Mitotic inhibition of corneal endothelium in neonatal rats

PURPOSE. Corneal endothelium in humans does not divide to any significant e xtent after birth; therefore, with age there is a gradual loss of cells. Wh en cell density is reduced to a critical level, the endothelium cannot func tion to maintain corneal clarity, and the cornea becomes permanently cloudy . Currently, the blindness that results can be treated only by corneal tran splantation. The long-term goal is to find methods to stimulate corneal end othelial proliferation in a clinically relevant manner. The first step towa rd achieving this goal is to identify mechanisms responsible for the induct ion and maintenance of mitotic inhibition of the corneal endothelium in viv o. During corneal development, the endothelium is formed by migration and p roliferation of mesenchymal cells from the ocular periphery. Soon after the monolayer is formed, proliferation ceases. In tissue culture, many cell ty pes cease proliferating upon formation of stable cell-cell and cell-substra te attachments. The goal of the present studies was to determine whether es tablishment of stable contacts correlates with cessation of endothelial pro liferation during corneal development in vivo. METHODS. Corneas from neonatal (days 1, 3, 7, 10, 13, 14, 17, 21, 28, and 4 2) and adult rats were used for immunolocalization of the following: bromod eoxyuridine (BrdU), an S-phase marker; p27kip1 and p21cip1, G1-phase inhibi tors; connexin-43 and ZO-1, proteins associated with gap and tight junction s, respectively; Na+/K+-ATPase and beta 3-integrin, markers of plasma membr ane polarity; and fibronectin and collagen type IV, constituents of Desceme t's membrane. Nuclei staining positively for BrdU were counted to determine the relative number of S-phase cells at various times after birth, Marker protein expression and localization were determined by conventional fluores cence microscopy and by confocal microscopy. RESULTS. The number of endothelial cells staining positively for BrdU gradu ally decreased between postnatal days 1 and 13. After postnatal day 13, pos itive BrdU staining was no longer detectable. During the first postnatal we ek, cells stained positively for the G1-phase inhibitor p27kip1 but not for p21cip1. Connexin-43 achieved its mature location by postnatal day 1. ZO-1 , Na+/K+-ATPase, beta 3-integrin, fibronectin, and collagen type IV achieve d their mature localization patterns between postnatal days 14 and 21. CONCLUSIONS. In neonatal rat, corneal endothelial cells are still entering the cell cycle at birth, but cell cycle entry gradually decreases, so that by postnatal day 13 cells are no longer entering the S-phase. The G1-phase inhibitor p27kip1, but not p21cip1, map help mediate this inhibition. Stabl e cell-cell and cell-substrate contacts gradually form, and monolayer matur ation is complete between postnatal days 14 and 21. The results lead to the hypothesis that, in developing rat cornea in vivo, the establishment of st able cell-cell and cell-substrate contacts initiates a cascade of events, m ediated by p27kip1, which induces mitotic inhibition in the endothelial mon olayer.