Genomic signatures of selection associated with benzimidazole drug treatments in Haemonchus contortus field populations

Abstract

Genome-wide methods offer a powerful approach to detect signatures of drug selection in parasite populations in the field. However, their application to parasitic nematodes has been limited because of both a lack of suitable reference genomes and the difficulty of obtaining field populations with sufficiently well-defined drug selection histories. Consequently, there is little information on the genomic signatures of drug selection for parasitic nematodes in the field and on how best to detect them. This study was designed to address these knowledge gaps using field populations of Haemonchus contortus with well-defined and contrasting benzimidazole-selection histories, leveraging a recently completed chromosomal-scale reference genome assembly. We generated a panel of 49,393 ddRADseq markers and used this resource to genotype 20 individual H. contortus adult worms from each of four H. contortus populations: two from closed sheep flocks that had an approximately 20-year history of frequent treatment exclusively with benzimidazole drugs, and two populations with a history of little or no drug treatment. The populations were chosen from the same geographical region to limit population structure in order to maximize the sensitivity of the approach. A clear signature of selection was detected on the left arm of chromosome I centered on the isotype-1 {beta}-tubulin gene in the benzimidazole-selected but not the unselected populations. Two additional, but weaker, signatures of selection were detected; one near the middle of chromosome I and one near the isotype-2 {beta}-tubulin locus on chromosome II. We examined genetic differentiation between populations, and nucleotide diversity and linkage disequilibrium within populations to define these two additional regions as encompassing five genes and a single gene. We also compared the relative power of using pooled versus individual worm sequence data to detect genomic selection signatures and how sensitivity is impacted by sequencing depth, worm number, and population structure. In summary, this study used H. contortus field populations with well-defined drug selection histories to provide the first direct genome-wide evidence for any parasitic nematode that the isotype-1 {beta}-tubulin gene is the quantitatively most important benzimidazole resistance locus. It also identified two additional genomic regions that likely contain benzimidazole-resistance loci of secondary importance. Finally, this study provides an experimental framework to maximize the power of genome-wide approaches to detect signatures of selection driven by anthelmintic drug treatments in field populations of parasitic nematodes. AUTHOR SUMMARYBenzimidazoles are important anthelmintic drugs for human and animal parasitic nematode control with [~]0.5 billion children at risk of infection treated annually worldwide. Drug resistance is common in livestock parasites and a growing concern in humans. Haemonchus contortus is the most important model parasite system used to study anthelmintic resistance and a significant livestock pathogen. It is also one of the few parasitic nematodes with a chromosomal-scale genome assembly. We have undertaken genome-wide scans using a dense RADseq marker panel on worms from natural field populations under differing levels of benzimidazole selection. We show that there is a single predominant genomic signature of selection in H. contortus associated with benzimidazole selection centred on the isotype-1 {beta}-tubulin locus. We also identify two weaker signatures of selection indicative of secondary drug resistance loci. Additionally, we assess the minimum data requirements for parameters including worm number, sequence depth, marker density needed to detect the signatures of selection and compare individual to Poolseq analysis. This work is the first genome-wide study in a parasitic nematode to provide direct evidence of the isotype-1 {beta}-tubulin locus being the single predominant benzimidazole resistance locus and provides an experimental framework for future population genomic studies on anthelmintic resistance.