A population of keratinocyte stem cells in defined locations governs the renewal of human stratified epithelia (1–7). These stem cells generate transient amplifying (TA) cells that terminally differentiate after a discrete number of cell divisions (1–7). Corneal stem cells are segregated in the basal layer of the limbus (1–3,8), which is the transitional zone of the epithelium located between the cornea and the bulbar conjunctiva. Therefore, corneal epithelium is formed exclusively by TA cells that continuously migrate from the limbus (7).
Ocular burns are characterized by depletion of limbal cells thereby leading to corneal re-epithelialization by bulbar conjunctival cells (9–11). This abnormal wound healing induces neovascularization, chronic inflammation, and stromal scarring. Total limbal stem cell deficiency eventually leads to corneal opacification and visual loss (9–11). Allogeneic corneal grafts, aimed at replacing the corneal stroma, are not successful in these patients unless the limbal stem cell population has been restored by limbal grafts taken from the uninjured eye (12). This procedure, however, requires a large limbal withdrawal from the healthy eye and is not suitable for bilateral lesions.
We have shown that autologous cultured limbal cells restored the corneal surface of two patients with limbal cell deficiency (13). The possibility of using limbal cultures to restore damaged corneal surfaces was further suggested by Schwab et al (14). Recently, Tsai et al. (15) cultured limbal epithelial cells onto amniotic membrane, in the absence of feeder-layer. Maintenance of stem cells has not been evaluated in their new culture system. This is particularly relevant, because preservation of epithelial stem cells has been demonstrated only in the presence of feeder-layers (3,16–23). Furthermore, amniotic membrane-cultured limbal cells were applied predominantly on patients affected by partial limbal deficiency (15), not ruling out the possibility that cultures were not permanently engrafted but were stimulating proliferation of resident limbal cells. Collectively, the absence of controls of stem cell maintenance in culture, the inadequate number of patients with a precise diagnosis of limbal stem cell deficiency, and the scarce handiness of cultured epithelial sheets is banishing this technology at a developmental stage far from daily practice.
In this study, we selected a homogeneous group of patients whose limbal stem cell deficiency was evaluated by scoring the gravity of the clinical picture and the expression pattern of keratins. We then investigated whether a careful control of stem cells in cultured limbal grafts allowed a permanent restoration of corneal integrity in patients with total limbal stem cell deficiency.
MATERIALS AND METHODS
Fibrin Sealant and Cell Culture
The fibrin sealant (Tissucol, Baxter-Immuno, Wien, Austria) was prepared by mixing 330 μl of thrombin solution and 330 μl of fibrinogen solution (prepared as described, 21) into a plastic ring (diameter 3 cm) previously placed on a Petri dish not treated for cell culture (Fig. 1A, arrows).
Limbal keratinocytes were cultivated on a feeder-layer of lethally irradiated 3T3-J2 cells as described (3). Colony forming efficiency and number of cell generations were evaluated as described (3). Clonal analysis was performed from sub-confluent secondary cultures as described (3,18). Briefly, single cells were inoculated onto multi-well plates containing feeder layers. After 7 days, clones were identified under a microscope, photographed, and transferred into three dishes. Two dishes (three-quarters of the clone) were used for serial propagation and further analysis. The third dish (one-quarter of the clone) was fixed 9–12 days later and stained with Rhodamine B for the classification of clonal type, which was determined as described (3,18).
Definition of Limbal Stem Cell Deficiency and Selection of Patients
We defined the degree of limbal stem cell deficiency by arbitrarily scoring the gravity of the clinical picture and the cytological state of the corneal surface. Normal controls were scored 0.
Patients with haziness of the corneal epithelium and/or recurrent epithelial defects were scored 1; patients with persistent epithelial defects were scored 2; patients with total corneal conjunctivalization were scored 3.
