Normally-activated Cdk5 functions in essential neuronal regulatory processes. The occurrence of significant amounts of p25 in neurons appears to result from the abnormally high calpain protease activity in stressed neurons. P25-activated kinase within neurons produces hyperphosphorylated tau protein, which causes tau to dissociate from microtubules and accumulate as a major component of the neurofibrillary tangles that occur in Alzheimer disease. When CIP is expressed in cultured neurons by transfection, it selectively inhibits this pathological activity of cdk5/p25 but not the normal cdk5/p35 activity. Thus when neurons in cullture are stressed to produce p25 and subsequently neurofibrillary tangles, similarly-treated neurons do not produce neurofibrillary tangles if they were first transfected to produce CIP (Zheng et al, 2005). We are engaged in studies to further define the mechanism of this selective inhibition. Cdk5 is a member of the cyclin-dependent kinase family. Although cdk5 activator proteins are not cyclins, the activator interfaces with the kinases in this family are largely similar and they induce similar conformational transitions. Thus we are also exploring computational means to obtain information about the interactions of CIP and related inhibitors with the cdk5 kinase. From examination of the crystal structure of the cdk5/p25 complex (Tarricone, et al), when complexed with cdk5, it is evident that,the N- and C- terminals of p25 interact with each other at the molecular surface opposite to that of the p25 interface with cdk5. The truncations that produce CIP delete this interaction but do not eliminate any part of the p25 binding interface with cdk5. These observations suggest that a flexibility hypothesis can explain how truncations can convert p25 activation to selective inhibition. We are employing molecular dynamics simulations of cdk5/p25 and cdk5/CIP complexes to examine the effects of these truncations on the conformations of cdk5/CIP and of the active kinase. Work in progress: Truncations that form CIP from p25 do not remove residues directly involved in the cdk5/p25 interface. Preliminary investigations, some of which were described in the last report, indicate that small changes in this interface can alter the conformation of the kinase substrate-binding groove. This, in turn, may alter the catalytic activity at the kinase phosphorylation site. We are currently acquiring and analysing molecular dynamics data designed to test this and related issues. Large-scale molecular dynamic simulations of the cdk5-p25/CIP complexes, and the corresponding monomers, have been conducted to gain insight into the structure, dynamics, and thermodynamics of the system. Preliminary analyses show that truncation of the N- and C-terminus of p25 leads to structural relaxation and gentle rearrangement of some of the helices in CIP. This relaxation, in turn, elicits larger structural changes in two key loops located at the cdk5-CIP interface, which have been implicated in both cdk5-p25 complex formation, and binding and stabilization of the substrate. In particular, one of these loops contains E240, which is known to interact electrostatically with a conserved charged residue in the substrate (position +3 from the phosphorylation site). The structural changes of this loop repositions E240 relative to the substrate-binding pocket, which may prevent its direct interaction with the substrate. These changes may explain, at least partially, the inhibitory properties of CIP. In addition, crystallographic data show that the active conformation of cdk5 is stabilized by specific interfacial interactions with p25. These stabilizing centers are perturbed in the cdk5-CIP complex, which may destabilize the active conformation of the kinase and further contribute to the inhibitory action of CIP. A thermodynamic study is underway to quantify these processes. Truncation of the N- and C-termini of p25 may thus affect both the active conformation of the kinase and the binding of the substrate, both deleterious to enzyme activity.
