Recent Approaches of Collagen Biomatrix Modification as a Corneal Biomatrix and its Cellular Interaction

1. Introduction

The cornea is a transparent window to the eyes that maintains the refractive properties of light transmission to the retina [1,2,3]. The cornea is a multilayered component and is enclosed by a non-keratinized stratified epithelium layer, continuously providing a smooth ocular surface [1,2,3,4], as shown in Figure 1. The smoothness and integrity of corneal epithelium are essential for transparency, providing adequate light refraction, and homeostasis [1,5]. These important roles depend on a balanced corneal epithelial cells (CECs) turnover, as the new CECs originated from the limbal epithelial stem cells (LESCs), which were located at the periphery of the cornea and replaced the older CECs [1,2,6]. LESCs maintain the self-renewal of CECs by continuous and slow epithelization processes which are involved in the proliferation and differentiation of LESCs into CECs, followed by centripetal migration of CECs towards the central region of the cornea [1,2,6].

The cornea is also responsible for the frontal barrier that allows the diffusion of oxygen and essential nutrient from the tear film, against invading pathogens, debris, chemical agents, and trauma [7,8,9,10]. Thus, the corneal epithelium is vulnerable to external injury, which compromises its first line defense against corneal damage and can be overcome by a rapid healing process through re-epithelialization activity [7,9,10,11,12,13,14]. The stages of the wound healing process following corneal epithelial injury were illustrated in Figure 1.

However, various potential problems such as delay in cell migration, epithelial hypertrophy and recurrent corneal erosion are prone to occur during the epithelialization process, which eventually leads to the scarring of the stroma, reducing the vision quality and corneal damage [11,15,16,17]. It may be due to the dysfunction, destruction, or deficiency of LESCs, which is also known as limbal stem cell deficiency (LSCD). Recent studies also reported that LSCD can induce unstable production of corneal epithelium, followed by corneal ulceration, conjunctival invasion into the cornea and neovascularization on the corneal surface that leads to inflammation and chronic pain, thus, ultimately to vision loss [18,19].

The only recognized treatment strategy for corneal blindness or vision loss is through corneal transplant but is limited due to the corneal shortage [20,21]. Another alternative solution was by tissue engineering approach which replaces the damaged cornea with a biomaterial-based biomatrix combined with cells to replicate the corneal tissue [22,23,24,25]. Many corneal biomatrices have been developed to replace all or only a part of the cornea depending on the patient’s requirement [26]. The development of cornea biodegradable biomatrix (specific for the epithelial layer damage treatment) was focusing on regenerating the damaged epithelial layer either from transplanted CECs or differentiated from transplanted LESCs. Thus, the LESCs biomatrix is important in supporting the expansion, stratification and maintaining LESCs functions [27].

Biological biomatrices such as the human amniotic membrane (HAM) [28,29], fibrin [30,31,32] and feeder layers such as 3T3 fibroblast [33] are gold standard treatments and widely used as cell carriers as they promote cell expansion. However, as natural carriers, their potential drawbacks such as the tendency to carry infection, not being optically transparent and inadequate structural compaction and rigidity as a corneal biomatrix were reported [34,35]. The high economical cost of these natural biomatrix needs to be overcome by discovering new biomaterials for CECs and LESCs [36].

Many studies were focusing on biomaterial for CECs and LESCs including collagen, silk fibroin [37], gelatin [38], chitosan [39], alginate [40], hyaluronic acid [41] and also decellularized cornea [27]. Collagen is one of the well-known biomaterials for corneal biomatrix which mimics native corneal structure [4,42,43,44]. Collagen makes up about 70% of the dry-wet of the cornea and plays an important role in supporting CECs, LESCs and corneal fibroblast cell growth [4,6,9,10,23]. Three forms of collagen biomatrix are normally used in tissue engineering research, such as collagen sponge [45,46,47,48], collagen hydrogel [49,50,51] and collagen film [52,53,54].

