If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Epithelial Biology Unit, Departments of Dermatology, Pharmacology and Urology, NYU Cancer Institute, New York University School of Medicine, New York, New York, USA
Corneal epithelium is a self-renewing tissue. Recent studies indicate that corneal epithelial stem cells reside preferentially in the basal layer of peripheral cornea in the limbal zone, rather than uniformly in the entire corneal epithelium. This idea is supported by a unique limbal/corneal expression pattern of the K3 keratin marker for corneal-type differentiation; the preferential distribution of the slow-cycling (label-retaining) cells in the limbus; the superior proliferative capacity of limbal cells as compared with central corneal epithelial cells in vitro and in vivo; and the ability of limbal basal cells to rescue/reconstitute severely damaged or completely depleted corneal epithelium upon transplantation. The limbal/stem cell concept provides explanations for several paradoxical properties of corneal epithelium including the predominance of tumor formation in the limbal zone, the centripetal migration of peripheral corneal cells toward the central cornea, and the “mature-looking” phenotype of the corneal basal cells. The limbal stem cell concept has led to a better understanding of the strategies that a stratified squamous epithelium uses in repair, to a new classification of various anterior surface epithelial diseases, to a repudiation of the classical idea of “conjunctival transdifferentiation”, and to a new surgical procedure called limbal stem cell transplantation.
Stem cells are a subpopulation of cells capable of extensive self-renewal that upon division gives rise to progeny (transit amplifying or TA cells) that have limited renewal capability (
). Additionally, stem cells divide relatively infrequently in mature tissues and are structurally and biochemically primitive. In cases of tissue injury, stem cells can proliferate to repopulate the tissue. The TA cell divides more frequently than the stem cell and ultimately all of the TA cells differentiate in the scheme of “stem cell→TA cell→terminally differentiated cell” (
). In many cases the identification of stem cells provide new insights into the growth and differentiation properties of the tissue in question. In the case of corneal epithelium, this tissue has long been known to have several unusual and puzzling features. For example, almost all corneal epithelial neoplasias are associated with the peripheral cornea in a rim called the limbus, which represents the transitional zone between the transparent cornea and the white conjunctiva (
). Another well known and peculiar feature of corneal epithelium is that the peripheral corneal epithelial cells seem to be able to migrate centripetally toward the center of the cornea (
While studying the growth and differentiation of rabbit corneal epithelial cells in vivo and in cell culture, Sun and coworkers discovered in the early to mid-1980s that corneal epithelial cells synthesized two major tissue-restricted keratins called K3 and K12 (
noted that K3 was associated with the upper, more differentiated, cell layers, indicating that K3 was a marker for an advanced stage of corneal epithelial differentiation. When the expression of the K3 keratin was examined in vivo, it was observed that this keratin was also expressed in the upper cell layers in corneal epithelium in the limbal zone; this was consistent with the concept that K3 was a marker for an advanced stage of differentiation. Unexpectedly, however, K3 was found to express uniformly in central rabbit corneal epithelium (i.e., even the supposedly undifferentiated basal cells of central corneal epithelium express the K3 differentiation marker). This uniform expression suggests that, although the basal cells in the limbal zone were undifferentiated, those of the central corneal epithelium are more differentiated as far as the expression of the K3 marker is concerned. This finding, coupled with several other biological considerations (see below), led
to propose that corneal epithelial stem cells were not uniformly dispersed across the entire corneal epithelial basal layer, as had been thought; instead, these stem cells were concentrated in the peripheral limbal zone (Figure 2).
Figure 1Cultured rabbit corneal epithelial cells. Appearance of rabbit corneal epithelial cells cultured in the presence of 3T3 feeder cells, double-stained with antibodies to keratin (red) and bromodeoxyuridine (BrdU) (green) demonstrating the epithelial nature and proliferative activities of these cells.