Conjunctival and limbal-corneal epithelia express keratin type 19 and keratin type 3, respectively (1). Impression cytology was performed by applying a Millipore filter acetate on the corneal surface and by immunostaining the impressed cells with K3-specific AE5 monoclonal antibody (mAb) and K19-specific RCK108 mAb (DAKO). Corneal surfaces covered predominantly by K3+ cells (≥80%) were scored 1, whereas those covered predominantly by K19+ cells (≥80%) were scored 3. Score 2 indicates that the percentage of K3+ and K19+ cells ranged between 20% and 80%.
We defined three grades of limbal deficiency: total, when both clinical and cytological scores were 3; severe, when both scores were 2; slight, when both scores were 1. Eighteen patients (mean age 48±12 years) with damaged corneas due to chemical burns were included in this study (Table 1). Fifteen patients had total limbal deficiency, while three patients had severe limbal deficiency unresponsive to conventional surgical therapy.
Transplantation of Autologous Fibrin-Cultured Limbal Cells
A 1–2 mm2 limbal biopsy was taken from the contralateral eye of each patient. Limbal keratinocytes were cultivated as above. Sub-confluent primary cultures were trypsinized and plated at a cell density of 5×103/cm2 on the circular fibrin sealant, in the presence of feeder-layer. At confluence, fibrin-cultured epithelial sheets were washed in Dulbecco’s minimum essential medium (DMEM) containing 4 mM of glutamine and penicillin-streptomycin (50 IU-50 μg/ml), placed in sterile holders for contact lenses (Bausch & Lomb, Fig. 1A) filled with DMEM, equilibrated at 37°C in 5% CO2 humidified atmosphere for 15 min, and transferred at room temperature to the hospital.
Limbal biopsies were processed within 24 hr from withdrawal. Primary cultures and preparation of fibrin-cultured grafts required 14–16 days. Grafts were applied within 24–36 hr after their transfer in the contact lens holder.
A 360-degree conjunctival peritomy was performed under retrobulbar anesthesia to remove abnormal epithelium and fibrovascular tissue. Fibrin-cultured epithelial sheets were placed on the prepared corneal-scleral surface and fitted under the dissected conjunctiva. The conjunctiva was sutured with two 8/0 vicryl sutures. Two 4/0 silk sutures were used to close the eyelids. Three days later, the eyelid sutures were removed and prophylactic antibiotics eye drops were given three times a day for 2 weeks. All patients were examined at 1, 2, and 4 weeks after surgery. Patients were then examined approximately every 3 months up to the present. Impression cytology was performed at approximately 1-year follow-up.
For immunohistochemistry, specimens were fixed in paraformaldehyde (4% in phosphate-buffered saline) 30 min at room temperature and embedded in paraffin. Sections were stained with K3-specific AE5 mAb and K19-specific RCK108 mAb (DAKO), as described (3).
Cultivation of Limbal Cells on Fibrin and Clonal Analysis
Fibrin is a highly manageable and quickly degradable natural substrate. Clonogenic ability, overall size, and morphology and growth rate of limbal keratinocytes grown on plastic or on fibrin were identical. Limbal keratinocytes (6 strains) cultivated on plastic or on fibrin underwent 93±8 and 95±7 divisions before senescence, respectively. Fibrin-cultured cells, isolated from a 1-mm2 limbal biopsy, generated an epithelial sheet 3-cm in diameter in 14–16 days (Fig. 1A, arrows). The epithelial sheet was composed of a well-conserved basal layer formed by cuboidal cells and of several suprabasal layers (Fig. 1B).
Human keratinocyte stem and TA cells when isolated in culture give rise to holoclones and paraclones, respectively (3,18,19,21,22). The proliferative compartment of stratified epithelia also contains a third type of cell, the meroclone, which is considered a “young” TA cell endowed with a greater proliferative capacity than the paraclone (3,18). Preservation of stem cells on fibrin was therefore evaluated by clonal analysis of sub-confluent secondary limbal cultures. We analyzed a total of 118 and 105 clones obtained from limbal cells cultivated on plastic and fibrin, respectively. In both culture conditions, the majority of clones (84.9% and 83.7%, respectively) were classified as TA cells (meroclones and paraclones) (Fig. 1, panels C and D, bars M and P), whereas stem cells (holoclones) represented 15.1% and 16.3% of total clones, respectively (Fig. 1, panels C and D, bar H). Ten randomly selected holoclones (5 for each condition) were serially propagated and each holoclone produced more than 90 generations before senescence, accounting for the entire proliferative capacity of the original mass culture. Figure 1E shows the progeny of a quarter of a colony generated by a single fibrin-cultured holoclone. This colony generated 318 large and smooth daughter colonies, each of which is potentially able to give rise to enough epithelium to cover an entire corneal surface in approximately 14 days (Fig. 1F).