The important aspect that needs to be considered during the collagen biomatrix development is the collagen biomatrix interaction with CECs and LESCs which is influenced by the physicochemical properties of the biomatrix [26,55]. The source of biomaterial, mechanical strength, biodegradation rate, optical characteristics and biocompatibility properties need to be tailored during the development of corneal biomatrix [55,56,57]. Considering the aforementioned characterization of the biomatrix, modification or optimization process is crucial for achieving optimal biomatrix designs [26]. Modifications of collagen biomatrix such as crosslinking, physical modification, incorporation of other biomolecules or cells into collagen biomatrix or incorporation of the collagen into another biomatrix, give different effects in terms of interaction between the biomatrix and cells. Since the main outcome of the produced corneal biomatrix is a prolonged effect after implantation, thus it is important to ensure its ability to regenerate into corneal native tissue.

The mechanical strength of the biomatrix is an important aspect that needs to be optimized as it must resist the high tension during implant, and also in vivo dynamic environments such as physiologic intraocular pressure and constant eyelid motion [58], which is closely related to the biocompatibility towards LESCs and CECs [59,60,61]. Modification through the cross-linking process will produce a mechanically strong cross-linked biomatrix [62], which results in multidimensional polymeric chain extension of the biomatrix. Unfortunately, the crosslinker could reduce biomatrix’s transparency and cause cell toxicity that will overshadow their cross-linking potential [51].

Recently, several studies have reported the improvement of cell–biomatrix interactions via surface alterations of the biomatrix. The biomatrix surface with moderate hydrophilicity, irregular structure and the cationic charge was likely to be attached and grown by the LESCs and CECs [63,64]. Some studies have exploited the biocompatibility properties of collagen by incorporating or coating the collagen on another biomatrix to increase the biomatrix biocompatibility [40,65]. All modified biomatrix with their desirable properties for corneal engineering are shown in Figure 2.

A literature search was conducted to identify the recent modifications that have been performed on collagen biomatrix within recent 10 years. This systematic review aims to discover the modification strategies to optimize collagen biomatrix for CECs and LESCs carriers in the treatment of LSCD. This review will also provide insight to further explore a better and safe modification to the collagen biomatrix for CECs and LESCs in future studies.

2. Materials and Methods

2.1. Search Strategy

This systematic review was performed to identify the relevant studies of recent modifications on collagen biomatrix and the efficacy of these modifications on the physicochemical and biocompatibility of the biomatrix toward CECs or LESCs, both in vitro and in vivo. Briefly, it was constructed based on PRISMA guidelines to ensure the quality and transparency of this review [66]. Five separate databases including Scopus (Elsevier, Amsterdam, NH, The Netherlands), Web of Science (WoS) (Clarivate Analytics, Philadelphia, PA, USA), PubMed (National Center for Biotechnology Information, NCBI, Bethesda, MD, USA), Wiley (John Wiley and Sons, Inc, New Jersey, USA) and EBSCOhost (EBSCO Information Services, Massachusetts, USA) were systematically searched to discover the studies related to the biomaterial or bioengineering, especially in collagen biomatrix as a biomatrix of CECs and LESCs in corneal therapy.

These databases have screened all related published journal articles. This article search was guided by the focus question formulated using the PICO strategy whereby Population (P) was in vitro and in vivo study on collagen biomatrix for CECs and LESCs transplantation; Intervention (I) was different modification strategies on the collagen biomatrix; Comparison (C) with other biomaterials was not applicable; and Outcome (O) was physicochemical and cellular characteristics of the biomaterials studied towards CECs or LESCs (in vitro and in vivo).

The combination of three sets of keywords (corneal epithelial cells OR CECs OR corneal epithelium OR limbal epithelium OR limbal epithelial cells OR limbal epithelial stem cells OR LESCs) AND (limbal deficiency OR limbal stem cell deficiency OR LSCD OR corneal limbal stem cell deficiency OR corneal epithelial injury) AND (collagen OR collagen biomatrix OR collagen bio scaffold OR collagen scaffold) were used during the searching process of the relevant articles published.

