Indian J Med Res 128, August 2008, pp 149-156

149

Human amniotic membrane (HAM) is an importantsubstratum for the growth of corneal stem cells and alsofor ocular surface reconstruction in clinical conditionswith partial and total limbal stem cell deficiency(LSCD)1-6. HAM is preferred because it facilitatesepithelialization and maintains an epithelial phenotype

by the basement membrane and has anti-inflammatory,anti-angiogenic and anti-scarring effects7,8.

There are two different methods in the preparationof HAM; HAM with its epithelium removed (denudedHAM)9 and intact HAM10,11 (i.e., retaining devitalized

Putative stem cell markers in limbal epithelial cells cultured onintact & denuded human amniotic membrane

Balasubramanian Sudha, Guruswamy Sitalakshmi*, Geetha Krishnan Iyer* & Subramanian Krishnakumar

L&T Department of Ocular Pathology, Vision Research Foundation & *Department of Cornea ServicesMedical Research Foundation, Sankara Nethralaya, Chennai, India

Received July 10, 2007

Background & objectives: The ocular surface is an ideal region to study the epithelial stem cell (SC)biology because of the unique spatial arrangement of stem cells and transient amplifying cells. Amajor challenge in corneal SC biology is the ability to identify SC in vitro and in situ, and one of themajor controversies in the field relates to reliable SC markers. This study was carried out to evaluateand compare the expression of the stem cell associated marker: ABCG2, keratinocyte stem cellmarker: p63 and corneal differentiation markers: Cnx43 and K3/K12 on limbal explants culturedon human amniotic membrane (HAM) with intact epithelium and HAM denuded of its epithelium.Methods: Human limbal biopsies obtained from the cadaveric donor eyes were used in this study.The cells were cultured over the HAM with intact and denuded epithelium. Reverse transcriptasePCR, immunohistochemistry, Western blotting for ABCG2, P63, Cnx43 and K3/K12 were done.Results: The limbal epithelial cells cultured over intact HAM expressed the stem cell associatedmarkers (ABCG2, p63) and showed reduced expression of the differentiation markers (Cnx43 andK3/K12) when compared to limbal epithelial cells cultured over denuded HAM, which expressedmore differentiation markers at the end of three weeks. BrdU label retaining cells were observed inthe limbal epithelial cells cultured over HAM with epihelium only.Interpretation & conclusions: Our results showed that the intact HAM supported the growth oflimbal epithelial cells expressing stem cell associated markers, and allowing little differentiation ofthe limbal cells to cornea phenotype. Further studies are needed to understand the properties of theamniotic epithelium that retains the stemness in the cultured limbal stem cells.

Key words ABCG2 - amniotic epithelium - Cnx43 - human amniotic membrane - K3/K12 - limbal phenotype

HAM epithelium) are utilized for corneal stem cellcultures as a substrate. There are numerous studies11-13

showing the advantages of ex vivo expanding of thelimbal explants on the intact HAM as it provides a morefavourable microenvironment for expansion of cornealstem cells, present in the limbus.

Earlier studies10-13 showed that limbal epithelial cellsexpanded on intact HAM retained their limbalphenotype, i.e., they express stem cell associatedmarker: ABCG2, keratinocyte stem cell marker: p63and are negative for the gap junction markers: connexin26 and connexin 43 (Cnx26, Cnx43), which allows thecells to be segregated from the transient amplifyingcells14, and also are negative for the cornea phenotypemarkers, keratin3/keratin12 (K3/K12)11,13.

ABCG2 is a member of the ATP-binding cassette(ABC) family of cell surface transport proteins thatincludes more than 50 members and mediates thetransfer of a diverse array of substrates across cellularmembranes. ABCG2 expression occurs in a variety ofnormal tissues and is relatively limited to primitive stemcells. De Paiva et al15 identified this transporter proteinin the population of the clonogenic human limbalepithelial cells. There is not much information availableon the expression of the stem cell associated markerABCG2 on the limbal epithelial cells expanded overintact HAM and denuded HAM.

Therefore in this study, we investigated theexpression of the stem cell associated marker, ABCG2on the limbal epithelial cells expanded on thecryopreserved intact and epithelially denudedmembrane. In addition, we also compared the p63,Cnx43 and K3/K12 expressions on these cultured limbalepithelial cells over the intact and denuded HAM.

