Whole genome sequence analysis of Cryptococcus gattii from the Pacific Northwest reveals unexpected diversity

Abstract

A recent emergence of Cryptococcus gattii in the Pacific Northwest involves strains that fall into three primarily clonal molecular subtypes: VGIIa, VGIIb and VGIIc. Multilocus sequence typing (MLST) and variable number tandem repeat analysis appear to identify little diversity within these molecular subtypes. Given the apparent expansion of these subtypes into new geographic areas and their ability to cause disease in immunocompetent individuals, differentiation of isolates belonging to these subtypes could be very important from a public health perspective. We used whole genome sequence typing (WGST) to perform fine-scale phylogenetic analysis on 20 C. gattii isolates, 18 of which are from the VGII molecular type largely responsible for the Pacific Northwest emergence. Analysis both including and excluding (289,586 SNPs and 56,845 SNPs, respectively) molecular types VGI and VGIII isolates resulted in phylogenetic reconstructions consistent, for the most part, with MLST analysis but with far greater resolution among isolates. The WGST analysis presented here resulted in identification of over 100 SNPs among eight VGIIc isolates as well as unique genotypes for each of the VGIIa, VGIIb and VGIIc isolates. Similar levels of genetic diversity were found within each of the molecular subtype isolates, despite the fact that the VGIIb clade is thought to have emerged much earlier. The analysis presented here is the first multi-genome WGST study to focus on the C. gattii molecular subtypes involved in the Pacific Northwest emergence and describes the tools that will further our understanding of this emerging pathogen.

Figure 1. Phylogenetic analysis comparison of MLST and WGST data from 22 C. gattii isolates.

Maximum parsimony phylogenetic analysis was performed in MEGA4 on 22 C. gattii genomes with MLST (panel A, one of 270 most parsimonious trees shown) and WGST (panel B, one of 190 most parsimonious trees shown) data. The R265 whole genome sequence was used as the reference for SNP discovery. Both trees are rooted on the VGI/VGIII branch. Bootstrap values less than 50% and for intra-VGIIa, VGIIb and VGIIc branches are not shown. The taxa nomenclature includes a four digit unique identifier, the location of origin (BC = British Columbia, CA = California, ID = Idaho, OR = Oregon and WA = Washington state), and the molecular subtype as determined by MLST analysis (I, IIa, IIb, IIc and III). The number of SNPs included in each data set is indicated (number of parsimony-informative SNPs in parentheses), as is the consistency index (CI) as calculated by MEGA4. While WGST analysis found unique genotypes for all isolates, they are not visible on this tree due to the large numbers of SNPs separating the VGII from VGI and VGIII isolates.

Figure 2. Phylogenetic analysis of WGST data from C. gattii VGII molecular type isolates.

Maximum parsimony phylogenetic analysis was performed in MEGA4 on WGST SNP data from 18 VGII C. gattii genomes using the R265 whole genome sequence as the reference for SNP discovery (R265a-H-BC-02). The analysis found 76 most parsimonious trees, one of which is shown. The total number of SNPs included in the analysis is indicated (number of parsimonious SNPs in parentheses) as is the consistency index (CI) as calculated by MEGA4. Bootstrap values less than 50% are not shown. The tree shown is rooted on the mid-point. Branch lengths as calculated by MEGA4 are indicated above the three major branches. All other branch lengths are less than 500. The taxa nomenclature include a unique four digit identifier, the molecular subtype (a, b or c), the source of the isolate (A = alpaca, C = cat, D = dog, E = environmental, H = human and P = porpoise), the location of origin (BC = British Columbia, CA = California, ID = Idaho, OR = Oregon and WA = Washington state), and the year of collection. For example, B7395a-D-WA-08 indicates that isolate B7395 is a VGIIa subtype collected from a dog in Washington state in 2008.

Figure 3. Comparison of MLST and WGST analysis of VGIIa, VGIIb and VGIIc C. gattii isolates.

Maximum parsimony phylogenetic analysis was performed in MEGA4 on MLST and WGST SNP data from VGIIa (panel A), VGIIb (panel B) and VGIIc (panel C) genomes. Scale bars indicate relative branch lengths for each analysis. The R265 whole genome sequence was used as the reference for SNP discovery (R265a-H-BC-02). VGIIa (panel A) analysis is shown both with and without R265, VGIIb and VGIIc shown only without R265. The numbers of SNPs included in each analysis is indicated. The consistency index value (CI) was calculated in MEGA4. Meaning of taxa nomenclature is described in legends for Figures 1 (MLST) and 2 (WGST). Trees shown are mid-point rooted, bootstrap values less than 50% are not shown.