Hydrogel membranes based on genipin-cross-linked chitosan blends for corneal epithelium tissue engineering
- First Online:
- Cite this article as:
- Grolik, M., Szczubiałka, K., Wowra, B. et al. J Mater Sci: Mater Med (2012) 23: 1991. doi:10.1007/s10856-012-4666-7
- 2.2k Downloads
Novel polymeric hydrogel scaffolds for corneal epithelium cell culturing based on blends of chitosan with some other biopolymers such as hydroxypropylcellulose, collagen and elastin crosslinked with genipin, a natural substance, were prepared. Physicochemical and biomechanical properties of these materials were determined. The in vitro cell culture experiments with corneal epithelium cells have indicated that a membrane prepared from chitosan–collagen blend (Ch–Col) provided the regular stratified growth of the epithelium cells, good surface covering and increased number of the cell layers. Ch–Col membranes are therefore the most promising material among those studied. The performance of Ch–Col membranes is comparable with that of the amniotic membrane which is currently recommended for clinical applications.
Ocular diseases and wounds requiring treatment affect more than 15 million people worldwide each year . A substantial fraction of them are mechanical, thermal, or chemical injuries of cornea. It is estimated that more than 10 million people in the world suffer from problems with cornea  which are currently the second most common cause of blindness in the world, with only cataract being more frequent. Cornea is the outermost transparent five-layer part of the eyeball covering iris and pupil. It plays three main important roles. First, it acts as a physical barrier against pathogenic microorganisms, dirt, and other noxious physical factors. Second, it plays an active role in the process of vision by refracting light onto lens and retina. It is estimated that cornea is responsive for 70 % of the refracting power of an eye . Third, it absorbs UV radiation between 200 and 295 nm preventing the damage of other elements of the optical system of an eye. Corneal transparency and optical refraction is preserved as a consequence of the continuous renewal of the epithelium, the outermost layer of the cornea . Epithelium is made up of 5–7 layers of very regularly arranged cells . The thickness of human corneal epithelium is about 50–52 μm while overall thickness of the cornea is about 600 μm. The renewal of corneal epithelium is maintained by the proliferation and differentiation of the corneal epithelial stem cells, or limbal stem cells (LSCs) located in the basal layer of the cornea, known as the limbus, located at the border of cornea and sclera [6, 7]. Cornea is quite resistant to minor injuries or abrasions due to the ability of the corneal epithelium to undergo continuous renewal. In the case of injury, the epithelial cells migrate at a rate of 60–80 μm/h until wound is closed . Dysfunction or loss of the LSCs resulting from chemical or thermal burns, contact lenses related or microbial infections, inflammatory eye diseases, hereditary or iatrogenic disorders can cause the cornea surface opaqueness [6, 7, 9].
There are several approaches to the treatment of seriously injured cornea. One of them is a replacement of the cornea. Corneal blindness may be treated by transplantation of donor cadaver corneas, known as penetrating keratoplasty . In fact, it is cornea which was the first allografted human tissue  and penetrating keratoplasty is still one of the most successful types of transplantations. However, the availability of donor corneas is very limited. Moreover, in some cases such as severe chemical burns, ocular pemphigoid, Stevens–Johnson syndrome, trachoma, severe dry eye syndrome, severe herpes zoster, aniridia, certain metabolic opacities, ectodermal dysplasia, and vascularized traumatic injuries, penetrating keratoplasty gives poor results .
Damage of the surface epithelia and corneal stroma leads to the severe cicatrisation of the ocular surface. In such cases combination of artificial materials (PMMA) and human solid tissue is used to restore vision. Currently, in severely destroyed corneas application of keratoprosthesis is recommended . That device, applicable in clinical practice, is built from optical cylinder and its carrier. Currently, two types of keratoprostheses are used. The most popular are the Boston type 1 and 2 keratoprostheses carried by the donor’s cornea , while the second type, called osteo-odonto-keratoprosthesis , is carried by the skeletal bone or dental laminate. Also the artificial corneas, e.g., Alphacor made from PHEMA are available. They have a sponge-like peripheral region with interconnecting pores allowing biointegration with surrounding corneal tissue . Corneal replacements made of animal tissues, usually porcine, are also used [15, 16].
