Concise Review: Comparison of Culture Membranes Used for Tissue Engineered Conjunctival Epithelial Equivalents

The conjunctival epithelium plays an important role in ensuring the optical clarity of the cornea by providing lubrication to maintain a smooth, refractive surface, by producing mucins critical for tear film stability and by protecting against mechanical stress and infectious agents. A large number of disorders can lead to scarring of the conjunctiva through chronic conjunctival inflammation. For controlling complications of conjunctival scarring, surgery can be considered. Surgical treatment of symblepharon includes removal of the scar tissue to reestablish the deep fornix. The surgical defect is then covered by the application of a tissue substitute. One obvious limiting factor when using autografts is the size of the defect to be covered, as the amount of healthy conjunctiva is scarce. These limitations have led scientists to develop tissue engineered conjunctival equivalents. A tissue engineered conjunctival epithelial equivalent needs to be easily manipulated surgically, not cause an inflammatory reaction and be biocompatible. This review summarizes the various substrates and membranes that have been used to culture conjunctival epithelial cells during the last three decades. Future avenues for developing tissue engineered conjunctiva are discussed.


Conjunctiva
Conjunctival epithelium is non-keratinized and is at least two cell layers thick [1]. The number of cell layers depends on the degree of conjunctival stretching [2]. The conjunctival epithelium consists of two phenotypically distinct cell types, stratified squamous non-goblet cells (90%-95%) and goblet cells (5%-10%) (Figure 1), in addition to occasional lymphocytes [3] and melanocytes. The conjunctival epithelium plays an important role in ensuring the optical clarity of the cornea by providing lubrication to maintain a smooth, refractive surface, and by producing mucins critical for tear film stability [4]. The conjunctiva also protects the eye against mechanical stress and infectious agents. It, furthermore, contributes water and electrolytes to the tear fluid [5]. The squamous cells produce cell membrane-tethered mucins, while the goblet cells secrete the gel-forming mucins, both of which helps to maintain a protective tear film. The superficial surface of the squamous cells are covered by the membrane-tethered mucins mucin-1 (MUC1), mucin-4 (MUC4) and mucin-16 (MUC16) [6], which are essential for tear stability and make up the glycocalyx [6]. Original magnification of the HE photomicrograph: ×630. Immunofluorescence photomicrographs of forniceal conjunctival sections show conjunctival epithelial cell markers, which include the goblet cell markers anti-cytokeratin 7 (Ck7), Ulex europaeus agglutinin 1 (UEA-1) lectin and anti-mucin 5AC (MUC5AC), as well as the marker for stratified squamous non-goblet cells anti-cytokeratin 4 (Ck4). Nuclei were stained with DAPI (blue). Ck7 is expressed in the goblet cell body, whereas UEA-1 and MUC5AC stain the goblet cell mucin-contents. Ck4 is only detected in squamous cells between goblet cell clusters. The basal membrane is indicated by the white dotted line. Scale bars: 100 μm. Adapted from Fostad et al. 2012 [7].
The gel-forming mucin-5AC (MUC5AC) and mucin-2 (MUC2) are secreted by goblet cells into the aqueous layer of the tear film [8,9] (Figure 2). The squamous conjunctival cells also contribute to the hydration of the ocular surface through ion transport across the apical cell membrane with accompanying osmotic water transfer [5]. Goblet cells contain mucin-granules and have traditionally been identified through their secretory product using markers, including the ulex europaeus agglutinin-1 (UEA-1) lectin, anti-mucin-5AC (MUC5AC) and anti-AM3 antibodies, and periodic acid-Schiff (PAS) reagent that target the goblet cell gel-forming mucins [10]. In addition to cytokeratin 4 (Ck4) (Figure 1), squamous conjunctival epithelial cells can be identified by Ck13, a binding pair of Ck4 [11].
Stem cells are surrounded and influenced by a three-dimensional microenvironment known as a niche [27]. The niche comprises of numerous components, including stromal cells, soluble factors, extracellular matrix (ECM), mechanical/spatial cues and signaling molecules that dictates stem cell function [28]. The limbal stem cell niche has been reported to contain specific ECM proteins. Moreover, the specific composition of the ECM shows topographical variations throughout the ocular surface [29]. Thus, the specific composition of the ECM in the substrate may affect the preservation of conjunctival stem cells in culture.

