Comparative Analysis of <em>Mycobacterium tuberculosis pe</em> and <em>ppe</em> Genes Reveals High Sequence Variation and an Apparent Absence of Selective Constraints

<div><p><em>Mycobacterium tuberculosis</em> complex (MTBC) genomes contain 2 large gene families termed <em>pe</em> and <em>ppe</em>. The function of pe/ppe proteins remains enigmatic but studies suggest that they are secreted or cell surface associated and are involved in bacterial virulence. Previous studies have also shown that some <em>pe</em>/<em>ppe</em> genes are polymorphic, a finding that suggests involvement in antigenic variation. Using comparative sequence analysis of 18 publicly available MTBC whole genome sequences, we have performed alignments of 33 <em>pe</em> (excluding <em>pe_pgrs</em>) and 66 <em>ppe</em> genes in order to detect the frequency and nature of genetic variation. This work has been supplemented by whole gene sequencing of 14 <em>pe</em>/<em>ppe</em> (including 5 <em>pe_pgrs</em>) genes in a cohort of 40 diverse and well defined clinical isolates covering all the main lineages of the <em>M. tuberculosis</em> phylogenetic tree. We show that nsSNP's in <em>pe</em> (excluding <em>pgrs</em>) and <em>ppe</em> genes are 3.0 and 3.3 times higher than in non-<em>pe</em>/<em>ppe</em> genes respectively and that numerous other mutation types are also present at a high frequency. It has previously been shown that non-<em>pe</em>/<em>ppe M. tuberculosis</em> genes display a remarkably low level of purifying selection. Here, we also show that compared to these genes those of the <em>pe</em>/<em>ppe</em> families show a further reduction of selection pressure that suggests neutral evolution. This is inconsistent with the positive selection pressure of “classical” antigenic variation. Finally, by analyzing such a large number of genes we were able to detect large differences in mutation type and frequency between both individual genes and gene sub-families. The high variation rates and absence of selective constraints provides valuable insights into potential <em>pe/ppe</em> function. Since pe/ppe proteins are highly antigenic and have been studied as potential vaccine components these results should also prove informative for aspects of <em>M. tuberculosis</em> vaccine design.</p>