Characterization of dextrin-based hydrogels: Rheology, biocompatibility, and degradation

Characterization of dextrin-based hydrogels: Rheology,biocompatibility, and degradation

Joana Carvalho,1 Susana Moreira,1 Joao Maia,2 Francisco M. Gama11IBB, Institute for Biotechnology and Bioengineering, Department of Biological Engineering, Universidade do Minho,Campus de Gualtar, 4710-057 Braga, Portugal2I3N, Department of Polymer Engineering, Institute for Nanostructures, Nanomodelling and Nanofabrication,Universidade do Minho, 4800-058 Guimaraes, Portugal

Received 1 October 2008; revised 26 March 2009; accepted 7 April 2009Published online 30 June 2009 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jbm.a.32553

Abstract: A new class of degradable dextrin-based hydro-gels (dextrin-HEMA) was developed. The hydroxyethylmethacrylate ester (HEMA) hydroxyl groups were acti-vated with N,N0-carbonyldiimidazole (CDI), followed bytheir coupling to dextrin, yielding a derivatized materialthat can be polymerized in aqueous solution to formhydrogels. A comparative study of the stability of the dex-trin-HEMA hydrogels and dextrin-vinyl acrylate (dextrin-VA, produced in previous work) revealed that only thefirsts are effectively hydrolyzed under physiological condi-tions. A severe mass loss of dextrin-HEMA gels occursover time, culminating in the complete dissolution of thegels. Rheologic analysis confirmed that physical structur-

ing is less pronounced when dextrin is modified withmethacrylate side groups. The biocompatibility resultsrevealed that the dextrin hydrogels have negligible celltoxicity, irrespective of the hydrogel type (HEMA andVA), allowing cell adhesion and proliferation. Gatheringthe biocompatibility and the ability to tailor the releaseprofiles, we consider dextrin a promising biomaterial forbiomedical applications, namely for controlled release.� 2009 Wiley Periodicals, Inc. J Biomed Mater Res 93A:389–399, 2010

Key words: dextrin; hydrogels; biocompatibility; degrada-tion kinetics

INTRODUCTION

Hydrogels are a class of three-dimensional, hydro-philic polymeric networks that can absorb substan-tial amounts of water. In general, hydrogels have agood biocompatibility and are capable of resemblingnatural tissues, allowing for a favorable controlledinteraction with living systems.1 For this reason,these materials have found widespread applicationin the biomedical, biotechnological, and pharmaceu-tical fields.2–8 Hydrogels can be used in tissue engi-neering, as scaffolds to support and promote tissueregeneration and also as attractive systems for thecontrolled release of pharmaceutically active mole-cules.9–14

An important class of hydrogels is obtained by thepolymerization of acrylate esters [e.g., vinyl acrylate,methacrylate, hydroxyethyl methacrylate] throughthe polymerization of the water-soluble functional-

ized polymers.15–21 In previous work, we havedescribed a new dextrin-based hydrogel in whichthe polysaccharide is functionalized with VA. Thereactive double bonds cross-links by free-radical po-lymerization in aqueous solution.22 The degradationof the ester bonds present in the crosslinks of thosepolymeric networks is very slow under physiologicalconditions. However, the hydrogels could be ren-dered degradable through the incorporation of anenzyme, which prove to be an effective route tomodulate the gel degradation and the release ofentrapped molecules. Alternatively, the functionali-zation of the polymer with hydroxyethyl methacry-late ester (HEMA) was attempted, as alreadydescribed for dextran, as an efficient approach toobtain hydrogels with desirable spontaneous (nonen-zymatic) degradation kinetics.16–19,23 The productionof dextrin-HEMA hydrogels is reported in this work.It should be remarked that dextrin, an a-(1?4) glu-cose polymer, was used instead of dextran [a-(1?6)]. A comparative study of the stability of thetwo different acrylated dextrins (dextrin-VA anddextrin-HEMA) was also carried out. The rheologicanalysis allowed exploring the time evolution of the

Correspondence to: F. M. Gama; e-mail: [email protected]

� 2009 Wiley Periodicals, Inc.

viscoelastic behavior of the hydrogels, providing aqualitative characterization of the network, as wellas the structural changes occurring during degrada-tion. The biocompatibility of the hydrogels is anessential issue regarding their pharmaceutical andbiomedical applicability.2,24–28 In this context, in vitrotoxicity of the hydrogels was evaluated (followingthe guidelines from ISO10993) and observations withboth fluorescence and phase-contrast inverted lightmicroscopy were also carried out as indicators ofcell viability and morphology.

MATERIALS AND METHODS

Materials

Dextrin-Koldex 60 starch was a generous gift from Tate& Lyle (Decatur, IL). VA was from Aldrich, N,N,N0,N0-tetramethylenethylenediamine (TEMED) and ammoniumpersulfate (APS) were purchased from BioRad, dimethyl-sulfoxide (DMSO) and acetone were from AppliChem.2-Hydroxyethyl methacrylate (HEMA), hydroquinonemonomethyl ether (>98%), 4-(N,N-dimethylamino)pyridine(DMAP, 99%), N,N0-carbonyldiimidazole (CDI, 98%), andtetrahydrofuran (THF) were from Fluka. Dulbecco’s modi-fied Eagle medium (DMEM) with 4500 mg glucose/liter,L-glutamine, 110 mg sodium pyruvate/liter and NaHCO3

(Sigma) was supplemented with 10% calf bovine serum(CBS) (Gibco, UK) and 1% of a penicillin/streptomycin so-lution (Sigma). This medium is further referred to asDMEM complete medium. MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl-tetrazolium bromide) was obtained fromSigma and transwell plates were purchased from OrangeScientific (Belgium). For the fluorescence microscopyassays, the LIVE/DEAD1 Viability/Cytotoxicity Kit formammalian cells, from Molecular Probes (USA), was used.Regenerated cellulose tubular membranes with 3500 mo-lecular weight cutoff (MWCO) were obtained from Mem-brane Filtration Products (Seguin, USA). All other chemi-cals and solvents used in this work were of the highestpurity commercially available.

