Identification of New Helicobacter pylori Subpopulations in Native Americans and Mestizos From Peru

Abstract

Region-specific Helicobacter pylori subpopulations have been identified. It is proposed that the hspAmerind subpopulation is being displaced from the Americans by an hpEurope population following the conquest. Our study aimed to describe the genomes and methylomes of H. pylori isolates from distinct Peruvian communities: 23 strains collected from three groups of Native Americans (Asháninkas [ASHA, n = 9], Shimaas [SHIM, n = 5] from Amazonas, and Punos from the Andean highlands [PUNO, n = 9]) and 9 modern mestizos from Lima (LIM). Closed genomes and DNA modification calls were obtained using SMRT/PacBio sequencing. We performed evolutionary analyses and evaluated genomic/epigenomic differences among strain groups. We also evaluated human genome-wide data from 74 individuals from the selected Native communities (including the 23 H. pylori strains donors) to compare host and bacterial backgrounds. There were varying degrees of hspAmerind ancestry in all strains, ranging from 7% in LIM to 99% in SHIM. We identified three H. pylori subpopulations corresponding to each of the Native groups and a novel hspEuropePeru which evolved in the modern mestizos. The divergence of the indigenous H. pylori strains recapitulated the genetic structure of Native Americans. Phylogenetic profiling showed that Orthogroups in the indigenous strains seem to have evolved differentially toward epigenomic regulation and chromosome maintenance, whereas OGs in the modern mestizo (LIM) seem to have evolved toward virulence and adherence. The prevalence of cagA ⁺/vacA s1i1m1 genotype was similar across populations (p = 0.32): 89% in ASHA, 67% in PUNO, 56% in LIM and 40% in SHIM. Both cagA and vacA sequences showed that LIM strains were genetically differentiated (p < 0.001) as compared to indigenous strains. We identified 642 R-M systems with 39% of the associated genes located in the core genome. We found 692 methylation motifs, including 254 population-specific sequences not previously described. In Peru, hspAmerind is not extinct, with traces found even in a heavily admixed mestizo population. Notably, our study identified three new hspAmerind subpopulations, one per Native group; and a new subpopulation among mestizos that we named hspEuropePeru. This subpopulation seems to have more virulence-related elements than hspAmerind. Purifying selection driven by variable host immune response may have shaped the evolution of Peruvian subpopulations, potentially impacting disease outcomes.

Keywords: Amerindians; Peru; ancestry; hspAmerind; indigenous; mestizo.

FIGURE 1

Hierarchical clustering analysis of ANIb values. Each line represents the similarity score of each H. pylori genome (32 study and 95 references). Left, ancestral H. pylori populations. Right, H. pylori study populations. Heavily admixed genomes (ASHA-003, PUNO-003, PUNO-009, PUNO-010, and LIM-007) are shown in red font.

FIGURE 2

Phylogenomic relationships and population structures of H. pylori and human populations. (A) Global phylogenomic tree for 127 H. pylori strains. The tree was constructed from a total of 930,403 SNPs with a K value of 29 using KSNPV3.0. The shading blue (ASHA) and red (SHIM) are the strains obtained from the Amazon, the light green (PUNO) are the strains isolated from Puno in the Andes, and the light yellow (LIM) represent the strains isolated from Lima. The Peruvian map shows the regions where the samples were collected. The asterisks represent the heavily admixed strains (ASHA-003, PUNO-003, PUNO-009, PUNO-010, and LIM-007). (B) Ancestry profiles inferred using Chromopainter v2/fineSTRUCTURE for 127 H. pylori strains (32 study and 95 references; Supplementary Table S4). (C) ADMIXTURE results for 10 human populations from 1000 Genomes (n = 250) and Borda et al. (2020) (n = 74; Supplementary Table S1). We plot admixture results for the K value with the lowest cross validation error (K = 6; Supplementary Figure S2).

FIGURE 3

Genomic consensus of the study populations and hpEurope references. These circular views were obtained using the method developed by Tada et al. (2017), which creates a consensus genome that is used as a template for alignment. Each ring represents one complete genome and each block in the ring represents a genomic region. Different colors represent the genes according to their genomic position in the consensus. The shifts in the color represent the rearrangements. The outermost ring is the distant genome from the consensus. The names alongside each circle indicate the genomes going from outward to inward direction. The indicated areas (black circles) in the ASHA (ring 2) and PUNO (rings 1, 5, and 6) genomes show the regions in which these genomes are similar to the LIM (rings 2 and 6). The green circles indicate the similarities between the LIM genomes and the hpEurope references. The genomes ASHA-003, PUNO-003, PUNO-009, and PUNO-010 had an inversion from 10 o’clock to 11 o’clock which was not observed in SHIM genomes.

FIGURE 4

(A) Restriction-modification systems content in core and accessory H. pylori genomes from indigenous (9 ASHA, 5 SHIM, and 9 PUNO) and mestizos (9 LIM). (B) Venn diagram of methylation motifs with at least 80% methylation fractions.