Molecular Characterization of Explant Cultured Human Oral Mucosal Epithelial Cells

Abstract

Purpose.: To culture and characterize oral mucosal epithelial cells (OMEC) grown on de-epithelialized human amniotic membrane (HAM) to explore their suitability as autografts in patients with bilateral ocular surface disease (OSD) and limbal stem cell deficiency.

Methods.: Oral biopsy samples were obtained from 20 patients undergoing oral reconstructive surgery, with informed consent and Institutional Ethics Committee approval. Morphologic studies, transmission electron microscopy (TEM), reverse transcriptase (RT) PCR and immunocytochemistry were used to characterize the OMEC.

Results.: Morphologic studies and TEM revealed a confluent sheet of proliferating, stratified oral epithelial cells connected to each other by desmosomes, containing intracellular cytokeratins and abundant mucin granules. These characteristics were further corroborated and elucidated by RT-PCR and immunocytochemistry. The presence of markers of differentiated, stratified epithelial cells (cytokeratin K3, K4, K13, and connexin 43), progenitor stem cell cell markers (p63, p75, β1-integrin/CD29, and ABCG2), and a variety of predominantly membrane-bound and a few gel-forming mucins (MUC 1, 5B, 6, 13, 15, and 16) was established.

Conclusions.: Cultured OMEC have the potential to act as autografts for ocular surface reconstruction in patients with bilateral ocular surface disease and can prove to be particularly beneficial to ameliorate the mucin deficiency state in dry eye associated with OSD.

The ocular surface is composed of three types of specialized epithelial cells: conjunctival, limbal, and corneal epithelium. ^1,2 A stable and healthy ocular surface facilitates clear vision and protects the eye from pathogenic organisms and xerosis. The basal layer of the limbal epithelium located in between the cornea and conjunctiva harbors stem cells, which have the ability to differentiate into corneal epithelial cells. ^3,4 Certain conditions like chemical and thermal injury, Stevens-Johnson syndrome (SJS), and ocular cicatricial pemphigoid (OCP) result in destruction and deficiency of limbal stem cells leading to ocular surface disease (OSD). OSD is marked by conjunctival invasion of the corneal surface, neovascularization, chronic inflammation, and stromal scarring leading to corneal opacification, loss of vision, and severe dry eye. ^5,6  

Therapeutic outcome of penetrating keratoplasty in patients with severe OSD is often poor due to underlying chronic inflammation and severe dry eye. Reconstructive surgery like keratoepithelioplasty and limbal stem cell transplantation (LSCT) have improved clinical outcome in these patients. ^7,8 LSCT has been proven to be an effective treatment modality, which is universally accepted and practiced ^{9 –11} However, in unilateral OSD, autologous LSCT or a whole tissue transplant technique like conjunctival limbal autograft (CLAU) may put the healthy donor eye at risk and in patients with bilateral OSD, LSCT from an allogenic donor or a keratolimbal allograft (KLAL) is associated with the risk of graft rejection, toxicity, and high cost of prolonged postoperative immunosuppressant use. ¹² Ex vivo cultured cell transplantation techniques are considered to be better than whole tissue transfer techniques especially CLAU, where graft size is a limiting factor. ¹² Thus, alternative transplant material such as oral mucosal epithelial cells (OMEC) and conjunctival cells can also be considered in bilateral OSD. ^13,14 As OMEC can be easily harvested, it is indeed a viable alternative compared with LSCT. ¹⁵ Different techniques and variations have been tried for the use of OMEC as grafts in OSD. ^{16 –18} Few studies have reported different molecular characteristics of OMEC but extensive studies including functional characterization are still awaited. ^14,17,19  

Mucin in the tear film, plays an active role in maintaining tear film stability. A qualitative and/or a quantitative decline in mucin composition of the tear film results in pathologic changes of the ocular surface. Mucins are high molecular weight glycosylated molecules found at the ocular surface, which have potential to attract and hold water because of their hydrophilic character. This allows them not only to act as a barrier to pathogenic infiltration but also maintain a smooth refractive ocular surface by lubricating it and preventing it from drying up. ²⁰ Very few studies have looked into the expression of mucins in OMEC. ^21,22  

