Population genomic and evolutionary modelling analyses reveal a single major QTL for ivermectin drug resistance in the pathogenic nematode, Haemonchus contortus

- Results: Here, we present a genome-wide analysis of two genetic crosses between ivermectin resistant and sensitive isolates of the parasitic nematode Haemonchus contortus, an economically important gastrointestinal parasite of small ruminants and a model for anthelmintic research.
- 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0.
- 6 Department of Comparative Biology and Experimental Medicine, Faculty of Veterinary Medicine, University of Calgary, Calgary, Alberta, Canada Full list of author information is available at the end of the article Doyle et al.
- The ability to utilise these molecular ap- proaches is complemented by extensive information about the structure of the genome and transcriptional differ- ences between the major life stages [19 – 21]..
- contortus populations, together with the limited number of well-characterised ivermectin resistant isolates, at best only circumstantial and inconsistent support is available for the involvement of any of the leading candidate genes.
- The lack of consistency between studies had led to much discussion and debate regarding the complexity of the genetic basis of ivermectin resistance, both in terms of the number of loci involved and the extent of genetic variation between geographically distinct resistant iso- lates .
- Sub- sequent genetic studies showed that none of the leading candidate genes from the literature - Hco-avr-14, Hco- glc-5, Hco-lgc-37, Hco-pgp-9, Hco-pgp-2 and Hco-dyf-7 - showed evidence of introgression [31], and therefore, the genetic mediator(s) of resistance remain unresolved..
- 1 Outline of the initial crosses, backcrosses, in vivo passage and selection experiments.
- Resistance alleles were maintained in the backcross population by selection with ivermectin, before seeding the next round of the backcross with cross-derived heterozygous females and new susceptible MHco3(ISE) males.
- This process was repeated for four rounds of backcrossing, resulting in the backcrossed population becoming genetically similar to the MHco3(ISE) parental line in all regions of the genome not linked to ivermectin resistance.
- This work represents the most comprehensive analysis of the genome-wide impact of selection and genetics of anthelmintic resistance in a parasitic nematode to date.
- 2b) revealed high levels of nucleotide diversity throughout the five autosomes of the genome (Fig.
- of the.
- 7.6 million biallelic SNPs distributed in the genomes of the samples analysed.
- Genetic diversity between each of the three parental populations was assessed by pairwise analysis (measured by F ST of single nucleotide polymorphisms [SNPs] in 10 kbp windows), which confirmed significant genome-wide differentiation (Fig.
- Genome-wide analysis of genetic diversity reveals the same single large introgressed region in both backcross lines To map genetic variation linked to ivermectin resistance in the two parental resistance populations, we sequenced pre- and post ivermectin treatment samples of the initial back- cross between each of MHco10(CAVR) and MHco4(WRS) populations with the MHco3(ISE) susceptible genetic back- ground [39], as well as after subsequent passage of the introgressed populations with further treatment.
- Consistent with the backcross design, pairwise comparisons were made between specific stages of the experiment and the MHco3(ISE) susceptible parent.
- MHco3/4.BC 4 .noIVM), there would be relatively little gen- etic differentiation throughout the majority of the genome between the defined population and MHco3(ISE).
- 97% of the population due to the repeated backcrossing with MHco3(ISE) (resistant alleles comprise half of the ini- tial cross population and reduce by a half with each back- cross), making them largely indistinguishable from the susceptible parent.
- no more than 50% of the genetic material at any location may originate from the resistant parent..
- No other re- gion of the genome displayed a marked and progressive in- crease in differentiation throughout the crosses.
- Geographic origin of the susceptible MHco3(ISE) and ivermectin resistant MHco10(CAVR) and MHco4(WRS) populations used in the genetic crosses.
- Within-population nucleotide diversity for each of the parental populations, calculated as mean diversity per 100 kbp windows throughout the genome using npstat.
- Characterisation of the introgression region located on chromosome V.
- Output of the single-locus evolutionary model, describing sites which were inconsistent with a model of neutral evolution, measured using a likelihood threshold.
- Genetic differentiation (F ST ) measured in 10 kbp windows throughout chromosome V is presented between MHco3(ISE) parent and both MHco3/10.BC 4 .IVM.P 4 (light) and MHco3/4.BC 4 .IVM.P 4 (dark) representing the final time point of the crosses.
- Inset presents the smoothed F ST distribution of the two comparisons.