Thus, fibrin preserves stem cells and represents a suitable carrier for human limbal keratinocytes destined to autologous transplantation.
Transplantation of Autologous Fibrin-Cultured Limbal Cells on Patients with Total Limbal Stem Cell Deficiency
At a clinical level, limbal deficiency is characterized by persistent epithelial defects, corneal conjunctivalization, and visual loss (9–11). Conjunctival and limbal-corneal epithelia express keratin type 19 and keratin type 3, respectively (1). Thus, at a cytological level, limbal deficiency is characterized by a predominance of K19+ cells and by a strong decrease, or even absence, of K3+ cells on the corneal surface (10,24,25). We graded limbal stem cell deficiency by scoring the gravity of the clinical picture and the keratin expression pattern (Table 1, see Methods). Eighteen patients, all suffering from chemical burns, were enrolled in this study. Fifteen patients had total limbal deficiency, whereas three patients had severe limbal deficiency (Table 1, baseline). Note that (a) nine patients (including the three patients with severe limbal deficiency) underwent single or multiple unsuccessful lamellar or penetrating keratoplasty; (b) in all patients, visual acuity was compromised; (c) all patients reported persistent burning, pain and photophobia, and dependence on artificial tears.
Figure 2 shows the grafting procedure and the clinical results obtained in patient 2. This patient had an alkali burn on his right eye 4 years earlier and two unsuccessful keratoplasties afterward. At admittance, conjunctivalization was complete (Fig. 2A) and patient’s visual acuity was reduced to 0.1. A 360-degree conjunctival peritomy and dissection of the fibrovascular pannus (Fig. 2B, arrow) were performed. Autologous fibrin-cultured limbal epithelium was then applied on the prepared wound bed (Fig. 2, C and D, arrows) and fitted under the dissected conjunctiva, which was sutured in place (Fig. 2E). Figure 2F shows the appearance of the patient’s eye 12 months later. His cornea was covered with a transparent, normal-looking epithelium without vascularization and epithelial defects. The patient recovered a visual acuity of 0.6.
Table 1 summarizes clinical results obtained in 18 patients after 12–27 months follow-up. Clinical success was determined by the improvement of patients’ symptoms (clinical signs) and by stable regeneration of the corneal epithelium (corneal transparency and corneal cytology). Limbal cultures were successful in 14 of 18 patients, whereas negative results were observed in 4 patients who did not show any clinical or cytological improvement. Persistent inflammation and bleeding, observed during the early postoperative course, were probably the cause of failure in these patients. Results were further analyzed in the remaining 14 patients by evaluating clinical signs and corneal cytology (Table 1).
In all patients, complete re-epithelialization occurred within the first week. Inflammation and vascularization regressed within the first 3–4 weeks. By the first month, the corneal surface appeared clear, smooth, wettable, and covered by a transparent normal-looking epithelium, as also demonstrated by fluorescein staining (not shown). Furthermore, all patients reported a stable improvement of their symptoms (burning, pain, and photophobia). At 12–27 months follow up, the average score for clinical signs improved from 2.7±0.5 to 0.07±0.2.
Restoration of the corneal surface was also analyzed by impression cytology and evaluation of the relative percentage of K3+ and K19+ cells. At 12–27 months follow up, the average score for corneal cytology improved from 2.8±0.4 to 0.8±0.5. Thus, the average total (clinical and cytological) score of these 14 patients improved from 5.5±0.8 to 0.9±0.7. Statistical analysis showed a direct correlation between clinical signs and corneal cytology at baseline as well as at 1-year follow-up (P <0.01, Rho=0.70 and P <0.001, Rho=0.99, respectively). Although few patients showed an improvement of their vision (see patients 2, 3, and 16), overall visual acuity was not significantly improved at 1-year follow-up, compared to the baseline value (Table 1). This is expected because visual acuity also depends on the integrity of the corneal stroma (see below).