Material & Methods

Grading donor eyes: Eye tissues were obtained fromthe C.U. Shah eye bank of Medical ResearchFoundation, Sankara Nethralaya, Chennai, with theconsent of donor or donor family. This study protocolwas approved by the institutional ethics committee.

Corneal limbal biopsy of 2mm3 from the cadavericdonor eye was collected in Dulbecco’s minimumessential medium (DMEM) with 3 per cent fetal calfserum (FCS) and antibiotics as the transport mediumand was transported for further processing. The donorblood samples were screened for humanimmunodeficiency virus (HIV) type1 and 2, hepatitisB virus (HBV), hepatitis C virus (HCV) and Treponema

pallidum infections. Data have been collected on age,sex, cause of death, time of death, time of eye donationand time of biopsy collection.

Chemicals and reagents: Slightly modified DMEM,F12, Hanks balanced salt solution (HBSS), HEPESbuffer, amphotericin B, and foetal bovine serum (FBS)were purchased from Hi-Media, Mumbai, India. Themouse monoclonal IgG antibody against p63 (1:100)(clone 4A4) was obtained from Santa Cruz, USA, andABCG2 (1:50) (clone BXP 21) was obtained fromChemicon, USA and FITC-conjugated goat anti-mouseIgG and IgM antibodies from LSAB, Dakocytomation,and Glostrup, Denmark. Hydrocortisone, epidermalgrowth factor (EGF), insulin-transferrin-sodium selenitemedia supplement were all from Sigma Chemicals,USA. ABC kit was obtained from DAKO, USA. Thetissue culture plastic plates and culture plate inserts werefrom Becton Dickinson (Lincoln Park, NJ). RNAextraction kit and cDNA conversion (sensiscript reversetranscriptase kit) were obtained from Qiagen, Germany.The specific primer sequences were obtained fromSigma Chemicals, USA; proprep protein extraction kitfrom InTron Biotechnology, USA, and Brdu labelingkit from Roche, Germany.

Preperation of human amniotic membrane (HAM): HAMwas kindly provided by Vijaya Health Centre, Chennai,India at the time of cesarean section, after proper informedconsent. The HAM was processed as describedpreviously16 and stored at -80°C for at least three months.HAM was devitalized by freezing and thawing andwashed three times with HBSS before being fastenedonto a 12 well culture insert as previously described17.

Preparation of human corneolimbal tissue: Limbalexplants were used from 20 donors ranging from 8-85 yrof age. After removal of excessive sclera, cornealendothelium, conjunctiva, and Tenon’s capsule, thelimbal ring was separated by a 7.5 mm trephine fromdonor corneas that were obtained from the MedicalResearch Foundation, Chennai, and were of transplantquality but had been excluded from clinical use fornonocular reasons.

Human limbal explants culture on intact and denudedHAM: HAM (3-4 mm/well) with the epithelial side facingupwards was fastened on the culture insert17. Similarly,denuded membrane was fastened with the basementmembrane side facing upward. The limbal tissue wasgently washed thrice with culture growth medium. Aftercareful removal of excessive sclera the tissue was cutinto multiple bits using sterile sharp curved scissors/Bard-

150 INDIAN J MED RES, AUGUST 2008

Parker blade. The bits were placed over the centre of thesubstrates in all the wells and about 0.5 ml of the limbalmedium (equal volume of DMEM and F12 supplementedwith 10 per cent FBS, 50 ng/ml of streptomycin, 1.25ng/ml of amphotericin B 2 ng/ml of mouse epidermalgrowth factor (EGF), 5 ng/ml of insulin, 5 ng/ml oftransferrin, 5 ng/ml of selenium, 5 mg of keratinocytegrowth supplement, 0.5 mg/ml of hydrocortisone) wasadded to cover the explants and plate was incubated at370C under 95 per cent humidity and 5 per cent CO2 for30-45 min. After incubation remaining 1.5 ml of themedium was added to cover the entire well completely.The medium was changed once in two days and cellgrowth was monitored daily for three weeks with aninverted phase contrast microscope (Nikon, Tokyo,Japan). Confluent cells were harvested for furthermolecular characterization and all the experiments wereperformed in triplicates.