Corneal structure reconstruction can be also proceeded in a layer by layer approach. Transparent corneal surface may be restored by transplantation of autologous limbal or oral mucosa epithelia cultured ex vivo on a proper support which is then implanted together with the confluent sheet of expanded epithelial cells. This procedure is sufficient to reconstruct the ocular surface, however, for the reconstruction of deeper corneal layers penetrating or lamellar keratoplasty techniques are required. Various materials have been used as LSCs culture supports, the amniotic membrane (AM) being the clinical standard due to the content of growth factor and low immunogenicity . However, this material is costly and is associated with a high risk of disease transmission. Therefore, alternative materials for AM are strongly desired. Both synthetic and natural polymers are considered as AM replacements for the ex vivo culturing of corneal cells. The examples of the former are modified and unmodified copolymers of 2-hydroxyethyl methacrylate (HEMA) and methyl methacrylate (MMA) . Natural polymers used for the fabrication of scaffolds include gelatin and chondroitin sulfate , silk fibroin , recombinant human collagen , and argon plasma treated collagen .
The biopolymer, which has gained great and still rapidly rising interest in ophthalmology, is chitosan (Ch). Chitosan is a linear polysaccharide derived by partial N-deacetylation of chitin, which is the primary structural polymer in arthropod exoskeletons, shells of crustaceans, or the cuticles of insects . It is extensively studied due to its unique biocompatibility, biodegradability, biological inertness, stability in the natural environment as well as antifungal and anti-bacterial properties [22, 23]. It has found numerous pharmaceutical applications, primarily as a component of drug delivery systems including ocular ones [24, 25, 26, 27, 28, 29, 30, 31]. We have also studied the application of chitosan-based materials as a drug-carrier  and as antiheparin agents [33, 34]. What is important from the point of view of the studies presented here, chitosan is successfully used for constructing supports for adhesion, proliferation, and differentiation of cells [35, 36, 37].
To increase their mechanical strength cell culture supports based on chitosan are chemically crosslinked usually using glutaraldehyde [38, 39], but also with reagents such as glyoxal  and epichlorohydrin . However, these substances are toxic and may impair the biocompatibility of the crosslinked biomaterials. Therefore, much interest is now directed toward natural crosslinking substances with low toxicity such as genipin (Gp), which was used to crosslink all the materials described in this paper. Genipin is naturally found in the Gardenia jasminodes Ellis fruit. Genipin-crosslinked chitosan is a fluorescent bluish hydrogel, which has been intensively studied recently [29, 42, 43, 44, 45] since it is reported to be about 5,000–10,000 times less cytotoxic than glutaraldehyde  and genipin-cross-linked materials have comparable mechanical strength to the glutaraldehyde-cross-linked ones .
The purpose of the current studies was to obtain genipin-crosslinked chitosan-based scaffolds and to determine their applicability as alternatives for AM in reepithelialization of the cornea. Although chitosan and its derivatives, both as a single polymer and in blends with other polymers, have been already used as supports for corneal epithelial cells, they were not chemically crosslinked [14, 48], or crosslinked with toxic  or costly  crosslinkers. To the best of our knowledge, this is the first report on the application of genipin-crosslinked chitosan scaffolds for culturing corneal epithelium. We have studied the chitosan supports containing additions of other biopolymers, i.e. hydroxypropyl cellulose (HPC), collagen (Col), and elastin (Ela) frequently used for the fabrication of scaffolds.
