| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TISSUE-SPECIFIC STEM CELLS |
aCells for Sight Transplantation and Research Programme,
bOcular Repair and Regeneration Biology Unit, UCL Institute of Ophthalmology, London, United Kingdom;
cMoorfields Eye Hospital NHS Foundation Trust, London, United Kingdom
Key Words. Stem cell • Cornea • Limbus cornea • Somatic stem cell biology • Human • Stem cell transplantation • Three-dimensional imaging
Correspondence: Alex J. Shortt, M.Sc., M.R.C.Ophth., MRC Clinical Research Fellow, Institute of Ophthalmology and Moorfields Eye Hospital, 11–43 Bath Street, London, EC1V 9EL, United Kingdom. Telephone: +442076086894; Fax: +442076086887; e-mail: a.shortt{at}ucl.ac.uk
Received September 14, 2006;
accepted for publication February 18, 2007.
First published online in STEM CELLS EXPRESS March 1, 2007.
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionDisclosure of Potential...AcknowledgmentsReferences |
|---|
and ABCG2; however, current methods cannot identify for certain which exact cells within this cell population are truly LESCs. These structures could be located and observed in vivo in normal human subjects, but not in patients with clinically diagnosed corneal LESC deficiency. The distribution of these structures around the corneal circumference is not uniform. Biopsies targeted to limbal regions rich in LESC niche structures yielded significantly higher numbers of LESCs in culture. Our findings demonstrate how adult stem cell niches can be identified and observed in vivo in humans and provide new biological insight into the importance of LESC niche structures in maintaining normal LESC function. Finally, the concept of targeted biopsy of adult SC niches improves stem cell yield and may prove to be essential for the successful development of novel adult stem cell therapies. Disclosure of potential conflicts of interest is found at the end of this article.
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDisclosure of Potential...AcknowledgmentsReferences |
|---|
One key area of research is the identification and investigation of adult stem cell niches. Adult stem cells exhibit intrinsic properties that influence their behavior, but they also depend on specialized environmental niches to maintain them in an undifferentiated state and regulate their functions. Understanding these stem cell-niche interactions is essential if the goal of developing new adult stem cell therapies is to be fully realized. Some characteristics of adult stem cell niches in the gut, skin, and bone marrow have already been identified (reviewed in [1–3]), but progress in this area is hampered by the fact that, to date, it has not been possible to identify and observe adult stem cells within their niches in vivo in humans. Indeed, most studies in this area employ animal models and/or histopathological analysis of tissue specimens.
In skin, the continuous desquamation of superficial cells and replenishment of basal cells is made possible by interfollicular epidermal keratinocyte stem cells [4]. In the cornea, an equivalent population of limbal epithelial stem cells (LESCs) generate and renew the corneal epithelium throughout life and are thus essential for maintaining corneal transparency and vision [5–7]. Destruction of the LESC population by disease or injury results in the absence of corneal epithelium. This manifests as chronic discomfort and visual impairment caused by the overgrowth of the cornea by functionally incompatible conjunctival epithelial cells and blood vessels [8–10]. LESC deficiency can be treated successfully by transplantation of ex vivo cultured LESCs, as first reported by Pellegrini et al. [11] in 1997 and subsequently by others [12–20]. This pioneering adult stem cell therapy has placed ophthalmology at the forefront of stem cell research and therapeutic delivery.
The cornea possesses two unique characteristics that make it ideally suited as a model system for studying adult stem cells in humans. Firstly, it is optically transparent, and, hence, noninvasive imaging of LESCs in humans is possible. Secondly, LESCs are found at the corneal limbus (the peripheral extent of the cornea) and are anatomically segregated from their transient amplifying cell progeny, which migrate centrally to cover the paracentral and central cornea [7]. It has been shown in histological tissue sections that LESCs are discretely located in the basal layer of the corneal limbal epithelium, at the junction between the transparent cornea and the opaque sclera [6, 21, 22]. However, the total number and distribution of LESCs within the niche is unknown. Furthermore the presence of a specialized adult stem cell niche within the human corneal limbus has not been conclusively demonstrated to date. The aims of this study were to use state-of-the-art imaging techniques to identify the structural characteristics and regional distribution of the LESC niche and to determine whether this niche can be located and observed in vivo in humans, thus allowing limbal biopsies to be targeted to potentially LESC-rich limbal regions.
