Novel porous soy protein‐based blend structures for ...meitalz/Articles/B22.pdf · Preparation of...

Novel porous soy protein-based blend structures for biomedicalapplications: Microstructure, mechanical, and physical properties

Hilla Barkay-Olami, Meital Zilberman

Department of Biomedical Engineering, Faculty of Engineering, Tel-Aviv University, Tel-Aviv 69978, Israel

Received 11 December 2014; revised 31 March 2015; accepted 13 May 2015

Published online 00 Month 2015 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jbm.b.33459

Abstract: Use of naturally derived materials for biomedical

applications is steadily increasing. Soy protein has advan-

tages over various types of natural proteins employed for

biomedical applications due to its low price, nonanimal ori-

gin, and relatively long storage time and stability. In the cur-

rent study, blends of soy protein with other polymers

(gelatin, alginate, pectin, polyvinyl alcohol, and polyethylene

glycol) were developed and studied. The mechanical tensile

properties of dense films were studied in order to select the

best secondary polymer for porous three-dimensional struc-

tures. The porous soy–gelatin and soy–alginate structures

were then studied for physical properties, degradation behav-

ior, and microstructure. The results show that these blends

can be assembled into porous three-dimensional structures

by combining chemical crosslinking with freeze-drying. The

soy–alginate blends are advantageous over soy–gelatin

blends, demonstrated better stability, and degradation time

along with controlled swelling behavior due to more effective

crosslinking and higher water uptake than soy–gelatin blends.

Water vapor transmission rate experiments showed that all

porous blend structures were in the desired range for burn

treatment [2000–2500 g/(m2 d)] and can be controlled by the

crosslinking process. We conclude that these novel porous

three-dimensional structures have a high potential for use as

scaffolds for tissue engineering, especially for skin regenera-

tion applications. VC 2015 Wiley Periodicals, Inc. J Biomed Mater

Res Part B: Appl Biomater 00B: 000–000, 2015.

Key Words: soy protein, natural polymers, gelatin, alginate,

tissue regeneration

How to cite this article: Barkay-Olami H, Zilberman M. 2015. Novel porous soy protein-based blend structures for biomedicalapplications: Microstructure, mechanical, and physical properties. J Biomed Mater Res Part B 2015:00B:000–000.

INTRODUCTION

Various natural, synthetic, and semisynthetic materials arecurrently being utilized in the fabrication of implantablescaffold devices.1 There is a growing tendency to replacesynthetic polymers for biomedical uses with natural, abun-dant, nonanimal origin, and low-cost biodegradable productssuch as soy protein.2,3 These plant proteins can beemployed to meet the high demand for new materialsneeded for various tissue engineering applications. Theyhave several advantages over animal proteins (e.g., collagen,fibrinogen, and silk), including lower immunogenicity. Theseplant proteins, therefore, have a great potential for use asbiomaterials for medical applications.

Soy proteinSoy protein presents good water resistance, a low price, andstorage stability and is also very versatile. The properties ofsoy protein can be modified by physical, chemical, or enzy-matic treatments.4 This enables tailoring soy protein for thediverse requirements of different biomedical applications.

The suitability of soy protein for biomedical applicationshas recently been investigated, due to its abovementionedadvantages and bioactive properties. Proposed applications

range from bone cement5 to hydrogels,6 wound dressingmembranes for promoting wound healing,7 and, morerecently, drug delivery carrier films,7,8 temporary replace-ment implants, and tissue engineering scaffoldings.9

Soy protein and plant proteins, in general, have lowermolecular weights than animal proteins and, as a result,are highly susceptible to enzymatic degradation in thehuman body. Chemical modifications such as crosslinkingare, therefore, essential for achieving the necessarymechanical properties and for ensuring integrity andwater stability of plant protein-based biomaterials duringthe implantation period.10 Crosslinking agents, includingformaldehyde11 and glutaraldehyde,12 can form bridgesbetween protein chains by reacting with the amino groupsof lysine residues and lead to improvement of the mechan-ical and other properties of crosslinked soy protein isolate(SPI) films.

However, glutaraldehyde has been shown to be toxiceven at a relatively low concentration, such as 3.0 ppm,when it is released into the host following biodegradationof biomaterials. Promising solutions for biomedical purposesare glyoxal,9 a dialdehyde with lower toxicity comparedwith similar agents, and water-soluble carbodiimide 1-ethyl-

Correspondence to: M. Zilberman; e-mail: [email protected] grant sponsor: Israel Science Foundation, CISF; contract grant number: 1055/11

VC 2015 WILEY PERIODICALS, INC. 1

3-[3-dimethylaminopropyl]carbodiimide hydrochloride(EDC).

Another approach to improving the physical propertiesof SPI materials is to blend the soy protein with differentpolymers. It is assumed that blending with polymers resultsin materials with better performance than their separatecomponents. To date, SPI have been blended with variousplasticizers and biodegradable polymers such as polysaccha-rides, proteins, and synthetic polymers in order to achievethe desired properties.13

Additional natural polymers include sodium alginate,14

casein,15 cellulose,16 chitin,17 chitosan,18 gluten,19 whey pro-tein isolate,20 and gelatin.4 Despite numerous studies oncomposite films based on soy protein, there have been onlya few reports on the preparation of porous three-dimensional SPI-based blend structures. Furthermore, thereare no published studies on the preparation of chemicallycrosslinked porous soy–natural polymer blend structuresfor biomedical applications. Our proposed novel porous soy-based blend structure provides a biostable and cost-effective nonanimal origin solution for biomedical applica-tions and for cell delivery systems in tissue regeneration.

Additional natural polymers used in this studyIn the current study, we assumed that combining soy pro-tein with other proteins and polysaccharides and syntheticpolymers may result in physical and chemical interactionsthat may improve the material’s properties and biologicalperformance for better cell affinity. Gelatin and alginatewere selected for this study as secondary components dueto their suitable range of mechanical properties and bioac-tive characteristics, as well as their proven history as bio-materials. The polysaccharide pectin was also chosen forexamination, because it has recently been reported that itpromote cell adhesion, thus stimulating cell growth andproliferation, and for due to its physical properties, whichresemble those of alginate. However, despite their manyadvantages, these types of biomaterials suffer from thesame limitation as other biodegradable polymers, namelyuncontrolled rate of degradation and antigenic potential.After crosslinking with EDC or glyoxal, the soy–naturalpolymer hybrid material will become even more biostableand resistant to degradation, due to the formation of newstable covalent bonds and higher molecular weight, asdesired for porous structures. Finally, we chose freeze-drying as our fabrication method for new soy proteinblend materials, based on an attempt to create biomimeticthree-dimensional structures. It is important to note thatthe various blends are composed mainly of soy protein,with a relatively small amount of gelatin or polysaccharide(pectin and alginate).