2 Experimental section
Low-molecular-weight chitosan (Ch) was purchased from Sigma. The degree of deacetylation of the chitosan was approximately 77 %, as determined by elemental analysis. Genipin (Gp) powder (98 %) was obtained from Challenge Bioproducts Co. Hydroxypropyl cellulose (HPC), elastin (Ela), and boric acid (ACS reagent) were obtained from Sigma. Solution of collagen type I (0.3 %, Col) from rat tail was obtained from BD Biosciences. Disodium hydrogen phosphate (analytical grade) and potassium dihydrogen phosphate (analytical grade), hydrochloric acid, ethanol (analytical grade) were obtained from Polskie Odczynniki Chemiczne (Gliwice, Poland). Sodium tetraborate decahydrate (analytical grade) was obtained from Fluka. Sodium chloride (analytical grade) was obtained from Lach:Ner. All chemicals were used without further purification. Water was distilled twice.
2.2 UV–Vis absorption spectra
The UV–Vis absorption spectra of the membranes supported on 1-mm thick quartz plates were measured using a 8452A Hewlett-Packard spectrophotometer.
2.3 Preparation of membranes based on chitosan
Chitosan (Ch) solution (2 % w/v) was prepared by dissolving 0.8 g of Ch in 40 mL of 0.1 M hydrochloric acid. Genipin (Gp) solution (5 % w/v) was prepared by dissolving 0.1 g of Gp powder in 2 mL of 70 % v/v ethanol. 6 % w/v Hydroxypropylcellulose (HPC) solution was prepared by dissolving 0.9 g of HPC powder in 15 mL of water. Elastin (Ela) solution (13.3 % w/v) was obtained by dissolving 2 g of Ela in 15 mL of 0.25 M oxalic acid. Collagen solution was used as received. The hydrogel membranes were prepared using 1.5 mL of clear, slightly yellowish mixture of equal volumes of Ch solution and the solutions of HPC, Col and Ela, respectively. The polymeric mixtures were stirred for 5 min and then 40 μL of Gp solution was added to initiate the crosslinking reaction. The mixture was homogenized by vigorous stirring for 10 min at room temperature and then poured onto a 60 mm plastic Petri dish and placed in an incubator for 48 h at 45 °C. After a few of hours the solution became lightly blue and increasingly viscous due to the started crosslinking reaction.
2.4 Swelling ratio measurements
2.5 Contact angle measurements
The values of the contact angle of water on polymer membranes were measured using Surftens Universal instrument (OEG GmbH, Frankfurt, Germany) at room temperature. A small drop of doubly distilled water was deposited onto the membrane and the contact angle was measured immediately. The contact angle values reported are the averages of five consecutive measurements for each sample.
2.6 Optical microscopy
The Nikon Eclipse LV 1000 optical microscope was employed to observe the morphologies of the membranes based on Ch crosslinked with Gp. The membranes were imaged at room temperature.
2.7 Atomic force microscopy (AFM)
The surface topography of the membranes was analyzed using a Nanoscope IVA atomic force microscope. AFM images in air were obtained using tapping mode technique. The root mean square (RMS) roughness was calculated from data obtained.
2.8 Mechanical testing
2.9 Cell culture assays
The agreement of the Bioethical Commission of Silesian Medical University was obtained (agreement number: NN-6501-184/I/05/06).
Culture media and chemicals were purchased from Sigma (Germany). Reagents for immunostaining were purchased from Santa Cruz Biotechnology Inc. (USA). All parts of the experiment were performed under tenets of Declaration of Helsinki.
The cells used in the study were human corneal epithelial cells collected for cultivated epithelium transplantation procedure. The limbal epithelium source were the eyes of healthy donors. Before donation each eye was examined to detect pathology which could pose a potential risk of visual acuity decrease in the future. All patients were informed about transplantation procedure, experimental assays, and signed agreement forms.
