In rodent studies, p27
Kip1 has been shown to play critical roles in the timing and fidelity of RPE differentiation as well as in the interdigitation of the RPE monolayer with the outer segments.
9 33 Mice nullizygous for the p27
Kip1 gene have multiple changes in their RPE monolayer compared with their wild-type counterparts, including an increased thickness in the actual monolayer,
10 a higher percentage of binucleated cells, and a decreased association with photoreceptor cells.
9 These findings reveal that p27
Kip1 is essential for the structure and function of the RPE and firmly establish the protein as a primary regulator of the balance between RPE cell proliferation and terminal differentiation. Building on these studies using human RPE-1 cells, we have found that the levels of p27
Kip1 are kept relatively low in proliferating cells (e.g.,
Fig. 3 , middle column, top panel), but escalate in response to UBE2E3 depletion (e.g.,
Figs. 3 4 ). Along with the loss of Ki-67 staining and the doubling of cell area, the changes induced by UBE2E3 depletion are consistent with those observed in vivo for RPE cells as they transition from a state of proliferation to one of terminal differentiation.
9 33 Of note, we did not observe an upregulation of RPE-specific differentiation markers after UBE2E3 knockdown (data not shown). This finding indicates that, at least in cultured RPE-1 cells, depletion of the enzyme may be necessary to drive cells toward differentiation, but is apparently not sufficient to induce terminal differentiation. Nonetheless, a potential implication of the siRNA findings is that during development of the retina, UBE2E3 levels are sustained in RPE cells until enough cells are generated to form the RPE monolayer. UBE2E3 expression then decreases, which results in p27
Kip1 accumulation and triggers cell cycle exit and terminal differentiation. Because of the inherent limitations of using an immortalized tissue culture cell line to test a developmental model, we created a unique mouse strain to evaluate expression of the enzyme in proliferating versus differentiated RPE cells. This mouse strain expresses β-gal-neo driven by the endogenous UbcM2 promoter. A comparison of β-gal activity in proliferating RPE cells from E13 embryos versus differentiated RPE cells from P17 mice revealed a striking, age-dependent decrease in transcription of the UbcM2 gene
(Fig. 6) . We interpret these findings with the β-gal reporter to indicate that UbcM2 gene expression was decreased by greater than threefold in RPE cells that had exited the cell cycle and undergone terminal differentiation. Together, these studies couple UBE2E3 with p27
Kip1 stability and in turn implicate the enzyme as a regulator of the balance between RPE cell proliferation and differentiation. This is the first demonstration that a specific E2 is linked with this process in RPE cells.