Synthesis of HEMA-imidazolylcarbamate (HEMA-CI)

The synthesis of HEMA-CI was performed as describedby vanDijkWolthuis et al.18 CDI (1.62 g, 10 mmol) was dis-solved in about 15 mL anhydrous THF in a nitrogenatmosphere and HEMA (1.30 g, 10 mmol) was added. Thereaction mixture was stirred for 16 h at ambient tempera-ture. After addition of a small amount of hydroquinonemonomethyl ether (60 mg), the solvent was evaporatedand the crude product was dissolved in ethyl acetate,extracted with water and dried on anhydrous MgSO4. Af-ter filtration, hydroquinone monomethyl ether (60 mg) wasadded once more and the solvent evaporated.

Synthesis of dextrin-HEMA

The modification of dextrin with HEMA was performedas described by vanDijkWolthuis et al.18 with few modifi-cations. Briefly dextrin (10.0 g) was dissolved in driedDMSO (90 mL) in a nitrogen atmosphere. After dissolutionof DMAP (2.0 g), HEMA-CI was added, in the molar ratiosof 0.25 and 0.50 of HEMA-CI to glucopyranose residues indextrin. The solution was stirred at room temperature for4 days, after which the reaction was stopped by adding 2mL of concentrated HCl. The reaction mixture was dia-lyzed for 2–3 days against demineralized water at 48C. Themethacrylated dextrin was lyophilized and the final whitefluffy product was stored at 2208C until use. The degreeof acrylate substitution (DS) of the methacrylated dextrinwas determined by proton nuclear resonance spectroscopy(1H NMR) in D2O. Briefly, 1H NMR spectra were recordedin D2O (10 mg in 1 mL) and DS was calculated usingEq. (1):

DS% ¼ x

y3100 ð1Þ

in which x, is the average integral of the protons at thedouble bond (d at 5.8 and 6.2 ppm) and y is the integral ofthe anomeric proton (d around 5.4 ppm).

General procedure for the preparation ofdextrin-VA and dextrin-HEMA hydrogels

Dextrin-VA monomers were synthesized from dextrin inDMSO in the presence of different amounts of VA and thedegree of acrylate substitution (DS) was determined byproton nuclear resonance spectroscopy (1H NMR) in D2Oas previously described.22 The hydrogel slabs with differ-ent DS (20 and 40%) were prepared by radical polymeriza-tion of aqueous solution of either dextrin-VA or dextrin-HEMA (300 mg/mL). The gelation reactions were initiatedby adding 90 lL APS (80 mg/mL in phosphate buffer)and 90 lL TEMED (13.6% (v/v) in water, pH adjusted to8.0 with HCl) and allowed to occur for 30 min at roomtemperature.

Standard degradation conditions

Two sets of hydrogels were used. After the polymeriza-tion reaction, gels were removed from the casting recipi-ents. One set of hydrogels was immediately lyophilized todetermine the initial dry weight (Wi). The other one wasimmersed in 50 mL of phosphate buffer saline (PBS) pH7.4 and kept in a thermostated room at 378C. At time inter-vals, gels were removed, blotted with filter paper toremove water excess, weighted (Ws, weight in the swollenstate) and lyophilized to determine the dry weight, Wd.The swelling ratio (SR) and the mass loss (mloss) werecalculated according to Eqs. (2) and (3), respectively:

SR ¼ Ws �Wd

Wdð2Þ

390 CARVALHO ET AL.

Journal of Biomedical Materials Research Part A

mloss ¼ 100� Wi

Wd3100

� �ð3Þ

The hydrogel degradation time is defined as the timeneeded for complete degradation (W 5 0).

Rheologic measurements

The rheologic measurements were performed in a con-trolled-stress Rheologica rheometer at a controlled tempera-ture (258C). A plate–plate measuring system was used (Ø25 mm) and the gap between the plates was 1 mm. For theexperiments, polymerized hydrogels with 25-mm diameterand 1-mm thickness, were immersed in 50 mL of PBS atpH 7.4 and kept in a thermostated room at 378C to allowdegradation, with the exception of the initial time mea-surements (t 5 0), in which hydrogels were used immedi-ately after polymerization. At set time intervals, gels wereremoved from the containers and placed between theplates of the measuring system. Two sets of experimentswere performed:

(a) Stress sweep experiments were carried out to opti-mize the applied stress used in the frequency-oscil-lation experiments. The lower limit of the appliedstress is determined by the sensitivity of the instru-ment. On the other hand, the upper limit of theapplied stress will be influenced by the linear visco-elastic range and the presence of slip effects.