The purpose of this study was to elucidate the molecular and ultrastructural characteristics of OMEC after culturing them on de-epithelialized human amniotic membrane (HAM) to explore their suitability as autografts in patients with bilateral ocular surface disease (OSD) and limbal stem cell deficiency. Although different scaffolds are available, we chose to grow OMEC on HAM as it has multiple advantages ^23,24 and can be grown without the support of 3T3 mouse fibroblast feeder layer. ¹⁸  

Materials and Methods

Preparation of HAM

In accordance with the tenets of the Declaration of Helsinki and with proper informed consent, HAM was obtained under sterile conditions after caesarian section delivery from a seronegative donor (HIV, hepatitis B surface antigen, hepatitis C virus, syphilis) with informed consent. HAM is being routinely prepared for tissue grafting for ophthalmic indications by the National Eye Bank, Dr Rajendra Prasad Centre for Ophthalmic Sciences, All India Institute of Medical Sciences (AIIMS), New Delhi. The placenta is washed thoroughly with normal saline containing antibiotics. The amniotic membrane is peeled off using blunt dissection and blood clots removed. It is cut into small pieces (4 cm × 4 cm) and placed on previously sterilized nitrocellulose membrane in sterile cryopreservative medium containing DMEM and glycerol (1:1) and frozen at −80°C. Cryopreserved HAM was thawed an hour before usage by immersing in a sterile water bath at 37°C for 20 minutes. After thawing, the HAM was treated with 0.25% Trypsin-EDTA at 37°C for 30 minutes to denude the epithelial lining of the HAM. The denuded or de-epithelialized HAM was used as a carrier or scaffold for OMEC explant tissue. 

Harvesting of Oral Mucosal Epithelial Tissue

Oral mucosal epithelial tissue (4 mm × 4 mm × 2 mm) was harvested from 20 patients undergoing reconstructive surgery for nonmalignant conditions in the Department of Oral and Maxillofacial Surgery, Centre for Dental Education and Research, AIIMS, New Delhi, after taking informed consent and Institutional Ethics Committee clearance. Preoperative assessment was performed by the dental surgeon to evaluate the feasibility of obtaining 4 mm × 4 mm tissue from the inner surface of the oral cavity (cheek) of the patient after appropriate antisepsis and anesthesia. Patients with malignant lesions of the oral cavity and those who were long-term tobacco users (either smokers or chewers) were excluded from the study. Twenty oral biopsies were used for the study. Routine dental surgery protocols were followed for oral antisepsis, which included pre- and postoperative maintenance of oral hygiene by regular brushing of teeth followed by cyclohexidine mouthwash, per operative 1% betadine mouth paint, and pre- and postoperative antibiotics. 

Oral Mucosa Epithelial Cell Culture (Explant Technique)

The oral biopsy specimen was transported to the stem cell laboratory in a sterile vial containing DMEM medium with antibiotics (penicillin, streptomycin, and amphotericin B) but without fetal calf serum (FCS). It was washed (five cycles of 5 minutes each) with normal saline containing antibiotics (penicillin, streptomycin, metronidazole, and amphotericin B). Explant culture technique was used for growing OMEC. A mechanical rather than enzymatic removal of epithelial cell sheet was done. Briefly, the oral mucosal layer was manually sliced off from adjoining connective tissue using a sterile surgical blade and then cut into small 1 mm ² pieces (explant tissue) and placed on de-epithelialized HAM. The de-epithelialized HAM was stretched out using an autoclaved plastic ring. The HAM and explant tissue were kept in a sterile petri plate in an incubator at 37°C with 95% air and 5% carbon dioxide. Explant tissue pieces were allowed to adhere to HAM for 1 to 2 hours before addition of media. Growth media contained DMEM and Ham's F12 nutrient media in the ratio 1:1 with antibiotics (penicillin, streptomycin, and amphotericin B), EGF (10 ng/mL), insulin (5 μg/mL), and FCS (10% vol/vol). Growth media was changed every third day. Oral mucosal epithelial cells (as explants) were grown on HAM for approximately 2 weeks (until confluence) while monitoring their growth under a phase contrast inverted microscope (Nikon, Tokyo, Japan) and photographs of growth and expansion of OMEC were taken with appropriate magnification. 