2.2. Criteria of Selection

Only English articles were included due to limited resources for translation. Studies that provide free full-text articles published within 10 years, with a limit ranging from 2012 to 2022 were considered. Titles and abstracts that have fulfilled the topic requirements were systematically screened. The articles related to humans were included as the relevant basis for the scope of this review. All research articles related to collagen as a component or a part of the biomatrix for CECs and LESCs were also included. All secondary literature and any original article that involved clinical studies were removed. Any studies that focused on the other fields except for physicochemical properties and biocompatibility (in vitro and in vivo) were also omitted.

2.3. Management of Data Extraction Table

Articles were screened and underwent three phases to be selected as part of this systematic review. The first phase involved article title screening that meets the requirement of the topic of interest. The title that did not match the inclusion criteria was removed. The next phase involved the elimination of unrelated articles based on inclusion criteria followed by the removal of all identical articles. The last steps involved omitting the articles that did not meet the inclusion criteria after full-text reading by two independent reviewers. Two reviewers were independently assessing the specified inclusion and exclusion criteria of selected published articles to guarantee neutrality in the selection of the final articles.

This was accompanied by a discussion among the reviewers to get a consensus on the discrepancies that emerged during the assessment of the articles. Extracted information as outlined in the data extraction table for in vitro study are as follows (1) Author and year published (2) Type of biomatrix (3) Modification techniques (4) Type of cells used (5) Test and result (physicochemical properties) (6) Test and result (in vitro biocompatibility) and (7) Conclusion. The data extraction table for in vivo study was outlined as follows (1) Author and year published (2) Type of biomatrix (3) Modification techniques (4) Animal model/injury (5) Test and result (in vivo) and (6) Conclusion. This review is not suitable to be published in PROSPERO as it is included in vitro studies. For quality assessment, this review was carried out systematically, in which employing the critical appraisal instrument [67]. Each item in the appraisal instrument for each selected study was also discussed by independent reviewers.

3. Results

3.1. Searching Result

The combination of three sets of keywords during the searching process successfully identified 444 articles as potentially relevant. A total of 409 articles that have not fulfilled the inclusion criteria and identical were removed during the title and abstract screening process. From the remaining 35 articles, the reviewers omitted six more articles that did not meet the requirement of the inclusion criteria. After the selection process, 29 articles were included in the data extraction table, of which 10 articles were acquired from WOS, seven articles from Scopus, 10 articles from PubMed, one article from Wiley and one article from EBSCOhost, respectively. The article's screening and selection process were summarized in Figure 3.

3.2. Study Characteristics

In this review, the search was successfully finalized the studies related to the modification of collagen biomatrix and its effect on CECs or LESCs. To summarize the selected articles from 2012 to 2022, eight studies aimed to develop a new formulation for collagen biomaterials [68,69,70,71,72,73,74,75], whereas 12 studies aimed in improving or characterizing current collagen biomaterials, including the improvement of fabrication methods to produce better collagen biomatrix [27,35,51,58,59,60,61,76,77,78,79,80].

Four studies investigated the interaction of corneal cells on collagen biomaterials with other biological molecules and cells [81,82,83,84]. Meanwhile, four studies exploited the biological benefit function of collagen by combination of different biomatrix to improve its function as a cell carrier [40,85,86,87]. From these articles, several types of cells were used such as CECs, LESCs alone or co-culture of both cells and the combination of one of these cells with stromal cells (corneal stromal stem cell or limbal fibroblast). Most of these cells are primarily acquired from the cornea of the human cadaveric donor, rabbit, porcine, mini pig (Gottingen), bovine and mouse. It was either freshly obtained from the corneal rim, immortalized, primary or cell line.

All these studies reported various modifications on the collagen biomatrix either by extraction of collagen from new sources, physical modification, crosslinkers, or incorporation with other cells and biomolecules. Some researchers also incorporate collagen into other biomatrix. The new source of collagen biomatrix that has been obtained is through the decellularization of the bovine eyeballs, porcine conjunctiva and fish scale, production of the synthetic collagen peptide or modification of collagen’s methacrylate group. The physical modification that has been done was compression by using the different compressors, embedding the decellularized corneal lenticule (dCL) with compressed collagen, surface patterning or vitrification process.

