinnarizine

solubilisation when model LCFA or Brij 97 colloids were exposed to the acidic microclimate in

vivo. The equilibrium solubility of cinnarizine was assessed in a series of LCFA colloids with

decreasing system pH and decreasing lipid concentration (to simulate exposure to the acidic

microclimate and lipid absorption); as well as in a series of Brij 97 colloids with decreasing

system pH (to simulate exposure to the acidic microclimate only, as Brij 97 is not absorbed).

The role of the acidic microclimate in the absorption of cinnarizine from supersaturated, LCFA-

containing colloids was also assessed in rat jejunal perfusion studies via co-perfusion of donor bile

with cinnarizine-loaded LCFA colloids (conditions previously shown to induce cinnarizine

supersaturation in situ and to promote intestinal drug absorption – Chapter 4), in the absence and

presence of 2 mM amiloride.


199

5.3.3 Preparation of LCFA-containing intestinal colloids

Model LCFA colloids used in in situ rat perfusion studies consisted 0.1% w/v oleic acid and 0.06%

w/v monoolein solubilised in simulated endogenous intestinal fluid (SEIF)39 at pH 6.30 ± 0.01.

SEIF comprised of 4 mM total bile salt (25 mol% glycocholate, 17.5 mol% glycodeoxycholate, 25

mol% glycochenodeoxycholate, 12.5 mol% taurocholate, 7.5 mol% taurodeoxycholate, 12.5 mol%

taurochenodeoxycholate), 1 mM LPC and 0.25 mM cholesterol. The oleic acid:monoolein molar

ratio was kept at 2:1, reflecting the ratio of digestion products expected from digestion of 1 mole

of triolein. To model the effect of colloid interaction with the acidic microclimate and the

absorption of lipid components on cinnarizine solubility, systems were prepared at decreasing pH

(pH 6.3, 5.8, 5.3, 4.8) and containing decreasing quantities of lipids (0.1, 0.05, 0.025, 0% w/v

oleic acid, with a proportional decrease in monoolein concentrations) for cinnarizine equilibrium

solubility determination studies.

The preparation of SEIF and LCFA colloids were as described in Chapter 4. pH adjustment of

colloids to 6.30, 5.80, 5.30, 4.80 was achieved by drop wise addition of H3PO4 solution. For the

preparation of drug-loaded LCFA colloids (for in situ jejunal perfusion studies), cinnarizine was

pre-dissolved in oleic acid and allowed to equilibrate overnight at a concentration of 61 mg/g and

115 mg/g, prior to preparation of colloids, such that the final concentration of cinnarizine in the

colloids was 65 µg/mL (~ 40% saturated solubility) and 130 µg/mL (~ 80% saturated solubility),

respectively. As a final step, trace quantities of 3H-oleic acid (to achieve 0.25 µCi/mL) were added

to the drug-loaded colloids, followed by a 1-min vortex. When amiloride was included in the

LCFA colloids, the appropriate mass of amiloride was dissolved in the prepared colloids at 37 °C,

and used within 30 min of preparation. The total sodium strength in all the prepared colloids was

kept constant at 150 mM.


200

5.3.4 Preparation of Brij 97 colloids

Brij 97 (a liquid at 37 °C) was weighed into a volumetric flask and made to volume with

phosphate buffer (18 mM NaH2PO4.2H2O, 12 mM Na2HPO4, 108 mM NaCl), followed by pH

adjustment to 6.30 ± 0.01 with H3PO4 solution. From a plot of cinnarizine solubility vs. Brij 97

concentration, 3.09% w/v Brij 97 was identified as the concentration required to provide equal

cinnarizine solubilisation capacity as the model LCFA colloid (i.e. 157.7 µg/mL). This

concentration (3.09% w/v) was therefore used to form the model Brij 97 colloids used in jejunal

perfusion experiments. Solutions of 3.09% w/v Brij 97 were also prepared at pH 5.80, 5.30, 4.80

(pH adjustment via drop wise addition of H3PO4 solution) for cinnarizine equilibrium solubility

determination studies. For the preparation of drug-loaded Brij 97 colloids (for in situ jejunal

perfusion studies), 100 µL of a 130 mg/mL cinnarizine in transcutol stock solution was spiked into

10 mL of model Brij colloids to achieve a final concentration of 130 µg/mL cinnarizine (~ 80%

saturated solubility). When amiloride was included in the Brij 97 colloids, the appropriate mass of

amiloride was dissolved in the prepared colloids at 37 °C, and used within 30 min of preparation.

The total sodium concentration in all the prepared colloids was kept constant at 150 mM.

5.3.5 Equilibrium solubility studies of cinnarizine in colloids

To determine the equilibrium solubility of cinnarizine in LCFA colloids and Brij 97 colloids,

excess solid drug was added to 2 mL of the colloids in glass vials. Vials were briefly vortexed,

incubated at 37 °C, and samples taken every 24 h over a period of 120 h. During sampling, vials

were centrifuged (2,200 xg, 10 min, 37 °C), 50 µL of supernatant sampled, and vials re-vortexed.

Cinnarizine concentration in the supernatant was determined via HPLC. Equilibrium solubility

was defined when cinnarizine concentrations in consecutive supernatant samples varied by ≤ 5%

on three separate occasions. The equilibrium solubility of cinnarizine was also determined when 2

mM amiloride was included in model LCFA colloids, model Brij 97 colloids, and 1:1 v/v mixture


201

of model LCFA colloids and fasted rat bile, to confirm that solubilisation capacity for cinnarizine

was unaltered by the inclusion of amiloride. Based on physical examination and the maintenance

of consistent drug solubilisation capacity, model LCFA and Brij 97 colloids were stable for 5 days

after preparation.

For some of the LCFA colloids prepared at system pHs < 5.8, phase separation into a highly

dispersed aqueous phase and an undispersed oil phase was evident. Since an oil phase is not likely

to be retained at the UWL in vivo (unionised LCFA is expected to be readily absorbed in close

proximity to the absorptive membrane), the drug solubilisation capacity of the aqueous micellar

phase rather than the oil phase was deemed relevant in assessing expected changes to drug

solubilisation and thermodynamic activity in the UWL. Therefore, to accurately determine the

equilibrium solubility of cinnarizine in the aqueous phase for all LCFA colloids, solubility

samples were left to equilibrate at 37 ºC for 120 h (as equilibrium solubility was attained by this

time), and ultracentrifuged for 30 min at 37 ºC and 400,000 xg (Optima xL-100 K centrifuge, SW-

60 rotor, Beckman, Palo Alto, CA, USA) to separate samples into an aqueous phase, an

undispersed oil phase (if any), and a pellet phase (containing excess solid drug) as described

previously254. Samples from the aqueous phase were assayed for cinnarizine content by HPLC.

5.3.6 Animals






202

5.3.7.1 Single-pass rat jejunum perfusion






Surgical procedures for the collection of fasted bile from donor rats have been described in

Section 4.3.9.1.


After surgery, animals were equilibrated for 30-min, during which time blood was collected from

the cannulated mesenteric vein (~ 0.3 mL/min) to enable re-infusion via the jugular vein.

Perfusion buffer (150 mM Na+, 18 mM H2PO4-, 12 mM HPO42-, 108 mM Cl-, adjusted to pH 6.30

± 0.01) was pumped through the jejunal segment at a rate of 0.1 mL/min and outflowing buffer

discarded to waste. For experiments where amiloride was administered, perfusion buffer

containing 2 mM amiloride was perfused during the equilibration period. The exposed jejunal

segment was kept moist by covering with saline-soaked gauze throughout the experiment.

Perfusate flow through the jejunal segment was maintained at 0.1 mL/min in all experiments to

minimise variations in the thickness of the unstirred water layer which may influence drug flux206.

Following the equilibration period, perfusion buffer was replaced with model colloids. The

perfused colloids were sampled at t = 0 to confirm cinnarizine and/or 3H-oleic acid concentrations.

After this time, the outflowing perfusate was continuously collected at 10-min intervals, and

briefly vortexed before samples were taken for analysis of drug and/or 3H-oleic acid content.


203

Blood draining the perfused jejunal segment was also collected at 5-min intervals, plasma

separated by centrifugation (10,000 xg, 5 min), and samples taken for analysis of cinnarizine

and/or 3H-oleic acid concentrations.

The concentration of cinnarizine flowing into the perfused jejunal segment was held at 130 µg/mL

in studies assessing the role of the acidic microclimate in the absorption of cinnarizine from model

LCFA-containing or Brij 97 colloids. When amiloride was included in these experiments, it was

pre-dissolved in the model colloids at 37 ºC at a concentration of 2 mM amiloride. In studies that

assessed the role of the acidic microclimate in the absorption of cinnarizine from supersaturated

LCFA-containing colloids (i.e. where co-perfusion of donor rat bile triggered drug supersaturation

in the model LCFA colloids), the concentration of cinnarizine flowing into the perfused jejunal

segment was held at 65 µg/mL. Therefore, in experiments where model colloids were perfused

alone, cinnarizine was loaded into the perfusate at 65 µg/mL (~ 40% saturated solubility). In

experiments where model colloids were co-perfused in a 1:1 v/v ratio with a secondary perfusate

of bile, cinnarizine was loaded into the primary perfusate at 130 µg/mL (~ 80% saturated

solubility), such that 1:1 v/v dilution led to a final perfusate concentration of 65 µg/mL cinnarizine.

When amiloride was included, it was pre-dissolved in bile at 37 ºC at 4 mM, such that 1:1 v/v

dilution led to a final perfusate concentration of 2 mM amiloride. Both model LCFA colloids and

bile were pumped at 0.05 mL/min, and mixed via a three-way “T” connector immediately prior to

entry into the jejunal segment, such that total perfusate flow was maintained at 0.1 mL/min. Since

luminal cinnarizine supersaturation was generated in these experiments, outflowing perfusate

samples were taken before and after centrifugation (2,200 xg, 2 min), to obtain an indication of the

degree of drug precipitation within the jejunal segment.


204


5.3.9.1 Sample preparation and HPLC assay conditions for cinnarizine

Samples of LCFA colloids and Brij 97 colloids were prepared for HPLC assay by a minimum 20-

fold dilution with mobile phase (50% v/v acetonitrile:50% v/v 20 mM NH4H2PO4). Plasma

samples were prepared for HPLC using a validated extraction procedure, with flunarizine as an

internal standard, as reported previously227. Replicate analysis of n = 4 quality control samples

revealed acceptable accuracy and precision (± 10%, ± 15% at the limit of quantification) for

cinnarizine concentrations between 20–1000 ng/mL for colloids, and 10–320 ng/mL for plasma.


Quantification of 3H-oleic acid in perfusate and collected plasma was performed via scintillation

counting on a Packard Tri-Carb 2000CA liquid scintillation analyser (Packard, Meriden,

Connecticut, USA). Perfusate samples (100 µL) and plasma samples (200 µL) were added to 2 mL

Irga-safe Plus scintillation fluid followed by a 10-sec vortex. Samples were corrected for

background radioactivity by the inclusion of a blank sample in each run.

5.3.9.3 Blood:plasma ratio determination for cinnarizine, oleic acid

The blood:plasma ratios for cinnarizine and oleic acid have been determined as described in

Chapter 2. The mean blood:plasma ratio was used to convert plasma concentrations to blood

concentrations in perfusion experiments, enabling quantification of total compound transport into

mesenteric blood.


205


In the single-pass rat jejunum perfusion model, permeability coefficients were calculated from the

flux data obtained after attainment of steady state drug transport into mesenteric blood. Two

apparent permeability coefficients (Papp) were calculated as described previously184:



∆.

Equation 5.3

where ‘Disappearance’ Papp is the apparent permeability coefficient calculated from drug loss from

the perfusate (cm/sec); ‘Appearance’ Papp is the apparent permeability coefficient calculated from

drug appearance in the mesenteric blood (cm/sec); Q is the perfusate flow rate (mL/sec); A is the

surface area of the perfused jejunal segment (cm2), which is calculated by multiplying the

diameter and the length of the perfused intestinal segment as described previously207; C1 is the



appearance in mesenteric blood at steady state (ng/sec); and <C> is the logarithmic mean drug

concentration in the lumen (ng/mL), where <C> = (C1 – C0) / (ln C1 – ln C0).


Statistically significant differences were determined by ANOVA followed by Tukey's test for

multiple comparisons at a significance level of α = 0.05 using SPSS v19 for Windows (SPSS Inc.,

Chicago, IL, USA).


206

5.4 RESULTS

5.4.1 Attenuation of the acidic microclimate using amiloride reduces oleic

acid and cinnarizine absorption from model LCFA colloids; but has no effect

on the absorption of cinnarizine from fatty acid-free Brij 97 colloids

Figure 5.2 shows the intestinal absorptive flux vs. time profiles of oleic acid (from LCFA

containing colloids), and cinnarizine (from both LCFA and Brij 97 colloids) in the absence and

presence of 2 mM amiloride during in situ rat jejunum perfusion experiments. Corresponding

steady state-absorptive flux, disappearance Papp, and appearance Papp data are reported in Table 5.1

Perfusate disappearance profiles are shown in Figure 5.3.

Administration of amiloride reduced the absorption of oleic acid, resulting in a significant (α <

0.05) 3.0-fold, 1.4-fold, and 3.0-fold reduction in absorptive flux, disappearance Papp, and

appearance Papp of oleic acid, respectively (Figure 5.2A, Table 5.1). Amiloride administration also

led to a significant reduction in cinnarizine absorption from LCFA colloids, resulting in 6.5-fold,

2.8-fold, and 5.7-fold reductions (α < 0.05) in the absorptive flux, disappearance Papp, and

appearance Papp of cinnarizine, respectively (Figure 5.2B, Table 5.1). In contrast, co-

administration of amiloride with Brij 97 colloids did not lead to significant changes in the

absorptive flux, disappearance Papp, and appearance Papp of cinnarizine (Figure 5.2C, Table 5.1).

The model colloids were chosen such that the cinnarizine solubilisation capacity in each was the

same (157.7 ± 3.0 µg/mL and 153.2 ± 2.5 µg/mL for LCFA colloids and Brij 97 colloids

respectively, mean ± SEM of n = 3 determinations), and drug was loaded at the same

concentration (130 µg/mL). Cinnarizine thermodynamic activity was therefore also matched and

cinnarizine was dissolved at ~ 80% saturated solubility in both colloidal solutions. In the absence


207

of amiloride (i.e. in the presence of an intact acidic microclimate), the absorptive flux of

cinnarizine from model LCFA colloids was 4.9-fold higher than from Brij 97 colloids (Figure

5.2B & C, Table 5.1). This difference was abolished in the presence of amiloride (i.e. under

conditions where the acidic microclimate was attenuated and lipid absorption inhibited). The data

suggest that cinnarizine absorption is significantly more efficient from LCFA-containing intestinal

colloids than from Brij 97 colloids, in spite of matched initial thermodynamic activity in both

colloids, and that this is dependent on the presence of an acidic microclimate at the intestinal

UWL. For the model LCFA colloids, the increase in cinnarizine absorption in the absence of

amiloride also occurred coincidentally with an increase in oleic acid absorption (Figure 5.2A & B,

Table 5.1).


208

(A)

(B)

(C)

Figure 5.2: Mesenteric blood appearance profiles of (A) oleic acid (OA) from model LCFA

colloids, (B) cinnarizine (CIN) from model LCFA colloids, and (C) cinnarizine (CIN) from model

Brij 97 colloids, after 70 min single-pass perfusion of ~ 10 cm2 segments of rat jejunum with and

without co-administration of 2 mM amiloride, which attenuates the acidic microclimate of the

intestinal unstirred water layer. Model LCFA colloids and model Brij 97 colloids had equal

solubilisation capacity for CIN (157.7 ± 3.0 µg/mL and 153.2 ± 2.5 µg/mL, respectively; average

± SEM of n = 3 determinations). CIN was loaded into both colloids at constant concentration (130

µg/mL) and thermodynamic activity (~ 80% saturated solubility). Data represent mean ± SEM of

n = 3-4 experiments. Steady state-absorptive flux, disappearance Papp, and appearance Papp of CIN

and OA from this series of experiments are tabulated in Table 5.1.

Time (min)

0 10 20 30 40 50 60 70

OA

flux

into

mes

ent

eric

b

loo

d (

ng

/5 m

in/1

0 c

m2

)0

2000400060008000

100001200014000 LCFA colloids

LCFA colloids + amiloride

Time (min)

0 10 20 30 40 50 60 70

CIN

flu

x in

to m

esen

teric

bl

ood

(ng/

5 m

in/1

0 cm

2 )

0

200

400

600

800

1000LCFA colloidsLCFA colloids + amiloride

Time (min)

0 10 20 30 40 50 60 70

CIN

flu

x in

to m

esen

teric

bl

ood

(ng/

5 m

in/1

0 cm

2 )

0

200

400

600

800

1000 Brij 97 colloids Brij 97 colloids + amiloride


209

Table 5.1: Cinnarizine (CIN) and oleic acid (OA) disappearance Papp (x 106 cm/sec) from the

intestinal perfusate, appearance Papp (x 106 cm/sec) in the mesenteric blood, and steady state

absorptive flux into mesenteric blood (ng/5 min/10 cm2) after 70 min of single-pass perfusion of ~

10 cm2 segments of rat jejunum with model long chain fatty acid (LCFA) colloids or model Brij

97 colloids, with and without co-administration of 2 mM amiloride. Values calculated using data

obtained after steady state attainment (t = 55-70 min). Data represent mean ± SEM of n = 3-4

experiments.

Perfusate CIN conc.

(µg/mL)

OA conc.

(µg/mL)

Disappearance Papp (x

106 cm/sec)

Appearance Papp

(x 106 cm/sec) Mesenteric blood

flux (ng/5min/10cm2)

CIN LCFA colloid 130 - 35.5 ± 2.2 1.7 ± 0.4 600 ± 129

LCFA colloid + amiloride

130 - 12.7 ± 4.1 a 0.3 ± 0.1 a 93 ± 25 a

Brij 97 colloid 130 - 14.0 ± 6.7 a 0.3 ± 0.0 a 123 ± 14 a

Brij 97 colloid + amiloride

130 - 5.4 ± 1.2 a 0.2 ± 0.0 a 90 ± 3 a

OA LCFA colloid - 1000 22.1 ± 2.0 3.3 ± 0.4 9144 ± 991

LCFA colloid + amiloride

- 1000 15.3 ± 0.9 b 1.1 ± 0.1 b 3045 ± 304 b

a Significant difference (α < 0.05) from model LCFA colloid group (containing 130 µg/mL CIN) in the absence of amiloride

b Significant difference (α < 0.05) from model LCFA colloid group (containing 1000 µg/mL OA) in the absence of amiloride


210

(A)

(B)

(C)

Figure 5.3: Perfusate disappearance profiles of (A) oleic acid (OA) from model LCFA colloids,

(B) cinnarizine (CIN) from model LCFA colloids, and (C) cinnarizine (CIN) from model Brij 97

colloids, after 70 min single-pass perfusion of ~ 10 cm2 segments of rat jejunum with and without

co-administration of 2 mM amiloride, which attenuates the acidic microclimate of the intestinal

unstirred water layer. Model LCFA colloids and model Brij 97 colloids had equal solubilisation

capacity for cinnarizine (157.7 ± 3.0 µg/mL and 153.2 ± 2.5 µg/mL, respectively; average ± SEM

of n = 3 determinations). CIN was loaded into both colloids at constant concentration (130 µg/mL)

and thermodynamic activity (~ 80% saturated solubility). Data represent mean ± SEM of n = 3-4

experiments.

Time (min)

0 30 40 50 60 70

% O

A d

ose

pass

ing

thro

ugh

jeju

num

60

70

80

90

100

LCFA colloids LCFA colloids + amiloride

Time (min)

0 30 40 50 60 70

% C

IN d

ose

pass

ing

thro

ugh

jeju

num

60

70

80

90

100

LCFA colloids LCFA colloids + amiloride

Time (min)

0 30 40 50 60 70

% C

IN d

ose

pass

ing

thro

ugh

jeju

num

60

70

80

90

100

Brij 97 colloids Brij 97 colloids + amiloride


211

5.4.2 Exposure of model LCFA colloids to the acidic microclimate, and

absorption of lipid components, leads to cinnarizine supersaturation and

enhanced thermodynamic activity

To examine the possible mechanisms by which pH changes at the UWL (due to the acidic

microclimate), and the process of lipid absorption, result in changes to cinnarizine absorption, the

equilibrium solubility of cinnarizine in LCFA colloids was assessed in systems with decreasing

pH and decreasing lipid concentration (to simulate exposure to the acidic microclimate and lipid

absorption, Figure 5.4A). Equivalent data were also generated for Brij 97 colloids with decreasing

system pH, in this case to simulate exposure to the acidic microclimate only, as Brij 97 is not

absorbed (Figure 5.4B).

As system pH decreased, the LCFA colloids became increasingly turbid, ultimately leading to

phase separation into an aqueous phase and an undispersed oil phase (Figure 5.5). Phase

separation occurred earlier (i.e. at higher pHs) in systems containing higher concentrations of lipid.

Systems that phase separated are asterisked in Figure 5.4A. Phase separation occurred at pH 4.8

for all LCFA-based colloids and at and below pH 5.3 for colloids containing the highest lipid load

(i.e. 0.1% w/v oleic acid). Even where phase separation did not occur, the turbidity of the LCFA

colloidal solutions increased with increasing lipid concentration. Development of turbidity was not

evident in SEIF (i.e. 0% oleic acid) or the Brij 97 colloids (Figure 5.5).

Changes in turbidity of the LCFA-containing colloids appeared to correlate with changes to

solubilisation capacity, such that drug solubility was higher in systems with increasing turbidity.

As depicted in Figure 5.4A, cinnarizine solubility in LCFA colloids increased significantly with

increasing lipid concentration and also increased (albeit more moderately) with decreasing pH.

These trends continued until phase separation occurred, at which point cinnarizine solubility in the


212

micellar phase was reduced due to partitioning of lipids from the micellar phase into a poorly

dispersed lipid phase. At 0% incorporated lipid (i.e. SEIF only), cinnarizine solubility also

increased slightly with decreasing pH. However, this increase was small when compared to the

increase in solubility afforded by increasing lipid concentration or decreasing pH in the LFCA-

containing colloids.

In vivo, exposure of LCFA colloids to the intestinal acidic microclimate results in a decrease in pH

and an increase in LCFA absorption. Changes to cinnarizine solubility under the same

circumstances are therefore expected to be predicted by assessment of solubility changes under

conditions where system pH and lipid concentration are decreased simultaneously. ‘O’ in Figure

5.4 identifies the maximum cinnarizine solubilisation capacity of the LCFA and Brij 97 colloids.

Cinnarizine was loaded into model colloids at 80% saturated solubility, i.e. 130 µg/mL – blue

dotted line. At the highest lipid load (0.1% w/v OA), exposure of the LCFA containing colloids to

decreases in pH that are consistent with conditions in the acidic microclimate, initially increased

cinnarizine solubility (at pH 5.8), but ultimately reduced cinnarizine solubility at pH 5.3 (the

lowest reported microclimate pH45) (Figure 5.4). The solubility of cinnarizine was also highly

dependent on the concentration of incorporated lipids, and a reduction in oleic acid content

(consistent with the reductions expected on lipid absorption) significantly reduced cinnarizine

solubility. Indeed, the drop in cinnarizine solubility seen at pH 5.3 most likely reflects a loss of

micellar lipid content due to phase separation rather than an effect of pH alone. A theoretical line

in Figure 5.4A (red broken arrow) depicts the changes in cinnarizine solubility expected under

conditions where pH is reduced to pH 5.30 and lipids are fully absorbed. Under these

circumstances, cinnarizine solubility is reduced dramatically and in the absence of precipitation, is

expected to lead to supersaturation, enhanced thermodynamic activity and improved absorption.


213

Conversely, in Brij 97 colloids, a decrease in system pH led to increased cinnarizine solubility

(Figure 5.4B), presumably due to increased cinnarizine ionisation at lower pH. Since Brij 97 is not

absorbed, the concentration of micellar surfactant is expected to remain constant during passage

across the UWL, and the only driver of changes to drug solubility at the UWL is the reduction in

pH due to the acidic microclimate. As depicted by the red broken arrow in Figure 5.4B,

cinnarizine solubilisation is therefore expected to increase during passage of Brij 97 colloids

across the UWL, leading to reduced thermodynamic activity.


214

(A)

LCFA colloids

(B)

3.09% w/v Brij 97 colloids

Figure 5.4: (A) Three-dimensional plot of cinnarizine (CIN) solubility (37 °C) in the aqueous

phase of LCFA-containing intestinal colloids as a function of oleic acid and monoolein

concentration in bile micelles (concentration on y-axis refers to oleic acid concentration) and

system pH (B) CIN solubility (37 °C) in 3.09% w/v Brij 97 colloids as a function of system pH.

Concentration of bile components in (A) was kept constant. Black solid arrows depict the

expected change in CIN solubilisation capacity when (1) colloids are exposed to the acidic

microclimate of the UWL, and (2) as oleic acid and monoolein are absorbed and removed from

colloids (absorption of colloidal components only applies to LCFA colloids as Brij 97 is non-

absorbable). * in (A) denotes colloidal systems where phase separation into an aqueous phase and

undispersed oil phase was evident during the 120 h equilibrium solubility determination study (see

Figure 5.5). ‘O’ denotes the solubilisation capacity of the model LCFA and Brij 97 colloids used

in Figure 5.2, and the blue dotted line shows the concentration of CIN (i.e. 130 µg/mL – 80%

saturated solubility) in the model colloids used in Figure 5.2 and Table 5.1. The red broken arrows

depict (A) the theoretical solubility trend when pH of system is reduced by 1 unit to pH 5.3

(lowest pH reported in the literature45) and lipids are completely removed, and (B) when pH of

system is reduced by 1 unit to pH 5.3 such as on entry into the UWL. The passage of model LCFA

colloids across the UWL and subsequent absorption of lipid digestion products is therefore

expected to reduce cinnarizine solubility, leading to supersaturation, enhanced thermodynamic

potential and absorption; while the passage of model Brij 97 colloids across the UWL leads to a

small increase in CIN solubility and reduced thermodynamic potential.

0

50

100

150

200

0.1000.1000.1000.100

0.0500.0500.0500.0500.0250.0250.0250.025

0.0000.0000.0000.0004.8

5.35.8

6.3

* **

*

O

(1)(2)

pH

6.3 5.8 5.3 4.8

CIN

so

lub

ility

(µ

g/m

L)

0

50

100

150

200

250

300

O

(1)


215

(A)

0% OA 0% MO

(B)

0.025% w/v OA 0.016% w/v MO

(C)

0.050% w/v OA 0.032% w/v MO

(D)

0.10% w/v OA 0.06% w/v MO

(E)

3.09% w/v Brij 97

Figure 5.5: The appearance of LCFA-containing intestinal colloids (panel A-D) and 3.09% w/v

Brij 97 colloids (panel E) 120 h into cinnarizine equilibrium solubility determination studies, after

vials were centrifuged at 2,200 xg for 10 min at 37 °C. LCFA-containing colloids consisted oleic

acid (OA) and monoolein (MO) (concentrations as labelled) solubilised in simulated endogenous

intestinal fluid (SEIF) containing 4 mM total bile salt, 1 mM lysophosphatidylcholine and 0.25

mM cholesterol. In each panel, the system pH of colloids was, from left to right, 6.30, 5.80, 5.30


216

and 4.80, respectively. In LCFA-containing systems, the turbidity of samples was observed to

increase when system pH was decreased, up to a point where colloids were destabilised and

underwent phase separation into an aqueous phase and an undispersed oil phase (undispersed oil

phase floats on top of aqueous phase and may not be seen clearly in pictures), by which point the

samples appeared clear again. * denotes colloidal systems where phase separation into an aqueous

phase and undispersed oil phase was evident. Since the concentration of bile components was held

constant, phase separation of colloidal system with 0.10% w/v OA (Panel D) occurred at a higher

pH than systems containing 0.05 and 0.025% w/v OA (Panel B and C). Turbidity did not develop

in SEIF containing 0% OA (Panel A) or 3.09% w/v Brij 97 colloids (Panel E). White mass at the

bottom of vials is excess solid cinnarizine added in the solubility determination studies. The

pictures illustrate the impact of the acidic microclimate on the microstructure of LCFA-containing

intestinal colloids, where exposure of LCFA colloids to the microclimate pH leads to protonation

of LCFA, increasing the ratio of unionised LCFA:ionised LCFA and resulting in the formation of

larger colloids with reduced thermodynamic stability in vitro. In vivo, the increase in LCFA

thermodynamic activity is expected to lead to enhanced micellar dissociation and absorption.


217

5.4.3 Attenuation of the acidic microclimate using amiloride abolishes bile-

induced, supersaturation-enhanced, cinnarizine absorption from model

LCFA colloids

In Chapter 4, co-perfusion of fasted rat bile with cinnarizine-loaded model LCFA colloids

triggered cinnarizine supersaturation in the GI lumen and enhanced cinnarizine absorptive flux by

3.2-fold in an in situ rat jejunal perfusion model (Figure 5.6B, Table 5.2). In the current chapter,

inclusion of 2 mM amiloride in the same perfusate (i.e. 1:1 v/v mixture of model LCFA colloids

and fasted rat bile) reduced the absorptive flux and appearance Papp of oleic acid by 4.6-fold and

4.4-fold, respectively (Table 5.2, Figure 5.6A). A coincident decrease in the absorptive flux and

appearance Papp of cinnarizine (a decrease of 5.2-fold and 5.0-fold, respectively) was also

observed. Cinnarizine appearance permeability in the presence of bile and amiloride was therefore

not significantly different to that in the absence of bile (and therefore in the absence of

supersaturation) (Figure 5.6B, Table 5.2). Analysis of the outflowing perfusate in the permeability

experiment suggested that cinnarizine supersaturation was maintained in the bulk GI fluids, in

both the absence and presence of amiloride, (i.e. precipitation did not occur, Figure 5.7B).

