Jones, 2001. Liofilización. Sacarosa y Tween

7/29/2019 Jones, 2001. Liofilizacin. Sacarosa y Tween

1/12

The Effects of Tween 20 and Sucrose on the Stability ofAnti-L-Selectin during Lyophilization and Reconstitution

LATOYA S. JONES,1 THEODORE W. RANDOLPH,2 ULRICH KOHNERT,3 APOLLON PAPADIMITRIOU,3 G. WINTER,3

MARIE-LUISE HAGMANN,3 MARK C. MANNING,1 JOHN F. CARPENTER1

1School of Pharmacy, University of Colorado Health Sciences Center, Denver, Colorado 80262

2Department of Chemical Engineering, University of Colorado at Boulder, Boulder, Colorado 80309

3Boehringer Mannheim, Penzberg, Germany

Received 19 December 2000; revised 24 May 2001; accepted 25 May 2001

ABSTRACT: We have chosen an anti-L-selectin antibody as a model protein toinvestigate the effects of sucrose and/or Tween 20 on protein stability during

lyophilization and reconstitution. Native anti-L-selectin secondary structure is

substantially retained during lyophilization in the presence of sucrose (1 or 0.125%).

However, aggregation of the protein during reconstitution of lyophilized protein

powders prepared without sucrose is not reduced by the presence of sucrose in the

reconstitution medium. Aggregate formation upon reconstitution is completely

inhibited by freeze drying the protein with sucrose and reconstituting with a 0.1%

Tween 20 solution. Tween 20 (0.1%) also partially inhibits loss of native anti-L-selectin

secondary structure during lyophilization. However, upon reconstitution the formula-

tions lyophilized with Tween 20 contain the highest levels of aggregates. The presence of

Tween in only the reconstitution solution appears to inhibit the transition from dimers

to higher order oligomers. Potential mechanism(s) for the Tween 20 effects were

investigated. However, no evidence of thermodynamic stabilization of anti-L-selectin

conformation (e.g., by Tween 20 binding) could be detected. 2001 Wiley-Liss, Inc. and the

American Pharmaceutical Association J Pharm Sci 90:1466 1477, 2001

Keywords: lyophilization; reconstitution; antibody formulation; infrared spectro-

scopy; protein aggregation

INTRODUCTION

To optimize protein stability during storage andshipping, lyophilized formulations are often

employed. However, a lyophilized product has its

drawbacks: the possibility for protein denatura-

tion exists during both the freeze drying16 andreconstitution7,8 processes. Excipients are oftencritical for maintaining native protein con-

formation during freezing and drying, and mini-mizing levels of protein aggregation in thereconstituted product.25,914 The purpose of thepresent study is to investigate the effects ofTween 20 and sucrose on the stability of an anti-L-selectin antibody during lyophilization andreconstitution.

Lyophilization involves both freezing and

dehydration steps, each capable of promotingprotein denaturation. Sucrose inhibits protein un-folding during freezing and drying.1,5,6,9,12,1518

However, studies of structural preservation of

1466 JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 90, NO. 10, OCTOBER 2001

Ulrich Kohnert's present address is Scil BiomedicalsGmbH, Fraunhofer Str. 15, D-82152, Martinsried, Germany.

Apollon Papadimitriou's present address is Roche Diagnos-tics GmbH, Pharma Research, Penzberg, Germany.

G. Winter's present address is Lehrstuhl fu r Pharmazeu-tische Technologie und Biopharmazie, Ludwig MaximiliansUniversita t Mu nchen, Germany.

Correspondence to: J. F. Carpenter (Telephone: 303-315-6075; Fax: 303-315-6281; E-mail: [email protected])

Journal of Pharmaceutical Sciences, Vol. 90, 1466 1477 (2001) 2001 Wiley-Liss, Inc. and the American Pharmaceutical Association


2/12

antibodies during lyophilization are limited.19

During freezing, sucrose inhibits proteindenaturation by increasing the free energy ofprotein unfolding.2022 During dehydration,

sucrose stabilizes a protein by replacing the

hydrogen bonds between the protein and watermolecules that are lost during drying.1,23 Thecurrent study employs infrared spectroscopy tocharacterize the secondary structure of anti-L-selectin in the initial aqueous and lyophilizedstates. The experiments test the hypothesis thatsucrose inhibits lyophilization-induced unfolding.

The stabilization of proteins by surfactants isoften attributed to limiting the extent of proteinadsorption at various potentially denaturing

interfaces (e.g., ice/liquid, air/liquid, vial/liquid).14,2427 In addition, nonionic surfactantscan stabilize some proteins, such as human

growth hormone24,29,30 and tissue factor,28 bybinding to solvent exposed hydrophobic regionsof the native state protein. Surfactants have beenshown to inhibit protein denaturation duringfreeze thawing.3,15,18,26 However, unlike excipi-ents such as sucrose, surfactants have a limitedcapacity to inhibit lyophilization-induced unfold-ing via the water replacement mechanism.