Since last 10 years, some researchers used the following crosslinker: polyethene glycol (PEG), N-hydroxysuccinimide (NHS), 1–Ethyl–3-(3-dimethylaminopropyl) carbodiimide (EDC), 1–Ethyl–3-(3-dimethylaminopropyl) carbodiimide-N-hydroxysuccinimide (EDC-NHS), 4-(4,6-dimethoxy-1,3,5-triazin-2-yl)-4-methyl-morpholinium chloride (DMTMM), methacyloyloxyethyl phosphorylcholine (MPC) or hybrid crosslinker. Few researchers used other components such as fibronectin (FN), laminin, stromal cells, ascorbic acid, or stem cell factor (SCF)/ C- kit into the collagen biomatrix. Some studies also incorporated the collagen into another biomatrix such as incorporation of collagen into poly(lactide-co-glycolide) (PLGA) biomatrix, dopamine hydrazone biomatrix-crosslinked hyaluronic acid (HA-DOPA), poly-L/DL-lactic acid (PLA) films or silk film.

Ten studies were conducted both in vitro and in vivo, two articles only involved in vivo study, whereas the rest of the articles only involved in vitro study. Ultimately, the selected articles have shown optimization of collagen biomatrix and its effect on the physicochemical and biocompatibility of biomatrix towards the CECs and LESCs which affect its cellular biocompatibility differently. All articles were summarized in Table 1 for in vitro study and Table 2 for in vivo study.

4. Discussion

4.1. New Bioresource and Its Efficacy on CECs and LESCs

It is vital to tailor the collagen source during the development of the biocompatible biomatrix due to the presence of various amino acids in the collagen depending on the species and tissues sources [1,6,9,10]. This affects the final characteristic, physical properties and biocompatibility of the biomatrix [90,91]. Decellularization is one of the current approaches in producing a new collagen source for the biomatrix.

A previous study by Zhao and colleagues decellularized conjunctiva to produce an acellular conjunctiva matrix (aCM) as LESCs carrier [70]. The conjunctiva has a high degree of similarity to the cornea as both are derived from the epidermal ectoderm. Compared to the denuded amniotic membrane (dAM), the aCM biomatrix has better physical characteristics and is biocompatible with CECs. In vivo, the aCM could reconstruct the ocular surface in LSCD rabbits without neovascularization, inflammation or oedema [70].

A study by Park and colleagues also decellularized corneal stromal tissue from bovine eyeballs and produced three dimensional (3D) bioprinting decellularized collagen sheet (3D-BDCS), which could re-epithelialize the damaged epithelial layer within a few days [71]. Moreover, the decellularization of the porcine limbus and re-cellularized Statens Seruminstitut Rabbit Cornea (SIRC) limbal epithelial cell line and human adipose-derived mesenchymal stem cells (hADSCs) could produce a biomatrix with high content of collagen IV, and are able to regenerate stratified epithelium [75]. This is due to collagen IV affecting CECs transcriptional factor, which plays a significant role in CECs adhesion and migration properties [92,93].

Fish scale collagen was isolated from fish scale of the fresh fish (L. calcarifer) of sea catch, was one of the other new collagen bio sources that is rich in collagen I, followed by coated with polyethene (FSC-PE) possesses favorable physical strength, transparent and biocompatible with corneal cells [68,94]. FSC-PE allows proper epithelialization due to collagen I up-regulation of the specific gene, which regulates cell viability, attachment, and differentiation. The porous nature of FSC-PE also supported the viability and differentiation of LESCs and enhanced CECs migration and proliferation [35,68,94,95].

Collagen mimetic peptides (CMP) or collagen-like peptide (CLP) is another alternative of collagen source for the development of biomatrix [72]. CMP promotes realignment of damaged collagen, which accelerates wound closure in vivo. This is due to CMP is a short synthetic collagen peptide that able intercalate into damaged endogenous collagen I in vivo [96,97]. CMP is also recognized to enhance CECs density and re-epithelialization process with a better organization of epithelial layers [69,72,77].