|
METHODS |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDisclosure of Potential...AcknowledgmentsReferences |
|---|
In Vitro Confocal Microscopy and Three-Dimensional Reconstruction of Limbal Structure
Ten cadaveric human corneas were obtained from eight donors (age range, 18–77 years; mean, 56.9 years; median, 66 years). Corneas were oriented, dissected into quarters, fixed (within 72 hours of death), and stained with propidium iodide (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com) to label nuclei and fluorescein isothiocyanate (FITC)-phalloidin (Sigma-Aldrich) to label the actin cytoskeleton. Segments were then whole-mounted on glass slides and coverslipped. Z-stack image series were collected through the entire corneal limbal epithelium and superficial limbal stroma using a Zeiss LSM510 confocal microscope (Carl Zeiss, Jena, Germany, http://www.zeiss.com). Image series were converted into AVI movie format using Windows Movie Maker (Microsoft, Redmond, WA, http://www.microsoft.com) and also reconstructed in three-dimensional (3D) imaging using Volocity 3.0 software (Improvision, Coventry, U.K., http://www.inprovision.com) for evaluation. To examine regional differences in the structure of the limbus, a continuous array of overlapping images collected at the level of the basal epithelial layer were combined to create a 360° montage of the corneal limbus (Adobe Photoshop 7.0, Adobe, San Jose, CA, http://www.adobe.com).
Whole Mount Immunofluorescence
Two human corneas from different donors were fixed with 4% paraformaldehyde for 1 hour at room temperature then divided into quarters, washed, and blocked using 20% goat serum in phosphate-buffered saline (PBS) at room temperature for 2 hours. They were then incubated with primary antibodies to CD31/platelet endothelial cell adhesion molecule-1 (PECAM-1; Abcam, Cambridge, U.K., http://www.abcam.com) at 4°C overnight. After washing, specimens were incubated with tetramethyl rhodamine isothiocyanate (TRITC)-conjugated goat anti-mouse IgG 1:200 (Jackson Immunoresearch Laboratories, West Grove, PA, http://www.jacksonimmuno.com) and 50 µg/ml FITC-phalloidin (Sigma-Aldrich) for 2 hours at room temperature. The specimens were washed and mounted in Mowiol 4-88 (Sigma-Aldrich), coverslipped, and examined using a Zeiss LSM 510 confocal microscope.
Scanning Electron Microscopy of Corneal Limbal Stroma
Corneoscleral rims were decellularized by incubation in 100 mM EDTA for 2 hours at 37°C followed by gentle mechanical abrasion of the epithelium. Specimens were fixed, processed for scanning electron microscopy, coated with gold palladium, and imaged in a JEOL 6100 SEM (Jeol Ltd., Tokyo, http://www.jeol.com) operating at 15 kV. This protocol effectively removed virtually all epithelial cells. The integrity of the underlying basal lamina and stroma were assessed by scanning and transmission electron microscopy and were found to be unaffected by the decellularization protocol. Eighteen rims from 15 donors (age range, 3–70 years; mean, 46.5 years; median, 52 years) were examined. A continuous array of overlapping images collected at x25 magnification were combined to create a 360° montage of the corneal limbus (Adobe Photoshop 7.0).