Accordingly, Kreilgaard et al. found that Tween20 (0.002%) did not inhibit lyophilization-inducedunfolding of Factor XIII.5 In contrast, Chang et al.

found that Tween 80 (0.1%) partially inhibited

lyophilization-induced unfolding/aggregation ofinterleukin-1 receptor antagonist, but this studywas conducted under grossly destabilizing condi-tions that led to 50% aggregation after reconstitu-tion.3 There appears to be no published study thatdirectly determines the effect of nonionic surfac-tants on the lyophilization-induced unfolding ofantibodies. Thus, a second goal of the currentstudy is to test the hypothesis that Tween 20should not inhibit the lyophilization-induced

unfolding of anti-L-selectin.Lyophilized drugs must be reconstituted prior

to administration. Aggregates have been pro-posed to form during reconstitution from afraction of nonnative protein molecules that arepresent in the dried solid.17 Minimizing proteinunfolding during lyophilization by inclusion ofexcipients such as sucrose can reduce the level ofaggregates present after reconstitution.16,31,32

Inclusion of surfactants (e.g., Tween 20) in thereconstitution solution might alter kinetics tofavor refolding over aggregation. Typically, wateralone is used to reconstitute the lyophilizedproduct. Prestrelski et al. found that some

excipients (e.g., 0.1 and 0.5% Tween 20, 0.05%EDTA) in the reconstitution solutions reducedaggregation of keratinocyte growth factor andinterleukin-2, whereas others (e.g., N-octylgluco-side, Pluronic) had no signicant effect on

keratinocyte growth factor but promoted aggrega-tion of interleukin-2.7,8 They also found thatadditives in the reconstitution medium decreasedthe amount of soluble ribonuclease A aggregates.8

These studies focused only on proteins thathad been stored at 458C for 2 or more weeks,prior to reconstitution.7,8 There are two publishedreports on the effects of surfactant solution onreconstitution of proteins immediately after lyo-philization. Chang et al. report that the presence

of 0.1% Tween 80 in the reconstitution mediumdecreased aggregation of interleukin 1-receptorantagonist.3 A reconstitution solution of 0.02%

Tween 80 was equally as effective at inhibitingaggregation of bovine IgG as the inclusion ofeither 0.02 or 0.1% Tween 80 in the initiallyophilized formulation.14 However, separateeffects of the surfactant on inhibiting lyophiliza-tion-induced unfolding and fostering refoldingduring reconstitution were not tested. Wehypothesize that, although Tween 20 might notcompletely inhibit structural perturbations of theprotein during lyophilization, it will fosterreduced levels of aggregation during reconstitu-

tion. Furthermore, the effects of low concentra-

tions of sucrose in the reconstitution solution willbe evaluated.

EXPERIMENTAL SECTION

Materials

Puried anti-L-selectin antibody was produced byBoehringer Mannheim (Mannheim, Germany)

and was stored at 808C until needed. Theprotein is a humanized murine IgG4 antibody.

Upon thawing and dialysis, our SE-HPLC analy-sis (see below) revealed approximately 98%monomer and 2% dimer. Potassium phosphatemonobasic, potassium phosphate dibasic, guani-dine hydrochloride (GdnHCl), infrared gradepotassium bromide (KBr), and Tween 20 (SigmaUltra) were purchased from Sigma ChemicalCompany. High-purity sucrose was purchasedfrom Pfansteihl. Total protein assay (BCA) solu-tions were purchased from Pierce ChemicalCompany. All protein and excipient solutionswere freshly prepared using distilled deionized

TWEEN 20, SUCROSE, AND ANTI-L-SELECTIN 1467

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 90, NO. 10, OCTOBER 2001


3/12

water. Buffer solutions were ltered through0.45-micron nylon lters.

Lyophilization Study

Anti-L-selectin was concentrated to 2.3 mg/mLusing an Amicon stir cell concentrator with a

YM10 lter. The protein was then dialyzedagainst three changes of 600 mL 10 mM potas-sium phosphate buffer (pH 7.2) for at least 18 h.This buffer was chosen to avoid pH decreasesduring freezing. However, it should be noted that

Anchordoguy et al. reported that 10 mM potas-sium phosphate buffer adjusted to pH 7.5 at 228Calkalinized to pH 8.1 upon freezing.33 Follow-

ing dialysis, the protein was reconcentrated to4.8 mg/mL as above. Solutions of 5% (w/v) sucrosein buffer and 1.0% (w/v) Tween 20 in buffer were

prepared in buffer. Anti-L-selectin formulationsto be lyophilized were rst prepared in eppendorftubes by combining the appropriate buffer andexcipient solutions, then adding protein stocksolution to obtain 2 mg/mL anti-L-selectin in atotal volume of 300 mL. The samples were thentransferred to 1-mL lyophilization vials (WestCompany) and placed in a freeze dryer (FTSSystems DuraStop). Table 1 provides a summaryof the excipients and their concentrations in theve lyophilized formulations.

Vials were placed on a room temperature

freeze-dryer shelf. The shelf temperature wasreduced at a rate of 18C per minute until thetemperature of representative sample reached308C. The shelf was held at this temperature for2 h. Then the shelf temperature was reduced to408C at 18C per minute. The chamber pressurewas reduced to 60 mTorr, and shelf temperaturewas maintained at 408C for 12 h. Next, the shelftemperature was increased to 208C at 18C per

minute and held for an additional 6 h. Finally, the

shelf temperature was raised to 308C at a rate of0.58C per minute, and held at this temperature forapproximately 8 h. Vials were stoppered whileunder vacuum, and analysis of the lyophilized

formulations began the same day they were

removed from the freeze dryer. Any vials thatcould not be analyzed that day were stored at808C for no more than 2 days.