Moreover, Qin and friends (2021) [80] produced Collagen Methacrylate (ColMA) by modifying collagen with a methacrylate group, followed by photo crosslinking. ColMA is a transparent biomatrix, with high-pressure overload capacity and is compatible with hCECs. Nanogranules from dislodging ColMA adhere to stromal tissue promoting re-epithelization, reducing myofibroblast activation and decreasing scar formation.

Within these recent 10 years, all collagen new bioresources including FSC-PE, aCM and 3D-BDCS potential to be developed as CECs or LESCs carriers. The presence of alternative extracellular matrix (ECM) to the macro-molecule collagen including CMP and ColMA is an attractive biomaterial and suitable to be developed as a biocompatible biomatrix for CECs or LESCs.

4.2. Physical Modification of Biomatrix

From the past 10 years, some researchers have performed physical modifications to improve the mechanical stability of the collagen biomatrix. The collagen biomatrix is also modified through compression technique by Jones and colleagues (2012) [59] and Xeroudaki and colleagues (2020) [58] modified. They found that compression of collagen hydrogel bioengineered porcine collagen (BPC) crosslinked with EDC-NHS reduced the water content, thus permitting control of the collagen concentration, stiffness, mechanical strength and surface topography, which contributes to its biocompatibility with corneal cells [49,59,60,61]. The compressed BPC supports the proliferation and maintains the normal morphology of hCECs (in vitro and in vivo) [58,59].

A study by Gouveia and colleagues [60] have reported that plastic compression of the collagen hydrogel which is Real Architecture for 3-Dimensional Tissue Equivalent (RAFT TE) also likely supported the differentiation of the LESCs via mechanotransduction-dependent pathways, Yes-associated protein (YAP) support the viability of the differentiated CECs. However, RAFT TE promotes the migration of LESCs which maintained a single monolayer, but few stratification cells became round and detached from the basal sheet. These cause RAFT TE to have a low viable differentiated CECs number and flat stretched CECs morphology compared to uncompressed softer collagen hydrogel [60].

Therefore, LESCs maintenance is highly dependent on the softer biomechanical limbus niche region properties which are opposed to the relatively stiff corneal central [98,99,100,101]. This is due to the LESCs niche requiring a specialized microenvironment to maintain the undifferentiated LESCs by slowing their migration rate and preserving their proliferative and stratification capabilities [60,79]. But these are opposed by studies by Massie et al., (2015) [61] and Kureshi et al., (2014) [78]. They showed the compressed RAFT TE supported the attachment, viability and proliferation of undifferentiated LESCs in vitro [78], [59].

Hong and colleagues (2018) [27] have conducted another physical modification of collagen bio-composite which embedded the decellularized corneal lenticule (dCL) with compressed collagen as a LESCs carrier known as COLLEN. COLLEN take the biological advantages from the compressed collagen and the mechanical properties of the dCL. COLLEN support the LESCs and hCECs attachment, expansion, morphology, and functions which were similar to the compressed collagen. In vivo, COLLEN-based limbal graft was stably grafted with native limbal region tissues without neovascularization, oedema, conjunctivalization and inflammation. COLLEN also support multilayered differentiated CECs after 2 weeks post-implantation while maintaining the putative stem cell markers on the limbal region compared to central cornea [27].

Another study by Haagdorens and colleagues [77] has reported that the surface modification which is FN pattern on CLP hydrogels influenced the pattern of cell proliferation. Yuncin and friends (2021) [102] stated that parallel ridge on silk film-coated collagen I enhanced CECs growth, and spreading and promotes wound recovery. Surface patterning and surface topography including the curvature of the biomatrix also affects the adhesion, proliferation, and gene expression of CECs [103].

The vitrification process is another physical modification approach to develop a rigid glassy material collagen vitrigel (CV) biomatrix. It is a thin membrane composed of high-density organized meshwork type I collagen fibrils with superior mechanical properties, non-degradable, stable water content, and optically transparent and supporting corneal cells [35,104]. In vivo, CV promotes the regeneration of healthy CECs and LESCs with low inflammatory response and reduced neovascularization (5 weeks post-surgery) [35].