Culture of Limbal Epithelial Holoclone (Stem Cell) Colonies from Corneal Limbal Biopsies
Clock hours were marked out on normal corneas using gentian violet. Confocal imaging of the limbus was then performed in an aseptic manner using the Heidelberg Retinal Tomograph (HRT) II and Rostock corneal module (Heidelberg Gmbh, Heidelberg, Germany, http://www.heidelberg.com). Areas that contained large numbers of proposed niche structures—limbal crypts (LCs) and focal stromal projections (FSPs)—and areas in which these were absent were identified. The locations of these were noted using the clock hours. The features used to designate and area as containing LCs and FSPs were the presence of their distinct morphological characteristics (as demonstrated in Figs. 1, 2, and 6) and the presence of multiple structures in close proximity. This information was then used to target punch biopsies, 3 mm in diameter, to regions that were either rich in, or devoid of, these proposed niche structures. In total, 30 biopsies (16 from niche rich regions and 14 from non-niche regions) were taken from five donor corneas (age range, 27–79 years; mean, 63.2; median, 66 years). Biopsies were incubated with 2.4 units/ml dispase II (Roche Diagnostics, Basel, Switzerland, http://www.roche-applied-science.com) at 37°C for 60 minutes followed by gentle scraping to separate the epithelium from the limbal stroma. A single-cell suspension was obtained by incubation with 0.5 g of trypsin and 0.2 g of EDTA-4Na/L (Invitrogen, Carlsbad, CA, http://www.invitrogen.com) at 37°C for 3–8 minutes. Feeder layers of growth-arrested 3T3-J2 cells were prepared by incubation with 4 µg/ml mitomycin-C for 2 hours and seeded at 4 x 104 cells per cm2. All epithelial cells isolated from each biopsy were seeded onto separate feeder layers and cultured in a humidified atmosphere and 5% CO2 in Dulbecco's modified Eagle's medium and Ham's F12 basal medium (2:1 mixture) containing glutamine (4 mM), fetal calf serum (10%), hydrocortisone (0.4 µg/ml), cholera toxin (0.1 nM), insulin (5 µg/ml), penicillin/streptomycin (50 units/ml each), adenine (0.18 mM), triiodothyronine (2 nM), transferrin (5 µg/ml), and epidermal growth factor (10 ng/ml). After 10–12 days, cultures were fixed with 100% methanol at –20°C for 1 hour, rehydrated, and stained with 2% Rhodamine B. Colonies were examined under a dissecting microscope. Holoclone colonies formed by clonal expansion of single limbal epithelial cells were identified and counted as previously described [23, 24]. Results are expressed as the number of holoclone colonies per square millimeter of tissue biopsy.
|
|
(Cell Signaling Technology, Beverly, MA, http://www.cellsignal.com) at 4°C overnight. After washing, sections were incubated with secondary antibodies (FITC-conjugated goat anti-mouse IgG or TRITC-conjugated goat anti-rabbit IgG, 1:200; Jackson Immunoresearch) and 1 µg/ml 4'-6-diamidino-2-phenylindole (DAPI; Sigma-Aldrich) for 2 hours at room temperature. Slides were washed and mounted in Mowiol 4-88 (Sigma-Aldrich) and coverslipped. Sections were examined and images collected using a Zeiss LSM 510 confocal microscope.
In Vivo Confocal Imaging of the Human LESC Niche Structures
In vivo confocal microscopy was performed on the corneas of normal subjects to determine whether LCs and FSPs could be identified in vivo. Ethics committee approval and informed consent were obtained for all in vivo human studies. The corneal limbus was examined in vivo using the HRT II with the Rostock corneal module [25]. This scanning laser confocal microscope allows the collection of en face optical sections at varying depths within the cornea and corneal limbus. Initially, the corneal limbus of 10 normal volunteers (six male, four female; age range, 22–59 years; mean, 44.7 years; median, 47 years) were examined. Subsequently, we examined the corneal limbus of eight patients with LESC deficiency (six male, two female; age range, 32–80 years; mean, 47.5 years; median, 42 years) due to aniridia (n = 4) or chemical burns (n = 4). A minimum of 10 volume scans were collected from each of the superior, inferior, nasal, and temporal corneal limbal quadrants. Each volume scan comprised a series of images starting at the limbal epithelial surface and continuing to a depth of 180 µm, with images collected at 1.5-µm intervals.
Quantification of Findings
Extent of Proposed Niche Structures. The 360° corneal limbal montages constructed from confocal and SEM images were imported into Adobe Photoshop 7.0. The extent of corneal limbus containing putative niche structures was quantified in degrees of arc using the geometric center of the montage as the center point of the arc.