Cell proliferation assay: Thymidine or BrdU labellinghas been successfully used to identify ‘label retaining’stem cells that are slow cycling or mitoticallyquiescent18,19. Cell proliferation kinetic study wasassessed by measuring 5-bromo-2-deoxyuridine (BrdU-Roche applied sciences, Germany) incorporation duringDNA synthesis in proliferating cultured cells overvarious substrates with the outgrowth reached 5-8 mmin diameter. The staining was done according to themanufacturer’s instructions and chased for 1-21 days.The BrdU labelling indices were assessed by countingthe nuclei through a microscope using 40X objective.The labelling index was expressed as the number ofpositively labelled nuclei/the total number of nuclei X100 per cent.

RNA isolation and RT- PCR: Total RNA was isolatedusing Qiagen-kit according to the manufacturer’srecommended protocol from the cultured cells at theend of 21 days of incubation. cDNA conversion wasdone with sensiscript RT (Qiagen) to see the expressionof different putative stem cell markers specific forcorneal limbal stem cells by reverse transcriptase PCR(RT-PCR). Glyceraldehyde-3-phosphate dehydrogenase(GAPDH), as an internal control, the mRNA expressionof different molecular markers was analyzed withrespective annealing temperature as shown in the Tableby semiquantitative RT-PCR20-22. PCR products werefractionated by electrophoresis using 2 per cent agarosegel (SRL, India) containing 0.5 per cent ethidiumbromide (Sigma, India) with molecular marker HinfIϕX digest to confirm the size of the resultant productof the amplification curve. The semiquantitation was

done with Quantity G software in Bio-Rad geldocumentation system. The fidelity of the RT-PCRproducts was verified by comparing their size with theexpected cDNA bands and by sequencing the PCRproducts.

Immunohistochemistry: Immunostaining for the p63 andABCG2 were done on the cultured cells at the end ofthe 21 days. The limbal epithelial cells cultured overintact and denuded HAM were fixed in neutral bufferedformaldehyde, processed in graded alcohol and xyleneand embedded paraffin wax. The tissue sections weretaken in poly L-lysine slides. The slides weredeparaffinized and antigen retrieval was done by thetrypsin digestion method23. Sections of samples werewashed in PBS and incubated for 1 h with the p63antibody clone 4A4 (1:50 dilution) and ABCG2antibody clone 5D3 (1:40), or with 1 per cent BSA-PBSas a negative control. After washing, the slides wereincubated with biotinylated anti-mouse immunoglobulinfor overnight, washed again, and incubated withhorseradish peroxidase-conjugated streptavidin(dilution as per the instruction manual) for 30 min. Thereaction was revealed by 3,3’-diaminobenzidine andcounterstained with hematoxylin. For the negativecontrol, the primary antibody was omitted andimmunostaining was performed.

Western blotting: The confluent cells on the 21st daywere harvested and protein was extracted. Proteins wereseparated by SDS-PAGE, blotted onto nitrocellulosemembrane and probed with the antibodies dilutionrecognizing anti p63 1:1000 (clone 4A4) and anti-ABCG2 dilution 1:500 (clone BXP21), and incubatedfor overnight at 4oC. Peroxidase conjugated anti-mouse

SUDHA et al: LIMBAL STEM CELL MARKERS EXPRESSION OVER INTACT/DENUDED HAM 151

Table. Primer sequence and reaction condition for the reverse transcriptasePCR

Gene Primer sequences Annealing PCRtemp., product

oC size, bp

∆Np63 FP: CAGACTCAATTTAGTGAGRP: AGCTCATGGTTGGGGCAC 54 440

ABCG-2 FP: AGTTCCATGGCACTGGCCATARP: TCAGGTAGGCAATTGTGAAGG 62 379

Connexin FP: CCTTCTTGCTGATCCAGTGGTAC43 RP: ACCAAGGACACCACCAGCAT 66 154K3 FP: GGCAGAGATCGAGGGTCTC

RP: GTCATCCTTCGCCTGCTGTAG 64 145K12 FP: CATGAAGAAGAACCACGAGGATG

RP: TCTGCTCAGCGATGGTTTCA 63 150GAPDH FP: GCCAAGGTCATCCATGACAAC

RP: GTCCACCACCCTGTTGCTGTA 63 498

Fig. 1. The growth of the limbal epithelial cells from the explants after 2 wk of incubation. 1A shows the growth on the intact HAM and 1B onthe denuded HAM. The expanded cells appear as a monolayer of small uniform with nucleus cytoplasmic ration of approximately 1:1 (20X).

