Biochimica et Biophysica Acta 1830 (2013) 4030–4039
Contents lists available at SciVerse ScienceDirect
Biochimica et Biophysica Acta
j ourna l homepage: www.e lsev ie r .com/ locate /bbagen
Rejoining of cut wounds by engineered gelatin–keratin glue
S. Thirupathi Kumara Raja a, T. Thiruselvi a, G. Sailakshmi a, S. Ganesh b, A. Gnanamani a,⁎a Microbiology Division, CSIR-CLRI, Adyar, Chennai, Tamil Nadu, Indiab Bio-organic Division, CSIR-CLRI, Adyar, Chennai, Tamil Nadu, India
Background: Rejoining of cut tissue ends of a critical site challenges clinicians. The toxicity, antigenicity, low ad-hesive strength, flexibility, swelling and cost of the currently employed glue demands an alternative. Engineeredgelatin–keratin glue (EGK-glue) described in the present study was found to be suitable for wet tissue approxi-mation.Methods: EGK-gluewas prepared by engineering gelatinwith caffeic acid using EDC and conjugatingwith keratinbyperiodate oxidation. UV–visible, 1HNMR and circular dichroismanalyses followed by experiments on gelationtime, rheology, gel adhesive strength (in vitro), wet tissue approximation (in vivo), H&E staining of tissue sec-tions at scheduled time intervals and tensile strength of the healed skin were carried out to assess the effective-ness of the EGK-glue in comparison with fibrin glue and cyanoacrylate.Results: Results of UV–visible, NMR and CD analyses confirmed the functionalization and secondary structuralchanges. Increasing concentration of keratin reduces the gelation time (b15 s). Lap-shear test demonstratesthe maximum adhesive strength of 16.6 ± 1.2 kPa. Results of hemocompatibility and cytocompatibility studies
suggested the suitability of the glue for clinical applications. Tissue approximation property assessed using theincisionwoundmodel (Wistar strain) in comparisonwith cyanoacrylate andfibrin glue suggested, that EGK-glueexplicitly accelerates the rejoining of tissue with a 1.86 fold increase in skin tensile strength after healing.Conclusions: Imparting quinone moiety to gelatin–keratin conjugates through caffeic acid and a weaker oxidiz-ing agent provides an adhesive glue with appreciable strength, and hemocompatible, cytocompatible and bio-degradable properties, which, rejoin the cut tissue ends effectively.General significance: EGK-glue obtained in the present study finds wide biomedical/clinical applications.
Adhesives in modern surgical practices are unavoidable and they aremajorly used as an adjunct in several surgical procedures (tissue approx-imation, wound closure/sealant, grafting and hemostatic procedures)[1,2]. Tissue approximation using glues/sealants of both natural andsynthetic origin are currently in use. Cyanoacrylate has long been recog-nized as a performing synthetic adhesive for external use. Despite theenormous adhesive strength, release of toxic components, insufficientflexibility and lipid membrane damage prevents cyanoacrylate from in-ternal applications [3,4]. Although adhesives of biological origin (fibringlue, gelatin resorcinol formaldehyde, glutaraldehyde stabilized bovineserum albumin) were found helpful, the biocompatibility, low adhesivestrength, immune response and potential infection restricts the clinicalapplications and thus demands high strength, non-toxic and safe agentsfor tissue approximation. Since, the internal woundmilieu is completelydifferent from the external; agentswith additional characteristic features(high adhesive strength and compatible with wet tissue surface) arehighly required. Hence, approaches were made to mimic the natural
ani).
l rights reserved.
adhesives (adhesives of sessile organisms, mussels, and sandcastle)[5–8]which should also ensure that the demands on efficacy, safety, use-fulness, ease of preparation, cost and regulatory approvability in additionto different surgical applications could be met. It has been found thatthese organisms exhibit adhesive properties because of the presence ofdiversified functional groups (–NH2, –SH, –COOH, –OH, Aromatic-OH)of proteins present at the interface of glue and substrate. Thus, proteinswith a wide variety of functional groups are well suited for the fabrica-tion of adhesives. However, the utility of many proteins is limited dueto inconsistencies in composition, performance, limited supply andhigh cost compared to synthetic adhesives. Despite this, engineeringthe proteins for tissue approximation for both external and internal ap-plications, especially in wet tissue surface is still going on.
Gelatin, a denatured product of thematrix protein collagenwaswide-ly explored in the field of food and biomedical technology [9]. Theresorbable, low immunogenicity nature of this protein is used to preparescaffolds, wound dressing material and hydrogels. Gelatin/collagen haslong been used to prepare surgical adhesives however, without acrosslinker gelatin alone cannot be used as tissue glue [10]. But, in thepresence of enzymes (transglutaminase) and chemical crosslinkers(EDC/NHS, glutaraldehyde, resorcinol/formaldehyde, genepin) gelatincan be transformed into an adhesive product, however, the major
4031S. Thirupathi Kumara Raja et al. / Biochimica et Biophysica Acta 1830 (2013) 4030–4039
disadvantages are the presence of aldehydes and byproducts, low adhe-sive strength, flexibility and swelling.
Caffeic acid, is a bi-functional plant phenolic, widely present in veg-etables, fruits, coffee, tea, olive oil, red wine etc. The presence of di-hydroxyl groups and α,β unsaturated carboxylic group of caffeic acidimparts several functions like wound healing, antimicrobial, antioxi-dant, anti-inflammatory, antitumor, antianxiety, antimetastatic andMMP 2,9 inhibitor activities [11–13]. Most importantly, caffeic acidhas a similar structural and functional resemblance to DOPA, a potentialcandidate for imparting adhesive strength and curing rate tomussel ad-hesives [14].
Keratin is a fibrous structural protein providing outer coverings suchas hair,wool, feathers, nail, and the horns ofmammals, reptiles and birds.The disulfide bonds in keratin contribute strength to the hair. It has beenemployed as a biomaterial in the form of 2D and 3D scaffolds andhydrogels [15–17]. Keratin and gelatin blended films are also used toprepare 2D film and skin graft mesh to treat wounds [18,19]. However,for the preparation of surgical glues no reports are available on the useof keratin alone or in combinationwith natural proteins/polysaccharides.