Nuclear to Cytoplasmic Ratio and Cell Size Measurements. Five representative images of each of the proposed niche structures identified (LCs and FSPs) were randomly selected from each of the 10 donors. The nuclear to cytoplasmic ratio (NCR) and cell size of 250 of the cells lining the bases and edges of LCs (as demonstrated in Fig. 1F and 1G) were measured using ImageJ (NIH, Rockville, MD, http://www.nih.gov) and Zeiss LSM 510 software. Similarly, the NCR and cell size of 250 basal cells immediately attached to the edges of focal stromal projections (as demonstrated in Fig. 2E and 2F) were measured. The NCR and cell sizes of these two groups of cells were compared first with 250 suprabasal cells that were located immediately adjacent to them within LCs and FSPs, and second with basal limbal cells that did not reside within LCs and FSPs.
Statistical Analysis
When analyzing differences in cell sizes and nuclear to cytoplasmic ratios, and differences in the number of holoclone colonies cultured from different biopsy regions, a two-tailed t test was used to compare groups. Where means are reported, they are followed by the SE.
|
RESULTS |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDisclosure of Potential...AcknowledgmentsReferences |
|---|
Limbal Crypts. Distinct invaginations of epithelial cells extended from the peripheral cornea into the corneal limbal stroma (Fig. 1B; supplemental online data). We termed these structures limbal crypts. They are similar to the rete pegs of the epidermis in that they were downward projections of the limbal epithelium into the limbal stroma, but in other respects, they have similarities to the epithelial crypts of the gut in that they are extremely well circumscribed, are shorter in length than rete pegs, and are polarized so far as they opened onto the corneal rather than the conjunctival surface. The stroma that surrounds LCs is highly cellular and contains a distinct vascular supply that is closely associated with the crypts (Fig. 1E; supplemental online data). SEM images of decellularized corneal limbus in a region rich in LCs demonstrated the manner in which the limbal stroma is specialized to enclose the LCs (Fig. 1C, 1D). The previously described limbal palisades of Vogt [26–28] enclose the LCs laterally and are continuous posteriorly, with a less prominent ridge of limbal stroma that encircles the posterior aspects of the palisades. Limbal stroma borders the inferior aspect of the LCs.
Focal Stromal Projections. These are fingerlike projections of stroma containing a central blood vessel (Fig. 2B), that extend upward into the corneal limbal epithelium and are surrounded by small, tightly packed basal cells (Fig. 2A; supplemental online data). The structure of FSPs was further demonstrated on SEM (Fig. 2B, 2C) and 3D reconstruction of the confocal optical sections (supplemental online data). FSPs are located at the corneal end of, but are distinct from, the LCs (Fig. 3).
|
The cells that are directly adherent to FSPs (highlighted in Fig. 2E, 2F) are significantly smaller (mean diameter, 9.25 ± 0.10 µm; range, 6.21–15.22 µm) and have a higher nuclear to cytoplasmic ratio (mean, 0.76 ± 0.01; range, 0.50–0.98) than suprabasal cells that were located immediately adjacent to them (mean diameter, 18.09 ± 0.23 µm; range, 10.1–29.76 µm; p < .001; and mean NCR, 0.44 ± 0.21; range, 0.21–0.70; p < .001). The cells that are directly adherent to FSPs were also found to be significantly smaller and have a larger NCR than basal limbal cells that did not reside within LCs or FSPs (p < .001).
There was no significant difference in cells size or NCR between the cells lining the edges or bases of LCs (Fig. 1F, 1G) and FSPs (Fig. 2E, 2F). These values suggest that these cells may represent LESCs.