Fig. 2. A significantly faster growth rate was observed on thedenuded membrane (solid line) than in intact membrane (dottedline). Limbal epithelial cells on the intact membrane started to growonly on the 3rd day compared to that of the denuded membranewhich started to growth by the end of 2nd day and after approximatelyof 3 wk they reached the confluence.

Fig. 3A. BrdU labelling index showing significantly higherpercentage of label retention on cells cultured over the intactmembrane at the end of 7th, 14th and 21st day (*P<0.05). Fig. 3B.Identification of label retaining cells in human corneal epithelialcultures. After 24 h of BrdU labeling at one week of early growthstage, 64.0 per cent on the cells cultured over the denudedmembrane (E) and 51 per cent on the cells cultured over the intactmembrane (A) the cultured limbal epithelial cells from the explantshave positively stained nucleus. In the cultures the labelling indexdecreased on the chasing. The index was 15.3, 10.9, 2.3 per centafter chasing for 7 (B), 14 (C), and 21 days (D) respectively. Thecells on the denuded membrane showed 7.8, 3.2, 0.6 per cent onthe 7th day (F), 14th day (G) and 21st day (H). Arrows indicate thepositive nucleus.

IgG as secondary antibody was added. The protein bandswere visualized using ECL detection kit.

Results

Under microscopic observation epithelial migrationwas noted from the limbal biopsies at the end of 48 h inintact and denuded HAM (Fig. 1). The cells culturedover the denuded HAM showed significantly a highergrowth rate than those on the intact HAM and reachedalmost a confluent growth after 21 days (Fig. 2). Thecultures on the denuded HAM showed growth from 2nd

day onwards whereas those on intact HAM began togrown only by 3rd day. By the end of the 15th day 90-100 per cent confluent growth was seen and cells wereharvested.

Cell proliferation assay: In this study, we labelled boththe cultures continuously for 24 h at their early growth

152 INDIAN J MED RES, AUGUST 2008

3A

3B

Inta

ctD

enud

ed

stage. The labelling index was high on denuded HAM(64.0 ± 7.76%) when compared to the intact HAM after24 h. The cultures were chased continuously at the endof day 7, 14 and 21. On the 7th day it was 15.3 ± 2.24per cent and on the 14th day it was 10.9 ± 2.0 per centand it was 2.3 ± 0.2 per cent on the 21st day on the cellscultured over the intact HAM. Similarly on the denudedmembrane the labelling index was 7.8 ± 3.2 per cent onthe 7th day, on 14th day it was 3.2 ± 0.33 per cent and onthe 21st day it was 0.6 ± 0.002 per cent (Fig. 3).

ABCG2 and p63 expression by immunohistochemistry:After 3-4 wk of culture, ABCG2 expression was seenonly on the cells cultured over the intact HAM whereasit was completely absent on the cells cultured over thedenuded HAM. Positive expression for ABCG2 showedthe staining of the cytoplasm on the basal cells on the

Fig. 4. The immunohistochemistry picture of the limbal epithelial cells cultured on the intact and denuded membrane harvested at the endof the third week. 4A shows the ABCG2 expression on the cells cultured over the intact membrane basal layer of cells showing the positiveexpression. 4B shows cells on the denuded membrane without expression. 4C shows p63 expression on the cells cultured over the intact.Positive cells are seen as the brown nuclei. 4D shows p63 staining on the denuded amniotic membrane.

cells cultured over the intact HAM. For the p63expression strong nuclear staining was seen on the cellscultured over both intact and denuded HAM (Fig. 4).

Western blot analysis: Expression of the p63 andABCG2 was confirmed with the Western blot. Proteinwas extracted from the cells cultured over the intactand denuded HAM at the end 3-4 wk of incubation.The cells cultured over the intact HAM showed theexpression of ABCG2 whereas the expression wascompletely absent on the cells cultured over the denudedHAM. p63 gave positive expression on both cellscultured over the intact and denuded HAM (Fig. 5).