In the present study, an attemptwasmade to prepare a protein basedadhesive with high strength and compatible for wet tissue approxima-tion. In brief, both gelatin and keratin proteins were interacted via (i)functionalization using bi-functional phenolics, followed by (ii) oxida-tion (either chemical/enzymatic), and the resultant glue like material,hereby named as EGK-glue (engineered gelatin–keratin-glue) wassubjected to wet tissue approximation using both in vitro and in vivomodels. Mechanical property (rheology and tensile strength), in vitrobiodegradability and biocompatibility (cyto and hemo-compatibility)were also assessed in addition to 1H NMR and circular dichroism analy-ses to demonstrate functionalization and interaction.
2. Materials and methods
2.1. Materials
Gelatin, type A (300 bloom strength), EDC (N-(3-di-methylaminopropyl)-N′-ethylcarbodiimide hydrochloride), NHS (N-hydroxysuccinimide), caffeic acid and collagenase, type I (Clostrid-ium histolyticum), were purchased from Sigma Aldrich, USA. MES(2-(N-morpholino)ethane sulfonic acid), and HEPES (N-2-hydroxyethylpiperazine-N′-2-ethane-sulfonic acid) were purchasedfrom HiMedia, India. All other chemical reagents were of analyt-ical grade and commercially available. Periodate was obtainedfrom Sd Fine Chem, India. MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromide)was purchased from Sigma, USA. Ke-tamine hydrochloride was obtained from Neon Laboratories, Mumbai,India. Fibrin glue was obtained from Reliance Life Sciences, Mumbai,India, in the form of Reliseal and cyanoacrylate was obtained fromEthicon Inc. in the form of Dermabond.
2.2. Preparation of EGK-glue
2.2.1. Functionalization of gelatin using caffeic acidCaffeic acid functionalized gelatin was prepared using EDC–NHS
[20]. In brief, caffeic acid (0.1 M) dissolved in 0.1 M MES buffer(pH 5.5) was treated with EDC (0.2 M) and NHS (0.2 M) and stirredfor 45 min at 25 °C and then mixed with gelatin dissolved in HEPESbuffer (pH 7.0). The resulting solution was extensively dialyzed against5 mMHCl and then against the samemedium containing 1% NaCl. Finaldialysis was made against 0.1 M phosphate buffer (pH 6.5) and the di-alyzed samples were freeze dried, stored at 4 °C and hereby named asfunctionalized gelatin.
Proton NMR spectra of gelatin, functionalized gelatin and caffeicacid was recorded at an operating frequency of 500 MHz in a JEOLECA-500 FT NMR spectrometer. Ten milligrams of gelatin, function-alized gelatin and 1 mg of caffeic acid was dissolved in 750 μl D2O
individually and 1H-spectrawas recorded by keeping Tetramethylsilane(0.01%) as an internal reference and at 45 °C (for better solubility)with-out suppressing the water signal. The acquisition parameters for (1H)NMR were as follows: 32 scans for signal to noise averaging, relaxationtime of 5 s, sweep width of 20 ppm, number of data points was 16,384and 45 pulse width is 6.5 μs.
UV–visible spectrum for gelatin, functionalized gelatin and caffeicacid was recorded using UV-2450 (Shimadzu, Japan). In brief, caffeicacid (1 μg), gelatin and functionalized gelatin (1 mg) were dissolvedseparately in 1 ml of phosphate buffer (pH 6.5). The spectra wererecorded in the wavelength region of 200 to 600 nm by keeping thebuffer as a blank.
2.2.2. Keratiene (reduced form of keratin) preparationHuman hair keratin was extracted according to the Shindai meth-
od [21]. In brief hair was obtained from a local hair salon, intensivelywashed with water containing 0.5% SDS, rinsed with fresh water andair-dried and then treated with chloroform and hexane to remove theexternal lipids and dried at 50 °C. About 10 g of hair was cut intosmall pieces and mixed with the 250 ml of extraction mediumcontaining 10 M urea, 5% SDS, 5% 2-mercaptoethanol. The mixturewas kept at 50 °C for 24 h and centrifuged at 5000 rpm for 10 min.The supernatant was dialyzed against water, until the keratin solutionwas free from mercaptoethanol. The dialyzed solution was concen-trated using an ultraconcentrator using 10 kDa cut off membrane.The resulting thiol functionalized keratin (keratiene) solution wasused for conjugation with functionalized gelatin.
2.2.3. Conjugation of caffeic acid functionalized gelatin andthiol-functionalized keratin
Different percentage weight ratios, 10:0, 10:2, 10:3, 10:4, 10:5 and10:6 respectively to functionalized gelatin and keratin in 0.1 M phos-phate buffer (pH 7.5) were taken and then treated individually with1.0% sodium meta periodate. The time taken by the proteins to trans-form from solution to gel state (by inverting the reaction tube) wasmeasured as curing (gelation) time. The resulting product was namedas EGK-glue.
2.3. Surface morphology of EGK-glue
The surface morphology of the EGK-glue was studied using a scan-ning electron microscopy (SEM). The EGK-glue sample was freeze-dried and the cross sections of the freeze-dried samples were placedon the carbon ribbon and gold coated. The cross sectional view was ob-served and captured at different magnifications using a HITACHI-S3400N SEM operated at 5 kV.
2.4. Circular dichroism
All circular dichroism (CD)measurements were performed using aJasco J715 spectropolarimeter at room temperature, using a circularquartz cell with a path-length of 0.1 cm. The instrument was calibrat-ed with ammonium D-camphor-10-sulfonate as described by the in-strument manufacturer. All CD spectra were measured between 190and 300 nm with a scanning speed of 100 nm/min. The bandwidth,response time and data pitch were set to 1 nm, 1 s and 0.5 nm, re-spectively. All CD spectra represent the average of three individualscans and all the spectra were solvent subtracted.