Conjunctival Scarring Diseases
A large number of disorders can lead to scarring of the conjunctiva through chronic conjunctival inflammation. Scarring varies in severity and can be self-limited, such as in chemical/thermal burns and infectious diseases due to adeno-and herpes viruses, or progressive, as in cicatrizing conjunctivitis, which consists of several diseases including ocular cicatricial pemphigoid, Stevens-Johnson syndrome, atopic keratoconjunctivitis and Sjögren's syndrome [30]. Cicatrizing conjunctivitis is rare, and in total these disorders have an incidence of 1.2 in 1 million in the United Kingdom [30]. Treatment depends on the disease etiology and severity, but can include various anti-inflammatory, immunomodulatory and immunosuppressive drugs [31].
Surgical treatment of symblepharon includes removal of the scar tissue to reestablish the deep fornix [32]. The surgical defect is then covered with a tissue substitute to prevent re-obliteration. These include mechanical [33], physical [34] or chemical [35] approaches and the grafting of conjunctival or mucous membranes [32]. Surgical techniques for restoration of a diseased conjunctiva have utilized different conjunctival substitutes, including conjunctival autografts [36]. An obvious limiting factor when using autografts is the size of the defect to be covered, as the amount of healthy conjunctiva is limited. These drawbacks have led scientists to develop tissue engineered conjunctival equivalents.

Tissue Engineered Conjunctival Equivalents
A tissue engineered conjunctival epithelial equivalent needs to be easily manipulated surgically, not cause an inflammatory reaction, be biocompatible and contain a mix of stratified squamous cells, goblet cells and undifferentiated cells. Unlike tissue engineered corneal equivalents, conjunctival equivalents do not need to be transparent, which increases the range of suitable culture membranes.
In addition to conjunctival epithelial cells (CEC) cultured on amniotic membrane (AM) [4], there is likely a wide range of cell types that can be used for developing a tissue engineered conjunctival equivalent. This assumption is based on multiple studies demonstrating successful restoration of the cornea with cultured non-limbal cells. Tissue engineered corneal equivalents share many of the same prerequisites as conjunctival equivalents, e.g., with regard to barrier function and tear film support. Besides limbal stem cells, corneal equivalents have been developed from oral mucosal epithelial cells [37,38], embryonic stem cells (ESC) [39], bone-marrow-derived mesenchymal stem cells (MSC) [40], immature dental pulp stem cells [41], hair follicle-derived stem cells [42] and umbilical cord lining stem cells [43]. For conjunctival reconstruction, epidermal keratinocytes have been cultured on AM and transplanted to restore the conjunctiva in rhesus monkeys [44]. Although the conjunctival stratified squamous cell markers MUC4 and Ck4 were present in the transplant, goblet cells were absent. In a recent study, a tissue engineered conjunctival equivalent was developed from cultured AM epithelial cells [45]. The conjunctival equivalent contained PAS-positive cells, indicative of goblet cells, and successfully restored the conjunctiva in a rabbit model. Transplants containing goblet cells could also be developed from nasal mucosa, which harbors goblet cells [46]. Thus, there are multiple possible cell sources for developing conjunctival equivalents, though no comparative studies have defined the optimal choice of donor cells.
A number of different substrates and membranes have been attempted for tissue engineering conjunctival epithelial equivalents. These can be categorized into: (1) biological membranes; (2) extracellular matrix protein-containing membranes; and (3) synthetic polymer membranes.