The main goal of this research was to develop and studynovel porous soy protein-based blend structures as poten-tial new materials for biomedical applications. These newmaterials can be used either as an acellular matrix (e.g.,wound dressings) or as a carrier system for cells or activebiomolecules in the field of tissue engineering, especially forskin and cartilage.

MATERIALS AND METHODS

MaterialsPrimary component. Soy protein source: Non-GMO SPI(Solpro 910TM; minimum 90% wt/wt protein, on dry basis)was obtained from SolbarTM (Ashdod, Israel) as a donation.

Additional natural and synthetic polymers. Gelatin “typeA” from porcine skin (90–100 and 300 bloom) was pur-chased from Sigma-Aldrich, Rehovot, Israel (G6144 andG2500, respectively). Pectin from apple (76282), alginic acidsodium salt [viscosity �250 cps, 2% (258C); A2158], poly-ethylene glycol (PEG; MW 35,000), and polyvinyl alcohol(PVA; MW 23,000) were purchased from Sigma-Aldrich.

Crosslinkers. N-(3-dimethylaminopropyl)-N’-ethylcarbodiimidehydrochloride, a water-soluble carbodiimide (EDC) E7750, andglyoxal (50649) were purchased from Sigma-Aldrich. Sodiumchloride was purchased from Merck (1064041000).

Preparation of dense films for the first step of the studyThe effects of a secondary polymer on the mechanical prop-erties of dense blend films were studied. The films wereprepared using the solvent-casting method as describedelsewhere.21 The formulation and process are presented inTable I. After completing this step, selected secondary poly-mers were chosen for the rest of the study (on porousblend structures).

Preparation of porous soy protein-based blendstructuresSoy protein–gelatin and soy protein–alginate (or pectin)blends were prepared by hydrating the second componentat room temperature for 20 min in double-distilled water ata concentration of 5 wt %. For all blends, a final soy proteinconcentration of 5 wt % was reached by slowly suspendingthe soy protein powder under constant stirring in double-distilled water. The soy protein slurry and the second com-ponent solution were then mixed together to obtain a seriesof blends with a final soy protein content of 5 wt % andsecond biopolymer content of 12, 25, and 50 wt % relativeto the SPI (8:1, 4:1, and 2:1 weight compositions, respec-tively). All blend preparation parameters are presented inTable II.

In order to create porous structures, a crosslinker solutioncontaining EDC or glyoxal in double-distilled water (1 or 4 wt% relative to the soy protein) was cast into a 3-cm diameterPetri dish. The crosslinked samples were then washed threetimes with saline. Finally, the porous soy protein-based blendsamples were obtained by freezing at 2408C for 4 h followedby freeze-drying at 2208C for 18–24 h.

Tensile mechanical propertiesThe tensile mechanical properties were measured in a drystate, at room temperature, using a 5500 Instron universaltesting machine (Instron Engineering, Corp.) with 2-kN loadcell, under unidirectional tension at a rate of 50 mm/minaccording to the standard method ASTM D638-03. The sam-ples were cut into a dumbbell-shape (neck length 2 cm,

2 BARKAY-OLAMI AND ZILBERMAN SOY PROTEIN-BASED BLEND STRUCTURES

width 5 mm). The maximal strain was defined as the break-ing strain. Five samples were tested for each specimen.Means and standard deviations (SDs) were calculated usingExcel.

Water vapor transmission rate (WVTR)The moisture permeability of the porous structures wasdetermined by measuring the water vapor transmission rate(WVTR) through the material. A Sheen Payne permeabilitycup with an exposure area of 10 cm2 was filled with 5 mLdouble-distilled water and was then covered with a circularsponge (n5 3 specimens per group). The cup was placed inan upright position inside an oven, which contained 1 kg offreshly dried silica gel at 378C. The weight of the assemblywas measured every hour and a graph of the evaporatedwater versus time plotted. The WVTR was calculatedaccording to Eq. (1):

WVTR ¼ slope324

area

g

m23day

� �(1)

Water uptakeRound specimens (1.5 cm average diameter) of untreatedand porous crosslinked soy structures (n5 3 specimens pergroup) were preweighed separately and immersed in asaline solution, pH 7, at 378C. The weight of the sampleswas measured after 30 min, 1 h, 2 h, and up to 2 days, by

removing the saline and blotting them gently to removeexcess fluid with a filter article. The water uptake was cal-culated using Eq. (2):

Water uptake %ð Þ ¼ Wwet2Wdry

Wdry3100 (2)

In vitro degradability studyThe degradability behavior of the porous soy-based blendstructures was studied by both quantitative and qualitativemethods.

Quantification of noncovalent soluble protein by 10Murea. Urea in concentrations up to 10M is a powerful pro-tein denaturant since it disrupts their noncovalent bonds.This property can be exploited to increase proteins’ solubil-ity. A degradation test was performed by immersing thesoy-based samples in urea (10M) for up to 2 days. An ali-quot of each sample’s supernatant was taken at differenttime points. The quantity of the soluble protein was ana-lyzed using the Bradford protein assay. A calibration curvefor correlating absorbance at 595 nm to protein contentwas prepared using soy protein and bovine serum albumin(BSA). The calibration curve was then used to calculate thesoluble protein content in the sample. The assay was per-formed in triplicate.

In vitro qualitative degradation. Porous soy protein blendstructures of 30 mm diameter (n5 3 specimens per group)were placed in a 30-mm diameter Petri dish, and the sam-ples were immersed into 4.0 mL Dulbecco’s Modified Eagle’sMedium (DMEM) at pH 7.4, supplemented with 1% PEN-STREP Nystatin solution, all from Biological Industries, andthen incubated at 378C for up to 21 days. Samples werequalitatively checked for robustness every 3 days, using atweezers.