Infrared (IR) spectroscopy was used to comparesecondary structure of anti-L-selectin in lyophi-lized formulations to that of native aqueous anti-L-selectin. Spectra were obtained using a BomemProta infrared spectrometer. For the aqueousnative protein sample, dialyzed anti-L-selectinwas concentrated to approximately 20 mg/mL

with a Centricon 10 concentrator. The concen-trated protein was then injected into a cell withCaFl

2windows separated by a 6-micron mylar

spacer, and spectra were collected as described byDong and colleagues.34 For each freeze-driedformulation, the contents of a single vial contain-ing lyophilized anti-L-selectin were combinedwith approximately 300 mg KBr and pressed intoa pellet as described by Kreilgaard et al.5 Thespectra were corrected for background (andbuffer, in the case of the liquid control), convertedto second derivative spectra, and smoothed usinga seven-point function. Using Grams software,the smoothed spectra were area normalized and

overlaid for comparison.35

Reconstitution Study

The excipient solutions for reconstitution wereprepared in water instead of buffer to avoidincreasing the nal buffer salt concentrations ofthe samples. Samples were reconstituted bypipetting 300 mL of the reconstitution solution orwater (room temperature) directly onto the cakesin the vials. Contents were gently swirled by hand

until the solutions were clear. Table 2 shows thereconstitution schemes. Lyophilized formulations

lacking sucrose (buffer alone or Tween 20 as thesole excipient) were reconstituted with water, 1%sucrose, and 0.125% sucrose solutions. Analo-gously, the lyophilized formulations with bufferalone or sucrose as the excipient were reconsti-tuted with water and a 0.1% Tween 20 solution.The formulation lyophilized in buffer alone wasthe only one reconstituted with an aqueous so-lution containing both 0.1% Tween 20 and 0.125%sucrose. Following reconstitution, the sampleswere transferred to eppendorf tubes and centri-fuged in a benchtop microfuge at 48C for 10 min.

Table 1. Final Excipient Concentrations of Anti-L-

Selectin Lyophilization Formulationsa

Formulation

0.125%

Sucrose

1.0%

Sucrose

0.1%

Tween 20

1

2

3

4

5

aAll formulations had a nal anti-L-selectin concentrationof 2 mg/mL. The buffer was 10 mM potassium phosphate(pH 7.2). ()-Indicates the formulation contains the excipient.

1468 JONES ET AL.



4/12

The tubes were then checked visually for pelleted

protein. A total protein assay was performed onthe supernatant (Pierce BCA assay, using anti-L-selectin that had not been lyophilized to construct

the standard curve) to quantify the amount ofsoluble protein remaining. The concentration ofthe anti-L-selectin standard was determined byUV spectroscopy (e 1.45 cm1 g1L).

Size-exclusion high-performance liquid chro-matography (SE-HPLC) was utilized to quantifylevels of monomeric protein and soluble aggre-gates. The various aggregation states of theprotein were separated using a Tosohaas TSK3000SWLx gel ltration column connected to aDionex chromatography system (Sunnyvale, CA).

The mobile phase was 200 mM potassium phos-

phate (pH 6.9) with 150 mM KCl, and the ow ratewas 0.2 mL/min. The samples, chromatographysystem, column, and buffer were all maintained at48C throughout the analysis. Anti-L-selectin thathad not been lyophilized was used as a control.Data collected were imported into Grams softwareprogram, and curve tting was used to deconvolveareas of overlapping peaks. The percentages ofdimer and oligomer contents were calculated bydividing the areas under the curves (AUCs) of the

respective curve-tted peaks by the total AUCs ofall peaks in the chromatogram of a given sample.

Statistical signicance of differences between theamounts of aggregates of the various samples wasobtained using a Student's t-test with a 95%condence interval.

Tween 20 Anti-L-Selectin Interactions

The GdnHCl-induced unfolding of 0.3 mg/mLanti-L-selectin in 10 mM potassium phosphate(pH 7.2) was followed using an Aviv circulardichroism (CD) spectrometer (Model 62DS). Stocksolutions of 7.3 M GdnHCl in buffer, with and

without 0.1% Tween 20, were prepared as

described by Pace et al.36 For the Tween 20experiment, 0.1% Tween 20 was added to thebuffer, protein, and GdnHCl stock solutions prior

to combining the stock solutions to prepare thesamples with various GdnHCl concentrations.Samples lacking anti-L-selectin were used asblanks to correct for non-anti-L-selectin contribu-tions to far UV CD signal. A 1-mm pathlengthsample cell was used for all CD spectroscopicmeasurements. Sixty-second averages of far UVCD signals at 220 nm of samples having 07 MGdnHCl were collected to obtain protein-unfold-ing curves. The fraction-unfolded curves wereconstructed using the linear extrapolation

method previously described by Pace et al.36

Finally, full far UV CD scans of anti-L-selectinin 0 and 7.0 M GdnHCl solutions, with andwithout 0.1% Tween 20, were taken from 260 to211 nm at a step size of 0.5 nm with a 3-saveraging time. These spectra were corrected bysubtracting the wavelength scans of the fourcorresponding solutions without protein.