Biological Membranes
Seventy-six years after it was first used in ophthalmology, AM, which constitutes the innermost layer of the fetal membranes, has a prominent role in ocular surface reconstruction [47]. AM is particularly suited for clinical use as it supports epithelialization [48], reduces scaring [49], suppresses the immune response [50], reduces pain, and decreases inflammation [51]. Prior to AM transplantation (AMT), the AM is cryopreserved, which kills all the AM cells [52]. Hence, AM grafts function primarily as a matrix and not by virtue of transplanted functional cells. The membranes have most commonly been cryopreserved in a basal cell medium at −80 °C [53], but a technique for freeze-drying the AM has also been developed [54]. Freeze-dried AM can be sterilized by gamma-irradiation [54], however, AM treated this way may release a less amount of growth factors than conventionally cryopreserved membranes [55]. In addition, the AM can be sterilized with per-acetic acid/ethanol and air-dried [56]. The latter technique is, on the other hand, reported to yield inferior results compared to cryopreserved AM with respect to rate of cell outgrowth, release of wound-healing factors, and preservation of the AM basement membrane (BM) [57]. In patients with chronic inflammation there is a tendency for recurrent shrinkage and symblepharon formation after restoring the ocular surface with AM [58]. The success of transplanting AM is therefore dependent on the underlying disease [4].
Twelve studies have described culture of CEC on AM, of which eight used denuded AM (dAM) ( Table 1). Meller et al. first reported the use of dAM for cell culture of CEC since they noticed that the devitalized AM epithelium inhibited adhesion and growth of the CEC [59]. All later studies using intact AM have utilized explant culture.
Eight out of ten studies confirmed the presence of goblet cells on AM (detected either by their mucin content or by Ck7), irrespective of whether dAM or intact AM had been used [59]. Data on actual percentages of goblet cells in CEC cultures on AM are sparse, although one study reported that between 25% and 75% of the cells were MUC5AC positive [60]. Although Ck7 positive goblet cells have been demonstrated under serum-free conditions, addition of 10% FBS improved the preservation of goblet cells [61]. This is in line with a study showing that FBS promotes expression of conjunctival epithelial cytokeratins due to the effect of vitamin A [62]. Development of mucin-containing goblet cells have also been achieved on AM independent of feeder cells, air-lifting or high calcium [60]. Thus, AM generally promotes goblet cell development.
Stratified CEC were obtained in all studies using AM, except one [63]. Culture techniques to induce stratification include the use of explants, air-lifting, feeder layer, and high calcium. Air-lifting promotes cell polarity by increasing the number of microvilli, tight junctions, and hemidesmosomes in CEC cultures [59]. The molecular mechanisms involved in air-lifting include the p38 mitogen-activated protein kinase and Wnt signaling pathways [64]. Stratification was achieved when including a feeder layer [50], air-lifting [59] and/or high calcium [65] in cell cultures.
Stratified CEC cultures were also generated on cadaveric acellular dermis (AlloDerm) coated with collagen type 4 (COL4) [65]. The latter study employed a serum-free culture protocol without feeder cells. Goblet cells, however, were not reported. Hence, except for the latter study on acellular dermis, culture of CEC on biological membranes generally promotes stratified cultures with goblet cells.
Goblet cells were seen when CEC were grown inside a collagen gel [78], but not always when grown as a monolayer on top of the collagen gel [78]. Compared to Matrigel, CEC grown on COL1 expressed more MUC5AC RNA than Matrigel cultures [76]. Five percent PAS positive goblet cells were detected when culturing CEC on top of a COL1:COL3 mix in serum-free medium [71]. The latter study also achieved stratification. When cultured without feeder cells, air-lifting or high calcium, the CEC formed monolayer cultures on COL1 [66]. Stratified cultures were achieved with the addition of feeder cells [67], air-lifting [67], or high calcium [76].
Use of fibronectin, either alone or in a mix with COL1, was reported in four studies [69][70][71][72]. The CEC formed monolayer cultures [70], but the presence of goblet cells were not reported. Elastin-like polymer has been used to grow Ck7 positive cells of the cell line IOBA-NHC [79]. Gelatin-chitosan yielded stratified cultures with Ck4 positive squamous cells when using explant culture [77]. Of all the ECM protein substrates, collagen gels and compressed collagen appear the most useful for conjunctival tissue engineering due to their mechanical properties and potential promotion of goblet cell formation.