Microstructure of porous blendsThe structure of dry porous structures was observed usingan environmental scanning electron microscope (ESEM).The cryogenically fractured surfaces were Au-sputteredbefore observation. Surfaces of selected samples wereobserved using a Quanta 200 FEG ESEM in a high vacuummode with an acceleration voltage of 10 kV (n5 2 pergroup). The approximate porosity of the observed

TABLE I. The Formulation and Parameters of Preparation of the Dense Blend Films Used in the First Step of the Study

Parameters Component and Concentrations

Primary component Soy protein, 5% wt/volSecondary component Natural polymers: high- (HMW) and low- (LMW) molecular weight

gelatin and alginateSynthetic polymers: PEG (MW 35,000) and PVA (MW 23,000)

Concentrations of secondary component 15%, 25%, and 50% wt/wt relative to soy protein (soy protein:secondpolymer, 8:1, 4:1, and 2:1 weight ratios, respestively)

Temperature of the aqueous solution (8C) 25, 60, 80pH and buffer solutions DDW, PBS, and saline: at pH 7 and 10

TABLE II. The Formulation and Parameters of Preparation of

the Soy–Natural Polymer Blend Porous Structures

ParameterValues Usedin the Study

Soy protein concentration 5 wt %Natural polymer Alginate, pectin,

and gelatinSoy protein:natural

polymer ratio2:1, 4:1, and 8:1

Temperature of soy proteindispersion in double-distilledwater

808C

Temperature of soyprotein–naturalpolymer mixture

608C

Crosslinking agent andconcentration

Glyoxal, EDC, 1–4 wt% relative to SPI

ORIGINAL RESEARCH REPORT

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | MONTH 2015 VOL 00B, ISSUE 00 3

morphologies was analyzed using the Scandium software.Porosity was determined as the area occupied by the poresdivided by the total area.

Statistical analysisAll data were processed using Excel software. Statisticalcomparison between more than two groups was performedusing the ANOVA (Tukey Kramer) method via SPSS (version15) software. A value of p< 0.05 was considered statisticallysignificant. All errors and error bars indicate the SD fromthe mean.

RESULTS

Mechanical properties of dense soy protein blend filmsThe first step in this study was to evaluate the effect of asecondary polymer on the mechanical properties of densesoy protein films. Our assumption was that nonporous soyprotein blend film compositions with improved mechanicalproperties (compared with a reference soy protein film)would be able to maintain mechanical stability also asporous constructs. They could thus serve as the basicselected formulation for the second phase of the study.

Our results showed that the polymer mixture obtainedfrom the 4:1 soy protein:natural polymer composition pro-duced the most suitable blend mixtures, which yielded suit-able porous structures. We, therefore, chose this ratio as thepreferred composition. In order to choose the most suitable

secondary biopolymers, we prepared dense blend films con-taining soy protein and a secondary biopolymer at a 4:1weight ratio and measured their tensile properties com-pared with those of the soy protein reference film. Theresults are presented in Figure 1.

Reference films prepared from the pH 7 solutionresulted in a tensile strength of 5.67 Mpa. The addition oflow-molecular weight (LMW) gelatin did not show a signifi-cant change in the strength and maximal strain, while addi-tion of high-molecular weight (HMW) gelatin and alginateincreased the tensile strength to 7.71 and 7.82 Mpa, respec-tively. The addition of the synthetic polymers PVA and PEGto the soy protein films decreased the tensile strength ofthe soy protein films, while addition of alginate drasticallydecreased the maximal strain of the soy protein films. Addi-tion of PVA and PEG had almost no effect on the maximalstrain. Based on the tensile strength results of the densefilms, HMW gelatin and alginate were chosen as secondarypolymers for the porous structures.

Physical propertiesWater vapor transmission rate. WVTR was measured forvarious types of noncrosslinked and crosslinked porous soy-based blend formulations, and the results are presented inFigure 2. Evaporative water loss through the various poroussoy-based blend formulations was linearly dependent ontime (R2� 0.99 in all cases, results not shown), resulting ina constant WVTR. Noncrosslinked samples exhibited WVTRvalues in the range of 2500–2700 g/(m2 d), with no signifi-cant difference. The WVTR of an exposed aqueous surfacewas also determined as control. Crosslinking of the soy–algi-nate substantially increased its WVTR, while crosslinkedporous soy–gelatin and pure soy protein structuresremained unchanged.

Water uptake. The fluid adsorption capacity was studied onblends containing soy protein and a second component (gel-atin or alginate) in relative quantities of 4:1 (wt/wt). Cross-linked samples were used to measure the fluid absorptioncapacity over time and the results presented in Figure 3(a).The glyoxal crosslinked soy–gelatin and EDC crosslinked

FIGURE 1. Tensile properties of soy-based blend films based on a 4:1

soy protein:polymer ratio, compared with soy reference film: (a) ten-

sile strength and (b) maximal strain. The secondary component is

indicated in the x axis. Data represent the average 6 standard devia-

tion of five independent blend samples. [Color figure can be viewed

in the online issue, which is available at wileyonlinelibrary.com.]

FIGURE 2. Water vapor transmission rate of 4:1 soy:natural polymer

blends: , noncrosslinked blends; , crosslinked blends, using 1 wt.

% EDC. Data represent the average 6 standard deviation of three inde-

pendent samples.


http://wileyonlinelibrary.com

soy–alginate exhibited a similar water uptake pattern, asfollows:

a. A rapid initial water uptake within the first 30 min,reaching the maximum swelling.

b. A decrease in water content during the following 2 h.c. Constant water uptake between 2 and 25 h.

The water uptake values of the soy–gelatin blends dur-ing stages (a) and (b) were higher than those obtained forthe soy–alginate blends [Figure 3(a)].

The EDC crosslinked soy–gelatin blends displayed asteady increase in water content over time until equilibriumwas achieved after 2 h [Figure 3(a)]. The final water uptake

(after 25 h) of this blend was significantly lower than thatof the glyoxal crosslinked soy–gelatin or of the EDC cross-linked soy–alginate blends [Figure 3(a) and Table III].