- yellow) of the crosses (top panels).
- The null expectation is that variance between these samples will be low in regions of the genome not under selection.
- The mean likelihood of the best fitting single-locus model at each locus position is shown by the solid black line.
- The position of the maximum likelihood value is shown by the vertical black dotted line.
- The mean likelihood of the best fitting constrained two-locus model is shown by the black dashed line.
- the value shown represents the best value of the model given that one of the selected loci is at the given position.
- A multi-locus population genetic model provided clar- ity on the extent to which the location of the selected allele could be determined by our data.
- The maximum likeli- hood of the allele under selection was in a broadly.
- Example structures generated using the maximum likelihood parameters for each population show how selection for resistance drives the accumulation of genetic material from the resistant par- ent through the course of the experiment (Fig.
- the large size of each block reduces the precision with which the location of the allele under selection can be identified..
- We further used the multi-locus single driver model to infer the strength and manner of selection in favour of the drug resistance allele.
- In each case the value of the dominance coefficient, h, was less than one, indicating either an additive effect for the MHco10(CAVR) case, whereby each copy of the.
- Within our model, the fitness of a genotype is determined by the alleles at a single resistant allele, with homozygous susceptible worms having fitness 1, homozygous resistant worms having fitness 1 + s, and heterozygous worms having fitness 1 + hs in the presence of the drug.
- blue = low likelihood) for the position in the genome of the resistant allele.
- Changes of the order of 100 likelihood units indicate substantial differences in the extent to which the model fits to the data.
- This is consistent with the pheno- typic characterisation of the initial crosses, whereby the F 1.
- As such, having two copies of the resistant allele had a greater phenotypic effect than one.
- these data therefore have im- plications for the mechanism of the drug resistance allele, and evolution of drug resistance in general terms, as dis- cussed further below..
- We have analysed whole genome sequence data collected from the most recent phase of a multi-generational cross- ing experiment conducted in populations of the parasitic worm H.
- of the genome), each of our analytical approaches consistently identified the same region under selection, with no other comparable regions of introgression else- where in the genome in either of the two backcross exper- iments.
- The annotation of the H.
- contortus genome is ongoing, and hence we can only make a best guess estimate of the number of putative genes in this region:.
- There has been much discussion and debate regarding the complexity of the genetic basis of ivermectin resist- ance, both in terms of the number of loci involved and the extent of geographical variation .
- adjacent to the introgression region, none were found within the peak of the region suggesting that these also are unlikely to be major direct targets of ivermectin-mediated selection.
- Given the frag- mented nature of the T.
- circumcincta assembly, it is intriguing to speculate whether many of the signals ob- served in the earlier study, including one or both of these genes, are not direct mediators of resistance, but rather show evidence of selection due to being linked to nearby driver mutation, as is likely the case in H.
- Population genetic and evolutionary modelling defined the boundaries of the ivermectin resistance QTL.
- Specifically, the random location of recombination events, in the context of strong selection, and genetic drift imposed by population bottlenecks in the experimental design, leads to potential variation in the final outcome of the experiment.
- Our finding of additivity arises from the evolution of the worm popula- tions during passage following the backcross experiment (i.e., in BC 4 .IVM.P 3/4 samples versus the BC 4 .IVM data), and particularly the fixation of alleles from the resistant parent at putative segregating sites near the inferred pos- ition of resistance.
- Our ability to resolve the precise location of the vari- ant under selection was limited by the number of re- combination events in the worm population.
- Precise identification of the location of an allele under selection.
- These bottlenecks were due to the limited num- ber of L 4 worms that it was possible to collect from each sheep and transplant in the successive generation of the cross.
- Genetic drift induced by successive population bottlenecks introduced considerable uncertainty in the outcome of the experiment, such that replicate sets of allele frequencies from our model showed considerable differences between each other (Additional file 11:.
- beyond the clear large-scale patterns observable in the data and detailed above, more minor details of the output might not be seen again were the experiment to be repeated.
- The structure of the experi- ment thus imposes a limit on our ability to infer the location of a selected allele.
- populations, for example by the simultaneous passage of the population through multiple animals.
- While standard metrics can identify sites of maximum differentiation in a population with great pre- cision, neglect of the inherent stochasticity in the out- come of an experiment can lead to an overconfidence in the extent to which they provide an accurate diagnosis of the causative variant of selection..