Binding of Tween 20 to native anti-L-selectinwas investigated using electron paramagneticresonance (EPR) spectroscopy (Bruker ESP300)

as described by Bam et al.30 Briey, 16-doxylstearic acid serves as a hydrophobic spin probe

capable of partitioning into hydrophobic environ-ments. Partitioning affects the spectral signal bybroadening the peaks, resulting from the decreasein rotational mobility of probe molecules inmicelles. The spectrum for any given sample is acombination of signals from freely rotating androtationally hindered probe populations (cf. ref.30). Both pure surfactant micelles and surfac-tantprotein complexes can provide hydrophobicenvironments that result in hindered proberotation. In the present study, the spin probewas formulated with 0 to 1630 mM Tween 20 in

Table 2. Reconstitution Schemes for Lyophilized Anti-L-Selectina

Lyophilization

Formulation

Reconstitution Solutions

Water

0.125% Sucrose

in Water

1.0% Sucrose in

Water

0.1% Tween 20 in

Water

0.125% Sucrose and

0.1% Tween 20 in

Water

1

2

3

4

5

a()-Indicates the lyophilized formulation was reconstituted with this solution.




5/12

both the absence and presence of 10 mg/mL anti-L-selectin. The Tween 20 concentrations spannedbelow and above the critical micelle concentra-tion, CMC, (&60 mM) and provided investigations

of potential interactions from 0:1 to 24:1 molar

ratios of Tween 20:anti-L-selectin. Spectra werecollected at a constant frequency of 9.75 GHz andcentered at a magnetic eld of 3470 G. Data wereanalyzed using DaDisp and Microsoft Excelspreadsheets as previously described.30

RESULTS

Effects of Excipients on Lyophilization-InducedProtein Unfolding

The conformationally sensitive amide I region ofinfrared spectra34 was used to compare the

secondary structure of native anti-L-selectin anti-body in aqueous solution to that found inlyophilized formulations. Infrared spectroscopywas also used to monitor the secondary struc-ture of anti-L-selectin (20 mg/mL) during freezethawing. Within the resolution of this technique,anti-L-selectin's secondary structure was notperturbed during freezing or thawing (data notshown). In addition, freeze-thawing experimentsof 1 mg/mL solutions documented that proteinaggregates were not induced by this treatment.

When anti-L-selectin was freeze dried without

excipients there was a decrease in the peak depthat the main b-sheet band, 1640 cm1, relative tothe spectrum for native, aqueous anti-L-selectin,indicating a loss of nativeb-sheet (Figure 1A). Thebroadening of the bands at 1660 and 1675 cm1

was due to perturbation of native turn structures(Figure 1A). Including 1.0 or 0.125% sucrose(Figure 1A and B ) prevented loss of native b-sheet structures. However, the presence ofsucrose did not completely prevent perturbation

of native turn structures (Figure 1A and B).Tween 20 (0.1%) only partially inhibited loss of

native b-sheet and did not prevent the perturba-tion of turn structures (Figure 1C). Sampleslyophilized with both 0.125% sucrose and 0.1%Tween had secondary structural retention thatwas essentially the same as that for antibodyformulated with sucrose alone (Figure 1D).

Reconstitution Study

After reconstitution, all formulations containedonly soluble protein. Precipitated protein wasnot detected by either visual inspection or by

centrifugation and analysis of the supernatant fortotal protein content (results not shown). How-ever, soluble aggregates were identied usingsize-exclusion high-performance liquid chromato-graphy. Chromatograms of the reconstituted

protein and that of the native control materialwere used for determining recovery of solubleprotein (Figure 2). Two peaks, instead of a singlepeak, were t to the monomer peak to account forits asymmetry. The relative curve-t areas wereused to calculate the fractions of each aggregatetype and monomeric protein.

The soluble aggregate content was measuredfor anti-L-selectin samples that were lyophilizedin buffer alone and reconstituted in various

solutions (Figure 3A). It is important to note thatthe starting material contained 2% dimer and nodetectable higher order oligomers. Water andaqueous sucrose solutions were equivalent recon-

stitution media for this formulation. Dimer levelsremained at 2%, but there was also almost 1.5%oligomer. Formation of oligomers was inhibited byreconstituting with 0.1% Tween 20 solutions, withor without sucrose. However, there was a con-comitant increase in dimers, relative to the levelin the control solution.

Addition of 0.125% or 1.0% sucrose to thereconstitution solution had no effect on the levelof protein aggregation (Figure 3A and C). How-

ever, anti-L-selectin freeze dried with 0.125 or

1.0% sucrose, which inhibited lyophilization-induced unfolding, and reconstituted with waterhad signicantly lower oligomer content thanprotein samples lyophilized with buffer alone andreconstituted with either water or sucrose solu-tions. Reconstitution of either sucrose formula-tion with a 0.1% Tween 20 solution resulted incontrol levels of dimer and no detectable oligo-mers (Figure 3B).

The presence of 0.1% Tween 20 in the lyophi-

lized formulation caused the highest levels ofaggregates noted in this study (Figure 3C). For

both Tween 20 formulations tested, there werehigher levels of dimers than noted for sampleslyophilized with only buffer, independent of thereconstitution solutions tested. Oligomer levelswere approximately the same as those for thesample lyophilized in buffer alone.

