Abstract
Emerging fungal diseases of wildlife are on the rise worldwide (3) and the best lens on the evolution of the fungal pathogens is population genomics. Our genome-wide analysis shows that the newly introduced North American population of Pseudogymnoascus destructans, the causal agent of White-Nose Syndrome (WNS) in bats, has expanded in size, has begun to accumulate variation through mutation, and presents no evidence as yet for genetic exchange and recombination among individuals.
Emerging fungal diseases of wildlife are on the rise worldwide (3) and the best lens on the evolution of the fungal pathogens is population genomics. Our genome-wide analysis shows that the newly introduced North American population of Pseudogymnoascus destructans, the causal agent of White-Nose Syndrome (WNS) in bats, has expanded in size, has begun to accumulate variation through mutation, and presents no evidence as yet for genetic exchange and recombination among individuals. DNA Fingerprinting (6) and MultiLocus Sequence Typing (MLST, 8), support the hypothesis that introduction to N. America of P. destructans involved a genotype of one mating type from Europe, where the fungus is genotypically diverse and both mating types are found (10). WNS was first reported in bats at a single location in New York State in 2006 and has since spread to a substantial portion of eastern North America and an outlier location in Washington State (9). WNS is associated with widespread bat mortality, and has driven one of the most common N. American bat species, Myotis lucifugus, and at least one other species, M. septentrionalis, to the brink of local extinction in eastern N. America (2). The causal agent of WNS, the fungus Pseudogymnoascus destructans, is cave-adapted and cold-tolerant, and causes cutaneous infection during hibernation, leading to disruption in bats’ torpor-arousal cycles, and ultimately the depletion of fat reserves crucial to winter survival (11). European isolates are infectious and cause cutaneous lesions in captive N. American and European bats, but in Europe mass mortality of bats has not been observed and host and pathogen apparently are able to coexist (11). This is consistent with the hypothesis that the hardest-hit N. American bat species were naïve hosts with little intrinsic resistance to P. destructans.
Our primary question was whether or not the clonal population of P. destructans in N. America has begun to accumulate genetic variability through mutation. Given variation, we then asked whether or not the population displays a signature of recombination. The answers to these questions were not accessible through MLST analysis and required whole-genome analysis. We therefore subjected 36 strains of the fungus ‐ representative of its geographical distribution in North America ‐ to whole-genome sequencing (Supplemental Information). We also sequenced one European strain of the same MLST haplotype that occurs in N. American (8). We then aligned the individual sequence reads to a genome reference sequence (NCBI Accession PRJNA39257) to an average read depth of ca. 160X. Lastly, we identified all variants in the genome, including SNPs and indels, that passed a specific set of filtering criteria, with verification by viewing alignments around each variant site.
Based on five lines of evidence, population-genomic observations fit the expectations for a young and expanding clonal population (Figure 1 and Supplemental Information). 1. Variation is exceedingly rare ‐ only 83 variants were discovered among the 36 N. American strains over the 31 MB genome. 2. All of the mutant alleles are infrequent: 75 of 83 variants exist as singletons, one exists in two strains, five exist in three strains, one exists in four strains, and one exists in seven strains. 3. Tajima’s D, a measure of the difference between the observed average pairwise differences among individuals and that expected in a population of constant size in equilibrium for drift and mutation, is strongly negative (-2.7). 4. Recombination is not detected – none of the pairwise comparisons of the variant (biallelic) sites revealed all four possible genotypes, the criterion of the four-gamete test for recombination (5). 5. While the European strain carried the ancestral allele at each of the 83 variant sites identified among the N. American strains, it was by far the most divergent from the others, with 15,793 variants at other positions scattered throughout the genome. Our European strain is therefore substantially different from the strain that originally founded the N. American population, a difference not evident in the previous MLST data (8).
In parsimony analysis of the 83 variant sites, a single, minimum-length tree (83 steps) of the 37 strains was identified (Figure 1). This tree has no internal conflict (consistency index 1.0); branches therefore represent character-state changes that occurred only once in the tree. The tree illustrates how the mutant alleles, in addition to being rare, are locally distributed. All of the branches leading to strains represent singleton mutations, which by definition occurred in only one place. Also, the strains within the three internal clades defined by non-singleton alleles (represented by internal branches) were geographically restricted within the range of the overall sample.
Our results conclusively show that P. destructans represents an asexual clone of single origin. Multiple origins from the diverse European population, even if from the same Haplotype as identified by MLST analysis (8), are not plausible because groups of strains distinguished by hundreds or thousands of variants would have easily been detected. Of course, groups of independent origin may yet appear in the North American population in the future. Our overriding conclusion is that N. America has begun to accumulate new alleles through mutation and is in an early stage of diversification.
Two questions of biological importance emerge for the future of this population of critical conservation importance. First, does the population undergo genetic exchange and recombination? Recombination is always a possibility, even in the absence of two mating types and a sexual cycle, because of the well-known capacity of fungi for parasexuality (4). Given the extreme uniformity of the North American population, individuals of the clone can presumably undergo hyphal anastomosis with mixing of genotypically different nuclei in a common cytoplasm without triggering a somatic incompatibility response. With the accumulation of additional variability in the future, the probability of detecting recombination, if it exists, should go up. Second, how will the further accumulation of variability impact virulence of the fungus on the host bats? The answer here will not only depend on the ongoing population dynamics of the fungus, but also on those of the bats. Evidence suggests that at least one common N. American bat species (Eptesicus fuscus) is resistant to, or tolerant of, infection and has apparently not suffered mass mortality (1). Impacts vary widely for populations of other infected species (2). Moreover, a slight rebound in M. lucifugus populations hit first by WNS, suggests the possibility of some level of increasing resistance in some N. American bat populations (7). Because the ancestry of the North American population of P. destructans is about as clear as it ever gets in a natural arena, this system represents a unique opportunity to follow the evolution of an emerging disease epidemic as it unfolds. Of particular concern is that new introductions may yet add variability to the N. American population and increase the potential for recombination; such changes in the fungus population could affect the durability and strength of newly appearing resistance in bats.
Footnotes
↵* email: jb.anderson{at}utoronto.ca