Cytopathology of Cultivated OMEC

Confluent, multilayered OMEC were mechanically detached from the HAM scaffold using a sterile cell scraper. The procedure was validated by visualizing the HAM scaffold under the microscope and it was confirmed that no cells remained adherent to the HAM after the procedure in any of the areas. The cells were washed with 0.1 M phosphate buffer (pH 7.4), fixed in a 3:1 acetic acid-methanol mixture and spread on a polylysine-coated glass slide. Multiple slides of a single specimen were prepared and stored at 4°C. Specimens were stained with hematoxylin-eosin (H&E). 

Transmission Electron Microscopy (TEM)

The OMEC growing on the HAM scaffold were gently washed in 0.1 M phosphate buffer (pH 7.4) twice and then fixed with Karnovsky's fixative in 0.1 M phosphate buffer (pH 7.4) for 2 hours at 4°C. After washing, these were postfixed in 1% osmium tetroxide for 2 hours at 4°C, dehydrated in ascending grades of acetone and embedded in araldite CY 212. Thin sections (70 nm) were cut with a glass knife and mounted onto nickel grids. They were contrasted with uranyl acetate and lead citrate and viewed under a transmission electron microscope (Philips CM10; Eindhoven, Holland). 

Reverse Transcription PCR (RT-PCR)

Total RNA was isolated from confluent cultures of cultivated OMEC after mechanically detaching them from the HAM scaffold using a sterile cell scraper (as described previously), in biological replicates, using reagent (Tri-Reagent; Sigma, St. Louis, MO). RNA was spectrophotometrically quantified and RT-PCR was set-up with 2 μg total RNA to yield cDNA. 5′ and 3′ gene-specific primers were designed against the coding region for the selected genes and synthesized (MWG Biotech, Cork, Republic of Ireland; Table 1). Synthesis of the first strand was carried out using MMLV-RT (MBI Fermentas, Glen Burnie, MD) and oligo (dT) primer. PCR amplification was carried out in a total volume of 25 μL using 1 μL cDNA, 10X PCR buffer, 10 mM dNTPs, 15 pmol/L of each gene-specific primer, 15 pmol/L of GAPDH primers and 0.75 U polymerase (Taq; Biotools, Madrid, Spain). After 5 minutes of initial denaturation, 35 amplification cycles of 1 minute at 94°C, 1 minute at 60°C, and 1 minute at 72°C were carried out followed by a 10-minute elongation at 72°C. GAPDH and β2-microglobulin housekeeping genes were used to check for integrity of cDNA. Molecular size of PCR product was determined with the help of 100 base pair ladder or other appropriate molecular weight marker. 

Table 1.

 