2.5. Rheological study
Rheological analysis was carried out using an oscillatory rheome-ter (Anton Paar Rheometer MCR-301) with a cone and plate geome-try of 1°, 25 mm diameter with 0.1 mm gap. The experiments werecarried out at 37 °C using 1% strain and the frequency was variedfrom 0.1 to 100 rad/s. The change in viscoelasticity was recorded as
4032 S. Thirupathi Kumara Raja et al. / Biochimica et Biophysica Acta 1830 (2013) 4030–4039
storage (G′) and loss modulus (G″). In brief, the experiments werecarried out by oxidizing different weight ratios of functionalized gel-atin and functionalized keratin mixtures and then cast on to a Petridish and allowed to cure for 60 min. The cured gel was carefully re-moved and placed in the rheometer stage. After adjusting the gap,the excess sample was removed or wiped out using tissue paper.
2.6. Determination of adhesive strength of EGK-glue: lap shear test
The chicken skin model was used to determine the adhesive bond-ing strength of the fabricated adhesive product (ASTM F2255-03).Skin was obtained from the local slaughtering house and washed to re-move the blood. Followed by washing, skin samples were cut into a di-mension of 3 × 2 cm (L × B) and subjected to lap shear test. In brief,two skin samples with equal dimensions were taken and to one endof the dermal side, EGK-glue was applied and overlapped with the sec-ond piece of the skin with the same geometry. After giving a load of100 g/cm2 for 60 min, the test skin samples were subjected to a lapshear test in an Instron instrument with a 100 N load cell and a cross-head speed of 10 mm/min at 37 °C. The force required to detach thetwo different surfaces has been considered as adhesive strength.
2.7. Biodegradation studies
Ten milligrams of oven dried samples of EGK-glue was exposed toan enzyme buffer at pH 7.5 containing 50 mM Tricine, 10 mM calci-um chloride and 400 mM sodium chloride. Collagenase enzyme wasadded (10 U/ml) and kept for incubation at 37 °C. The prolonged re-lease of amino acid upon degradation was analyzed using TNBS assay[22] for a period of 96 h.
2.8. Biocompatible studies
To study the biocompatibility of EGK-glue, both hemo andcytocompatibility assays were performed. Hemocompatibility of thematerials was studied by both direct and indirect contact assays. Inthe indirect method, EGK-glue sample was disinfected by immersingin 70% alcohol [23] and rinsing several times with sterile PBS and thenincubated at 25 °C for the period of 5 and 10 days in static condition.At scheduled time intervals, the PBS extract was recovered and treatedwith 1 ml of RBC cells for 1 h at room temperature. The direct methodinvolves, exposing 1 ml of blood to the EGK-glue material for 1 h atroom temperature. Followed by incubation, the samples were cen-trifuged at 3000 rpm for 5 min. The hemolytic potential of the materialwas recorded bymeasuring the optical density (OD) of the supernatantat 540 nm. The percentage hemolysiswas calculated from the followingequation: {(OD of positive control − OD of sample) / OD of positivecontrol} × 100. According to the percentage of hemolysis the scoreswere given as hemolytic (>5%), slightly hemolytic (2–5%) andnonhemolytic (b2%). Plain water and PBS were taken as a positive anda negative control respectively.
Cytocompatibility of EGK-glue was studied using NIH-3T3 fibroblastcells [23]. In brief, EGK-glue sample (1 g ± 0.1 g) was incubatedwith10 ml of sterile RPMI medium without FBS (fetal bovine serum)for the scheduled period of 5 and 10 days. Followed by incubation, thesamples were centrifuged and the supernatant with 10% FBS was usedfor culturing thefibroblast cell. Cell culture plates (96well)were seededwith 1 × 104 cells per well with the prepared medium and incubatedfor 24 and 48 h at 37 °C with 5% CO2. After incubation, the supernatantof each well was replaced with MTT diluted in serum-free medium andthe plates incubated at 37 °C for 4 h. After removing the MTT solution,dimethyl sulfoxide was added to each well and aspirated to dissolveall of the crystals and then left at room temperature for a few minutesto ensure the solubility of the crystals. Finally, absorbance was mea-sured at 570 nm using UV plate reader (Epoch, BioTek). The resultswere compared with cells treated with RPMI medium without the
EGK-glue extract. Biocompatible studies for both caffeic acid andperiodate alone has not been performed as described in detail in theResults and discussion section.
2.9. In vivo wet tissue approximation study: incision wound model
Wet tissue approximation property of EGK-glue was assessed usingan incision wound model followed by approval from the ethical com-mittee, vide no. 466/01a/CPCSEA – IAEC No. 08/02/2011b. Thirty sixmale albino (Wistar strain) rats with an average weight of 275 ± 5 gwere segregated into four experimental groups as detailed below instandard animal cages and fed with pelletized feed and surplus water.
Group I untreated animals (control);Group II animals treated with cyanoacrylate (positive control I);Group III animals treated with EGK-glue;Group IV animals treated with fibrin glue (positive control II).
During experiments, animals were anesthetized with ketamine(100 mg/kg body weight) intraperitoneally. The surgical area wasshaved and sterilized with 70% ethanol. For all the rats a paravertebralincision (3 cm long and 0.3–0.4 cm depth) was made on the skinusing standard surgical blades. Except for the untreated group, eachwound area was exposed to chosen tissue approximation agents,which ultimately closed the wound surface. Wound samples were col-lected on days 4, 10 and 20 of the experimental period. On the day ofsampling, the test animals were euthanized by cervical dislocationand the pelt (6 × 6 cm) was collected for mechanical strength mea-surement and histological studies. To measure the tensile strength ofthe rejoined wound, skin section was cut (4 × 1 cm) perpendicular tothe wound surface. The skin sections were loaded on an Instron instru-ment of a 100 N load cell, and a crosshead speed of 10 mm/min and themaximum force required to break the rejoined wound was calculated.For histological studies the skin sections were stored at 10% bufferedformalin for 1 week. The tissue sections were processed for hematoxy-lin and eosin staining (H&E) to access the histomorphological analysis.Images were captured using a Nikon Eclipse 80i microscope and the re-duction in wound area was calculated using Nikon NIS element D-3.2software. To measure the collagen synthesis, picrosirus red stainingwas performed and the images were captured using a microscope.The density of collagen produced is directly proportional to the in-tensity of red stain produced and measured using Adobe Photoshopsoftware. The percentage rise in collagen productionwas compared rel-atively with the healthy skin.
