The Genetic Architecture of Methotrexate Toxicity Is Similar in Drosophila melanogaster and Humans

Abstract

The severity of the toxic side effects of chemotherapy varies among patients, and much of this variation is likely genetically based. Here, we use the model system Drosophila melanogaster to genetically dissect the toxicity of methotrexate (MTX), a drug used primarily to treat childhood acute lymphoblastic leukemia and rheumatoid arthritis. We use the Drosophila Synthetic Population Resource, a panel of recombinant inbred lines derived from a multiparent advanced intercross, and quantify MTX toxicity as a reduction in female fecundity. We identify three quantitative trait loci (QTL) affecting MTX toxicity; two colocalize with the fly orthologs of human genes believed to mediate MTX toxicity and one is a novel MTX toxicity gene with a human ortholog. A fourth suggestive QTL spans a centromere. Local single-marker association scans of candidate gene exons fail to implicate amino acid variants as the causative single-nucleotide polymorphisms, and we therefore hypothesize the causative variation is regulatory. In addition, the effects at our mapped QTL do not conform to a simple biallelic pattern, suggesting multiple causative factors underlie the QTL mapping results. Consistent with this observation, no single single-nucleotide polymorphism located in or near a candidate gene can explain the QTL mapping signal. Overall, our results validate D. melanogaster as a model for uncovering the genetic basis of chemotoxicity and suggest the genetic basis of MTX toxicity is due to a handful of genes each harboring multiple segregating regulatory factors.

Fine-Mapping Nicotine Resistance Loci inDrosophilaUsing a Multiparent Advanced Generation Inter-Cross Population

“…Since strain AB8, which was used to found both panels of RILs, is only present at an appreciable frequency at the Q4 locus in population pB, clarifying the relationship between the groups we identify in pA and in pB is difficult. It is plausible that the effect at Q4 is due to an allelic series, as we (Kislukhin et al 2013) and others (Baud et al 2013) have observed previously for QTL mapped in panels founded by multiple parental genotypes. However, given that Q4 maps to a region containing a family of detoxification genes (see below), it is perhaps more likely that multiple genes are collectively responsible for generating the signal at Q4.…”

“…These populations were maintained for 50 generations before initiating inbred lines, and the multiple rounds of recombination to which the populations were subjected allows high QTL mapping resolution (Darvasi and Soller 1995;Cheng et al 2010). In addition, the increased number of founders relative to traditional two-way QTL mapping gives better representation of the allelic diversity in the species and the potential to characterize the phenotypic effects of multiple alleles at a locus (for example, Kislukhin et al 2013).Using a larval nicotine resistance assay, we screened .800 DSPR RILs, and succeeded in mapping QTL that collectively explained almost 70% of the broad-sense heritability for nicotine resistance, including one QTL that alone contributed 50% of the genetic variation. Following genomewide expression profiling via RNAseq, we identified genes within mapped QTL intervals that showed differential gene expression between founders predicted to harbor lowand high-resistance alleles and/or between nicotine-free and nicotine-supplemented media.…”

“…Using the estimated mosaic haplotype structure of each RIL, we inferred for every variant present in the founders the probability that each RIL harbors the minor allele. For each variant within mapped QTL intervals we then fit a model associating phenotype with genotype, while simultaneously correcting for subpopulation, to test whether individual variants are capable of explaining the QTL peaks (see Kislukhin et al 2013). …”

“…These populations were maintained for 50 generations before initiating inbred lines, and the multiple rounds of recombination to which the populations were subjected allows high QTL mapping resolution (Darvasi and Soller 1995;Cheng et al 2010). In addition, the increased number of founders relative to traditional two-way QTL mapping gives better representation of the allelic diversity in the species and the potential to characterize the phenotypic effects of multiple alleles at a locus (for example, Kislukhin et al 2013).…”

From Nature to the Lab: Establishing Drosophila Resources for Evolutionary Genetics

“…For example, the increased use of isogenic lines propelled by the DGRP-Drosophila melanogaster Genetic Reference Panel , has inaugurated an era of unprecedented studies in Drosophila GWAS studies (Huang et al, 2012;Jordan et al, 2012;Magwire et al, 2012;Weber et al, 2012;Kislukhin et al, 2013;Swarup et al, 2013). In parallel, the development of recombinant inbred lines that constitute the Drosophila Synthetic Population Resource (DSPR) have provided another important resource for the dissection of the genetic basis of complex traits (King et al, 2012;Burke et al, 2014;Marriage et al, 2014;Cogni et al, 2016).…”