RT-PCR: Semiquantitative RT-PCR results showed theexpression of various markers such as p63, ABCG2,Cnx43 and K3/K12 at the end of 21st day of cultivationon the cells grown over the intact and denuded HAM.

SUDHA et al: LIMBAL STEM CELL MARKERS EXPRESSION OVER INTACT/DENUDED HAM 153

4A

4C 4D

4B

Fig. 5. The Western blot data on the cells harvested at the end of3rd wk on incubation on the intact and denuded membrane. Thepositive expression of p63 was observed on both intact and denudedmembrane. ABCG2 expression was completely absent on the cellsharvested from the denuded membrane. Lane 1: Control (p63-SiHacell lysate; ABCG2-MCF7 cell lysate); Lane 2: cell cultured overintact HAM; Lane 3: cells cultured over denude HAM. β-actin asan internal control.

The expression of ∆Np63 was more on the cells culturedover the intact HAM when compared to the cells culturedover the denuded HAM (P<0.001). ABCG2 expressionwas observed only on the cells cultured over the intactHAM and it was completely absent on the cells culturedover the denuded HAM. The cells cultured over the intactHAM showed the less expression of the Cnx43 whereascells cultured over the denuded HAM showed highexpression of the cnx43 (P<0.05). Expression of K3/K12was high on the cells cultured over the denuded HAMand less on cells cultured on the intact HAM (Fig. 6).

Discussion

Amniotic membrane is the matrix, which iscurrently used for growing limbal epithelial cells andthen used for ocular surface reconstruction along withthe cultured cells11-13. HAM with intact epithelium hasbeen considered to be ideal for the growth of limbalepithelial cells and for ocular surface reconstruction24-

27. In this study, we compared the growth rate, expressionof putative stem cell markers, expression of cornealphenotype markers and presence of label retaining cellsover limbal epithelial cells cultured over HAM withintact and with denuded epithelium.

Our study showed that limbal epithelial cellscultured over the intact HAM had a slower growth ratewhen compared to that grown over the denuded HAM.

Fig. 6A. The semiquantitative RT-PCR results on the cells harvestedfrom the intact and denuded membrane. ∆Np63 and ABCG2expression was more on the cells cultured over the intact amnioticmembrane (*P<0.001). Cnx43, K3/K12 expression was more on thecells harvested from the denuded amniotic membrane. Fig. 6B. showsthe electrophoretogram of the semiquantitative RT-PCR. Lane 1: cellscultured over intact HAM; Lane 2: cells cultured over denuded HAM;Lane3: 100bp DNA ladder. GAPDH as an internal control.

154 INDIAN J MED RES, AUGUST 2008

Total of 20 limbal tissues were processed and theoutgrowth rates were measured at the end of every 2nd

day and the mean was plotted. The rate of growth ofcells on the denuded amniotic membrane seemed to behigher when compared to the cells cultured on the intactmembrane. Our results were consistent with the earlierresults10,12 where rabbit limbal and corneal explants wereused.

Immunohistochemical analysis showed a positiveABCG2 expression only in the basal cells on the intactHAM and not in the cells cultured over the denudedHAM. Semiquantitative RT-PCR showed that stem cellassociated marker ABCG2 was seen on the cellscultured over the intact membrane harvested at the endof 21 days, whereas it was completely absent on thecells cultured over the denuded membrane. Theexpression of ∆Np63 isoform of the keratinocyte stemcell maker p63 was expressed more on the cells culturedover the intact HAM compared to the cells cultured overthe denuded HAM. Similarly the expression ofdifferentiation markers i.e., gap junction marker: Cnx43

and cornea phenotype marker: K3/K12 were expressedin low levels on the cells over the intact membrane.Expression of p63 and ABCG2 was confirmed by theWestern blot. Thus our data were similar to that reportedearlier10-13,28-32.