Tween 20 Interactions With Anti-L-Selectin

In an attempt to gain insight into potentialmechanisms of these effects of Tween 20 on therecovery of native protein during reconstitution,

1470 JONES ET AL.



6/12

we examined the interactions between Tween 20and anti-L-selectin. Based on far UV CD spectro-scopy the native conformation of this b-sheetprotein was not affected by the presence of 0.1%Tween 20 (Figure 4A). Moreover, the spectra forthe completely denatured state in 7 M guanidineHCl (Figure 4B) were also identical, in thepresence and absence of 0.1% Tween 20. Thus,any Tween bound to the protein did not alter thesecondary structure of the protein in either thenative or denatured state.

EPR spectroscopy was used to determine ifTween 20 bound to anti-L-selectin.30 Analysis ofthe EPR data indicated that Tween 20 did notbind to native anti-L-selectin (Figure 5). If Tween20 bound to anti-L-selectin, we would expect asignicantly higher fraction of the probe mole-cules to partition into micellar environments inthe presence of anti-L-selectin than in its absence(cf. ref. 30), which was not the case (Figure 5).

To determine if Tween 20 alters the thermo-dynamic stability of anti-L-selectin, GdnHCl

Figure 1. Comparison of the effects of excipients on the secondary structure of anti l-

selectin (anti-L-selectin) in the freeze-dried solid. (AC) solid line: Liquid control; dotted

line: anti-L-selectin freeze dried in buffer alone. (A) Dashed line: anti-L-selectin freeze-

dried with 1.0% sucrose in buffer. (B) Dashed line: anti-L-selectin freeze dried with

0.125% sucrose in buffer. (C) Dashed line: anti-L-selectin freeze dried with 0.1%Tween

20 in buffer. (D) Solid line: liquid control; dotted line: anti-L-selectin freeze dried with

0.1%Tween 20 in buffer; dashed line: anti-L-selectin freeze dried with 0.125% sucrose in

buffer; dashed-dotted line: anti-L-selectin freeze dried with 0.1% Tween 20 and 0.125%

sucrose in buffer.




7/12

unfolding curves were measured in the presenceand absence of 0.1% Tween 20 (Figure 6A). Thecurves were superimposable (Figure 6A). Thecalculated free energy of unfolding of anti-L-

selectin was approximately 3.5 kcal/mol in eitherformulation (Figure 6B).

DISCUSSION

Effects of Excipients on Lyophilized-InducedUnfolding of Anti-L-Selectin

In general, retention of native protein structure inlyophilized formulations requires protection ofthe protein during both freezing and drying.16,17

Freezing protection depends on the initial bulkconcentration of sucrose.10 The stabilization ofproteins by sucrose during freezing is explainedby the preferential exclusion mechanism, because

the stabilization actually involves the solutes

in the non-ice phase.37

In contrast, proteinstabilization by sucrose during drying dependson the mass ratio of sugar to protein.38 For alyophilized protein solution, a much lower initialsucrose concentration is needed for protectionduring dehydration than during freezing. Forexample, Allison et al. found that maximumprotection of lyophilized actin, a freeze-labileprotein, was achieved with an initial sucrose:protein mass ratio of 5:1. In contrast, a 1:1 initial

mass ratio of sucrose:actin was sufcient for airdrying, which does not have the need for proteinstabilization prior to dehydration.38 A relatively

low initial concentration of sucrose is sufcient tostabilize a protein during drying because sucroseinhibits dehydration-induced unfolding by repla-cing the hydrogen bonds between the protein andwater, which are lost upon water removal.9,10,13

Under conditions used in the current study,anti-L-selectin did not appear to be freeze labile,and by including as little as 0.125% sucrose (0.5:1sucrose:protein mass ratio), anti-L-selectin in thedried solid had essentially native b-sheet struc-ture. When freeze-thaw labile proteins (e.g.,

rFXIII, actin, lactate dehydrogenase) were lyo-

philized, much higher sugar concentrations or theaddition of a cryoprotectant (e.g., polyethyleneglycol) was necessary to inhibit lyophilization-induced unfolding maximally.2,5,10,12,13,38 Thus,only protection against dehydration stress wasneeded to prevent lyophilization-induced unfold-ing of the anti-L-selectin.

During lyophilization, the presence of Tween20 alone only partially inhibits loss of native b-sheet structure. Because no evidence of freeze-

thaw lability is observed for the anti-L-selectin,Tween is likely providing some protection during

drying. Tween 20 might be preventing some puta-tive surface-mediated damage during drying, butthe mechanisms for this protection are not clear.

Effect of Formulation on Protein Stability DuringLyophilization and Reconstitution in Water

Preservation of native protein secondary struc-ture during lyophilization usually results in na-tive protein upon reconstitution because theprotein does not need to alter its conforma-tion.13,32 Unfolded protein in the freeze-dried

Figure 2. (A) Representative chromatogram and

curve t of reconstituted anti l-selectin (anti-L-selectin

lyophilized in buffer and reconstituted with 1%sucrose). The actual chromatogram is displayed as a

solid line. The curve ts are as follows: dotted line:

oligomer; dashed line: dimer; long-dashed line and

dashed-double dotted line: monomer; and dashed-

dotted line: total curve t. (B) Chromatogram of control.

