Objectives: Candida glabrata is the second most common causative agent of systemic candidiasis in North America. Some strains of C. glabrata are resistant to multiple classes of antifungals, severely limiting treatment options. Although it is generally accepted that a majority of C. glabrata infections occur as a result of autoinfection from endogenous populations present in patient microbial flora and a smaller fraction of cases are hypothesized to be nosocomially transmitted patient-to-patient, strain typing is needed to distinguish between the two scenarios. Therefore, when faced with high prevalence of infections with drug-resistant C. glabrata, accurate information of possible transmission in healthcare settings is essential. C. glabrata exhibits extensive genetic variation and rapid evolutionary dynamics, which makes whole-genome sequencing (WGS) ideal for the analysis of the molecular epidemiology and phylogeography. Herein, we investigated the utility of WGS to assess transmission of C. glabrata within a single facility. To better understand the genomic variation and population genetics of C. glabrata in North America, we used multiple WGS methods on 73 isolates from multiple healthcare facilities across North America, including 33 from a single facility.

Methods: We performed WGS using Illumina short-read as well as PacBio and MinION long-read sequencing technology on 73 North American isolates. Thirty-three isolates from 20 patients in a single facility and 40 additional isolates from 40 separate patients at multiple facilities in North America were sequenced. A C. glabrata genome sequenced on PacBio and Illumina was used to generate a de novo reference genome assembly. SNP variants were identified using SAMtools and the NASP SNP analysis pipeline. Phylogenies were inferred using a maximum parsimony algorithm.

Results: All isolates had at least 50X sequence coverage. Several major clades were identified with our phylogenetic analysis. The tightest clustering was observed within pairs of isolates from the same patient. Pairs of isolates from different patients within a single clade had a slightly larger range of genetic variants or SNPs. Of the 20 patients from a single facility, 10 had isolates that grouped in a single clade. The strain-to-strain differences between pairs of isolates from different patients within this clade were lower than the variation seen between isolates from any other clade. High levels of genetic diversity were observed among isolates from the single facility, which included isolates from all seven major clades.  

Conclusions: Substantial genomic diversity of C. glabrata existed even within isolates collected from a single facility. Conversely we observed evidence for potential transmission and clonal expansion of a shared single population of C. glabrata within patients, as half the patient isolates from the facility carried a genetically isolated population of C. glabrata. Continued WGS and phylogenetic analysis may help track the spread of Candida species and improve infection control.


8 / The use of long and short-read sequencing technology to elucidate genomic variation and epidemiology of Candida glabrata in North America