A nematode-specific gene underlies bleomycin-response variation in Caenorhabditis elegans

Shannon C. Brady; Stefan Zdraljevic; Karol W. Bisaga; Robyn E. Tanny; Daniel E. Cook; Daehan Lee; Ye Wang; Erik C. Andersen

Tukey HSD), suggesting that interacting loci could underlie bleomycin responses in a 479 background-dependent manner. We performed a two-factor genome scan to map potential 480 epistatic loci but did not identify a significant interaction between the QTL on chromosome V 481 and other loci (Figure S7, File S8) . However, the failure to detect significant interacting QTL 482 could be because we have too few recombinant strains or because too few replicates of each 483 RIAIL were phenotyped. Alternatively, more than two loci might underlie the transgressive 484 phenotype of ECA230 and a two-factor genome scan might not be able to capture this 485 complexity. 486 487 Nonetheless, because ECA230 recapitulated the expected QTL phenotype, we generated two 488 NILs (ECA411 and ECA528) that narrowed this introgressed region to more precisely locate the 489 causal variant (File S5, File S6 ). In addition, the N2 region on the left of chromosome V was 490 removed from both NIL strains to ensure that this region of introgression did not underlie the 491 phenotypic difference between ECA230 and CB4856. The genotypes of ECA411 and ECA528 492 differ in a small region of chromosome V that includes the QTL confidence interval (Figure 2 , 493 . CC-BY 4.0 International license is made available under a The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It . https://doi.org/10.1101/565218 doi: bioRxiv preprint File S5, File S6). Both of these strains were more sensitive to bleomycin than the background 494 parental strain, CB4856. This result could suggest that the introgressed region shared between 495 these strains, which does not include the QTL, conferred some bleomycin-response variation 496 between the N2 and CB4856 strains (Figure 2) . Alternatively, the hypersensitivity of these NILs 497 could suggest the presence of Dobzhansky-Muller incompatibilities between the N2 and CB4856 498 genotypes (Snoek et al. 2014 ) that might affect stress responses of the NILs. However, ECA528 499 was much more sensitive to bleomycin than ECA411 (Figure 2, File S7) . Because ECA528 has 500 the N2 genotype across the QTL region and ECA411 has the CB4856 genotype, these results 501 suggest that the QTL genotype strongly affects bleomycin sensitivity (Figure 2 The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It . https://doi.org/10.1101/565218 doi: bioRxiv preprint cnc-10, contain predicted non-synonymous variants between the N2 and CB4856 strains ( Figure 517 3A, File S10, File S11). These variants could cause differential bleomycin sensitivity between 518 the N2 and CB4856 strains. 519 520 To test these genes in bleomycin-response variation, we systematically deleted each of the 521 candidate genes in both the N2 and CB4856 backgrounds. We used CRISPR-Cas9 mediated 522 genome editing to generate two independent deletion alleles of each gene in each genetic 523 background to reduce the possibility that off-target mutations could cause phenotypic differences 524 (Materials and Methods, Supplemental Information). We tested the bleomycin response of each 525 deletion allele in comparison to the N2 and CB4856 parental strains ( Figure 3B, File S3) . The 526 deletion alleles of C45B11.8, C45B11.6, srg-42, and cnc-10 each had a bleomycin response 527 similar to the respective parent genetic background, which suggested that the functions of each 528 of these genes did not affect bleomycin respo