Restoration of Normal Vision by Grafting Of Autologous Fibrin-Cultured Limbal Cells Followed by Penetrating Keratoplasty
Figure 3 shows the clinical results obtained in patient 14. This patient had an acid burn on his right eye 18 months earlier and an unsuccessful lamellar keratoplasty afterward. At admittance, slit-lamp examination showed complete conjunctivalization (Fig. 3A). Visual acuity was reduced to light perception, and his corneal surface was covered almost entirely (≥90%) by K19+ conjunctival cells (Fig. 3, B and C). After grafting of fibrin-cultured cells, his cornea was covered by a stable and transparent epithelium, with absence of vascularization (Fig. 3D). The patient referred a dramatic improvement of his discomfort. Impression cytology, performed 1 year after grafting, confirmed that his cornea was covered by K3+ corneal cells (≥90%) (Fig. 3, E and F). However, because of the residual stromal scarring (Fig. 3D, asterisks), the patient had only a partial improvement of his visual acuity (0.2). A penetrating keratoplasty was therefore performed 13 months after grafting of limbal cultures. At 3-month follow-up, the corneal stroma was restored (Fig. 3G), and the patient fully recovered his visual acuity (1.0). At the last follow-up, 24 months after grafting of limbal cultures and 11 months after keratoplasty, the patient was clinically and cytologically stable and maintained a visual acuity of 1.0. Histologic examination of the engrafted corneal epithelium removed at the time of keratoplasty showed a stratified epithelium resembling a normal cornea (Fig. 3, H and I). As expected, the epithelium uniformly expressed K3 (Fig. 3H) but not K19 (Fig. 3I), further demonstrating that cultured stem cells have been engrafted and were able to permanently restore the corneal surface. Similar results were obtained in patients 4 and 7 (Table 1).
We showed that limbal stem cells are preserved when cultivated on fibrin and that transplantation of autologous fibrin-cultured limbal cells permanently restores corneal integrity of patients with total limbal deficiency unresponsive to conventional surgical therapy. The multicenter nature of this study and the handiness and ease of long-distance transportation of the fibrin-cultured epithelial sheets suggest that this technology can be widely applied.
Preservation of Limbal Stem Cells
Holoclones have been identified as stem cells for surface epithelia (3,18,19,21,22,26). The great proliferative capacity of the holoclone (3,18,19,22), the ability of a single-cultured holoclone to generate a mature epithelium in vivo (27) and to differentiate into distinct cellular lineages (3), and the permanent epidermal regeneration obtained with autologous cultured epithelial grafts bearing holoclones in burn victims (17,21,23) provide compelling evidence that that keratinocyte “stem-ness” can be preserved in culture. However, irreversible stem to TA cell conversion can occur very rapidly under incorrect culture conditions. For instance, keratinocyte clonal growth and/or preservation of epidermal holoclones have been demonstrated when keratinocytes were cultivated in the presence (3,16–23) but not in the absence (our unpublished data) of a proper feeder-layer of lethally irradiated fibroblasts. Replacement of self-renewing tissues requires permanent engraftment of stem cells. Therefore, the proposal of a new culture system for epithelial cells destined to cell therapy of stem cell deficiency should be preceded by the demonstration of maintenance of holoclones in culture. Evaluation of stem cells is probably not essential when cultured cells are used as biological medications for the repair of partial epithelial defects, because in this clinical context permanent engraftment is not required. For instance, allogeneic cultured keratinocytes induce healing of chronic leg ulcers and of partial-thickness burns (28), whereas amniotic membrane itself can induce re-epithelialization of recurrent corneal epithelial defects (29). In both cases, these devices presumably stimulate proliferation of resident stem cells.