Studies on labelling index showed that cells culturedover the denuded membrane had a higher rate ofproliferation and the chasing experiments revealed thatthe cells cultured over the intact membrane were ableto retain the label till the end of 21 days when comparedto that on the denuded membrane. The high labellingindex after 24 h indicated that BrdU was incorporatedinto the DNA during S phase in all mitotic cells,including the stem cells and transient amplifying cells.The decreased labelling index after different intervalsof chasing indicated that the BrdU-labelled transientamplifying cells (rapid cycling cells) were reduced innumber or disappeared, while the BrdU labelled stemcells still remained. Some of them had a cell cycle lengthof at least 21 days. The low labelling index noted afterchasing for 7-21 days was indeed the result of slowcycling cells.

In conclusion, amniotic membrane with intactepithelium allowed the persistence of the limbalepithelial cells, which express the putative stem cellmarkers, while the limbal epithelial cells cultured overamniotic membrane with denuded epithelium had moreof differentiated cornea phenotype cells. Further studiesare needed to understand the properties of the amnioticepithelium that allows the persistence of stem cells inthe limbal epithelial cells cultured over them.

AcknowledgmentAuthors acknowledge the Indian Council of Medical Research,

New Delhi for financial support.

References1. Dogru M, Tsubota K. Current concepts in ocular surface

reconstruction. Semin Ophthalmol 2005; 20 : 75-93.2. Gomes JA, Romano A, Santos MS, Dua HS. Amniotic

membrane use in ophthalmology. Curr Opin Ophthalmol 2005;16 : 233-40.

3. Koizumi N, Inatomi T, Suzuki T, Sotozono C, Kinoshita S.Cultivated corneal epithelial stem cell transplantation in ocularsurface disorders. Ophthalmology 2001; 108 : 1569-74.

4. Tseng SCG, Tsubota K. Amniotic membrane transplantationfor ocular surface reconstruction. In: Holland EJ, Mannis MJ,editors. Ocular surface diseases: Medical and surgicalmanagement, Sringer, 2001.

5. Grueterich M, Espana EM, Tseng SC. Ex vivo expansion oflimbal epithelial stem cells: amniotic membrane serving as astem cell niche. Surv Ophthalmol 2003; 48 : 631-46.

6. Tsai RJF, Li L-M, Chen J-K. Reconstruction of damagedcorneas by transplantation of autologous limbal epithelial cells.N Engl J Med 2000; 343 : 86-93.

7. Schwab IR. Cultured corneal epithelia for ocular surfacedisease. Trans Am Ophthalmol Soc 1999; 97 : 891-986.

8. Lee S-H, Tseng SCG. Amniotic membrane transplantation forpersistent epithelial defects with ulceration. Am J Ophthalmol1997; 123 : 303-12.

9. Meller D, Pires RTF, Tseng SCG. Ex vivo preservation andexpansion of human limbal epithelial stem cells on amnioticmembrane cultures. Br J Ophthalmol 2002; 86 : 463-71.

10. Grueterich M, Espana E, Tseng SC. Connexin 43 expressionand proliferation of human limbal epithelium on intact anddenuded amniotic membrane. Invest Ophthalmol Vis Sci 2002;43 : 63-71.

11. Grueterich M, Tseng SC. Human limbal progenitor cellsexpanded on intact amniotic membrane ex vivo. ArchOphthalmol 2002; 120 : 783-90.

12. Koizumi N, Fullwood NJ, Bairaktaris G, Inatomi T, KinoshitaS, Quantock AJ. Cultivation of corneal epithelial cells on intactand denuded human amniotic membrane. Invest OphthalmolVis Sci 2000; 41 : 2506-13.

13. Hernandez Galindo EE, Theiss C, Steuhl KP, Meller D. Gapjunctional communication in microinjected human limbal andperipheral corneal epithelial cells cultured on intact amnioticmembrane. Exp Eye Res 2003; 76 : 303-14.

14. Wolosin JM, Xiong X, Schutte M, Stegman Z, Tieng A. Stemcells and differentiation stages in the limbo-corneal epithelium.Prog Retinal Eye Res 2000; 19 : 223-55.

15. de Paiva CS, Chen Z, Corrales RM, Pflugfelder SC, Li DR.ABCG2 Transporter identifies a population of clonogenichuman limbal epithelial cells. Stem Cells 2005; 23 : 63-73.

16. Lee S-H, Tseng SCG. Amniotic membrane transplantation forpersistent epithelial defects with ulceration. Am J Ophthalmol1997; 123 : 303-12.

