The role of p27 Tyrosine phosphorylation in the G1-S phase cell cycle transition.

Abstract

The oncogenes Cyclin D and cdk4 drive proliferation of cancer cells, and Cyclin D is overexpressed in 80% of breast cancers. The Cyclin D-cdk4 (hereafter D-K4) complex phosphorylates the cell cycle gate-keeper Rb, thereby inactivating it and thus, enabling the G1-S phase progression. p27kip1 (hereafter p27) is required to assemble and activate the D-K4 complex. Depending on the phosphorylation state of Tyrosine (Y) 88 and 89 residues, p27 can inhibit or activate the complex. When Y88 and Y89 residues are phosphorylated, p27 vacates the ATP binding pocket of cdk4, permitting catalytic activation of the D-K4 complex. We identified Brk (Breast tumor Related Kinase) as the kinase that activates p27 by phosphorylation on residues Y88 and Y89 in breast cancer cell lines. Overexpression of WT Brk in MCF7 cells led to increased Y88 phosphorylation, resulting in increased D-K4 activity and cell proliferation. Reducing Brk expression using siRNA reduced Y88 phosphorylation, suggesting that Brk is the physiological kinase in MCF7 cells; moreover other Y kinases were not able to compensate for its loss. We show that a naturally occurring catalytically inactive Alternatively-spliced variant (Alt Brk or ALT) was expressed in these cells, and it competitively inhibited WT Brk’s activity, blocked p27 pY88 phosphorylation, and caused G1 arrest. Given its role in tumorigenesis, inhibition of D-K4 is an attractive therapeutic strategy. Several cdk4 specific small molecule inhibitors (cdk4i) have been approved for clinical use recently. Treatment with cdk4i caused proliferation arrest for a period of time, but cells became resistant with several days. The mechanism of this adaptation and resistance was unknown. We showed that the other G1 cyclin dependent kinase, cdk2, directly compensated for the loss of D-K4 activity, reestablishing Rb phosphorylation, promoting cell proliferation, and was responsible for the observed drug resistance. This suggested that inhibition of cdk4 and cdk2 at the onset of therapeutic intervention would inhibit both the kinase that promotes proliferation and the kinase that promotes resistance. We observed that blocking pY88 inhibited cdk4, but also stabilized p27, which in turn inhibited cdk2. Using ALT to block Y88 phosphorylation, we were able to inhibit both cdk4 and cdk2 activity. Treatment with ALT or ALT+Palbociclib caused a cell cycle arrest that was durable for >30 days and induced cellular senescence. In vivo, using a a breast cancer xenograft model, we showed that ALT+Palbociclib treatment reduced p27 Y88 phosphorylation and caused tumor regression. Thus, our data supports the concept that targeting p27 pY88 might be an effective novel strategy to treat breast cancers. The clinical use of cdk4is has raised the requirement for a predictive marker of D-K4 activity. Several studies have shown lack of statistical correlation between Palbociclib response and p16, Cyclin D or Cyclin E expression. We developed a dual pY/p27 immunohistochemistry assay and demonstrated a positive correlation between p27 pY88, D-K4 and Palbociclib sensitivity in cell lines. Studies using archival, formalin-fixed, paraffin- embedded breast cancer patient material support our observations and suggest that pY88 status may have predictive power to stratify patients who would respond to this type of therapy.