1472 JONES ET AL.



8/12

solid does not necessarily result in nonnativeprotein because recovery of native protein mole-cules upon reconstitution is largely governed bythe kinetic competition between protein refolding

and aggregation.3 In dried formulations where

most of the native secondary structure is pre-served, some aggregates can still be formed from asmall fraction of the protein population that hasperturbed secondary or tertiary structure. Thus,even for formulations with apparent maximalinhibition of lyophilization-induced perturbationsof secondary structure, it may be necessary toemploy conditions that inhibit aggregation duringreconstitution to optimize recovery of nativeprotein.

As expected, based on the IR spectra of thedried samples, anti-L-selectin lyophilized in thepresence of sucrose contains the least amount of

aggregates after reconstitution with water.Nevertheless, even the formulations with thehighest degree of native b-sheet structure in thedried solid still have detectable oligomers afterreconstitution with water (Figure 2B). This resultis consistent with that of Ressing et al., who founda 10% loss of activity of mouse IgG MN12lyophilized in either the presence or absence ofstabilizers, including sucrose, when reconstitutedimmediately with water.39 In the present study,both sucrose formulations required 0.1% Tween

20 in the reconstitution solution to prevent

oligomer formation and reduce dimers to controllevels.

Lyophilization from Tween 20 solutions resultsin more aggregation of anti-L-selectin than lyo-philization in buffer alone. Sarciaux et al. foundwith bovine IgG that the presence of Tween 80 inthe lyophilized formulation reduced turbidity

Figure 3. Aggregate content of reconstituted sam-

ples. The reconstitution solution, water or excipient in

water, is given on the x-axis. Filled bars: dimer. White

bars: oligomer. Results are means SD for triplicatesamples (A) Anti-L-selectin freeze dried in buffer. (B)

Anti-L-selectin freeze dried in buffer containing

sucrose. (C) Anti-L-selectin freeze dried in buffer

containing 0.1% Tween 20. (Note: the symbols indicate

statistical signicance using Student's t-test with a 95%

condence interval. *Signicantly different from the

control. For a given lyophilization formulation, recon-

stitution with the excipient solution yields signicantly

different results than reconstituting with water. # For

the given reconstitution solution, the result is signi-

cantly different from samples lyophilized in the pre-

sence of buffer alone.)




9/12

after reconstitution.14 Reduced turbidity couldhave been due to a lower level of aggregates and/or a reduction in the size of the aggregates.

The secondary structure of anti-L-selectin inthe solid state was most perturbed in theformulation lyophilized in buffer alone. Consider-

ing only this factor, anti-L-selectin lyophilized inthe presence of buffer alone would be predicted to

have the greatest aggregate content after recon-stitution, which was not observed. Therefore,caution should be taken when using solid-statesecondary structural data to predict the nalaggregation levels. A possible explanation for thediscrepancy is that IR spectroscopy cannot detectchanges that involve only a small fraction of thetotal protein population.23 Maximum aggregatelevels in this study were less than 10%, makingdetection by IR spectroscopy problematic.Furthermore, IR spectroscopy is not sensitive totertiary structure.

Effect of Excipients in the Reconstitution Solution

The presence of sucrose in the reconstitutionmedium did not affect the level of aggregatesmeasured. In contrast, reconstituting formula-tions lyophilized in sucrose with Tween solutionsreduced the levels of aggregates relative to thosemeasured after reconstitution in water. Interest-

ingly, for samples lyophilized in buffer alone,reconstitution with Tween inhibited formation ofoligomers, but increased the levels of dimers. Asimilar result was obtained with recombinanthuman Factor XIII during agitation and freezethawing.26 Tween 20 prevented soluble aggre-gates from assembling further to form precipi-tates.

To investigate potential mechanisms by whichTween 20 fostered increased recovery of native

anti-L-selectin, we tested for effects of the surfac-tant on protein thermodynamic stability and for

potential binding of the surfactant to the nativeprotein. Neither of these effects was detectable.

Addition of 0.1% Tween 20 did not affect the farUV CD ellipticity spectra of either native orcompletely denatured forms of anti-L-selectin,indicating that the presence of 0.1% Tween 20does not alter the secondary structure of theprotein in aqueous solutions (Figure 4), nor did0.1% Tween 20 affect the thermodynamic stability(Figure 6). Furthermore, no detectable bindingof Tween 20 to native anti-L-selectin was obser-ved in EPR spectroscopic studies (Figure 5).

Figure 4. Far UV CD of native (A) and denatured

(B) anti-L-selectin. Solid line: anti-L-selectin in buffer.

Dotted line: anti-L-selectin in buffer containing 0.1%Tween 20.

Figure 5. 16-Doxyl stearic acid partitioning in the

presence and absence of anti-L-selectin. Filled circle:

partitioning in the absence of anti-L-selectin. Opencircles: partitioning in the presence of 10 mg/mL anti-L-

selectin.

1474 JONES ET AL.



10/12

Therefore, the mechanisms for the Tween 20

effects during reconstitution remain unclear.