(b) Once the range of linear viscoelastic response wasdetermined, experiments were performed in SmallAmplitude Oscillatory Shear, SAOS, at a fixed fre-quency of 1 Hz and a stress known to yield a goodquality signal in the linear viscoelastic regime.

Cell culture

Mouse embryo fibroblasts 3T3 (ATCC CCL-164) weregrown in DMEM complete medium, at 378C, in a fullyhumidified air containing 5% CO2 (IR auto Flow). The cellswere fed every 2–3 days. When cells reached confluence,the culture medium was discarded and the cells weredetached with 2 mL of 0.25% (w/v) trypsin-EDTA [1:250,from porcine pancreas (Sigma)] solution for 15 min at378C, and 6 mL of DMEM complete medium was addedto inactivate the trypsin after cell detachment. The cellswere then centrifuged (10 min, 1000 rpm) and resus-pended in culture medium before use.

Mitochondrial metabolic activity assay

MTT solution (0.45 mg/mL in DMEM complete mediumwithout phenol red) in an amount equal to 10% of the cul-ture volume was added to each well. After 3 h of incuba-tion at 378C, the MTT solution was removed and the insol-uble formazan crystals formed in the bottom of the wellswere dissolved in a dimethylsulfoxide volume equal to theoriginal culture volume.29 The absorbance was measured

at 570 nm using a plate reader (SynergyHT, BioTek). TheCPII was calculated using Eq. (4):

CPII ¼ 100� DO570 of test culture

DO570 of control culture3 100

� �ð4Þ

Direct contact assay

A fibroblast suspension (3 mL) containing 3 3 104 cells/mL was plated into each well of a six-well plate, to yield afinal density of 104 cells/cm2. After reaching a state of sub-confluence (after 24 h), hydrogel discs (Ø 4 mm, 2-mmthickness) were placed on the wells, in direct contact withcells. At regular time intervals, the cell morphology andviability was assessed with the Live/Dead fluorescencelabeling. Positive and negative controls (discs of latex andagar gel) were also used.

Adhesion assay

The cell adhesion in dextrin-VA hydrogels was eval-uated. After 2 h of polymerization, the hydrogels formedat the bottom of the wells were washed, first with PBS andsecondly with DMEM complete medium (3 3 200 lL);then 100 lL of a fibroblast suspension containing 3 3 104

cells/mL was plated into each well, to yield a final densityof 104 cells/cm2. The culture medium was refreshed (100lL) every 2 days. For the control assays, cells were growndirectly in the bottom of the wells. At regular time inter-vals, the hydrogels were washed with PBS (3 3 100 lL) toremove floating cells, and the cell layer was trypsinizedbefore conducting the colorimetric (MTT) and countingassays, an approach that resulted in more accurate andreliable results.

Microscopy assays

At regular time intervals, cell growth and morphologywere evaluated by observation with a phase-contrastinverted light microscope. Cytotoxic effects were ranked incellular death, cell adhesion ability, and induction of mor-phological changes. Additionally, cells were labeled withLIVE/DEAD Viability/Cytotoxicity Kit and observed byfluorescence microscopy.

RESULTS AND DISCUSSION

In previous work, we have demonstrated the pro-duction of a new dextrin hydrogel, through thetransesterification of the polymer with VA, followedby free-radical polymerization.22 In that work, it wasdemonstrated that the esterification occurs on the 2-(preferentially) and 3-hydroxyl groups in the gluco-pyranose ring. Furthermore, a comprehensive studywas performed, by solid state NMR, allowing thedetection of adsorbed, linked and polymerized

CHARACTERIZATION OF DEXTRIN-BASED HYDROGELS 391


acrylic molecules. The dextrin-VA hydrogel degradesslowly and to a limited extent, under physiologicalconditions, which is a limitation in light of theirapplication as controlled delivery systems or in tis-sue engineering. Several studies reported the suc-cessful functionalization of dextran with a methacry-late ester (HEMA) and it was shown that hydrogelsobtained by polymerization of this compound, areable to degrade in desirable timeframes.16–19,23 Inthis work, a similar dextrin modification wasattempted using the same chemistry. As dextrin israther distinct from dextran (different size, differentstructure), distinct properties are expected and char-acterized in this work.

Activation of HEMA with CDI and synthesisof dextrin-HEMA

As previously described for dextran,18 it was alsopossible to couple HEMA to the hydroxyl groups ofdextrin via a mixed carbonate ester. This linkage can

be effectively introduced by using N,N0-CDI as anintermediate activation step. In this activation reac-tion, HEMA reacts with an equimolar amount ofCDI, in THF, to produce HEMA-imidazolyl carba-mate, hereafter referred to as HEMA-CI (Fig. 1). Inthe next step, HEMA-CI reacts with dextrin, usingDMSO as reaction medium. 1H RMN analysis showsthat dextrin-HEMA with different DS (ranging fromca. 10 to 40%) can be obtained, by varying the molarratio of HEMA-CI to dextrin (Fig. 2). It is assumedthat the dextrin esterification with HEMA occurspreferentially at the 2- and 3-hydroxyl groups of theglucopyranose ring, as previously shown for dextrin-VA. In addition, the modified polymer was success-fully recovered by lyophilization, after dialysis.

Swelling and degradation behavior of dextrin-VAand dextrin-HEMA hydrogels

On exposure to an aqueous solution, the watermolecules are drawn into the hydrogel network. The

Figure 1. Synthesis of dextrin-HEMA.