Amiloride co-administration therefore negated the increase in cinnarizine absorption stimulated by

bile-mediated supersaturation. Amiloride did not directly influence cinnarizine solubility in model

LC colloids in the presence or absence of bile (cinnarizine solubility in LC colloids + amiloride

was 151.6 ± 17.0 µg/mL, cinnarizine solubility in LC colloids + bile + amiloride was 10.3 ± 0.3

µg/mL respectively; mean ± SEM; n =3).


218

(A) (B)

Figure 5.6: Absorptive flux-time profiles of (A) oleic acid (OA) and (B) cinnarizine (CIN) when

model LCFA colloids were perfused via single-pass through an isolated rat jejunal segment (~ 10

cm2), with and without 1:1 v/v co-perfusion with donor rat bile, in the absence and presence of 2

mM amiloride. * Data in the absence of amiloride are reproduced from Chapter 4. Co-perfusion of

rat bile generates CIN supersaturation in situ within the perfused jejunal segment, and increased

cinnarizine absorptive flux by 3.2-fold. Inclusion of 2 mM amiloride in the perfusate, however,

abolished the absorption-enhancing effects of bile-induced supersaturation. SS denotes drug

supersaturation in perfusate. Data represent mean ± SEM of n = 3-4 rats. Steady state-absorptive

flux, disappearance Papp, and appearance Papp of CIN and OA from this series of experiment are

tabulated in Table 5.2.

Time (min)

0 10 20 30 40 50 60 70% O

A d

ose

into

me

sen

teric

b

loo

d (

%/5

min

/10

cm

2)

0

1

2

3

4

5

Time (min)

0 10 20 30 40 50 60 70

CIN

flu

x in

to m

esen

teric

bl

ood

(ng/

5 m

in/1

0 cm

2 )

0

100

200

300

400

500

600


219

Table 5.2: Cinnarizine (CIN) and oleic acid (OA) disappearance Papp (x 106 cm/sec) from the

intestinal perfusate, appearance Papp (x 106 cm/sec) in the mesenteric blood, and steady state

absorptive flux into mesenteric blood (ng/5 min/10 cm2) after 70 min of single-pass perfusion of ~

10 cm2 segments of rat jejunum with model LCFA colloids, with and without 1:1 v/v co-perfusion

with donor rat bile, in the absence and presence of 2 mM amiloride. Data in the absence of

amiloride are obtained from Chapter 4. Values calculated using data obtained after steady state

attainment (t = 55-70 min). Data represent mean ± SEM of n = 3-4 experiments.

Perfusate CIN conc.

(µg/mL)

OA conc.

(µg/mL)

DisappearancePapp

(x 106 cm/sec)

Appearance Papp

(x 106 cm/sec)

Mesenteric blood flux

(ng/5min/10cm2)

CIN LCFA colloid * 65 - 12.2 ± 2.8 0.6 ± 0.2 138 ± 28

LCFA colloid + Bile SS * 65 - 36.9 ± 5.5 a 2.5 ± 0.2 a 443 ± 34 a

LCFA colloid + Bile + amiloride SS 65 - 52.9 ± 6.3 a 0.5 ± 0.0 b 85 ± 4 b

OA LCFA colloid * - 1000 18.0 ± 5.9 2.0 ± 0.3 5607 ± 863

LCFA colloid + Bile * - 500 26.1 ± 3.8 6.2 ± 0.8 a 8557 ± 1205

LCFA colloid + Bile + amiloride - 500 35.0 ± 4.1 a 1.4 ± 0.4 b 1853 ± 441 b

a Significant difference (α<0.05) from model LCFA colloid group in the absence of amiloride b Significant difference (α<0.05) from model LCFA colloid + Bile group in the absence of amiloride

* Papp and flux data reproduced from Chapter 4 SS denotes drug supersaturation in perfusate, induced by bile dilution of model LCFA colloids


220

(A) (B)

Figure 5.7: Perfusate disappearance (% drug dose passing through jejunum) profiles of (A) oleic

acid (OA) and (B) cinnarizine (CIN) when model LCFA colloids were perfused via single-pass

through an isolated rat jejunal segment (~ 10 cm2), with and without 1:1 v/v co-perfusion with

donor rat bile, in the absence and presence of 2 mM amiloride. * Data in the absence of amiloride

are reproduced from Chapter 4. Co-perfusion of model LCFA colloids with rat bile generates drug

supersaturation in situ within the perfused jejunal segment. The degree of drug precipitation

within the perfusate is represented by the difference in CIN perfusate concentration pre- and post-

centrifugation. SS denotes drug supersaturation in perfusate. Data represent mean ± SEM of n = 3-4

rats.

Time (min)

0 30 40 50 60 70

% 3

H-O

A d

ose

pa

ssin

gth

rou

gh

jeju

nu

m

50

60

70

80

90

100

Time (min)

0 30 40 50 60 70

% C

IN d

ose

pas

sin

g

thro

ug

h je

jun

um

50

60

70

80

90

100


221

5.5 DISCUSSION

The mechanism of absorption of long-chain fatty acid (LCFA) from intestinal mixed micelles is

well described in the literature. Westergaard and Dietschy’s seminal studies initially proposed that

solubilisation of LCFA within bile salt micelles increased LCFA diffusion across the intestinal

unstirred water layer (UWL), and increased the concentration of LCFA presented to the absorptive

membrane51. Shiau and colleagues later suggested that the acidic microclimate within the UWL

further enhanced LCFA absorption by protonating ionised LCFA, leading to enhanced micellar

dissociation and absorption56. These initial studies have also been followed by several studies

describing the role of active transport systems such as CD3661, FATP62 and SR-BI63 in lipid

uptake across the apical absorptive membrane, although the quantitative importance of active vs.

passive transport as a means of lipid absorption under differing lipid loads remains contentious58-60,

64, 255.

In contrast to lipid absorption, the mechanism of absorption of poorly water-soluble drugs (PWSD)

from intestinal mixed micelles has been less clearly defined. We recently described the potential

for drug supersaturation to be induced during the interactions of lipid colloidal phases (mixed

micelles and vesicles) with secreted bile (Chapter 3 and 4), and suggested that supersaturation-

enhanced absorption may be an endogenous mechanism to reverse the reduction in

thermodynamic activity inherent in drug solubilisation within colloidal phases. During these

studies, however, it became apparent that multiple mechanisms may underpin the generation of

drug supersaturation in intestinal mixed micelles. For weak bases with high solubility in LCFA

containing micelles, interaction with bile appeared to decrease micellar drug solubility with

increasing bile concentration. This in turn resulted in transient supersaturation and absorption

promotion. However, interaction with bile also promoted (albeit to a lesser extent) the absorption

of a non-ionic drug (danazol), where micellar solubility increased rather than decreased with


222

increasing bile concentrations. This occurred simultaneously with an increase in lipid absorption.

These previous studies suggest that lipid absorption and drug absorption may be linked, and

stimulated the current hypothesis, i.e. that lipid absorption from intestinal mixed micelles reduces

drug solubility in micellar structures, thereby inducing supersaturation and promoting drug

absorption.

The possibility that protonation and absorption of LCFA from intestinal mixed micelles during

passage across the UWL can trigger drug supersaturation and enhance drug absorption has been

studied in two ways. Firstly, via an examination of changes to cinnarizine solubility in mixed

micelles as a function of changes to pH and lipid concentration; and secondly by modifying the

conditions within the UWL such that LCFA protonation and absorption are suppressed, and

examining the effects of suppressed lipid absorption on cinnarizine absorption. The latter was

achieved by inhibiting the action of the Na+/H+ antiporter at the enterocyte brush border, thereby

reversing the acidity of the microclimate adjacent to the absorptive surface46. The acidic

microclimate has long been suggested to be important in the absorption of LCFA56. The pKa of

oleic acid, a common LCFA component (or digestion product of dietary or formulation-derived

lipids), is 9.85256. When solubilised in bile salt micelles, however, the pKa of oleic acid decreases

to 6.3–6.5257. As such pH changes from 6.5 in the GI lumen, to 5.3 in the UWL45, significantly

increase the unionised:ionised fraction for LCFA. A decrease in ionisation is expected to increase

LCFA absorption by 1) increasing cellular partitioning due to the pH-partition effect and 2)

increasing thermodynamic activity due to a reduction in amphiphilicity and micellar solubility,

and stimulation of micellar dissociation (as described by Shiau and coworkers56, 57). In contrast,

conjugated bile salts are not absorbed in the upper GI tract. The pKa values of taurine and glycine

conjugated bile salts are <1 and ~3.8, respectively258. Bile salts therefore remain ionised during

passage across the UWL, and in the absence of absorption, bile salt micellar structures are


223

expected to persist. The lipid content of the micelles however, is expected to reduce (due to

enhanced LCFA micellar dissociation and absorption), reducing drug solubilisation capacity and

generating the conditions required for drug supersaturation. In the current studies cinnarizine was

employed as a model PWSD to examine the impact of the acidic UWL on drug absorption from

mixed micelles. Cinnarizine was chosen for two reasons. Firstly, previous studies have shown that

cinnarizine is highly soluble in bile salt-fatty acid mixed micelles; and solubility is highly

dependent on micellar oleic acid content (Chapter 4). Thus, cinnarizine is likely to be amenable to

enhancements in drug thermodynamic activity induced by oleic acid absorption. Secondly, as a

base, exposure to the UWL microenvironment acidity is expected to increase drug ionisation, and

therefore decrease drug absorption based on pH partition relationships. As such, any increase in

drug absorption that occurs due to microenvironment acidity is expected to reflect mechanisms

unrelated to drug ionisation/partitioning.

Under pH conditions reflective of the acidic UWL, the LCFA containing colloids examined here

became increasingly turbid, suggesting increases in colloid particle size consistent with the

observations of Shiau et al.57. The increase in turbidity with decreasing pH only occurred in

systems containing LCFA, and developed instantly (see Figure 5.8 for the appearance of model

LCFA colloids after the addition of one drop of 20% v/v H3PO4). Reduced LCFA solubility in bile

salt micelles manifests as an increase in system turbidity or phase separation in vitro. However, in

vivo where an absorptive sink is present directly adjacent to the UWL, increased LCFA

thermodynamic activity and an increase in unionised:ionised LCFA is expected to contribute to

enhanced absorption56, 57. Consistent with this suggestion, the absorptive flux of oleic acid was

significantly reduced when the acidic microclimate of the UWL was attenuated by co-

administration with amiloride (Figure 5.2A). This is consistent with previous data that have shown

reduced oleic acid uptake into rat jejunal brush border vesicles in the presence of amiloride259.


224

(A)

Model LCFA colloids

(B)

Model Brij 97 colloids

Figure 5.8: Appearance of cinnarizine-loaded (A) model LCFA colloids and (B) model Brij 97

colloids immediately after the addition of one drop of 20% v/v H3PO4. Turbidity instantly

developed in model LCFA colloids while model Brij 97 colloids remained optically clear. The

addition of H3PO4 reduced the system pH of model LCFA colloids and Brij 97 colloids from 6.30

to 5.92, and 6.30 to 6.01, respectively. Magnetic stirrers are present at the bottom of vials. This

figure visually depicts the effect of protonation of oleic acid (OA) when model LCFA colloids are

exposed to acidic conditions. An increase in the ratio of unionised OA:ionised OA is thought to

enhance the thermodynamic potential of OA, and induce a micelle to emulsion transition in the

colloids.

In light of the significant reduction in lipid absorption in the presence of amiloride, this

experimental system was utilised to explore the potential link between lipid absorption and drug

absorption. The data in Figure 5.2B show a very clear reduction in cinnarizine absorption in the

presence of amiloride. The reduction in drug absorption occurred coincidently with the reduction

in LCFA absorption (Figure 5.2A). In contrast, perfusion of Brij 97 colloids containing the same

concentrations of cinnarizine at the same thermodynamic activity resulted in drug flux that was

significantly lower than that from LCFA-containing colloids in the absence of amiloride, and was

independent of the co-administration of amiloride. The data suggest that drug absorption from

micellar systems containing LCFA is inherently more effective than from solubilising systems that


225

lack fatty acids, and that fatty acid absorption is critical in mobilising the solubilised fraction to

maximise drug absorption. These findings are significant in the context of the design of lipid-

based formulations, and suggest that formulations containing absorbable lipids may have inherent

advantages over systems containing non-absorbable surfactants and co-solvents. Consistent with

this suggestion, previous studies comparing danazol absorption from LBF comprising surfactants

and cosolvents alone (i.e. LFCS Type IV formulations) and formulations where the same

surfactants and cosolvents were combined with glyceride lipids (i.e. a more traditional LFCS

Type III self-emulsifying formulation), suggest increased drug absorption from the lipid

containing formulations260. Thus, despite similar solubilisation properties in in vitro dispersion

and digestion tests, a glyceride lipid-containing formulation (55% w/w Cremophor RH 40, 7.5%

w/w ethanol, 37.5% w/w Soybean oil:Maisine (1:1 w/w)) outperformed (~ 2-fold difference in in

vivo exposure of danazol) a similar surfactant/cosolvent formulation (55% w/w Cremophor RH 40,

7.5% w/w ethanol, 37.5% w/w Pluronic L121). We were previously unable to explain these

observations, but the current studies provide some justification for the apparent advantage of

formulations containing absorbable lipids.

To better understand the impact of microclimate pH and lipid absorption on drug absorption from

LCFA-containing intestinal colloids, a series of studies was undertaken to detail changes in drug

solubility under conditions of both reduced pH and decreased lipid content. As a weak base with a

pKa of 7.47, cinnarizine is increasingly ionised at lower pH, and this was reflected in increases in

solubility when the pH of blank SEIF (i.e. bile micelles with 0% solubilised lipids) and model Brij

97 colloids was decreased from 6.3 to 4.8 (Figure 5.4). In LCFA-containing colloids, decreasing

pH initially resulted in the formation of swollen colloidal particles with increased turbidity and

higher cinnarizine solubilisation capacity (Figure 5.4), consistent with previous studies that

suggest increases in drug solubility in larger colloidal particles39. The expected changes to


226

cinnarizine solubility resulting from pH changes in the acidic UWL alone (i.e. an increase in

ionisation and solubility, leading to a decrease in thermodynamic activity), are therefore unlikely

to promote drug absorption and are unable to explain the in vivo data in Figure 5.2B.

However, a second major change to colloidal structure was expected on entry into the UWL in

vivo, namely a significant reduction in micellar lipid content due to the promotion of LCFA

micellar dissociation and absorption. At lower pHs in vitro, decreases in micellar lipid content

were also stimulated by phase separation out of the micellar phase (as indicated by asterisks in

Figure 5.4A, Figure 5.5). Thus, during passage of LCFA colloids across the acidic UWL in vivo,

the lipid content of the micellar phase is expected to drop due to lipid absorption and potentially

due to lipid phase separation, although in vivo, in the presence of a large absorption sink, phase

separation may be less prevalent.

Since cinnarizine solubility in intestinal colloids is highly dependent on the concentration of

incorporated lipids (cinnarizine solubility in LCFA colloids decreased dramatically with

decreasing lipid load at each pH studied - Figure 5.4A), decreasing lipid concentration is likely to

be the predominant factor in determining cinnarizine solubilisation at the UWL. Therefore, in

predicting the impact of both a decrease in pH and a coincident decrease in micellar lipid

concentration in the UWL on drug solubilisation (Figure 5.4A), the net change to the solubility of

cinnarizine in the aqueous phase is likely to be a reduction due to the loss in lipid content. In the

absence of precipitation, this will drive supersaturation and enhance drug thermodynamic activity,

consistent with the significant increase in drug flux seen in vivo under conditions where lipid

absorption was enhanced (Figure 5.2A & B).


227

The ability of the LCFA colloids to harness acidic microclimate-induced drug supersaturation

therefore stems from the ability of LCFA to respond to a decrease in pH at the UWL, leading to

enhanced LCFA absorption, subsequently decreasing colloid solubilisation capacity for PWSD.

That is, the presence of oleic acid confers pH sensitivity to LCFA-containing colloids and serves

as a trigger for acidic microclimate-induced drug supersaturation, ultimately enabling the

absorption of solubilised drug via enhancements in thermodynamic activity. Similar events might

also be expected for non-ionisable, but absorbable micellar components such as monoglycerides,

although in this case, lipid absorption is not stimulated by changes in pH and therefore the effects

of acidic microclimate attenuation on drug absorption may not be as significant. Conversely, for

colloids that lack an absorbable (and/or pH-responsive) component, potential increases in drug

thermodynamic activity due to entry into the acidic microclimate are unlikely, and thermodynamic

activity (and absorption) of drug is expected to remain constant (or to decrease as drug is absorbed)

during colloidal passage across the UWL. The data obtained here for cinnarizine absorption from

Brij 97 colloids (a non-ionisable, non-absorbable surfactant) is consistent with this suggestion

(Figure 5.2C). Indeed, when the acidic microclimate was attenuated, and when lipid absorption

was effectively inhibited, cinnarizine absorption from LCFA and Brij 97 micelles was comparable

and low (Figure 5.2B & C).

The intercalation of lipid digestion products into bile-derived intestinal colloids has previously

been shown to increase solubilisation capacity for a range of PWSD39, 91, 99. Therefore, although

cinnarizine was used as a model PWSD in the current study, acidic microclimate-induced drug

supersaturation is likely to be a common endogenous mechanism of absorption enhancement for

lipophilic PWSD after co-administration with absorbable lipids. The degree of supersaturation that

is induced by entry into the acidic microclimate will be dictated by the sensitivity of drug

solubility to micellar lipid content, and the efficiency of lipid absorption. Under these


228

circumstances the combination of weak bases, with LBF comprising fatty acids or glyceride that

generate fatty acids in situ, may be particularly beneficial since the colloid solubility of weak

bases in structures containing fatty acids is high (refer Chapter 4), and the efficiency of fatty acid

absorption is also high.

Previously, we have shown that the interaction of biliary components (i.e. bile salts, phospholipid,

cholesterol) with similar lipid colloidal phases to those examined here leads to changes in colloid

microstructure that dramatically reduce cinnarizine solubilisation capacity (Chapters 3 and 4).

Similar to the events described in the UWL here, precipitation in the previous studies was delayed

sufficiently that supersaturation occurred in the bulk luminal fluids and ultimately enhanced the

intestinal absorptive flux of cinnarizine. In an attempt to link the current studies at the UWL, with

these previous studies that describe luminal supersaturation, a final experiment was performed to

explore the importance of a functioning acidic microclimate on increases in drug absorption

resulting from supersaturation in the bulk luminal fluids. From the data in Figure 5.6B it is

apparent that the increase in cinnarizine absorptive flux that occurred due to interaction of LCFA-

containing colloids with bile (and the period of drug supersaturation in the luminal fluids that

ensued), were abolished when the acidic microclimate was inhibited with amiloride. It seems

likely therefore that even where supersaturation is generated in the intestinal lumen by, for

example, initiation of lipid digestion4, 91, 176 or interaction with bile, micellar structures are still

required to promote transport across the UWL, and the acidity of the UWL is critical to allowing

efficient absorption of LCFA and PWSD from these micellar structures. Thus, on approach to the

absorptive surface, LFCA protonation and absorption is required to translate luminal

supersaturation into enhanced drug absorption, presumably via enhancements in thermodynamic

activity that enable drug absorption from the supersaturated solubilised reservoir.


229

As a final caveat, the generation of drug supersaturation during the processing of LBF is not

always expected to be beneficial. Indeed, supersaturation is also a precursor to crystal nucleation

and drug precipitation, events that are likely to reduce drug absorption for PWSD with

dissolution-rate limited absorption. The benefits of supersaturation generation must therefore be

weighed against the potential for precipitation. However, supersaturation at the UWL may be

particularly effective in enhancing drug absorption from intestinal colloids, as proximity to the

absorptive membrane is likely to provide an effective sink for effective removal (i.e. absorption)

of supersaturated drug, thereby minimising the risk of drug precipitation. Similar concepts in non-

lipid based systems have recently been elegantly exemplified by the Augustijns group where the

likelihood of drug precipitation from supersaturated solutions (formed in this case by solvent shift)

was dramatically reduced by the presence of an absorptive membrane261. The viscous, mucus layer

that is present at the UWL may also play a role in stabilising drug supersaturation, as increased

viscosity has been suggested to delay nucleation and crystal growth of solutes from supersaturated

solutions262, 263.

5.6 CONCLUSION

Lipid absorption is an endogenous mechanism that triggers drug supersaturation and facilitates

effective drug absorption from intestinal mixed micelles. Formulations containing absorbable

lipids or the pre-digestive precursors to absorbable lipids may therefore be more effective in

enhancing PWSD absorption when compared to formulations that lack absorbable components.

The degree of drug supersaturation (and therefore absorption enhancement potential) generated by

lipid absorption is likely maximised under conditions where the dependency of drug solubility on

micellar lipid content is high, and where lipid absorption is highly efficient. Combinations of drug

and excipients that are formulated according to the above paradigm (e.g. weak bases and lipid-


230

based formulations comprising precursors of fatty acids) may therefore maximise the absorptive

benefits associated with lipid absorption-induced drug supersaturation. These data reinforce

previous suggestions that lipid-based formulations are able to interact with the dynamic GI

environment, and in doing so promote both drug solubilisation and thermodynamic activity. This

provides unique benefits for drug absorption when compared to formulations that promote

solubilisation or supersaturation alone.

231

CHAPTER 6 : SUMMARY AND

PERSPECTIVES

Chapter 6: Summary and perspectives

232

The ability of lipids in food or lipid-based formulations (LBF) to enhance the oral bioavailability

of poorly water-soluble drugs (PWSD) has long been recognised. However, confident application

of LBF, and accurate simulation or prediction of oral drug absorption from LBF, is limited by the

lack of a holistic understanding of the mechanisms by which lipids enhance the absorption of

PWSD. The studies described in this thesis aimed to address this gap in understanding, and

focussed on elucidating the mechanism of drug absorption from the colloidal micellar and

vesicular species that are formed in the gastrointestinal (GI) tract during lipid digestion.

Co-administration with lipids promotes drug solubilisation within intestinal lipid colloidal phases

and increases the apparent solubility of PWSD in the small intestine (SI). The mechanism of drug

absorption from these solubilising species, however, remains poorly-defined. Current

understanding suggests that drug flux across absorptive membrane is the product of the free drug

concentration and drug permeability across the membrane. For solubilised systems, therefore, drug

solubilised within colloidal phases serves to rapidly replenish the free drug fraction via re-

establishment of the equilibrium between solubilised and free drug. This results in a significant

increase in dissolution rate when compared to dissolution from a typical solid dose form. Since

solubilisation does not directly enhance the inter-micellar free drug concentration, however, lipid

co-administration may not increase membrane flux for drugs where absorption is solubility-

limited (rather than dissolution-limited). Indeed, it has recently been suggested that solubilising

formulation approaches result in a ‘solubility-permeability interplay’ where potential

enhancements in drug absorption due to solubilisation are compromised by decreases in free drug

fraction100, 146. This suggestion, however, runs counter to a body of evidence that demonstrates that

solubilising formulations such as LBF commonly increase the oral bioavailability of PWSD87.

Whilst the advantages of LBF may stem from increases in dissolution rate alone, the available data

also suggest that drug flux in excess of that predicted by the free drug concentration is likely. This


233

apparent contradiction between robust drug absorption in the presence of LBF or food, and the

realisation that this occurs even though the free drug concentration is often extremely low,

stimulated the work described in this thesis. The principle hypothesis that underpins the work is

that drug absorption from intestinal colloidal species is not simply a function of the free drug

concentration in equilibrium with the solubilised reservoir. The alternative hypotheses that have

been explored include the possibility that drug absorption occurs directly from the solubilised

fraction and/or that transient changes to colloidal structure in situ may lead to colloidal

supersaturation and an increase in the free drug concentration that subsequently increases drug

absorption.

In Chapter 3 the prospect for the solubilised fraction to contribute directly to drug absorption has

been explored via examination of the potential role of collisional drug absorption. This was

investigated by comparing the intestinal absorption of a model PWSD, cinnarizine (CIN), from

micelles and vesicles that had equal CIN solubilisation capacities and were loaded with drug at the

same concentration and therefore that had the same thermodynamic activity/free concentration.

The micelles and vesicles employed were assembled using differing quantities of medium-chain

fatty acid and monoglyceride solubilised in simulated intestinal fluid. Whilst both species had

identical solubilising properties, the structure of the two colloids was notably different and the

particle size of the vesicles was significantly larger than that of the micelles (443 nm vs. 9 nm

respectively). Since the rate of collisional transfer is directly related to the number of particles,

collisional absorption of CIN from micelles was expected to be higher since the number of

micellar particles was substantially higher than that of vesicular particles. In contrast, the data in

Chapter 3 show that intestinal absorption of CIN from vesicles and micelles was the same when

the free concentration was matched, suggesting limited collisional transfer and instead reliance on

the free concentration.


234

Receptor-mediated collisional absorption was also investigated by assessing CIN bioavailability in

the absence and presence of inhibitors of common lipid uptake transporters (e.g. SR-BI, CD36,

NPC1L1). These transporters were hypothesised to interact directly with intestinal colloids to

mediate the absorption of solubilised lipid and drug since they have similar interactions with other

colloidal particles (e.g. HDL) in the blood161. Consistent with the micelle/vesicle data, however,

no significant differences in CIN absorption were seen in the presence and absence of the

inhibitors, providing further support for the suggestion that collision-mediated uptake was not a

significant driver for drug absorption from intestinal lipid colloidal phases, and that drug

absorption occurred largely from the free fraction.

Subsequently, attention turned to the possibility that drug absorption from intestinal colloidal

phases may be promoted, not by direct absorption from the solubilised phase, but by boosting

thermodynamic activity, thereby rendering solubilised drug more available for absorption via the

free fraction. In the first instance, the potential for increases in thermodynamic activity based on

the generation of supersaturation due to interaction with biliary fluids was explored. Bile dilution

was hypothesised to stimulate drug supersaturation since the addition of bile to intestinal lipid

colloidal phases has previously been shown to generate colloids that are less lipid-rich with lower

solubilisation capacities39, 194. To investigate the impact of biliary dilution on drug absorption from

lipid colloidal phases, fasted bile was collected from donor rats and added to CIN-loaded (CIN

was loaded at sub-saturated concentrations) micelles and vesicles in a 1:1 v/v ratio. Bile addition

resulted in CIN supersaturation, and supersaturation was maintained for longer time periods in

micellar systems when compared to vesicles. The ability of bile-induced supersaturation to

enhance drug absorption was subsequently evaluated in situ and in vivo via the assessment of CIN

absorptive flux from colloids with and without the co-perfusion of donor bile in rat jejunal

perfusion studies, and the assessment of CIN bioavailability in bile diverted vs. bile-intact rats.


235

The results showed that bile-induced supersaturation led to enhanced intestinal absorption and

systemic exposure of CIN from micelles, but failed to enhance the absorption of CIN from

vesicles due to rapid precipitation. Thus, bile-induced supersaturation was identified as a novel

mechanism to increase drug thermodynamic activity and to better mobilise the solubilised drug

fraction for absorption.

The ability of bile to enhance the absorption of PWSD is well-described in the literature, however

the mechanism by which this occurs, has previously been assumed to reflect increases in drug

solubilisation. The novelty of the data in chapter 3 therefore lies in showing that bile may also

enhance drug absorption via promotion of drug supersaturation. The difference in the absorption

enhancement of CIN from micelles and vesicles also highlights the need to achieve an optimal

balance between drug supersaturation and precipitation.

Chapter 4 extended the findings in Chapter 3 to include micellar systems containing long-chain

lipids (which have the added advantage of increasing relevance beyond LBF to also include

typical food-related lipids), and to a range of drugs with differing physicochemical properties.

Initially, the sensitivity of the solubility of five PWSD (2 basic drugs, 2 neutral drugs, 1 acidic

drug) to colloidal bile or long-chain lipid content was evaluated. Consistent with the data obtained

with CIN and the medium-chain lipid containing colloids in Chapter 3, the solubility of the basic

drugs in long-chain lipid containing micelles was found to decrease with increasing bile

concentration. In contrast, and more consistent with typical solubilisation behaviour, the solubility

of the neutral and acidic drugs increased with increasing bile (surfactant) concentration. The data

suggest that the potential for drug supersaturation to be triggered during bile dilution is highest for

basic drugs. The basis for supersaturation stimulation was suggested to be related to drug affinity

for the core lipids within the colloids. Thus, high affinity for colloidal lipids resulted in high initial


236

solubility, and the addition of bile served to disrupt intermolecular interactions between drug and

lipid, leading to reductions in solubilisation capacity. Consistent with this suggestion, in situ

evaluation of the absorption enhancement afforded by 1:1 v/v co-perfusion of bile with drug-

containing colloids showed that bile enhanced the absorption of cinnarizine (a basic drug) to a

greater extent than danazol (a neutral drug).

Interestingly, however, bile did stimulate some increase in danazol absorption, in spite of a

reduction in thermodynamic activity (bile addition increased danazol solubility in colloidal species

thereby decreasing thermodynamic activity). Since oleic acid absorption was also increased by

bile co-perfusion, the increase in danazol absorption was suggested to be due to the increase in

lipid absorption in the presence of bile. Thus lipid absorption was hypothesised to reduce micellar

drug solubilisation capacity, and therefore to increase thermodynamic activity and absorption

(assuming that absorption enhancement precedes drug precipitation).

In Chapter 5, the potential for lipid absorption (i.e. a reduction in colloidal lipid content) to

stimulate drug supersaturation and absorption was examined directly. The acidic microclimate of

the intestinal unstirred water layer (UWL) facilitates the absorption of long-chain fatty acids

(LCFA) by protonating LCFA and increasing thermodynamic activity and membrane

partitioning56, 57. In this series of experiments, therefore, the acidity of the UWL was attenuated

using amiloride in order to inhibit oleic acid absorption. This enabled a comparison of CIN

absorption from oleic acid-containing colloids under conditions of normal lipid absorption vs.

conditions of inhibited lipid absorption. In the rat jejunal perfusion model the absorptive flux of

CIN was dramatically attenuated (6.5-fold) when oleic acid absorption was suppressed. In vitro

assessment of CIN solubilisation behaviour under conditions that simulate lipid absorption at the

UWL subsequently indicated that supersaturation was likely to be responsible for the enhancement


237

in CIN absorption during normal lipid absorption. Importantly, in this chapter, it was shown that

the intrinsic solubilisation capacity of colloids did not dictate the efficiency of drug absorption.

