P5-03-03: Identification of New Substrates for Breast Tumor Specific Low-Molecular-Weight Cyclin E Cyclin-Dependent-Kinase 2

Background: Cyclin E overexpression occurs in 25% of breast cancer tumors and is linked to poor prognosis. In tumor cells full length cyclin E (FL-E) is processed by an elastase-like protease into low-molecular weight isoforms (LMW-E) that are biochemically hyperactive. When overexpressed in breast cancer cells, LMW-E induced genomic instability and resistance to p21, p27, and antiestrogens. LMW-E are tumor specific and are associated with the aggressive triple negative breast cancer (TNBC). Transgenic mice overexpressing LMW-E had increased incidence of mammary tumors and distant metastasis when compared to FL-E. Furthermore, when the role of CDK2 in LMW-E mediated mammary tumorigenesis was interrogated by crossing LMW-E transgenic mice with CDK2 knock out mice, we found that the mice are protected against mammary tumor formation. Additionally, treatment of LMW-E transgenic mice with roscovitine (a CDK2 inhibitor) delays mammary tumor formation. These results indicate that kinase inhibition may have therapeutic activity in LMW-E overexpressing tumors as shown in this preclinical model. These results lead us to hypothesize that the biological and biochemical differences between FL-E and LMW-E may be due to the phosphorylation of a distinct set of substrates when complexed with CDK2. Our goal is to identify distinct LMW-E/CDK2 substrates on a proteome-wide scale that could serve as novel therapeutic targets for the treatment of the aggressive LMW-E expressing TNBC.

Material and Methods: We used two different approaches to identify LMW-E/CDK2 substrates. 1) We generated an analog sensitive CDK2 kinase, (F80A or F80G)-CDK2 to specifically radiolabel its substrates in cell extracts followed by their identification by mass spectroscopy. 2) We incubated ProtoArrayMicroarrays spotted with 3000 GST-tagged human protein on high density glass slides either with recombinant active FL-E/CDK2 or LMW-E/CDK2.

Results: In the first approach, we expressed and purified wild-type CDK2 in complex with FL-E or LMW-E and CDK2 (F80A) and CDK2 (F80G) from insect cells. Although all 3 CDK2 kinases can use ATP to phosphorylate GST-Rb protein, only the F80G mutant can use PE-ATP-γ-S. In the second approach, we incubated protein arrays either with recombinant active FL-E/CDK2 or LMW-E/CDK2 at a concentration of 50 nM. Our first screen identified a total of 122 potential substrates to both FL-E/CDK2 or LMW-E/CDK2 kinase complexes. We only identified 1 protein that is phosphorylated by FL-E/CDK2 significantly more than by LMW-E/CDK2 as compared to the 32 potential substrates specific to LMW-E/CDK2 suggesting that by losing the N-terminal portion, the LMW-E/CDK2 kinase complex is able to specifically interact and phosphorylate novel proteins.

Discussion: The identification of new physiological LMW-E/CDK2 substrates will lead to the development of novel targets for therapeutics and the identification of the biological function for the treatment of the aggressive LMW-E expressing TNBC.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P5-03-03.