"Select and Resequence" Methods Enable a Genome-Wide Association Study of the Dimorphic Human Fungal Pathogen Coccidioides posadasii

Next-generation sequencing has unlocked a wealth of genotype information for wild populations, but interpreting it in the context of phenotypes remains a bottleneck, particularly for nonmodel organisms that are difficult to manipulate. To meet this challenge, we pioneered a method for the mapping of genotype to phenotype in natural populations for the thermally dimorphic pathogenic fungus Coccidioides posadasii, using temperature-responsive growth as a proof of concept. We first sequenced the genomes of 66 natural C. posadasii isolates. We then mixed these strains into pools, competed them in growth assays at 37 °C and room temperature, and sequenced the resulting DNA mixtures. We inferred the abundance of each strain in the pool from the sequence coverage of polymorphisms across their genomes in each competition. Ultimately, we used these abundance measures for genome-wide association tests to find loci predictive of, and potentially causal for, temperature-dependent growth as it varied across strains. Emerging from the top hits were variants in the gene D8B26_001557, which we identified from omics resources as a part of the regulatory network controlled by the thermally responsive dimorphism transcription factor Ryp1. Together, our data underscore the power of pooled strain phenotyping and association mapping as a tool for the genetic dissection of trait variation in nonmodel systems.