news and views

were taken in a l-mm2 biopsy from the good , then they were cultured and, on second Cultivating a cure passage, they formed a tightly packed and communicating (confluent) monolayer of cells. For grafting, the ! epi­ for blindness thelium was completely removed from the and limbus of the recipient eye, and replaced with a slightly larger monolayer of Stuart Hodson cultured limbal . The were Damaged can often be repaired using donor grafts, but if the then covered with therapeutic soft contact damage is too great the graft wlll be rejected. This may change with the lenses, and tightly patched for several days. development of a method to Inhibit rejection which uses cultivated cells. The results after the limbal epithelial graft were very promising. The grafted epithelium he human cornea has a special proper­ epithelium can mould a smooth apical and was stable, transparent, multi-layered and Tty known as immune privilege, which basal surface, the limbal epithelial stem cells smooth. One of the patients had suffered an allows tissue grafts from donors to be do not form such a smooth surface. alkali burn to his left eye ten years earlier, and carried out without the usual problems of If the corneal epithelium is lost it can be had undergone three previous unsuccessful immune rejection. However, immune privi­ functionally regenerated by the limbal stem corneal grafts. Before the treatment he had lege is lost at the perimeter of the cornea - cells. But if both the corneal and limbal continual severe corneal vascularization the limb us of the eye - where the transpar­ epithelia are lost, the corneal surface is re­ ( development of blood vessels) and persis­ ent corneal stroma meets the opaque colonized by the other neighbour of the tent ulceration, and the eye was painful and (Fig. 1). So, if too large a corneal button is limbal epithelium - the conjunctiva! epi­ blind. But after the conjunctiva! epithelium grafted and it approaches the limbus of thelium'-5. After a cornea has been grafted that was covering his cornea was replaced the recipient, the graft is usually rejected. into a recipient, its outer surface will eventu­ with cultured limbal epithelium, a stable and Similarly, the prognosis is not good if the ally be covered by a host-derived epithelium transparent corneal epithelium was recon­ recipient's cornea is so badly damaged that that migrates in from the limbus. So, if the stituted, and there were no further problems. there is no sign of healthy corneal tissue. recipient has limbal epithelial cells, a normal His vision improved so much that he could A solution to these problems may now corneal epithelium is generated. But if the count fingers from one metre away. How­ be at hand with the publication of a report by eye has been severely damaged and the recip­ ever, because of his previous experiences he Pellegrini et al. 1 in The Lancet. They have ient's limbal epithelium is missing, the donor refused the next logical step, which was to extended immune privilege in an unexpected cornea will be covered, by cell migration, check whether his cornea had recovered its way with the aid of human stem cells, and with cells from the conjunctiva! epithelium. immune privilege and could now take a full­ their results promise a wider use of corneal Conjunctiva! epithelia and corneal/limbal thickness graft. transplantation in what are currently thought epithelia represent two different cell lines6, The other patient, whose epithelial graft to be relatively untreatable conditions. and the conjunctiva! epithelium cannot using the new method had also been a The outer layer of the cornea comprises form smooth apical and basal surfaces if it is success, agreed to a second operation of a the regularly dividing and sloughing corneal allowed to colonize the cornea. The result is a full-thickness graft. The result was excellent epithelial cells. Studies into the origin of loss ofvisual acuity by scarring of the stroma, - the eye that had been blind for many these cells showed that they are not derived as well as severe discomfort. years, and had been perforated after an earli­ by self-perpetuation - instead they come Pellegrini et al. 1 now describe a method er unsuccessful graft, was restored to a visual from stem cells in the limbal epithelium2•3, for using limbal stem cells to advantage in acuity of a high enough standard to allow which occupies an annulus (ring) about 1.5 corneal epithelial grafts. Two patients, who him to drive a vehicle safely using the mm wide between the corneal epithelium each had one severely damaged cornea that restored eye alone. His previously untreat­ and the conjunctiva! epithelium (Fig. 1). The was covered with conjunctiva! epithelium, able blindness was cured after 29 years. transition from corneal to limbal epithelial were willing subjects after conventional The clinical results are surprising and cells is quite abrupt, but whereas the corneal grafting had failed. Limbal epithelial cells encouraging. Clearly, much of the cornea's immune privilege is vested in the limbal/ corneal epithelium, and other epithelial cell lines seem to be unable to inherit this privi­ lege. It would be nice to see the development oflimbal epithelial allografts (for which tis­ sue matching would be necessary), to find out whether they could also extend the immune privilege of the cornea. It could then be possible to restore full vision to millions of totally blind people through limbal epithelial primary grafts. D Stuart Hodson is in th e Department of Optometry and Vision Sciences, Ca rdiff University of Wales, PO Box 905, Cardiff CF/ 3XF, UK. I. Pellegrini, G. et al. Lancet 349, 990- 993 ( 1997). 2. Cotsarelis, G., Cheng, S. Z., Dong, G., Sun, T. T. & Lavker, R. M. ------corneal epithelium Figure I Donor grafts can often be used to repair damaged corneas, as the donor's Cell 57, 20 1-209 ( 1989). can be replaced by stem cells of the limbal epithelium. If the recipient's cornea is too badly damaged 3. Dua, H. & Forrester, J. V. Am./. Ophthalmol. 110, 646-656 ( l990 ). and the neighbouring limbal epithelial cells are also lost, conjunctiva! epithelium will extend to 4. Shapi ro, M. S., Friend, J. & Tho ft , R. A. Invest. Opht/H1/111ol. Vi$ . replace the corneal epithelium and the grafted cornea's 'immune privilege' will be compromised. Sci. 21, 135-142 I 198 1 ). S. Shermer, A., Ga lvin, S. & Sun, T. T. /. Cell Biol. 103, 49-62 ( 1986). Pellegrini et al.' have now developed a method to inhibit rejection, by restoring the corneal surface 6. We i, Z. G., Sun, T. T. & Lavker, R. M. Jm •est. Ophrhalrnol. Vis. using sheets of cultivated host limbal epithelium as a primary graft prior to full-thickness grafting. Sci. 37. 523-533 ( 1996).

NATURE IVOL387 l29MAY 1997 449