In order to elucidate the effect of crosslinking on thewater uptake of the blends, the water uptake of the threetypes of soy-based blends was compared with that of non-crosslinked blends. The result for the soy–gelatin blend ispresented as an example in Figure 3(b). The noncrosslinkedblend exhibited rapid water uptake, reaching a value of435% after 30 min, followed by a decrease in water uptakeduring the following 2 h, probably due to weight loss of thesoluble noncrosslinked fraction. After an additional hour, itwas completely dissolved. In contradistinction, the EDCcrosslinked soy–gelatin blend exhibited an increase in wateruptake with time, reaching an equilibrium after 2 h. It isimportant to note that the pure crosslinked and noncros-slinked soy samples, as well as noncrosslinked soy–alginateblends, practically dissolved within a very short time of 30min (results not shown).

The dimensional increase of the crosslinked blends due towater uptake is presented in Table III. It should be noted thatthe gelatin-containing blends exhibited a greater increase indiameter compared with the alginate-containing blend.

In vitro degradation tests.a. Quantification of noncovalent soluble protein

Various types of soy blends were immersed in 10Murea. The in vitro degradation rate was determined bydetecting the soluble protein content in water superna-tants over time using the Bradford assay. As shown inFigure 4, the crosslinked soy–gelatin [Figure 4(a)] andsoy–alginate [Figure 4(b)] samples were significantly lesssoluble in 10M urea than their noncrosslinked counter-parts. The effect of the EDC concentration on the degreeof soy protein solubility of both types of samples arealso presented in Figure 4(a,b). Both types of samplesshowed a tendency toward a decrease in protein solubil-ity, <10% soluble protein, with an increase in the EDCconcentration of up to 4 wt %. The in vitro degradationprofiles of the soy blend samples crosslinked with 1 wt% EDC are presented in Figure 5. The solubility curve ofpure soy protein was higher than that of the soy–gelatinand soy–alginate blends. Minimum solubility wasobserved for the soy–alginate group.

b. Qualitative in vitro degradationThe degradation time of soy–natural polymer blends at

various time points was determined qualitatively byobserving robustness using forceps to maneuver thestructure. The results are presented in Tables IV and V.

FIGURE 3. Water uptake over time (wt %) for: (a) porous crosslinked

soy blends using 1 wt % crosslinker: , soy–gelatin crosslinked with

EDC; , soy–alginate crosslinked with EDC; , soy–gelatin crosslinked

with glyoxal. (b) , Noncrosslinked soy–gelatin blend; , porous

crosslinked soy–gelatin blend, using 1 wt % crosslinker. Data repre-

sent the average 6 standard deviation of three independent samples.

[Color figure can be viewed in the online issue, which is available at

wileyonlinelibrary.com.]

TABLE III. Maximal Water Uptake for Crosslinked (1 wt %) Soy Porous Blends After 24 h and the Dimensional Increase During

Swelling After 48 h of Incubation, for Crosslinked Soy Porous Blends

Dimensional Increase (Diameter, %) Maximum Water Uptake (wt %) Crosslinked Soy Porous Blend

27.2 6 1.9a 308.5 6 14.2b Soy–gelatin (1 wt % EDC)26.5 6 1.7a 609.0 6 51.4 Soy–gelatin (1 wt % glyoxal)18.6 6 1.0 429.4 6 51.1b Soy–alginate (1 wt % EDC)

Results marked with the same lowercase letter are not significantly different (p> 0.05).




In order to evaluate the effect of the natural additiveconcentration on degradation time, a series of cross-linked soy-based blends containing soy protein:naturalpolymer at ratios of 2:1, 4:1, and 8:1 were prepared. Theresults for the soy–alginate blends are presented as anexample in Table IV. Qualitative robustness observationsrevealed that a higher alginate concentration in the soy–alginate samples resulted in a decreased degradationrate. Two types of crosslinking agents were used on soy–gelatin samples in order to study the effect of crosslink-ing agents on degradation time: 1 wt % glyoxal and 1 wt% EDC solutions. Qualitative robustness observationsrevealed that crosslinked soy–gelatin samples with EDCremained intact and maintained their structure for a lon-ger time (14 days) than samples crosslinked with glyoxal(7 days) (Table V). The noncrosslinked 4:1 soy–naturalpolymer samples in both groups were completelydegraded within 3 days.

MicrostructureEffect of the polymer concentration. Figure 6 shows ESEMfractographs for crosslinked pure soy protein [Figure 6(a,b)]

and crosslinked 4:1 and 2:1 soy–gelatin blends [Figure6(c,d) and 6(e,f), respectively]. A pure soy protein spongeexhibited an irregular porous structure, while the blendsexhibited a more organized structure. The 4:1 blend exhib-ited a smaller and more uniform pore structure than the2:1 blend.

Effect of the natural polymers. Figure 7 shows three dif-ferent blends that were used in the current study (soy–gelatin, soy–alginate, and soy–pectin), all crosslinked with1 wt % EDC and based on the 4:1 soy protein:naturalpolymer blend formulation. The three blends yielded dif-ferent resultant matrix microstructures. Analysis of theblends under ESEM revealed a homogenous dispersion ofthe soy in the gelatin, alginate, and pectin at 4:1 ratios[Figure 7(b,d,f)]. Porous soy–alginate-samples [Figure7(c,d)] yielded a well-interconnected morphology com-pared with the porous soy–gelatin blend [Figure 7(a,b)].The pore size and distribution were relatively similar forboth soy–gelatin and soy–alginate samples. The totalporosity of soy–gelatin and soy–alginate sponges was high(a porosity of approximately >70%). The achieved blendstructures combine suitable porosity for skin tissue engi-neering applications with large pore size (approximatelyin the range of 100–300 mm). However, the porous struc-ture was not observed for the soy–pectin blend, whichpresented a different structure than all other soy naturalpolymer groups, with significantly lower porosities thanall other groups [Figure 7(e,f)].

Effect of the crosslinking agent. Figure 8 shows a cross-section of two types of crosslinking agents that were usedon soy–gelatin blend samples based on a 4:1 ratio [1 wt% EDC, Figure 8(a), and 1 wt % glyoxal, Figure 8(b)]. Poreorientation structure was found in the soy–gelatin blendsamples using 1% glyoxal as the crosslinking agent.

FIGURE 4. Protein content (%) in supernatants over time of porous

soy blend samples based on a 4:1 soy protein:natural polymer ratio

for: (a) soy–gelatin and (b) soy–alginate; , noncrosslinked; , cross-

linked using 1 wt % EDC; , crosslinked using 4 wt % EDC. Data rep-

resent the average 6 standard deviation of three independent soy

protein-based blend samples. [Color figure can be viewed in the

online issue, which is available at wileyonlinelibrary.com.]