Instead, the ability of colloids to lose solubilisation capacity and to generate drug supersaturation

at the UWL was seemingly more important. Thus, head-to-head comparison of CIN absorption

from two colloidal systems with equal solubilisation capacity (colloids were loaded with CIN at

the same concentration such that total, solubilised, and free concentrations were equal) showed

that drug absorption was dramatically more efficient when micelles comprised bile salt and oleic

acid, when compared to micelles comprising Brij 97 (a non-ionisable and non-absorbable

surfactant). These differences were abolished when oleic acid absorption was inhibited. The data

support the importance of lipid absorption-induced supersaturation as a means to mobilise the

solubilised drug fraction and to enable effective drug absorption from intestinal mixed micellar

species.

The major findings from this thesis therefore support a consensus view that LBF inherently

generate drug supersaturation when incorporated into endogenous lipid processing pathways.

Specifically, supersaturation has previously been shown to be generated when solubilisation

capacity is lost during LBF dispersion and digestion5, 91, 176, 177, 193. The current studies further

suggest that supersaturation is generated during biliary dispersion of lipid digestion products, and

during the absorption of lipid digestion products. The role of bile and the acidic UWL in

promoting drug supersaturation and absorption has not been described previously, and represent

interesting new avenues for further research. Thus, LBF have traditionally been regarded as

formulations that promote the absorption of PWSD by increasing solubilisation. The current data,

however, suggest that in vivo, LBF are capable of promoting both drug solubilisation and drug

supersaturation.


238

The unique ability of LBF to simultaneously increase the solubilised reservoir and to also enhance

the thermodynamic activity of co-administered drug may distinguish LBF from other enabling

formulations and help explain why lipids are often very effective in enhancing the oral absorption

of a range of lipophilic PWSD. Firstly, the formation of intestinal mixed micelles and vesicles

during lipid digestion increases the solubilisation capacity of the SI, and allows for higher

amounts of drug to be maintained in solution. Secondly, although drug thermodynamic activity is

initially lowered by solubilisation, the ensuing interaction between drug-containing colloids and

endogenous lipid processing pathways triggers drug supersaturation, and provides a means to

reverse the reduction in thermodynamic activity and to mobilise the solubilised drug fraction for

absorption. Lastly, the multi-step, sequential manner by which drug supersaturation is triggered in

the SI, coupled with the increased drug solubilisation capacity of the SI after lipid co-

administration, ensures that the supersaturation generated during LBF processing is rate and

extent-controlled, therefore providing a balance between solubilisation and supersaturation that

maximises the potential for drug absorption while minimising the risk of drug precipitation. In this

way, LBF possess inherent advantages when compared to other enabling formulations that either

promote drug solubilisation only (e.g. cyclodextrins) or drug supersaturation only (e.g. solid

dispersions), and may provide a means to finely balance the need to solubilise the drug dose, and

to maintain or increase drug thermodynamic activity.

The findings from this thesis provide significant new insights into the fundamental mechanisms by

which lipids enhance the oral absorption of PWSD. The studies show that supersaturation is an

important driving force for drug absorption after co-administration with lipids, and identify two

novel mechanisms (i.e. bile dilution and lipid absorption) by which drug supersaturation may be

triggered during intestinal lipid processing. Better understanding of the mechanisms of

supersaturation generation is expected to contribute to an increasingly rational basis for the


239

selection of lipid excipients. Although more work is needed, practical applications related to the

findings of this thesis are conceivable, and relate to the opportunity to generate colloidal

assemblies with built-in sensitivity to bile dilution and lipid absorption mediated supersaturation.

For example, the studies in Chapter 4 suggest that the use of lipid-drug combinations where drug

possesses high affinity for lipid digestion products may predispose post-digestion colloids to lose

solubilisation capacity during bile dilution via the disruption of drug-lipid intermolecular

interactions. Studies in Chapter 5 further suggest that the utilisation of absorbable lipids (or

precursors of absorbable lipids) may lead to potent stimulation of drug supersaturation at the UWL,

especially where the drug has high affinity for the lipids and where lipid absorption is efficient.

Collectively, these studies suggest that combinations of drug and excipients that are formulated

according to the above principles (e.g. weak bases with LBF comprising precursors of fatty acids)

may maximally harness the absorptive benefits associated with bile and lipid absorption-induced

drug supersaturation. The in vitro studies in Chapter 4 that delineate drug solubility relationships

with bile components or lipid digestion products also serve to demonstrate that the likelihood of

supersaturation generation in vivo may be predicted in vitro, and may be amenable to high

throughput screening.

Finally, the data in Chapter 5 demonstrate a crucial role of the UWL acidic microclimate in the

translation of luminal supersaturation into enhanced absorption (inhibition of UWL acidity

blocked increases in drug absorption due to bile-mediated promotion of luminal supersaturation).

This suggests that supersaturated luminal colloids alone are not sufficient to promote drug

absorption and that micellar dissociation via a decrease in UWL pH is an important mechanism to

mobilise supersaturated solubilised drug for absorption. This may be an important consideration

for solubilising formulations that lack pH-sensitivity.


240

In conclusion, this thesis has demonstrated for the first time that effective drug absorption from

intestinal colloidal species is a result of dynamic interactions between the solubilised drug

reservoir and the GI environment that lead to multiple opportunities for supersaturation-enhanced

absorption. This improved understanding of the mechanism of drug absorption is ultimately

expected to inform rational design criteria for LBF and to improve oral absorption simulation

models. Future work might be usefully directed towards further characterising the parameters that

dictate the propensity of LBF to generate drug supersaturation during formulation processing, and

in particular an assessment of the balance between drug supersaturation and drug precipitation.

241

REFERENCES

1. Lipinski, C. A. Drug-like properties and the causes of poor solubility and poor permeability. Journal of Pharmacological and Toxicological Methods 2000, 44, (1), 235-249.

2. Williams, H. D.; Trevaskis, N. L.; Charman, S. A.; Shanker, R. M.; Charman, W. N.; Pouton, C. W.; Porter, C. J. H. Strategies to address low drug solubility in discovery and development. Pharmacological Reviews 2013, 65, (1), 315-499.

3. Amidon, G. L.; Lennernäs, H.; Shah, V. P.; Crison, J. R. A theoretical basis for a biopharmaceutic drug classification: The correlation of in vitro drug product dissolution and in vivo bioavailability. Pharmaceutical Research 1995, 12, (3), 413-420.

4. Williams, H. D.; Anby, M. U.; Sassene, P.; Kleberg, K.; Bakala-N’Goma, J.-C.; Calderone, M.; Jannin, V.; Igonin, A.; Partheil, A.; Marchaud, D.; Jule, E.; Vertommen, J.; Maio, M.; Blundell, R.; Benameur, H.; Carrière, F.; Müllertz, A.; Pouton, C. W.; Porter, C. J. H. Toward the establishment of standardized in vitro tests for lipid-based formulations. 2. The effect of bile salt concentration and drug loading on the performance of Type I, II, IIIA, IIIB, and IV formulations during in vitro digestion. Molecular Pharmaceutics 2012, 9, (11), 3286-3300.

5. Pouton, C. W.; Porter, C. J. H. Formulation of lipid-based delivery systems for oral administration: Materials, methods and strategies. Advanced Drug Delivery Reviews 2008, 60, (6), 625-637.

6. Tokumura, T.; Tsushima, Y.; Tatsuishi, K.; Kayano, M.; Machida, Y.; Nagai, T. Enhancement of the oral bioavailability of cinnarizine in oleic acid in beagle dogs. Journal of Pharmaceutical Sciences 1987, 76, (4), 286-288.

7. Myers, R. A.; Stella, V. J. Systemic bioavailability of penclomedine (NSC-338720) from oil-in-water emulsions administered intraduodenally to rats. International Journal of Pharmaceutics 1992, 78, (1–3), 217-226.

8. Charman, W. N.; Rogge, M. C.; Boddy, A. W.; Berger, B. M. Effect of food and a monoglyceride emulsion formulation on danazol bioavailability. The Journal of Clinical Pharmacology 1993, 33, (4), 381-386.

9. Dahan, A.; Hoffman, A. The effect of different lipid based formulations on the oral absorption of lipophilic drugs: The ability of in vitro lipolysis and consecutive ex vivo intestinal permeability data to predict in vivo bioavailability in rats. European Journal of Pharmaceutics and Biopharmaceutics 2007, 67, (1), 96-105.

10. Bates, T. R.; Sequeira, J. A. Bioavailability of micronized griseofulvin from corn oil-in-water emulsion, aqueous suspension, and commercial tablet dosage forms in humans. Journal of Pharmaceutical Sciences 1975, 64, (5), 793-797.

11. Hauss, D. J.; Fogal, S. E.; Ficorilli, J. V.; Price, C. A.; Roy, T.; Jayaraj, A. A.; Keirns, J. J. Lipid-based delivery systems for improving the bioavailability and lymphatic transport of a poorly water-soluble LTB4 inhibitor. Journal of Pharmaceutical Sciences 1998, 87, (2), 164-169.

12. Caliph, S. M.; Charman, W. N.; Porter, C. J. H. Effect of short-, medium-, and long-chain fatty acid-based vehicles on the absolute oral bioavailability and intestinal lymphatic transport of halofantrine and assessment of mass balance in lymph-cannulated and non-cannulated rats. Journal of Pharmaceutical Sciences 2000, 89, (8), 1073-1084.

242

13. Wei, L.; Sun, P.; Nie, S.; Pan, W. Preparation and evaluation of SEDDS and SMEDDS containing carvedilol. Drug Development & Industrial Pharmacy 2005, 31, (8), 785-794.

14. Mueller, E. A.; Kovarik, J. M.; van Bree, J. B.; Tetzloff, W.; Grevel, J.; Kutz, K. Improved dose linearity of cyclosporine pharmacokinetics from a microemulsion formulation. Pharmaceutical Research 1994, 11, (2), 301-304.

15. Woo, J. S.; Song, Y.-K.; Hong, J.-Y.; Lim, S.-J.; Kim, C.-K. Reduced food-effect and enhanced bioavailability of a self-microemulsifying formulation of itraconazole in healthy volunteers. European Journal of Pharmaceutical Sciences 2008, 33, (2), 159-165.

16. Wu, W.; Wang, Y.; Que, L. Enhanced bioavailability of silymarin by self-microemulsifying drug delivery system. European Journal of Pharmaceutics and Biopharmaceutics 2006, 63, (3), 288-294.

17. Gao, P.; Rush, B. D.; Pfund, W. P.; Huang, T.; Bauer, J. M.; Morozowich, W.; Kuo, M.-S.; Hageman, M. J. Development of a supersaturable SEDDS (S-SEDDS) formulation of paclitaxel with improved oral bioavailability. Journal of Pharmaceutical Sciences 2003, 92, (12), 2386-2398.

18. Pouton, C. W. Lipid formulations for oral administration of drugs: non-emulsifying, self-emulsifying and ‘self-microemulsifying’ drug delivery systems. European Journal of Pharmaceutical Sciences 2000, 11, Supplement 2, (0), S93-S98.

19. Pouton, C. W. Formulation of poorly water-soluble drugs for oral administration: Physicochemical and physiological issues and the lipid formulation classification system. European Journal of Pharmaceutical Sciences 2006, 29, (3–4), 278-287.

20. Hauss, D. J. Oral lipid-based formulations. Advanced Drug Delivery Reviews 2007, 59, (7), 667-676.

21. Yoshiya, Y.; Takeshi, N.; Hiroshi, T.; Tadao, M. Biopharmaceutical studies of lipid-containing oral dosage forms: Relationship between drug absorption rate and digestibility of vehicles. International Journal of Pharmaceutics 1979, 3, (1), 23-31.

22. Borovicka, J.; Schwizer, W.; Mettraux, C.; Kreiss, C.; Remy, B.; Asal, K.; Jansen, J. B.; Douchet, I.; Verger, R.; Fried, M. Regulation of gastric and pancreatic lipase secretion by CCK and cholinergic mechanisms in humans. American Journal of Physiology - Gastrointestinal and Liver Physiology 1997, 273, (2), G374-G380.

23. Moreau, H.; Bernadac, A.; Gargouri, Y.; Benkouka, F.; Laugier, R.; Verger, R. Immunocytolocalization of human gastric lipase in chief cells of the fundic mucosa. Histochemistry 1989, 91, (5), 419-423.

24. Hamosh, M.; Scanlon, J. W.; Ganot, D.; Likel, M.; Scanlon, K. B.; Hamosh, P. Fat digestion in the newborn. Characterization of lipase in gastric aspirates of premature and term infants. Journal of Clinical Investigation 1981, 67, (3), 838-846.

25. Abrams, C. K.; Hamosh, M.; Lee, T. C.; Ansher, A. F.; Collen, M. J.; Lewis, J. H.; Benjamin, S. B.; Hamosh, P. Gastric lipase: localization in the human stomach. Gastroenterology 1988, 95, (6), 1460-1464.

26. Ransac, S.; Rogalska, E.; Gargouri, Y.; Deveer, A. M.; Paltauf, F.; de Haas, G. H.; Verger, R. Stereoselectivity of lipases. I. Hydrolysis of enantiomeric glyceride analogues by gastric and pancreatic

243

lipases, a kinetic study using the monomolecular film technique. Journal of Biological Chemistry 1990, 265, (33), 20263-70.

27. Canaan, S.; Roussel, A.; Verger, R.; Cambillau, C. Gastric lipase: Crystal structure and activity. Biochimica et Biophysica Acta (BBA) - Molecular and Cell Biology of Lipids 1999, 1441, (2–3), 197-204.

28. Armand, M.; Borel, P.; Pasquier, B.; Dubois, C.; Senft, M.; Andre, M.; Peyrot, J.; Salducci, J.; Lairon, D. Physicochemical characteristics of emulsions during fat digestion in human stomach and duodenum. American Journal of Physiology - Gastrointestinal and Liver Physiology 1996, 271, (1), G172-G183.

29. Jensen, M. S.; Jensen, S. K.; Jakobsen, K. Development of digestive enzymes in pigs with emphasis on lipolytic activity in the stomach and pancreas. Journal of Animal Science 1997, 75, (2), 437-45.

30. Carey, M. C.; Small, D. M.; Bliss, C. M. Lipid digestion and absorption. Annual Review of Physiology 1983, 45, (1), 651-677.

31. Ohtani, N.; Sasaki, I.; Naito, H.; Shibata, C.; Matsuno, S. Mediators for fat-induced ileal brake are different between stomach and proximal small intestine in conscious dogs. J Gastrointest Surg 2001, 5, (4), 377-382.

32. Van Citters, G. W.; Lin, H. C. The ileal brake: A fifteen-year progress report. Curr Gastroenterol Rep 1999, 1, (5), 404-409.

33. Nordskog, B. K.; Phan, C. T.; Nutting, D. F.; Tso, P. An examination of the factors affecting intestinal lymphatic transport of dietary lipids. Advanced Drug Delivery Reviews 2001, 50, (1–2), 21-44.

34. Mu, H.; Høy, C.-E. The digestion of dietary triacylglycerols. Progress in Lipid Research 2004, 43, (2), 105-133.

35. Ros, E. Intestinal absorption of triglyceride and cholesterol. Dietary and pharmacological inhibition to reduce cardiovascular risk. Atherosclerosis 2000, 151, (2), 357-379.

36. Lowe, M. E. Structure and function of pancreatic lipase and colipase. Annual review of nutrition 1997, 17, 141-158.

37. Borgström, B.; Dahlqvist, A.; Lundh, G.; Sjövall, J. Studies of intestinal digestion and absorption in the human. Journal of Clinical Investigation 1957, 36, (10), 1521-1536.

38. Meyer, J. G. Lipolytic enzymes of the human pancreas. II. Purification and properties of cholesterol ester hydrolase. Biochimica et Biophysica Acta (BBA) - Lipids and Lipid Metabolism 1989, 1002, (1), 89-92.

39. Kossena, G. A.; Charman, W. N.; Boyd, B. J.; Dunstan, D. E.; Porter, C. J. H. Probing drug solubilization patterns in the gastrointestinal tract after administration of lipid-based delivery systems: A phase diagram approach. Journal of Pharmaceutical Sciences 2004, 93, (2), 332-348.

40. Salentinig, S.; Sagalowicz, L.; Leser, M. E.; Tedeschi, C.; Glatter, O. Transitions in the internal structure of lipid droplets during fat digestion. Soft Matter 2011, 7, (2), 650-661.

244

41. Warren, D. B.; Anby, M. U.; Hawley, A.; Boyd, B. J. Real time evolution of liquid crystalline nanostructure during the digestion of formulation lipids Using synchrotron small-angle X-ray scattering. Langmuir 2011, 27, (15), 9528-9534.

42. Fatouros, D. G.; Deen, G. R.; Arleth, L.; Bergenstahl, B.; Nielsen, F. S.; Pedersen, J. S.; Mullertz, A. Structural development of self nano emulsifying drug delivery systems (SNEDDS) during in vitro lipid digestion monitored by small-angle X-ray scattering. Pharmaceutical Research 2007, 24, (10), 1844-1853.

43. Kossena, G. A.; Charman, W. N.; Boyd, B. J.; Porter, C. J. H. A novel cubic phase of medium chain lipid origin for the delivery of poorly water soluble drugs. Journal of Controlled Release 2004, 99, (2), 217-229.

44. Nguyen, T.-H.; Hanley, T.; Porter, C. J. H.; Boyd, B. J. Nanostructured liquid crystalline particles provide long duration sustained-release effect for a poorly water soluble drug after oral administration. Journal of Controlled Release 2011, 153, (2), 180-186.

45. Shiau, Y. F.; Fernandez, P.; Jackson, M. J.; McMonagle, S. Mechanisms maintaining a low-pH microclimate in the intestine. American Journal of Physiology - Gastrointestinal and Liver Physiology 1985, 248, (6), G608-G617.

46. Ikuma, M.; Hanai, H.; Kaneko, E.; Hayashi, H.; Hoshi, T. Effects of aging on the microclimate pH of the rat jejunum. Biochimica et Biophysica Acta (BBA) - Biomembranes 1996, 1280, (1), 19-26.

47. Li, C.-Y.; Zimmerman, C. L.; Wiedmann, T. S. Diffusivity of bile salt/phospholipid aggregates in mucin. Pharmaceutical Research 1996, 13, (4), 535-541.

48. Khanvilkar, K.; Donovan, M. D.; Flanagan, D. R. Drug transfer through mucus. Advanced Drug Delivery Reviews 2001, 48, (2–3), 173-193.

49. Norris, D. A.; Sinko, P. J. Effect of size, surface charge, and hydrophobicity on the translocation of polystyrene microspheres through gastrointestinal mucin. Journal of Applied Polymer Science 1997, 63, (11), 1481-1492.

50. Sanders, N. N.; De Smedt, S. C.; Demeester, J. The physical properties of biogels and their permeability for macromolecular drugs and colloidal drug carriers. Journal of Pharmaceutical Sciences 2000, 89, (7), 835-849.

51. Westergaard, H.; Dietschy, J. M. The mechanism whereby bile acid micelles increase the rate of fatty acid and cholesterol uptake into the intestinal mucosal cell. Journal of Clinical Investigation 1976, 58, (1), 97-108.

52. Carey, M. C.; Small, D. M. The characteristics of mixed micellar solutions with particular reference to bile. The American Journal of Medicine 1970, 49, (5), 590-608.

53. Garcıa, R. A.; Pantazatos, S. P.; Pantazatos, D. P.; MacDonald, R. C. Cholesterol stabilizes hemifused phospholipid bilayer vesicles. Biochimica et Biophysica Acta (BBA) - Biomembranes 2001, 1511, (2), 264-270.

54. Mansbach II, C. M.; Tso, P.; Kuksis, A.; eds, Intestinal lipid metabolism. Springer: 2000.

55. Israelachvili, J. N., Intermolecular and surface forces: Revised third edition. Academic press: 2011.

245

56. Shiau, Y. F. Mechanism of intestinal fatty acid uptake in the rat: the role of an acidic microclimate. The Journal of Physiology 1990, 421, (1), 463-474.

57. Shiau, Y. F.; Kelemen, R. J.; Reed, M. A. Acidic mucin layer facilitates micelle dissociation and fatty acid diffusion. American Journal of Physiology - Gastrointestinal and Liver Physiology 1990, 259, (4), G671-G675.

58. Chow, S. L.; Hollander, D. A dual, concentration-dependent absorption mechanism of linoleic acid by rat jejunum in vitro. Journal of Lipid Research 1979, 20, (3), 349-56.

59. Ling, K.-Y.; Lee, H.-Y.; Hollander, D. Mechanisms of linoleic acid uptake by rabbit small intestinal brush border membrane vesicles. Lipids 1989, 24, (1), 51-55.

60. Goré, J.; Hoinard, C.; Couet, C. Linoleic acid uptake by isolated enterocytes: Influence of α-linolenic acid on absorption. Lipids 1994, 29, (10), 701-706.

61. Nassir, F.; Wilson, B.; Han, X.; Gross, R. W.; Abumrad, N. A. CD36 is important for fatty acid and cholesterol uptake by the proximal but not distal intestine. Journal of Biological Chemistry 2007, 282, (27), 19493-19501.

62. Stahl, A.; Hirsch, D. J.; Gimeno, R. E.; Punreddy, S.; Ge, P.; Watson, N.; Patel, S.; Kotler, M.; Raimondi, A.; Tartaglia, L. A.; Lodish, H. F. Identification of the major intestinal fatty acid transport protein. Molecular Cell 1999, 4, (3), 299-308.

63. Bietrix, F.; Yan, D.; Nauze, M.; Rolland, C.; Bertrand-Michel, J.; Coméra, C.; Schaak, S.; Barbaras, R.; Groen, A. K.; Perret, B.; Tercé, F.; Collet, X. Accelerated lipid absorption in mice overexpressing intestinal SR-BI. Journal of Biological Chemistry 2006, 281, (11), 7214-7219.

64. Stremmel, W. Uptake of fatty acids by jejunal mucosal cells is mediated by a fatty acid binding membrane protein. Journal of Clinical Investigation 1988, 82, (6), 2001-2010.

65. Schoeller, C.; Keelan, M.; Mulvey, G.; Stremmel, W.; Thomson, A. B. R. Role of a brush border membrane fatty acid binding protein in oleic acid uptake into rat and rabbit jejunal brush border membrane. Clinical and Investigative Medicine 1995, 18, (5), 380-380.

66. Van Bennekum, A.; Werder, M.; Thuahnai, S. T.; Han, C.-H.; Duong, P.; Williams, D. L.; Wettstein, P.; Schulthess, G.; Phillips, M. C.; Hauser, H. Class B Scavenger Receptor-mediated intestinal absorption of dietary β-carotene and cholesterol. Biochemistry 2005, 44, (11), 4517-4525.

67. Hauser, H.; Dyer, J. H.; Nandy, A.; Vega, M. A.; Werder, M.; Bieliauskaite, E.; Weber, F. E.; Compassi, S.; Gemperli, A.; Boffelli, D.; Wehrli, E.; Schulthess, G.; Phillips, M. C. Identification of a receptor mediating absorption of dietary cholesterol in the intestine. Biochemistry 1998, 37, (51), 17843-17850.

68. Altmann, S. W.; Davis, H. R.; Zhu, L.; Yao, X.; Hoos, L. M.; Tetzloff, G.; Iyer, S. P. N.; Maguire, M.; Golovko, A.; Zeng, M.; Wang, L.; Murgolo, N.; Graziano, M. P. Niemann-Pick C1 Like 1 protein is critical for intestinal cholesterol absorption. Science 2004, 303, (5661), 1201-1204.

69. Niot, I.; Poirier, H.; Tran, T. T. T.; Besnard, P. Intestinal absorption of long-chain fatty acids: Evidence and uncertainties. Progress in Lipid Research 2009, 48, (2), 101-115.

246

70. Mansbach, C. M.; Siddiqi, S. A. The biogenesis of chylomicrons. Annual Review of Physiology 2010, 72, (1), 315-333.

71. Johnston, J. M.; Rao, G. A. Triglyceride biosynthesis in the intestinal mucosa. Biochimica et Biophysica Acta (BBA) - Lipids and Lipid Metabolism 1965, 106, (1), 1-9.

72. Tso, P.; Karlstad, M. D.; Bistrian, B. R.; DeMichele, S. J. Intestinal digestion, absorption, and transport of structured triglycerides and cholesterol in rats. American Journal of Physiology - Gastrointestinal and Liver Physiology 1995, 268, (4), G568-G577.

73. Bloom, B.; Chaikoff, I. L.; Reinhardt, W. O. Intestinal lymph as pathway for transport of absorbed fatty acids of different chain lengths. American Journal of Physiology -- Legacy Content 1951, 166, (2), 451-455.

74. Chaikoff, I. L.; Bloom, B.; Stevens, B. P.; Reinhardt, W. O.; Dauben, W. G. Pentadecanoic acid-5-C14; its absorption and lymphatic transport. Journal of Biological Chemistry 1951, 190, (1), 431-435.

75. Kiyasu, J. Y.; Bloom, B.; Chaikoff, I. L. The portal transport of absorbed fatty acids. Journal of Biological Chemistry 1952, 199, (1), 415-419.

76. Porter, C. J. H.; Trevaskis, N. L.; Charman, W. N. Lipids and lipid-based formulations: optimizing the oral delivery of lipophilic drugs. Nat Rev Drug Discov 2007, 6, (3), 231-248.

77. Nilsson, Å. Intestinal absorption of lecithin and lysolecithin by lymph fistula rats. Biochimica et Biophysica Acta (BBA) - Lipids and Lipid Metabolism 1968, 152, (2), 379-390.

78. Sato, Y. The metabolic fate of lysolecithin administered into rat duodenal lumen. The Tohoku Journal of Experimental Medicine 1970, 100, (3), 277-287.

79. Scow, R. O.; Stein, Y.; Stein, O. Incorporation of dietary lecithin and lysolecithin into lymph chylomicrons in the rat. Journal of Biological Chemistry 1967, 242, (21), 4919-4924.

80. Le Kim, D.; Betzing, H. Intestinal absorption of polyunsaturated phosphatidylcholine in the rat. Hoppe Seylers Z Physiol Chem 1976, 357, (9), 1321-1331.

81. Ottolenghi, A. Estimation and subcellular distribution of lecithinase activity in rat intestinal mucosa. Journal of Lipid Research 1964, 5, (4), 532-537.

82. Clark, S. B.; Tercyak, A. M. Reduced cholesterol transmucosal transport in rats with inhibited mucosal acyl CoA: Cholesterol acyltransferase and normal pancreatic function. Journal of Lipid Research 1984, 25, (2), 148-159.

83. Gallo, L. L.; Newbill, T.; Hyun, J.; Vahouny, G. V.; Treadwell, C. R. Role of pancreatic cholesterol esterase in the uptake and esterification of cholesterol by isolated intestinal cells. Proceedings of the Society for Experimental Biology and Medicine. Society for Experimental Biology and Medicine (New York, N.Y.) 1977, 156, (2), 277-281.

84. Borja, C. R.; Vahouny, G. V.; Treadwell, C. R. Role of bile and pancreatic juice in cholesterol absorption and esterification. American Journal of Physiology -- Legacy Content 1964, 206, (1), 223-228.

247

85. Gao, P.; Guyton, M. E.; Huang, T.; Bauer, J. M.; Stefanski, K. J.; Lu, Q. Enhanced oral bioavailability of a poorly water soluble drug PNU-91325 by supersaturatable formulations. Drug Development and Industrial Pharmacy 2004, 30, (2), 221-229.

86. Carrier, R. L.; Miller, L. A.; Ahmed, I. The utility of cyclodextrins for enhancing oral bioavailability. Journal of Controlled Release 2007, 123, (2), 78-99.

87. Porter, C. J. H.; Pouton, C. W.; Cuine, J. F.; Charman, W. N. Enhancing intestinal drug solubilisation using lipid-based delivery systems. Advanced Drug Delivery Reviews 2008, 60, (6), 673-691.

88. Noyes, A. A.; Whitney, W. R. The rate of solution of solid substances in their own solutions. Journal of the American Chemical Society 1897, 19, (12), 930-934.

89. Naylor, L. J.; Bakatselou, V.; Dressman, J. B. Comparison of the mechanism of dissolution of hydrocortisone in simple and mixed micelle systems. Pharmaceutical Research 1993, 10, (6), 865-870.

90. Miyazaki, S.; Inoue, H.; Yamahira, T.; Nadai, T. Interaction of drugs with bile components. I. Effects of bile salts on the dissolution behavior of indomethacin and phenylbutazone. Chemical & pharmaceutical bulletin 1979, 27, (10), 2468-2472.

91. Kaukonen, A. M.; Boyd, B.; Porter, C.; Charman, W. Drug solubilization behavior during in vitro digestion of simple triglyceride lipid solution formulations. Pharmaceutical Research 2004, 21, (2), 245-253.

92. Hörter, D.; Dressman, J. B. Influence of physicochemical properties on dissolution of drugs in the gastrointestinal tract. Advanced Drug Delivery Reviews 1997, 25, (1), 3-14.

93. Crison, J. R.; Shah, V. P.; Skelly, J. P.; Amidon, G. L. Drug dissolution into micellar solutions: Development of a convective diffusion model and comparison to the film equilibrium model with application to surfactant-facilitated dissolution of carbamazepine. Journal of Pharmaceutical Sciences 1996, 85, (9), 1005-1011.

94. Balakrishnan, A.; Rege, B. D.; Amidon, G. L.; Polli, J. E. Surfactant-mediated dissolution: Contributions of solubility enhancement and relatively low micelle diffusivity. Journal of Pharmaceutical Sciences 2004, 93, (8), 2064-2075.