- different versions of the MHco3(ISE).N1 reference gen- ome assembly (Fig.
- In the fragmented draft genome assemblies, the signals of genetic differentiation be- tween susceptible and resistant populations were nu- merous and dispersed across many assembly scaffolds, suggesting that many discrete regions of the genome may be under selection (Fig.
- In this study, we have used only a subset of the total variation present, i.e., only single nu- cleotide polymorphisms, to characterise the introgression;.
- however, understanding the functional consequences of population-specific diversity will rely on a more compre- hensive description of all of the variation that defines a population.
- 7 Impact of genome contiguity on the resolution of the introgression region.
- 10.BC 4 .P 4 – is presented using three versions of the H.
- This work represents the most comprehensive analysis of the genetics of anthelmintic resistance in a parasitic nematode to date, and demonstrates the power of genetic crossing and a contiguous genome assembly to eliminate false positive genetic signals typically linked to resistance.
- In the first generation of each cross, female worms of the ivermectin resistant parental populations [MHco10(- CAVR) or MHco4(WRS)] were crossed with male MHco3(ISE) worms, to generate lines designated MHco3/.
- L 4 /immature adult worms into worm-free recipient lambs all within 2 h of the original collection.
- In vivo iver- mectin selection was applied after mating and before collection of eggs to enrich for ivermectin resistant adults to be used in the subsequent round of the backcross.
- Pfizer) and closely monitored on completion of the surgery.
- and homozygous susceptible (q 0 ) individuals throughout the course of the experiment under the assumption of selective neutrality..
- if the next generation of the population contains N individuals, the probability of obtaining n 1 , n h , and n 0 individuals with each diploid variant is given by:.
- Evaluating this process gives a probability distribution for the frequency of the resistant allele at any given point in the experiment, dependent upon the number of resistant alleles in the initial resistant population.
- We first suppose that the frequency of the resistant allele in the resistant parental population is equal to.
- however, using the data an esti- mate can be made for this statistic, expressed as a distribution of the allele frequency.
- This process generated the neutral expectation of the re- sistant allele frequency conditional on the observation of this frequency in the resistant population.
- If at the sam- pling point t, a total of N t worms were collected, the dis- tribution of the number of worms n t with the resistant allele at that point is given by:.
- A multi-locus model was developed to describe the manner in which allele frequencies would be expected to change over the course of the experiment.
- Firstly, we identified genomic sites for which the frequency of the resistant allele was 95% or greater in the resistant parent, requiring a read depth of at least 50x coverage for such sites to achieve a good level of statistical certainty.
- Secondly, we noted that, following the backcross performed at the start of the experiment, no locus can be homozygous for the resistant allele and as such, the resistant allele frequency can be no greater than 50% in the population.
- Conservatively, sites no more than eight standard deviations from the mean of the posterior frequencies were included in the analysis.
- We suppose that selection acts in favour of the resistance allele at a given locus when worms are in a sheep being treated with ivermectin, such that the fitness of individuals that are homozygous for the resistant allele (w 1.
- We assumed that all worms have equal fitness in the ab- sence of the drug..
- Our model is defined in terms of the manner in which selection acts upon the population, by the location in the genome of the selected allele, and the extent to which that allele conveys a fitness advantage to worms in the presence of drug treatment, this fitness advantage being characterised in terms of s and h.
- Accounting for the stochasticity of the system, a set of at least 250 simulations were generated for each set of parameters, calculating the mean likelihood fit to the data across the simulations.
- Confidence intervals for the maximum likelihood loca- tions of the selected allele were calculated.
- we calculated a test likelihood equal to the mean plus one standard deviation of the replicate likelihoods generated by these parameters.
- We believe this gives a conservative estimate of the uncertainty in the location of the selected allele, given the stochasticity in- herent in the experiment..
- The slope of the regression was plotted.
- red) and passages 3 (orange) and 4 (yellow) of the crosses.
- The repeated backcross removes most of the resistant genotypes from the population..
- a distinct line is shown for each of the 250 replicate simulations run for the parameters generating the maximum likelihood fit.
- The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript..
- The genome and developmental transcriptome of the strongylid nematode Haemonchus contortus.
- Genetic analysis of inbreeding of two strains of the parasitic nematode Haemonchus contortus.

Xem thử không khả dụng, vui lòng xem tại trang nguồn
hoặc xem Tóm tắt