CONCLUSIONS

We conclude that the stage at which an excipientis present in a formulation (i.e., during lyophiliza-tion or only during reconstitution) can greatlyaffect the recovery of native protein measure afterreconstitution. Optimal protein recovery canrequire the addition of a surface-active agent,such as Tween 20, in the reconstitution solutionas well as minimization of structural perturba-tions in the lyophilized product by the inclusion of

sucrose. However, the presence of a nonionicsurfactant during lyophilization may be detri-mental to protein stability.

ACKNOWLEDGMENTS

We gratefully acknowledge Boehringer Mann-heim for supplying anti-l-selectin and nancialsupport, grants from the Colorado Institute for

Research in Biotechnology (CIRB) and NSF(Grant BES9816975), and an advanced predoc-toral fellowship to L.S.J. from the PharmaceuticalResearch and Manufacturing Association.

REFERENCES

1. Allison SD, Chang B, Randolph TW, Carpenter JF.1999. Hydrogen bonding between sugar and pro-

tein is responsible for inhibition of dehydration-

induced protein unfolding. Arch Biochem Biophys

365:289298.

2. Chang BS, Fischer NL. 1995. Development of an

efcient single-step freeze-drying cycle for protein

formulations. Pharmaceut Res 12:831 837.

3. Chang BS, Kendrick BS, Carpenter JF. 1996.

Surface-induced denaturation of proteins during

freezing and its inhibition by surfacants. J Phar-

maceut Sci 85:13251330.

4. Chang BS, Reeder G, Carpenter JF. 1996. Devel-

opment of a stable freeze-dried formulation of

recombinant human interleukin-1 receptor antago-nist. Pharmaceut Res 13:243249.

5. Kreilgaard L, Frokhaer S, Flink JM, Randolph TW,

Carpenter JF. 1998. Effects of additives on the

stability of recombinant human factor XIII during

freeze-drying and storage in the dried solid. Arch

Biochem Biophys 360:121134.

6. Vermuri S, Yu CD, Roosdorp N. 1994. Effect of

cryoprotectants on freezing, lyophilization, and

storage of lyophilized recombinant alpha 1-anti-

trypsin formulations. PDA J Pharmaceut Sci

Technol 48:241 246.

7. Zhang MZ, Pikal K, Ngyuen T, Arakawa T,

Prestrelski SJ. 1996. The effect of the reconstitu-tion medium on aggregation of lyophilized recom-

binant interleukin-2 and ribonuclease A.

Pharmaceut Res 13:643 646.

8. Zhang MZ, Wen J, Arakawa T, Prestrelski SJ.

1995. A new strategy for enhancing the stability of

lyophilized protein: The effect of the reconstitution

medium on keratinocyte growth factor. Pharma-

ceut Res 12:14471452.

9. Carpenter JF, Pikal MH, Chang BS, Randolph TW.

1997. Rational design of stable lyophilized protein

formulations: Some practical advice. Pharmaceut

Res 14:969975.

Figure 6. (A) Guanidine hydrochloride unfolding of

anti-L-selectin in the presence and absence of 0.1%

Tween 20. (B) Free energy of unfolding of anti-L-

selectin in the presence and absence of 0.1% Tween 20.

Filled circles: anti-L-selectin in buffer. Open circles:

anti-L-selectin in buffer containing 0.1% Tween 20.




11/12

10. Carpenter JF, Prestrelski SJ, Arakawa T. 1993.

Separation of freezing- and drying-induced dena-

turation of lyophilized proteins using stress-specic

stabilization. I. Enzyme activity and calorime-

tric studies. Arch Biochem Biophys 303:456

464.

11. Chang BS, Beauvais RM, Dong A, Carpenter JF.

1996. Physical factors affecting the storage stability

of freeze-dried interleukin-1 receptor antagonist:

Glass transition and protein conformation. Arch


12. Kreilgaard L, Frokhaer S, Flink JM, Randolph TW,

Carpenter JF. 1999. Effects of additives on the

stability of Humicola lanuginosa lipase during

freeze-drying and storage in the dried solid. J

Pharmaceut Sci 88:281 290.

13. Prestrelski SJ, Arakawa T, Carpenter JF. 1993.

Separation of freezing-and drying-induced dena-

turation of lyophilized proteins using stress-specic

stabilization. II. Structural studies using infraredspectroscopy. Arch Biochem Biophys 303:465

473.

14. Sarciaux J-M, Mansour S, Hageman MJ, Nail SL.

1999. Effects of buffer composition and processing

conditions on aggregation of bovine IgG during

freeze-drying. J Pharmaceut Sci 88:13541361.

15. Arakawa T, Kita Y, Carpenter JF. 1996. Polymers

protect lactate dehydrogenase during freeze-drying

by inhibiting dissociation in the frozen state. Arch


16. Carpenter JF, Chang BS. 1996. Lyophilization of

protein pharmaceuticals. In: Avis KE, Wu VL,

editors. Biotechnology and biopharmaceutical man-

ufacturing, processing, and preservation. Buffalo

Grove, IL: Interpharm Press, Inc. pp. 199264.

17. Carpenter JF, Kendrick BS, Chang BS, Manning

MC, Randolph TW. 1999. Inhibition of stress-

induced aggregation of protein therapeutics. Meth-

ods Enzymol 309:236255.