95. Bates, T. R.; Gibaldi, M.; Kanig, J. L. Solubilizing properties of bile salt solutions I: Effect of temperature and bile salt concentration on solubilization of glutethimide, griseofulvin, and hexestrol. Journal of Pharmaceutical Sciences 1966, 55, (2), 191-199.

96. Kassem, M. A.; Mattha, A. G.; El-Nimr, A. E. M.; Omar, S. M. Study of the influence of sodium taurocholate (STC) and sodium glycocholate (SGC) on the mass transfer of certain drugs. Digoxin. International Journal of Pharmaceutics 1982, 12, (1), 1-9.

97. Kararli, T. T.; Gupta, V. W. Solubilization and dissolution properties of a leucotriene-D4 antagonist in micellar solutions. Journal of Pharmaceutical Sciences 1992, 81, (5), 483-485.

98. Luner, P. E.; Babu, S. R.; Radebaugh, G. W. The effects of bile salts and lipids on the physicochemical behavior of gemfibrozil. Pharmaceutical Research 1994, 11, (12), 1755-1760.

248

99. Kossena, G. A.; Charman, W. N.; Boyd, B. J.; Porter, C. J. H. Influence of the intermediate digestion phases of common formulation lipids on the absorption of a poorly water-soluble drug. Journal of Pharmaceutical Sciences 2005, 94, (3), 481-492.

100. Miller, J. M.; Beig, A.; Krieg, B. J.; Carr, R. A.; Borchardt, T. B.; Amidon, G. E.; Amidon, G. L.; Dahan, A. The solubility–permeability interplay: Mechanistic modeling and predictive application of the impact of micellar solubilization on intestinal permeation. Molecular Pharmaceutics 2011, 8, (5), 1848-1856.

101. Sugano, K. Estimation of effective intestinal membrane permeability considering bile micelle solubilisation. International Journal of Pharmaceutics 2009, 368, (1–2), 116-122.

102. Heerklotz, H. Interactions of surfactants with lipid membranes. Quarterly reviews of biophysics 2008, 41, (3-4), 205-264.

103. Aungst, B. J. Intestinal permeation enhancers. Journal of Pharmaceutical Sciences 2000, 89, (4), 429-442.

104. Sakai, M.; Imai, T.; Ohtake, H.; Azuma, H.; Otagiri, M. Effects of absorption enhancers on the transport of model compounds in Caco-2 cell monolayers: Assessment by confocal laser scanning microscopy. Journal of Pharmaceutical Sciences 1997, 86, (7), 779-785.

105. Tomita, M.; Sawada, T.; Ogawa, T.; Ouchi, H.; Hayashi, M.; Awazu, S. Differences in the enhancing effects of sodium caprate on colonic and jejunal drug absorption. Pharmaceutical Research 1992, 9, (5), 648-653.

106. Chao, A. C.; Nguyen, J. V.; Broughall, M.; Recchia, J.; Kensil, C. R.; Daddona, P. E.; Fix, J. A. Enhancement of intestinal model compound transport by DS-1, a modified Quillaja saponin. Journal of Pharmaceutical Sciences 1998, 87, (11), 1395-1399.

107. Rege, B. D.; Kao, J. P. Y.; Polli, J. E. Effects of nonionic surfactants on membrane transporters in Caco-2 cell monolayers. European Journal of Pharmaceutical Sciences 2002, 16, (4–5), 237-246.

108. Martin, G. P.; Marriott, C. Membrane damage by bile salts: The protective function of phospholipids. Journal of Pharmacy and Pharmacology 1981, 33, (1), 754-759.

109. Seelig, A.; Gerebtzoff, G. Enhancement of drug absorption by noncharged detergents through membrane and P-glycoprotein binding. Expert Opinion on Drug Metabolism & Toxicology 2006, 2, (5), 733-752.

110. Constantinides, P. P.; Wasan, K. M. Lipid formulation strategies for enhancing intestinal transport and absorption of P-glycoprotein (P-gp) substrate drugs: In vitro/In vivo case studies. Journal of Pharmaceutical Sciences 2007, 96, (2), 235-248.

111. Werle, M. Polymeric and low molecular mass efflux pump inhibitors for oral drug delivery. Journal of Pharmaceutical Sciences 2008, 97, (1), 60-70.

112. Goole, J.; Lindley, D. J.; Roth, W.; Carl, S. M.; Amighi, K.; Kauffmann, J.-M.; Knipp, G. T. The effects of excipients on transporter mediated absorption. International Journal of Pharmaceutics 2010, 393, (1–2), 17-31.

249

113. de Lannoy, I. A. M.; Silverman, M. The MDR1 gene product, P-glycoprotein, mediates the transport of the cardiac glycoside, digoxin. Biochemical and Biophysical Research Communications 1992, 189, (1), 551-557.

114. Gottesman, M. M.; Pastan, I. Biochemistry of multidrug resistance mediated by the multidrug transporter. Annual Review of Biochemistry 1993, 62, (1), 385-427.

115. Augustijns, P. F.; Bradshaw, T. P.; Gan, L. S. L.; Hendren, R. W.; Thakker, D. R. Evidence for a polarized efflux system in Caco-2 cells capable of modulating cyclosporine A transport. Biochemical and Biophysical Research Communications 1993, 197, (2), 360-365.

116. Saeki, T.; Ueda, K.; Tanigawara, Y.; Hori, R.; Komano, T. P-glycoprotein-mediated transcellular transport of MDR-reversing agents. FEBS Letters 1993, 324, (1), 99-102.

117. Williams, G. C.; Sinko, P. J. Oral absorption of the HIV protease inhibitors: A current update. Advanced Drug Delivery Reviews 1999, 39, (1–3), 211-238.

118. Friche, E.; Jensen, P. B.; Sehested, M.; Demant, E. J.; Nissen, N. N. The solvents Cremophor EL and Tween 80 modulate daunorubicin resistance in the multidrug resistant Ehrlich ascites tumor. Cancer communications 1990, 2, (9), 297-303.

119. Spoelstra, E. C.; Dekker, H.; Schuurhuis, G. J.; Broxterman, H. J.; Lankelma, J. P-glycoprotein drug efflux pump involved in the mechanisms of intrinsic drug resistance in various colon cancer cell lines: Evidence for a saturation of active daunorubicin transport. Biochemical Pharmacology 1991, 41, (3), 349-359.

120. Zordan-Nudo, T.; Ling, V.; Liu, Z.; Georges, E. Effects of nonionic detergents on P-glycoprotein drug binding and reversal of multidrug resistance. Cancer Research 1993, 53, (24), 5994-6000.

121. Woodcock, D. M.; Linsenmeyer, M. E.; Chojnowski, G.; Kriegler, A. B.; Nink, V.; Webster, L. K.; Sawyer, W. H. Reversal of multidrug resistance by surfactants. Br J Cancer 1992, 66, (1), 62-68.

122. Loe, D. W.; Sharom, F. J. Interaction of multidrug-resistant Chinese hamster ovary cells with amphiphiles. Br J Cancer 1993, 68, (2), 342-351.

123. Regev, R.; Assaraf, Y. G.; Eytan, G. D. Membrane fluidization by ether, other anesthetics, and certain agents abolishes P-glycoprotein ATPase activity and modulates efflux from multidrug-resistant cells. European Journal of Biochemistry 1999, 259, (1-2), 18-24.

124. Risovic, V.; Sachs-Barrable, K.; Boyd, M.; Wasan, K. M. Potential mechanisms by which Peceol® increases the gastrointestinal absorption of amphotericin B. Drug Development & Industrial Pharmacy 2004, 30, (7), 767-774.

125. Sachs-Barrable, K.; Thamboo, A.; Lee, S. D.; Wasan, K. M. Lipid excipients Peceol and Gelucire 44/14 decrease P-glycoprotein mediated efflux of rhodamine 123 partially due to modifying P-glycoprotein protein expression within Caco-2 cells. J Pharm Pharmaceut Sci 2007, 10, (3), 319-331.

126. Barta, C. A.; Sachs-Barrable, K.; Feng, F.; Wasan, K. M. Effects of monoglycerides on P-glycoprotein: Modulation of the activity and expression in Caco-2 cell monolayers. Molecular Pharmaceutics 2008, 5, (5), 863-875.

250

127. Kivisto, K. T.; Kroemer, H. K.; Eichelbaum, M. The role of human cytochrome P450 enzymes in the metabolism of anticancer agents: Implications for drug interactions. British Journal of Clinical Pharmacology 1995, 40, (6), 523-530.

128. Wacher, V. J.; Wu, C.-Y.; Benet, L. Z. Overlapping substrate specificities and tissue distribution of cytochrome P450 3A and P-glycoprotein: Implications for drug delivery and activity in cancer chemotherapy. Molecular Carcinogenesis 1995, 13, (3), 129-134.

129. Benet, L. Z.; Wu, C.-Y.; Hebert, M. F.; Wacher, V. J. Intestinal drug metabolism and antitransport processes: A potential paradigm shift in oral drug delivery. Journal of Controlled Release 1996, 39, (2–3), 139-143.

130. Benet, L. Z.; Cummins, C. L. The drug efflux–metabolism alliance: Biochemical aspects. Advanced Drug Delivery Reviews 2001, 50, Supplement 1, (0), S3-S11.

131. Watkins, P. B. The barrier function of CYP3A4 and P-glycoprotein in the small bowel. Advanced Drug Delivery Reviews 1997, 27, (2–3), 161-170.

132. Benet, L. Z.; Izumi, T.; Zhang, Y.; Silverman, J. A.; Wacher, V. J. Intestinal MDR transport proteins and P-450 enzymes as barriers to oral drug delivery. Journal of Controlled Release 1999, 62, (1–2), 25-31.

133. Patel, J. P.; Brocks, D. R. The effect of oral lipids and circulating lipoproteins on the metabolism of drugs. Expert Opinion on Drug Metabolism & Toxicology 2009, 5, (11), 1385-1398.

134. Buggins, T. R.; Dickinson, P. A.; Taylor, G. The effects of pharmaceutical excipients on drug disposition. Advanced Drug Delivery Reviews 2007, 59, (15), 1482-1503.

135. Wasan, K. M.; Brocks, D. R.; Lee, S. D.; Sachs-Barrable, K.; Thornton, S. J. Impact of lipoproteins on the biological activity and disposition of hydrophobic drugs: Implications for drug discovery. Nat Rev Drug Discov 2008, 7, (1), 84-99.

136. Humberstone, A. J.; Porter, C. J. H.; Edwards, G. A.; Charman, W. N. Association of halofantrine with postprandially derived plasma lipoproteins decreases its clearance relative to administration in the fasted state. Journal of Pharmaceutical Sciences 1998, 87, (8), 936-942.

137. Trevaskis, N. L.; Charman, W. N.; Porter, C. J. H. Lipid-based delivery systems and intestinal lymphatic drug transport: A mechanistic update. Advanced Drug Delivery Reviews 2008, 60, (6), 702-716.

138. Charman, W. N. A.; Stella, V. J. Effects of lipid class and lipid vehicle volume on the intestinal lymphatic transport of DDT. International Journal of Pharmaceutics 1986, 33, (1–3), 165-172.

139. Khoo, S.-M.; Shackleford, D. M.; Porter, C. J. H.; Edwards, G. A.; Charman, W. N. Intestinal lymphatic transport of halofantrine occurs after oral administration of a unit-dose lipid-based formulation to fasted dogs. Pharmaceutical Research 2003, 20, (9), 1460-1465.

140. White, K. L.; Nguyen, G.; Charman, W. N.; Edwards, G. A.; Faassen, W. A.; Porter, C. J. H. Lymphatic transport of methylnortestosterone undecanoate (MU) and the bioavailability of methylnortestosterone are highly sensitive to the mass of coadministered lipid after oral administration of MU. Journal of Pharmacology and Experimental Therapeutics 2009, 331, (2), 700-709.

251

141. Tsuji, A.; Tamai, I. Carrier-mediated intestinal transport of drugs. Pharmaceutical Research 1996, 13, (7), 963-977.

142. Lee, V. H. L. Membrane transporters. European Journal of Pharmaceutical Sciences 2000, 11, Supplement 2, S41-S50.

143. Poelma, F. G. J.; Breäs, R.; Tukker, J. J.; Crommelin, D. J. A. Intestinal absorption of drugs. The influence of mixed micelles on the disappearance kinetics of drugs from the small intestine of the rat. Journal of Pharmacy and Pharmacology 1991, 43, (5), 317-324.

144. Katneni, K.; Charman, S. A.; Porter, C. J. H. Permeability assessment of poorly water-soluble compounds under solubilizing conditions: The reciprocal permeability approach. Journal of Pharmaceutical Sciences 2006, 95, (10), 2170-2185.

145. Amidon, G. E.; Higuchi, W. I.; Ho, N. F. H. Theoretical and experimental studies of transport of micelle-solubilized solutes. Journal of Pharmaceutical Sciences 1982, 71, (1), 77-84.

146. Dahan, A.; Miller, J. M.; Hoffman, A.; Amidon, G. E.; Amidon, G. L. The solubility–permeability interplay in using cyclodextrins as pharmaceutical solubilizers: Mechanistic modeling and application to progesterone. Journal of Pharmaceutical Sciences 2010, 99, (6), 2739-2749.

147. Miller, J. M.; Beig, A.; Carr, R. A.; Spence, J. K.; Dahan, A. A win–win solution in oral delivery of lipophilic drugs: Supersaturation via amorphous solid dispersions increases apparent solubility without sacrifice of intestinal membrane permeability. Molecular Pharmaceutics 2012, 9, (7), 2009-2016.

148. Parrott, N.; Lave, T. Applications of physiologically based absorption models in drug discovery and development. Molecular Pharmaceutics 2008, 5, (5), 760-775.

149. Hilgers, A. R.; Smith, D. P.; Biermacher, J. J.; Day, J. S.; Jensen, J. L.; Sims, S. M.; Adams, W. J.; Friis, J. M.; Palandra, J.; Hosley, J. D.; Shobe, E. M.; Burton, P. S. Predicting oral absorption of drugs: A case study with a novel class of antimicrobial agents. Pharmaceutical Research 2003, 20, (8), 1149-1155.

150. Wootan, M. G.; Bernlohr, D. A.; Storch, J. Mechanism of fluorescent fatty acid transfer from adipocyte fatty acid binding protein to membranes. Biochemistry 1993, 32, (33), 8622-8627.

151. Thumser, A.; Tsai, J.; Storch, J. Collision-mediated transfer of long-chain fatty acids by neural tissue fatty acid-binding proteins (FABP). Journal of Molecular Neuroscience 2001, 16, (2), 143-150.

152. Narayanan, V. S.; Storch, J. Fatty acid transfer in taurodeoxycholate mixed micelles. Biochemistry 1996, 35, (23), 7466-7473.

153. Storch, J.; Kleinfeld, A. M. Transfer of long-chain fluorescent free fatty acids between unilamellar vesicles. Biochemistry 1986, 25, (7), 1717-1726.

154. Nauli, A. M.; Nassir, F.; Zheng, S.; Yang, Q.; Lo, C. M.; VonLehmden, S. B.; Lee, D.; Jandacek, R. J.; Abumrad, N. A.; Tso, P. CD36 Is important for chylomicron formation and secretion and may mediate cholesterol uptake in the proximal intestine. Gastroenterology 2006, 131, (4), 1197-1207.

155. Werder, M.; Han, C.-H.; Wehrli, E.; Bimmler, D.; Schulthess, G.; Hauser, H. Role of scavenger receptors SR-BI and CD36 in selective sterol uptake in the small intestine. Biochemistry 2001, 40, (38), 11643-11650.

252

156. Hansen, G. H.; Niels-Christiansen, L.-L.; Immerdal, L.; Danielsen, E. M. Scavenger receptor class B type I (SR-BI) in pig enterocytes: Trafficking from the brush border to lipid droplets during fat absorption. Gut 2003, 52, (10), 1424-1431.

157. Altmann, S. W.; Davis Jr, H. R.; Yao, X.; Laverty, M.; Compton, D. S.; Zhu, L.-j.; Crona, J. H.; Caplen, M. A.; Hoos, L. M.; Tetzloff, G.; Priestley, T.; Burnett, D. A.; Strader, C. D.; Graziano, M. P. The identification of intestinal scavenger receptor class B, type I (SR-BI) by expression cloning and its role in cholesterol absorption. Biochimica et Biophysica Acta (BBA) - Molecular and Cell Biology of Lipids 2002, 1580, (1), 77-93.

158. Smart, E. J.; De Rose, R. A.; Farber, S. A. Annexin 2–caveolin 1 complex is a target of ezetimibe and regulates intestinal cholesterol transport. Proceedings of the National Academy of Sciences of the United States of America 2004, 101, (10), 3450-3455.

159. Valasek, M. A.; Weng, J.; Shaul, P. W.; Anderson, R. G. W.; Repa, J. J. Caveolin-1 is not required for murine intestinal cholesterol transport. Journal of Biological Chemistry 2005, 280, (30), 28103-28109.

160. Pohl, J.; Ring, A.; Ehehalt, R.; Schulze-Bergkamen, H.; Schad, A.; Verkade, P.; Stremmel, W. Long-chain fatty acid uptake into adipocytes depends on lipid raft function. Biochemistry 2004, 43, (14), 4179-4187.

161. Krieger, M. Scavenger receptor class B type I is a multiligand HDL receptor that influences diverse physiologic systems. Journal of Clinical Investigation 2001, 108, (6), 793-797.

162. Abumrad, N.; Coburn, C.; Ibrahimi, A. Membrane proteins implicated in long-chain fatty acid uptake by mammalian cells: CD36, FATP and FABPm. Biochimica et Biophysica Acta (BBA) - Molecular and Cell Biology of Lipids 1999, 1441, (1), 4-13.

163. Drover, V. A.; Abumrad, N. A. CD36-dependent fatty acid uptake regulates expression of peroxisome proliferator activated receptors. Biochemical Society Transactions 2005, 33, 311-315.

164. Sudhop, T.; Lütjohann, D.; von Bergmann, K. Sterol transporters: Targets of natural sterols and new lipid lowering drugs. Pharmacology & Therapeutics 2005, 105, (3), 333-341.

165. Hui, D. Y.; Howles, P. N. Molecular mechanisms of cholesterol absorption and transport in the intestine. Seminars in Cell & Developmental Biology 2005, 16, (2), 183-192.

166. Davies, J. P.; Levy, B.; Ioannou, Y. A. Evidence for a Niemann–Pick C (NPC) gene family: Identification and characterization of NPC1L1. Genomics 2000, 65, (2), 137-145.

167. Iyer, S. P.; Yao, X.; Crona, J. H.; Hoos, L. M.; Tetzloff, G.; Davis, H. R.; Graziano, M. P.; Altmann, S. W. Characterization of the putative native and recombinant rat sterol transporter Niemann-Pick C1 Like 1 (NPC1L1) protein. Biochimica et biophysica acta 2005, 1722, (3), 282-292.

168. Kirkham, M.; Parton, R. G. Clathrin-independent endocytosis: New insights into caveolae and non-caveolar lipid raft carriers. Biochim Biophys Acta 2005, 1745, (3), 273-286.

169. Ehehalt, R.; Füllekrug, J.; Pohl, J.; Ring, A.; Herrmann, T.; Stremmel, W. Translocation of long chain fatty acids across the plasma membrane – lipid rafts and fatty acid transport proteins. Molecular and Cellular Biochemistry 2006, 284, (1-2), 135-140.

253

170. Patel, H. H.; Murray, F.; Insel, P. A. Caveolae as organizers of pharmacologically relevant signal transduction molecules. Annual review of pharmacology and toxicology 2008, 48, 359-391.

171. Ring, A.; Pohl, J.; Völkl, A.; Stremmel, W. Evidence for vesicles that mediate long-chain fatty acid uptake by human microvascular endothelial cells. Journal of Lipid Research 2002, 43, (12), 2095-2104.

172. Babitt, J.; Trigatti, B.; Rigotti, A.; Smart, E. J.; Anderson, R. G. W.; Xu, S.; Krieger, M. Murine SR-BI, a high density lipoprotein receptor that mediates selective lipid uptake, is N-glycosylated and fatty acylated and colocalizes with plasma membrane caveolae. Journal of Biological Chemistry 1997, 272, (20), 13242-13249.

173. Brouwers, J.; Brewster, M. E.; Augustijns, P. Supersaturating drug delivery systems: The answer to solubility-limited oral bioavailability? Journal of Pharmaceutical Sciences 2009, 98, (8), 2549-2572.

174. Augustijns, P.; Brewster, M. E. Supersaturating drug delivery systems: Fast is not necessarily good enough. Journal of Pharmaceutical Sciences 2012, 101, (1), 7-9.

175. Gao, P.; Akrami, A.; Alvarez, F.; Hu, J.; Li, L.; Ma, C.; Surapaneni, S. Characterization and optimization of AMG 517 supersaturatable self-emulsifying drug delivery system (S-SEDDS) for improved oral absorption. Journal of Pharmaceutical Sciences 2009, 98, (2), 516-528.

176. Anby, M. U.; Williams, H. D.; McIntosh, M.; Benameur, H.; Edwards, G. A.; Pouton, C. W.; Porter, C. J. H. Lipid digestion as a trigger for supersaturation: Evaluation of the impact of supersaturation stabilization on the in vitro and in vivo performance of self-emulsifying drug delivery systems. Molecular Pharmaceutics 2012, 9, (7), 2063-2079.

177. Cuiné, J.; Charman, W.; Pouton, C.; Edwards, G.; Porter, C. Increasing the proportional content of surfactant (Cremophor EL) relative to lipid in self-emulsifying lipid-based formulations of danazol reduces oral bioavailability in beagle dogs. Pharmaceutical Research 2007, 24, (4), 748-757.

178. Mohsin, K.; Long, M. A.; Pouton, C. W. Design of lipid-based formulations for oral administration of poorly water-soluble drugs: Precipitation of drug after dispersion of formulations in aqueous solution. Journal of Pharmaceutical Sciences 2009, 98, (10), 3582-3595.

179. Warren, D. B.; Benameur, H.; Porter, C. J. H.; Pouton, C. W. Using polymeric precipitation inhibitors to improve the absorption of poorly water-soluble drugs: A mechanistic basis for utility. Journal of Drug Targeting 2010, 18, (10), 704-731.

180. Singhal, D.; Ho, N. F. H.; Anderson, B. D. Absorption and intestinal metabolism of purine dideoxynucleosides and an adenosine deaminase-activated prodrug of 2′ ,3′ -dideoxyinosine in the mesenteric vein cannulated rat ileum. Journal of Pharmaceutical Sciences 1998, 87, (5), 569-577.

181. Schurgers, N.; Bijdendijk, J.; Tukker, J. J.; Crommelin, D. J. A. Comparison of four experimental techniques for studying drug absorption kinetics in the anesthetized rat in situ. Journal of Pharmaceutical Sciences 1986, 75, (2), 117-119.

182. Grassi, M.; Cadelli, G. Theoretical considerations on the in vivo intestinal permeability determination by means of the single pass and recirculating techniques. International Journal of Pharmaceutics 2001, 229, (1–2), 95-105.

254

183. Salphati, L.; Childers, K.; Pan, L.; Tsutsui, K.; Takahashi, L. Evaluation of a single-pass intestinal-perfusion method in rat for the prediction of absorption in man. Journal of Pharmacy and Pharmacology 2001, 53, (7), 1007-1013.

184. Cummins, C. L.; Salphati, L.; Reid, M. J.; Benet, L. Z. In vivo modulation of intestinal CYP3A metabolism by P-glycoprotein: Studies using the rat single-pass intestinal perfusion model. Journal of Pharmacology and Experimental Therapeutics 2003, 305, (1), 306-314.

185. Ho, N. F. H.; Higuchi, W. I. Theoretical model studies of intestinal drug absorption IV: Bile acid transport at premicellar concentrations across diffusion layer—membrane barrier. Journal of Pharmaceutical Sciences 1974, 63, (5), 686-690.

186. Madara, J. L.; Pappenheimer, J. R. Structural basis for physiological regulation of paracellular pathways in intestinal epithelia. J. Membrain Biol. 1987, 100, (1), 149-164.

187. Fagerholm, U.; Nilsson, D.; Knutson, L.; Lennernäs, H. Jejunal permeability in humans in vivo and rats in situ: Investigation of molecular size selectivity and solvent drag. Acta Physiologica Scandinavica 1999, 165, (3), 315-324.

188. Lindahl, A.; Krondahl, E.; Grudén, A.-C.; Ungell, A.-L.; Lennernäs, H. Is the jejunal permeability in rats age-dependent? Pharmaceutical Research 1997, 14, (9), 1278-1281.

189. Pappenheimer, J. R.; Reiss, K. Z. Contribution of solvent drag through intercellular junctions to absorption of nutrients by the small intestine of the rat. J. Membrain Biol. 1987, 100, (1), 123-136.

190. Yano, K.; Masaoka, Y.; Kataoka, M.; Sakuma, S.; Yamashita, S. Mechanisms of membrane transport of poorly soluble drugs: Role of micelles in oral absorption processes. Journal of Pharmaceutical Sciences 2010, 99, (3), 1336-1345.

191. Gao, P.; Morozowich, W. Development of supersaturatable self-emulsifying drug delivery system formulations for improving the oral absorption of poorly soluble drugs. Expert Opinion on Drug Delivery 2006, 3, (1), 97-110.

192. Gu, X.; Trigatti, B.; Xu, S.; Acton, S.; Babitt, J.; Krieger, M. The efficient cellular uptake of high density lipoprotein lipids via Scavenger Receptor Class B Type I requires not only receptor-mediated surface binding but also receptor-specific lipid transfer mediated by its extracellular domain. Journal of Biological Chemistry 1998, 273, (41), 26338-26348.

193. Porter, C. J. H.; Kaukonen, A. M.; Boyd, B. J.; Edwards, G. A.; Charman, W. N. Susceptibility to lipase-mediated digestion reduces the oral bioavailability of danazol after administration as a medium-chain lipid-based microemulsion formulation. Pharmaceutical Research 2004, 21, (8), 1405-1412.

194. Kossena, G. A.; Boyd, B. J.; Porter, C. J. H.; Charman, W. N. Separation and characterization of the colloidal phases produced on digestion of common formulation lipids and assessment of their impact on the apparent solubility of selected poorly water-soluble drugs. Journal of Pharmaceutical Sciences 2003, 92, (3), 634-648.

195. Harmon, C. M.; Luce, P.; Beth, A. H.; Abumrad, N. A. Labeling of adipocyte membranes by sulfo-N-succinimidyl derivatives of long-chain fatty acids: Inhibition of fatty acid transport. Journal of Membrane Biology 1991, 121, (3), 261-268.

255

196. Nieland, T. J. F.; Penman, M.; Dori, L.; Krieger, M.; Kirchhausen, T. Discovery of chemical inhibitors of the selective transfer of lipids mediated by the HDL receptor SR-BI. PNAS 2002, 99, (24), 15422-15427.

197. Coort, S. L. M.; Willems, J.; Coumans, W. A.; van der Vusse, G. J.; Bonen, A.; Glatz, J. F. C.; Luiken, J. J. F. P. Sulfo-N-succinimidyl esters of long chain fatty acids specifically inhibit fatty acid translocase (FAT/CD36)-mediated cellular fatty acid uptake. Molecular and Cellular Biochemistry 2002, 239, (1), 213-219.

198. Garcia-Calvo, M.; Lisnock, J.; Bull, H. G.; Hawes, B. E.; Burnett, D. A.; Braun, M. P.; Crona, J. H.; Davis, H. R.; Dean, D. C.; Detmers, P. A.; Graziano, M. P.; Hughes, M.; MacIntyre, D. E.; Ogawa, A.; O'Neill, K. A.; Iyer, S. P. N.; Shevell, D. E.; Smith, M. M.; Tang, Y. S.; Makarewicz, A. M.; Ujjainwalla, F.; Altmann, S. W.; Chapman, K. T.; Thornberry, N. A. The target of ezetimibe is Niemann-Pick C1-Like 1 (NPC1L1). Proceedings of the National Academy of Sciences of the United States of America 2005, 102, (23), 8132-8137.

199. Staggers, J. E.; Hernell, O.; Stafford, R. J.; Carey, M. C. Physical-chemical behavior of dietary and biliary lipids during intestinal digestion and absorption. 1. Phase behavior and aggregation states of model lipid systems patterned after aqueous duodenal contents of healthy adult human beings. Biochemistry 1990, 29, (8), 2028-2040.

200. Hay, D. W.; Cahalane, M. J.; Timofeyeva, N.; Carey, M. C. Molecular species of lecithins in human gallbladder bile. Journal of Lipid Research 1993, 34, (5), 759-68.

201. Duane, W. C.; Ginsberg, R. L.; Bennion, L. J. Effects of fasting on bile acid metabolism and biliary lipid composition in man. Journal of Lipid Research 1976, 17, (3), 211-9.

202. Boyd, B. J.; Porter, C. J. H.; Charman, W. N. Using the polymer partitioning method to probe the thermodynamic activity of poorly water-soluble drugs solubilized in model lipid digestion products. Journal of Pharmaceutical Sciences 2003, 92, (6), 1262-1271.

203. Waynforth, H. B.; Flecknell, P. A. Experimental and surgical technique in rat. Academic Press 1992.

204. Van Heek, M.; Farley, C.; Compton, D. S.; Hoos, L.; Alton, K. B.; Sybertz, E. J.; Davis, H. R. Comparison of the activity and disposition of the novel cholesterol absorption inhibitor, SCH58235, and its glucuronide, SCH60663. British Journal of Pharmacology 2000, 129, (8), 1748-1754.

205. Kyle, M. E.; Nakae, D.; Sakaida, I.; Miccadei, S.; Farber, J. L. Endocytosis of superoxide dismutase is required in order for the enzyme to protect hepatocytes from the cytotoxicity of hydrogen peroxide. Journal of Biological Chemistry 1988, 263, (8), 3784-3789.