Parameters for RT-PCR Analysis of Different Genes in Cultivated OMEC

Serial Number	Gene	Primer Sequence	Annealing Temperature (°C)	Product Size (bp)
1.	Cytokeratin K3	F 5′ GTCCTGGAGACCAAGTGGAA 3′	60	177
		R 5′ CACCAGGTCCTCCATGTTCT 3′
2.	Cytokeratin K12	F 5′ ACATGAAGAAGAACCACGAGGATG 3′	55	150
		R 5′ TCTGCTCAGCGATGGTTTCA 3′
3.	Cytokeratin K4	F 5′ AGGAGGTCACCATCAACCAG 3′	58	228
		R 5′ GCTCAAGGTTTTTGCTGGAG 3′
4.	Cytokeratin K13	F 5′ CCAACACTGCCATGATTCAG 3′	58	228
		R 5′ TCTGGCACTCCATCTCACTG 3′
5.	Connexin 43	R 5′ TCTGGCACTCCATCTCACTG 3′	58	200
		R 5′ TCTTTCCCTTAACCCGATCC 3′
6.	p63	F 5′ CAGACTCAATTTAGTGAG 3′	55	440
		R 5′ AGCTCATGGTTGGGGCAC 3′
7.	β1-integrin (CD29)	F 5′ TTGCAAGTGTCGTGTGTGTG 3′	59	202
		R 5′ ACAGACACCAAGGCAGGTCT 3′
8.	ABCG2	F 5′ AGTTCCATGGCACTGGCCATA 3′	56	379
		R 5′ TCAGGTAGGCAATTGTGAGG 3′
9.	p75	F 5′ TGAGTGCTGCAAAGCCTGCAA 3′	61	230
		R 5′ TCTCATCCTGGTAGTAGCCGTAG 3′
10.	MUC1	F 5′ GTGCCATTTCCTTTCTCTGC 3′	60	207
		R 5′ GTAGGTGGGGTACTCGCTCA 3′
11.	MUC2	F 5′ AGCCCGGTTCTCCAGTTTAT 3′	57	200
		R 5′ GGGCCGTTTGATGATACAGT 3′
12.	MUC3/17	F 5′ TGCAGAACAGGACCTCAGTG 3′	59	204
		R 5′ GTCATCTCAGGGTTGGTGCT 3′
13.	MUC4	F 5′ CTTCAGATGCGATGGCTACA 3′	57	200
		R 5′ GTTTCATGCTCAGGTGCTCA 3′
14.	MUC5AC	F 5′ AACGTGAGCATACCCTGACC 3′	60	243
		R 5′ AAGACGCAGCCCTCATAGAA 3′
15.	MUC5B	F 5′ CACTTCCTCCTCCAGTCCAA 3′	60	198
		R 5′ GTGTGGGTGGTCTCTGGAGT 3′
16.	MUC6	F 5′ ACACTTCCCACTCACGTTCC 3′	59	197
		R 5′ CTGGTGGTCACTGTCATTGG 3′
17.	MUC7	F 5′ TCCTCACCAGCCACCTAAAC 3′	57	222
		R 5′ GGGGAATTCACTGGTGCTAA 3′
18.	MUC12	F 5′ CGTCAGTGAAGAATCCAGCA 3′	57	201
		R 5′ CTGCTGTGGGAAGTTGTTGA 3′
19.	MUC13	F 5′ AAATGCGTGCTGATGACAAG 3′	56	201
		R 5′ AACCATTGAGGCAGTCATCC 3′
20.	MUC15	F 5′ CATCTCAGCACATCCCAATG 3′	56	185
		R 5′ TGTTTGTGGTAAGCCATCCA 3′
21.	MUC16	F 5′ CTCTCAGCCTCCCAAGTGTC 3′	59	199
		R 5′ GTGTCCATGGTGGGGAATAC 3′
22.	MUC19	F 5′ CACTGCTTCCACAGGAGTCA 3′	57	199
		R 5′ CAATGGACCGTGACAAACTG 3′
23.	β2-microglobulin	F 5′ TAGCTGTGCTCGCGCTACT 3′	58	90
		R 5′ TCTCTGCTGGATGACGTGAG 3′
24.	GAPDH	F 5′ TGCACCACCAACTGCTTAGC 3′	57	297
		R 5′ TTTCTAGACGGCAGGTCAGG 3′

 F, forward; R, reverse.