In conclusion, the compression on the collagen biomatrix was able to increase the physical strength and affected the behavior of cultivated CECs or LESCs. Most of the researchers reported that compressed collagen biomatrix supported the viability of CECs, but it is not suitable for LESCs growth. This is due to the LESCs required soft biomechanical region properties of the collagen biomatrix to preserve proliferative and stratification capabilities of the undifferentiated LESCs. But these were opposed to the result obtained by [83]. Thus, further study is needed to explore the mechanism behind these corneal cells' behavior. In addition, the surface modification of collagen hydrogel also affects the growth and behavior of cultivated CECs and LESCs. CV and COLLEN are other creative approaches in developing optimal designs of CECs or LESCs biomatrix.

4.3. Crosslinking Effect on the Biocompatibility of the Construct towards CECs/LESCs

Crosslinking is another approach in improving the mechanical stability of the collagen biomatrix. Different cross-linkers have different effects on the mechanical properties, in terms of biocompatibility of the biomatrix towards corneal cells, especially CECs and LESCs. Figure 4 showed the overview of the chemical cross-linker used in recent 10 years based on selected articles. Amide-based crosslinker such as EDC-NHS is commonly used with collagen biomatrix as it mimics the lysine-based crosslinker which is naturally present in collagen [105]. EDC-NHS affect the GxOGER sequences of collagen molecules [58], not remained as a part of the protein structure post-crosslinking [106,107] and improved the porous and interconnected structure of the collagen biomatrix [43], promoting cell attachment, proliferation and viability of the attached cells.

EDC-NHS enhances the physical properties of the collagen biomatrix which has sufficient mechanical strength during subcutaneous implantation and implant in the rabbit cornea but was not strong enough for penetrating keratoplasty (PKP) [58,107,108,109]. This is due to the functional group located on the adjacent collagen microfibril being too far to be bridged by EDC-NHS as EDC-NHS only can link within 1.0 nm from each other [109]. These drawbacks could be overcome by hybridization of EDC-NHS (amide-type crosslinker) with a bifunctional cross-linker such as PEG, which provides synergistic effects on the physical and biological properties of collagen biomatrix [110,111].

Study by Jangamreddy and colleagues [69], Rafat and colleagues [112] have managed to conjugate the CLP-EDC-NHS to PEG (4 arms or 8 arms) to further improve the mechanical strength and promote a stable biomatrix for regeneration. This hybrid crosslinked hydrogel (CLP-PEG-EDC-NHS) enhanced the mechanical strength and elasticity by 100% and 20% respectively, compared to the non-hybrid biomatrix. The hydrogel comprised over 90% water, compared to the cornea that has 78% water, which contributes to its’ biocompatibility. It also did not cause any cytotoxicity effects toward CECs as there was minimal CECs death at 48 hours (h) post cell culture and was able to regenerate the neo cornea with healthy regenerated functional CECs in vivo [69]. Although CLP-PEG-EDC-NHS did not have high tensile strength as control recombinant human collagen type III (RHCIII) conjugated to MPC (RHCIII-MPC) implants, but it was more elastic compared to the RHCIII-MPC, which allowed them to withstand the grafting procedure [69,113].

Regarding Fernandes-Cunha and colleagues [51], the concentration and arm number of PEG affect the transparency, directly proportional to the storage modulus and degradation profile of PEG crosslinked collagen biomatrix. The highest storage moduli achieved was at 8% (8-arm of PEG) but decreased at 16% PEG content. It may be due to the saturation effect, where beyond a certain concentration of PEG, no further crosslinking was achieved, and any additional PEG reduced the crosslinking density and transparency via inadequate macromolecular mixing leading to matrix heterogeneity.

