角膜上皮干细胞定位特征的免疫组织化学研究

利用单克隆抗体ＡＥ５与分化型角膜上皮细胞中角蛋白Ｋ３特异性结合,研究缺乏分化标志特征的角膜上皮干细胞定位特点,应用免疫组织化学方法显示Ｋ３阳性表达的区域分布于除角膜缘上皮基底部以外的所有角膜上皮细胞中,角膜上皮干细胞存在于角膜缘基底部ＡＥ５抗体反应阴性细胞中,即角膜干细胞位于角膜缘上皮层基底部。     

Regional heterogeneity in human corneal and limbal epithelia: an immunohistochemical evaluation.

The authors studied the distribution of specific keratins within the superior, inferior, medial, and lateral regions of human limbus and cornea to determine whether the limbal epithelium exhibits regional heterogeneity in its microstructure. A corneal epithelial basic keratin (K3), recognized by monoclonal antibody AE5, was immunohistochemically undetectable in the basal layers of the limbus in these four regions, but was seen in all layers in the central cornea. The pattern of immunostaining with another monoclonal antibody, AE1, which recognizes several acidic keratins, was complementary to AE5 staining in that AE1 recognized a similar heterogeneity in the limbal epithelial cells. AE1 immunoreacted with the basal cells of the limbus, but not those of the central corneal epithelium. Limbal characteristics, as defined by AE1-positive and AE5-negative staining, extended deeply into peripheral cornea in the superior and inferior regions, but to a lesser extent in the lateral and medial regions. The broader regions of epithelium with limbal characteristics in the superior and inferior regions raises the possibility that these regions play an important role in corneal epithelial maintenance and wound healing.     

Differentiation in cultured limbal epithelium as defined by keratin expression.

The authors investigated differentiation of cultured corneal and limbal epithelial cells by immunochemically evaluating the changes in the profiles of keratins recognized by two monoclonal antibodies: AE5, which recognizes K3, and AE1, which recognizes a group of acidic keratins including K16, which is present in the hyperproliferative cells. After 1 and 2 weeks in culture, the human epithelial cells did not react with AE5 but did react strongly with AE1. At 3 weeks, only suprabasal cells exhibited a moderate reactivity with AE5, whereas AE1 binding was seen in all of the cells. After 5 to 6 weeks in culture, all of the cells reacted moderately with AE5 and AE1. Treatment of 2-week-old limbally derived cultures with mitomycin C (mitosis inhibitor) did not inhibit subsequent K3 expression. Thus, K3 expression was associated with maturation or a later stage of differentiation that did not require an additional cell division. Unlike human epithelial cells, rabbit suprabasal epithelial cells expressed K3 (reactivity to AE5) after only ten days in culture. The epithelium derived from central human cornea lost K3 by 1 week in tissue culture but expressed keratin(s) recognized by AE1. Even after 4-6 weeks, cells derived from the central cornea did not become confluent and did not react with AE5. Thus, limbally derived human and rabbit epithelial cells undergo chronological changes in K3 expression similar to that seen in rabbit epithelial cells derived from central cornea. However, cultured human limbal epithelial cells take a significantly longer time to express K3 (a phenotypic characteristic of differentiated corneal epithelium) than do rabbit epithelial cells.     

角膜上皮干细胞定位特征的免疫组织化学研究

利用单克隆抗体ＡＥ５与分化型角膜上皮细胞中角蛋白Ｋ３特异性结合,研究缺乏分化标志特征的角膜上皮干细胞定位特点,应用免疫组织化学方法显示Ｋ３阳性表达的区域分布于除角膜缘上皮基底部以外所有角膜上皮细胞中,角膜上皮干细胞存在于角膜缘基底部ＡＥ５抗体反应阴性细胞中,即角膜干细胞位于角膜缘上皮层基底部。     

Ex vivo preservation and expansion of human limbal epithelial stem cells on amniotic membrane cultures