Figure 2The limbal stem cell concept. Panel (a) depicts the expression of the K3 keratin marker for an advanced stage of corneal-type differentiation in the limbal and corneal epithelium. In Panel (b), corneal epithelial stem cells are proposed to be situated in the basal layer of the limbal epithelium. The TA cells (stem cell progeny) migrate centripetally towards the central cornea
Strong support for the limbal stem cell concept has come from several approaches. The observation that slow-cycling cells were restricted to the limbal basal layer provided compelling evidence in support of the limbal/corneal stem cell hypothesis (Figure 3;
). One of the most reliable ways to identify epithelial stem cells in vivo takes advantage of the fact that these cells are relatively slow cycling, and thus can be identified experimentally as “label-retaining cells” (LRC) (
). To detect the slow-cycling cells, one can perfuse a tissue continuously with tritiated thymidine (3H-T) or bromodeoxyuridine (BrdU) to label as many dividing cells as possible, including some of the occasionally dividing stem cells. During a chase period, which is typically 4–8 wk, the rapidly dividing TA cells lose most of their labels due to dilution whereas the slow-cycling stem cells still retain their label; this way some of the stem cells can be detected experimentally as the LRC. Application of this labeling technique to mouse corneal epithelium revealed that central corneal epithelium contained no LRC; such cells were found exclusively in the basal layer of peripheral corneal epithelium in the limbal area (
Figure 3Limbal location of label-retaining cells. Label-retaining cells (with red-stained nuclei) are preferentially located in the basal layer of the limbal (L) epithelium. The TA cells (silver grains over nuclei; arrows) are primarily located in the corneal (C) epithelium.
Further support of the limbal stem cell concept has come from cell and explant culture studies showing that limbal cells have a higher in vitro proliferative potential than central corneal epithelial cells (
In vitro growth and differentiation of rabbit bulbar, fornix, and palpebral conjunctival epithelia. Implications on conjunctival epithelial transdifferentiation and stem cells.
). In vivo experiments have demonstrated that when limbal and corneal epithelia were continuously stimulated with phorbol myristate, limbal epithelium maintained a significantly greater proliferative response than the corneal epithelium (
), thus providing additional support to the idea that limbal cells have a greater proliferative capacity than corneal epithelial cells.
An interesting and previously not well-understood phenomenon about corneal epithelium is that its squamous cell carcinomas, which are particularly abundant in cattle and are known as “cancer eye” (
) and since the limbal epithelium is enriched in stem cells, it makes sense that tumors originate from this region.
Perhaps some of the most striking biological data in support of the limbal stem cell concept are the transplantation studies pioneered by Tseng and colleagues, who demonstrated that limbal stem cells can be used to reconstitute the entire corneal epithelium (Figure 4;
). This procedure, known as limbal stem cell transplantation, has restored the eyesight of many patients and is being practiced by ophthalmologists all over the world (
noted that limbal transplantation, in cases of severe ocular surface injury caused by chemical and thermal burns, resulted in rapid corneal re-epithelialization with an optically smooth, stable surface that did not subsequently erode or persistently breakdown. These findings clearly demonstrated that the limbal epithelium can be used to restore the lost stem cell population. The next advance in limbal transplantation was the successful use of limbal allografts, in conjunction with immunosuppression, to restore eyesight in patients with severe corneal epithelial damage (
). This technique reduces the risk of causing limbal cell deficiency in the healthy donor eye after the removal of a relatively large limbal autograft (
). Another way to minimize the damage of the donor eye is to expand, under in vitro cell culture conditions, human limbal epithelial cells for the purpose of transplantation (
were the first to demonstrate that such in vitro expanded limbal cells can be successfully transplanted to the severely damaged eye with subsequent restoration of the corneal surface and vision. More recently, it has been reported that human amniotic membrane provides a substrate that not only supports the in vitro propagation of limbal stem cells, but also has a striking anti-inflammatory effect on the recipient site. Limbal stem cells propagated this way have been used successfully in patients with a variety of ocular surface disorders (
Figure 4Restoration of eyesight after limbal stem cell transplantation. Clinical pictures of a chemical burnt eye before (left) and after (right) the reconstitution of corneal epithelium by limbal transplantation, followed by corneal transplantation
(courtesy of Dr Scheffer Tseng, Ocular Surface Foundation, Miami, Florida).