In addition, this biomatrix has low cytotoxicity effects as almost 100% of human immortalized corneal epithelial cells (iCECs) confluent after 2 days post cultivated on all biomatrix. PEG arms concentration and arm number were directly proportional to the iCECs adhesion and proliferation except in 16% of PEG respectively. The high PEG concentration causes the reduction of biomatrix porosity which restrains the mobility of the polymer network and thus, reduced the proliferation of corneal cells [112]. The presence of iCECs improved the transmittance in the 4 and 6 arms by 16% PEG. PEG also affect the alignment of the collagen fiber which affects the iCECs behaviors [113].

Haagdorens and colleagues [77] also used an EDC-NHS crosslinker for the biomatrix of human corneal epithelial cells (hCECs) and human limbal epithelial stem cells (hLESCs). They used seven different collagen-derived hydrogels (recombinant human collagen type I (RHCI) and CLP hydrogel) with EDC-NHS or DMTMM cross-linker as a carrier of immortalized human corneal epithelial cells (ihCECs) and primary hLESCs. All these collagen hydrogels met the physical criteria of a good biomatrix for CECs and LESCs carriers. RHCI and CLP hydrogel irrespective of the crosslinker type except for CLP-12-EDC is biocompatible towards ihCECs as there was minimal cell death that supported the metabolic activity, attachment, and proliferation of ihCECs and primary hLESCs. However, these biomatrix did not promote LESCs differentiation except for CLP [77].

Overall, the crosslinker use on collagen biomatrix as a carrier of CECs or LESCs within these recent 10 years at its optimal crosslinker concentration was able to improve the mechanical stability of the biomatrix and biocompatible with CECs or LESCs. The use of high concentration and arm numbers of the crosslinker leads to the saturation effect which could cause a decrease in the transparency and porosity of the biomatrix and reduce the proliferation of the corneal cells. More novel crosslinkers and techniques need to be explored, either alone or in combination between chemical, enzymatic or physical crosslinker methods, which may give synergistic effects as a CECs or LESCs carrier.

4.4. Interaction of CECs/LESCs with Other Cells and Molecules in a Collagen Biomatrix

Other components such as FN, laminin, stromal cells, ascorbic acid, and SCF/C-kit have been incorporated into the biomatrix as they are able to promote re-epithelization [6,10,13,79]. Wilson and colleagues (2014) [82] investigated the effects of FN on the cultured CECs by seeding the adult porcine CECs on the FN-coating rat tail collagen type 1 hydrogel (RTCI-gels) encapsulated adult human derived corneal stromal (AHDCS) (RTCI-gels-FN-coated–AHDCS). RTCI-gels-FN-coated–AHDCS supported the CECs viability and maintained the normal cobblestone morphology with tight cell-cell junction. This is due to FN-binding integrin a5b1 promoting the CECs adhesion and migrating to cover the uncovered surface FN [13,114,115,116,117]. It mimics the physiological cornea as FN is a temporary ECM that presents abundantly during early corneal wound healing and is progressively replaced by collagen and laminin from the basal membrane as wound healing progressed [10,13,77,118,119,120].

In contrast, CECs appeared to be much smaller, less flattened and lacking the tight cell-cell junction when cultured on the biomatrix treated with wortmannin (epithelial stroma interaction inhibitor) [82]. These indicated the mutual interaction of CECs with stromal cells encapsulated in the hydrogel that was needed to support the CECs growth and enhanced epithelium multilayered organization [82,83,121].

This is also supported by Kureshi et al., 2010 [122], Massie et al., 2015 [84] and Zhang et al., 2015 [123]. Their studies showed that stromal cells/limbal fibroblast cells successfully enhanced the LESCs growth on the collagen biomatrix, which mimics in vivo corneal arrangement. It also was due to the LESCs differentiation depended on the mediated expression of the Wnt/B catenin of bone morphogenetic protein (proliferative marker) secreted by the stromal cell [60]. Massie and colleagues [84] have reported that human limbal fibroblast (hLF) (quiescent) was quite safe to be transplanted with hLESCs on the collagen biomatrix. hLF was less activated in terms of stroma and basement membrane remodeling and did not progress toward a scarring-like phenotype compared to diseased fibroblasts (dFib).

