Use of Population Genetics to Assess the Ecology, Evolution, and Population Structure of Coccidioides

Abstract

During the past 20 years, a general picture of the genetic diversity and population structure of Coccidioides, the causal agent of coccidioidomycosis (Valley fever), has emerged. The genus consists of 2 genetically diverse species, C. immitis and C. posadasii, each of which contains 1 or more distinct populations with limited gene flow. Genotypic data indicate that C. immitis is divided into 2 subpopulations (central and southern California populations) and C. posadasii is divided into 3 subpopulations (Arizona, Mexico, and Texas/South America populations). However, admixture within and among these populations and the current paucity of environmental isolates limit our understanding of the population genetics of Coccidioides. We assessed population structure of Coccidioides in Arizona by analyzing 495 clinical and environmental isolates. Our findings confirm the population structure as previously described and indicate a finer scale population structure in Arizona. Environmental isolates appear to have higher genetic diversity than isolates from human patients.

Keywords: Arizona; Coccidioides immitis; Coccidioides posadasii; United States; Valley fever; coccidioidomycosis; fungi; fungus; microsatellites; population genetics; population structure.

Figure 1

Results of principal coordinate analysis and STRUCTURE analyses of Coccidioides spp. populations. A) Principal coordinate analysis using Nei’s unbiased genetic distance estimates supports 3 main groupings: C. immitis, C. posadasii TX/SA/MX, and C. posadasii Arizona (see also Technical Appendix 3 Figure 2). The greatest separation occurs between species and is reflected in principal coordinate 1 (93.92% of variance). Color-coding for populations: lime green, San Joaquin Valley (SJV); aqua, San Diego/Mexico (SDMX); pink, Texas/South America (TXSA); purple, Mexico (MEXICO); red, Tucson (TUCSON); yellow, Phoenix/Yuma (PHOENIX); orange, soil (AZSOIL). B) STRUCTURE analysis. Microsatellite matrices were analyzed with STRUCTURE 2.3.4 to determine population structure within Coccidioides populations (30). The running length of burn-in period was 100,000 repetitions with 1 million Markov chain Monte Carlo repetitions. Default settings in STRUCTURE 2.3.4 were as follows: the admixture model was used to infer α along with the previous sampling location information model (LOCPRIOR) (30). We used CLUMPP, a cluster matching and permutation program (https://web.stanford.edu/group/rosenberglab/clumpp.html), to define populations within the STRUCTURE algorithm. K is the number of significant populations in each main group. A consensual STRUCTURE plot was generated from the admixture values by using the Clustering Markov Packager Across K (CLUMPAK) server, and final plots were built with STRUCTURE PLOT (32,33).

Figure 2

Population tree of Coccidioides subspecies population splits and mixtures. Maximum-likelihood population tree and presence of gene flow between diverged Coccidioides populations were inferred by using TreeMix software and microsatellites data (34). Direction of arrow indicates migration or gene flow based on admixture models; migration weights are shaded according their importance, supporting gene flow from a soil-derived population (AZSOIL) recovered from animal passage to a clinical-associated population (TUCSON). Color-coding for populations: lime green, San Joaquin Valley (SJV); aqua, San Diego/Mexico (SDMX); pink, Texas/South America (TXSA); purple, Mexico (MEXICO); red, Tucson (TUCSON); yellow, Phoenix/Yuma (PHOENIX); orange, soil (AZSOIL). The drift parameter, represented by horizontal scale, measures the variance in allele frequency change along each branch of the tree. The actual source of each evaluated isolate (clinical, veterinary, or soil) is represented proportionally in the pie chart.

Figure 3

Coccidioides subspecies distribution for North, Central, and South America. The frequency of assignment for each Coccidioides population was plotted in a pie chart for each location, and numbers of isolates from each location are displayed. For example, patients from Mexico were infected with isolates from Texas, San Diego, and Mexico populations, as determined by analysis with STRUCTURE. Because each of the patients’ location is the hospital, no fine-scale population is assessed.