Our goal is to gain a better understanding of the control of cell cycle progression in normal cells; and then to utilize the information to better understand the unregulated growth of tumors. This laboratory has for many years been a leader in the development of novel cytometric techniques and their use in the study of cell proliferation. Recently, we have utilized this approach to demonstrate that cyclin D1 levels vary dramatically through the cell cycle, with low levels during S phase required for DNA synthesis, and stimulation of these levels during G2 phase necessary for continued cell cycle progression. At the same time, p27Kip1 levels determine the rate of passage through G1 phase, and the rate of cell cycle progression. Importantly, the expression patterns of these proteins seen in normal cells are dramatically altered in many tumor cells.
Many past studies of cell cycle regulation have relied upon cells synchronized by manipulating serum levels. While those studies have considerable utility, they are not suitable for studies of continuous cell cycle progression, or of cell cycle termination; two processes critical for understanding tumor cell growth. To address these critical issues we developed the means to quantitated protein levels in individual cellshc. The levels of these proteins were then manipulated following microinjection of expression plasmids or suppressors of activity. These studies, which avoid treatments to interfere with cell cycle progression, have provided a totally new understanding of cell cycle control, and promise to open new avenues for studies in tumors (6, 13).
With the use of time-lapse and microinjection of neutralizing anti-Ras antibodies, we found that it is during G2 phase that a cell makes the commitment to continue active cell cycle progression (1). With the use of quantitative image analysis it was demonstrated that cyclin D1 levels increase dramatically during G2 phase, and that it is this increase that commits a cell to continued cell cycle progression (2). Subsequent studies revealed that cyclin D1 levels sharply decline upon entry into S phase. This decline is required for DNA synthesis, and is due to protein degradation following phosphorylation of Thr 286 (8). The increase of cyclin D1 upon entry into G2 phase is dependent upon the stabilization of cyclin D1 mRNA by proliferative signaling molecules (9). This pattern has been observed in all monolayer cells analyzed, and in solid tissues in paraffin sections (unpublished data). The requirement that cyclin D1 be high in some cell cycle phases and low in others ensures that major alterations in its expression, which might otherwise lead to uncontrolled cell growth, are highly unlikely (6).
From the above considerations it is clear that cyclin D1 plays a central role in control of proliferation, and that its expression level through the cell cycle is vital to this control. It is not surprising; therefore, that in tumor cell lines the regulation of cyclin D1 through the cell cycle is often dramatically altered. In few if any of the tumor lines analyzed is cyclin D1 suppressed during S phase as dramatically as in normal cells. In many tumor cells there is little suppression of cyclin D1 during S phase. Moreover, in at least breast tumors the expression pattern of cyclin D1 appears to reflect the signing properties of the cell (unpublished data). Thus, the likelihood that cellular Ras proteins are required for growth f the tumor is directly related to the likelihood that cyclin D1 levels fall during S phase. We are anxious to determine if this pattern can be utilized as a means to assess the signaling properties of primary tumors in an effort to better predict the most effective treatment for that tumor.
The stability of cyclin D1 is determined to a large extent by phosphorylation of Thr 286. It has been reported that glycogen synthase kinase 3 is responsible. Our efforts to confirm this observation, however, have failed. We find no evidence for the involvement of this kinase in the control of cyclin D1 in NIH3T3 mouse or MRC5 human diploid fibroblasts (8). Rather, it appears that a process intimately associated with DNA synthesis itself is involved. Our search for the kinase involved is focused upon mechanisms directly tied to S phase, rather than any kinase whose activity ight be controlled by proliferative signaling like glycogen synthase kinase 3.
It has long been known that p27Kip1 (p27) and cyclin D1 share the ability to link proliferative signaling to the control of cell cycle progression. We found that proliferative signaling is able to suppress p27 levels in all cell cycle phases, and that the signaling pathway involved is different for each cell cycle phase (7). Importantly, MEK and phosphatidylinositol-3 kinase are required for suppression during G1 phase. This is critical since p27 levels, which are profoundly inhibitory to cell cycle progression, must be critically regulated during G1 phase (10). For example, when p27 levels were manipulated by microinjection of siRNA or an expression plasmid, we found that the rate of passage through G1 phase, and through the entire cell cycle, is determined by p27 levels (10). The restriction point:
In order to explain how p27 might function to regulate the rate of passage through G1 phase, we found a direct link between p27 levels and the activity of CDK2, the kinase which regulates passage through S phase. In quiescent cells stimulated to re-enter the cell cycle, p27 levels are gradually suppressed. When p27 levels reach a certain level, the cell becomes committed to enter S phase (12). This is apparently due to the fact that p27 is able to inhibit activation of CDK2, since if p27 levels are artificially increased during this time period, CDK2 fails to become activated and the cells fail to enter S phase. An artificial suppression of p27, on the other hand, results in CDK2 activation and S phase entry in cells that had not otherwise passed the restriction point (12).
The above data have been combined into a working model to explain the control of cell cycle progression in normal cells. From our data it is clear that alterations in cyclin D1 levels take place primarily during G2 phase (4), where it determines the proliferative fate of the cell. P27 levels, on the other hand, are suppressive to cell growth. As p27 begins to accumulate during G1 phase it first binds cyclin D1, which masks its inhibitory activity. Continued elevation of p27 above cyclin D1 levels is either blocked by degradation, or the added p27 will inhibit CDK2 activity and inhibit entry into S phase. Therefore, the ability of a cell to proliferate is determined during G2 phase by elevation of cyclin D1, while the rate of cell cycle progression is determined during G1 phase by suppression of the level of p27. In both cases, these levels are directly dependent upon the extracellular proliferative signaling environment of the cell. While certain aspects of the model are familiar, it is in general totally unique to this laboratory.