Immunocytochemical Analysis

Expression of target proteins in the cultured OMEC was assessed by immunocytochemistry. OMEC fixed and stored for cytology were placed at room temperature and washed with 0.1 M phosphate buffer (pH 7.4). Endogenous peroxidases were blocked with hydrogen peroxide in PBS containing 70% methanol and nonspecific binding blocked using 5% bovine serum albumin (BSA). The cells were then incubated separately with primary antibodies against cytokeratin K3/K12, AE5 clone (1:100 dilution) (Chemicon, Billerica, MA), p63 4A4 clone (1:200 dilution) that recognizes all isoforms of p63 (Millipore, Billerica, MA). The anti-p75 antibody was obtained from Abcam (Cambridge, MA) and used at 1:200 dilution. Connexin-43 (Millipore) and β1-integrin clone (LM534; CD29; Millipore) primary antibodies were also used at a dilution of 1:100. All primary antibodies were incubated for 12–16 hours at 4°C. Immunodetection was achieved by an avidin-biotin horse radish peroxidase-based colorimetric method (Novocastra, Newcastle, UK) according to the manufacturer's protocol, with 3,3′-diaminobenzidine tetrahydrochloride (DAB) as a chromogen and H₂O₂ as the substrate, followed by light counterstaining with hematoxylin and examination under a microscope. Specimens were considered as immunopositive if at least 5% of the cells displayed distinct immunostaining. 

Results

Morphologic Features of OMEC

Cell migration and growth from explant edge was initiated within 2 to 3 days (Figs. 1A, 1B) and a multilayered confluent sheet of OMEC was formed on HAM in 1–2 weeks (Figs. 1C, 1D). Optimal transplantable cell growth took 1–2 weeks. On H&E staining, cultured OMEC at the explant edge were found to be small oval shaped cells with comparatively larger nuclei, some of which differentiated into classically large irregular polygonal epithelial cells with small nuclei (Figs. 1E, 1F). This demonstrated that a heterogeneous population of OMEC was growing on HAM. Conventional procedures of air-lifting and growing OMEC on 3T3 feeder layer ²⁴ were not followed, yet stratified confluent sheets of epithelial cells were seen. 

Figure 1.

 

(A, B) Morphologic findings of OMEC growing on HAM for 2–3 days. Magnification: (A) ×20; (B) ×100. (C, D) OMEC as a confluent sheet on HAM (after 1–2 weeks). Magnification: (C) ×20; (D) ×100. (E) OMEC (without HAM) (after 1–2 weeks) stained with H&E (magnification, ×200). (F) OMEC (with HAM) (after 1–2 weeks); stained with H&E as whole mount (on nitrocellulose membrane). Magnification, ×200.

TEM

TEM was useful in demonstrating the ultrastructural features of OMEC as shown in Figures 2A and 2B. Cultured OMEC predominantly formed desmosomes with each other. Few hemidesmosomes were seen at the basal layer of cultivated OMEC sheet, aiding in attachment of the cultured OMEC to HAM. Gap junctions or tight junctions could not be visualized with the help of TEM. Intracytoplasmic cytokeratin bundles were evident within the epithelial cell (Fig. 2C). Mucin granules, an ultrastructural feature of the cultured OMEC graft (Fig. 2D) were seen intracellularly. Figures 2E and 2F show presence of euchromatin and multiple nucleoli within the nucleus of the cells in the basal layers of the cell sheet respectively, indicating actively proliferating basal epithelium in the cultivated OMEC sheet. 

Figure 2.

 

Ultrastructural features of cultivated OMEC growing on HAM for 1–2 weeks under TEM. (A, B) Desmosomes in between epithelial cells (OMEC) seen under lower (A) and higher (B) magnification. (C) Intracytoplasmic cytokeratin bundles seen within the OMEC. (D) Intracellular mucin granules seen within the OMEC. (E) Euchromatin within the nucleus indicates proliferating state of the OMEC. (F) Multiple nucleoli within the nucleus showing actively proliferating basal epithelium.