17. Meller D, Dabul V, Tseng SC. Expansion of conjunctivalepithelial progenitor cells on amniotic membrane. Exp EyeRes 2002; 74 : 537-45.

18. Cotsarelis G, Cheng SZ, Dong G, Sun TT, Lavker RM.Existence of slow-cycling limbal epithelial basal cells that canbe preferentially stimulated to proliferate: implications onepithelial stem cells. Cell 1989; 57 : 201-9.

19. Lauweryns B, van den Oord JJ, Missotten L. The transitionalzone between limbus and peripheral cornea. Animmunohistochemical study. Invest Ophthalmol Vis Sci 1993;34 : 1991-9.

20. Li DQ, Tseng SC. Three patterns of cytokine expressionpotentially involved in epithelial-fibroblast interactions ofhuman ocular surface. J Cell Physiol 1995; 163 : 61-79.

21. Li DQ, Lokeshwar BL, Solomon A, Monroy D, Ji Z,Pflugfelder SC, et al. Regulation of MMP-9 production byhuman corneal epithelial cells. Exp Eye Res 2001; 73 :449-59.

22. Chen Z, de Paiva CS, Luo L, Kretzcr FL, Plugfelder SC, LiDQ, et al. Characterization of putative stem cell phenotype inhuman limbal epithelia. Stem Cells 2004; 22 : 355-66.

23. Shi SR, Cote RJ, Taylor CR. Antigen retrievalimmunohistochemistry: past, present, and future. J HistochemCytochem 1997; 45 : 327-43.

SUDHA et al: LIMBAL STEM CELL MARKERS EXPRESSION OVER INTACT/DENUDED HAM 155

24. Li DQ, Tseng SC. Three patterns of cytokine expressionpotentially involved in epithelial-fibroblast interactions ofhuman ocular surface. J Cell Physiol 1995; 163 : 61-79.

25. Goodson WH III, Moore DH, Ljung BM, Chew K, FlorendoC, Mayall B, et al. The functional relationship between invivo bromodeoxyuridine labeling index and Ki-67 proliferationindex in human breast cancer. Breast Cancer Res Treat 1998;49 : 155-64.

26. Schwab IR, Reyes M, Isseroff RR. Successful transplantationof bioengineered tissue replacements in patients with ocularsurface disease. Cornea 2000; 19 : 421-6.

27. Tsai RJF, Li L-M, Chen J-K. Reconstruction of damagedcorneas by transplantation of autologous limbal epithelial cells.N Engl J Med 2000; 343 : 86-93.

28. Koizumi N, Inatomi T, Suzuki T, Sotozono C, Kinoshita S.Cultivated corneal epithelial stem cell transplantation in ocularsurface disorders. Ophthalmology 2001; 108 : 1569-74.

29. Koizumi N, Inatomi T, Suzuki T, Sotozono C, Kinoshita S.Cultivated corneal epithelial transplantation for ocular surfacereconstruction in acute phase of Stevens-Johnson syndrome.Arch. Ophthalmol 2001; 119 : 298-300.

30. Meller D, Pires RTF, Tseng SCG. Ex vivo preservation andexpansion of human limbal epithelial stem cells on amnioticmembrane cultures. Br J Ophthalmol 2002; 86 : 463-71.

31. Koizumi N, Rigby H, Fullwood NJ, Kawasak S, Tanioka H,Kinoshita S. Comparison of intact and denuded amnioticmembrane as a substrate for cell-suspension culture of humanlimbal epithelial cells. Graefe’s Arch Clin Exp Ophthalmol2007; 245 : 123-34.

32. Kim HS, Jun Song X, de Paiva CS, Chen Z, Pflugfelder SC,Li DQ, Phenotypic characterization of human cornealepithelial cells expanded ex vivo from limbal explant and singlecell cultures. Exp Eye Res 2004; 79 : 41-9.

Reprint requests: Dr Subramanian Krishnakumar, Department of Ocular Pathology, Vision Research FoundationSankara Nethralaya, 18 College Road, Chennai 600 006, Indiae-mail: drkrishnakumar_2000@yahoo.com

156 INDIAN J MED RES, AUGUST 2008