3. Results and discussions
3.1. Preparation of EGK-glue
3.1.1. Functionalization of gelatinIn the first step, the –COOH of caffeic acid was activated with water
soluble carbodiimide and stabilized with NHS, according to Kuijpers[24] and interacted with gelatin via ε-NH2 group of a lysine residue. Toconfirm the modification, 1H NMR spectrum was recorded for caffeicacid, gelatin and functionalized gelatin. 1H NMR spectrum of caffeicacid displayed chemical shifts at δ 6.34 (d, J = 16 Hz, H7) and 7.42 (d,J = 16 Hz, Hα) corresponds to α,β-unsaturation (Fig. 1a). The aromaticchemical shift values were given as, δ 6.65 (d, J = 8 Hz, H5), δ 7.09 (dd,J = 2, J = 1.5 Hz, H6), and δ 7.17 (d, J = 1.5 Hz, H2). The 1HNMR spec-trum of gelatin, and the functionalized gelatin displayed a broad chemi-cal shift from 7.2 to 7.4 corresponding to residual aromatic amino acidpresent in theprotein backbone. However, functionalized gelatin has sig-nificant chemical shift values at 6.3, 6.9 and 7.1 ppm similar to caffeicacid.
The UV–visible spectral analysis of gelatin, caffeic acid and function-alized gelatin showed the less absorbance value at 280 nm for gelatin
Fig. 1. Proton NMR and UV–visible spectroscopic confirmation of engineered gelatin.
(a) 1H NMR spectra of caffeic acid, gelatin and functionalized gelatin.(b) UV–visible spectra of caffeic acid, gelatin, and functionalized gelatin.
4033S. Thirupathi Kumara Raja et al. / Biochimica et Biophysica Acta 1830 (2013) 4030–4039
because of the presence of few aromatic amino acid residues and caffeicacid displayed the maximum absorbance at 290 and 315 nm (Fig. 1b).The functionalized gelatin displayed a new absorbance peak at290 nm and a shoulder at 315 nm.
3.1.2. Functionalization of keratinHuman hair is the cheapest source to extract keratin. In the pres-
ent study, a reduction method was followed to extract keratin fromhuman hair. During this process, the disulfide bridge of the keratinwas reduced to form free sulfhydryl (thiols) which will act as a strongnucleophile.
3.1.3. Preparation of EGK-glueAs said, there are different percentage weight ratios, viz., 10:0, 10:2,
10:3, 10:4, 10:5 and 10:6 respectively to caffeic acid functionalized gel-atin: functionalized keratin conjugated in the presence of periodate.Based on the curing time, optimization of the percentage weight ratio
was observed [25]. The addition of periodate oxidizes the vicinal diolsof caffeic acid and sulfhydryl groups of keratin and the generation ofquinone moiety mediates the interaction of functionalized gelatinwith keratin, resulting in the formation of a crosslinked polymer, withadhesive property.
The inter and intramolecular crosslinking between functionalizedgelatin and keratin may be facilitated through: Michael addition reac-tion of free sulfhydryl (–SH) and amines (NH2) group to caffeic acid,biaryl formation (coupling of semiquinone radicals), formation ofintermolecular keratin disulfide bridges and imine formation [26–28].During conjugation in the presence of periodate, phase transition (solu-tion to gel state) occurs, however, the time taken (curing time) for tran-sition varied significantly with respect to keratin concentration.Fig. 2(a) displays the tube inversion assay of the sample before andafter oxidation. Fig. 2(b) displayed the curing time for the experimentalsamples of different weight ratios. The curing time for functionalizedgelatin after oxidation with periodate was measured as 20 s and upon
Fig. 2. Phase transform, rate of curing and surface morphology of EGK-glue.
(a) Tube inversion assay for different ratios of functionalized gelatin and keratin.(b) Rate of curing of different weight ratios of functionalized gelatin and keratin(c) SEM image for the surface morphology of EGK-glue obtained by oxidizing a
10:4 percentage weight ratio of functionalized gelatin and keratin.
4034 S. Thirupathi Kumara Raja et al. / Biochimica et Biophysica Acta 1830 (2013) 4030–4039
addition of increasing weight ratios of keratiene, the curing time re-duced to 10 s for the samples of 10:4 and 10:5 weight ratios and at a10:6 ratio, the curing time increases (>15 s). The significant reductionin curing time in 10:4 and 10:5 ratios could be explained by the increasein the number of active reactive sites of keratiene (sulfhydryl groups ofcysteine, tryptophan, lysine, histidine groups) which facilitates quickgelling. The increase in curing time observed for the 10:6 ratio samplecould be explained by the non-availability of quinone moieties of func-tionalized gelatin which further interact with the higher concentrationof keratin.
In the present study, though, functionalized gelatin alone in the pres-ence of periodate transformed to a gel due to the effect of a few strategiccross-links between single peptide chains, however, addition of keratinwith more active crosslinking sites results in the formation of a networkwith the said interactions as summarized above within a short period oftime (b15 s). In general, the gelation mechanism of gelatin may beeffected through; (i) temperature changes [29]; (ii) presence of alde-hydes [30]; (iii) presence of enzymes like transglutaminase [31]; (iv)presence of EDC/NHS and (v) presence of genipin [32]. In all the saidmechanisms, except temperature, interactions were mediated throughthe –NH2 group of gelatin. However, with regard to caffeic acid function-alized gelatin, interaction may also be facilitated by the presence of anynucleophilic moieties.
It is well known that there is no possibility of interaction between –
NH2 and –SH groups of the chosen proteins, however, in the presentstudy, interaction between amino groups of gelatin and sulfhydrylgroups of keratiene was mediated by the introduction of phenolic moi-ety through caffeic acid and upon oxidation the formation of quinonemoiety initiates the stable interactions. When mimicking the adhesivesofMussel foot protein, quinonemoiety led the role in the preparation ofthe adhesive [6]. The diphenol moiety of a phenolic acid or polyphenolis readily oxidized to orthoquinone, either enzymatically or by molecu-lar oxygen and quinone forms a dimer in a side reaction, or reacts withamino or sulfhydryl side chains of polypeptides to form covalent C\N
or C\S bonds [33,34]. Crosslinking may take place when a second pro-teinmolecule reacts with the regenerated hydroquinone. Thesemecha-nisms have been described based on the observation of phenolic aciddimers and formation of gels and coacervates of gelatin. Dryhurst [35]observed, that thiol group interactions were more favorable forcrosslinking than NH2 groups when the reaction pH was maintainedat 6.5 to 8.5. This could be explained by the difference in the pKavalue of –SH (pKa—8.3) and –NH2 (pKa—10.3) groups.