206. Winne, D. Rat jejunum perfused in situ: Effect of perfusion rate and intraluminal radius on absorption rate and effective unstirred layer thickness. Naunyn-Schmiedeberg's Archives of Pharmacology 1979, 307, (3), 265-274.

207. Johnson, B. M.; Chen, W.; Borchardt, R. T.; Charman, W. N.; Porter, C. J. H. A kinetic evaluation of the absorption, efflux, and metabolism of verapamil in the autoperfused rat jejunum. Journal of Pharmacology and Experimental Therapeutics 2003, 305, (1), 151-158.

256

208. Wagner, J. G.; Sedman, A. J. Quantitation of rate of gastrointestinal and buccal absorption of acidic and basic drugs based on extraction theory. Journal of Pharmacokinetics and Biopharmaceutics 1973, 1, (1), 23-50.

209. Dahan, A.; Miller, J. M.; Hilfinger, J. M.; Yamashita, S.; Yu, L. X.; Lennernäs, H.; Amidon, G. L. High-permeability criterion for BCS classification: Segmental/pH dependent permeability considerations. Molecular Pharmaceutics 2010, 7, (5), 1827-1834.

210. Haikal, Z.; Play, B.; Landrier, J.-F.; Giraud, A.; Ghiringhelli, O.; Lairon, D.; Jourdheuil-Rahmani, D. NPC1L1 and SR-BI are involved in intestinal cholesterol absorption from small-size lipid donors. Lipids 2008, 43, (5), 401-408.

211. Tran, T. T. T.; Poirier, H.; Clément, L.; Nassir, F.; Pelsers, M. M. A. L.; Petit, V.; Degrace, P.; Monnot, M.-C.; Glatz, J. F. C.; Abumrad, N. A.; Besnard, P.; Niot, I. Luminal lipid regulates CD36 levels and downstream signaling to stimulate chylomicron synthesis. Journal of Biological Chemistry 2011, 286, (28), 25201-25210.

212. Yu, L.; Bharadwaj, S.; Brown, J. M.; Ma, Y.; Du, W.; Davis, M. A.; Michaely, P.; Liu, P.; Willingham, M. C.; Rudel, L. L. Cholesterol-regulated Translocation of NPC1L1 to the Cell Surface Facilitates Free Cholesterol Uptake. Journal of Biological Chemistry 2006, 281, (10), 6616-6624.

213. Moussa, M.; Landrier, J.-F.; Reboul, E.; Ghiringhelli, O.; Coméra, C.; Collet, X.; Fröhlich, K.; Böhm, V.; Borel, P. Lycopene absorption in human intestinal cells and in mice involves Scavenger Receptor Class B Type I but not Niemann-Pick C1-Like 1. The Journal of Nutrition 2008, 138, (8), 1432-1436.

214. Reboul, E.; Goncalves, A.; Comera, C.; Bott, R.; Nowicki, M.; Landrier, J.-F.; Jourdheuil-Rahmani, D.; Dufour, C.; Collet, X.; Borel, P. Vitamin D intestinal absorption is not a simple passive diffusion: Evidences for involvement of cholesterol transporters. Molecular Nutrition & Food Research 2011, 55, (5), 691-702.

215. Reboul, E.; Klein, A.; Bietrix, F.; Gleize, B.; Malezet-Desmoulins, C.; Schneider, M.; Margotat, A.; Lagrost, L.; Collet, X.; Borel, P. Scavenger Receptor Class B Type I (SR-BI) Is involved in vitamin E transport across the enterocyte. Journal of Biological Chemistry 2006, 281, (8), 4739-4745.

216. Goldstein, J. L.; Brown, M. S.; Anderson, R. G. W.; Russell, D. W.; Schneider, W. J. Receptor-mediated endocytosis: Concepts emerging from the LDL receptor system. Annual Review of Cell Biology 1985, 1, (1), 1-39.

217. Mukerjee, P.; Mysels, K. J. Critical micelle concentrations of aqueous surfactant systems. Washington, DC: National Bureau of Standards 1971, 88-170.

218. Schubert, R.; Schmidt, K. H. Structural changes in vesicle membranes and mixed micelles of various lipid compositions after binding of different bile salts. Biochemistry 1988, 27, (24), 8787-8794.

219. Hofmann, A. F. Bile acids: The good, the bad, and the ugly. Physiology 1999, 14, (1), 24-29.

220. Vertzoni, M.; Fotaki, N.; Nicolaides, E.; Reppas, C.; Kostewicz, E. S.; Stippler, E.; Leuner, C.; Dressman, J. Dissolution media simulating the intralumenal composition of the small intestine: physiological issues and practical aspects. Journal of Pharmacy and Pharmacology 2004, 56, (4), 453-462.

257

221. Järvinen, T.; Järvinen, K.; Schwarting, N.; Stella, V. J. β-cyclodextrin derivatives, SBE4-β-CD and HP–β-CD, increase the oral bioavailability of cinnarizine in beagle dogs. Journal of Pharmaceutical Sciences 1995, 84, (3), 295-299.

222. Tønsberg, H.; Holm, R.; Mu, H.; Boll, J. B.; Jacobsen, J.; Müllertz, A. Effect of bile on the oral absorption of halofantrine in polyethylene glycol 400 and polysorbate 80 formulations dosed to bile duct cannulated rats. Journal of Pharmacy and Pharmacology 2011, 63, (6), 817-824.

223. Ericzon, B. G.; Todo, S.; Lynch, S.; Kam, I.; Ptachcinski, R. J.; Burckart, G. J.; Van Thiel, D. H.; Starzl, T. E.; Venkataramanan, R. Role of bile and bile salts on cyclosporine absorption in dogs. Transplant Proc. 1987, 19, (1 Pt 2), 1248-1249.

224. Behrens, D.; Fricker, R.; Bodoky, A.; Drewe, J.; Harder, F.; Heberer, M. Comparison of cyclosporin A absorption from LCT and MCT solutions following intrajejunal administration in conscious dogs. Journal of Pharmaceutical Sciences 1996, 85, (6), 666-668.

225. Miller, J. M.; Beig, A.; Carr, R. A.; Webster, G. K.; Dahan, A. The solubility–permeability interplay when using cosolvents for solubilization: Revising the way we use solubility-enabling formulations. Molecular Pharmaceutics 2012, 9, (3), 581-590.

226. Dahan, A.; Beig, A.; Ioffe-Dahan, V.; Agbaria, R.; Miller, J. M. The twofold advantage of the amorphous form as an oral drug delivery practice for lipophilic compounds: Increased apparent solubility and drug flux through the intestinal membrane. The AAPS Journal 2012, In press.

227. Yeap, Y. Y.; Trevaskis, N. L.; Quach, T.; Tso, P.; Charman, W. N.; Porter, C. J. H. Bile secretion promotes drug absorption from lipid colloidal phases via induction of supersaturation. Submitted.

228. Van Speybroeck, M.; Williams, H. D.; Nguyen, T.-H.; Anby, M. U.; Porter, C. J. H.; Augustijns, P. Incomplete desorption of liquid excipients reduces the in vitro and in vivo performance of self-emulsifying drug delivery systems solidified by adsorption onto an inorganic mesoporous carrier. Molecular Pharmaceutics 2012, 9, (9), 2750-2760.

229. Trevaskis, N. L.; Porter, C. J. H.; Charman, W. N. Bile increases intestinal lymphatic drug transport in the fasted rat. Pharmaceutical Research 2005, 22, (11), 1863-1870.

230. Bergström, C. A. S.; Wassvik, C. M.; Johansson, K.; Hubatsch, I. Poorly soluble marketed drugs display solvation limited solubility. Journal of Medicinal Chemistry 2007, 50, (23), 5858-5862.

231. McIntosh, M. P.; Charman, W. N.; Campbell, M.; Porter, C. J. H. Influence of physicochemical properties on the patterns of association of a series of aliphatic esters of halofantrine with plasma lipoproteins. Journal of Controlled Release 2004, 95, (2), 275-289.

232. Van Speybroeck, M.; Mellaerts, R.; Mols, R.; Thi, T. D.; Martens, J. A.; Van Humbeeck, J.; Annaert, P.; Van den Mooter, G.; Augustijns, P. Enhanced absorption of the poorly soluble drug fenofibrate by tuning its release rate from ordered mesoporous silica. European Journal of Pharmaceutical Sciences 2010, 41, (5), 623-630.

233. Sanphui, P.; Bolla, G.; Nangia, A. High solubility piperazine salts of the nonsteroidal anti-Inflammatory drug (NSAID) meclofenamic acid. Crystal Growth & Design 2012, 12, (4), 2023-2036.

258

234. Guichard, J. P.; Blouquin, P.; Qing, Y. A new formulation of fenofibrate: Suprabioavailable tablets. Current Medical Research and Opinion 2000, 16, (2), 134-138.

235. Gu, C.-H.; Rao, D.; Gandhi, R. B.; Hilden, J.; Raghavan, K. Using a novel multicompartment dissolution system to predict the effect of gastric pH on the oral absorption of weak bases with poor intrinsic solubility. Journal of Pharmaceutical Sciences 2005, 94, (1), 199-208.

236. Peeters, J. Determination of ionization constants in mixed aqueous solvents of varying composition by a single titration. Journal of Pharmaceutical Sciences 1978, 67, (1), 127-129.

237. Khoo, S.-M.; Prankerd, R. J.; Edwards, G. A.; Porter, C. J. H.; Charman, W. N. A physicochemical basis for the extensive intestinal lymphatic transport of a poorly lipid soluble antimalarial, halofantrine hydrochloride, after postprandial administration to dogs. Journal of Pharmaceutical Sciences 2002, 91, (3), 647-659.

238. Marriner, S.; Bogan, J. A. The influence of the rumen on the absorption of drugs: studies using meclofenamic acid administered by various routes to sheep and cattle. Journal of Veterinary Pharmacology and Therapeutics 1979, 2, (2), 109-115.

239. Hernell, O.; Staggers, J. E.; Carey, M. C. Physical-chemical behavior of dietary and biliary lipids during intestinal digestion and absorption. 2. Phase analysis and aggregation states of luminal lipids during duodenal fat digestion in healthy adult human beings. Biochemistry 1990, 29, (8), 2041-2056.

240. Humberstone, A. J.; Charman, W. N. Lipid-based vehicles for the oral delivery of poorly water soluble drugs. Advanced Drug Delivery Reviews 1997, 25, (1), 103-128.

241. O'Connor, C. J.; Wallace, R. G.; Iwamoto, K.; Taguchi, T.; Sunamoto, J. Bile salt damage of egg phosphatidylcholine liposomes. Biochimica et Biophysica Acta (BBA) - Biomembranes 1985, 817, (1), 95-102.

242. Schubert, R.; Beyer, K.; Wolburg, H.; Schmidt, K. H. Structural changes in membranes of large unilamellar vesicles after binding of sodium cholate. Biochemistry 1986, 25, (18), 5263-5269.

243. Rosoff, M.; Serajuddin, A. T. M. Solubilization of diazepam in bile salts and in sodium cholate-lecithin-water phases. International Journal of Pharmaceutics 1980, 6, (2), 137-146.

244. Spernath, A.; Yaghmur, A.; Aserin, A.; Hoffman, R. E.; Garti, N. Food-grade microemulsions based on nonionic emulsifiers: Media to enhance lycopene solubilization. Journal of Agricultural and Food Chemistry 2002, 50, (23), 6917-6922.

245. Spernath, A.; Yaghmur, A.; Aserin, A.; Hoffman, R. E.; Garti, N. Self-diffusion nuclear magnetic resonance, microstructure transitions, and solubilization capacity of phytosterols and cholesterol in Winsor IV food-grade microemulsions. Journal of Agricultural and Food Chemistry 2003, 51, (8), 2359-2364.

246. Garti, N.; Amar, I.; Yaghmur, A.; Spernath, A.; Aserin, A. Interfacial modification and structural transitions induced by guest molecules solubilized in U-type nonionic microemulsions. Journal of Dispersion Science and Technology 2003, 24, (3-4), 397-410.

247. Rane, S. S.; Anderson, B. D. What determines drug solubility in lipid vehicles: Is it predictable? Advanced Drug Delivery Reviews 2008, 60, (6), 638-656.

259

248. Feldman, S.; Gibaldi, M. Bile salt-induced permeability changes in the isolated rat intestine. Experimental Biology and Medicine 1969, 132, (3), 1031-1033.

249. Feldman, S.; Reinhard, M.; Willson, C. Effect of sodium taurodeoxycholate on biological membranes: Release of phosphorus, phospholipid, and protein from everted rat small intestine. Journal of Pharmaceutical Sciences 1973, 62, (12), 1961-1964.

250. Day, J. P. R.; Rago, G.; Domke, K. F.; Velikov, K. P.; Bonn, M. Label-free imaging of lipophilic bioactive molecules during lipid digestion by multiplex coherent anti-stokes raman scattering microspectroscopy. Journal of the American Chemical Society 2010, 132, (24), 8433-8439.

251. Williams, H. D.; Sassene, P.; Kleberg, K.; Calderone, M.; Igonin, A.; Jule, E.; Vertommen, J.; Blundell, R.; Benameur, H.; Mullertz, A.; Pouton, C. W.; Porter, C. J. H. Toward the establishment of standardized in vitro tests for lipid-based formulations: 3) Understanding supersaturation versus precipitation potential during the in vitro digestion of Type I, II, IIIA, IIIB and IV lipid-based formulations. Pharmaceutical Research 2013, This issue.

252. Lucas, M. L.; Schneider, W.; Haberich, F. J.; Blair, J. A. Direct measurement by pH-microelectrode of the pH microclimate in rat proximal jejunum. Proceedings of the Royal Society of London. Series B. Biological Sciences 1975, 192, (1106), 39-48.

253. Mahnensmith, R. L.; Aronson, P. S. The plasma membrane sodium-hydrogen exchanger and its role in physiological and pathophysiological processes. Circulation Research 1985, 56, (6), 773-88.

254. Williams, H. D.; Sassene, P.; Kleberg, K.; Bakala-N'Goma, J.-C.; Calderone, M.; Jannin, V.; Igonin, A.; Partheil, A.; Marchaud, D.; Jule, E.; Vertommen, J.; Maio, M.; Blundell, R.; Benameur, H.; Carrière, F.; Müllertz, A.; Porter, C. J. H.; Pouton, C. W. Toward the establishment of standardized in vitro tests for lipid-based formulations, part 1: Method parameterization and comparison of in vitro digestion profiles across a range of representative formulations. Journal of Pharmaceutical Sciences 2012, 101, (9), 3360-3380.

255. Chow, S. L.; Hollander, D. Linoleic acid absorption in the unaesthetized rat: Mechanism of transport and influence of luminal factors on absorption. Lipids 1979, 14, (4), 378-385.

256. Kanicky, J. R.; Shah, D. O. Effect of degree, type, and position of unsaturation on the pKa of long-chain fatty acids. Journal of Colloid and Interface Science 2002, 256, (1), 201-207.

257. Hofmann, A. F., Molecular association in fat digestion. In Molecular Association in Biological and Related Systems, AMERICAN CHEMICAL SOCIETY: 1968; Vol. 84, pp 53-66.

258. Hofmann, A. F.; Mysels, K. J. Bile salts as biological surfactants. Colloids and Surfaces 1987, 30, (1), 145-173.

259. Schoeller, C.; Keelan, M.; Mulvey, G.; Stremmel, W.; Thomson, A. B. R. Oleic acid uptake into rat and rabbit jejunal brush border membrane. Biochimica et Biophysica Acta (BBA) - Biomembranes 1995, 1236, (1), 51-64.

260. Cuiné, J. F.; McEvoy, C. L.; Charman, W. N.; Pouton, C. W.; Edwards, G. A.; Benameur, H.; Porter, C. J. H. Evaluation of the impact of surfactant digestion on the bioavailability of danazol after oral administration of lipidic self-emulsifying formulations to dogs. Journal of Pharmaceutical Sciences 2008, 97, (2), 995-1012.

260

261. Bevernage, J.; Brouwers, J.; Annaert, P.; Augustijns, P. Drug precipitation–permeation interplay: Supersaturation in an absorptive environment. European Journal of Pharmaceutics and Biopharmaceutics 2012, 82, (2), 424-428.

262. Raghavan, S. L.; Trividic, A.; Davis, A. F.; Hadgraft, J. Crystallization of hydrocortisone acetate: Influence of polymers. International Journal of Pharmaceutics 2001, 212, (2), 213-221.

263. Machefer, S.; Huddar, M. M.; Schnitzlein, K. Effect of polymer admixtures on the growth habit of ionic crystals. Study on crystal growth kinetics of potassium dihydrogen phosphate in water/polyol mixtures. Journal of Crystal Growth 2008, 310, (24), 5347-5356.

261

APPENDIX 1

Re-printed with permission from Yeap YY, Trevaskis NL, Quach T, Tso P, Charman WN,

Porter CJH. Intestinal bile secretion promotes drug absorption from lipid colloidal phases

via induction of supersaturation. Molecular Pharmaceutics, 2013, in press.

1 Intestinal Bile Secretion Promotes Drug Absorption from Lipid2 Colloidal Phases via Induction of Supersaturation3 Yan Yan Yeap,† Natalie L. Trevaskis,*,† Tim Quach,‡ Patrick Tso,§ William N. Charman,†

4 and Christopher J. H. Porter*,†

5†Drug Delivery, Disposition and Dynamics and ‡Medicinal Chemistry and Drug Action, Monash Institute of Pharmaceutical Sciences,

6 Monash University, 381 Royal Parade, Parkville, Victoria, 3052, Australia

7§Department of Pathology and Laboratory Medicine, University of Cincinnati, Cincinnati, Ohio 45221, United States

8 *S Supporting Information

9 ABSTRACT: The oral bioavailability of poorly water-soluble10 drugs (PWSD) is often significantly enhanced by coadminis-11 tration with lipids in food or lipid-based oral formulations.12 Coadministration with lipids promotes drug solubilization in13 intestinal mixed micelles and vesicles, however, the mecha-14 nism(s) by which PWSD are absorbed from these dispersed15 phases remain poorly understood. Classically, drug absorption16 is believed to be a product of the drug concentration in free17 solution and the apparent permeability across the absorptive18 membrane. Solubilization in colloidal phases such as mixed19 micelles increases dissolution rate and total solubilized drug20 concentrations, but does not directly enhance (and may21 reduce) the free drug concentration. In the absence of changes to cellular permeability (which is often high for lipophilic,22 PWSD), significant changes to membrane flux are therefore unexpected. Realizing that increases in effective dissolution rate may23 be a significant driver of increases in drug absorption for PWSD, we explore here two alternate mechanisms by which membrane24 flux might also be enhanced: (1) collisional drug absorption where drug is directly transferred from lipid colloidal phases to the25 absorptive membrane, and (2) supersaturation-enhanced drug absorption where bile mediated dilution of lipid colloidal phases26 leads to a transient increase in supersaturation, thermodynamic activity and absorption. In the current study, collisional uptake27 mechanisms did not play a significant role in the absorption of a model PWSD, cinnarizine, from lipid colloidal phases. In28 contrast, bile-mediated dilution of model intestinal mixed micelles and vesicles led to drug supersaturation. For colloids that were29 principally micellar, supersaturation was maintained for a period sufficient to promote absorption. In contrast, for primarily30 vesicular systems, supersaturation resulted in rapid drug precipitation and no increase in drug absorption. This work suggests that31 ongoing dilution by bile in the gastrointestinal tract may invoke supersaturation in intestinal colloids and promote absorption,32 and thus presents a new mechanism by which lipids may enhance the oral absorption of PWSD.

33 KEYWORDS: absorption, poorly water-soluble drug, lipid-based formulation, supersaturation, bile, micelles, food effect,34 membrane permeability, cinnarizine

35 ■ INTRODUCTION

36 The potential for lipid-based formulations (LBF) to enhance

37 the oral bioavailability of poorly water-soluble drugs (PWSD)

38 has been recognized for over 40 years.1 Lipid coadministration

39 is thought to enhance the oral absorption of PWSD by

40 providing mechanisms to overcome both slow dissolution and

41 low solubility of PWSD in the aqueous gastrointestinal (GI)

42 milieu. First, LBF present drug to the GI tract in a molecularly

43 dispersed form (i.e., in solution in the formulation), thereby

44 circumventing the need for dissolution from the solid to the

45 liquid state. Subsequently, the intercalation of formulation

46 lipids into endogenous lipid digestion pathways results in the

47 generation of intestinal lipid colloidal phases (such as vesicular48 and micellar species) that enhance the solubilization capacity of

49the small intestine, promote drug solubilization and reduce the50risk of drug precipitation.51Solubilization within lipid colloidal phases therefore increases52the apparent solubility of PWSD in the intestinal fluids and53promotes dissolution. However, in the absence of solid drug,54solubilization also results in a reduction in thermodynamic55activity.2 In simple micellar systems this reduction in56thermodynamic activity is manifest in a decrease in the free57concentration of drug. Where solubilized drug exists in58equilibrium between the free concentration (Cfree) and the

Received: November 15, 2012Revised: March 9, 2013Accepted: March 12, 2013

Article

pubs.acs.org/molecularpharmaceutics

© XXXX American Chemical Society A dx.doi.org/10.1021/mp3006566 | Mol. Pharmaceutics XXXX, XXX, XXX−XXX

tms00 | ACSJCA | JCA10.0.1465/W Unicode | research.3f (R3.5.i1:3915 | 2.0 alpha 39) 2012/12/04 10:21:00 | PROD-JCAVA | rq_2253907 | 3/20/2013 16:18:31 | 15

pubs.acs.org/molecularpharmaceutics

59 concentration solubilized in intestinal colloids (Ccolloid), the60 total solubilized drug concentration (Ctotal) is the sum of Cfree61 and Ccolloid,

= +C C Ctotal free colloid62 (1)

63 Under these circumstances, the solubility of drug in the64 intermicellar phase (effectively the aqueous solubility of the65 drug) provides the upper limit for Cfree and the presence of66 solubilizing colloidal species typically increases the total drug67 concentration but does not increase, and often reduces, Cfree.68 Classical models of passive drug absorption suggest that drug69 flux across an absorptive membrane is the product of the free70 drug concentration and the drug permeability across the71 membrane. Therefore, where solubilization reduces Cfree (but72 does not alter permeability), absorption is expected to be73 reduced. Indeed, in solubilized systems, even at saturation, Cfree74 does not exceed the equilibrium solubility of drug in75 (nonmicellar) aqueous solution. Solubilization in intestinal76 colloidal phases therefore provides no practical advantage in77 free drug concentration and, in the absence of changes to78 permeability, is unlikely to lead to appreciable increases in79 membrane flux when compared to an aqueous solution80 containing drug at close to saturated solubility. In support of81 this suggestion, many authors have shown that increasing the82 total concentration of a PWSD by solubilization does not83 necessarily result in proportional increases in absorptive flux.2−6

84 Dahan et al. and Miller et al. recently proposed a model to85 quantify this phenomenon and referred to the existence of a86 “solubility-permeability interplay” where potential increases in87 membrane flux due to increases in solubilized drug88 concentration were offset by a reduction in the apparent89 permeability, the latter being, in large part, a function of90 decreases in free fraction.5,6

91 The dispersion and digestion of LBF therefore creates a92 solubilized reservoir that is in rapid equilibrium with drug in93 free solution and provides significant advantage in the effective94 rate of dissolution of a PWSD when compared to traditional95 dissolution from a solid dose form. In contrast, the inherent96 solubility limitations to flux (rather than dissolution-rate97 limitations) are seemingly unchanged when viewed from the98 perspective of the free concentration or may be made worse.99 This appears at odds with the wealth of experimental and100 practical observations that suggest the ability of lipids (either101 formulation- or dietary-derived) to enhance the oral absorption102 of a range of PWSD.1 A possible explanation for this anomaly is103 that the traditional view of drug absorption from colloidal104 dispersions may not adequately describe the dynamic manner105 in which LBF exert absorption-enhancing effects in the GI tract.106 Two alternative potential mechanisms of drug absorption are

f1 107 therefore examined here (Figure 1).108 The first mechanism evaluated was the potential for drug109 absorption to occur via direct collisional transfer from lipid110 colloidal phases to the absorptive membrane, and thus to be111 mediated not only by Cfree but also by the solubilized fraction,112 Ccolloid. Previously, studies by Storch and colleagues have shown113 that the transfer of poorly water-soluble fatty acids between114 model cell membranes and proteins may occur via collisional115 transfer.7,8 More recently, the possibility of drug absorption via116 collisional uptake has been suggested by Yano et al. and Gao et117 al.9,10 Collisional uptake may or may not be receptor-118 mediated,11 however, lipid uptake receptors such as CD36,12

119 SR-BI13 and NPC1L114 have been suggested to facilitate the120 absorption of cholesterol and fatty acids, poorly water-soluble

121molecules that are also solubilized in intestinal colloidal phases.122In the case of SR-BI and CD36, direct interaction of the123receptor with colloidal structures such as HDL (high-density124lipoprotein), bile salt micelles and phospholipid vesicles13,15 has125also been suggested, raising the possibility that lipid uptake126receptors may interact directly with intestinal lipid colloidal127phases to facilitate collisional absorption of solubilized contents,128including PWSD.129The second mechanism evaluated was the potential for130endogenous lipid processing pathways to lead to drug131supersaturation in lipid colloidal phases. Supersaturation132increases the thermodynamic activity of solubilized drug and,133in the solubilization model described by eq 1, will increase Cfree134above the equilibrium aqueous drug solubility. This in turn is135expected to enhance drug flux. The potential for super-136saturation to enhance the oral bioavailability of PWSD has137received increasing recent interest.10,16 For LBF, supersatura-138tion may be generated by the loss of drug solubilization139capacity resulting from the digestion of triglycerides17−19 and/140or surfactants,20 and the dilution of cosolvents21 during GI141processing.19 In contrast, the possibility that supersaturation142may result directly from interactions between lipid colloidal143phases and biliary fluids has been almost entirely ignored.144Although traditional micellar solubilization models suggest that145increases in bile salt concentrations increase drug solubilization,146previous studies have also shown that dilution of lipid colloidal147phases with model intestinal fluids (containing bile salts,148phospholipid and cholesterol) may lead to the generation of149less lipid-rich colloidal phases with lowered solubilization150capacities.22 This provides a plausible mechanism for super-151saturation generation in the small intestine and has been152examined in detail here.

Figure 1. Alternative mechanisms of drug absorption from intestinallipid colloidal phases. In collisional drug absorption (left panel), inaddition to the diffusion of free drug molecules (a), lipid colloidalphases collide with the absorptive membrane and facilitate directtransfer of solubilized drug into absorptive cells (b). Collisionaltransfer may or may not be receptor-mediated. In supersaturation-enhanced drug absorption (right panel), the interaction betweensecreted bile and lipid colloidal phases leads to drug supersaturation(possibly via stimulation of phase changes to less lipid-rich colloidswith lowered drug solubilization capacity). The increase inthermodynamic activity manifests in increases in free drugconcentration, and enhanced diffusional flux (a) across the absorptivemembrane.

Molecular Pharmaceutics Article

dx.doi.org/10.1021/mp3006566 | Mol. Pharmaceutics XXXX, XXX, XXX−XXXB

153 The data suggest, that under the conditions explored,154 collisional drug absorption has little impact on drug absorption155 from intestinal colloidal species. In contrast, supersaturation-156 enhanced drug absorption, mediated by the interaction between157 secreted bile and lipid colloidal phases, may provide an158 endogenous mechanism to promote supersaturation and to159 facilitate drug absorption from lipid colloidal phases.

