Facile generation of <i>Hermes</i> insertion mutants in prototrophic <i>Candida glabrata</i> for use in nutrient-limited environments

ABSTRACT Transposon mutagenesis coupled with deep sequencing (Tn-seq) is currently being deployed in microbial eukaryotes, including the opportunistic yeast pathogen Candida glabrata , for functional genomics research. This method depends on the generation of highly diverse pools of transposon insertion mutants to cover all genes while minimizing the presence of markers and remnants of engineering. Up to now, pools of Hermes transposon insertion mutants in C. glabrata were generated in uracil-requiring ura3∆ auxotrophs, limiting their use in nutrient-restricted environments, such as those of the host. Indeed, we found that ura3∆ mutants were outcompeted by URA3+ prototrophs during colonization of the mouse gastrointestinal tract. To avoid using auxotrophs in Tn-seq experiments, a new scheme was developed for generating prototrophic pools of Hermes insertion mutants. The scheme involved introducing a recessive cycloheximide resistance mutation in the chromosomal RPL28 gene, which did not alter fitness during mouse colonization. When implemented in several different strains of C. glabrata , high insertion densities were obtained, and differences in subtelomeric chromatin compaction were observed that correlated with natural variation in the silencing gene, SIR3 . However, all the strains lacked insertions in the PDR1 and CDR1 genes, which are necessary for resistance to cycloheximide and other antifungals. We directly tested the effect of pdr1∆ mutants and found that they exhibited moderate fitness defects in the gastrointestinal tract of mice even in the absence of antifungals. Thus, the new scheme easily generates high-quality pools of insertion mutants in prototrophic C. glabrata with only minor and knowable limitations. IMPORTANCE Treatment of fungal infections may be improved by a deeper understanding of the genetic mechanisms of colonization within host organisms. Current approaches to deep genetic sequencing in eukaryotic microbes often involve engineered components that have significant biases, minimize microbial complexity, or alter the normal in vivo fitness of opportunistic fungal pathogens. This study designs a new method for developing transposon insertion mutants in Candida glabrata that does not innately introduce altered fitness in mouse models of gastrointestinal tract infection. This scheme is also portable across strains and possibly even fungal species. The findings show that the new method can be used in this pathogenic yeast to yield highly complex pools and reliably identify genetic components of colonization in mouse models of infection.