BACKGROUND/AIM: Amniotic membrane (AM) transplantation effectively expands the remaining limbal epithelial stem cells in patients with partial limbal stem cell deficiency. The authors investigated whether this action could be produced ex vivo. METHODS: The outgrowth rate on AM was compared among explants derived from human limbus, peripheral cornea, and central cornea. For outgrowth of human limbal epithelial cells (HLEC), cell cycle kinetics were measured by BrdU labelling for 1 or 7 days, of which the latter was also chased in primary cultures, secondary 3T3 fibroblast cultures, and in athymic Balb/c mice following a brief treatment with a phorbol ester. Epithelial morphology was studied by histology and transmission electron microscopy, and phenotype was defined by immunostaining with monoclonal antibodies to keratins and mucins. RESULTS: Outgrowth rate was 0/22 (0%) and 2/24 (8.3%) for central and peripheral corneal explants, respectively, but was 77/80 (96.2%) for limbal explants (p <0.0001). 24 hour BrdU labelling showed a uniformly low (that is, less than 5%) labelling index in 65% of the limbal explants, but a mixed pattern with areas showing a high (that is, more than 40%) labelling index in 35% of limbal explants, and in all (100%) peripheral corneal explants. Continuous BrdU labelling for 7 days detected a high labelling index in 61.5% of the limbal explants with the remainder still retaining a low labelling index. A number of label retaining cells were noted after 7 day labelling followed by 14 days of chase in primary culture or by 21 days of chase after transplantation to 3T3 fibroblast feeder layers. After exposure to phorbol 12-myristate 13-acetate for 24 hours and 7 day labelling, HLEC transplanted in athymic mice still showed a number of label retaining basal cells after 9 days of chase. HLEC cultured on AM were strongly positive for K14 keratin and MUC4 and slightly positive in suprabasal cells for K3 keratin but negative for K12 keratin, AMEM2, and MUC5AC. After subcutaneous implantation in athymic mice, the resultant epithelium was markedly stratified and the basal epithelial cells were strongly positive for K14 keratin, while the suprabasal epithelial cells were strongly positive for K3 keratin and MUC4, and the entire epithelium was negative for K12 keratin and MUC5A/C. CONCLUSIONS: These data support the notion that AM cultures preferentially preserve and expand limbal epithelial stem cells that retain their in vivo properties of slow cycling, label retaining, and undifferentiation. This finding supports the feasibility of ex vivo expansion of limbal epithelial stem cells for treating patients with total limbal stem cell deficiency using a small amount of donor limbal tissue.     

Cytokeratin patterns in corneal, limbal, and conjunctival epithelium. An immunofluorescence study with PKK-1, 8.12, 8.60, and 4.62 anticytokeratin antibodies.

The authors examined immunofluorescently the specific cytokeratin staining patterns of corneal, limbal, and conjunctival epithelium with PKK-1, 8.12, 4.62, and 8.60 monoclonal anticytokeratin antibodies. Observations were made on unfixed frozen postmortem human tissue. The PKK-1 antibody stain was observed in all layers of corneal epithelium but only in suprabasal layers of limbal and conjunctival epithelium. By contrast, the 8.12 antibody stain was observed only in the superficial layer of corneal epithelium but through all layers of limbal and conjunctival epithelium. The 4.60 antibody stain was seen in focal areas of corneal and limbal epithelium and through all layers of conjunctival epithelium. The 8.60 antibody stain was not present in the three epithelia. These immunofluorescence studies showed unique cytokeratin patterns among layers in corneal, limbal, and conjunctival epithelium.     

Expression of K12 keratin in alkali-burned rabbit corneas.

