THE USE OF SIDE-CHAIN PACKING METHODS IN MODELING BACTERIOPHAGE REPRESSOR AND CRO PROTEINS

In recent years, it has been repeatedly demonstrated that the coordina tes of the main-chain atoms alone are sufficient to determine the side -chain conformations of buried residues of compact proteins. Given a p erfect backbone, the side-chain packing method can predict the side-ch ain conformations to an accuracy as high as 1.2 Angstrom RMS deviation (RMSD) with greater than 80% of the (chi) angles correct. However, si milarly rigorous studies have not been conducted to determine how well these apply, if at all, to the more important problem of homology mod eling per se. Specifically, if the available backbone is imperfect, as expected for practical application of homology modeling, can packing constraints alone achieve sufficiently accurate predictions to be usef ul? Here, by systematically applying such methods to the pairwise mode ling of two repressor and two cro proteins from the closely related ba cteriophages 434 and P22, we find that when the backbone RMSD is 0.8 A ngstrom, the prediction on buried side chain is accurate with an RMS e rror of 1.8 Angstrom and approximately 70% of the (chi) angles correct ly predicted. When the backbone RMSD is larger, in the range of 1.6-1. 8 Angstrom, the prediction quality is still significantly better than random, with RMS error at 2.2 Angstrom on the buried side chains and 6 0% accuracy on (chi) angles. Together these results suggest the follow ing rules-of-thumb for homology modeling of buried side chains. When t he sequence identity between the modeled sequence and the template seq uence is > 50% (or, equivalently, the expected backbone RMSD is < 1 An gstrom), side-chain packing methods work well. When sequence identity is between 30-50%, reflecting a backbone RMS error of 1-2 Angstrom, it is still valid to use side-chain packing methods to predict the burie d residues, albeit with care. When sequence identity is below 30% (or backbone RMS error greater than 2 Angstrom), the backbone constraint a lone is unlikely to produce useful models. Other methods, such as thos e involving the use of database fragments to reconstruct a template ba ckbone, may be necessary as a complementary guide for modeling.