FIGURE 5. Protein content (%) in supernatants over time for different

types of 4:1 porous soy protein:natural polymer blends crosslinked

with 1 wt % EDC: , pure soy protein porous structure; , soy–gela-

tin; , soy–alginate. Data represent the average 6 standard deviation

of three independent soy protein-based blend samples. [Color figure

can be viewed in the online issue, which is available at wileyonlineli-

brary.com.]





DISCUSSION

Blended soy protein-based materials that are chemicallycrosslinked with other natural polymers were developedand studied. Gelatin (protein) and alginate or pectin (poly-saccharides) were added and chemically crosslinked to soyprotein using the crosslinking agents EDC or glyoxal inorder to obtain a new porous three-dimensional networkthrough freeze-drying. After completing the first stage of thestudy, three formulations were selected in order to studytheir effects on materials characteristics with relation toskin tissue engineering applications.

Mechanical properties of the dense filmsThe average tensile strength and strain of normal humanskin are 7.7 MPa and 100%, respectively.22 Good mechanicalproperties are crucial for routine handling and functionalstability of biomedical products, especially for tissueengineered-porous construct, as it needs to provide an ini-tial biomechanical profile for the cells before new tissue canbe formed. Performance under tensile can affect the ulti-mate function of the product in vivo. Therefore, in the firststep of the study, we compared the mechanical properties(ultimate tensile yield and maximal strain) of a series ofdense soy protein blend films to those of the soy proteinreference film in order to choose the best secondarycomponents.

The addition of HMW gelatin and alginate to soy proteinfilms at a 4:1 soy:gelatin or soy:alginate ratio showedimprovement in tensile strength compared with the refer-ence films. This can be explained by two possible factors:(i) the formation of physical and chemical interactions suchas hydrogen bonds or other intermolecular attractions, thatis, van der Waals; (ii) chain entanglement between thepolymers.

Soy and gelatin have good compatibility, and both areproteins. In the soy–alginate blends, there is an increasein stiffness and brittleness as result of the presence ofsugar rings in the polysaccharide chains that limit theflow and sliding movement of polymeric chains over eachother. The addition of synthetic polymers such as PVA andPEG at the same ratio decreased the tensile strength. This

may be due to a typical plasticizing effect, which causesweakening of the blend film. The plasticizer can interposeitself between the polymer chains and decrease the forcesholding the chains together.23,24 Furthermore, it can bespeculated that the synthetic polymers PEG and PVA mightbe less soluble, thus causing a decrease in the ability ofsoy protein to interact with polymer chains, which makethem less compatible with soy protein, and they might beeasily expelled from the blend mixture.25 It is also possi-ble that they can crystalize, which could also cause phaseseparation.

Physical properties. Successful functional biomaterialsshould have adequate physical properties to be able toperform in a moist environment. For example, an idealscaffold for wound healing must provide a control of fluidbalance over time in order to enable accelerated skinregeneration and should absorb body fluid for transfer ofcell nutrients and metabolites through the material. Twoparameters must, therefore, be determined: the WVTRthrough the material and the water uptake ability. Thiswas tested in order to elucidate the affinity of the porousstructures to water and the effect of the addition of a sec-ond component and crosslinking on the structure’s behav-ior in an aqueous environment. The degradation behaviorof the porous soy blend in aqueous solutions was alsostudied.

Water vapor transmission rate. An effective wound dress-ing provides good WVTR management that retains a moistwound bed at the desired levels for the healing course. Anexcessive WVTR may lead to wound dehydration, whereas alow WVTR might lead to maceration and bacterial contami-nation.26 An ideal dressing or graft would control the lossof water from the skin at an optimal rate. It has beenclaimed that burn wound dressings should ideally possess aWVTR in the range of 2000–2500 g/(m2 d).27

The WVTR is directly proportional to the degree ofcrosslinking and is related to the structural properties

TABLE IV. Effect of the Soy Protein:Alginate Ratio on Degradation Time of Crosslinked Soy–Alginate With 1 wt % EDC,

Determined by Qualitative Robustness Observations at Various Time Points

Soy Protein:Alginate(Weight Ratio) 3 Days 7 Days 14 Days 21 Days

8:1 Dissolved – – –4:1 Stable Stable Stable Softened2:1 Stable Stable Stable Stable

TABLE V. Effect of Crosslinking agent on Degradation of Crosslinked Soy–Gelatin 4:1, Determined by Qualitative Robustness

Observations at Various Time Points

CrosslinkingAgent (1 wt %) 3 Days 7 Days 14 Days 21 Days

Glyoxal Stable Dissolved – –EDC Stable Stable Softened Dissolved



(thickness, porosity) and to the hydrophilicity of thematerial.28–30

Our results show that all tested porous noncrosslinkedsoy blends demonstrate a WVTR in the range of 2500–2700 g/(m2 d) which is sufficient to prevent exudate accu-mulation and wound dehydration. These relatively high val-ues, which match the demands for treating burns, areattributed to the hydrophilic nature of the protein polymerand polysaccharide used. Our results show that the WVTRpractically does not change due to crosslinking for pure soyprotein and soy–gelatin groups. This is probably related tothe relatively low crosslinking density in these two cases.When the degree of crosslinking is low, porosity is high,allowing fast passage of water vapor through the material.With a high degree of crosslinking the structure is dense,which may limit the passage of water vapor through thematerial. Only crosslinking of soy–alginate substantially

increased its water vapor resistance. Our WVTR results sug-gest that soy–alginate blends are more effectively cross-linked with EDC than soy–gelatin and pure soy protein.These results are in agreement with the degradation behav-ior study, which is presented later. The WVTR of our soy-based scaffolds is similar to the WVTR of commerciallyavailable skin graft.31

Water uptake. Functional properties such as swelling andwater holding capacity are directly related to the manner inwhich the polymer interacts with water and are usually pro-portional to the crosslinking density. Water absorptionoccurs as the hydrophilic polymer gradually absorbs water,followed by chain relaxation. Controlled swelling behavior isof great importance for various applications in the field ofbiomedical materials because it enables prediction of thematerials’ behavior in the hydrated environment, which

FIGURE 6. ESEM fractograhs of: (a,b) pure soy protein, (c,d) 4:1 porous soy–gelatin blends and (e,f) 2:1 porous soy–gelatin blends. All cross-

linked using 1 wt % EDC.


