Tracing histoplasmosis genomic epidemiology and species occurrence across the USA

Abstract

ABSTRACTHistoplasmosis is an endemic mycosis in North America frequently reported along the Ohio and Mississippi River Valleys, although autochthonous cases occur in non-endemic areas. In the United States, the disease is provoked by two genetically distinct clades of Histoplasma capsulatum sensu lato, Histoplasma mississippiense (Nam1) and H. ohiense (Nam2). To bridge the molecular epidemiological gap, we genotyped 93 Histoplasma isolates (62 novel genomes) including clinical, environmental, and veterinarian samples from a broader geographical range by whole-genome sequencing, followed by evolutionary and species niche modelling analyses. We show that histoplasmosis is caused by two major lineages, H. ohiense and H. mississippiense; with sporadic cases caused by H. suramericanum in California and Texas. While H. ohiense is prevalent in eastern states, H. mississipiense was found to be prevalent in the central and western portions of the United States, but also geographically overlapping in some areas suggesting that these species might co-occur. Species Niche Modelling revealed that H. ohiense thrives in places with warmer and drier conditions, while H. mississippiense is endemic to areas with cooler temperatures and more precipitation. In addition, we predicted multiple areas of secondary contact zones where the two species co-occur, potentially facilitating gene exchange and hybridization. This study provides the most comprehensive understanding of the genomic epidemiology of histoplasmosis in the USA and lays a blueprint for the study of invasive fungal diseases.

Keywords: Histoplasma mississippiense; Histoplasma ohiense; Histoplasmosis; genomics; molecular epidemiology; species distribution modelling.

Figure 1.

Whole-genome phylogenetic tree of Histoplasma isolates collected in the United States. The branches are proportional to the number of mutations and 1000 ultrafast bootstraps and SH-aLRT were used supporting the major branches of the tree. The nodes in the tree represent common ancestors, and the branching points indicate the divergence of lineages. The tips represent each Histoplasma isolate and are colour-coded according to its source; the percentage of each source of isolation was plotted in a pie chart. The state or country of isolation were added next to each taxa; two isolates had no information regarding the place of isolation (NI).

Figure 2.

Diversity analysis and population genetic analysis of Histoplasma sp. in the USA. (A) Violin plot explaining the genome-wide dispersion of nucleotide diversity in Histoplasma ohiense and H. mississippiense (π). Red dots represent the average π values within each species. Structure plots and Principal Coordinate Analysis (PCA) suggests different patterns of population structure in Histoplasma ohiense and H. mississippiense together (B) or H. ohiense (C) and H. mississippiense (D) separately. Structure analysis revealed the presence or absence of cryptic populations which was inferred by the maximization of log-likelihood scores simulated in the fastSTRUCTURE software and is displayed in the upper panel of each comparisons. Each row represents an individual and the heights and colours of percentage of each population represent the probability of belonging to a given cluster. PCA plots are displayed in the lower panels and depicts the genetic variation distributed in two coordinates, PC1 and PC2, based on polymorphisms and similarities in the genomes of Histoplasma ohiense and H. mississippiense analyzed together or separately.

Figure 3.

Geographic distribution of Histoplasma ohiense, H. mississippiense and H. suramericanum in the US territory. Eighty two isolates were genotyped using whole genome phylogenetic typing and were plotted as pie charts to each respective state of origin. Histoplasma ohiense is respresented by black while H. mississippiense and H. suramericanum are represented by grey and blue colours respectively. Two isolates had no information regarding the location of isolation. The background map is coloured in purple representing the known endemic area of histoplasmosis in the mid-eastern of US; hachure area represents potential endemic areas.

Figure 4.

Species Niche Modelling of Histoplasma ohiense and H. mississippiense. The location of 152 Histoplasma typed by whole-genome or other methods were used as presence-only geographic data to generate species distribution models (SDM) based on the Maximum Entropy (MaxEnt) algorithm (25). Separate MaxEnt models were ran for the distribution of H. ohiense (red area) and H. mississippiense (green area) and performance was assessed using the area under the curve (AUC) of the receiver operating characteristic (ROC). In order to detect areas of niche overlap between the two species of Histoplasma, the predicted distributions were overlayed onto a map (yellow area). A threshold of predicted probability of 0.75 was set to examine areas of that have highest probability of species-specific areas as well as the overlapping areas.