Since the term corneal epithelial stem cell was first used in 1986, the concept of corneal epithelial stem cells residing in the limbus has spawned a fast-growing field of research. Unlike other epithelial stem cells that are physically adjacent to their progeny thus complicating their analysis (see papers on epidermal and hair follicular stem cells in this volume), limbal stem cells are well separated from their progeny cells. Therefore, the corneal/limbal epithelium, as a model system, offers unique advantages for studying the properties of stem cells versus their progeny TA cells and terminally differentiated cells (
). For example, the manner in which stem cell and TA cell proliferation is modulated during corneal epithelial repair has provided information on the strategies that a stratified squamous epithelium adopts to expand during wound healing. Using a double-labeling technique that permits the detection of two or more rounds of DNA synthesis in a given cell, we demonstrated that a large number of normally slow-cycling limbal epithelial stem cells could be induced to replicate in response to a single physical or chemical perturbation of the central corneal epithelium (
). In addition, we showed that corneal epithelial TA cells, located in unperturbed peripheral cornea, replicate at least twice and have a relatively long cell cycle time of about 72 h. When induced to proliferate, however, these TA cells can reduce their cell cycle time and undergo additional cell divisions. In contrast, corneal epithelial TA cells can usually divide only once prior to becoming post-mitotic even after TPA stimulation, suggesting a reduced proliferative capacity. These results indicate that corneal epithelium uses three strategies to expand its cell population during wound healing: (i) recruitment of stem cells to produce more TA cells; (ii) increasing the number of times a TA cell can replicate; and (iii) increasing the efficiency of TA cell replication by shortening the cell cycle time (Figure 5;
). Similar repair strategies may be used by epidermis and other stratified squamous epithelia.
Figure 5Three strategies of epithelial proliferation. In a resting or “normal” situation, stem cells (S) located in the limbus rarely cycle (large curved arrow). Upon division, stem cells produce regularly cycling TA cells (vertical arrows) located in the peripheral (pp) and central (cc) corneal epithelium. Young TA cells (TA1, 2, 3) are preferentially located in peripheral cornea, whereas more mature TA cells (TA4) reside in central cornea. Under these conditions, not every TA cell will utilize its full potential to divide, illustrated by those TA cells that give rise to terminally differentiated cells (TD; squares), cells 5–8. Upon wounding, corneal epithelium adopts three means to expand its cell population. It can recruit more stem cells to divide with a more rapid cell cycle time (small curved arrows) thus producing more TA cells. It can induce the young TA cells to utilize more fully their replicative capacity thus producing more (mature) TA cells (TA4). Finally, by shortening the cell cycle time (short vertical arrows), it can increase the efficiency of TA cell replication.
Many other important basic questions about the properties of corneal epithelial stem cells have been studied intensively, including basement membrane heterogeneity (
). Impressive clinical advances have taken advantage of the limbal stem cell concept, including limbal stem cell transplantation and a new way of classifying the anterior ocular epithelial deficiencies and abnormalities (
). In addition, in vitro and in vivo studies have demonstrated that corneal/limbal epithelium and conjunctival epithelium belong to two distinct lineages, thus refuting the classical concept of conjunctival transdifferentiation (
Conjunctival epithelial cells can resurface denuded cornea, but do not transdifferentiate to express cornea-specific keratin 12 following removal of limbal epithelium in mouse.
Figure 6Molecular regulation of the K3 keratin gene. A 300-bp 5′-upstream sequence of rabbit keratin 3 gene was shown to be able to function as a keratinocyte-specific promoter in transient transfection assays (
Lineage-specific and differentiation-dependent expression of K12 keratin in rabbit corneal/limbal epithelial cells: cDNA cloning and northern blot analysis.
). Mutations of an “E-motif”, that contains overlapping Sp1 and AP-2 sites, reduce K3 gene promoter activity by 70%. We showed that Sp1 activates whereas AP-2 represses the K3 promoter. Although the undifferentiated corneal epithelial basal cells express equal amounts of Sp1- and AP-2-binding activities, the differentiated corneal epithelial cells downregulate drastically their AP-2 activity thus resulting in a 6–7-fold increase in the activator to repressor ratio. Such an increased activator/repressor ratio in differentiated corneal epithelial cells provides an explanation for the differentiation-dependent expression of the K3 keratin gene. The fact that basal cells contains a high concentration of polyamine, which preferentially inhibits the binding of the Sp-1 activator to its DNA motif, provides an additional mechanism for the selective expression of K3 gene in the suprabasal cellular compartment (
The corneal epithelial stem cell concept has enhanced our understanding of the biology, biochemistry, and diseases of anterior ocular epithelia. Many challenges still face epithelial stem cell biologists, however. For example, the generation of stem cell-specific surface markers will greatly facilitate the physical isolation and molecular characterization of stem cells. Some of the currently available markers for limbal stem cells, e.g., enolase (
) are expressed not only by limbal basal cells, but also by a majority of basal cells of various stratified squamous epithelia making them unlikely to be stem cell specific. Another important area is the characterization of the microenvironment that forms the stem cell niche. One of the first examples of biochemical heterogeneity between limbal and corneal epithelial basal cells was the K3 keratin data, which demonstrated the suprabasal expression of K3 keratin in the limbal zone, but uniform expression in corneal epithelium (
Identification of a cytosolic NADP+-dependent isocitrate dehydrogenase that is preferentially expressed in bovine corneal epithelium: A corneal epithelial crystallin.