Regional Variation in Distribution of LCs and FSPs
The degrees of arc containing these proposed niche structures were quantified in ten 360° confocal montages (Fig. 3A, 3C) and 18 SEM montages (Fig. 3B, 3D). The results are summarized in Table 1. These results demonstrate that LCs and FSPs are not uniformly distributed around the corneal circumference; rather, they are predominantly located in the superior and inferior corneal limbal quadrants and extend to a varying degree temporally and nasally. They were absent in the horizontal meridian of all donors.
|
|
were examined and compared between the central cornea, LC-/FSP-rich limbal regions and non-LC/-FSP limbal regions. ABCG2 and p63 were expressed only by limbal basal epithelial cells and not by the corneal epithelium. Levels of ABCG2 and p63 expression were markedly higher in LC-/FSP-rich regions than in non-LC/-FSP regions (Fig. 5). The highest levels of p63 and ABCG2 expression were observed in the cells highlighted in Figures 1F and 1G (LCs) and 2E and 2F (FSPs).
|
|
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDisclosure of Potential...AcknowledgmentsReferences |
|---|
The limbal palisades of Vogt have previously been proposed as the site of this niche [26]. Clinically, these can be observed using a slitlamp microscope and appear as radial linear structures measuring up to 1 mm in length [27, 28]. Photomicrographic, angiographic, and histological studies have demonstrated the fibrovascular nature of the palisades and the presence of "ridges of thickened epithelium" in the interpalisade zone [27, 28]. The advent of confocal microscopy has provided the opportunity to optically section the corneal limbus and create 3D reconstructions of the tissue. By applying these techniques together with SEM, we have for the first time performed a detailed 3D study of the entire human corneal limbus. This approach allowed us to identify previously unrecognized candidates for the LESC niche, LCs and FSPs, and has significantly advanced our understanding of the structure of this adult stem cell niche. Our findings show that the structures traditionally ascribed as the limbal niche (the limbal palisades of Vogt) are only one facet of the limbal niche architecture, forming the lateral walls of the stromal structures that encircle LCs. These results demonstrate how state-of-the-art 3D imaging can be used to search for and detect adult stem cells niches.
It is widely accepted that LESCs are located in the basal layer of the corneal limbal epithelium. Although there is functional evidence that the rabbit corneal limbus serves as a niche [29], to date, there is no direct evidence that the same applies to human corneal limbus. We have, for the first time, provided such evidence. Firstly, regions rich in LCs and FSPs contain a larger number of cells that express higher levels of putative LESC markers than do regions of the corneal limbus that do not exhibit these structures. Secondly, a small cell size [30, 31] and a high nuclear to cytoplasmic ratio [22, 32] are accepted characteristics of LESCs. Our results indicate that both of these structures contain populations of cells that have a morphology in keeping with published values for LESCs [30, 32]. Thirdly, biopsies from LC-/FSP-rich regions give rise to significantly more limbal holoclone (stem cell) colonies in culture than do biopsies from non-LC/-FSP regions. Finally, we have demonstrated that the presence of these structures is associated with normal corneal epithelial function but that, in patients with LESC deficiency, because environmental insult (chemical injury) or genetic abnormalities (aniridia/pax6 haploinsufficiency) these structures are undetectable. These data provide a unique biological insight into the importance of this novel adult stem cell niche in maintaining normal LESC function.
The absence of a definitive marker for LESC has proved challenging to the study of the biology of and in particular to the localization of these adult stem cells [33]. Our finding that the putative LESC markers ABCG2 and p63 were expressed only by limbal basal cells and not by the corneal epithelium is consistent with previous studies [22, 34]. Recently Chen et al. [22] demonstrated that the entire basal cell population of the palisades of Vogt at the superior corneal limbus demonstrate characteristics consistent with a LESC phenotype, including the expression of the putative LESC markers ABCG2 and p63. However, other reports have demonstrated that ABCG2 [35–37] and p63 [34] are not uniformly expressed by limbal basal cells but rather are expressed by clusters of cells. Our demonstration of regional variation in limbal structure may explain these apparently contradictory results. Furthermore, the search for a definitive LESC marker may have been hampered by the investigation of potential markers in inappropriate limbal regions.
