Evolutionary and physiological importance of hub proteins

Abstract

It has been claimed that proteins with more interaction partners (hubs) are both physiologically more important (i.e., less dispensable) and, owing to an assumed high density of binding sites, slow evolving. Not all analyses, however, support these results, probably because of biased and less-than reliable global protein interaction data. Here we provide the first examination of these issues using a comprehensive literature-curated dataset of well-substantiated protein interactions in Saccharomyces cerevisiae. Whereas use of less reliable yeast two-hybrid data alone can reject the possibility that local connectivity correlates with measures of dispensability, in higher quality datasets a relatively robust correlation is observed. In contrast, local connectivity does not correlate with the rate of protein evolution even in reliable datasets. This perhaps surprising lack of correlation with evolutionary rate appears in part to arise from the fact that hub proteins do not have a higher density of residues associated with binding. However, hub proteins do have at least one other set of unusual features, namely rapid turnover and regulation, as manifest in high mRNA decay rates and a large number of phosphorylation sites. This, we suggest, is an adaptation to minimize unwanted activation of pathways that might be mediated by adventitious binding to hubs, were they to actively persist longer than required at any given time point. We conclude that hub proteins are more important for cellular growth rate and under tight regulation but are not slow evolving.

Figure 1. Relationship between Connectivity and Dispensability

Scatter plots of growth defect upon homozygous deletion of a gene and the natural log of connectivity of the gene in the interaction network. Growth rate data are from Steinmetz et al. [45].

Figure 2. Relationship between Connectivity and Evolutionary Rate

Scatter plots of natural log of connectivity and evolutionary rate of S. cerevisiae proteins. Shown here are the plots for singleton genes. For analysis of all genes and those with duplicates see Figures S1 and S2.