Figure 2. Dextrin-HEMA 1H NMR spectrum.

392 CARVALHO ET AL.


initial and relaxed state of the hydrogel (dry state)is, thereby, replaced by a swollen state. As the watermolecules are going to occupy some space, themeshes of the network will start expanding, allowingfor the uptake of more water molecules. The stretch-ing caused by the swelling process is counterbal-anced by the covalent or physical crosslinking of thenetwork, preventing its destruction and creating anequilibrium swelling state.30 The functionality andnumber of linkages in the hydrogel, as well as themonomer concentration, influence not only the SR,but also the erosion profile and the overall time fordegradation.5 In this work, a comparative study onthe swelling and degradation behavior, was per-formed in hydrogels with two distinct ester func-tions [acrylate (dextrin-VA) and methacrylate (dex-trin-HEMA)], and the influence of the DS on theseparameters was also evaluated. Dextrin-VA hydro-gels with DS 20 or DS 40 were compared withdextrin-HEMA with the same DS.

As can be seen in Figure 3(A), for dextrin-VA, atphysiological pH, in the initial period of the incuba-tion (around 5 days), a very slight increase in the SRis detected for both degrees of substitution (DS 20 orDS 40). Nevertheless, neither gel showed a signifi-cant increase in swelling over 100 days, indicatingthat no significant hydrolysis of the polymerizedacrylate esters occurred. As expected, greater swel-ling values (in the range 2.5–3) were found for the DS20 networks, compared with those (in the range 1.6–2.1) obtained with the higher DS, because in highlycrosslinked hydrogels the mobility of the polymerchains is hindered, lowering the ability to swell.

Contrarily to the dextrin-VA hydrogels, in whichthe SR remains almost constant during the timecourse of the experiment, in the case of dextrin-HEMA, a progressive swelling occurred over thetime, until the complete dissolution of the hydrogelsis accomplished. The increased swelling is likely aconsequence of the hydrolysis of the methacrylate

bonds in the crosslinked structure. The cleavage ofthe ester bonds leads to a decrease in the crosslink-ing density, allowing for the network to expand andswell, absorbing more water. As can be seen in Fig-ure 3(A), the SR values reach a maximum at thetime that the gels starts to dissolve and dextrinhydrogels most likely bear few crosslinks per poly-mer chain. The mass loss of dextrin-VA/HEMAhydrogels with different DS (DS 20 and DS 40) wasalso monitored as a function of time. Figure 3(B)summarizes the degradation times of all dextrin gelsinvestigated. This figure shows that for both gel sys-tems (dextrin-VA or dextrin-HEMA), the degrada-tion time increased with an increasing DS. It seemsobvious that an increasing DS at a fixed initial watercontent (all the hydrogels have the same initial watercontent of about 70%), results in a network with ahigher crosslinking density. In these cases, to accom-plish the network dissolution, more cross-links haveto be hydrolyzed, which requires more time. Whencomparing the two hydrogels systems, a linear massloss profile can be observed in dextrin-VA, until day20, with a rate of roughly 0.5% of gel solubilized perday. However, following this initial period, the rateslows down, and at day 100, only about 40 and 25%of mass loss is detected for DS 20 and DS 40 hydro-gels, respectively. Probably, the mass losses observedcorrespond to the release of nonpolymerized mate-rial.

Contrarily, as shown in Figure 3(B), for dextrin-HEMA hydrogels, the mass loss was dramatic,reaching about 30% for DS 20 and 50% for DS 40,after 20 days of incubation. In this case, hydrogelsstart to disintegrate quickly, until a complete disso-lution is achieved. As also expected, while for a DS20 dextrin-HEMA hydrogel, the dissolution tookplace in �35 days, in case of DS 40 hydrogels, fur-ther time was needed to accomplish dissolution (60days). These results suggest that, in the case of dex-trin-HEMA hydrogels, the degradation occurs via a

Figure 3. Degradability of dextrin-based hydrogels. Swelling ratio (A) and mass loss as a function of time, whenimmersed in PBS pH 7.4, at 378C (data representative of three experiments; the standard deviation was always inferior to5%). (*From day 40 the hydrogel structure was severely damaged, preventing the SR measurements).



homogeneous bulk erosion process. In this case, theSR would be expected to increase with the incuba-tion time, because of the loosened polymer networkbecause, although hydrolyzed, the polymer chainsremain in the matrix, allowing the network tobecome looser and to continuously swell, until thedissolution is complete.31–34 In the case of dextrin-VA hydrogels, it seems that the initial mass loss canbe attributed to the release of a fraction of unboundmolecules which were physically entrapped in thematrix at the time of polymerization. As diffusionoccurs, the small molecules are released out of thenetwork. A stabilization of the swelling values is,thus, observed at the time that all the unbound mol-ecules are released. In contrary to dextrin-HEMA,which is hydrolytically labile, dextrin-VA hydrogelis very stable under physiological conditions.