160 ■ EXPERIMENTAL SECTION161 Materials. Cinnarizine, flunarizine dihydrochloride, mon-162 ensin sodium, sodium taurocholate, sodium taurodeoxycholate,163 sodium glycocholate, sodium glycochenodeoxycholate, choles-164 terol, L-α-lysophosphatidylcholine (LPC, from egg yolk), L-α-165 phosphatidylcholine (PC, from dried egg yolk), oleic acid,166 caprylic acid, monocaprylin, N-hydroxysulfosuccinimide so-167 dium, dicyclohexylcarbodiimide solution (60% w/v in xylenes),168 N,N-dimethylformamide, TWEEN 80, potassium dihydrogen169 phosphate (KH2PO4) and sodium chloride (NaCl) were170 obtained from Sigma-Aldrich, Australia. Sodium taurocheno-171 deoxycholate, sodium glycodeoxycholate, ortho-phosphoric172 acid 85% (H3PO4), sodium hydroxide pellets (NaOH), tert-173 butyl methyl ether (TBME), dimethyl sulfoxide (DMSO),174 glacial acetic acid and absolute ethanol were from Merck,175 Australia. Disodium hydrogen orthophosphate (Na2HPO4),176 sodium dihydrogen orthophosphate (NaH2PO4·2H2O) and177 ammonium dihydrogen orthophosphate (NH4H2PO4) (Ajax178 Finechem, Australia), cholesterol, [4-14C] (49.8 mCi/mmol),179 and Irga-Safe Plus (Perkin-Elmer Life Sciences, Waltham, MA),180 oleic acid, [9,10-3H(N)] (60 Ci/mmol) (American Radio-181 labeled Chemicals, St. Louis, MO), Block Lipid Transport-1182 (BLT-1, i.e., 2-hexyl-1-cyclopentanone thiosemicarbazone)183 (Chembridge, San Diego, CA), ezetimibe (Jai Radhe Sales,184 AMD, India), heparin sodium injection BP (1000 IU/mL,185 Hospira, Australia), xylazine (100 mg/mL, Troy Laboratories,186 Australia), acepromazine (10 mg/mL, Ceva Delvet, Australia),187 ketamine (100 mg/mL, Provet, Australia) and pentobarbitone188 sodium (325 mg/mL, Virbac, Australia) were obtained from189 listed suppliers. Acetonitrile, methanol and chloroform used190 were analytical reagent grade. Water was obtained from a191 Millipore Milli-Q Gradient A10 water purification system192 (Millipore, Billerica, MA).193 Sulfo-N-succinimidyl Oleate (SSO) Synthesis. SSO syn-194 thesis was adapted from the procedure of Harmon et al.23

195 Briefly, dicyclohexylcarbodiimide (1.26 mmol) and N-hydrox-196 ysulfosuccinimide (sodium salt, 1.20 mmol) were added to a197 solution of oleic acid (1.20 mmol) dissolved in N,N-198 dimethylformamide (2 mL), and the reaction was stirred at199 room temperature overnight. The precipitated dicyclohexylurea200 was removed by filtration and ethyl acetate (2 mL) added to201 the filtrate, which was left to stand at 4 °C overnight. SSO202 (precipitate) was then collected by filtration and dried under203 vacuum (1 mmHg). SSO identity was confirmed by NMR and204 mass spectrometry.205 Experimental outline. To assess the role of receptor-206 mediated collisional drug absorption, cinnarizine bioavailability207 was assessed after administration of a lipid emulsion208 formulation in the absence and presence of BLT-1, SSO and209 ezetimibe. BLT-1, SSO and ezetimibe are chemical inhibitors of210 SR-BI,24 CD3625 and NPC1L1,26 respectively. The possibility211 of endocytosis-mediated uptake was also investigated by the use212 of a general endocytosis inhibitor, monensin.213 The role of collisional drug absorption was assessed more214 generically using an in situ rat jejunum perfusion model to

215compare the absorptive flux of cinnarizine from two distinctly216different lipid colloidal phases (micelles vs vesicles) with217matched drug solubilization capacities. Colloidal systems with218the same total solubilization capacity, loaded with drug at the219same concentration, have the same thermodynamic activity and220therefore Cfree is the same in both cases. Under these221circumstances, comparison of the flux profiles obtained from222two structurally different colloids, but with identical Cfree,223provides a means of determining whether the nature of the224colloid, or Cfree, is the principal determinant of absorption.225Generation of identical flux profiles from both systems would226therefore confirm the dependence of flux on thermodynamic227activity and free concentration, whereas a significant difference228in flux would indicate a role for factors beyond Cfree in229determining flux. These include the potential for collisional230drug absorption mechanisms since collision rates are a231statistical function of particle number and are expected to be232markedly higher for micelles (where the smaller particle size233results in higher particle numbers) when compared to vesicles.234To assess the potential for intestinal fluids to enhance drug235absorption from lipid colloidal phases via the induction of drug236supersaturation, whole bile was collected from fasted rats and237mixed with model micelles and vesicles to simulate the process238of interaction with bile in vivo. The potential for bile to generate239drug supersaturation was evaluated in vitro by assessing changes240in cinnarizine solubility, and by monitoring the kinetics of241cinnarizine solubilization and precipitation, following bile242addition to cinnarizine-loaded micelles and vesicles. Sub-243sequently, the impact of drug supersaturation on the intestinal244absorptive flux of cinnarizine from micelles and vesicles was245assessed in an in situ rat jejunum perfusion model, with and246without coperfusion of donor bile. Finally, the relevance of bile-247induced drug supersaturation in vivo was assessed via248examination of changes to cinnarizine bioavailability after249administration of drug-loaded micelles and vesicles (with250matched thermodynamic activity) in bile-intact vs bile-diverted251rats.252Formulation Preparation. Lipid Emulsion. The lipid253emulsion (3 mL per dose) consisted of 1 mg of cinnarizine254and 49 mg of oleic acid solubilized in 8 mM sodium255taurocholate, 2 mM phosphatidylcholine, 2 mM cholesterol256and trace amounts of 14C-cholesterol (1 μCi/3 mL) and/or 3H-257oleic acid (3 μCi/3 mL). The emulsion was prepared in 7.5 mL258batches by weighing appropriate masses of cinnarizine in oleic259acid stock solution (20 mg/g), phosphatidylcholine and260cholesterol into a glass vial, and the mixture made up to261volume with a buffered sodium taurocholate solution (buffer262consisted 18 mM Na2H2PO4·2H2O and 12 mM Na2HPO4).263Appropriate volumes of 14C-cholesterol, 3H-oleic acid, 5 mg/264mL BLT-1 in ethanol (for SR-BI inhibition experiments only),26525 mg/mL SSO in DMSO (for CD36 inhibition experiments266only) and 10 mg/mL monensin in ethanol (for endocytosis267inhibition experiments only) were spiked into the vial and268vortexed for 1 min. The formulation was emulsified by269ultrasonification with a Misonix XL 2020 ultrasonic processor270(Misonix, Farmingdale, NY) equipped with a 3.2 mm271microprobe tip running at an amplitude of 240 μm and a272frequency of 20 kHz for 1.5 min. The total solvent273concentration in the emulsion was ≤2.5% v/v. The emulsion274was used within 4 h of preparation, and the concentration of275drug and labeled cholesterol and/or oleic acid was assayed276before dosing (in duplicate) to confirm compound content in277the emulsion and to allow for dose normalization between rats.


dx.doi.org/10.1021/mp3006566 | Mol. Pharmaceutics XXXX, XXX, XXX−XXXC

278 Intravenous Formulation. The formulation (1 mL per279 dose) used for intravenous administration of cinnarizine and280

14C-cholesterol comprised 0.5% w/v Tween 80 in buffer (36281 mM Na2HPO4 and 22 mM KH2PO4, adjusted to pH 4 with282 acetic acid). Cinnarizine and cholesterol were added to the283 formulation by spiking 5% v/v DMSO (containing 10 mg/mL284 cinnarizine) and 5% v/v ethanol (containing 5 mg/mL285 cholesterol and 40 μCi/mL 14C-cholesterol) into the micellar286 solution. The formulation was mixed by vortexing, and the287 concentration of drug and labeled cholesterol was assayed288 before dosing to confirm compound content in the formulation289 and to allow for dose normalization between rats. The290 formulation was used within 1 h of preparation.291 Model Micelles and Vesicles. The preparation of model292 micelles and vesicles was guided by the methods and phase293 diagram published by Kossena et al.27 Medium-chain lipid294 containing colloids were chosen over long-chain systems since295 the former have previously been shown to generate mono-296 phasic micellar and vesicular systems.27 Relatively high lipid297 concentrations were chosen to reflect the species that are298 expected to initially form during the digestion of medium-chain299 triglycerides.18 The model colloids consisted of tricaprylin300 digestion products (caprylic acid and monocaprylin) solubilized301 in simulated endogenous intestinal fluid (SEIF). SEIF302 comprised the six most prevalent bile salts in human bile,28

303 lysophosphatidylcholine (LPC) and cholesterol. The total bile304 salt:LPC:cholesterol molar ratio was maintained at 16:4:1,305 reflecting known ratios within fasted human bile.29,30 The306 combination of bile salts used here comprised 25 mol % sodium307 glycocholate, 17.5 mol % sodium glycodeoxycholate, 25 mol %308 sodium glycochenodeoxycholate, 12.5 mol % sodium taur-309 ocholate, 7.5 mol % sodium taurodeoxycholate and 12.5 mol %310 sodium taurochenodeoxycholate. The concentration ratios of311 the bile salts were chosen based on average concentrations of312 the six most prevalent bile salts found in human bile.28 The313 caprylic acid:monocaprylin molar ratio was kept at 2:1,314 reflecting the ratio of digestion products expected on digestion315 of 1 mol of triglyceride. The concentration of micellar and316 vesicular components was varied by trial and error (but317 maintaining the ratios described above) to identify systems with318 similar drug solubilization capacities. It has previously been319 shown that the thermodynamic activity (i.e., free concen-320 tration) of drug in a solubilized system may be estimated via321 assessment of solubility behavior, such that having different322 colloidal solutions containing drug at a fixed proportion of the323 saturated solubility results in matched free concentrations.31

324 Thus, drug was loaded into either micellar or vesicular systems325 at the same concentration (and the same proportion of326 saturated solubility) and was therefore present at the same327 thermodynamic activity (i.e., Ctotal, Ccolloid and Cfree were the328 same in both micellar and vesicular systems). The compositions329 of the identified micellar and vesicular systems are shown in

t1 330 Table 1.331 SEIF (8 mM total bile salt:2 mM LPC:0.5 mM cholesterol)332 was prepared in 50 mL batches. Briefly, LPC and cholesterol333 were dissolved in 1 mL of chloroform in a round-bottom flask,334 followed by solvent evaporation under vacuum. The thin film335 formed by solvent evaporation was reconstituted with buffered336 bile salts solution (8 mM total bile salt, 18 mM337 NaH2PO4·2H2O and 12 mM Na2HPO4, 100 mM NaCl),338 vortexed for 1 min and allowed to equilibrate at room339 temperature overnight. When vesicles were prepared, a similar340 procedure was adopted, but in this case SEIF was diluted 4-fold

341with buffer (18 mM NaH2PO4·2H2O and 12 mM Na2HPO4,342108 mM NaCl) to reduce the bile salt:lipid concentration ratio.343Micelles and vesicles were prepared in 10 mL batches by adding344caprylic acid and monocaprylin (quantities in Table 1) to SEIF,345followed by pH adjustment to 6.30 with solid NaOH and346vortexing for 1 min. The phases were then ultrasonicated as347described earlier (30 s continuous ultrasonication followed by348pulsatile, 1 s on/1 s off ultrasonication for 5 min). When349included in the colloids, cinnarizine was predissolved in caprylic350acid, and the drug/fatty acid solution was allowed to equilibrate351overnight prior to micelle/vesicle preparation.352Particle Sizing. The particle size of model micelles and353vesicles was determined by photon correlation spectroscopy354(Malvern Instruments Nano-ZS Zetasizer, Malvern, U.K.).355Micelles had a mean particle size of 9 ± 1 nm [polydispersity356index 0.128 ± 0.040], and vesicles had a mean particle size of357443 ± 32 nm [polydispersity index 0.456 ± 0.038]. Data358reported are mean ± SEM of n = 3 determinations.359Equilibrium Solubility of Cinnarizine in the Model360Micelles and Vesicles. Excess solid cinnarizine was added to3612 mL micelles or vesicles in glass vials. Vials were briefly362vortexed and incubated at 37 °C and samples taken every 24 h363over a period of 120 h. During sampling, vials were centrifuged364(2200g, 10 min, 37 °C), 50 μL of supernatant was sampled, and365vials were revortexed. Equilibrium solubility was defined when366drug concentrations in consecutive samples varied by ≤5% and367was determined on three separate occasions.368Equilibrium solubility of cinnarizine was also determined369after 1:1 addition of bile/bile pH 6.30/buffer pH 6.30 to the370different colloidal phases. The pH of fresh bile was adjusted to3716.30 with H3PO4. Buffer pH 6.30 consisted of 18 mM372NaH2PO4·2H2O, 12 mM Na2HPO4 and 108 mM NaCl.373Kinetics of Cinnarizine Precipitation after the374Addition of Bile to Model Micelles and Vesicles. The375kinetics of cinnarizine precipitation was monitored after376addition of bile to the micellar and vesicular phases, to377determine whether a period of drug supersaturation existed378prior to drug precipitation. In a temperature and stirring rate-379controlled vessel, 2.5 mL of bile was added to 2.5 mL of380micelles or vesicles containing 0.2 mg/mL cinnarizine (∼80%381saturated solubility). Samples (100 μL) were taken before the382addition of bile, and at 1, 10, 20, 30, 40, 50, 60, 80, 100, 120383min after bile addition. Samples were immediately centrifuged384(2200g, 5 min, 37 °C) to separate precipitated drug, and 50 μL385of supernatant was assayed for drug content. The proportion of386the initial solubilized cinnarizine concentration that remained387solubilized after bile addition was assessed as the percent of the388drug mass remaining in solution (i.e., concentration in the

Table 1. Composition of Model Micelles and Vesiclesa

concn (mM)

total bilesaltb LPCc cholesterol

caprylicacid monocaprylin

micelles 8 2 0.5 69.3 34.7vesicles 2 0.5 0.125 52.0 26.0

aMicelles and vesicles also consist of 18 mM NaH2PO4·2H2O and 12mM Na2HPO4. Sodium strength was adjusted to 150 mM with NaCl.Final pH of phases was adjusted to 6.30 ± 0.01. bTotal bile salt consistof 25 mol % sodium glycocholate, 17.5 mol % sodium glycodeox-ycholate, 25 mol % sodium glycochenodeoxycholate, 12.5 mol %sodium taurocholate, 7.5 mol % sodium taurodeoxycholate, 12.5 mol% sodium taurochenodeoxycholate. cLPC is lysophosphatidylcholine.


dx.doi.org/10.1021/mp3006566 | Mol. Pharmaceutics XXXX, XXX, XXX−XXXD

389 supernatant multiplied by the volume remaining in vessel)390 relative to the total drug mass in the vessel at each time point.391 Solid-State Analysis of the Cinnarizine Precipitate392 Using Polarized Light Microscopy. Selected cinnarizine393 pellets from the precipitation kinetics experiments were394 analyzed using a Zeiss Axiolab microscope (Carl Zeiss,395 Oberkochen, Germany) equipped with crossed polarizing396 filters. At the end of the precipitation kinetics experiments,397 1.5 mL of remaining bile + colloid mixture was centrifuged398 (2200g, 10 min, 37 °C), the supernatant was discarded, and a399 small amount of pellet was carefully placed on a microscope400 slide. Samples were analyzed under cross-polarized light, and401 images were recorded using a Canon PowerShot A70 digital402 camera (Canon, Tokyo, Japan).403 Animals. All rat studies were approved by the institutional404 animal ethics committee and were conducted in accordance405 with the guidelines of the Australian and New Zealand Council406 for the Care of Animals in Research and Teaching. Male407 Sprague−Dawley rats (280−330 g) were used in all experi-408 ments and were allowed to acclimatize in the institutional409 animal housing facility for at least 7 days with free access to410 standard chow and water. All animals were fasted overnight411 (12−18 h) prior to surgery.412 Surgical Procedures. Anesthesia was induced in rats by413 subcutaneous injection of 1.0 mL/kg of “Cocktail I” (37.3 mg/414 mL ketamine, 9.8 mg/mL xylazine, 0.4 mg/mL acepromazine415 in saline), and maintained throughout the study with416 subcutaneous doses of 0.44 mL/kg of “Cocktail II” (90.9417 mg/mL ketamine, 0.9 mg/mL acepromazine) when required.418 Rats were maintained on a 37 °C heated pad throughout419 surgery and experiments. At the end of all experiments, rats420 were euthanized via an intravenous or intracardiac injection of421 100 mg of sodium pentobarbitone.422 Cinnarizine Bioavailability Studies following Intraduode-423 nal Administration. The surgical procedures for the conduct of424 bioavailability studies included cannulations of the right carotid425 artery, right jugular vein, duodenum (1 cm below pylorus) and426 common bile duct (only for bile-diverted rats). The surgical427 procedures for the cannulations were as described previ-428 ously.32,33

429 Fasted Rat Bile Collection. The bile duct was cannulated430 near the hilum of the liver (where the duct is free of pancreatic431 tissue) in order to facilitate the collection of bile fluid without432 contamination by exocrine pancreatic secretions.34 Rats were433 rehydrated via saline infusion (1.5 mL/h) into a cannula434 inserted into the right jugular vein, and bile was continuously435 collected for 5 h. The concentration of total bile salt in436 collected bile was assayed using a validated enzymatic437 colorimetric assay (Total Bile Acids kit #431-15001; Wako438 Pure Chemical Industries, Osaka, Japan) on a plate reader439 (Fluostar Optima plate reader, BMG Labtechnologies,440 Germany) measuring absorbance at a wavelength of 540 nm.441 In all subsequent experiments, bile was used within 24 h of442 collection.443 Single-Pass Rat Jejunum Perfusion. The model employed444 to assess flux across rat jejunum involved in situ perfusion445 (single-pass) of an isolated jejunal segment and simultaneous446 blood collection from the corresponding mesenteric vein447 branch. The surgical procedures for the perfusion studies are448 similar to those described elsewhere, with slight modifica-449 tions.35 Briefly, the right jugular vein was cannulated to enable450 infusion of donor blood. A piece of jejunal segment (∼10 cm2)451 was isolated and cannulated at the proximal and distal ends

452with sections of Teflon tubing (0.03 in. i.d. proximal/inlet,453Upchurch Scientific, Oak Harbor, WA; 0.0625 in. i.d. distal/454outlet, Shimadzu, Kyoto, Japan). Jejunal contents were initially455flushed with warm perfusion buffer (150 mM Na+, 18 mM456H2PO4

−, 12 mM HPO42−, 108 mM Cl−, adjusted to pH 6.30 ±4570.01). The mesenteric vein draining the jejunal segment was458then isolated, the rat heparinized (90 IU/kg) via the jugular459vein and the mesenteric vein immediately catheterized. A drop460of superglue was placed over the site of catheterization, and461silicone tubing (0.025 in. i.d., Helix Medical, CA) attached to462the catheter tip for the collection of venous blood. Immediately463following catheterization of the mesenteric vein and for the464remainder of the experiment, rats were infused with heparinized465donor rat blood (5 IU/mL) via the jugular vein, and the rate466(0.3 mL/min) was adjusted based on the outflow from the467mesenteric blood.468Cinnarizine Bioavailability after Intraduodenal Infu-469sion in Anesthetized Rats. A 30 min equilibration period470was allowed between the end of surgery and drug dosing. To471examine the impact of lipid uptake receptors on drug472absorption, studies were conducted in the presence or absence473of lipid uptake inhibitors. To examine the impact of bile-474induced drug supersaturation on drug absorption, studies were475conducted in bile-intact or bile-diverted rats. Colloidal systems476(lipid emulsion, micelles, vesicles) containing cinnarizine were477infused into the duodenum of rats at a rate of 1.5 mL/h for 2 h.478When the dose infusion was complete, saline was infused at a479rate of 1.5 mL/h for 10 min to flush any remaining formulation480in the tubing into the duodenum. Blood samples (0.3 mL) were481collected via the carotid artery cannula up to 8 h after infusion482initiation into tubes containing 3 IU heparin. The sampling483intervals were as follows: t = 0, 1, 2, 2.5, 3, 4, 6, 8 h for the484receptor inhibition studies; and t = 0, 1, 1.5, 2, 3, 4, 6, 8 h for485the bile-induced drug supersaturation studies. After each blood486sample was taken, the cannula was flushed with 0.3 mL of 2 IU/487mL heparinized saline to ensure cannula patency, and to replace488the volume of blood removed. Plasma was separated by489centrifugation (10000g, 5 min) to enable analysis of drug and490labeled lipid content.491Administration of Lipid Uptake Inhibitors (BLT-1, SSO,492Ezetimibe) and Endocytosis Inhibitor (Monensin). Ezetimibe493has previously been dosed at 0.3 mg/kg intravenously into rats,494and has been shown to inhibit cholesterol absorption (from an495intraduodenally dosed lipid emulsion) without reports of496toxicity.36 Therefore, in our study, intravenous administration497of ezetimibe (0.3 mg/kg via the jugular vein) was selected as498the route to administer the inhibitor at the beginning of the 30499min equilibration period. An appropriate volume of 5 mg/mL500ezetimibe in ethanol was spiked into blank rat plasma, and 0.8501mL of resultant plasma was dosed into rats as an intravenous502bolus. The total ethanol concentration was less than 2.5% v/v.503In the case of BLT-1 and SSO, the inhibitors had not been504previously administered intravenously, therefore local (i.e.,505intestinal) administration was selected to limit the systemic506effects of the inhibitors. Monensin was also administered507directly into the intestine to minimize systemic endocytosis508inhibition. Here, 100 μM BLT-1, 1 mM SSO and 100 μM509monensin were preinfused intraduodenally (in saline) at a rate510of 1.5 mL/h during the 30 min equilibration period, and511subsequently coinfused at the same concentration as part of the512lipid emulsion. BLT-1 and SSO have previously been shown to513inhibit lipid uptake in cell-based studies at concentrations of 1−


dx.doi.org/10.1021/mp3006566 | Mol. Pharmaceutics XXXX, XXX, XXX−XXXE

514 10 μM24 and 400 μM,23 respectively. Monensin has previously515 been shown to inhibit endocytosis in cultured cells at 10 μM.37

516 Intravenous administration studies were also conducted in517 control rats and monensin-treated rats to assess the effect of518 endocytosis inhibition on the systemic distribution and519 clearance of cinnarizine and cholesterol. In these studies,520 blank (i.e., not containing cinnarizine and 14C-cholesterol) lipid521 emulsion (with or without 100 μM monensin) was infused522 intraduodenally as described above, and intravenous infusion of523 the cinnarizine and 14C-cholesterol containing intravenous (iv)524 formulation was commenced at the same time. The iv525 formulation (1 mL) was infused into the right jugular vein at526 a rate of 0.05 mL/15 s (total infusion period <5 min). Blood527 samples (0.3 mL) were collected via the carotid artery cannula528 at t = 5, 15, 30, 60, 120, 180, 240, 360, 480 min after infusion529 initiation into tubes containing 3 IU of heparin. After each530 blood sample was taken, the cannula was flushed with 0.3 mL of531 2 IU/mL heparinized saline. Plasma was separated by532 centrifugation (10000g, 5 min) to enable analysis of drug and533 labeled cholesterol content.534 Intestinal Absorptive Flux Assessment via in Situ535 Single-Pass Rat Jejunum Perfusion. After surgery, animals536 were equilibrated for 30 min, during which time heparinized537 donor rat blood was infused via the jugular vein as described538 above. During re-equilibration, blood from the cannulated539 mesenteric vein (∼0.3 mL/min) was collected for reinfusion.540 Perfusion buffer was pumped through the jejunal segment at a541 rate of 0.1 mL/min and outflowing buffer discarded to waste.542 The exposed jejunal segment was kept moist by covering with543 saline-soaked gauze throughout the experiment.544 In all experiments, the concentration of cinnarizine in the545 perfusate was held at 0.1 mg/mL (∼40% saturated solubility).546 Therefore, in experiments where micelles or vesicles were547 perfused alone, cinnarizine was loaded into the perfusate at 0.1548 mg/mL. In experiments where micelles or vesicles were549 coperfused in a 1:1 v/v ratio with a secondary perfusate of550 bile/bile pH 6.30/buffer pH 6.30, cinnarizine was loaded into551 the primary perfusate at 0.2 mg/mL, such that 1:1 v/v dilution552 led to a final perfusate concentration of 0.1 mg/mL.553 Perfusate flow was maintained at 0.1 mL/min in all554 experiments to minimize variations in the thickness of the555 unstirred water layer.38 For experiments where 1:1 v/v556 coperfusion of the phases with bile/bile pH 6.30/buffer pH557 6.30 was required, micelles/vesicles and bile/buffer were558 pumped at 0.05 mL/min, and mixed via a three-way “T”559 connector immediately prior to entry into the jejunal segment,560 providing a total perfusate flow of 0.1 mL/min. Perfusate was561 sampled at t = 0 to confirm lipid and drug concentrations. After562 this time, the outflowing perfusate was continuously collected563 at 10 min intervals and briefly vortexed before samples were564 taken for analysis of drug and lipid content. For experiments565 where drug supersaturation was generated, perfusate samples566 were taken before and after centrifugation (2200g, 2 min), to567 assess the degree of drug precipitation within the jejunal568 segment. Blood draining the perfused jejunal segment was569 collected at 5 min intervals, plasma was separated by570 centrifugation (10000g, 5 min) and samples were taken for571 analysis of drug content by HPLC as described below.572 Analytical Procedures. Sample Preparation and HPLC573 Assay Conditions for Cinnarizine. Samples of lipid emulsion574 were prepared for HPLC assay by an initial 80-fold dilution575 with chloroform:methanol (2:1 v/v), followed by a 10-fold576 dilution with mobile phase (50% v/v acetonitrile:50% v/v 20

577mM NH4H2PO4). Samples of iv formulation and micelles/578vesicles were prepared for HPLC assay by a 900-fold and a 400-579fold dilution with mobile phase (50% v/v acetonitrile:50% v/v58020 mM NH4H2PO4), respectively. Plasma samples were581prepared for HPLC using a validated extraction procedure,582with flunarizine as an internal standard, as reported583previously.39

584Cinnarizine HPLC assay conditions were as described585previously,39 with slight modification to the mobile phase586employed for cinnarizine quantification in plasma, to 45% v/v587acetonitrile:55% v/v 20 mM NH4H2PO4 in this study. Replicate588analysis of n = 4 quality control samples revealed acceptable589accuracy and precision (±10%, ±15% at the limit of590quantification) for concentrations between 20 and 1000 ng/591mL for iv formulation, emulsion, micelles and vesicles, and 10592and 320 ng/mL for plasma.593Scintillation Counting. Quantification of 14C-cholesterol594and 3H-oleic acid in the plasma was performed via scintillation595counting on a Packard Tri-Carb 2000CA liquid scintillation596analyzer (Packard, Meriden, CT, USA). Plasma samples (50597μL) were added to 2 mL of Irga-safe Plus scintillation fluid598followed by a 10 s vortex. Samples were corrected for599background radioactivity by the inclusion of a blank plasma600sample in each run. The counting method was validated by601spiking blank plasma with low, medium and high concen-602trations of labeled cholesterol and oleic acid (in triplicate). The603measured concentrations were within 5% of the nominal604concentration.605Blood:Plasma Ratio Determination for Cinnarizine. The606blood:plasma ratio for cinnarizine was determined by spiking6070.5 mL of blank blood with known amounts of cinnarizine to608achieve low, medium and high concentrations (in triplicate).609Plasma was separated by centrifugation (10000g, 5 min) and610plasma drug concentration assayed by HPLC. The blood:-611plasma ratio was calculated from the ratio of known612concentration in spiked blood to the concentration measured613in plasma separated from spiked blood. The mean blood:-614plasma ratio was subsequently used to convert plasma615concentrations to blood concentrations in perfusion experi-616ments, enabling quantification of total drug transport into617mesenteric blood.618Calculations. In the single-pass rat jejunum perfusion model,619permeability coefficients were calculated using steady-state drug620concentrations in perfusate and blood. Two apparent621permeability coefficients (Papp) were calculated as described622previously:35

= − ·PCC

disappearanceQA

lnapp1

0 623(2)

=·⟨ ⟩

ΔΔ( )

PA C

appearance

Mt

app

B

624(3)

625where “disappearance Papp” is the apparent permeability626coefficient calculated from drug loss from the perfusate (cm/627s); “appearance Papp” is the apparent permeability coefficient628calculated from drug appearance in the mesenteric blood (cm/629s); Q is the perfusate flow rate (mL/s); A is the surface area of630the perfused jejunal segment (cm2), which is calculated by631multiplying the diameter by the length of the perfused intestinal632segment as described previously;40 C1 is the average steady-633state drug concentration exiting the perfused jejunal segment634(ng/mL); C0 is the drug concentration entering the jejunal


dx.doi.org/10.1021/mp3006566 | Mol. Pharmaceutics XXXX, XXX, XXX−XXXF

635 segment (ng/mL); ΔMB/Δt is the average rate of drug mass636 appearance in mesenteric blood at steady state (ng/s); and ⟨C⟩637 is the logarithmic mean drug concentration in the lumen (ng/638 mL), where ⟨C⟩ = (C1 − C0)/(ln C1 − ln C0).639 Noncompartmental Pharmacokinetic Analysis. The max-640 imum plasma concentration (Cmax), time to reach Cmax (Tmax),641 area under the plasma concentration−time curve from time642 zero to the last measured concentration (AUC0−8h), area under643 the plasma concentration−time curve extrapolated to infinity644 (AUC0−inf), elimination rate constant (Ke), volume of645 distribution (Vd) and clearance (Cl) were calculated using646 WINONLIN version 5.3 (Pharsight Inc., Apex, NC, USA).647 Statistical Analysis. Results were analyzed using Student’s648 t test. A P value of <0.05 was considered to be a significant649 difference. Analyses were performed using SPSS v19 for650 Windows (SPSS Inc., Chicago, IL, USA).

651 ■ RESULTS

652 SR-BI, CD36, NPC1L1 and Endocytosis Have Little653 Impact on Drug Absorption from Intestinal Colloidal654 Phases. Inhibition of the lipid uptake receptors SR-BI (by655 coinfusion of 100 μM BLT-1), CD36 (by coinfusion of 1 mM656 SSO) and NPC1L1 (by intravenous administration of 0.3 mg/657 kg ezetimibe) did not result in significant changes to the658 systemic plasma concentration−time profiles or systemic

f2t2 659 exposure of cinnarizine (Figure 2A, Table 2). Inhibition of660 endocytosis (by coinfusion of 100 μM monensin) unexpectedly

661led to significant increases in cinnarizine systemic plasma662concentration at t = 2.5 and 4 h (Figure 2A), and increased the663AUC0−8h 1.5-fold (Table 2). However, subsequent intravenous664dosing studies revealed significantly lower Vd and Cl values in665the monensin-treated rats when compared to control (Table S1666in the Supporting Information), suggesting that the increase in667cinnarizine systemic exposure was due to a decrease in668cinnarizine systemic distribution and clearance, rather than669changes to intestinal absorption. The data suggest a limited role670for SR-BI, CD36, NPC1L1 and endocytosis generally, in the671absorption of cinnarizine from micelles and vesicles.672In contrast, inhibition of NPC1L1 did lead to significantly673lower systemic plasma concentrations of exogenously dosed674cholesterol at t = 2, 2.5, 3, 4, 6, 8 h (Figure 2B). Plasma675concentrations of exogenously dosed oleic acid also appeared676lower in rats following inhibition of CD36 although differences677were not significantly different (Figure 2C). While inhibition of678SR-BI and CD36 did not lead to significant changes to the679systemic plasma concentration−time profile of exogenously680dosed cholesterol, inhibition of endocytosis significantly681increased the systemic plasma concentrations of exogenously682dosed cholesterol at t = 4, 8 h (Figure 2B). This increase in683exposure is likely explained by a decrease in systemic684distribution of 14C-cholesterol in the monensin-treated rats,685as plasma 14C-cholesterol concentrations at early sample time686points (which reflect the distribution phase) in intravenous687studies were significantly higher in the monensin-treated rats688(Figure S1B in the Supporting Information) when compared to689controls (Vd and Cl could not be calculated for cholesterol as a690typical elimination phase was not evident in the systemic691plasma concentration vs time profiles of cholesterol: Figure 2B692and Figure S1B in the Supporting Information).693Drug Absorption from Micelles and Vesicles Is694Determined by Cfree and Not Colloidal Structure. The695intestinal perfusion of colloidal media with markedly different696compositional profiles (micelles and vesicles) but with697comparable Cfree and thermodynamic activity did not result in698significant differences in steady-state absorptive flux, disappear-699 t3f3ance Papp or appearance Papp of cinnarizine (Table 3, Figure 3A:700filled symbols). Similarly, in spite of large differences in particle701size and composition, the intraduodenal infusion of the same702micellar or vesicular systems to bile-diverted rats did not result703in significant differences in systemic plasma concentration−704 f4time profiles (Figure 4: filled symbols) and pharmacokinetic705 t4t5parameters of cinnarizine (Table 5: micelles vs vesicles in bile-

Figure 2. Systemic plasma concentration−time profiles of (A)cinnarizine (CIN), (B) 14C-labeled cholesterol (Ch) and (C) 3H-labeled oleic acid (OA) following intraduodenal infusion of a 3 mLlipid emulsion consisting of 1 mg of cinnarizine emulsified in 59 mMoleic acid, 8 mM sodium taurocholate, 2 mM phosphatidylcholine, 2mM cholesterol, 1 μCi of 14C-cholesterol and/or 3 μCi of 3H-oleicacid. Experiments were performed in control rats (filled circle); ratstreated with 100 μM BLT-1 (open circle); rats treated with 1 mM SSO(open triangle); rats treated with 0.3 mg/kg ezetimibe (open square);and rats treated with 100 μM monensin (open diamond), to inhibitthe lipid uptake receptors SR-BI, CD36, NPC1L1 and endocytosis,respectively. BLT-1, SSO, and monensin were coinfused as part of thelipid emulsion; ezetimibe was administered intravenously. Datarepresent mean ± SEM of n = 4 rats for A and B; and n = 3 ratsfor C. Statistically significant difference with respect to control rats (p< 0.05) is denoted by the symbol *.