3.2. Surface morphology of EGK-glue
Scanning electron microscopic images revealed the presence of awell defined three dimensional matrix with a uniform pore structureattributed to the intra and intermolecular crosslinking (Fig. 2(c)). Theporous structure facilitated the cell adherence, proliferation and col-lagen synthesis when applied to the wounded tissues as evidencedfrom the picrosirus red staining and H&E staining of tissue sectionsin the in vivo studies.
3.3. Circular dichroism
Circular dichroism (CD) analysis of the samples was recorded to fur-ther confirm the interactions at themolecular level. CD, is awell-knownbiophysical technique for the elucidation of secondary structures of pro-teins and peptides in solution [36]. In the present study, CD spectrum ofgelatin exhibits a single negative minimum at 198 nm (Fig. 3a (i)),which is characteristic of proteins with a predominantly unorderedstructure [37]. The CD spectrum also has a weak negative minimum at230 nm accompanied by a weak positive band at 222 nm. Tamburroet al. [38] suggested that the peptides adopting β-turn conformationshows a positive band at 220 nm accompanied by a low energy nega-tive band around 230 nm. It should be noted that peptides and proteinswith left-handed poly proline II type structures also exhibit a positiveband at 222 nm, but without the high energy band at 230 nm [39,40].Since gelatin is derived from the thermal denaturation of collagen,which has a PPII-type structure, the CD spectrum thus suggests that gel-atin adopts a predominantly unordered structure with residual β-turnand PPII-type structures. Alternatively, gelatin exists in unorderedstructures with few local stretches of PPII type structures. The CD spec-trum of caffeic acid functionalized gelatin displays an increase in the in-tensity of the negative band at 198 nm (Fig. 3a (i)), implying that thestructure of gelatin undergoes further denaturation which results in ahighly unordered structure compared to that of native gelatin. In thecase of functionalized keratin, the CD spectrum shows negative doubleminima at 206 and 222 nm (Fig. 3a(i)), which are characteristic of anα-helical structure. However, the intensities of both the bands are notequal which is a characteristic feature of proteins with α/β and α + βtype structures [41]. Thus, the present CD spectrum indicates that kera-tin has both α-helical and β-sheet conformations, which are in goodagreement with previous results.
It is clear from Fig. 3a(ii) that conformational changes are dependentupon the ratio between functionalized gelatin and functionalized kera-tin. For example, at a 10:2 ratio, the CD spectra did not show anychanges, implying that oxidation did not affect the conformation of acomplex. In contrast, at 10:4 and 10:6 (shown in different colors), a sig-nificant change in the intensity of the band at 198 nm suggested thatoxidation induces crosslinking between the proteins, which result inthe change of the secondary structures of the corresponding complexes.
3.4. Rheological analysis
Since all the above said instrumental analyses suggested and con-firmed the interactions between functionalized gelatin and keratin,consequently, the phase transition observed in the present study wasconfirmed by performing the rheological analysis. For all the said exper-imental ratios, rheological analysis was performed at 37 °C with a 1%
Fig. 3. Circular dichroism, rheological and adhesive strength of EGK-glue
a)(i) CD spectrum of gelatin, keratin and functionalized gelatin(ii) Comparison of CD spectra of functionalized gelatin and functionalized keratin at different ratios before and after periodate oxidation. The continuous lines represent
before the oxidation process and the discontinued line represents after oxidation of the samples. The black colored line represents 10:2 ratios, the red colored linerepresents 10:4 ratios; and the blue colored line represents 10:6 ratios.
b) Rheological spectra of functionalized gelatin, different ratios of functionalized gelatin and keratin oxidized samples (values are represented as storage modulus).c) Adhesive strength (measured from lap shear test) of different ratios of EGK-glue prepared using different ratios of functionalized gelatin and keratin upon oxidation.
4035S. Thirupathi Kumara Raja et al. / Biochimica et Biophysica Acta 1830 (2013) 4030–4039
strain (since 1% strain falls within the linear viscoelastic region) and byvarying the oscillatory frequency. The storage/elastic (G′) and loss/vis-cous moduli (G″) were calculated. Observations on rheological analysis(Fig. 3c) suggested that the G′ value for functionalized keratin alonewas less than G″ for all the frequencies tested which implies that func-tionalized keratin exits in a viscous state (results not shown). In contra-diction, functionalized gelatin showed G′ was less than G″ at lowerfrequencies, and at higher frequencies, it was greater than G″which, im-plies that at lower frequencies, functionalized gelatin behaves as a vis-cous liquid and as the frequency increases, it behaves like a gel. Fig. 3cillustrates a change in G′ values of the samples with 10:3, 10:4, 10:5and 10:6 ratios. Interestingly, for all the samples, G′ value was greaterthan G″ and suggested the gel nature of the sample. The significantchange in G′ was observed with 10:4 and 10:5 ratios compared toother ratios. Though the curing time for 10:6 ratio was comparativelyhigher than 10:4 and 10: ratios, rheological analysis revealed that italso behaves as a gel. These observations on storage and lossmoduli cor-relate well with the observationsmade during the gelation experiments.
3.5. Adhesive strength—in vitro assessment
Adhesive strength of the EGK-glue obtained in the present study wasexamined using a chicken skin model (Fig. 3d). Adhesive strength in-creases with an increase in the concentration of functionalized keratinup to 10:4 ratios, and an additional increase in keratin concentration re-duces the adhesive strength significantly. These observations correlatewell with the results observed fromgelation experiments. Themaximum
adhesive strength of 16.48 kPa was observed with 10:4 ratios and it wasonly 2.25 kPa for samples of 10:6 ratios. Adhesives prepared from fibringlue showed variable adhesive strength from 1 to 6 kPa [42–44] whentested using different skin tissues. Thus, it seems EGK-glue displayedcomparatively higher adhesive strength. Though the adhesive strengthof cyanoacrylate was 31% more than EGK-glue, the release of toxic com-ponents and flexibility may restrict its application.
