Purpose: To determine whether TGF-beta2 exerts inhibitory action in a density dependent manner in primary, first passage, and second passage corneal endothelial cells (CEC).
Methods: Fifty percent confluent cultures were used for actively cycling cells and monolayers were used as contact inhibited cultures. Half of the experiments were performed in cells treated with TGF-beta2 at 10 ng/ml for 24 h. Subcellular localization of cyclin dependent kinase 4 (Cdk4), p27Kip1 (p27), and phosphorylated p27 (pp27) was determined by immunofluorescent staining followed by confocal laser microscopic analysis. Expression of proteins were analyzed by immunoblotting.
Results: Before colocalization between Cdk4 and p27 was studied, the two proteins were respectively stained, either in growing cells for the presence of Cdk4 or in contact inhibited cultures for the presence of p27. Nuclear Cdk4 was observed in FGF-2 treated cells while nuclear staining of Cdk4 was lost in mitogen deprived or TGF-beta2 treated cells. On the other hand, a strong positive staining of nuclear p27 was observed in growth down regulated conditions, which was completely lost in growth up regulated conditions. When cells were double stained with Cdk4 and p27 antibodies, actively cycling cells contained nuclear Cdk4. Less than 10% of the primary cells were positive for Cdk4 staining, whereas all of the second passage CEC contained nuclear Cdk4. Conversely, p27 was not detected in actively cycling cells in either primary or passaged cells. Contact inhibited cells demonstrated nuclear p27 staining in all cells, but only a few cells were positive for nuclear Cdk4. Nuclear Cdk4 was absent when the actively cycling cells were treated with TGF-beta2, whereas TGF-beta2 did not induce the expression of nuclear p27 in the same cultures. In contact inhibited cells, TGF-beta2 did not affect the staining profiles of p27. In the first passage CEC, TGF-beta2 slightly increased the number of cells that were positive for nuclear Cdk4. When the effect of TGF-beta2 at the level of protein synthesis was determined, TGF-beta2 markedly downregulated Cdk4 synthesis and slightly upregulated p27 synthesis in actively cycling cells. On the other hand, TGF-beta2 did not exert the same effect on Cdk4 synthesis in contact inhibited cells as it did on actively cycling cells. Contact inhibited cells contained a high level of p27, and TGF-beta2 slightly upregulated p27 synthesis in these cells. When phosphorylated p27 was determined to be present, the nuclei of both actively cycling and contact inhibited cells contained phosphorylated p27 in the nuclei, regardless of the passage numbers. TGF-beta2 inhibited phosphorylation of p27 in actively cycling cells, but it had no effect on phosphorylation of p27 in contact inhibited cells.
Conclusions: These data suggest that Cdk4 and p27 expression is density dependent, and TGF-beta2 exerted its activity on actively cycling cells. In these cells, TGF-beta2 downregulated Cdk4 expression and prevented the phosphorylation of p27, which is a prerequisite for nuclear export of the inhibitor molecule for degradation. Thus, TGF-beta2 inhibits the G1/S transition while it maintains p27 in an active form in the nuclei during the exponential growth cell stage.