Table 2. Pharmacokinetic Parameters for Cinnarizine afterIntraduodenal Administration of a 3 mL Lipid Emulsion inRatsa

expl group AUC0−8h (ng h/mL) Cmax (ng/mL) Tmax (h)

control 978 ± 122 288.1 ± 26.6 2.4 ± 0.1BLT-1-treated 883 ± 21 219.0 ± 20.0 2.6 ± 0.5SSO-treated 924 ± 88 195.5 ± 33.6 2.4 ± 0.2ezetimibe-treated 1048 ± 92 333.9 ± 13.6 2.3 ± 0.1monensin-treated 1487 ± 103b 432.8 ± 45.9b 2.3 ± 0.1

aIn each case, the formulation type was emulsion and the CIN dosewas 3.33 mg/kg. Experiments were performed in control rats; ratstreated with 100 μM BLT-1; rats treated with 1 mM SSO; rats treatedwith 0.3 mg/kg ezetimibe; and rats treated with 100 μM monensin, toinhibit the lipid uptake receptors SR-BI, CD36, NPC1L1, andendocytosis, respectively. Values represent mean ± SEM of n = 4rats. bSignificant difference when compared to control group.


dx.doi.org/10.1021/mp3006566 | Mol. Pharmaceutics XXXX, XXX, XXX−XXXG

706 diverted rats). The results suggest that cinnarizine absorption707 from lipid colloidal phases is relatively insensitive to the708 physical nature of the colloidal milieu, and is instead controlled709 largely by Cfree.710 Bile-Mediated Dilution of Cinnarizine-Loaded Mi-711 celles and Vesicles Generates Drug Supersaturation.712 The equilibrium solubility and change in cinnarizine solubiliza-713 tion capacity of model micelles and vesicles before and after 1:1714 v/v dilution with bile, bile pH 6.30 or buffer pH 6.30 are

f5 715 tabulated in Table 4, and shown graphically in Figure 5.716 Dilution of the micellar or vesicular systems in a 1:1 v/v ratio717 with bile obtained from donor animals led to significant718 decreases in cinnarizine solubilization capacity. As a proportion719 of initial, the solubilization capacity of the micellar system720 dropped significantly from 100% to 14%, and for vesicles the721 decrease was even greater from 100% to 7%. This was in spite722 of the fact that the bile salt concentration in donor bile was723 higher than that in the micellar or vesicular system (average724 total bile salt concentration in donor bile was 14.7 ± 0.9 mM;725 mean ± SEM, n = 13), and therefore “dilution” with bile did726 not reduce bile salt concentrations below the critical micellar727 concentration (CMC) and instead increased the overall bile salt728 concentration.729 The pH of donor bile was higher than intestinal pH (average730 pH of donor bile was 8.02 ± 0.02; mean ± SEM, n = 5). As731 such additional studies were performed to examine whether the732 effects on solubility reflected a pH effect. 1:1 v/v dilution of the733 phases with pH adjusted bile (pH of bile adjusted to 6.30 to734 match the pH of the micelles and vesicles) also resulted in735 significant decreases in solubilization capacity (to 28% and 15%736 of initial for micelles and vesicles, respectively), although the737 decrease was slightly attenuated when compared to non-pH738 adjusted bile. Finally, the micellar and vesicular systems were

739diluted 1:1 v/v with buffer at pH 6.30 in an attempt to740uncouple simple dilution effects from pH effects and bile-741mediated effects. Dilution with pH 6.30 buffer also decreased742cinnarizine solubilization capacity significantly, although to a743lesser extent (41% and 21% in micelles and vesicles,744respectively).745 f6Analysis of the kinetics of drug precipitation (Figure 6)746demonstrated that bile addition to cinnarizine-loaded micelles747and vesicles (cinnarizine present in phases at 0.2 mg/mL,748∼80% saturated solubility) did not result in immediate drug749precipitation, and was preceded by a period of supersaturation.750The time taken for cinnarizine precipitation to occur was

Table 3. Cinnarizine Disappearance Papp (×106 cm/s) fromthe Intestinal Perfusate, Appearance Papp (×106 cm/s) in theMesenteric Blood and Steady-State Absorptive Flux intoMesenteric Blood (ng/5 min/10 cm2) after 60 min of Single-Pass Perfusion of ∼10 cm2 Segments of Rat Jejunum withModel Micelles and Vesicles, with and without 1:1 v/vCoperfusion with Rat Bile, Rat Bile pH 6.30 or Buffer pH6.30a

SSratiob

disappearancePapp (×10

6

cm/s)

appearancePapp (×10

6

cm/s)

flux into mesentericblood (ng/5 min/10

cm2)

micelles 0.4 19.4 ± 3.0 1.1 ± 0.2 310 ± 50micelles +bileSSc

6.1 29.3 ± 3.0d 3.7 ± 0.1d 979 ± 30d

micelles +bile pH6.30SS

3.0 39.5 ± 5.6d 4.2 ± 1.0d 1099 ± 279d

micelles +buffer pH6.30SS

2.1 38.9 ± 4.1d 4.4 ± 0.4d 1180 ± 100d

vesicles 0.4 22.0 ± 2.2 1.2 ± 0.1 340 ± 29vesicles +bileSS

11.6 66.0 ± 18.3e 2.0 ± 0.7 499 ± 155

aValues calculated using data obtained after steady-state attainment (t= 40−60 min). In all experiments, cinnarizine concentration inperfusate and total perfusate flow rate were kept constant at 0.1 mg/mL and 0.1 mL/min, respectively. Data represent mean ± SEM of n =3−4 experiments. bSS ratio = supersaturation ratio = (supersaturated)concentration of drug in perfusate/equilibrium solubility of drug inperfusate. cSS denotes drug supersaturation in perfusate. dSignificantincrease from micelles alone. eSignificant increase from vesicles alone.

Figure 3. (A) Absorptive flux of cinnarizine (CIN) into mesentericblood (ng/5 min/10 cm2) and (B, C) CIN disappearance fromintestinal perfusate (% drug dose passing through jejunum) whenmicelles/vesicles were perfused through an isolated rat jejunal segment(∼10 cm2), with (open symbols) and without (filled symbols) 1:1 v/vcoperfusion with rat bile. Coperfusion of micelles and vesicles withdonor bile generates drug supersaturation in situ within the perfusedjejunal segment. SS denotes experiments where drug is supersaturatedin the perfusate. The degree of drug precipitation within the perfusateis illustrated in B and C as the difference in perfusate concentrationbetween pre- and postcentrifugation data. Precipitation was significantin the vesicle, but not micellar groups. Experiments were performedusing an in situ single-pass rat jejunum perfusion model. In allexperiments, the concentration of cinnarizine in perfusate and the totalperfusate flow rate were kept constant at 0.1 mg/mL and 0.1 mL/min,respectively. Data represent mean ± SEM of n = 3−4 experiments.

Figure 4. Systemic plasma concentration−time profiles of cinnarizine(CIN) following a 2 h intraduodenal infusion of cinnarizine-loaded(0.2 mg/mL) micelles (left panel) and vesicles (right panel) to bile-intact and bile-diverted rats. Consistent with observations in ratjejunum perfusion studies, bile-induced supersaturation translated intoincreased in vivo exposure during the absorption phase in the case ofmicelles but not vesicles. Data represent mean ± SEM of n = 4 rats.Statistical significance (p < 0.05) is denoted by the symbol *.


dx.doi.org/10.1021/mp3006566 | Mol. Pharmaceutics XXXX, XXX, XXX−XXXH

751 variable, however supersaturation was maintained for longer752 periods in micelles (>20 min in all cases) when compared to753 vesicles (1−20 min).754 Bile-Induced Drug Supersaturation Increases Jejunal755 Absorptive Flux for Micelles but Not Vesicles. The756 absorptive flux vs time profiles for cinnarizine under super-757 saturated conditions (i.e., when bile was coperfused with the758 phases) and under nonsupersaturated conditions (when bile759 was not coperfused with the phases) are shown in Figure 3A.760 Steady-state absorptive flux, disappearance Papp and appearance761 Papp of cinnarizine in all perfusion experiments are reported in762 Table 3.763 Coperfusion of micelles with bile in a 1:1 v/v ratio increased764 the absorptive flux, disappearance Papp and appearance Papp of765 cinnarizine from micelles 3.2-fold, 1.5-fold and 3.4-fold,766 respectively. In contrast, 1:1 v/v coperfusion of vesicles with767 bile did not lead to significant changes in cinnarizine absorptive768 flux or appearance Papp. Disappearance Papp for cinnarizine did769 increase 3.0-fold when vesicles were coperfused with bile,770 however the drop in perfusate drug concentration was largely a771 result of rapid drug precipitation in the perfusate (see below).772 For the micellar preparation, bile-induced supersaturation773 was relatively stable throughout the experimental period. Thus774 the cinnarizine concentration in the perfusate was essentially775 the same before or after centrifugation (Figure 3B). In contrast,776 when vesicles were coperfused with bile, significant drug777 precipitation was observed during the time required for

778perfusate to transit the jejunal segment (as indicated by the779difference between the pre- and postcentrifugation data in780Figure 3C). This was consistent with the in vitro dilution781profiles in Figure 6.782To distinguish between bile-induced increases in cinnarizine783absorptive flux resulting from drug supersaturation, pH784increases (since an increase in pH might be expected to785increase the permeability of a weak base) and nonspecific786effects of bile components on membrane permeability, micelles787were also coperfused with pH-adjusted bile (pH 6.30) and788buffer (pH 6.30). 1:1 v/v coperfusion of micelles with bile pH7896.30 or buffer pH 6.30 increased cinnarizine absorptive flux 3.5-790fold and 3.8-fold, respectively, when compared to the perfusion791 f7of micelles alone (Figure 7, Table 3). Stable supersaturation

Table 4. Equilibrium Solubility (37 °C) and Percent OriginalSolubilization Capacity Values of Cinnarizine in ModelMicelles and Vesicles, before and after 1:1 v/v Addition ofRat Bile, Rat Bile pH 6.30 or Buffer pH 6.30a

equilibrium solubility(μg/mL)

percent original solubilizationcapacity (%)b

Micellesmicelles alone 236 ± 9.3 100

+ bile (1:1) 16.5 ± 0.8 14.0 ± 0.7+ bile pH6.30 (1:1)

33.1 ± 1.0 28.0 ± 0.9

+ buffer pH6.30 (1:1)

48.7 ± 0.8 41.2 ± 0.6

Vesiclesvesicles alone 232 ± 3.3 100

+ bile (1:1) 8.60 ± 0.5 7.4 ± 0.4+ bile pH6.30 (1:1)

17.4 ± 0.4 15.0 ± 0.3

+ buffer pH6.30 (1:1)

24.9 ± 0.3 21.4 ± 0.2

aData represent mean ± SEM of n = 3−4 determinations. bPercentoriginal solubilization capacity = (solubilityfinal × volumefinal/solubilityinitial × volumeinitial) × 100%.

Table 5. Pharmacokinetic Parameters for Cinnarizine after Intraduodenal Administration of Cinnarizine-Loaded (0.2 mg/mL)Micelles and Vesicles to Bile-Intact and Bile-Diverted Ratsa

exptl group formulation type CIN dose (mg/kg) AUC0−8h (ng h/mL) Cmax (ng/mL) Tmax (h)

micellesbile-intact micelles 2 502 ± 59b 138.8 ± 11.6b 1.6 ± 0.2bile-diverted micelles 2 362 ± 39 85.7 ± 14.3 2.0 ± 0.4

vesiclesbile-intact vesicles 2 411 ± 74 109.4 ± 10.6 2.3 ± 0.3bile-diverted vesicles 2 393 ± 82 97.6 ± 12.8 2.0 ± 0.4

aValues represent mean ± SEM of n = 4 rats. bSignificant increase when compared to micelles (bile-diverted) group.

Figure 5. Percent original cinnarizine (CIN) solubilization capacity ofmodel colloids (micelles or vesicles), before and after a 1:1 v/vaddition of rat bile, rat bile pH 6.30 or buffer pH 6.30. Data representmean ± SEM of n = 3−4 determinations. Statistical significantdifference (p < 0.05) to colloid only is denoted by the symbol *.

Figure 6. Kinetics of cinnarizine (CIN) precipitation from modelmicelles (filled circles, n = 5) and model vesicles (open circles, n = 4)upon addition of rat bile (in a 1:1 v/v ratio) at t = 0. Addition of bilereduces the equilibrium cinnarizine solubilization capacity of micellesand vesicles to 14% and 7% of initial, respectively (see Table 4, andshown here as the lines denoted equilibrium solubility). Cinnarizinesupersaturation appeared to be maintained for longer in micelles thanin vesicles. Cinnarizine was loaded into micelles and vesicles at 80%saturation (∼0.2 mg/mL). Each line represents individual experiments.


dx.doi.org/10.1021/mp3006566 | Mol. Pharmaceutics XXXX, XXX, XXX−XXXI

792 was also generated within the perfused jejunal segment in these793 experiments (Figure S2 in the Supporting Information:794 cinnarizine concentration in the perfusate was essentially the795 same before and after centrifugation).796 Since 1:1 v/v coperfusion of micelles with bile, bile pH 6.30797 and buffer pH 6.30 are all expected to generate cinnarizine798 supersaturation within the perfused jejunal segment (according799 to the solubility data in Table 4, and the lack of drug800 precipitation in outflow perfusate in all cases), the observation801 that absorptive flux enhancement was similar in all groups802 (Figure 7) suggests that the enhancement was attributable to803 drug supersaturation, and not an increase in system pH (by804 comparing micelles + bile group with micelles + bile pH 6.30805 group), or nonspecific effects of bile on membrane permeability806 (by comparing micelles + bile pH 6.30 group with micelles +807 buffer pH 6.30 group). However, the degree of flux808 enhancement did not correlate with the degree of super-809 saturation, as flux enhancement was similar in all groups despite810 the significantly higher supersaturation ratio generated in the811 micelles + bile group, when compared to the micelles + bile pH812 6.30 group and micelles + buffer pH 6.30 group (super-813 saturation ratio of 6 vs 2−3) (Table 3). Previous studies have814 shown that, for lipophilic drugs such as cinnarizine, the drug815 fraction extracted into octanol (and analogous to the permeable816 fraction) as a function of pH is shifted to lower pHs than would817 be expected based on the un-ionized fraction.41,42 This left shift818 would limit pH effects on permeability over the range of pH 6−819 8, consistent with the observations here.820 Bile-Induced Drug Supersaturation Increases in Vivo821 Cinnarizine Exposure after Intraduodenal Infusion. The822 systemic plasma concentration−time profiles of cinnarizine823 following intraduodenal infusion of cinnarizine-loaded micelles824 and vesicles (0.2 mg/mL; ∼80% saturated solubility) to bile-825 intact and bile-diverted rats are shown in Figure 4. The826 pharmacokinetic parameters of cinnarizine for the bioavail-827 ability studies are reported in Table 5. Consistent with the828 results from the rat jejunal permeability studies, intraduodenal829 administration of micelles to bile-intact rats resulted in

830significantly higher systemic plasma concentrations of cinnar-831izine at t = 1, 1.5 and 2 h (AUC0−8h increased 1.4-fold) when832compared to bile-diverted rats. In contrast, when vesicles were833dosed to bile-intact rats, the systemic plasma concentration−834time profiles and AUC0−8h of cinnarizine were not different835from that observed in bile-diverted rats.

836■ DISCUSSION837After oral administration, the absorption of poorly water-838soluble drugs (PWSD) is often limited by slow dissolution and839low solubility in the GI tract. LBF overcome many of the840dissolution limitations of PWSD (by providing a mechanism to841circumvent dissolution from the solid to liquid state), however,842the solubility limitations of PWSD are seemingly unaddressed,843since solubilization in the lipid-based colloidal phases that result844from the digestion of LBF does not typically enhance free drug845concentrations.2,43 Nonetheless, coadministration of lipids846(either formulation lipids or via coadministration with lipid-847rich foods) remains a highly effective means to promote the848absorption of PWSD. This suggests the potential for alternative849mechanisms by which LBF enhance drug absorption. In the850current communication, two possible alternatives to the851traditional model of drug absorption from LBF have been852explored: first, that drug absorption from lipid colloidal phases853may involve a collisional uptake component (i.e., drug854absorption directly from the solubilized phase); and second,855that flux across the absorptive membrane may be enhanced by a856transient increase in the thermodynamic activity of drug in857intestinal colloidal phases due to supersaturation.858The data describing drug absorption from micellar and859vesicular colloidal phases suggest that direct interactions860between colloids (or at least the systems examined here) and861the absorptive membrane do not play an important role in862cinnarizine absorption. Thus, comparable absorptive flux863(Figure 3A: filled symbols) and systemic plasma concen-864tration−time profiles (Figure 4: bile-diverted rats) were865observed following jejunal perfusion and intraduodenal infusion866of micelles and vesicles. The thermodynamic activity and Cfree867of both systems were held constant, but the large difference in868hydrodynamic radius (9 nm of micelles vs 443 nm of vesicles)869and higher bile salt, LPC and lipid concentrations in the870micellar system (Table 1) suggest that the number of871administered micellar particles was substantially higher than872that of vesicular particles. Collision-mediated absorption is873highly sensitive to increased particle number, since this874increases the statistical likelihood of collisions and collisional875transfer.7 As such collisional interactions did not appear to876dictate the degree of drug absorption from micelles and vesicles877and, instead, absorption was seemingly controlled by878thermodynamic activity (or Cfree). Differences in particle size879might also be expected to alter colloid diffusion across the880unstirred water layer (UWL). The similarity in absorption881profiles from micelles and vesicles (Figure 3A: filled symbols)882therefore suggests that diffusion across the UWL either was883nonlimiting or was not affected by particle size in this884experimental model. Similarly, the rates of replenishment of885Cfree (i.e., the rate of re-establishment of the equilibrium886between solubilized and free drug) might be expected to be887different between the two different particle size colloids,888however this was presumably sufficiently fast in both cases to889have little impact on drug absorption in the current model.890Data obtained from the lipid uptake receptor inhibition891studies further support the notion that collisional uptake

Figure 7. Absorptive flux−time profiles of cinnarizine (CIN) whenmicelles were perfused through an isolated rat jejunal segment (∼10cm2), with and without 1:1 v/v coperfusion with rat bile, rat bile pH6.30 or buffer pH 6.30. SS denotes experiments where drug issupersaturated in perfusate. Coperfusion of rat bile, rat bile pH 6.30 orbuffer pH 6.30 with micelles generates drug supersaturation in situwithin the perfused jejunal segment in all cases, and increasedcinnarizine absorptive flux by 3.2-fold, 3.5-fold and 3.8-fold,respectively. Experiments were performed using an in situ single-passrat jejunum perfusion model. In all experiments, the concentration ofcinnarizine in perfusate and the total perfusate flow rate were keptconstant at 0.1 mg/mL and 0.1 mL/min, respectively. This series ofexperiments provides further support for the suggestion thatsupersaturation was responsible for the increase in absorptive fluxseen in Figure 3A. Data represent mean ± SEM of n = 3−4experiments.


dx.doi.org/10.1021/mp3006566 | Mol. Pharmaceutics XXXX, XXX, XXX−XXXJ

892 mechanisms have a limited role in drug absorption from these893 systems. SR-BI, CD36 and NPC1L1 were examined since they894 have previously been shown to mediate the cellular uptake of895 fatty acids and/or cholesterol,12,15,44 the absorption of which is896 also facilitated by micellar solubilization. Whether SR-BI, CD36897 and NPC1L1 function as authentic transporters that directly898 mediate lipid absorption,12−15,44 or whether they act to899 facilitate intracellular lipid trafficking or to modify signaling900 processes that mediate lipid absorption,45−47 or both, remains901 contentious, but nonetheless all merit examination here. Recent902 reports also suggest the involvement of SR-BI, CD36 and903 NPC1L1 in the absorption of fat-soluble nutrients (such as904 carotenoids,15,48 vitamin D49 and vitamin E50), providing905 further support for a role in drug absorption. In the current906 study, consistent with previous reports in vivo, inhibition of907 NPC1L1 reduced cholesterol absorption (Figure 2B),14 and908 inhibition of CD36 reduced (albeit nonsignificantly) the909 absorption of oleic acid (Figure 2C).12 In contrast, inhibition910 of SR-BI and CD36 had little impact on cholesterol absorption911 (Figure 2B). Inhibition of SR-BI and CD36 was expected to912 reduce cholesterol absorption based on previous in vitro913 studies.12,13,15 However, in vivo evidence of a role of SR-BI914 and CD36 in cholesterol absorption is less clear and, for915 example, no significant differences in intestinal cholesterol916 absorption were reported in SR-BI knockout vs wild-type917 mice.15,51 The lack of effect of SR-BI and CD36 inhibition on918 cholesterol absorption in the current study therefore runs919 contrary to previous in vitro studies but is in agreement with920 some previous in vivo data. Inhibition of endocytosis pathways921 via administration of monensin also failed to reduce cholesterol922 and cinnarizine absorption. Rather, the volume of distribution923 and clearance of cholesterol and cinnarizine appeared to be924 reduced, resulting in increases in plasma exposure. Since925 exogenously dosed 14C-cholesterol and cinnarizine are likely to926 be present within lipoproteins in the systemic circulation, the927 changes in systemic disposition of 14C-cholesterol and928 cinnarizine in the monensin-treated rats may reflect inhibition929 of receptor-mediated endocytosis of LDL.52

930 Inhibition of SR-BI, CD36 and NPC1L1 did not result in931 significant changes to the systemic exposure of cinnarizine932 following intraduodenal infusion of a lipid emulsion for-933 mulation in rats (Figure 2A). The data suggest that while lipid934 colloidal phases may be capable of direct interaction with lipid935 receptors (as suggested previously), the transfer of solubilized936 content into absorptive cells is likely selective and more937 applicable to solubilized lipids and nutrients rather than drugs.938 Together with the data describing cinnarizine absorption from939 micelles and vesicles, the results indicate that cinnarizine940 absorption from lipid colloidal phases is largely independent of941 the physical nature of the infused colloid (realizing that in these942 first experiments cinnarizine was present at concentration well943 below the solubilization limit and therefore under conditions944 where solubility/precipitation-mediated events were avoided),945 is not influenced significantly by common lipid uptake946 receptors and, instead, appears to be primarily dependent on947 the free drug concentration in equilibrium with the solubilized948 reservoir.949 Subsequent studies addressed the possibility that bile950 secretion may enhance drug absorption via the induction of951 supersaturation during the intestinal processing of dietary or952 formulation-related lipids. The addition of bile to cinnarizine-953 loaded micelles led to sustained drug supersaturation that954 ultimately led to increased intestinal drug absorption and

955systemic drug exposure. In contrast, although addition of bile to956cinnarizine-loaded vesicles also led to supersaturation, the957metastable supersaturated state was less stable than that958generated by bile addition to micelles, resulting in more rapid959precipitation of solubilized drug and therefore a lack of increase960in drug absorption and systemic drug exposure. The data961indicate that bile-mediated dilution of lipid colloidal phases may962represent an endogenous mechanism of supersaturation963generation during lipid processing in the small intestine; and964that the transient increase in thermodynamic activity may lead965to enhanced drug absorption. The observations also highlight966both the potential for supersaturation to enhance drug967absorption and the need to achieve an optimal balance between968drug supersaturation and drug precipitation.969Interaction of intestinal colloidal phases with bile leads to970dilution, an increase in pH and an increase in the971concentrations of bile components (bile salts, phosphatidylcho-972line and cholesterol) associated with the colloidal species. For a973solubilized system above the critical micellar concentration974(CMC), simple 1:1 v/v dilution is expected to reduce the975solubilized concentration, but to maintain total solubilization976capacity (i.e., for the drug concentration to drop by 50% but977the volume to double and therefore for solubilization capacity978to remain unchanged). However, the data in Table 4 show that979dilution of the micelles or vesicles with bile or buffer results in a980drop in total solubilization capacity to only 7−41% of initial.981Greater proportional decreases were apparent for dilution of982vesicles (in all cases when compared to micelles), and after983dilution with bile rather than buffer (Table 4). The loss of984solubilization capacity on dilution suggests the likelihood of a985phase transition to structures with reduced solubilization986capacity. Although a complete explanation for these phase987transitions is not apparent at this time, for the micellar systems,988it may be related to the ability of ionized caprylic acid to self-989associate and form fatty acid micelles at high lipid990concentration.53 Thus, dilution of medium-chain colloidal991phases may reduce the concentration of caprylic acid below992the CMC, leading to a loss of cinnarizine solubilization capacity993in the micellar phase. For vesicles, previous studies have994suggested that increasing bile concentrations facilitate a995vesicular to micellar transition.54,55 Since micelles are expected996to have lower solubilization capacities for lipophilic drugs than997vesicles (micelles are smaller and less lipid-rich),22,56 a998reduction in cinnarizine solubilization capacity might therefore999be anticipated when bile is added to model vesicles. In the case1000of the micelles, therefore, it seems likely that the addition of1001bile disrupted the structure of swollen, mixed micelles leading1002to lower colloidal lipid content and lower solubilization1003capacity. Unfortunately, attempts to quantify changes to1004particle size on bile addition were unsuccessful due to high1005polydispersity. However, the broad trends observed were1006consistent with the suggestions above and decreases in particle1007size were apparent for vesicles (consistent with initiation of a1008vesicle to micelle transition) and increases in particle size for1009the mixed micelles (consistent with micellar destabilization and1010transformation to less dispersed structures) (data not shown).1011An additional complexity in these dilution studies was the1012realization that the pH of bile (pH 8.02) is higher than that1013normally employed for simulated intestinal fluids,27,57 and1014therefore incubation of the colloidal phases (prepared at pH10156.30) with bile increases system pH. Cinnarizine is a weak base1016with a pKa of 7.47

58 and is therefore expected to be less ionized,1017less soluble and potentially more permeable at higher pHs. The


dx.doi.org/10.1021/mp3006566 | Mol. Pharmaceutics XXXX, XXX, XXX−XXXK

1018 impact of pH in the current studies was therefore examined by1019 dilution of micelles and vesicles with pH-corrected bile at pH1020 6.30. Comparison to the data obtained with bile at pH 8.021021 suggests that the higher pH of bile provides an additional driver1022 for drug precipitation/supersaturation since the solubility drop1023 was greater after incubation with bile at pH 8.02 vs bile at pH1024 6.301025 While dilution of micelles and vesicles with bile resulted in1026 decreases in drug solubilization capacity, cinnarizine did not1027 precipitate immediately. A period of drug supersaturation was1028 evident for both micellar and vesicular systems, although drug1029 precipitation from vesicles was much more rapid than from1030 micelles. This in turn translated into increases in drug1031 absorption (Figure 3A) and systemic exposure (Figure 4A)1032 for the micellar systems. The difference in the capacity of1033 micelles and vesicles to maintain drug supersaturation (Figure1034 6) may be explained by the difference in the degree of drug1035 supersaturation induced by dilution with bile. The degree of1036 supersaturation is described by the supersaturation ratio, which1037 is the ratio of the (supersaturated) concentration of drug in1038 solution relative to the equilibrium solubility of the drug in the1039 same system.16 Crystallization theory suggests that the1040 thermodynamic drivers of precipitation from supersaturated1041 solutions increase with increasing supersaturation ratios, as the1042 likelihood of nucleation and crystal growth increases with1043 increases in (metastable) drug concentration in solution.16,59

1044 Here, 1:1 v/v addition of bile resulted in the attainment of1045 cinnarizine supersaturation ratios of 6 and 12, in micelles and1046 vesicles, respectively. Therefore, the faster rate of drug1047 precipitation from vesicles (when compared to micelles) may1048 be explained by the higher supersaturation ratio induced by bile1049 addition. It is also possible that micellar structures are more1050 effective in stabilizing supersaturation when compared to1051 vesicular structures, although this has not been examined1052 explicitly here. Notably, cinnarizine was found to precipitate in1053 the crystalline form in these experiments (Figure S3 in the1054 Supporting Information), precluding the possibility that the1055 enhanced cinnarizine absorption from micelles (when bile was