) also showed a similar limbal versus corneal epithelial expression pattern. This differential expression of proteins in the limbal versus corneal basal cells may be in part due to basement membrane heterogeneity. Using a monoclonal antibody AE27,
demonstrated strong staining of the corneal epithelial basement membrane zone and heterogenous staining of the limbal basement membrane zone. Interestingly, limbal basal cells in contact with those areas of the basement membrane that were strongly AE27 positive, expressed K3, whereas those cells resting on basement membrane that was AE27 negative or weak did not express K3 (
). This result strongly suggests basement membrane composition can influence K3 expression. Additional evidence for basement membrane heterogeneity was provided by
who showed that laminin chains α-2 and β-2 were present in limbal basement membrane but not in central corneal basement membrane. More recently, tissue recombination studies have demonstrated that the K3-negative phenotype of the limbal basal cells is mediated through the limbal stroma/basement membrane (
). Together, these data suggest that the regulation of the expression of many corneal differentiation-dependent genes may be influenced by (horizontal) basement membrane heterogeneity. Although such basement membrane heterogeneity undoubtedly contributes to the limbal and corneal epithelial phenotypes, many other mesenchymal signaling molecules are likely to be involved in maintaining the “stemness” of limbal stem cells. Some recent data suggest that amniotic membrane can support the replication of limbal stem cells and therefore provides an experimental stem cell niche (
Adult corneal epithelium basal cells possess the capacity to activate epidermal, pilosebaceous and sweat gland genetic programs in response to embryonic dermal stimuli.
showed that adult corneal epithelium, when combined with embryonic skin dermis, can gave rise to hair follicles. Since it is well accepted that TA cells comprise the proliferative population of the central corneal epithelium, these findings suggest that given appropriate signal(s), even the TA cells have the flexibility of being converted to epidermis and its appendages. More studies are clearly needed to fully define the flexibility of the stem cells of corneal epithelium, epidermis, and other stratified squamous epithelia.
This work was supported by National Institutes of Health Grants EY06769, EY13711 (RML), and EY04722 (T-TS).
Adult corneal epithelium basal cells possess the capacity to activate epidermal, pilosebaceous and sweat gland genetic programs in response to embryonic dermal stimuli.
Conjunctival epithelial cells can resurface denuded cornea, but do not transdifferentiate to express cornea-specific keratin 12 following removal of limbal epithelium in mouse.
Identification of a cytosolic NADP+-dependent isocitrate dehydrogenase that is preferentially expressed in bovine corneal epithelium: A corneal epithelial crystallin.
Classification, expression, and possible mechanisms of evolution of mammalian epithelial keratins: A unifying model.
in: Levine A. Topp W. Woude G Vance Watson J.D. A Unifying Model. New York,
Cold Spring Harbor1984: 169-176 (Classification, Expression, and Possible Mechanisms of Evolution of Mammalian Epithelial Keratins)
In vitro growth and differentiation of rabbit bulbar, fornix, and palpebral conjunctival epithelia. Implications on conjunctival epithelial transdifferentiation and stem cells.
Lineage-specific and differentiation-dependent expression of K12 keratin in rabbit corneal/limbal epithelial cells: cDNA cloning and northern blot analysis.
53009-1&id=gr1.jpg){kind=link}
53009-1&id=gr2.jpg){kind=link}
53009-1&id=gr3.jpg){kind=link}
53009-1&id=gr4.jpg){kind=link}
53009-1&id=gr5.jpg){kind=link}
53009-1&id=gr6.jpg){kind=link}