Molecular Characterization of OMEC Using RT-PCR and Immunocytochemistry

Molecular markers for epithelial cells of the oral mucosa and cornea, putative markers of stem cells and some membrane-bound and secreted mucins were evaluated using RT-PCR. Cultured OMEC expressed cytokeratin K3 while expression of cytokeratin K12 could not be established in these cells (Fig. 3A, Table 1 and Table 2). OMEC expressed sufficiently high quantities of cytokeratin K4 and K13 (Fig. 3B, Table 1 and Table 2). These cells were also seen to express connexin 43 (Fig. 3A, Table 1 and Table 2). Cultured OMEC cells were found to express selective stem cell markers, such as p63, p75, β1-integrin (CD29), and ABCG2 (Figs. 3A, 3B, Table 1, Table 2). A variety of membrane-bound and gel-forming mucins (MUC 1, 2, 3/17, 4, 5AC, 5B, 6, 7, 12, 13, 15, 16, and 19) were assessed and the following mucins were expressed: MUC 1, 5B, 6, 13, 15, and 16, while some mucins were not expressed (2, 5AC, 3/17, 4, 7, 12, and 19) (Figs. 3A–C, Table 1, Table 2). Normalization and validation of housekeeping genes was done with GAPDH and β2-microglobulin to confirm the integrity of the cDNA (Figs. 3B, 3D, Table 1, Table 2). 

Figure 3.

 

RT-PCR analyses of different genes in OMEC isolated from confluent sheets of OMEC cultivated on HAM for 1–2 weeks. (A) M, marker; L1, cytokeratin K3; L2, cytokeratin K12; L3, connexin 43; L4, p63; L5, β1-integrin (CD29); L6, p75; L7, MUC1; L8, MUC3; L9, MUC5AC. (B) M, marker; L1, MUC1; L2, cytokeratin K4; L3, cytokeratin K13; L4, MUC4; L5, MUC12; L6, MUC13; L7, β2-microglobulin; L8, MUC19; L9, ABCG2; L10, MUC15. (C) M, marker; L1, MUC5B; L2, MUC6; L3, MUC7; L4, MUC12; L5, MUC16; L6, MUC2. (D) GAPDH (housekeeping gene).

Table 2.

 

Expression Patterns and Functional Roles of Different Genes and Proteins in Cultivated OMEC as Determined by RT-PCR and/or Immunocytochemistry

Serial Number	Gene	Functional Role/Group	RT-PCR	Immunocytochemistry
1.	Cytokeratin K3	Corneal and oral epithelial cell marker	+	+*
2.	Cytokeratin K12	Corneal epithelial cell marker	−	+*
3.	Cytokeratin K4	Marker of nonkeratinized stratified oral epithelium	+	nd
4.	Cytokeratin K13	Marker of nonkeratinized stratified oral epithelium	+	nd
5.	Connexin 43	Marker for differentiated epithelial cells	+	+
6.	p63	Stem cell marker in stratified epithelium	+	+
7.	β1-integrin (CD29)	Epithelial stem cell and progenitor cell marker	+	+
8.	ABCG2	Stem cell and progenitor cell marker	+	nd
9.	p75	Marker for corneal progenitor cells/stem cell marker in oral epithelium	+	+
10.	MUC1	Mucin expressed by epithelial cells (membrane bound mucin)	+	+
11.	MUC2	Mucin usually expressed by myoepithelial cells of trachea; intestine (secreted mucin)	−	nd
12.	MUC3/17	Mucin usually expressed by intestinal epithelium (membrane bound mucin)	−	nd
13.	MUC4	Mucin usually expressed by trachea-bronchial and conjunctival epithelium (membrane bound mucin)	−	nd
14.	MUC5AC	Mucin usually expressed by myoepithelial cells and goblet cells (secreted mucin)	−	nd
15.	MUC5B	Mucin usually expressed by trachea-bronchial, salivary, sublingual, cervical epithelium (secreted mucin)	+	nd
16.	MUC6	Mucin usually expressed by gastric epithelium (secreted mucin)	+	nd
17.	MUC7	Mucin usually expressed by salivary gland epithelium (secreted mucin)	−	nd
18.	MUC12	Nonspecific epithelial cell surface associated or membrane bound mucins	−	nd
19.	MUC13	Nonspecific epithelial cell surface associated or membrane bound mucins	+	nd
20.	MUC15	Nonspecific epithelial cell surface associated or membrane bound mucins	+	nd
21.	MUC16	Nonspecific epithelial cell surface associated or membrane bound mucins	+	nd
22.	MUC19	Gel forming or secreted mucin	−	nd
23.	β2-microglobulin	Universally expressed housekeeping gene	+	nd
24.	GAPDH	Universally expressed housekeeping gene	+	nd

 +, positive expression; −, no expression; nd, not done.