exists in the human body. A scaffold that absorbs fluid toabout 80 times of its initial weight is suitable as skin tissueengineering.32 As expected, crosslinked soy–gelatin samplesenabled increased stability and decreased water absorption,compared with noncrosslinked soy–gelatin [Figure 3(b)].Chemical crosslinking of gelatin introduces covalent bondsinto the network, thereby decreasing the dissolution of thematrix. Soy–alginate had higher water uptake comparedwith soy–gelatin although not up to optimum value for skinregeneration scaffolds. These water adsorption results werenot completely in accordance with the degradation study. Itis usually expected that a higher crosslinking density (asexpressed by the degradation study) results in lower wateruptake during each of the stages, and enables reaching equi-librium, that is, constant water uptake, within a shorterperiod of time. An explanation for the deviating wateradsorption results might be that soy–alginate scaffolds

exhibit larger surface area due to the coarse porosity com-pared with soy–gelatin scaffolds (Figure 7).

The soy–gelatin crosslinked with glyoxal showedincreased water adsorption compared with soy–gelatincrosslinked with EDC, probably due to higher porosity(>80%) and surface area (Figure 8). The changes in thedimensions of the samples as swelling progressed to equi-librium are demonstrated in Table III. The dimensionalincreases were in accordance with the crosslinking density.A higher crosslinking density prevents expansion and swel-ling of the network. Soy–alginate samples demonstrated asignificant decrease in dimensional changes and in degrada-tion rate compared with both types of soy–gelatin samples.

In vitro degradation tests. An ideal tissue engineeringporous system requires that the material degrade at a rateat which cells can regenerate tissue to fill in pores. A

FIGURE 7. ESEM fractograhs of 4:1 soy:natural polymer blends: (a,b) soy–gelatin, (c,d) soy–alginate and (e,f) soy–pectin. All crosslinked using 1

wt % EDC.



degradation time of 25 days is suitable for healing acutewounds (burn and skin excision)33 or about 8 weeks forchronic wounds (diabetic ulcer, pressure ulcer) A morerapid or slower biodegradation rate would render it ineffec-tive as a wound closure device and would hinder the woundhealing process.34 In order to improve our new soy-basedmaterial’s stability toward biological, chemical, and physicaldegradations, we hoped to create synergy by combining twoapproaches: (1) blending soy protein with other naturalpolymers and (2) creating covalent bonds through chemicalcrosslinking. Two crosslinking agents suitable for proteinswere used in this study. The crosslinking agent EDC createsconjugation between -COOH groups (in polysaccharides andproteins) and amine groups (lysine) in proteins to createstable amide bonds and provides stability and resistance todegradation, while the reactive dialdehyde glyoxal may reactwith E-amine groups of lysine with the involvement of argi-nine residues, yielding advanced glycation or lipoxidationend products. It should be noted that glyoxal was used onlyfor the soy–gelatin blends in an attempt to improve protein-to-protein crosslinking.35,36

a. Quantification of noncovalent soluble proteinThe degradability behavior of the porous soy blend was

studied by quantifying the noncovalent soluble protein using10M urea. The soluble protein was analyzed by the Bradfordassay. The results indicate that a crosslinking reactionoccurred between selected polymeric components thatincrease stability toward degradation. As the concentration

of the crosslinking reagent increased, the material’s stabilityalso increased [Figure 4(a,b)]. The results also indicate thatsoy–alginate blends undergo a more effective crosslinkingreaction than soy–gelatin in the presence of EDC (Figure 5).This probably occurs as a result of numerous -COOH groupsthat are available for reaction in the linear polysaccharidealginate. It is known from the literature that carboxyl groupsare much more abundant in gelatin than amino groups, butthey appear to be less reactive in crosslinking reactions.37

The in vitro degradation behavior of the soy blend samplescould also be related to their corresponding crosslinkingdegree. It can be assumed from the results presented in Fig-ure 5 that soy–alginate blends are more crosslinked thansoy–gelatin blends and pure porous soy protein structures. Itis important to mention that soy protein is globular, so thatthese reactive groups may not be as accessible due to thefolding of the globular protein. Based on this result, it wasworth trying to crosslink soy–gelatin with glyoxal (protein-to-protein crosslinking).

b. Qualitative in vitro degradationThe effect of prolonged hydration on the degradability

of soy blend samples was evaluated qualitatively. Thiswas done in order to assess the ability of the hydratedsoy blend samples to hold structural stability, robustness,and integrity, which provide important information abouthandling the material in the hydrated state. Qualitativerobustness observations revealed that the higher alginateconcentration in the soy–alginate samples containing a2:1 soy:alginate ratio remained intact for longer (21

FIGURE 8. ESEM fractograhs of 4:1 soy:gelatin polymer blends:(a) crosslinked using 1 wt % EDC and (b) crosslinked using 1 wt % glyoxal.


days) than samples containing a 4:1 soy:alginate ratio(14 days) and containing an 8:1 ratio (3 days) (Table IV).This is due to the higher molar concentration of theCOOH groups in the polysaccharide alginate, which play akey role in protein crosslinking in the presence of EDC. Itshould be noted, however, that we selected 4:1 as thepreferred ratio since it provides satisfactory material sta-bility along with easy to use homogeneous mixtures. Theresults shown in Table V suggest that crosslinked soy–gelatin with EDC is more effective and resulted in adecreased degradation rate compared with samplescrosslinked with glyoxal, due to the fact that EDC reactsvia both lysine and COOH groups, whereas glyoxal reactsonly via lysine groups. This possibility needs to be exam-ined in order to confirm these results.

In conclusion, the physical properties of our soy protein-based blends demonstrated that crosslinking of soy–alginateblends is advantageous to the other studied blends due toits higher water vapor resistance and water absorption. Thesoy–alginate scaffolds also offer the advantage of relativelylow dimensional changes. Furthermore, these blendsremained intact and stable for longer than the soy–gelatingroup and pure soy protein and would thus be more suita-ble to acute wound healing.