18. Kerwin BA, Heller MC, Levin SH, Randolph TW.

1998. Effects of Tween 80 and sucrose on acute

short-term stability and long-term storage at

20 degrees C of a recombinant hemoglobin. J

Pharmaceut Sci 87:1062 1068.

19. Cleland JL, Lam X, Kensrick BS, Yang J, Yang T-Z,

Overcashier D, Brooks D, Hsu C, Carpenter JF.

2001. A specic molar ratio of stabilizer to protein isrequired for storage stability of a lyophilized mono-

clonal antibody. J Pharmaceut Sci 90:310326.

20. Lee JC, Timasheff SN. 1981. The stabilization of

proteins by sucrose. J Biol Chem 256:71937201.

21. Liu Y, Bolen DW. 1995. The peptide backbone plays

a dominant role in protein stabilization by natu-

rally occurring osmolytes. Biochemistry 34:12884

12891.

22. Qu Y, Bolen BL, Bolen DW. 1998. Osmolyte-driven

contraction of a random coil protein. Proc Natl Acad

Sci USA 95:92689273.

23. Allison SD, Dong A, Carpenter JF. 1996. Counter-

acting effects of thiocyanate and sucrose on chymo-

trypsinogen secondary structure and aggregation

during freezing, drying, and rehydration. Biophys J

71:20222032.

24. Bam NB, Cleland JL, Yang J, Manning MC,

Carpenter JF, Kelley RF, Randolph TW. 1998.

Tween protects recombinant human growth hor-

omone against agitation-induced damage via hydro-

photic interactions. J Pharmaceut Sci 87:1554

1559.

25. Kataham M, Bell LN, Banga AK. 1995. Effect of

surfacants on the physical stability of recombinant

human growth hormone. J Pharmaceut Sci 84:

713716.

26. Kreilgaard L, Jones LS, Randolph TW, Frokhaer S,

Flink JM, Manngin MC, Carpenter JF. 1998. Effect

of Tween 20 on freeze-thawing- and agitation-

induced aggregation of recombinant human factor

XIII. J Pharmaceut Sci 87:1597 1603.27. Maa YF, Nguyen PA, Hsu SW. 1998. Spray-drying

of airliquid interface sensitive recombinant

human growth hormone. J Pharmaceut Sci 87:

152159.

28. Jones LS, Cipolia D, Liu J, Shire SJ, Randolph TW.

1999. Investigation of proteinsurfactant interac-

tions of analytical ultracentrifugation and electron

paramagnetic resonance: The use of recombinant

human tissue factor as an example. Pharmaceut

Res 16:808812.

29. Bam NB, Cleland JL, Randolph TW. 1996. Molten

globule intermediate of recombinant human

growth hormone: Stabilization with surfactants.

Biotechnol Prog 12:801 809.

30. Bam NB, Randolph TW, Cleland JL. 1995. Stability

of protein formulations: Investigation of surfactant

effects by a novel EPR spectroscopic technique.

Pharmaceut Res 12:211.

31. Carpenter JF, Prestrelski SJ, Dong A. 1998.

Application of infared spectroscopy to develop-

ment of stable lyophilized protein formulations.

Eur J Pharmaceut Biopharmaceut 45:231

238.

32. Prestrelski S, Tedeschi N, Arakawa T, Carpenter

JF. 1993. Dehydration-induced conformational

changes in proteins and their inhibition by stabi-

lizers. Biophys J 65:661671.33. Anchordoguy TJ, Carpenter JF. 1996. Polymers

protect lactate dehydrogenase during freeze-drying

by inhibiting dissociation in the frozen state. Arch

Biochem Biophys 332:231 238.

34. Dong A, Huang P, Caughey WS. 1990. Protein-

secondary structures in water from second-deriva-

tive amide I infrared spectra. Biochemistry 29:

33033308.

35. Kendrick BS, Dong A, Allison SD, Carpenter

JF. 1996. Quantitation of area of overlap bet-

ween second derivative amide I infrared spectra

1476 JONES ET AL.



12/12

to determine structural similarity of a protein

in different states. J Pharmaceut Sci 85:155

158.

36. Pace BN, Shirley BA, Thomson JA. 1989. Measur-

ing the conformational stability of a protein. In:

Creighton TE, editor. Protein structure: A practice

approach. New York: IRL Press at Oxford Uni-

versity Press. pp. 311330.

37. Arakawa T, Prestrelski SJ, Kenney WC, Carpenter

JF. 1993. Factors affecting short-term and long-

term stabilities of proteins. Adv Drug Del Rev

10:128.

38. Allison SD, Randolph TW, Manning MC, Middleton

K, Davis A, Carpenter JF. 1998. Effects of

drying methods and additives on structure and

function of Actin: Mechanisms of dehydration-

induced damage and its inhibition. Arch Biochem

Biophys 358:171 181.

39. Ressing ME, Jiskoot W, Talsma H, van Ingen CW,

Beuvery EC, Crommelin DJ. 1992. The inuence of

sucrose, dextran, and hydropropyl-beta-cyclodex-

trin as lyoprotectants for a freeze-dried mouse

IgG2a monoclonal antibody (MN12). Pharmaceut

Res 9:266270.



Documents

Jones, 2001. Liofilización. Sacarosa y Tween