Permanent Restoration of Total Limbal Stem Cell Deficiency
Several techniques have been proposed to restore damaged corneas. Autologous limbal transplantation is successful (12) but requires a large limbal withdrawal from the donor eye (30–40% of the donor limbus). The possibility of covering the entire corneal surface using a 1-mm2 limbal biopsy avoids potential damage to the healthy eye, and in the presence of a spared limbal area, might allow treating also bilateral lesions. Furthermore, limbal cells can be cryopreserved and used in the case that the first graft fails. These benefits compensate the complexity and the cost of the cultivation. Allogenic limbal grafts require immunosuppression (30). Amniotic membranes promote re-epithelialization in partial limbal deficiency (29) but require resident limbal stem cells. Therefore, the use of autologous cultured cells is particularly relevant in clinical settings characterized by a total destruction of the limbal epithelium and is also essential for the clinical success of subsequent keratoplasty (12,13).
The diagnosis of limbal deficiency is based on the clinical and cytological evaluation of the patient. In this study, we propose a grading of limbal deficiency based on the gravity of clinical signs and on the keratin expression pattern. This helped us in selecting patients with total limbal deficiency and in evaluating their clinical response. Fourteen of 18 patients showed a permanent recovery of the corneal epithelium. Although we cannot formally prove that cultured cells were stably engrafted, the reproducible disappearance of conjunctivalization, the reconstitution of a normal-looking epithelium consisting of K3+ cells, and the restoration of visual acuity (see patients 2, 3, 4, 7, and 14) in patients otherwise unresponsive to conventional surgery strongly suggest a permanent engraftment of cultured cells.
We suggest that cultivation of limbal cells might offer an alternative to patients with unilateral lesion and a therapeutic chance to patients with severe bilateral corneal-epithelial loss, provided that stem cells are maintained in culture. Such cultivation also gives new perspectives on the treatment of ocular disorders, such as thermal burns, severe microbial infections, cicatricial pemphigoid, and Stevens-Johnson syndrome, characterized by stem cell deficiency. We suggest, however, that this approach should be limited to pathologies characterized by a total destruction of the limbal-corneal epithelium, because partial defects can be restored by alternative procedures, as for instance, a simple application of amniotic membranes (29).
The authors are indebted to Drs. Yann Barrandon and Vincent Ronfard for sharing information crucial for the preparation of the fibrin substrate. The authors thank Bausch & Lomb for providing contact lens holders free of charge. Prof. Howard Green kindly provided 3T3-J2 cells. AE5 mAb was kindly provided by Dr. Tung-Tien Sun. We thank Dr. Pietro Maria Donisi for impression cytology and histology.
1. Schermer A, Galvin S, Sun TT. Differentiation-related expression of a major corneal keratin in vivo and in culture suggests limbal location of corneal epithelial stem cells. J Cell Biol 1986; 103: 49.
2. Cotsarelis G, Cheng SZ, Dong G, Sun TT, Lavker RM. Existence of slow-cycling limbal epithelial basal cells that can be preferentially stimulated to proliferate: implications on epithelial stem cells. Cell 1989; 57: 201.
3. Pellegrini G, Golisano O, Paterna P, et al. Location and clonal analysis of stem cells and their differentiated progeny in the human ocular surface. J Cell Biol 1999; 145: 769.
4. Barrandon Y. The epidermal stem cell: an overview. Dev Biol 1993; 4: 209.
5. Fuchs E, Segre JA. Stem cells: a new lease on life. Cell 2000; 100: 143.
6. Watt FM, Hogan BLM. Out of Eden: stem cells and their niches. Science 2000; 287: 1427.
7. Lehrer MS, Sun TT, Lavker RM. Strategies of epithelial repair: modulation of stem cell and transient amplifying cell proliferation. J Cell Sci 1998; 111: 2867.
8. Dua HS, Azuara-Blanco A. Limbal stem cells of the corneal epithelium. Surv Ophthalmol 2000; 44: 415.
9. Shapiro MS, Friend J, Thoft RA. Corneal re-epithelialization from the conjunctiva. Invest Ophthalmol Vis Sci 1981; 21: 135.
