Why are the truncated cyclin Es more effective CDK2 activators than the full-length isoforms?

Abstract

Cell cycle regulating enzymes, CDKs, become activated upon association with their regulatory proteins, cyclins. The G1 cyclin, cyclin E, is overexpressed and present in low molecular weight (LMW) isoforms in breast cancer cells and tumor tissues. In vivo and in vitro studies have shown that these LMW isoforms of cyclin E hyperactivate CDK2 and accelerate the G1-S phase of cell division. The molecular basis of CDK2 hyperactivation due to LMW cyclin E isoforms in cancer cells is, however, unknown. Here, we employ a computational approach, combining homology modeling, bioinformatics analyses, molecular dynamics (MD) simulations, and principal component analyses to unravel the key structural features of CDK2-bound full-length and LMW isoforms of cyclin E1 and correlate those features to their differential activity. Results suggest that the missing N- and C-terminal regions of the cyclin E LMW isoforms constitute the Nuclear Localization Sequence (NLS) and PEST domains and are intrinsically disordered. These regions, when present in the full-length cyclin E/CDK2 complex, weaken the cyclin-CDK interface packing due to the loss of a large number of key interface interactions. Such weakening is manifested in the decreased contact area and increased solvent accessibility at the interface and also by the absence of concerted motions between the two partner proteins in the full-length complex. More effective packing and interactions between CDK2 and LMW cyclin E isoforms, however, produce more efficient protein-protein complexes that accelerate the cell division processes in cancer cells, where these cyclin E isoforms are overexpressed.