Mechanical properties

Mechanical properties of hydrogels are very im-portant for pharmaceutical applications. For exam-ple, the integrity of a drug delivery device duringthe lifetime of the application is crucial. The systemmust protect a sensitive therapeutic agent, such asprotein, maintaining its integrity until it is releasedout of the system. Changes in the crosslinking den-

sity of the hydrogels have been routinely applied toachieve the desired mechanical properties. Increasingthe crosslinking will result in a stronger gel with amore brittle structure, which sometimes is not desir-able because it shall interfere with the release pro-files. Hence, a compromise must be achievedbetween a relatively strong but yet elastic struc-ture.35 It is well established that the release of theproteins from a hydrogel is clearly influenced by itsdegradation rates.36 Dynamic mechanical analysis,performed on the hydrogel, provide quantitative in-formation on the viscoelastic and rheologic proper-ties of the material, by measuring the mechanicalresponse of the samples as they are deformed. Theelastic (also real or storage) modulus G0 and the vis-cous (also imaginary or loss) modulus G@ are pre-sented. The storage modulus [G0, Fig. 4(A)] of ahydrogel is proportional to the amount of elasticchains within the polymer network.37 Figure 4 showsthe rheologic results for all the hydrogels and it isclear that the results are in good agreement with thedegradability studies above. Dextrin-HEMA hydro-gels show lower G0 values, as compared with thedextrin-VA ones, indicating that physical structuringis less pronounced when dextrin is modified withmethacrylate side groups. Overall, it is possible tobroadly correlate an increase in SR with a decreasein both the elastic and viscous responses, G0 and G@

Figure 4. Rheologic measurements of the different hydrogel samples. Frequency is 1 Hz and stress varies to keep theresponse in the linear viscoelastic regime. Representative results selected among several assays are shown.

394 CARVALHO ET AL.


respectively, because the gels become softer as theyswell, which tends to decrease G0, whereas theincreased water absorption, because of the hydroly-sis of the methacrylate bonds in the crosslinkedstructure, leads to a decrease in G@ [Fig. 4(B)].

In the case of the dextrin-VA gels, G0 and G@decrease for short incubation times and stabilizeafter �30 days. As there is a large uptake of waterand these gels do not degrade, the viscous responseincreases slightly relatively to the elastic one, asreflected by the slight increase in tan(a), but theviscoelastic behavior clearly remains that of a softsolid [Fig. 4(C)]. The DS has an appreciable impacton the storage modulus. Dextrin-VA hydrogels witha higher DS (DS 40) has more crosslink sites, whichresults in a denser mesh structure, more networkchains, and an associated higher G0, compared withthe lower DS (DS 20). In the case of the dextrin-HEMA gels, the degradation is very severe, with theviscoelastic response quickly approaching that of ainelastic low-viscosity liquid, that is, low G0 and G@and tan(a) � 1. This degradation culminated in theimpossibility to perform any tests on the dextrin-HEMA gels, because of their extreme fragility above20 days of rest.

From these results, it could be concluded that agood correlation exists between the influence of theester function (methacrylate/acrylate) used in thepolymer derivatization, as well as the grafted grouppercentage (DS), on the rheologic degradation rates.

In vitro biocompatibility assessment

To evaluate the cytotoxicity of the two acrylate-derivatized dextrins, we have studied the effects ofthese materials in a mouse embryo fibroblast culture,namely in the cell proliferation and adhesion on thehydrogel, morphological features, and cell death.

Extraction assay

As previously described, on exposure to aqueoussolutions, hydrogels can release leachable productsand, therefore, the cytotoxicity of hydrogel extractsmust be evaluated. For that purpose, fibroblast cul-tures were exposed to extracts of hydrogels, pre-pared using dextrin-VA and dextrin-HEMA with dif-ferent DS (20 and 40%). Viability was quantifiedusing the MTT assay. The results are shown in TableI. The extracts induced a reduction inferior to 15% inthe mitochondrial metabolic activity of fibroblasts, ascompared with the control. No significant differenceswere found between the two classes of hydrogels(HEMA and VA). The DS of the hydrogel networkdoes also not produce major differences in the CPII.

Direct contact

The cytotoxicity of dextrin-VA and dextrin-HEMAhydrogels was also assessed using a direct contactmethod. For that purpose, hydrogel discs as well asa positive and negative control (latex rubber andagar, respectively) were used. Each sample wasplaced on top of a subconfluent cellular layer on asix-well plate, as described in the material and meth-ods section. Figure 5 shows the results obtained withthe fluorescence labeling of the cell layers.

As can be seen, the latex rubber is, as expected,highly cytotoxic. Some cells were floating in the me-dium and the ones that remained attached are deadas well (red labeled). Differences in the morphologywere found in the cells underneath the hydrogels.These cells appeared clustered, when compared withthe control cultures, but all of them remained alive,as shown by the green fluorescence (data notshown). The cell proliferation under the dextrinhydrogel was lower than the one observed in thecontrol plate. Massia and Stark,38 and more recentlyFerreira et al.39,40 using human skin fibroblasts, havealready reported this effect which is related to me-chanical compression of cells or poor oxygenationand nutrient diffusion, because of the physical pres-ence of the network. The cells in contact with thenegative control (agar) and the dextrin hydrogels,irrespective of the chemistry and DS, exhibited simi-lar morphology and proliferation rates, thus, thehydrogels can be considered nontoxic.

