Identification of PKC Phosphorylation Sites on AC7 using a Directed Bioinformatics Approach.

Eric J. Nelson1, John VanHoven2, Vlad Verkhusha, Tonny deBeer, and Boris Tabakoff., Univ. of Colorado;, Univ. of Colorado

Numerous methods of impairing the activity of protein kinase C-delta (PKC-delta) have been previously shown by us to attenuate the potentiation of adenylyl cyclase (AC) type-7 activity by ethanol. We have also demonstrated that recombinant full-length AC7 was selectively phosphorylated in vitro by PKC-delta. Thus we set out to use a directed bioinformatics approach utilizing comparative sequence analysis, molecular modeling, and machine learning techniques to assist in the discovery of PKC-sensitive phosphorylation site(s) within AC7. The combination of comparative sequence analysis of AC isotypes, species differences within AC7, and functional activity data significantly narrowed the number of potential PKC phosphorylation sites within AC7. Using both NetPhos 2.0, a neural network-based computational method for predicting phosphorylation propensity based on adjacent sequences, and a Hidden Markov Model developed and trained by us to identify potential sequences selective for PKC-delta phosphorylation, we identified a number of sequences that showed particular promise. These sites were then assessed, using secondary sequence algorithms and molecular 3-D modeling, as to whether they were solvent exposed or buried and to determine the proximity of these sites to the known functional domains of AC. The predicted phosphorylation sites from the bioinformatics approach were then tested directly using PCR-based site-directed mutagenesis for structure-function analysis. Here we present a bioinformatics approach to identifying suspected phosphorylation sites that bypasses the traditional method of phosphopeptide mapping a radiologically labeled target protein. The experimental data was consistent with the computationally generated hypothesis that identified novel PKC phosphorylation sites within AC7.