MicrostructureEffect of the polymer concentration. Soy–gelatin blendscontaining 50 wt % gelatin (2:1 soy:gelatin ratio) showedinconsistency in pore size, with areas of large and areas ofsmall pores [Figure 6(e,f)]. When the gelatin concentrationwas increased, the soy–gelatin blend exhibited a high viscos-ity, which may result in nonuniform concentrations. Conse-quently, it favors the formation of protein aggregates, andeven phase separation, leading to a lower efficiency of net-work formation. Thus, by changing the gelatin concentra-tion, it became possible to tailor both the pore size and thepore distribution. In addition, as demonstrated in Figure 6,more defined pores were obtained with the increase in gela-tin concentration compared with the pure porous soy pro-tein structure [Figure 6(a,b)]. This was probably affected bythe formation of a more structured and more rigid gel net-work before freeze-drying (it is known that gelatin beginsto set at <158C).

Effect of the natural polymers. Three natural polymerswere used in the current study, and their effect on themicrostructure was studied. All three soy blends showed arelatively uniform porous structure at the 4:1 soy:naturalpolymer ratio. The porosity of the soy–gelatin and soy–alginate blends was high, which is typical for scaffolds fab-ricated using freeze-drying resulting in pore diameters inthe range of 100–300 mm [Figure 7(b,d)].38 The porestructure of the soy–alginate samples showed greater con-nectivity than that of the soy–gelatin samples. High poros-ity and interconnected porous networks are desirable topermit the attachment and ingrowth of cells and regenera-

tion of new tissue, by providing a component and a three-dimensional extracellular environment similar to that ofnative tissue, with a preferred pore size of 50–500mm.39,40 In comparison with other studies,41,42 our poroussoy-based scaffolds exhibited the most suitable pore sizerange for skin tissue regeneration. As expected, the natureof the interactions between the different secondary poly-mers and soy protein has a dramatic effect on the struc-tural behavior and had a significant effect on the size anddistribution of the pores.

Effect of the crosslinking agent. SEM micrographs [Figure8(b)] showed that the porous soy–gelatin structures cross-linked with glyoxal possess a well-defined orientated micro-structure. Water interactions with polymeric chains have asignificant effect on network organization. The template of thearranged ice crystal assembly was imprinted in the soy–gela-tin blends during the freeze-drying process. Our results on theeffect of glyoxal on the morphology of the soy–gelatin blendsare similar to the results obtained by Wu et al.,43 who usedglutaraldehyde (also a dialdehyde) as a crosslinking agent.

SUMMARY AND CONCLUSIONS

In the current study, we demonstrated that soy proteinblends with gelatin, alginate, and pectin can be assembledinto porous three-dimensional structures by combiningchemical crosslinking with freeze-drying. The results of thisstudy show that a range of physical properties and porousstructures can be achieved through crosslinking and modifi-cations achieved by adding natural polymers. The structureof the soy-based blend sponges was highly dependent onproperties of the secondary component and on the soy pro-tein:natural polymer ratios. The achieved blend structurescombine suitable porosity with large pore size and adequateinterconnectivity.

The preferred weight ratio of soy protein:natural poly-mer is 4:1, which presented a uniform solution and porousstructures. The soy–alginate crosslinked blends are advanta-geous compared with soy–gelatin blends and showed betterstability and degradation time along with controlled swel-ling behavior and lower water uptake, all obtained due tomore effective crosslinking.

The WVTR experiments showed that the porous blendstructures were all in the desired range for burn treatment[2000–2500 g/(m2 d)] and can be controlled by the cross-linking process.

We conclude that these novel porous three-dimensionalstructures exhibit physicochemical properties that appear tohave a potential for use as scaffolds in tissue engineeringapplications for skin regeneration. In addition, all cross-linked samples showed promising biocompatibility in thecell cytotoxicity study and remained intact in the presenceof human fibroblasts. These results will be presented in aseparate publication.

ACKNOWLEDGMENTS

We would like to thank SolbarTM (Ashdod, Israel) for kindlyproviding the SPI used in this study.



REFERENCES1. Sachlos E, Czernuszka JT. Making tissue engineering scaffolds

work. Review: The application of solid freeform fabrication tech-

nology to the production of tissue engineering scaffolds. Eur Cell

Mater 2003;5:39–40.

2. Santin M, Ambrosio L. Soybean-based biomaterials: Preparation,

properties and tissue regeneration potential. Expert Rev Med

Devices 2008;5:349–358.

3. Silva GA, Vaz CM, Coutinho OP, Cunha AM, Reis RL. In vitro deg-

radation and cytocompatibility evaluation of novel soy and

sodium caseinate-based membrane biomaterials. J Mater Sci

Mater Med 2003;14:1055–1066.

4. Cao N, Fu Y, He J. Preparation and physical properties of soy pro-

tein isolate and gelatin composite films. Food Hydrocolloids 2007;

21:1153–1162.

5. Santin M, Morris C, Standen G, Nicolais L, Ambrosio L. A new

class of bioactive and biodegradable soybean-based bone fillers.

Biomacromolecules 2007;8:2706–2711.

6. Snyders R, Shingel KI, Zabeida O, Roberge C, Faure MP, Martinu L,

Klemberg-Sapieha JE. Mechanical and microstructural properties of

hybrid poly (ethylene glycol)-soy protein hydrogels for wound

dressing applications. J Biomed Mater Res A 2007;83:88–97.

7. Peles Z, Zilberman M. Novel soy protein wound dressings with

controlled antibiotic release: Mechanical and physical properties.

Acta Biomater 2011;8:209–17.

8. Chen L, Remondetto G, Rouabhia M, Subirade M. Kinetics of the

breakdown of cross-linked soy protein films for drug delivery.

Biomaterials 2008;29:3750–3756.

9. Vaz CM, de Graaf LA, Reis RL, Cunha AM. In vitro degradation

behaviour of biodegradable soy plastics: Effects of crosslinking

with glyoxal and thermal treatment. Polym Degrad Stabil 2003;81:

65–74.

10. Reddy N, Yang Y. Potential of plant proteins for medical applica-

tions. Trends Biotechnol 2011;29:490–498.

11. Ghorpade VM, Li H, Gennadios A, Hanna MA. Chemically modi-

fied soy protein films. Trans ASAE 1995;38:1805–1808.

12. Chabba S, Netravali AN. “Green” composites part 1: Characteriza-

tion of flax fabric and glutaraldehyde modified soy protein con-

centrate composites. J Mater Sci 2005;40:6263–6273.

13. Song F, Tang DL, Wang XL, Wang YZ. Biodegradable soy protein

isolate-based materials: A review. Biomacromolecules 2011;12:

3369–3380.