Cell adhesion and proliferation ondextrin hydrogels

The cell adhesion on the hydrogel surfaces wasevaluated using samples with different DS (20 and40%). The results were expressed as a percentage ofthe adhesion on cell culture polystyrene plates(TCPP), estimated by cell counting. Additionally, the

TABLE ICell Proliferation Inhibition Index of Mouse Embryo

Fibroblasts, Cultured in the Presence of ExtractsObtained from Dextrin Hydrogels

Hydrogel

CPII (%)a, DS (%)b

20 40

Dextrin-VA 14.9 6 1.6 14.8 6 2.1Dextrin-HEMA 14.2 6 1.3 13.7 6 1.7

aCell proliferation inhibition index assessed by MTTmeasurements 6 standard deviation.

bDegree of substitution defined as the amount of acry-late groups per 100 glucopyranose residues, determined byliquid NMR spectroscopy.



mitochondrial activity of the adherent cells was alsoassessed in a parallel experiment. This study showedthat the hydrogels, once again irrespective of the

chemistry (HEMA or VA) and DS, are as expected,less adhesive than TCPP. The resistance of thehydrogel surfaces to cell adhesion is, as described

Figure 5. Fluorescence labeling of mouse embryo fibroblasts. Micrographs show the control culture, the cell layers in con-tact with the positive (latex rubber) and negative (agar) controls, and DS 20 dextrin hydrogels (VA/HEMA), after 24 h ofincubation (magnification 310 and 340). [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Figure 6. Fluorescence labeling of mouse embryo fibroblasts. Micrographs show control culture and cells seeded on thesurface of DS 20 dextrin hydrogels (VA/HEMA), after 5, 24, and 48 h of incubation (magnification 310). A–C: Graphicscorrespond to the MTT absorbance values at 570 nm through time course of the assay. [Color figure can be viewed in theonline issue, which is available at www.interscience.wiley.com.]

396 CARVALHO ET AL.


for other polysaccharides such as dextran, presum-ably because of poor protein adsorption onto thehydrophilic and nonionic polymer.41–43 Indeed, cellu-lar adhesion on hydrophilic materials such as hyal-uronic acid and PEG-based hydrogels was reportedto be extremely sparse.44,45 In other cases, as recentlydescribed in pullulan hydrogels using smooth mus-cle cells, cellular adhesion, and moderate growth isobserved.46 Viability tests indicated that fibroblastsadhered to some extent (50–60%) when comparedwith the adhesion on TCPP and are capable to growon dextrin hydrogels. In the first stage of the culture,the cells appear round and then attach to the gel sur-face. As can be seen in Table II, after 1 h, <30% ofthe cells were attached to the hydrogel surface; after5 h, almost 60% of the initial cells remained associ-ated to the gels.

As the gels reticulated in the bottom of the wellsare slightly opaque, fluorescence labeling was usedto facilitate the evaluation of the cell morphology.The results are presented in Figure 6. After 24 h,small aggregates of round cells are visible. After 2days, larger cell aggregates are visible, and the cellsstart spreading out from these aggregates. The Live/Dead assay demonstrates the nontoxic character ofthe material, because no dead cells are observed.

The CPII evaluation is shown in Figure 7. As canbe seen, 48 h after the cell seeding on the hydrogelsurfaces, cell proliferation inhibition reaches about50% when compared with control cultures. Thedirect seeding of the cells onto the hydrogels surfa-ces resulted in a substantial reduction in the prolifer-ation rate. However, as can be seen by the non-nor-malized absorbance values [Fig. 6(A–C)], the resultsshow that although with a delay, cells are effectivelyable to grow, indicating that no deleterious effectsare produced by dextrin hydrogels.

CONCLUSIONS

It was possible to obtain a new class of polymeriz-able dextrins (dextrin-HEMA) by using a method

based on the activation of the hydroxyl groups ofHEMA with CDI, followed by their coupling withthe polymer, yielding a derivatized material that canbe polymerized in aqueous solution to form hydro-gels. The comparative study of the stability of dex-trin-HEMA/dextrin-VA hydrogels revealed that thefirsts are effectively hydrolyzed under physiologicalconditions. In dextrin-HEMA hydrogels, a progres-sive swelling is accompanied by a severe mass lossprocess, which culminates with the complete disso-lution of the gels. Accordingly, the rheologic analysisrevealed that physical structuring is less pronouncedwhen dextrin is modified with methacrylate sidegroups. It can be concluded that the dextrin func-tionalization with HEMA, may be an effective routeto modulate the degradation kinetics, tailoring therelease profiles to fit the purpose of the controlleddelivery.47 In light of their biomedical application,the cytotoxicity of the two classes of hydrogels wasinvestigated. The biocompatibility tests, expressed asdysfunction in the metabolic activity of the cells,gave similar results irrespective of the hydrogel type(HEMA and VA), revealing a negligible cell toxicity,allowing cell adhesion and proliferation. Theseinsights into the hydrogel structure-toxicity relation-ship could be helpful for optimizing the biocompati-bility of polymeric delivery systems.

Additionally, the modification of the surface of thehydrogels with RGD peptides, using recombinantstarch binding modules (SBM) fused to the bioactivepeptides, was proved to significantly improve thecell adhesion and surface spreading (ongoing work).This is a significant advantage on using a starchderived hydrogel (in this case, dextrin), because theavailability of SBMs allows a wide variety of bioac-tive peptides to be driven to the hydrogels surfaces.Furthermore, the interaction of this hydrogel with

TABLE IIMouse Embryo Fibroblasts Adhesion onto Dextrin-VA and

Dextrin-HEMAHydrogels with Different DSValues(% Estimated Taking the Cell Growth on TCPP as Reference)

Hydrogel DS (%)

Adhesion (%)

After 1 h After 5 h

Dextrin-VA 20 23.4 6 2.0 56.5 6 4.440 22.8 6 3.8 59.1 6 2.7

Dextrin-HEMA 20 25.7 6 4.6 59.2 6 3.740 23.9 6 3.1 59.7 6 2.6

Figure 7. Cell proliferation inhibition index (CPII)(mean% 6 SD) after 48 h for different dextrin hydrogeldiscs. CPII obtained using MTT. Data are from two repre-sentative experiments, each one performed in triplicate



different cell lines and living tissues—subcutaneousimplants—is currently being analyzed.

