Proteins mutability

N ucleic acids occupy a unique position in the biochemical world. Not only are they involved in many important reactions, but they carry genetic information, which must be faithfully duplicated so that it can be passed from one cell generation to the next and from one organism to the next. Nonetheless this information must be mutable to produce the variability on which the evolutionary selection process feeds. Finally the DNA must be selectively transcribed so that each cell can synthesize the proteins it needs. 

Now we demonstrate how to examine the mutability of specified residues in a protein. Two positions in an immunoglobulin structure are compared with respect to the number of stabilizing mutations among all possible substitutions. A singlesite mutability profile is generated for two regions of the structure, pointing to sites that may be more important for structural stability than others. 

To check single-site mutability for more than one residue, we specify a list of positions. Given a computing time of roughly 20 s per z-score calculation for a protein of this length on a 1.4 GHz cpu, the analysis takes approximately 1.75 h (15 residues x 20 amino acids x 20 secs.). The output is collected in four files Fd xm.pl. sip holds the z-scores for wild type and mutants and F d xmp 1. mut c omb, F d xm.p 1. mu t p ai r, and F d xmp 1. mut s u r f hold the mutability profiles for the three energy terms, respectively, exit... 

Scan the relative mutability of the protein with the following amino acid sequence ... 

Also, as is seen experimentally, many protein sequences adopt the same native-state conformation [56]. Once a sequence has selected its native-state structure, it is able to tolerate a significant degree of mutability except at certain key locations [54]. Furthermore, multiple protein functionalities can arise within the context of a single fold [57]. 

All pairwise comparisons of structures in each alignment produced by COMPARER were considered in the analysis, and all substitutions implied by pairwise comparisons were stored in tables as a function of the features identified in the three-dimensional structures. In order to avoid very sparse tables, we considered the structural features of only one of the two proteins compared. Secondly, in order to understand the general role of certain structural features in constraining the mutability, we accumulated the values across various features. In each case we display the data as 20 x 20 matrices where one dimension refers to the amino acid type restricted to a structural environment and the other is simply residue type. 

Amino acid residues Frequencies in proteins (%) Variability/ Mutability... 

Antigenic determinants were found across the top surface of the sialidase monomer and encircling the active site [62], Of importance for inhibitor development, the active site was seen to contain a large number of amino acids that were conserved in all influenza A and B virus sialidase sequences known to that time [34,62]. This suggested that the active site residues were under pressure to remain constant [69], and that this was an invariant area of the sialidase that could be targeted for inhibitor development against the otherwise highly mutable [4] protein. 

On this basis, one concludes that the rate limiting step for unfolding occurs in a considerably unfolded molecule, as measured by the number of buried histidine side-chains which are exposed in iIk transition from the native molecule to the (mutable) intermediate state. The rate limiting step can, of course, involve a further conformation change of the protein, but is more likely to be the dissociation of heme and protein, since the rates are so much lower than the rates for the conformation change of apomyoglobin. 

Third, the occurrence of genetic polymorphisms with respect to some of the proteins which otherwise meet all of the criteria will detract from the usefulness of these particular indicators. In effect, this means that certain classes of forward or backward mutations cannot be detected. It would appear that suitable mathematical allowances can be made for this fact, but obviously for monitoring purposes it would be wise to avoid the polymorphic proteins insofar as possible. Whether this decision would introduce an element of bias with respect to the mutability of the proteins concerned is a moot question. 

---