Synthetic Polymer Membranes
Included in this group are polymers of glycolic acid, lactic acid, ε-caprolactone, 1,3-trimethylene carbonate, ethyl acrylate, hydroxyethyl acrylate, and methacrylic acid. One of the benefits of using these polymers is that several of them, including poly(L-lactide-co-glycolide) (PLGA) and poly(ε-caprolactone) (PCL), are already approved by the Food and Drug Administration (FDA) for the use in the human body for specific applications. In addition, the biodegradability of these polymers can be adjusted by controlling the ratio and choice of polymers. For instance, PLGA degrades faster than PCL. Furthermore, in contrast to biological membranes, synthetic membranes can be manufactured under sterile conditions, thereby considerably reducing the risk of transferring infectious agents to the patient. Although biodegradable polymers have been investigated at length with various types of cells, only four studies reported biocompatibility with cultured CEC [80][81][82][83] (Table 3). Three of these explored growth of CEC on polymer substrates [80][81][82], whereas one investigated the toxicity of polymer extract on cells cultured on plastic [83]. One of the studies confirmed the presence of MUC5AC positive goblet cells of comparable density to that seen when culturing CEC on AM [80]. The remaining studies did not report presence of goblet cells. The extract study showed lowest to highest viability with 50:50 poly(DL-lactide-co-glycolide) (PDLGA); 85:15 PDLGA and Inion GTR TM , respectively [83]. In cell growth studies, substrates with all three polymers demonstrated high viability [82]. Equally high viability was also seen when growing CEC on poly(ethyl acrylate-co-hydroxyethyl acrylate) (P(EA-co-HEA)) copolymers or 90:10 poly(ethyl acrylate-co-methacrylic acid) (P(EA-co-MAAc)) copolymers [81]. Interestingly, the latter two polymer substrates showed increased adhesion, proliferation and viability when hydrophobicity was increased. In contrast, Ang, et al. demonstrated increased proliferation when decreasing hydrophobicity of their PCL membranes [80]. The latter authors also obtained stratified cultures, which became more stratified by increasing surface hydrophilicity with NaOH. Thus, surface modification of synthetic polymer membranes can affect adhesion, proliferation, viability and stratification. Obvious advantages of synthetic polymer membranes include existing FDA approval for specific uses in the human body, high mechanical strength and biodegradability.

Comparative Studies of the Effect of Different Substrates on Cultured Conjunctival Epithelial Cells
In 2010, Rama and associates described the importance of the phenotype for clinical success following transplantation of cultured limbal epithelial cells [84]. p63, which is a marker for undifferentiated cells, was a significant predictor of clinical outcome [84]. It is possible that the phenotype of cultured CEC will determine success following transplantation of CEC. Comparative studies on how various substrates affect the cell sheet with regard to the phenotype in particular are, therefore, warranted.

Storage and Transportation of Cultured Conjunctival Epithelial Cells
With steadily stricter regulations for cell therapy, which lead to centralization of culture units [85], storage technology of cultured CEC has become increasingly important to allow the tissue to be transported to eye clinics worldwide [86]. Keeping in mind the significance of the phenotype for clinical outcome [84], assessment of the phenotype among other parameters prior to surgery should ideally be performed during the storage period. Moreover, storage in a hermetically sealed container enables microbiological assessment [87]. Finally, storage technology has the advantage of offering increased flexibility in scheduling surgery [88]. Comparative studies on how various substrates influence the ability to store cultured CEC with regard to morphology, viability, and phenotype should be performed to enable worldwide access to cultured CEC.

Conclusion
Amniotic membrane is the most commonly used substrate for CEC culture. The majority of the studies demonstrated that AM support the growth of goblet cells, in contrast to several alternative substrates. A major weakness in the current literature is the lack of comparative studies, thus such studies should be prioritized to be able to identify the most ideal substrate for ocular surface repair. Considering the disadvantages inherent to the use of a foreign biological material such as AM, clinical studies involving alternative membranes should be carried out as currently only AM has so far been used for transplanting tissue engineered CEC in humans.