1056coperfused) observed in Figure 3A was due to accelerated1057cinnarizine dissolution from precipitated amorphous forms.1058The role of bile in enhancing drug absorption from LBF has1059been reported previously.60−62 In the majority of cases, bile-1060mediated bioavailability enhancement has been suggested to1061stem from the ability of bile to expand the solubilization1062reservoir for PWSD in the GI tract. This is typically assumed to1063occur via PWSD solubilization in simple bile micelles, or via the1064ability of bile to solubilize lipid digestion products and to1065generate more complex lipid colloidal phases with enhanced1066solubilization capacities. It seems likely that the ability to1067solubilize lipid digestion products and to promote colloid1068formation remains an integral part of the role of bile in1069supporting drug (and lipid) absorption. However, the data1070described here suggest that continued dilution of lipid colloidal1071phases with bile in the small intestinal lumen may also lead to1072physical changes that promote drug supersaturation, and1073ultimately promote drug absorption. In doing so, super-1074saturation induction may be a means by which the decrease1075in thermodynamic activity inherent in solubilization is reversed,1076such that the free concentration of drug available for absorption1077is maximized. Thus, a dual role of bile in facilitating drug1078 f8absorption from LBF may be conceived (see Figure 8). First,1079bile-mediated solubilization of lipid digestion products at the1080interface of a digesting lipid droplet results in the generation of1081lipid colloidal phases such as vesicles and micelles that promote1082drug solubilization during lipid digestion. Second, continued1083bile-mediated dilution of existing lipid colloidal phases1084promotes drug supersaturation, and enhances drug absorption1085by significantly increasing drug thermodynamic activity in1086colloidal phases. The combination of these two highly kinetic1087events likely contributes to the effective drug absorption often1088observed with lipid coadministration, as it affords a means to1089simultaneously increase solubilization capacity and promote1090thermodynamic activity of coadministered PWSD in the small1091intestine.1092Supersaturation induction via interaction with bile also1093provides a means of overcoming the recently described1094solubility−permeability interplay observed in studies where

Figure 8. The dual role of bile during lipid digestion and dispersion. (i) Bile-mediated solubilization of lipid digestion products at the interface of adigesting oil droplet results in the generation of lipid colloidal phases such as vesicles and micelles that maintain drug solubilization in the smallintestine. (ii) The continuing interaction of secreted bile with existing lipid colloidal phases in the lumen results in progressively less lipid-rich phaseswith lowered solubilization capacity. Thus, ongoing bile-mediated dilution of lipid colloidal phases promotes drug supersaturation and enhances drugabsorption by increasing drug thermodynamic activity in colloidal phases. The combination of the two highly kinetic, bile-mediated events affords ameans to simultaneously increase solubilization capacity and promote thermodynamic activity of coadministered drug in the small intestine, and maycontribute to the increase in drug absorption often observed with lipid coadministration. D represents the free concentration of drug available forabsorption. Dss is used to signify the increase in free concentration resulting from bile-mediated supersaturation that drives increases in drugabsorption.


dx.doi.org/10.1021/mp3006566 | Mol. Pharmaceutics XXXX, XXX, XXX−XXXL

1095 PWSD are coadministered with cosolvents, cyclodextrins or1096 surfactant systems.5,6,63 In these studies the authors describe1097 the reduction in thermodynamic activity common to most1098 solubilization technologies and show that this offsets the1099 potential increases in membrane flux that might be expected by1100 an increase in solubilized drug concentration.5,6 More recent1101 studies by the same authors have shown that this solubility−1102 permeability interplay can be addressed via the use of1103 amorphous solid dispersion formulations that stimulate super-1104 saturation, but do not promote solubilization.64,65 Here we1105 report that essentially similar outcomes are also possible with1106 solubilizing formulations, when the solubilizing formulations1107 contain lipids and when the kinetic changes that occur in the GI1108 lumen in the presence of bile secretion promote super-1109 saturation. The current data therefore suggest that endogenous1110 lipid processing pathways provide an exquisitely sensitive and1111 triggerable supersaturation mechanism that allows drug to1112 remain in a solubilized state during initial lipid digestion and at1113 high lipid:bile concentration ratios, but that ongoing bile1114 secretion subsequently provides a boost to thermodynamic1115 activity and in doing so supports enhanced absorption. This is1116 in contrast to other common solubilization strategies that may1117 not interact with the dynamic GI environment, or for which1118 interaction with bile typically reduces thermodynamic activity1119 by increasing solubilization capacity. Conversely, the ability of1120 LBF to promote drug solubilization until drug supersaturation1121 is triggered in the small intestine may confer an advantage over1122 formulation approaches that only utilize supersaturating1123 strategies, as the risk of drug precipitation in the GI tract1124 may be reduced by solubilization within lipid colloidal phases.

1125 ■ CONCLUSION

1126 Improved understanding of the mechanism of drug absorption1127 from lipid colloidal phases such as micelles and vesicles is1128 required to provide a platform for more rational design of lipid-1129 based formulations. Using medium-chain lipids, we have1130 demonstrated that the absorption of cinnarizine (a lipophilic,1131 poorly water-soluble drug) may be enhanced when drug1132 supersaturation is generated during bile-mediated dilution of1133 lipid colloidal phases. Previous studies suggest that a similar1134 induction of supersaturation may occur as a result of initiation1135 of digestion of some microemulsion-based LBFs.21 These1136 observations indicate that supersaturation and its associated1137 benefits in enhancing drug absorption may occur intrinsically1138 during LBF incorporation into endogenous lipid processing1139 pathways in the small intestine. Future work will be directed1140 toward assessing the impact of bile dilution on supersaturation1141 tendency, and thus absorption, for an extended range of PWSD1142 and in a series of different micellar and vesicular colloidal1143 systems.

1144 ■ ASSOCIATED CONTENT

1145 *S Supporting Information1146 Table S1 showing pharmacokinetic parameters for cinnarizine1147 after intravenous administration in rats. Figure S1 showing1148 systemic plasma concentration−time profiles of cinnarizine and1149

14C-labeled cholesterol following intravenous infusion in rats.1150 Figure S2 showing perfusate disappearance profiles of1151 cinnarizine when micelles were perfused through an isolated1152 rat jejunal segment, with and without 1:1 v/v coperfusion with1153 rat bile pH 6.30 or buffer pH 6.30. Figure S3 showing polarized1154 light microscopy images of the crystalline cinnarizine

1155precipitate following precipitation kinetics experiments. This1156material is available free of charge via the Internet at http://1157pubs.acs.org.

1158■ AUTHOR INFORMATION1159Corresponding Author1160*N.L.T.: phone, +61 3 9903 9138; fax, +61 3 9903 9583; e-1161mail, [email protected]. C.J.H.P.: phone, +61 311629903 9649; fax, +61 3 9903 9583; e-mail, [email protected] authors declare no competing financial interest.

1166■ ACKNOWLEDGMENTS1167Funding support from the National Health and Medical1168Research Council (NHMRC) is gratefully acknowledged.

1169■ ABBREVIATIONS USED1170LBF, lipid-based formulations; PWSD, poorly water-soluble1171drugs; GI, gastrointestinal; LPC, L-α-lysophosphatidylcholine;1172CD36, cluster of differentiation 36; SR-BI, scavenger receptor1173class B type 1; NPC1L1, Niemann-Pick C1 like 1; BLT-1, block1174lipid transport-1; SSO, sulfo-N-succinimidyl oleate; iv, intra-1175venous; HPLC, high performance liquid chromatography;1176NMR, nuclear magnetic resonance; DMSO, dimethyl sulfoxide;1177TBME, tert-butyl methyl ether; SEIF, simulated endogenous1178intestinal fluid; SEM, standard error of the mean; CMC, critical1179micellar concentration; UWL, unstirred water layer; HDL,1180high-density lipoprotein; LDL, low-density lipoprotein; CIN,1181cinnarizine; Ch, cholesterol; OA, oleic acid; SS ratio, super-1182saturation ratio

1183■ REFERENCES(1) 1184Williams, H. D.; Trevaskis, N. L.; Charman, S. A.; Shanker, R. M.;

1185Charman, W. N.; Pouton, C. W.; Porter, C. J. H. Strategies to address1186low drug solubility in discovery and development. Pharmacol. Rev.11872013, 65 (1), 315−499.

(2) 1188Poelma, F. G. J.; Breas, R.; Tukker, J. J.; Crommelin, D. J. A.1189Intestinal absorption of drugs. The influence of mixed micelles on the1190disappearance kinetics of drugs from the small intestine of the rat. J.1191Pharm. Pharmacol. 1991, 43 (5), 317−324.

(3) 1192Katneni, K.; Charman, S. A.; Porter, C. J. H. Permeability1193assessment of poorly water-soluble compounds under solubilizing1194conditions: The reciprocal permeability approach. J. Pharm. Sci. 2006,119595 (10), 2170−2185.

(4) 1196Amidon, G. E.; Higuchi, W. I.; Ho, N. F. H. Theoretical and1197experimental studies of transport of micelle-solubilized solutes. J.1198Pharm. Sci. 1982, 71 (1), 77−84.

(5) 1199Miller, J. M.; Beig, A.; Krieg, B. J.; Carr, R. A.; Borchardt, T. B.;1200Amidon, G. E.; Amidon, G. L.; Dahan, A. The solubility−permeability1201interplay: Mechanistic modeling and predictive application of the1202impact of micellar solubilization on intestinal permeation. Mol.1203Pharmaceutics 2011, 8 (5), 1848−1856.

(6) 1204Dahan, A.; Miller, J. M.; Hoffman, A.; Amidon, G. E.; Amidon, G.1205L. The solubility−permeability interplay in using cyclodextrins as1206pharmaceutical solubilizers: Mechanistic modeling and application to1207progesterone. J. Pharm. Sci. 2010, 99 (6), 2739−2749.

(7) 1208Wootan, M. G.; Bernlohr, D. A.; Storch, J. Mechanism of1209fluorescent fatty acid transfer from adipocyte fatty acid binding protein1210to membranes. Biochemistry 1993, 32 (33), 8622−8627.

(8) 1211Thumser, A.; Tsai, J.; Storch, J. Collision-mediated transfer of1212long-chain fatty acids by neural tissue fatty acid-binding proteins1213(FABP). J. Mol. Neurosci. 2001, 16 (2), 143−150.

(9) 1214Yano, K.; Masaoka, Y.; Kataoka, M.; Sakuma, S.; Yamashita, S.1215Mechanisms of membrane transport of poorly soluble drugs: Role of


dx.doi.org/10.1021/mp3006566 | Mol. Pharmaceutics XXXX, XXX, XXX−XXXM

http://pubs.acs.org

http://pubs.acs.org

mailto:[email protected]



1216 micelles in oral absorption processes. J. Pharm. Sci. 2010, 99 (3),1217 1336−1345.

(10)1218 Gao, P.; Morozowich, W. Development of supersaturatable self-1219 emulsifying drug delivery system formulations for improving the oral1220 absorption of poorly soluble drugs. Expert Opin. Drug Delivery 2006, 31221 (1), 97−110.

(11)1222 Gu, X.; Trigatti, B.; Xu, S.; Acton, S.; Babitt, J.; Krieger, M. The1223 efficient cellular uptake of high density lipoprotein lipids via Scavenger1224 Receptor Class B Type I requires not only receptor-mediated surface1225 binding but also receptor-specific lipid transfer mediated by its1226 extracellular domain. J. Biol. Chem. 1998, 273 (41), 26338−26348.

(12)1227 Nassir, F.; Wilson, B.; Han, X.; Gross, R. W.; Abumrad, N. A.1228 CD36 is important for fatty acid and cholesterol uptake by the1229 proximal but not distal intestine. J. Biol. Chem. 2007, 282 (27),1230 19493−19501.

(13)1231 Hauser, H.; Dyer, J. H.; Nandy, A.; Vega, M. A.; Werder, M.;1232 Bieliauskaite, E.; Weber, F. E.; Compassi, S.; Gemperli, A.; Boffelli, D.;1233 Wehrli, E.; Schulthess, G.; Phillips, M. C. Identification of a receptor1234 mediating absorption of dietary cholesterol in the intestine.1235 Biochemistry 1998, 37 (51), 17843−17850.

(14)1236 Altmann, S. W.; Davis, H. R.; Zhu, L.; Yao, X.; Hoos, L. M.;1237 Tetzloff, G.; Iyer, S. P. N.; Maguire, M.; Golovko, A.; Zeng, M.; Wang,1238 L.; Murgolo, N.; Graziano, M. P. Niemann-Pick C1 Like 1 protein is1239 critical for intestinal cholesterol absorption. Science 2004, 303 (5661),1240 1201−1204.

(15)1241 Van Bennekum, A.; Werder, M.; Thuahnai, S. T.; Han, C.-H.;1242 Duong, P.; Williams, D. L.; Wettstein, P.; Schulthess, G.; Phillips, M.1243 C.; Hauser, H. Class B Scavenger Receptor-mediated intestinal1244 absorption of dietary β-carotene and cholesterol. Biochemistry 2005,1245 44 (11), 4517−4525.

(16)1246 Brouwers, J.; Brewster, M. E.; Augustijns, P. Supersaturating1247 drug delivery systems: The answer to solubility-limited oral1248 bioavailability? J. Pharm. Sci. 2009, 98 (8), 2549−2572.

(17)1249 Porter, C. J. H.; Kaukonen, A. M.; Boyd, B. J.; Edwards, G. A.;1250 Charman, W. N. Susceptibility to lipase-mediated digestion reduces1251 the oral bioavailability of danazol after administration as a medium-1252 chain lipid-based microemulsion formulation. Pharm. Res. 2004, 211253 (8), 1405−1412.

(18)1254 Kaukonen, A. M.; Boyd, B.; Porter, C.; Charman, W. Drug1255 solubilization behavior during in vitro digestion of simple triglyceride1256 lipid solution formulations. Pharm. Res. 2004, 21 (2), 245−253.

(19)1257 Anby, M. U.; Williams, H. D.; McIntosh, M.; Benameur, H.;1258 Edwards, G. A.; Pouton, C. W.; Porter, C. J. H. Lipid digestion as a1259 trigger for supersaturation: Evaluation of the impact of supersaturation1260 stabilization on the in vitro and in vivo performance of self-emulsifying1261 drug delivery systems. Mol. Pharmaceutics 2012, 9 (7), 2063−2079.

(20)1262 Cuine, J.; Charman, W.; Pouton, C.; Edwards, G.; Porter, C.1263 Increasing the proportional content of surfactant (Cremophor EL)1264 relative to lipid in self-emulsifying lipid-based formulations of danazol1265 reduces oral bioavailability in beagle dogs. Pharm. Res. 2007, 24 (4),1266 748−757.

(21)1267 Pouton, C. W.; Porter, C. J. H. Formulation of lipid-based1268 delivery systems for oral administration: Materials, methods and1269 strategies. Adv. Drug Delivery Rev. 2008, 60 (6), 625−637.

(22)1270 Kossena, G. A.; Boyd, B. J.; Porter, C. J. H.; Charman, W. N.1271 Separation and characterization of the colloidal phases produced on1272 digestion of common formulation lipids and assessment of their1273 impact on the apparent solubility of selected poorly water-soluble1274 drugs. J. Pharm. Sci. 2003, 92 (3), 634−648.

(23)1275 Harmon, C. M.; Luce, P.; Beth, A. H.; Abumrad, N. A. Labeling1276 of adipocyte membranes by sulfo-N-succinimidyl derivatives of long-1277 chain fatty acids: Inhibition of fatty acid transport. J. Membr. Biol.1278 1991, 121 (3), 261−268.

(24)1279 Nieland, T. J. F.; Penman, M.; Dori, L.; Krieger, M.;1280 Kirchhausen, T. Discovery of chemical inhibitors of the selective1281 transfer of lipids mediated by the HDL receptor SR-BI. Proc. Natl.1282 Acad. Sci. U.S.A. 2002, 99 (24), 15422−15427.

(25)1283 Coort, S. L. M.; Willems, J.; Coumans, W. A.; van der Vusse, G.1284 J.; Bonen, A.; Glatz, J. F. C.; Luiken, J. J. F. P. Sulfo-N-succinimidyl

1285esters of long chain fatty acids specifically inhibit fatty acid translocase1286(FAT/CD36)-mediated cellular fatty acid uptake. Mol. Cell. Biochem.12872002, 239 (1), 213−219.

(26) 1288Garcia-Calvo, M.; Lisnock, J.; Bull, H. G.; Hawes, B. E.; Burnett,1289D. A.; Braun, M. P.; Crona, J. H.; Davis, H. R.; Dean, D. C.; Detmers,1290P. A.; Graziano, M. P.; Hughes, M.; MacIntyre, D. E.; Ogawa, A.;1291O’Neill, K. A.; Iyer, S. P. N.; Shevell, D. E.; Smith, M. M.; Tang, Y. S.;1292Makarewicz, A. M.; Ujjainwalla, F.; Altmann, S. W.; Chapman, K. T.;1293Thornberry, N. A. The target of ezetimibe is Niemann-Pick C1-Like 11294(NPC1L1). Proc. Natl. Acad. Sci. U.S.A. 2005, 102 (23), 8132−8137.

(27) 1295Kossena, G. A.; Charman, W. N.; Boyd, B. J.; Dunstan, D. E.;1296Porter, C. J. H. Probing drug solubilization patterns in the1297gastrointestinal tract after administration of lipid-based delivery1298systems: A phase diagram approach. J. Pharm. Sci. 2004, 93 (2),1299332−348.

(28) 1300Staggers, J. E.; Hernell, O.; Stafford, R. J.; Carey, M. C. Physical-1301chemical behavior of dietary and biliary lipids during intestinal1302digestion and absorption. 1. Phase behavior and aggregation states of1303model lipid systems patterned after aqueous duodenal contents of1304healthy adult human beings. Biochemistry 1990, 29 (8), 2028−2040.

(29) 1305Hay, D. W.; Cahalane, M. J.; Timofeyeva, N.; Carey, M. C.1306Molecular species of lecithins in human gallbladder bile. J. Lipid Res.13071993, 34 (5), 759−68.

(30) 1308Duane, W. C.; Ginsberg, R. L.; Bennion, L. J. Effects of fasting1309on bile acid metabolism and biliary lipid composition in man. J. Lipid1310Res. 1976, 17 (3), 211−9.

(31) 1311Boyd, B. J.; Porter, C. J. H.; Charman, W. N. Using the polymer1312partitioning method to probe the thermodynamic activity of poorly1313water-soluble drugs solubilized in model lipid digestion products. J.1314Pharm. Sci. 2003, 92 (6), 1262−1271.

(32) 1315Trevaskis, N. L.; Porter, C. J. H.; Charman, W. N. Bile increases1316intestinal lymphatic drug transport in the fasted rat. Pharm. Res. 2005,131722 (11), 1863−1870.

(33) 1318Caliph, S. M.; Charman, W. N.; Porter, C. J. H. Effect of short-,1319medium-, and long-chain fatty acid-based vehicles on the absolute oral1320bioavailability and intestinal lymphatic transport of halofantrine and1321assessment of mass balance in lymph-cannulated and non-cannulated1322rats. J. Pharm. Sci. 2000, 89 (8), 1073−1084.

(34) 1323Waynforth, H. B.; Flecknell, P. A. Experimental and surgical1324technique in rat; Academic Press: 1992.

(35) 1325Cummins, C. L.; Salphati, L.; Reid, M. J.; Benet, L. Z. In vivo1326modulation of intestinal CYP3A metabolism by P-glycoprotein:1327Studies using the rat single-pass intestinal perfusion model. J.1328Pharmacol. Exp. Ther. 2003, 305 (1), 306−314.

(36) 1329Van Heek, M.; Farley, C.; Compton, D. S.; Hoos, L.; Alton, K.1330B.; Sybertz, E. J.; Davis, H. R. Comparison of the activity and1331disposition of the novel cholesterol absorption inhibitor, SCH58235,1332and its glucuronide, SCH60663. Br. J. Pharmacol. 2000, 129 (8),13331748−1754.

(37) 1334Kyle, M. E.; Nakae, D.; Sakaida, I.; Miccadei, S.; Farber, J. L.1335Endocytosis of superoxide dismutase is required in order for the1336enzyme to protect hepatocytes from the cytotoxicity of hydrogen1337peroxide. J. Biol. Chem. 1988, 263 (8), 3784−3789.

(38) 1338Winne, D. Rat jejunum perfused in situ: Effect of perfusion rate1339and intraluminal radius on absorption rate and effective unstirred layer1340thickness. Naunyn-Schmiedeberg’s Arch. Pharmacol. 1979, 307 (3),1341265−274.

(39) 1342Kossena, G. A.; Charman, W. N.; Boyd, B. J.; Porter, C. J. H. A1343novel cubic phase of medium chain lipid origin for the delivery of1344poorly water soluble drugs. J. Controlled Release 2004, 99 (2), 217−1345229.

(40) 1346Johnson, B. M.; Chen, W.; Borchardt, R. T.; Charman, W. N.;1347Porter, C. J. H. A kinetic evaluation of the absorption, efflux, and1348metabolism of verapamil in the autoperfused rat jejunum. J. Pharmacol.1349Exp. Ther. 2003, 305 (1), 151−158.

(41) 1350Wagner, J. G.; Sedman, A. J. Quantitaton of rate of1351gastrointestinal and buccal absorption of acidic and basic drugs1352based on extraction theory. J. Pharmacokinet. Biopharm. 1973, 1 (1),135323−50.


dx.doi.org/10.1021/mp3006566 | Mol. Pharmaceutics XXXX, XXX, XXX−XXXN

(42)1354 Dahan, A.; Miller, J. M.; Hilfinger, J. M.; Yamashita, S.; Yu, L.1355 X.; Lennernas, H.; Amidon, G. L. High-permeability criterion for BCS1356 classification: Segmental/pH dependent permeability considerations.1357 Mol. Pharmaceutics 2010, 7 (5), 1827−1834.

(43)1358 Sugano, K. Estimation of effective intestinal membrane1359 permeability considering bile micelle solubilisation. Int. J. Pharm.1360 2009, 368 (1−2), 116−122.

(44)1361 Haikal, Z.; Play, B.; Landrier, J.-F.; Giraud, A.; Ghiringhelli, O.;1362 Lairon, D.; Jourdheuil-Rahmani, D. NPC1L1 and SR-BI are involved1363 in intestinal cholesterol absorption from small-size lipid donors. Lipids1364 2008, 43 (5), 401−408.

(45)1365 Tran, T. T. T.; Poirier, H.; Clement, L.; Nassir, F.; Pelsers, M.1366 M. A. L.; Petit, V.; Degrace, P.; Monnot, M.-C.; Glatz, J. F. C.;1367 Abumrad, N. A.; Besnard, P.; Niot, I. Luminal lipid regulates CD361368 levels and downstream signaling to stimulate chylomicron synthesis. J.1369 Biol. Chem. 2011, 286 (28), 25201−25210.

(46)1370 Nauli, A. M.; Nassir, F.; Zheng, S.; Yang, Q.; Lo, C. M.;1371 VonLehmden, S. B.; Lee, D.; Jandacek, R. J.; Abumrad, N. A.; Tso, P.1372 CD36 Is important for chylomicron formation and secretion and may1373 mediate cholesterol uptake in the proximal intestine. Gastroenterology1374 2006, 131 (4), 1197−1207.

(47)1375 Yu, L.; Bharadwaj, S.; Brown, J. M.; Ma, Y.; Du, W.; Davis, M.1376 A.; Michaely, P.; Liu, P.; Willingham, M. C.; Rudel, L. L. Cholesterol-1377 regulated Translocation of NPC1L1 to the Cell Surface Facilitates1378 Free Cholesterol Uptake. J. Biol. Chem. 2006, 281 (10), 6616−6624.

(48)1379 Moussa, M.; Landrier, J.-F.; Reboul, E.; Ghiringhelli, O.;1380 Comera, C.; Collet, X.; Frohlich, K.; Bohm, V.; Borel, P. Lycopene1381 absorption in human intestinal cells and in mice involves Scavenger1382 Receptor Class B Type I but not Niemann-Pick C1-Like 1. J. Nutr.1383 2008, 138 (8), 1432−1436.

(49)1384 Reboul, E.; Goncalves, A.; Comera, C.; Bott, R.; Nowicki, M.;1385 Landrier, J.-F.; Jourdheuil-Rahmani, D.; Dufour, C.; Collet, X.; Borel,1386 P. Vitamin D intestinal absorption is not a simple passive diffusion:1387 Evidences for involvement of cholesterol transporters. Mol. Nutr. Food1388 Res. 2011, 55 (5), 691−702.

(50)1389 Reboul, E.; Klein, A.; Bietrix, F.; Gleize, B.; Malezet-1390 Desmoulins, C.; Schneider, M.; Margotat, A.; Lagrost, L.; Collet, X.;1391 Borel, P. Scavenger Receptor Class B Type I (SR-BI) Is involved in1392 vitamin E transport across the enterocyte. J. Biol. Chem. 2006, 281 (8),1393 4739−4745.

(51)1394 Altmann, S. W.; Davis, H. R., Jr; Yao, X.; Laverty, M.; Compton,1395 D. S.; Zhu, L.-j.; Crona, J. H.; Caplen, M. A.; Hoos, L. M.; Tetzloff, G.;1396 Priestley, T.; Burnett, D. A.; Strader, C. D.; Graziano, M. P. The1397 identification of intestinal scavenger receptor class B, type I (SR-BI) by1398 expression cloning and its role in cholesterol absorption. Biochim.1399 Biophys. Acta, Mol. Cell Biol. Lipids 2002, 1580 (1), 77−93.

(52)1400 Goldstein, J. L.; Brown, M. S.; Anderson, R. G. W.; Russell, D.1401 W.; Schneider, W. J. Receptor-mediated endocytosis: Concepts1402 emerging from the LDL receptor system. Annu. Rev. Cell Biol. 1985,1403 1 (1), 1−39.

(53)1404 Mukerjee, P.; Mysels, K. J. Critical micelle concentrations of1405 aqueous surfactant systems; National Bureau of Standards: Washington,1406 DC, 1971; pp 88−170.

(54)1407 Schubert, R.; Schmidt, K. H. Structural changes in vesicle1408 membranes and mixed micelles of various lipid compositions after1409 binding of different bile salts. Biochemistry 1988, 27 (24), 8787−8794.

(55)1410 Hofmann, A. F. Bile acids: The good, the bad, and the ugly.1411 Physiology 1999, 14 (1), 24−29.

(56)1412 Horter, D.; Dressman, J. B. Influence of physicochemical1413 properties on dissolution of drugs in the gastrointestinal tract. Adv.1414 Drug Delivery Rev. 1997, 25 (1), 3−14.

(57)1415 Vertzoni, M.; Fotaki, N.; Nicolaides, E.; Reppas, C.; Kostewicz,1416 E. S.; Stippler, E.; Leuner, C.; Dressman, J. Dissolution media1417 simulating the intralumenal composition of the small intestine:1418 physiological issues and practical aspects. J. Pharm. Pharmacol. 2004,1419 56 (4), 453−462.

(58)1420 Jarvinen, T.; Jarvinen, K.; Schwarting, N.; Stella, V. J. β-1421 Cyclodextrin derivatives, SBE4-β-CD and HP-β-CD, increase the oral

1422bioavailability of cinnarizine in beagle dogs. J. Pharm. Sci. 1995, 84 (3),1423295−299.

(59) 1424Augustijns, P.; Brewster, M. E. Supersaturating drug delivery1425systems: Fast is not necessarily good enough. J. Pharm. Sci. 2012, 1011426(1), 7−9.

(60) 1427Tønsberg, H.; Holm, R.; Mu, H.; Boll, J. B.; Jacobsen, J.;1428Mullertz, A. Effect of bile on the oral absorption of halofantrine in1429polyethylene glycol 400 and polysorbate 80 formulations dosed to bile1430duct cannulated rats. J. Pharm. Pharmacol. 2011, 63 (6), 817−824.

(61) 1431Ericzon, B. G.; Todo, S.; Lynch, S.; Kam, I.; Ptachcinski, R. J.;1432Burckart, G. J.; Van Thiel, D. H.; Starzl, T. E.; Venkataramanan, R.1433Role of bile and bile salts on cyclosporine absorption in dogs.1434Transplant Proc. 1987, 19 (1 Part 2), 1248−1249.

(62) 1435Behrens, D.; Fricker, R.; Bodoky, A.; Drewe, J.; Harder, F.;1436Heberer, M. Comparison of cyclosporin A absorption from LCT and1437MCT solutions following intrajejunal administration in conscious1438dogs. J. Pharm. Sci. 1996, 85 (6), 666−668.

(63) 1439Miller, J. M.; Beig, A.; Carr, R. A.; Webster, G. K.; Dahan, A.1440The solubility−permeability interplay when using cosolvents for1441solubilization: Revising the way we use solubility-enabling formula-1442tions. Mol. Pharmaceutics 2012, 9 (3), 581−590.

(64) 1443Dahan, A.; Beig, A.; Ioffe-Dahan, V.; Agbaria, R.; Miller, J. M.1444The twofold advantage of the amorphous form as an oral drug delivery1445practice for lipophilic compounds: Increased apparent solubility and1446drug flux through the intestinal membrane. AAPS J. 2012,1447DOI: 10.1208/s12248-012-9445-3.

(65) 1448Miller, J. M.; Beig, A.; Carr, R. A.; Spence, J. K.; Dahan, A. A1449win−win solution in oral delivery of lipophilic drugs: Supersaturation1450via amorphous solid dispersions increases apparent solubility without1451sacrifice of intestinal membrane permeability.Mol. Pharmaceutics 2012,14529 (7), 2009−2016.


dx.doi.org/10.1021/mp3006566 | Mol. Pharmaceutics XXXX, XXX, XXX−XXXO

Documents

Supersaturation as a driving force for drug absorption ...€¦ · mucin interaction on drug absorption is likely difficult to predict and may be expected to vary for different drug