With regard to adhesion, in general, two different forces, viz., ad-hesive and cohesive forces are operating. Cohesive force is the forceof attraction between the molecules present in the adhesives andthe adhesive force is between the adhesive product and the substrate.The adhesive strength in turn depends on the cohesive force [45]. Inthe present study, an increase in cohesive force between the conju-gated molecules could be the reason for the increase in adhesivestrength realized. As discussed above, crosslinking concentrations atoptimum ratios of functionalized gelatin and keratin via quinonemoi-ety increases the cohesive forces.
3.6. Biodegradability studies
Biodegradability of EGK-glue was assessed by exposing the prod-uct to collagenase enzyme for a period of 96 h and examining the re-lease of amino acids. After 24 h of incubation only meager levels (notdetectable) of amino acid release were observed in the buffer solu-tion. Prolonging the incubation time for 48 and 96 h shows a gradualincrease in the amino acid release content, which confirms the grad-ual degradation of EGK-glue.
4036 S. Thirupathi Kumara Raja et al. / Biochimica et Biophysica Acta 1830 (2013) 4030–4039
3.7. Biocompatibility assessment
Hemocompatibility studies carried out under both indirect and di-rect contact methods using RBCs revealed that only 1.0 ± 0.5% hemo-lysis was observed with the indirect method and 1.5 ± 0.5% with thedirect method implying the non-hemolytic property of EGK-glue.
Further, the results of the cytocompatibility studies demonstratedthat more than 90% of the cells were viable for the experimental pe-riod of 24 and 48 h. With respect to the cytotoxicity studies of caffeicacid alone, it has been understandable that during the process ofEDC–NHS mediated incorporation of caffeic acid to gelatin backbone,the crucial dialysis steps make the material free from unreactedcaffeic acid and the possibility of free caffeic acid is found nil.According to Koganov [46], caffeic acid is one of the widely studiedgroups of phenolics in the field of food and agriculture and it is oneof the least toxic compounds amongst various phenolics screenedfor cytotoxicity studies using fibroblast cell lines.
With regard to the toxicity of periodate, it was used as an oxidiz-ing agent to convert a hydroxyl form of caffeic acid to the respectivequinone, which initiates the polymerization reaction. There may bea cytotoxicity issue due to periodate, however, when mixed withengineered gelatin–keratin solution, the consumption of electrons
Wound
Day 0
Day 4
Day 8
Day 10
Day 20
Untreated Cyanoacrylate(a)
(b)
Fig. 4. In-vivo incision wound model on albino Wistar rats
a) Tissue approximation of incision wound for untreated samples, cyanoacrylate, EGK-glub) Tensile strength measurements (kPa) on healed skin with respective chosen samples a
during the oxidation process will be accompanied by the reductionof periodate ion to iodate ion, which will be further reduced into aniodide ion, a low toxic compound [47].
3.8. Wet tissue approximation—in vivo studies
Wet tissue approximation of skin tissue of the incision wounds ofall the experimental group animals were photographed during days0, 4, 8, 10 and 20 (Fig. 4a). It has been observed that, similar to fibringlue and cyanoacrylate, EGK-glue stick onto the wound area and alsoapproximate the cut ends smoothly. Till day 4, no significant differ-ence between the test sample and the positive controls was observed,however, day 8 images showed significant variations. On a further ex-tension of the period, EGK-glue treated cut ends did not have any in-cision mark similar to the cyanoacrylate treatment, whereas, a clearrejoined mark was observed with fibrin glue.
Assessment on tensile strength of the incision skin tissue on days4, 10 and 20 in comparison with healthy skin of the animal suggestedthat the tensile strength of the healthy skin was 5100 kPa, whereas,the healed skin tissue of the test samples showed a tensile strengthin the range between 875 and 1675 kPa. Among the test samples,the maximum tensile strength of 1675 ± 50 kPa was observed with
EGK-Glue Fibrin glue
e and fibrin glue samples with respect to days.nd healing days in comparison with healthy skin.
Fig. 5. Histo-morphological analysis of rat skin sections using hematoxylin and eosin staining and picrosirus red staining.
a) H&E staining of tissue sections with respect to healing days for the untreated, cyanoacrylate, EGK-glue and fibrin glue samples.b) Histogram on wound contraction area (measured in mm2) with respect to healing days for the untreated, cyanoacrylate, EGK-glue and fibrin glue samples.c) Picrosirus red staining of tissue sections displaying the amount of collagen synthesized upon tissue approximation with the chosen samples. The images were for the untreated,
cyanoacrylate, EGK-glue and fibrin glue samples with respect to healing days.d) Histogram (intensity of red color) of collagen deposition of the wounded skin compared to that of the healthy skin.
4037S. Thirupathi Kumara Raja et al. / Biochimica et Biophysica Acta 1830 (2013) 4030–4039
EGK-glue treated skin tissue, whereas, it was only 1250 ± 25 kPa forcyanoacrylate and 875 ± 20 kPa for fibrin glue treated skin tissues(Fig. 4b). The reason for the high tensile strength for EGK-glue treatedtissues could be due to the appreciable amount of deposition of colla-gen during the tissue approximation process as evidenced from thestaining studies.
Fig. 6. Overall schematic representation of the EGK-glue. The figure represents the functionized gelatin and keratin mixtures and the molecular orchestra of EGK-glue. The pendant cafnumber indicates the possible interactions through Michael's addition reaction of 1) free am(coupling of semiquinone radicals), 4) Schiff base formation and 5) periodate mediated ox
Similar to the observations on tissue approximation of the inci-sion wound upon treatment with the chosen adhesives, histologicalexamination of tissue sections also demonstrated that approximation ofcut ends starts from the dermis level and a significant reduction in thewound area was observed with EGK-glue applied treatments comparedto cyanoacrylate and fibrin glue on day 10 and day 20. Day 20 sections
alization of gelatin with caffeic acid using EDC/NHS chemistry, oxidation of functional-feic acid undergoes inter and intramolecular crosslinking of the protein and the circledines (NH2), 2) sulfhydryl (–SH) groups to the pendant caffeic acid, 3) biaryl formationidation of intermolecular keratin disulfide bridge formation.