*  Cytokeratin K3 and K12 have a common antibody (AE5).

RT-PCR results were further corroborated by immunocytochemistry of the cultured cells for some proteins. Cytokeratin K3 and K12 had a common antibody (AE5) which gave positive results with cultured OMEC. As RT-PCR for K3 expression was positive while that for K12 was negative, we assume that the protein expression shown by the cells was due to K3 and not due to K12 (Fig. 4A, Table 2). A partial population of the cultured OMEC was also positive for connexin 43, believed to be a marker for differentiated epithelial cells thus again showing that the cultured cells comprised a heterogenous population (Fig. 4B, Table 2). Protein expression of stem cell markers, p63, p75, and β1-integrin (CD29) was also determined, all of which gave positive results in the OMEC (Figs. 4C–E, Table 2). The cell surface-associated mucin (MUC 1) was found to be expressed by immunocytochemistry (Fig. 4F, Table 2). 

Figure 4.

 

Expression of various marker proteins as assessed by immunocytochemistry in cultivated OMEC for 1–2 weeks. (A) Expression of cytokeratin K3/K12; magnification, ×400. (B) Expression of connexin 43; magnification, ×200. (C) Expression of p63; magnification, ×200. (D) Expression of β-1 integrin (CD29); magnification, ×200. (E) Expression of p75; magnification, ×200. (F) Expression of MUC1; magnification, ×200.

Discussion

As long-term sequelae of severe OSD are devastating, it is always beneficial to search for better therapeutic modalities especially in cases of bilateral OSD. The purpose of the present study was to establish the molecular, functional, and ultrastructural basis for autologous OMEC graft in patients with bilateral OSD. Our culture technique established time to confluence and optimal stratification to be approximately 2 weeks and circumvented use of the 3T3 feeder layer as well as the air-lifting technique. In our study, we have preferred to use HAM as a scaffold as it not only promotes growth and healing but also facilitates an easier and more successful transplant procedure. 

We found cultivated OMEC expressing cytokeratin K3, which is a marker for oral mucosal as well as corneal epithelial cells, but not cytokeratin K12, which is characteristic of corneal epithelium. This agrees with most other studies which stress that although OMEC resembles corneal epithelium, they do not differentiate into corneal epithelial cells. ^14,15,17,23 Concurrently, OMEC show presence of cytokeratin K4 and K13, which are distinctive of nonkeratinized stratified epithelial cells. This highlights the fact that OMEC have the ability to form a confluent stratified multilayered epithelial sheet similar to corneal epithelium, which forms the stable ocular surface essential for reconstructing the damaged ocular surface in patients with bilateral limbal stem cell deficiency. ^14,17,21  

For the first time, morphologic and molecular characteristics of the growing cell sheet in our study clearly identify the stratified epithelium and presence of a heterogeneous population of at least two kinds of cells comprising a small, rounded type of undifferentiated cells, with large nuclei that are rapidly proliferating, and differentiated epithelial cells with small nuclei and characteristic polygonal appearance. This finding also answers the pertinent question asked by many stem cell biologists and translational researchers as to whether transplant tissue should contain undifferentiated or differentiated cells. Our findings underline the need for a heterogeneous population of both differentiated and undifferentiated cells at the time of stem cell transplantation for optimum growth. 

In our study, the concomitant presence of connexin 43, a marker for differentiated epithelial cells as well as the presence of stem cell markers in stratified epithelium (p63, ABCG2, and β1-integrin) and stem cell marker in oral epithelium (p75) emphasize the presence of differentiated epithelial cells (K3, K12, Cx43) as well as undifferentiated stem cells in the cultivated OMEC sheet. Our findings are concordant with other published studies. ^16,19,25 The expression of both types of markers in in vitro culture conditions suggests that cultured OMEC contains heterogeneous populations of progenitor cells and matured epithelial cells. 

