Ancient diversity in host-parasite interaction genes in a model parasitic nematode

Abstract

Host-parasite interactions exert strong selection pressures on the genomes of both host and parasite. These interactions can lead to negative frequency-dependent selection, a form of balancing selection that is hypothesised to explain the high levels of polymorphism seen in many host immune and parasite antigen loci. Here, we sequence the genomes of several individuals of Heligmosomoides bakeri, a model parasite of house mice, and Heligmosomoides polygyrus, a closely related parasite of wood mice. Although H. bakeri is commonly referred to as H. polygyrus in the literature, their genomes show levels of divergence that are consistent with at least a million years of independent evolution. The genomes of both species contain hyper-divergent haplotypes that are enriched for proteins that interact with the host immune response. Many of these haplotypes originated prior to the divergence between H. bakeri and H. polygyrus, suggesting that they have been maintained by long-term balancing selection. Together, our results suggest that the selection pressures exerted by the host immune response have played a key role in shaping patterns of genetic diversity in the genomes of parasitic nematodes.

Fig. 1. Chromosome-level reference genomes for Heligmosomoides bakeri and Heligmosomoides polygyrus.

a The position of H. bakeri and H. polygyrus within Strongylida. Phylogeny was inferred using the amino acid sequences of 511 BUSCO genes that were present and single copy in at least 18 of the 20 species. H. bakeri and H. polygyrus are highlighted in bold. Suborders are indicated. Scale in substitutions per site is shown. Inset: a H. polygyrus female. b Cumulative length plots showing contiguity of all single nematode assemblies and both existing H. bakeri reference genomes (PRJEB1203 and PRJEB15396). The dotted line indicates that the curve for the H. bakeri PRJEB1203 assembly extends beyond 10,000 contigs (the total number of contigs in the assembly is 20,569). c Hi-C contact map for the nxHelBake1.1 reference genome. Chromosome names are indicated. The Hi-C contact map for the ngHelPoly1.1 reference genome is shown in Supplementary Fig. 3A. d Distribution of BUSCO genes in the six H. bakeri chromosomes in the nxHelBake1.1 reference genomes coloured by their allocation to the seven ancestral Nigon elements defined by Gonzalez de la Rosa et al. . Source data this figure can be found in the Source Data file and the GitHub repository.

Fig. 2. The H. bakeri and H. polygyrus genomes are highly divergent.

a Gene tree of the ribosomal internal transcribed spacer 2 (ITS2) in laboratory isolates of H. bakeri, wild isolates of H. polygyrus, and outgroup taxa. The sequences are from Cable et al. or the nuclear genomes of individuals sequenced as part of this work (highlighted in bold). The location of the laboratory or wild mouse colony from which each individual was collected is indicated. Bootstrap support values are shown for the branches subtending the H. bakeri and H. polygyrus clades. Branch lengths represent the number of substitutions per site; scale is shown. b The relative position of 2667 BUSCO genes in the H. bakeri and H. polygyrus genomes. Points are coloured by Nigon element. c Whole-genome alignment between H. bakeri and H. polygyrus from minimap2. The position of each alignment in the H. bakeri reference and the nucleotide identity is shown. Only alignments that are ≥5 kb are shown. d Synonymous site divergence in 2670 BUSCO genes in the H. bakeri and H. polygyrus genomes. Solid represent LOESS smoothing functions fitted to the data; standard error is shown using grey shading. Dotted line indicates mean synonymous site divergence (6.6%). Source data this figure can be found in the Source Data file and the GitHub repository.

Fig. 3. Hyper-divergent haplotypes are widespread in the H. bakeri genome.

Distribution of heterozygosity in the genomes of (a) H. polygyrus ngHelPoly1 and (b) H. bakeri nxHelBake1. Points represent the density of biallelic SNPs in 10 kb windows. nxHelBake1 was male and therefore had a hemizygous X chromosome; the SNP density peaks on the X chromosome are therefore erroneous and are derived from mismapped PacBio HiFi reads. ngHelPoly1 was female. SNPs called in repeat-containing regions were filtered out and SNP density was calculated as the number of non-repetitive SNPs per non-repetitive base. Blue lines represent LOESS smoothing functions fitted to the data. c H. bakeri nxHelBake1 PacBio HiFi read alignments to the nxHelBake1.1 reference genome in a 143 kb region on chromosome V (87.7 − 87.9 Mb) that contains a 56 kb hyper-divergent haplotype. The top panel shows the coverage and the bottom panel shows aligned PacBio HiFi reads. The coloured vertical lines indicate mismatched bases at that position. An alignment between the two assembled haplotypes in this region is shown in Supplementary Fig. 10. Source data this figure can be found in the GitHub repository.

Fig. 4. Ancient genetic diversity in two members of the Ancylostoma-secreted protein family.

A 287 kb region on chromosome I of H. bakeri (I:8.92–9.20 Mb) containing two ASP loci that show evidence of haplotype sharing. a Nucleotide alignments between the alternate H. bakeri haplotype (represented by nxHelBake2 primary), each H. polygyrus haplotype (ngHelPoly1 primary, ngHelPoly1 alternate, ngHelPoly2 primary), and the nxHelBake1 reference haplotype. Alignments are coloured by their nucleotide identity. Repetitive alignments are not shown. The position of the two ASP loci (nxHelBake1.g10196 and nxHelBake1.g10198) are indicated by dotted lines. b Gene contents of each haplotype. The alternate H. bakeri haplotype (represented by nxHelBake2 primary) has two homologues of nxHelBake1.g10195 and nxHelBake1.g10196, consistent with the divergent inverted duplication in (a). Green (H. bakeri) and orange (H. polygyrus) boxes represent genes that have a 1:1 relationship across all haplotypes; grey boxes represent genes that have non 1:1 relationship (nxHelBake1.g10195 and nxHelBake1.g10196). Links between genes represent homologous relationships. Gene trees of (c) nxHelBake1.g10196.t1 and its homologues and (d) nxHelBake1.g10198.t1 and its homologues. Gene trees were inferred using IQ-TREE under the LG + Γ substitution model. Scale is shown in substitutions per site. Outgroup not shown. Source data this figure can be found in the GitHub repository.