Fine-scale population structure of Japanese Helicobacter pylori provides new anthropological and epidemiological insights

Abstract

Helicobacter pylori is considered to contribute to gastric cancer and is also used as a marker to trace human migration due to its co-evolution with humans. To understand the recently proposed tripartite model suggesting three ancestral origins for the Japanese population and address the enigma of the high incidence of gastric cancer in Northeast Hondo (Hondo is mainland Japan), we conducted a fine-scale population structure analysis using a large Japanese H. pylori dataset, including 438 strains from 9 regions based on whole-genome sequences. As a result of fineSTRUCTURE analysis, it was found that H. pylori in Northeast Hondo is genetically distinct from hspEAsia subgroup 7 (sg7), which is widely distributed elsewhere in Hondo. We named this new subgroup hspEAsia-sg8 (Northeast Hondo). Ancestry analysis using ChromoPainter revealed that, while a large proportion of the genomes of hspEAsia-sg8 strains were painted by donors from their own population, the ancestry components of hspEAsia-sg7 showed a high proportion of Chinese and Korean components, suggesting that they were formed through admixture with continental hspEAsia subgroups. These results align with human genome studies, which indicate an original ancestry component in Northeast Hondo and a higher proportion of East Asian components in West Hondo, supporting the tripartite model. This also suggests novel potential for biogeographic ancestry inference in forensic science, as the H. pylori genome can distinguish Hondo populations. Furthermore, fixation index analysis comparing the genome of hspEAsia-sg8 with other Japanese hspEAsia subgroups revealed a high number of nonsynonymous mutations in hp0378 (ccsBA) and hp0377 (dsbC/ccmG). Because these genes are involved in cytochrome c maturation and disulphide bond formation, the detected mutations may affect bacterial survival, growth or pathogenicity. This study supports the tripartite model for the formation of modern Japanese people and suggests that the strain of H. pylori prevalent in the Northeast Hondo region may contribute to the high incidence of gastric cancer there.

Keywords: Helicobacter pylori; ancestry analysis; gastric cancer; human migration; population structure.

Fig. 1.. The geographical locations of the Japanese H. pylori strains used in this study (n=438). Japan is shaded in grey, consisting of Hondo and Okinawa. The numbers in parentheses represent the numbers of strains.

Fig. 2.. Co-ancestry matrix of all strains used in this study. The heatmap indicates the number of DNA chunks imported from a donor genome (column) to a recipient genome (row). The innermost bar plot represents the H. pylori populations, subpopulations or subgroups defined in this study. The isolation country and region of each strain are shown in the middle and top bar plot, respectively.

Fig. 3.. The distribution of H. pylori populations within Japan. The colours in the pie chart represent each population.

Fig. 4.. PCA of 560 H. pylori strains belonging to hspEAsia. Each strain is represented by the colour of the subgroups defined by fineSTRUCTURE analysis. The geographical relationships of each subgroup are represented on the map.

Fig. 5.. Ancestry of each strain estimated by chromosome painting. Each column shows the proportion of each ancestry, with colors indicating each population. The bar plot above the column indicates the populations defined by fineSTRUCTURE analysis. (a) Full chromosome painting analysis. All populations were used as donors. For hspEAsia, each subgroup was used as donor. (b) Ancestral chromosome painting analysis. Only the true historically ancestral populations, hpAfrica2, hpAfrica1, hpNEAfrica, hpAsia2, hpNorthAsia, hspUral and hspEAsia, were used as donors. The donor strains of hspEAsia were equally selected from all hspEAsia subgroups.

Fig. 6.. Fst analysis to identify the differentiated genomic regions in hspEAsia-sg8 compared with other Japanese hspEAsia subgroups (sg6 and sg7). The X-axis indicates the nt positions in the F57 reference strain, while the Y-axis shows the Fst values of each SNP. The positions of the top ten genes with the highest Fst values are highlighted in red, along with their gene names.

Fig. 7.. NS mutations fixed in hspEAsia-sg8. Fst values of each position were visualized with gradient colour (grey to red) on the 3D protein structure. The aa number is based on the F57 strain. (a) CcsBA (HP0378). The positions of TM-His1, TM-His2, P-His1 and P-His2 are indicated by spheres with carbon (yellow), nitrogen (blue) and oxygen (olive). (b) DsbC/CcmG (HP0377). The position of the active site (CXXC motif) is indicated by spheres the same as above and with sulphur (green). I55T mutation was also detected, but the protein structure of this region is not available. The position S90H (Fst 0.42) is indicated in parentheses, as stated in the main text. Compared with HP0377 of 26695, HPF57_1040 of F57 has an insertion of two aa between positions 133 and 134. Therefore, for example, aa 191 in F57 corresponds to aa 189 in strain 26695.