The healing of alkali-injured corneas is characterized by the persistence of polymorphonuclear leukocytes (PMN) in tissues and recurrent corneal epithelial defects. It has been suggested that the proteolytic enzymes secreted by PMN may account in part for the recurrent epithelial defects in the alkali-burned corneas. Cytoplasmic keratins, which form intracellular intermediate filaments, participate in the formation of hemidesmosomes and play a key role in the focal adhesion of epithelial cells to the basement membranes. The K3/K12 keratin pair is a major constituent of differentiated and stratified corneal epithelium. We have recently cloned the cDNA encoding the rabbit K12 keratin. In the present study we examined the expression of K12 keratin during the healing of alkali-burned rabbit corneas by slot-blot and in situ hybridization. Our results indicate that in normal cornea K12 keratin is equally expressed in all cell layers of stratified corneal epithelium and suprabasal layers of limbal epithelium, but not in bulbar conjunctival and other epithelia, i.e., lens, iris, and retinal pigment epithelium. The basal cells of the detached regenerating epithelium of the injured cornea express a very low level of K12 keratin. These observations are consistent with the notion that defective expression of K3/K12 keratins may play a role in the abnormal attachment of the regenerating epithelium to the basement membrane.     

Corneal epithelial stem cells at the limbus: looking at some old problems from a new angle

Corneal epithelium is traditionally thought to be a self-sufficient, self-renewing tissue implying that its stem cells are located in its basal cell layer. Recent studies indicate however that corneal epithelial stem cells reside in the basal layer of peripheral cornea in the limbal zone, and that corneal and conjunctival epithelia represent distinct cell lineages. These ideas are supported by the unique limbal/corneal expression pattern of the K3 keratin marker for corneal-type differentiation; the restriction of the slow-cycling (label-retaining) cells in the limbus; the distinct keratin expression patterns of corneal and conjunctival epithelial cells even when they are provided with identical in vivo and in vitro growth environments; and the limbal cells' superior ability as compared with central corneal epithelial cells in undergoing in vitro proliferation and in reconstituting in vivo an intact corneal epithelium. The realization that corneal epithelial stem cells reside in the limbal zone provides explanations for several paradoxical properties of corneal epithelium including its 'mature-looking' basal cells, the preponderance of tumor formation in the limbal zone, and the centripetal cellular migration. The limbal stem cell concept has led to a better understanding of the strategies of corneal epithelial repair, to a new classification of various anterior surface epithelial diseases, to the use of limbal stem cells for the reconstruction of corneal epithelium damaged or lost as a consequence of trauma or disease ('limbal stem cell transplantation'), and to the rejection of the traditional notion of 'conjunctival transdifferentiation'. The fact that corneal epithelial stem cells reside outside of the cornea proper suggests that studying corneal epithelium per se without taking into account its limbal zone will yield partial pictures. Future studies need to address the signals that constitute the limbal stem cell niche, the mechanism by which amniotic membrane facilitates limbal stem cell transplantation and ex vivo expansion, and the lineage flexibility of limbal stem cells.     

Relative proliferative rates of limbal and corneal epithelia. Implications of corneal epithelial migration, circadian rhythm, and suprabasally located DNA-synthesizing keratinocytes

An important element of the recently proposed limbal stem cell model is that corneal epithelial cells migrate centripetally. The driving force for this migration is unknown, although it has been suggested that limbal epithelium, proliferates at a higher rate than central corneal epithelium, thus creating a population pressure toward the central cornea. This hypothesis was tested by measuring the relative proliferative rates of limbal and central corneal epithelia using 3H-thymidine autoradiographic techniques. The results indicate that, in both the New Zealand white rabbit and SENCAR mouse, the labeling index (LI) of limbal epithelium is actually lower than that of central corneal epithelium. This difference in LI persists throughout the circadian rhythm cycle. These results suggest that population pressure per se cannot be responsible for the centripetal migration of corneal epithelium and raise the possibility that preferential desquamation of central corneal epithelium may "draw" peripheral cells toward the central cornea. In both epithelia, the LI peak precedes the mitotic index (MI) peak during circadian cycle by 4-6 hr. These data therefore are in close agreement with earlier results on several nonocular stratified epithelia but contradict an earlier suggestion that the LI and MI peaks of corneal epithelium coincide. Finally, although most of the 3H-thymidine incorporating cells in central cornea may appear to be suprabasally located, they are only partially displaced into the suprabasal compartment. In most cases, such cells are still connected with the basement membrane through a thin stalk of cytoplasm. Since corneal epithelium rests on an exceptionally flat and rigid substratum, an increase in cellular volume in DNA-synthesizing cells may not be tolerated well in an already crowded basal layer. This may explain why an unusually large proportion of DNA-synthesizing cells are expelled preferentially into either a "second tier basal layer" or into the suprabasal compartment.     