Limbal epithelium was collected under local anesthesia with local decontamination with 10 % solution of povidone–iodine for skin and 5 % povidone–iodine for conjunctiva. One minute after decontamination agent was washed out with a buffered salt solution (BSS). Limbal 2 mm2 specimen from upper limbus was gently cut with a crescent knife. Tissue was transferred to corneal storage medium at 4 °C. Tissue specimen was then trypsinized to obtain cell suspension with 1 % trypsin and 0.01 % EDTA for 10 min. Cells were gently scraped with the microscraper.
Culture dishes (Becton–Dickinson, USA) were covered with 3T3 fibroblasts (ATCC, USA) a week before the test. Cells were cultivated in Dulbecco’s Modified Eagle’s Medium (DMEM) with 10 % bovine serum and penicillin/streptomycin mixture. The monolayer of 3T3 fibroblasts was inactivated by incubation in regular medium containing 2 μg/ml of Mitomicin C for 2 h. The whole epithelial culture was carried out in the presence of 3T3 fibroblasts as a source of growth factors. The epithelial single cells were seeded on the membranes of two types (Ch–Col and Ch–Ela) in Petri dishes of 100 mm diameter. Cellular suspension with density of 1–4 × 104 cells for 1 mL were settled in the culture dishes (Cell counter, Coulter Z1, Miami, USA). Epithelial cultures were carried out in standard conditions in 37 °C in humidified atmosphere of 5 % CO2 and 95 % air. The medium was supplemented DMEM/HAM F12 mixture with 10 % bovine serum, 0.5 % dimethyl sulfoxide (DMSO), 10 ng/ml mouse epidermal growth factor (EGF), 5 μg/mL bovine insulin, 0.1 nM cholera toxin, 0.18 mM adenine, 2 nM triiodothyronine, 4 mM l-glutamine, 0.4 mg/mL hydrocortisone, and 100 μg/mL penicillin and streptomycin mixture. Culture medium was changed every 48 h. At the10th day of culture the plates were inspected under the light microscope for evaluation of epithelial growth .
The histological examinations of the samples were carried out. For these investigations the membranes with cultured cells were fixed with 10 % neutral buffered formalin (4 % formaldehyde in phosphate buffered saline) overnight at 4 °C. To remove fixative agent and water the samples were dehydrated in a graded series of alcohol solutions (10–20–50–95–100 %). Finally, in order to visualize and differentially identify microscopic structures of cultured epithelium the histological stains (hematoxylin—blue and eosin—pink) were used. Immunostaining for cytokeratin 3 (K3), cytokeratin 12 (K12), protein p63, and connexin 43 was performed to confirm corneal origin of the epithelium (K3, K12) and the presence of low differentiated cells.
3 Results and discussion
Since the membranes are expected to be resorbed after implantation, the fact that they are colored upon implantation should not pose a problem. On the contrary, their bluish tint should facilitate visual estimation of the degree of their resorption. The thickness of the membranes obtained was in the range from 6 to 23 μm.
3.1 Swelling of membranes
Swelling ratios, S (%), of the membranes at different pH values determined after 24 h of equilibration
pH = 6.0
pH = 7.4
pH = 9.0
All polymeric hydrogel membranes display significant water sorption ability. The samples prepared from the mixtures of Ch with proteins (Col and Ela) reveal similar degree of swelling. At pH = 7.4 the values of swelling ratio were very different for the three materials studied, while for pH of 6 and 9 the differences were much smaller. The swelling behavior of Ch crosslinked with Gp has been already well characterized . The degree of swelling of genipin-crosslinked Ch was found to increase with decreasing of pH value. That can be explained considering the pH effect on the protonation-deprotonation of the amino groups present in chitosan macromolecule inducing conformational changes of macromolecule in the networks. Protonation of amino groups in acidic solutions leads to the chain extension and chain repulsion. That increases the amount of water present in the polymeric network. HPC present in the chitosan gel (Ch–HPC) decreases the Ch sorption ability and lowers hydrogel sensitivity to the pH of solution. However, the membranes containing Col, Ela, and HPC do not follow the pH dependence of swelling characteristic of Ch.