References

1. Vyavahare N, Kohn J. Photocrosslinked hydrogels based oncopolymers of poly(ethylene glycol) and lysine. J Polym SciPart A: Polym Chem 1994;32:1271–1281.

2. Peppas NA. Biomaterials Science: An Introduction to Materi-als in Medicine. San Diego: Academic Press; 1996.

3. Kane JB, Tompkins RG, Yarmush ML, Burke JF. BiomaterialsScience: An Introduction to Materials in Medicine. San Diego:Academic Press; 1996.

4. McCulloch I, Shalaby SW. Tailored polymeric materials forcontrolled delivery systems. Washington, DC: AmericanChemical Society; 1998.

5. Peppas NA, Bures P, Leobandung W, Ichikawa H. Hydrogelsin pharmaceutical formulations. Eur J Pharm Biopharm 2000;50:27–46.

6. Chen TH, Embree HD, Brown EM, Taylor MM, Payne GF.Enzyme-catalyzed gel formation of gelatin and chitosan:Potential for in situ applications. Biomaterials 2003;24:2831–2841.

7. Drury JL, Mooney DJ. Hydrogels for tissue engineering: Scaf-fold design variables and applications. Biomaterials 2003;24:4337–4351.

8. Ishihara M, Ono K, Sato M, Nakanishi K, Saito Y, Yura H,Matsui T, Hattori H, Fujita M, Kikuchi M, Kurita A. Accelera-tion of wound contraction and healing with a photocrosslink-able chitosan hydrogel. Wound Repair Regen 2001;9:513–521.

9. Okay O, Gurun C. Synthesis and formation mechanism ofporous 2-hydroxyethyl methacrylate ethylene-glycol dimetha-crylate copolymer beads. J Appl Polym Sci 1992;46:401–410.

10. Walther DH, Sin GH, Blanch HW, Prausnitz JM. Pore-sizedistributions of cationic 2-hydroxyethyl methacrylate (Hema)hydrogels. Polym Gels Networks 1995;3:29–45.

11. Levy MC, Andry MC. Mixed-walled microcapsules made ofcross-linked proteins and polysaccharides—Preparation andproperties. J Microencapsul 1991;8:335–347.

12. Schacht E, Bogdanov B, VandenBulcke A, DeRooze N.Hydrogels prepared by crosslinking of gelatin with dextrandialdehyde. React Funct Polym 1997;33:109–116.

13. Disilvio L, Gurav N, Kayser MV, Braden M, Downes S. Bio-degradable microspheres—A new delivery system forgrowth-hormone. Biomaterials 1994;15:931–936.

14. Draye JP, Delaey B, van de Voorde A, van Den Bulcke A,Bogdanov B, Schacht E. In vitro release characteristics ofbioactive molecules from dextran dialdehyde cross-linkedgelatin hydrogel films. Biomaterials 1998;19:99–107.

15. vanDijkWolthuis WNE, Franssen O, Talsma H, Vansteenber-gen MJ, Vandenbosch JJK, Hennink WE. Synthesis, characteri-zation, and polymerization of glycidyl methacrylate derivat-ized dextran. Macromolecules 1995;28:6317–6322.

16. vanDijkWolthuis WNE, Hoogeboom JAM, vanSteenbergenMJ, Tsang SKY, Hennink WE. Degradation and releasebehavior of dextran-based hydrogels. Macromolecules 1997;30:4639–4645.

17. vanDijkWolthuis WNE, KettenesvandenBosch JJ, vanderKerk-vanHoof A, Hennink WE. Reaction of dextran with glycidylmethacrylate: An unexpected transesterification. Macromole-cules 1997;30:3411–3413.

18. vanDijkWolthuis WNE, Tsang SKY, KettenesvandenBosch JJ,Hennink WE. A new class of polymerizable dextrans withhydrolyzable groups: Hydroxyethyl methacrylated dextranwith and without oligolactate spacer. Polymer 1997;38:6235–6242.

19. vanDijkWolthuis WNE, vanSteenbergen MJ, Underberg WJM,Hennink WE. Degradation kinetics of methacrylated dextransin aqueous solution. J Pharm Sci 1997;86:413–417.

20. Ferreira L, Carvalho R, Gil MH, Dordick JS. Enzymaticsynthesis of inulin-containing hydrogels. Biomacromolecules2002;3:333–341.

21. Ferreira L, Gil MH, Dordick JS. Enzymatic synthesis of dex-tran-containing hydrogels. Biomaterials 2002;23:3957–3967.

22. Carvalho J, Goncalves C, Gil AM, Gama FM. Production andcharacterization of a new dextrin based hydrogel. Eur PolymJ 2007;43:3050–3059.