Our results demonstrate that putative LESCs are found lining the edges and bases of LCs. In addition, they were also located around the sides and tips of FSPs. Current methods cannot identify for certain which exact cells within this cell population are truly LESCs. The dual location and the difference in distribution of putative LESCs is interesting, but its significance is unclear. In the skin and the intestine, the stem cell compartment is located in the bases of crypts [38–41]. Our localization of putative LESCs to LCs is consistent with the distribution of stem cells in other epithelia. Our identification of FSPs and the presence of putative LESCs surrounding their sides and tips appears to be unique to the corneal limbus. The existence of stem cells outside of crypts appears to contradict the general organizational principles of epithelial stem cell compartments; however, there is evidence that epidermal keratinocyte stem cells are not confined solely to the bases of rete ridges [42, 43].
The presence of a rich vascular supply is a common feature of both LCs (Fig. 1E; supplemental online data) and FSPs (Fig. 2B). The presence of a dedicated vasculature network suggests that this is an essential component of the niche environment, supplying nutrients and possibly survival factors to LESCs.
In contrast to our findings, Dua et al. [45] recently described distinct subconjunctival structures, termed limbal epithelial crypts, which the authors propose as a niche structure and as the location of LESCs. The unique distinguishing features of these proposed structures are that they extended from the peripheral aspect of the undersurface of an interpalisade rete ridge and extended either radially into the conjunctival stroma parallel to the palisade or circumferentially along the corneal limbus at right angles to the palisade. The authors employed serial sectioning of the cornea and found an average of six limbal epithelial crypts per cornea, but the number of corneas examined (five) was small, and detail on the circumferential distribution and exact number present in each cornea was not reported [45]. These structures were not detected in any of the 38 corneas examined in the present study.
In vivo confocal microscopy has provided a significant advance in corneal imaging in human patients. We have used this technology to locate and observe an adult stem cell population in vivo for the first time. The ability to directly visualize LESC allowed us to study these cells in normal and in disease conditions. Although the identification of a definitive LESC-specific marker would be of immense value, there is no guarantee that this will be achieved. If, however, it were achieved, there would be extensive technical, safety, and regulatory obstacles to using an antibody to such a marker to identify stem cells in humans. In contrast, the parameters of structural niche characteristics, cell size, and nuclear to cytoplasmic ratio that we have employed are immediately translatable to clinical practice in the following ways. Firstly, documentation of the absence of LCs and FSPs can assist ophthalmologists in making the clinical diagnosis of LESC deficiency. Secondly, it enables the targeted biopsy of stem-cell-rich regions of corneal limbus for therapeutic stem cell cultures. At present, limbal biopsies are taken from areas in which limbal palisades can be identified clinically. However, in up to 20% of patients, limbal palisades cannot be identified clinically [28]. Ex vivo expansion and therapeutic transplantation of LESCs is dependent on harvesting LESCs from the corneal limbus of living (autologous or allogeneic) or cadaveric (allogeneic) donors. Our findings demonstrate how this variability can be overcome and how the presence or absence of these proposed niche structures can be clearly and reliably documented. We are currently investigating whether such targeted biopsies result in an improved clinical outcome after transplantation of cultured LESCs in humans.
Transplantation of ex vivo cultured LESCs is one of the few adult human stem cell therapies to have been successfully and widely employed [11–20]. Many of the issues and obstacles encountered during the development of this therapy are likely to be replicated during the development of future stem cell therapies for other tissues. The novel concept of applying high-resolution 3D imaging to locate adult stem cell niches may facilitate the identification and spatial localization of previously unrecognized niches in other tissues and organs. Our unique concept of targeted niche biopsy may prove to be essential for the successful development of novel adult stem cell therapies because harvesting an optimal population of cells will maximize culture efficiency and, possibly, clinical outcomes after transplantation.
|
DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDisclosure of Potential...AcknowledgmentsReferences |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDisclosure of Potential...AcknowledgmentsReferences |
|---|
|
REFERENCES |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDisclosure of Potential...AcknowledgmentsReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| STEM CELLS | THE ONCOLOGIST | CME | ALPHAMED PRESS JOURNALS |