3.2 Contact angle measurements
Water contact angle values for the studied membranes
Contact angle (°)
60.28 ± 4.13
60.40 ± 7.81
54.94 ± 5.48
3.3 Surface morphology studied with optical microscopy
For Ch–HPC (Fig. 3a) the surface seems to have a fibrous structure, while the morphology of Ch–Col surface is very smooth, with only some defects visible (Fig. 3b). The Ch–Ela membranes (Fig. 3c) are covered with droplet-like hemispherical features. Thus, by the addition of another biopolymer to the chitosan one can obtain genipin crosslinked membranes of very different morphologies. This is an important finding indicating that the surface morphology of the chitosan membranes may be easily modified and optimized for corneal epithelium growth and migration.
3.4 Surface morphology studied with AFM
Values of the RMS roughness (nm) of the studied membranes
RMS roughness (nm)
3.5 Biomechanical testing
Values of tensile strength, elongation at break, and Young’s modulus of the membranes
Tensile strength (MPa)
Elongation at break (%)
Young’s modulus (GPa)
31.70 ± 4.16
0.32 ± 0.04
19.93 ± 3.32
46.93 ± 5.72
0.36 ± 0.05
23.53 ± 4.22
48.10 ± 5.76
0.28 ± 0.05
33.03 ± 5.79
The values of tensile strength for the materials obtained (32–48 MPa) are much higher than those for the scaffolds obtained from amniotic membrane (2.3 MPa)  or decellularized porcine cornea (2.4–4.2 MPa) . The elongation at break expresses the elasticity of a material and it is very similar for all samples studied. It was concluded that the blends containing proteins are the most promising candidates as cell culture supports. Therefore, cell culture tests were performed using the Ch–Col and Ch–Ela membranes.
3.6 Epithelial cell culture tests
In the case of Ch–Ela membranes assay (Fig. 5c), growth was not regular with differences in the number of cell layers, poor attachment to the carrier surface and local areas covered only by epithelial colonies. Therefore, only Ch–Col carriers can be considered as eligible for grafting in humans. Membrane compounds require further studies to establish proper surface structure able to carry stratified epithelium. Collagen, which is a common component of basement membranes, seems to be more efficient in improving adhesive properties of Ch–Col membranes. The design of the artificial membranes should include superficial features of human basement membranes to obtain adequate and long-lasting cellular attachment.
Novel polymeric membranes based on blends of biopolymers Ch–HPC, Ch–Col and Ch–Ela crosslinked with natural substance, genipin, have been successfully prepared with the aim to use them as supports for corneal epithelium cell culturing. Due to the poor biomechanical performance of Ch–HCP that material was eliminated from the biological studies. The cell culture experiments carried out on Ch–Col and Ch–Ela membranes have indicated that Ch–Col is the most promising material. The results obtained with Ch–Col were comparable with these of standard cultures carried on the amniotic membrane, currently recommended for clinical applications. The good performance of Ch–Col can be explained considering the chemical properties of the biopolymers used but also good physicochemical and biomechanical characteristic of Ch–Col membrane, especially reasonable hydrophilicity, optimal morphology and reasonable mechanical parameters, all most likely resulted from good mixing of the blend components forming the homogenous mixture and the fact that both components undergo crosslinking process. Thus, genipin crosslinked Ch–Col hydrogel seems to be a promising material for further clinical tests directed towards the development of implantable corneal epithelium tissue.
Project operated within the Foundation for Polish Science Team Programme co-financed by the EU European Regional Development Fund, PolyMed, TEAM/2008-2/6. MG gratefully acknowledges a grant from National Science Centre 2011/01/N/ST5/05544. The authors thank Dr. Eng. Ewa Stodolak (Faculty of Materials Science and Ceramics, AGH University of Science and Technology, Cracow, Poland) for performing the mechanical tests.
This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.