The suprabasal layer of corneal epithelial cells represents the major barrier site to the passive movement of small molecules and trafficking leukocytes

AIMS: To investigate the site of barrier function to the passive diffusion of a small molecule (phalloidin) in the corneal epithelium in the mouse. METHODS: Penetration of phalloidin (molecular weight 1115 daltons) into the cornea was evaluated by studying fluorescent binding of phalloidin to actin in tissue sections, in whole mount preparations, and in the fixed intact globe by confocal microscopy. In addition, the location of tight junction proteins in the individual layers of the corneal epithelium was determined by immunohistochemistry. RESULTS: Phalloidin staining of corneal sections was positive in all corneal layers in tissue sections and in all layers of the corneal epithelium except the suprabasal layer in excised fixed whole mounts of the cornea. However, when phalloidin staining was attempted in intact fixed globes, before excision of the cornea for whole mount preparation, only the most superficial layer of cells was stained indicating that phalloidin could not penetrate the tissue beyond the suprabasal epithelial layer. Detergent (Triton X-100) treatment of the excised cornea and the intact fixed globe, allowed penetration of phalloidin into the suprabasal epithelial layer. Tight junction proteins occludin, ZO-1 and claudin were present in most layers of the cornea but while ZO-1 and occludin were distributed in a typical pericellular pattern, claudin seemed to be particularly prominent in the suprabasal layer and appeared only as a discontinuous punctate pericellular pattern in the superficial layer. Intraepithelial leukocytes were detected in the superficial epithelium and the basal epithelium but not in the suprabasal epithelium. CONCLUSION: The suprabasal epithelium cell layer appears to represent the main barrier site to the passage of small molecules and cells in the mouse cornea and this property may be attributable to prominent claudin expression in this layer.     

Autoimmunity against corneal antigens. I. Isolation of a soluble 54 Kd corneal epithelium antigen

Corneal epithelium antibodies were detected in patients with corneal melting disease and uveitis using an immunofluorescence technique with cryostat sections of corneas obtained from various species (man, guinea pig, rabbit, mouse, rat, cow, pig). No differences in results were found using these various substrates, indicating that the autoimmune response is directed against common non-species specific corneal epithelium antigens. The serum of a patient with corneal melting disease, containing a high antibody titer against corneal epithelium was used to identify and isolate one of the bovine corneal antigens. A 54,000 dalton protein was isolated, which was shown to be the major protein present in the corneal epithelium. Absorption studies with other tissues taken from human eyes showed that cornea epithelium, cornea devoid of epithelium, ciliary body and retina contained material which cross-reacted with the isolated bovine corneal epithelium antigen, whereas iris and sclera showed no detectable cross-reaction. The incidence of autoantibodies directed against this antigen was investigated in patients with corneal melting disease, corneal transplantion and in uveitis patients using an ELISA and comparing the results with those obtained with the immunofluorescence assay on rabbit cornea sections. A positive ELISA was always associated with a positive immunofluorescence test. The presence of antibodies against the 54 Kd antigen as detected by the ELISA could be confirmed by immunoblotting in 7 out of 9 positive sera tested. A large number of sera showed a positive immunofluorescence test but a negative ELISA against the 54 Kd corneal epithelium antigen.(ABSTRACT TRUNCATED AT 250 WORDS)     