TEM findings in our study revealed the predominance of desmosomes in cell-to-cell contact in cultivated OMEC. This resembles normal corneal epithelial intercellular connections. Hemidesmosomes were visualized in the basal layer between cultivated OMEC and HAM, by which the OMEC is attached to HAM. We could not detect any tight junctions or gap junctions. However, presence of connexin 43, a marker for differentiated epithelial cells as well as a gap junction protein, indicates presence of gap junctions which could not be visualized by TEM. These findings agree with other TEM studies done on OMEC. ^14,21 Our study is the first to show the presence of multiple nucleoli as well as euchromatin within the OMEC nuclei, which underlines the proliferative capability of these cells, further establishing their stem cell characteristics. Another ultrastructural finding evident from our study is the presence of cytokeratin bundles in the cytoplasm of OMEC that corroborate the presence of these characteristic molecular filaments determined by RT-PCR and immunocytochemistry. 

For the first time, in our TEM experiments, we could detect the presence of intracellular mucin granules. This lends credence to our hypothesis that mucins are an essential component of inducing symptomatic benefit for dry eye when cultivated OMEC autografts are transplanted onto the ocular surface for its regeneration. This also substantiates the fact that conjunctival goblet cells are not the only source of lubricating mucins and the cultivated OMEC have the potential to synthesize the lubricating and protective mucins. TEM results of our study reaffirm that the stratified epithelial layer formed in the cultivated OMEC is quite similar to the cornea, which firmly establishes the capability of cultivated OMEC to form a stable ocular surface. 

One of the salient findings of our study is the expression of certain mucins, which are present not only on the surface of OMEC but also intracellularly. Mucins are believed to attract and hold water because of their hydrophilic character, thus preventing drying of the epithelial surface. ²⁰ We checked for the presence of a diverse range of mucins using RT-PCR. Our findings show that OMEC express MUC 1, 5B, 6, 13, 15, and 16 while they do not express MUC 2, 5AC, 3/17, 4, 7, 12, and 19. The results partially agree with another similar study. ²² Our results show that the cultivated OMEC predominantly express membrane-bound/cell surface-associated mucins (MUC 1, 13, 15, and 16), however, they also express MUC 5B and 6, both of which are secreted/gel-forming mucins. Our findings do not totally corroborate with findings of Hori et al., ²² as we could not detect expression of another membrane-bound mucin, MUC 4. This can be explained by the fact that an oral mucosal cell line, KB, has been found to express MUC 13 and 16, thus validating our findings. ²² Maybe the ethnic diversity (the patients in the study by Hori et al. ²² were Japanese) can account for the slight variation in expression. Our findings indicate that OMEC exerts its beneficial effect with the help of both membrane-bound mucins as well as secreted or gel-forming mucins. This will be of great value in reducing the dry eye conditions highly prevalent in patients with OSD and rationalizes its use as a graft in OSD. 

In conclusion, it appears that autologous OMEC graft can serve as a potential graft for bilateral OSD as the molecular and ultrastructural characteristics of this graft support the development of a stable ocular surface and amelioration of dry eye symptoms. If the dry eye can be controlled to a certain extent in these patients, it may not only alleviate the discomfort associated with this condition but may also improve vision and make keratoplasty a realistic goal in the future. 

Footnotes

 Supported by a DST Fast Track Fellowship, Department of Science and Technology, Ministry of Science and Technology, Government of India (SR/FT/L-110/2005) (SuS).

Footnotes

 Disclosure: S. Sen, None; S. Sharma, None; A. Gupta, None; N. Gupta, None; H. Singh, None; A. Roychoudhury, None; S. Mohanty, None; S. Sen, None; T.C. Nag, None; R. Tandon, None;