4038 S. Thirupathi Kumara Raja et al. / Biochimica et Biophysica Acta 1830 (2013) 4030–4039
displayed more than 90% of restoration of skin morphology in EGK-gluetreatment. The reductions in the wound areas for all the experimentalgroupswas shown in Fig. 5a and b. Compared to cyanoacrylate and fibringlue, EGK-glue showed less wound area on day 20 with insignificant dif-ference. Furthermore, staining with picro-sirus red (Fig. 5c) indicates theamount of collagen deposited in the wound area. It has been observedthat an increased level of deposition of collagen in the tissue sectionstreated with EGK-glue compared to those treated with cyanoacrylateand fibrin glue (Fig. 5d) suggested the biocompatible nature ofEGK-glue allowing the cells to synthesize collagen,which, in turn acceler-ates the tissue approximation. The appreciable amount of collagen depo-sition during the tissue approximation process might be the reason forthe high tensile strength of EGK-glue treated tissues. Fig. 6 displayedthe schematic representation of the functionalization and conjugationprocess that took place in EGK-glue formation.
4. Conclusions
The present study emphasizes preparation, characterization and ap-plication of a protein based adhesive glue material which finds im-mense application in tissue approximation, wound healing andhemostasis. The glue named as EGK-glue was prepared by impartingquinone moiety to gelatin–keratin conjugates (in order to mimic mus-sel adhesive protein) through functionalizationwith caffeic acid andox-idizingwith aweaker oxidizing agent. Themaximum adhesive strengthof 16.6 kPa under in vitro condition and 1.67 MPa under in vivo condi-tion was observed for the EGK-glue. The wound healing efficacy of theglue was more than 30% compared to the conventional surgical glues.The intermolecular interactions between the conjugates impart athree dimensional network structure (SEM analysis) which facilitatesthe cell adherence and accelerates the healing. The biocompatibilitystudies demonstrated the suitability of the material for clinicalapplications.
Acknowledgements
The authors thank DBT, NewDelhi for funding this project through aTATA innovative Fellowship. The authors thankDr. Phani Kumar, Chem-ical Physics Department, CLRI, for his kind help in proton NMR spectros-copy. The first author thanks the Council of Scientific and IndustrialResearch (CSIR, New Delhi) for the financial support in the form of Se-nior Research Fellowship.
Appendix A. Supplementary Material
Results of experiments with enzymatic oxidation which weresummarized in the supplementary data associated with this articlecan be found, in the online version. Supplementary data to this arti-cle can be found online at http://dx.doi.org/10.1016/j.bbagen.2013.04.009.
References
[1] A. Cuschieri, Tissue adhesives in endosurgery, Surg. Innov. 8 (2001) 63–68.[2] M. Ryou, C.C. Thompson, Tissue adhesives: a review, Tech. Gastrointest. Endosc. 8
(2006) 33–37.[3] Y.-C. Tseng, Y. Tabata, S.-H. Hyon, Y. Ikada, In vitro toxicity test of 2-cyanoacrylate
polymers by cell culture method, J. Biomed. Mater. Res. 24 (1990) 1355–1367.[4] K. Kimura, K. Sugiura, Adhesive composition, in, Google Patents, 1982.[5] J.S. Otness, D.G. Medcalf, Chemical and physical characterization of barnacle ce-
ment, Comp. Biochem. Physiol. B 43 (1972) 443–449.[6] T.J. Deming, Mussel byssus and biomolecular materials, Curr. Opin. Chem. Biol. 3
(1999) 100–105.[7] B.D. Winslow, H. Shao, R.J. Stewart, P.A. Tresco, Biocompatibility of adhesive com-
plex coacervates modeled after the sandcastle glue of Phragmatopoma californicafor craniofacial reconstruction, Biomaterials 31 (2010) 9373–9381.
[8] N.F. Carderelli, Barnacle Cement as a Dental RestorativeMaterial, NIH, Bethesda, 1968.
[9] A.J. Kuijpers, G.H.M. Engbers, J. Krijgsveld, S.A.J. Zaat, J. Dankert, J. Feijen, Cross-linking and characterisation of gelatin matrices for biomedical applications, J.Biomater. Sci. Polym. Ed. 11 (2000) 225–243.
[11] S. Son, B.A. Lewis, Free radical scavenging and antioxidative activity of caffeic acidamide and ester analogues: structure–activity relationship, J. Agric. Food Chem.50 (2002) 468–472.
[12] W.M. Wu, L. Lu, Y. Long, T. Wang, L. Liu, Q. Chen, R. Wang, Free radical scavengingand antioxidative activities of caffeic acid phenethyl ester (CAPE) and its relatedcompounds in solution and membranes: a structure activity insight, Food Chem.105 (2007) 107–115.
[13] T.W. Chung, S.K. Moon, Y.C. Chang, J.H. Ko, Y.C. Lee, G. Cho, S.H. Kim, J.G. Kim, C.H.Kim, Novel and therapeutic effect of caffeic acid and caffeic acid phenyl ester onhepatocarcinoma cells: complete regression of hepatoma growth and metastasisby dual mechanism, FASEB J. 18 (2004) 1670–1681.
[14] K. Inoue, Y. Takeuchi, D. Miki, S. Odo, Mussel adhesive plaque protein gene is anovel member of epidermal growth factor-like gene family, J. Biol. Chem. 270(1995) 6698–6701.
[15] S. Reichl, Films based on human hair keratin as substrates for cell culture and tis-sue engineering, Biomaterials 30 (2009) 6854–6866.
[16] K. Katoh, T. Tanabe, K. Yamauchi, Novel approach to fabricate keratin spongescaffolds with controlled pore size and porosity, Biomaterials 25 (2004)4255–4262.
[17] P. Sierpinski, J. Garrett, J. Ma, P. Apel, D. Klorig, T. Smith, L.A. Koman, A. Atala, M. VanDyke, The use of keratin biomaterials derived from human hair for the promotion ofrapid regeneration of peripheral nerves, Biomaterials 29 (2008) 118–128.