Phenotypic study of a case receiving a keratolimbal allograft and amniotic membrane for total limbal stem cell deficiency

PURPOSE: To report the expression pattern of key molecules by the reconstructed corneal epithelium after a keratolimbal allograft (KLAL) and amniotic membrane transplantation (AMT) for total limbal stem cell deficiency. DESIGN: Interventional case report. METHOD: A 50-year-old woman with severe chemical burns in both eyes received an AMT as a temporary patch at the acute stage, and a KLAL with AMT as a graft at the chronic stage for total limbal stem cell deficiency. The corneal button removed during subsequent corneal transplantation was submitted for immunofluorescence staining with monoclonal antibodies against keratin K3, MUC5AC, connexin 43, integrins alpha3beta1 and alpha6beta4, and laminin 5 for comparison with a normal cornea. RESULTS: Histologically, a normal stratified corneal epithelium has five to six cell layers that lay on the thick amniotic membrane basement membrane. The phenotype was of a corneal origin, based on expression of positive keratin K3, negative MUC5AC, and positive connexin 43. Furthermore, intact basement membrane complexes were present, evidenced by positive staining to integrins alpha3beta1 and alpha6beta4 and to laminin 5. CONCLUSIONS: A normal corneal epithelial phenotype with normal basement membrane complexes was restored after a KLAL and AMT in a case with total limbal stem cell deficiency.     

Adaptation of homeostatic ocular surface epithelium to chronic treatment with the opioid antagonist naltrexone

PURPOSE: To determine how ocular surface epithelium adjusts to an increase in cell replication after treatment with the opioid antagonist naltrexone (NTX). METHODS: Adult male rats were given twice daily injections of 30 mg/kg NTX or vehicle for 7 days. Outcomes of NTX administration included DNA synthesis (monitored with BrdU), mitosis (assayed using colchicine), number of cell layers and cell diameter, apoptosis and necrosis, and packing density for the peripheral corneal epithelium, limbus, and conjunctiva. Also, transit time from basal to surface epithelial layers in the peripheral cornea was assessed with [H]thymidine as a marker. RESULTS: DNA synthesis and mitosis in the basal layer of the peripheral corneal epithelium of NTX-treated rats were increased 69% and 85%, respectively, from control levels; no changes in either parameter were recorded in the limbal or conjunctival epithelium (stem cell region). Epithelial thicknesses in the NTX group were increased by 8% to 38% from control subjects, without more cell layers. Packing density in NTX-treated rats was increased from control values by 26% in the basal layer of the limbus and by 12% to 28% in the suprabasal layers of the corneal epithelium, limbus, and conjunctiva. Epithelial cell diameters from corneas of NTX-exposed rats were subnormal in the basal and suprabasal cells of the limbus and conjunctiva. Apoptosis and necrosis were negligible in the epithelium of NTX-treated and control rats. Transit times of peripheral corneal epithelial cells of animals in the NTX group were shortened by 63% from control levels. CONCLUSIONS: These data show that a 1-week treatment with NTX does not induce proliferative pathology or toxicity in ocular surface epithelium, has a minimal effect on stem cell proliferation, and accelerates normal homeostatic processes. Topical application of NTX for stimulation of corneal epithelial wound healing results in no adverse sequelae, thereby supporting the therapeutic role for this drug in the treatment of ocular surface abnormalities.     