Another molecule that is important for corneal wound healing process is ascorbic acid. According to Chen and colleagues [81], the presence of L-ascorbic acid 2-phosphate (A2-P), the stable form of derivative ascorbic acid increased the stemness of mouse corneal epithelial stem/progenitor cells (TKE-2) by regulating the ECM components (collagen). A2-P also enhanced the stemness and proliferation of TKE2. SCF/C-Kit are another biomolecule that is present in normal mouse cornea and plays an important role in promoting corneal wound healing. According to Miyamoto and colleagues [89], SCF/C-kit enhanced cell attachment to FN, laminin and collagen type IV during corneal wound healing via the induction of the avidity and affinity of integrin members in vitro.

In conclusion, corneal stromal/keratocyte cells promoted CECs or LESCs growth on the collagen biomatrix. The presence of other components including FN, ascorbic acid, and SCF/C-kit enhanced the biocompatibility of collagen biomatrix towards CECs or LESCs.

4.5. Collagen as a Subtitute for Biomatrix

A study by Chakraborty and colleagues [76] reported that collagen IV coated surface improved LESCs attachment, growth, and proliferation when compared to the untreated plastic surface. Many researchers have ventured into the benefit of biocompatibility properties of collagen type IV as the main component of the biomatrix. This includes incorporating or coating the collagen IV into another biomatrix to develop the optimal corneal biomatrix design.

De la Mata and colleagues [85] studied the biological properties of collagen IV by incorporating the collagen into a poly-L/DL- lactic acid (PLA) biomatrix. Functionalizing PLA films with collagen IV (70:30) improved LESCs attachment, selection, and enrichment and maintained the undifferentiated LESCs phenotype with homogenous polygonal morphology [85]. This finding was supported by Wright, De Bank, Luetchford, Acosta and Connon [40] who incorporated collagen IV into an oxidized alginate biomatrix. Incorporating collagen IV further enhanced the CECs viability and provide niche environment that supports the re-epithelization process and may serve as viable wound healing bandages for the damaged cornea.

Moreover, a study by Kayiran Celebier and colleagues [86] exploited the biological benefits of collagen I by incorporating collagen I into poly(lactide-co-glycolide) (PLGA) polymers loaded with naproxen sodium (NS). This incorporation of collagen I did not affect the degradation period or the mechanical strength, but improved hydrophilicity and enhanced the water uptake capacity compared to the plain biomatrix [64,124]. As a result, the PLGA biomatrix coated with collagen I improved CECs attachment, proliferation, and viability [86,125], whereas incorporation of collagen I coating the silk film also enhanced biomatrix biocompatibility toward CECs [87].

In conclusion, exploitation of the biological benefit of collagen IV or collagen I by incorporating it into another biomatrix can improve the biomatrix biocompatibility toward CECs or LESCs, which have a great potential to be further used in in vivo study and clinically.

5. Conclusions

In conclusion, various modification strategies have been done to optimize the collagen biomatrix for CECs and LESCs carriers in the treatment of corneal defects. After summarizing and comparing the modification strategies and their effects on the biomatrix for each study included in this review. Despite the variability in the methodology approach, the reviewer postulated that the modification of the collagen biomatrix overall improves its mechanical and biocompatibility properties toward CECs and LESCs. The presence of the new bioresource of collagen within these 10 years potential to be developed as a biocompatible biomatrix. Physical modification and crosslinking methods improved the physical strength of the collagen biomatrix and affects the behavior of cultivated CECs and LESCs. The incorporation of the corneal stromal/keratocyte, FN, ascorbic acid, laminin, and SCF/C-kit into the collagen biomatrix enhanced its biocompatibility towards CECs and LESCs. Some studies also exploited the biological benefit of collagen by incorporating it into another biomatrix greatly improving the biocompatibility of the biomatrix towards CECs and LESCs. Our findings field provide insight into the field of current modifications on collagen biomatrix for the corneal defect. We believe the results highlight the need for modification of collagen biomatrix and these ultimately support the translation of collagen biomatrix-based therapies to the clinical setting.