23. Stubbe BG, Braeckmans K, Horkay F, Hennink WE, DeSmedt SC, Demeester J. Swelling pressure observations ondegrading dex-HEMA hydrogels. Macromolecules 2002;35:2501–2505.

24. Hansen MB, Nielsen SE, Berg K. Re-examination and furtherdevelopment of a precise and rapid dye method for meas-uring cell-growth cell kill. J Immunol Methods 1989;119:203–210.

25. Behl C, Davis JB, Lesley R, Schubert D. Hydrogen-peroxidemediates amyloid-beta protein toxicity. Cell 1994;77:817–827.

26. Berridge MV, Tan AS, McCoy KD, Wang R. The biochemicaland cellular basis of cell proliferation assays that use tetrazo-lium salts. Biochemica 1996;4:15–20.

27. Hern DL, Hubbell JA. Incorporation of adhesion peptidesinto nonadhesive hydrogels useful for tissue resurfacing.J Biomed Mater Res 1998;39:266–276.

28. Hoffman AS. Hydrogels for biomedical applications. AdvDrug Deliv Rev 2002;54:3–12.

29. Mosmann T. Rapid colorimetric assay for cellular growth andsurvival—Application to proliferation and cytotoxicity assays.J Immunol Methods 1983;65:55–63.

30. Flory PJ. Principles of Polymer Chemistry. New York: CornellUniversity Press; 1953.

31. Martens P, Holland T, Anseth KS. Synthesis and characteri-zation of degradable hydrogels formed from acrylatemodified poly(vinyl alcohol) macromers. Polymer 2002;43:6093–6100.

32. Suggs LJ, Kao EY, Palombo LL, Krishnan RS, Widmer MS,Mikos AG. Preparation and characterization of poly(propyl-ene fumarate-co-ethylene glycol) hydrogels. J Biomater Sci:Polym Ed 1998;9:653–666.

33. Suggs LJ, Krishnan RS, Garcia CA, Peter SJ, Anderson JM,Mikos AG. In vitro and in vivo degradation of poly(propyl-ene fumarate-co-ethylene glycol) hydrogels. J Biomed MaterRes 1998;42:312–320.

34. Wong WH, Mooney DJ. Synthesis and properties of biode-gradable polymers used as synthetic matrices for tissueengineering. In: Atala A, Mooney DJ, editors. Synthetic Bio-degradable Polymer Scaffolds. Boston: Birkhauser; 1997.p 49–82.

35. Anseth KS, Bowman CN, BrannonPeppas L. Mechanicalproperties of hydrogels and their experimental determination.Biomaterials 1996;17:1647–1657.

36. Meyvis TKL, De Smedt SC, Demeester J, Hennink WE. Influ-ence of the degradation mechanism of hydrogels on theirelastic and swelling properties during degradation. Macro-molecules 2000;33:4717–4725.

37. Ferry JD. Viscoelastic Properties of Polymers. New York:John Wiley; 1980.

38. Massia SP, Stark J. Immobilized RGD peptides on surface-grafted dextran promote biospecific cell attachment. J BiomedMater Res 2001;56:390–399.

39. Ferreira L, Rafael A, Lamghari M, Barbosa MA, Gil MH, Cab-rita AMS, Dordick JS. Biocompatibility of chemoenzymati-cally derived dextran-acrylate hydrogels. J Biomed Mater ResA 2004;68A:584–596.

398 CARVALHO ET AL.


40. Ferreira LS, Cabrita AMS, Rafael A, Gil MH, Dordick JS. Bio-compatible and degradable biocatalytic dextran hydrogels.Abstr Papers Am Chem Soc 2003;225:U239–U239.

41. Osterberg E, Bergstrom K, Holmberg K, Schuman TP, Riggs JA,Burns NL, Vanalstine JM, Harris JM. Protein-rejecting ability ofsurface-bound dextran in end-on and side-on configurations—Comparison to PEG. J Biomed Mater Res 1995;29:741–747.

42. Frazier RA, Matthijs G, Davies MC, Roberts CJ, Schacht E,Tendler SJB. Characterization of protein-resistant dextranmonolayers. Biomaterials 2000;21:957–966.

43. De Groot CJ, Van Luyn MJA, Van Dijk-Wolthuis WNE, CadeeJA, Plantinga JA, Den Otter W, Hennink WE. In vitro biocom-patibility of biodegradable dextran-based hydrogels testedwith human fibroblasts. Biomaterials 2001;22:1197–1203.

44. Mann P, Sussmann M, Sarbia M, Meyer-Kirchrath J, GrabitzK, Schror K, Fischer JW. Regulation of hyaluronic acid syn-thesis in human saphenous vein smooth muscle cells byinterleukin-1 and prostaglandins. Naunyn SchmiedebergsArch Pharmakol 2005;371:R87–R87.

45. Ramamurthi A, Vesely I. Smooth muscle cell adhesion oncrosslinked hyaluronan gels. J Biomed Mater Res 2002;60:196–205.

46. Autissier A, Letourneur D, Le Visage C. Pullulan-basedhydrogel for smooth muscle cell culture. J Biomed Mater ResA 2007;82A:336–342.

47. Carvalho J, Coimbra MA, Gama FM. New dextrin-vinylacry-late hydrogel: Studies on protein diffusion and release Carbo-hydr Polym 2009;75:322–327.



Documents

Characterization of dextrin-based hydrogels: Rheology, biocompatibility, and degradation