Particle-mediated gene transfer of opioid growth factor receptor cDNA regulates cell proliferation of the corneal epithelium

PURPOSE: This study was designed to determine at the molecular level whether interactions between the opioid growth factor (OGF) and OGF receptor (OGFr) play a role in regulating DNA synthesis in the homeostasis of the corneal epithelium. METHODS: The plasmid pcDNA3.1+OGFr-HA, carrying the rat OGFr cDNA epitope-tagged with a C-terminal hemagglutinin (HA), or the empty-vector (pcDNA3.1+), was delivered twice by the Helios Gene Gun System at 300 psi to the cornea of anesthetized rats. The contralateral (untreated) cornea served as the naive specimen. BrdU was used to determine whether the recombinant OGFr was effective in regulating DNA synthesis in the rat peripheral corneal epithelium. RESULTS: Within 18 hours of transfection, positive HA staining was apparent in both the basal and suprabasal layers (efficiency > 90% of the cells) throughout the central and peripheral cornea. Quantitative immunohistochemistry with rhodamine-conjugated anti-OGFr antibodies revealed twofold more OGFr expression in the central and peripheral epithelium of transfected corneas relative to naive corneas. The number of BrdU-positive basal cells in the peripheral epithelium of the transfected cornea was one-third of that in the naive cornea. CONCLUSIONS: These data demonstrate the direct role of the OGF-OGFr system in determining cellular renewal in the mammalian corneal epithelium. Moreover, the successful establishment of a novel delivery system of cDNAs to the ocular surface suggests a therapeutic role for gene therapy in the eye.     

Comparison of limbal and peripheral human corneal epithelium in tissue culture

Peripheral human corneal epithelium grows better in tissue culture than central epithelium, but it is not known whether ocular limbal epithelium grows even better than does the peripheral corneal epithelium. In this work we compared the growth kinetics of limbal and peripheral human corneal epithelial cells in tissue culture. Four 1-2 mm2 explants, removed from the limbus or from peripheral cornea (1-2 mm inside the limbus) of eye bank eyes, were grown to confluence in primary culture. Cells were then passaged at 2 X 10(5) cells per dish. At intervals thereafter, the cells were counted in a hemocytometer to determine plating efficiency and growth curves. Mitotic activity was determined 4 days after passaging by labeling cultures with 3H-thymidine and counting aliquots using the hemocytometer and scintillation counter. In the primary cultures, limbal epithelium grew as small, uniformly polygonal cells. Peripheral corneal cells grew to a variety sizes. The 24 hr plating efficiency and doubling time of limbal epithelial cells were 47 +/- 8% and 80 +/- 14 hr, respectively, while those of peripheral corneal cells were 41 +/- 10% (P less than 0.1) and 131 +/- 25 hr (P less than 0.001). The mitotic activity of limbal cells was significantly higher than that of peripheral (2.9 +/- 1.2 vs. 0.8 +/- 0.6) (P less than 0.01). These results indicate that human ocular limbal epithelium grows better in culture than does peripheral human corneal epithelium.     

Change in epithelial keratin expression during healing of rabbit corneal wounds

Corneal epithelial wound healing following full-thickness trephination and transcorneal freeze injury was studied by electron microscopy and immunofluorescent microscopy using monoclonal antibodies AE1, AE2, and AE3 to human epithelial keratin. Wounds were evaluated at various time intervals between 4 hr and 2 mo after injury. By scanning and transmission electron microscopy, epithelial migration was evident 4 hr after injury and was characterized by thinning of the epithelium and extension of filopodial processes. AE1 monoclonal antibody, which stains specifically the superficial cells of normal corneal epithelium, reacted to cells at the leading edge of the migrating epithelium. By 24 hr, all cells migrating over the wound displayed positive fluorescence with AE1 while the epithelium over the undamaged cornea exhibited normal fluorescence limited to the superficial epithelial cells. In full-thickness corneal wounds, reepithelialization was complete by 1-2 wk; however, all epithelial cells covering the wound remained positive for the AE1 antikeratin antibody. By 2 mo, the AE1 fluorescence returned to normal. In transcorneal freeze injuries, reepithelialization was complete by 4 to 7 days after injury, with all cells overlying the wound reacting with the AE1 antibody. By 2 wk after freeze injury, all epithelial cells appeared to express a normal AE1 staining pattern. No change was noted in the fluorescent distribution of either AE2 antibody, which did not react with the corneal epithelium, or AE3, which reacts with all corneal epithelial cells. These results suggest that healing of corneal epithelial wounds involves changes in keratin expression of the corneal epithelium.     