10. Tsai RJF, Sun TT, Tseng SCG. Comparison of limbal and conjunctival autograft transplantation in corneal surface reconstruction in rabbits. Ophthalmology 1990; 97: 446.
11. Dua HS. The conjunctiva in corneal epithelial wound healing. Br J Ophthalmol 1998; 82: 399.
12. Kenyon KR, Tseng SCG. Limbal autograft transplantation for ocular surface disorders. Ophthalmology 1989; 96: 709.
13. Pellegrini G, Traverso CE, Franzi AT, Zingirian M, Cancedda R, De Luca M. Long-term restoration of damaged corneal surfaces with autologous cultivated corneal epithelium. Lancet 1997; 349: 990.
14. Schwab IR, Reyes M, Isseroff RR. Successful transplantation of bioengineered tissue replacement in patients with ocular surface disease. Cornea 2000; 19: 421.
15. Tsai RJF, Li LM, Chen JK. Reconstruction of damaged corneas by transplantation of autologous limbal epithelial cells. N Engl J Med 2000; 343: 86.
16. Rheinwald JG, Green H. Serial cultivation of strains of human epidermal keratinocytes: the formation of keratinizing colonies from single cells. Cell 1975; 6: 331.
17. Gallico GG, O’Connor NE, Compton CC, Kehinde O, Green H. Permanent coverage of large burn wounds with autologous cultured human epithelium. N Engl J Med 1984; 311: 448.
18. Barrandon Y, Green H. Three clonal types of keratinocytes with different capacities for multiplication. Proc Natl Acad Sci U S A 1987; 84: 2302.
19. Rochat A, Kobayashi K, Barrandon Y. Location of stem cells of human hair follicles by clonal analysis. Cell 1994; 76: 1063.
20. Jones PH, Harper S, Watt FM. Stem cell patterning and fate in human epidermis. Cell 1995; 80: 83.
21. Pellegrini G, Ranno R, Stracuzzi G, et al. The control of epidermal stem cells (holoclones) in the treatment of massive full-thickness burns. Transplantation 1999; 68: 868.
22. Dellambra E, Golisano O, Bondanza S, et al. Downregulation of 14-3-3sigma prevents clonal evolution and leads to immortalization of primary human keratinocytes. J Cell Biol 2000; 149: 1117.
23. Ronfard V, Rives JM, Neveux Y, Carsin H, Barrandon Y. Long-term regeneration of human epidermis on third degree burns transplanted with autologous cultured epithelium grown on a fibrin matrix. Transplantation 2000; 70: 1588.
24. Kruse FE, Cheng JJY, Tsai RJF, Tseng SCG. Conjunctival transdifferentiation is due to incomplete removal of limbal basal epithelium. Invest Ophthalmol Vis Sci 1989; 30 (suppl): 520.
25. Wei ZG, Sun TT, Lavker RM. Rabbit conjunctival and corneal epithelial cells belong to two separate lineages: further evidence against the “transdifferentiation” theory. Invest Ophthalmol Vis Sci 1996; 37: 523.
26. Pellegrini G, Dellambra E, Golisano O, et al. p63 identifies keratinocyte stem cells. Proc Natl Acad Sci U S A 2001; 98: 3156.
27. Mathor MB, Ferrari G, Dellambra E, et al. Clonal analysis of stably transduced human epidermal stem cells in culture. Proc Natl Acad Sci U S A 1996; 93: 10371.
28. De Luca M, Albanese E, Cancedda R, et al. Treatment of leg ulcers with cryopreserved allogenic cultured epithelium. Arch Dermatol 1992; 128: 633.
29. Lee SH, Tseng SCG. Amniotic membrane transplantation for persistent epithelial defects with ulcerations. Am J Ophthalmol 1997; 123: 303.
30. Tsubota K, Satake Y, Kaido M, et al. Treatment of severe ocular-surface disorders with corneal epithelial stem-cell transplantation. N Engl J Med 1999; 340: 1697.