Analysis of complete genomes of Mycobacterium tuberculosis sublineage 2.1 (Proto-Beijing) revealed the presence of three pe_pgrs3-pe_pgrs4-like genes

Abstract

Mycobacterium tuberculosis Complex (MTBC), the etiological agent of tuberculosis (TB), demonstrates considerable genotypic diversity with distinct geographic distributions and variable virulence profiles. The pe-ppe gene family is especially noteworthy for its extensive variability and roles in host immune response modulation and virulence enhancement. We sequenced an Mtb genotype L2.1 isolate from Chiangrai, Northern Thailand, using second and third-generation sequencing technologies. Comparative genomic analysis with two additional L2.1 isolates and two L2.2.AA3 (Asia Ancestral 3 Beijing) isolates revealed significant pe-ppe gene variations. Notably, all L2.1 isolates harbored three copies of pe_pgrs3-pe_pgrs4-like genes (pe_pgrs3*, pe_pgrs4*, and pe_pgrs4), different from L2.2.AA3 and H37Rv strains. Additionally, ppe53 was duplicated in all but H37Rv, and ppe50 was deleted in L2.1 isolates, contrasting with an extended ppe50 in an L2.2 isolate (Mtb 18b), which contains an additional SVP motif. Complete deletion of ppe66 and loss of wag22 were observed in L2.1 isolates. These findings highlight the high structural variability of the pe-ppe gene family, emphasizing its complex roles in Mtb-host immune interactions. This genetic complexity offers potentially critical insights into mycobacterial pathogenesis, with significant implications for vaccine development and diagnostics.

Keywords: Mycobacterium tuberculosis; Pe-ppe; pe_pgrs3; pe_pgrs4; Lineage 2.1; Proto-Beijing genotype.

Fig. 1

Alignment of genes from ppe24 to pe_pgrs31 among the six isolates. The large, deleted region included 510 bp of cut1, wag22 and three hypothetical genes (Rv1760, Rv1761c and Rv1762c) in all L2.1 and 2.2 isolates. The pe_pgrs31 in Mtb N0031 was split into two ORFs possibly due to a homopolymeric error of long read sequencing data. However, short-read sequencing data is not available for verifying the sequence. Two coding sequences at the pe_pgrs31 locus is, therefore, presented as predicted.

Fig. 2

Alignment of genes from plcB to ppe39 among the six isolates in this study. The ppe38 was deleted in all L2.1 isolates, replaced with IS6110. The homologous region of CDC1551, which contains additional genes, esxN, esxJ and ppe71, is also included.

Fig. 3

Gene alignment of the region between ppe46 and pe29. The deleted region of two L2.1 isolates ranged from the 5’ end of ppe46, including pe27A, esxR, exsS, ppe47 and 3’ end of ppe48, resulting in a chimeric ppe46/48 gene. The other L2.1 isolate (Mtb Sea14117p6c4) shared a similar gene order to the two L2.2 (Mtb18b and 2279) isolates.

Fig. 4

Gene alignment from Rv3134 to ppe51. (A) The alignment among the six isolates The PPE50 was deleted in all L2.1 and a L2.2 isolate (Mtb 2279). In contrast, Mtb 18b had an insertion, containing the SVP motif, that resulted in ppe50e, a longer version of ppe50. (B) The alignment between H37Rv (above) and Mtb 18b (below).

Fig. 5

Gene alignment from Rv3737 to ppe67 among the six isolates. The ppe66 deletion was specific to L2.1. All L2.1 isolates had a duplication of a 128-bp-long segment at the position 1081–1208 of Rv3737, which resulted in frameshift and a premature stop codon. The remaining part of the coding sequence of Rv3737 was annotated as a hypothetical gene (black).

Fig. 6

Gene alignment of the pe_pgrs3 to ppe3 region among the six isolates. All L2.1 isolates harbor three copies of pe_pgrs3-pe_pgrs4-like genes, from left to right namely pe_pgrs3*, pe_pgrs4* and pe_pgrs4. The SNP distances between each pair of the three genes was approximately equal and much higher than the distances between the same genes among different isolates. None was closely similar to H37Rv pe_pgrs3.

Fig. 7

The phylogenetic tree and gene mapping of pe_pgrs3- pe_pgrs4-like genes. (A) The SNP-based maximum likelihood phylogenetic tree of pe_pgrs3, pe_pgrs3*, pe_pgrs4* and pe_pgrs4 of five L2 isolates, H37Rv, M. canetti CIPT, M. bovis AF2122/97 and M. bovis BCG Tokyo and Pasteur, showing that H37Rv pe_pgrs3 is only distantly similar to the three other genes. The pe_pgrs4 gene that is most dissimilar to the others belongs to M. canetti CIPT. (B) Gene alignment of the pe_pgrs3*, pe_pgrs4* and pe_pgrs4 of representative isolates of L2.1, L2.2.AA3, M. bovis, M. bovis BCG and M. canetti. H37Rv pe_pgrs3 and pe_pgrs4 are also shown for comparison.

Fig. 8

Gene alignment from ppe53 to moxR3 showing ppe53A (left) and ppe53B (right) in L2 isolates. ppe53B of Mtb CR170941 was truncated by insertion of IS6110. The split of ppe53B in Mtb 2279 into two coding sequences may be due to a homopolymeric run error in long-read sequencing. Its NGS data were not available for verification.