Phenotypic analyses of limbal epithelial cell cultures derived from donor corneoscleral rims

Grafted cultures of limbal epithelial cells aid repair of the corneal epithelium, but their phenotype is unclear. In this study, the phenotype of cultures that were similar in age to those used clinically were analysed. Limbal epithelial cells were isolated from donor corneoscleral rims and grown in various media, including those designed for keratinocytes. Successful cultures in each medium developed predominantly small (10 μm) tightly packed cells. Immunocytochemistry and western blotting revealed expression of keratins 3, 14 and 19. Expression of these keratins in situ was confirmed by immunohistochemistry. Basal limbal epithelial cells were positive for keratins 14 and 19, and suprabasal cells were positive for keratin 3. However, intense staining for keratin 14 was also observed at the inner cut edge of corneoscleral rims. These findings demonstrate the potential importance of keratins 14 and 19 as markers of epithelial cell differentiation in the human cornea.     

Localization of keratin in the cells of the cornea in aphakia and normal mouse embryos

Abnormal lens morphogenesis in the aphakia mutant in the mouse often results in a club-shaped elongated 'lens' that remains attached to the surface epithelium by a persistent connecting stalk, which is partially solid and partially cystic. Usually, the cells are continuous with the surface epithelium of the cornea and also with the cuboidal cells lining the corneal inner surface. Immunofluorescence with keratin antiserum not only gave positive reactions with the corneal epithelial cells, but also with many cells of the lens stalk, including its cysts, and with islands of cells on the inside of the cornea. These keratin-containing 'endothelial' cells may be the product of metaplasia of the endothelium into epithelium-like cells. Alternatively, they may also be the result of abnormal migration of epithelial cells into the eye or of abnormal differentiation of neural crest cells.     

Existence of small slow-cycling Langerhans cells in the limbal basal epithelium that express ABCG2

Despite the obvious importance of limbal stem cells in corneal homeostasis and tumorigenesis, little is known about their specific biological characteristics. The purpose of this study was to characterize limbal slow-cycling cells based on the expression of ABCG2 and major histocompatibility complex (MHC) class II and the cell size. Wistar rats were daily injected with 5-bromo-2-deoxyuridine (BrdU) at a dose of 5 mg/100 g for 2 weeks. After 4-week BrdU-free period, corneal tissues were excised, and immunofluorescence staining for ABCG2, BrdU, and MHC class II was performed by confocal microscopy. In another series, corneal tissues of normal rat were double immunostained for ABCG2, keratin 14, keratin 3, CD11c, and MHC class II. In addition, limbal, peripheral and central corneal epithelial sheets were isolated by Dispase II digestion and dissociated into single cell by trypsin digestion and cytospin preparations were double immunostained for ABCG2 and MHC class II. The cell size and nucleus-to-cytoplasm (N/C) ratio of limbal ABCG2+ cells were analyzed and compared with those of cells from other zones. BrdU label-retaining cells (LRCs) with expression of ABCG2 were found in the limbal epithelial basal layer, but not in other parts of the cornea. Approximately 20% of these cells were MHC class II positive. All MHC class II+ cells in the corneal epithelium were positive for CD11c, a marker for dendritic cells (DCs). Double labeling with ABCG2 and keratin 14 showed that nearly four-fifth of limbal ABCG2+ cells were positive for keratin 14 but negative for keratin 3, exhibiting an undifferentiated epithelial cell lineage. Cytospin sample analysis revealed the presence of a distinct population of smaller ABCG2+ cells with expression of MHC class II with a larger N/C ratio in the limbal epithelium. A new population of small slow-cycling cells with large N/C ratio has been found to express ABCG2 in the limbal epithelial basal layer. Some of these cells normally express MHC class II antigen. These findings may have important implications for our understanding of the characteristics of limbal slow-cycling cells.