14. Rhim JW, Lee JH. Effect of CaCl2,treatment on mechanical and

moisture barrier properties of sodium alginate and soy protein-

based films. Food Sci Biotechnol 2004;13:728–732.

15. Vaz CM, Fossen M, Van Tuil RF, De Graaf LA, Reis RL, Cunha AM.

Casein and soybean protein based thermoplastics and compo-

sites as alternative biodegradable polymers for biomedical appli-

cations. J Biomed Mater Res A 2003;65:60–70.

16. Wu RL, Wang XL, Wang YZ, Bian XC, Li F. Cellulose/soy protein

isolate blend films prepared via room-temperature ionic liquid.

Indus Eng Chem Res 2009;48:7132–7136

17. Zheng H, Tan Z, Ran Zhan Y, Huang J. Morphology and proper-

ties of soy protein plastics modified with chitin. J Appl Polym Sci

2003;90:3676–3682.

18. Silva SS, Santos MI, Coutinho OP, Mano JF, Reis RL. Physical

properties and biocompatibility of chitosan/soy blended mem-

branes. J Mater Sci Mater Med 2005;16:575–579.

19. Were L, Hettiarachchy NS, Coleman M. Properties of cysteine-

added soy protein-wheat gluten films. J Food Sci 1999;64:514–518.

20. Sabato SF, Ouattara B, Yu H, D’Aprano G, Le Tien C, Mateescu

MA, Lacroix M. Mechanical and barrier properties of cross-linked

soy and whey protein based films. J Agricult Food Chem 2001;49:

1397–1403.

21. Peles Z, Binderman I, Berdicevsky I, Zilberman M. Soy protein

films for wound healing applications: antibiotic release, bacterial

inhibition and cellular response. J Tissue Eng Regen Med 2012;

49:405–412.

22. Yamada H. Strength of Biological Materials. Baltimore, MD: Wil-

liams & Wilkins 1970.

23. Mo X, Sun X. Plasticization of soy protein polymer by polyol-

based plasticizers. J Am Oil Chemists Soc 2002;79:197–202.

24. Jia D, Fang Y, Yao K. Water vapor barrier and mechanical proper-

ties of konjac glucomannan–chitosan–soy protein isolate edible

films. Food Bioprod Process 2009;87:7–10.

25. Turhan K, Sahbaz F, G€uner A. A spectrophotometric study of

hydrogen bonding in methylcellulose-based edible films plasti-

cized by polyethylene glycol. J Food Sci 2001;66:59–62.

26. Boateng JS, Matthews KH, Stevens HN, Eccleston GM. Wound

healing dressings and drug delivery systems: A review.

J Pharmaceut Sci 2008;97:2892–2923.

27. Queen D, Gaylor J, Evans J, Courtney J, Reid W. The preclinical

evaluation of the water vapour transmission rate through burn

wound dressings. Biomaterials 1987;8:367–371.

28. Kester J, Fennema O. Edible films and coatings: A review. Food

Technol 1986;40:47–59.

29. Pascat B. Study of Some Factors Affecting Permeability. London:

Elsevier Applied Science Publishers; 1986. p. 7–24.

30. McHugh TH, Avena-Bustillos R, Krochta J. Hydrophilic edible

films: Modified procedure for water vapor permeability and expla-

nation of thickness effects. J Food Sci 1993;58:899–903.

31. Wu P, Fisher A, Foo P, Queen D, Gaylor J. In vitro assessment of

water vapour transmission of synthetic wound dressings. Bioma-

terials 1995;16:171–175.

32. Ma L, Gao C, Mao Z, Zhou J, Shen J, Hu X, et al. Collagen/chito-

san porous scaffolds with improved biostability for skin tissue

engineering. Biomaterials 2003;24:4833–4841.

33. Yannas I, Burke JF. Design of an artificial skin. I. Basic design

principles. J Biomed Mater Res 1980;14:65–81.

34. Galassi G, Brun P, Radice M, Cortivo R, Zanon GF, Genovese P,

Abatangelo G. In vitro reconstructed dermis implanted in human

wounds: Degradation studies of the HA-based supporting scaf-

fold. Biomaterials 2000;21:2183–2191.

35. Marqui�e C. Chemical reactions in cottonseed protein cross-linking

by formaldehyde, glutaraldehyde, and glyoxal for the formation

of protein films with enhanced mechanical properties. J Agricult

Food Chem 2001;49:4676–4681.

36. Odani H, Shinzato T, Usami J, Matsumoto Y, Brinkmann Frye E,

Baynes JW, et al. Imidazolium crosslinks derived from reaction of

lysine with glyoxal and methylglyoxal are increased in serum pro-

teins of uremic patients: evidence for increased oxidative stress

in uremia. FEBS Lett 1998;427:381–385.

37. Achet D, He X. Determination of the renaturation level in gelatin

films. Polymer 1995;36:787–791.

38. Hutmacher DW. Scaffolds in tissue engineering bone and carti-

lage. Biomaterials 2000;21:2529–2543.

39. Katoh K, Tanabe T, Yamauchi K. Novel approach to fabricate kera-

tin sponge scaffolds with controlled pore size and porosity. Bio-

materials 2004;25:4255–4262.

40. Thomson RC, Shung AK, Yaszemski M, Mikos AG. Polymer scaf-

fold processing. Principles Tissue Eng 2000;2:251–262.

41. Silva S, Oliveira JM, Mano JF, Reis RL. Physicochemical charac-

terization of novel chitosan-soy protein/TEOS porous hybrids for

tissue engineering applications. Mater Sci Forum 2006;514–516:

1000–1004.

42. Ma J, Wang H, He B, Chen J. A preliminary in vitro study on the

fabrication and tissue engineering applications of a novel chito-

san bilayer material as a scaffold of human neofetal dermal fibro-

blasts. Biomaterials 2001;22:331–336.

43. Wu X, Liu Y, Li X, Wen P, Zhang Y, Long Y, et al. Preparation of

aligned porous gelatin scaffolds by unidirectional freeze-drying

method. Acta Biomater 2010;6:1167–1177.


Novel porous soy protein‐based blend structures for ...meitalz/Articles/B22.pdf · Preparation of...

Documents

Transcript of Novel porous soy protein‐based blend structures for ...meitalz/Articles/B22.pdf · Preparation of...