A Targeted Capture Linkage Map Anchors the Genome of the Schistosomiasis Vector Snail, Biomphalaria glabrata

Abstract

The aquatic planorbid snail Biomphalaria glabrata is one of the most intensively-studied mollusks due to its role in the transmission of schistosomiasis. Its 916 Mb genome has recently been sequenced and annotated, but it remains poorly assembled. Here, we used targeted capture markers to map over 10,000 B. glabrata scaffolds in a linkage cross of 94 F1 offspring, generating 24 linkage groups (LGs). We added additional scaffolds to these LGs based on linkage disequilibrium (LD) analysis of targeted capture and whole-genome sequences of 96 unrelated snails. Our final linkage map consists of 18,613 scaffolds comprising 515 Mb, representing 56% of the genome and 75% of genic and nonrepetitive regions. There are 18 large (> 10 Mb) LGs, likely representing the expected 18 haploid chromosomes, and > 50% of the genome has been assigned to LGs of at least 17 Mb. Comparisons with other gastropod genomes reveal patterns of synteny and chromosomal rearrangements. Linkage relationships of key immune-relevant genes may help clarify snail-schistosome interactions. By focusing on linkage among genic and nonrepetitive regions, we have generated a useful resource for associating snail phenotypes with causal genes, even in the absence of a complete genome assembly. A similar approach could potentially improve numerous poorly-assembled genomes in other taxa. This map will facilitate future work on this host of a serious human parasite.

Keywords: Biomphalaria; Schistosoma; linkage disequilibrium; linkage map; mybaits.

Figure 1

Methodology summary. (A) We sequenced the complete genomes of two unrelated parents, which (B) guided our designing of targeted capture probes, which we used to genotype (C) the F1 offspring and (D) the parents. (E) Segregating targeted capture markers were mapped into LGs. We also performed both (F) targeted capture and (G) whole-genome sequencing on a set of unrelated outbred snails to (H) infer LD. (I) We integrated these data with each other to generate the final LGs. LD, linkage disequilibrium; LGs, linkage groups.

Figure 2

Linkage groups generated by linkage cross. For all 24 linkage groups, the map positions for parent A (red) and/or parent B (blue) are shown as black squares (centimorgan distance along y-axis; scale bar in upper left). At each map position, the x-axis width of the horizontal bar indicates the number of scaffolds that map to that location (scale bar in upper left; scaffolds subsequently added via linkage disequilibrium not shown). Gray lines indicate where the same scaffold maps in both parents.

Figure 3

Physical sizes of LGs. The cumulative size of all scaffolds on all LGs is shown in megabases. Colors indicate how scaffolds were assigned: via markers heterozygous in both parents (purple), just parent A (red), just parent B (blue), markers heterozygous in either parent but never in both (black), LD from targeted capture (yellow), or LD from whole-genome sequencing (white). The core linkage map consisting of the first 18 LGs represents the vast majority of mapped scaffolds. LD, linkage disequilibrium; LGs, linkage groups.

Figure 4

Sizes of scaffolds in BglaB1 and in the linkage map. Scaffolds are binned by size. Smaller scaffolds are less likely to have been mapped, but most of the genome is found on scaffolds > 30 kb and most of this sequence has been mapped. Thus, 56% of the genome has been assigned to linkage groups, including 75% of the genic and nonrepetitive regions.

Figure 5

LD across LG IX. All scaffolds are shown in mapped order along both the x- and y-axes. Scaffold color indicates whether placed in the linkage cross (blue) or by LD (green). Scaffolds are scaled relative to physical size; scaffolds shared with other LGs have been reduced in size accordingly. An arbitrary spacer of 10 kb is inserted between all scaffolds. For each scaffold pair, the highest observed pairwise r², for which Δ_AB ≥ 0.15, is plotted among all scaffolds using heat map colors (red: no Δ_AB ≥ 0.15 and/or r² = 0; white: r² = 1). Several notable immune-relevant loci occur on this LG, some of which are indicated. A large block of high LD stretching for several megabases (Table S7) includes the immune-relevant genes spondin-1, sod1, bmplys, and BgTLR (Mitta et al. 2005; Goodall et al. 2006; Galinier et al. 2013; Pila et al. 2016a). The linkage between prx4 and sod1 has been previously examined (Blouin et al. 2013). The Guadeloupe Resistance Complex contains several potential immunity genes, of which grctm6 is the most extensively studied candidate (Tennessen et al. 2015a; Allan et al. 2017). LD, linkage disequilibrium; LG, linkage group.

Figure 6

Synteny comparisons with other gastropods. Scaffolds from A. californica (A) and L. gigantea (B) are plotted along the y-axis (scale = count of scaffolds). Each ortholog to a mapped B. glabrata gene is plotted along the x-axis, in the column corresponding to the B. glabrata core LG. Colors alternate for ease of visualization. Most scaffolds match a single B. glabrata core LG that contains the majority of orthologs. Unlike A. californica, the more distantly-related L. gigantea has several scaffolds that tend to be shared between the same two B. glabrata LGs, likely indicating chromosomal rearrangements. LGs, linkage groups.