[18] S. Thilagar, N.A. Jothi, A.R.S. Omar, M.Y. Kamaruddin, S. Ganabadi, Effect of keratin–gelatin and bFGF–gelatin composite film as a sandwich layer for full-thickness skinmesh graft in experimental dogs, J. Biomed. Mater. Res. B 88B (2009) 12–16.
[19] S. Prasong, T. Wasan, Preparation and characterization of hair keratin/gelatinblend films, Pak. J. Biol. Sci. 14 (2011) 351–356.
[20] M.J.E. Fischer, Amine coupling through EDC/NHS: a practical approach, MethodsMol. Biol. 627 (2010) 55–73.
[21] A. Nakamura, M. Arimoto, K. Takeuchi, T. Fujii, A rapid extraction procedure ofhuman hair proteins and identification of phosphorylated species, Biol. Pharm.Bull. 25 (2002) 569–572.
[22] J. Adler-Nissen, Determination of the degree of hydrolysis of food protein hydroly-sates by trinitrobenzenesulfonic acid, J. Agric. Food Chem. 27 (1979) 1256–1262.
[23] B.B. Mandal, S. Kapoor, S.C. Kundu, Silk fibroin/polyacrylamide semi-interpenetratingnetwork hydrogels for controlled drug release, Biomaterials 30 (2009) 2826–2836.
[24] A.J. Kuijpers, G.H.M. Engbers, J. Feijen, S.C. De Smedt, T.K.L. Meyvis, J. Demeester, J.Krijgsveld, S.A.J. Zaat, J. Dankert, Characterization of the network structure ofcarbodiimide cross-linked gelatin gels, Macromolecules 32 (1999) 3325–3333.
[25] D. Gupta, C.H. Tator, M.S. Shoichet, Fast-gelling injectable blend of hyaluronanand methylcellulose for intrathecal, localized delivery to the injured spinal cord,Biomaterials 27 (2006) 2370–2379.
[27] S. Haemers, G.J.M. Koper, G. Frens, Effect of oxidation rate on cross-linking ofmussel adhesive proteins, Biomacromolecules 4 (2003) 632–640.
[28] J. Monahan, J.J. Wilker, Cross-linking the protein precursor of marine mussel adhe-sives: bulk measurements and reagents for curing, Langmuir 20 (2004) 3724–3729.
[29] R. Schrieber, H. Gareis, Gelatine Handbook, Wiley-VCH, Theory and industrial practice,2007.
[30] I. Migneault, C. Dartiguenave, M.J. Bertrand, K.C. Waldron, Glutaraldehyde: be-havior in aqueous solution, reaction with proteins, and application to enzymecrosslinking, Biotechniques 37 (2004) 790–806.
[31] M.K. McDermott, T. Chen, C.M.Williams, K.M. Markley, G.F. Payne, Mechanical proper-ties of biomimetic tissue adhesive based on the microbial transglutaminase-catalyzedcrosslinking of gelatin, Biomacromolecules 5 (2004) 1270–1279.
[32] H.W. Sung, D.M. Huang, W.H. Chang, L.L.H. Huang, C.C. Tsai, I.L.I.N. Liang, Gelatin-derived bioadhesives for closing skin wounds: an in vivo study, J. Biomater. Sci.Polym. Ed. 10 (1999) 751–771.
[33] M.C. Figueroa-Espinoza, M.H. Morel, A. Surget, M. Asther, S. Moukha, J.C. Sigoillot,X. Rouau, Attempt to cross-link feruloylatedarabinoxylans and proteins with afungal laccase, Food Hydrocolloids 13 (1999) 65–71.
[34] G. Strauss, S.M. Gibson, Plant phenolics as cross-linkers of gelatin gels and gelatin-basedcoacervates for use as food ingredients, Food Hydrocolloids 18 (2004) 81–89.
[35] G. Dryhurst, K.M. Kadish, F. Scheller, R. Renneberg, Biological Electrochemistry,Academic Press, New York, 1982.
[36] G.D. Fasman, Circular Dichroism and the Conformational Analysis of Biomole-cules, Plenum Press, New York, 1996.
[37] M.L. Tiffany, S. Krimm, Circular dichroism of the random polypeptide chain, Bio-polymers 8 (1969) 347–359.
[38] A.M. Tamburro, V. Guantieri, L. Pandolfo, A. Scopa, Synthetic fragments and ana-logues of elastin. II. Conformational studies, Biopolymers 29 (1990) 855–870.
[39] G. Shanmugam, P.L. Polavarapu, Structural transition during thermal denatur-ation of collagen in the solution and film states, Chirality 21 (2009) 152–159.
[40] R.W. Woody, Circular dichroism and conformation of unordered polypeptides,Adv. Biophys. Chem. 2 (1992) 37–79.
[41] P. Manavalan, W.C. Johnson, Sensitivity of circular dichroism to protein tertiarystructure class, Nature 305 (1983) 831–832.
[42] K. Ono, Y. Saito, H. Yura, K. Ishikawa, A. Kurita, T. Akaike, M. Ishihara, 289–295, Photocrosslinkable chitosan as a biological adhesive, J. Biomed. Mater.Res. A 49 (2000) 289–295.
4039S. Thirupathi Kumara Raja et al. / Biochimica et Biophysica Acta 1830 (2013) 4030–4039
[43] M. Ishihara, K. Nakanishi, K. Ono, M. Sato, M. Kikuchi, Y. Saito, H. Yura, T.Matsui, H. Hattori, M. Uenoyama, Photocrosslinkable chitosan as a dressingfor wound occlusion and accelerator in healing process, Biomaterials 23(2002) 833–840.
[44] B.H. Hu, P.B. Messersmith, Enzymatically cross linked hydrogels and their adhe-sive strength to biosurfaces, Orthod. Craniofac. Res. 8 (2005) 145–149.
[45] E.M. Petrie, Handbook of Adhesives and Sealants, second ed. McGraw-Hill, New York,2007.
[46] M.M. Koganov, O.V. Dueva, B.L. Tsorin, Activities of plant-derived phenols in a fi-broblast cell culture model, J. Nat. Prod. 62 (1999) 481–483.
[47] S.A. Burke, M. Ritter-Jones, B.P. Lee, P.B. Messersmith, Thermal gelation and tissueadhesion of biomimetic hydrogels, Biomed. Mater. 2 (2007) 203.
