First Detection of Bat White-Nose Syndrome in Western North America

White-nose syndrome (WNS) represents one of the most consequential wildlife diseases of modern times. Since it was first documented in New York in 2006, the disease has killed millions of bats and threatens several formerly abundant species with extirpation or extinction. The spread of WNS in eastern North America has been relatively gradual, inducing optimism that disease mitigation strategies could be established in time to conserve bats susceptible to WNS in western North America. The recent detection of the fungus that causes WNS in the Pacific Northwest, far from its previous known distribution, increases the urgency for understanding the long-term impacts of this disease and for developing strategies to conserve imperiled bat species.

America. The fungus presumably originated in Eurasia, where it occurs on bats but is not known to cause severe infections or population declines (5)(6)(7). Consistent with a point source introduction of an exotic pathogen, P. destructans has spread outward from the WNS epicenter in New York by approximately 200 to 900 km per year (Fig. 1). As of spring 2016, the fungus had reached as far west as eastern Oklahoma, eastern Nebraska, and eastern Minnesota, approximately 1,900 km from the presumed site of introduction.
On 11 March 2016, a moribund little brown bat (Myotis lucifugus) was found in King County, WA (United States), and submitted to a local wildlife rehabilitation center. The animal presented with dried and contracted areas of crusted skin on the wings and died 2 days later. Swab samples of the wings were positive for P. destructans by real-time PCR (8), and the bat was confirmed to have WNS in accordance with defined histopathologic criteria (9). An isolate of P. destructans was obtained by culturing a portion of wing skin on Sabouraud dextrose agar containing chloramphenicol and gentamicin at 13°C.
In eastern North America, P. destructans appears to be spreading clonally, with all isolates exhibiting no genetic diversity at the markers examined (10). However, isolates of the fungus from Europe display significant genetic variation (11). To determine whether the isolate of P. destructans from Washington matched the clonal lineage from eastern North America, we conducted whole-genome sequencing using the Ion Torrent Personal Genome Machine (PGM) on the Washington isolate (NWHC#27099-001), as well as on three additional isolates of P. destructans from eastern North America. These isolates originated from M. lucifugus bats collected in Albany County, NY, in 2008 (NWHC#20631-008) and in Iowa County, WI, in 2016 (NWHC#26994-002) and a tricolored bat (Perimyotis subflavus) in Jackson County, AL, in 2015 (NWHC#44797-145). For comparison to European isolates of P. destructans, we used whole-genome data from isolates from the Czech Republic (isolates CCF3941, CCF3942, CCF4124, and CCF4125) available in the NCBI SRA Database (accession numbers SRR3411506, SRR3411507, SRR3411508, and SRR3411509), as well as the North American type isolate (NWHC#20631-21; NCBI SRA accession number SRR1952982).
Sequencing followed the manufacturer's (Thermo Fisher Scientific) recommendations using the following kits: Ion plus fragment library kit (catalog number 4471252), Ion library TaqMan quantitation kit (catalog number 4468802), Ion PGM Hi-Q view OT2 (OneTouch 2) kit (catalog number A29811), Ion 318v2 (catalog number 4484354), and Ion PGM Hi-Q view sequencing kit (catalog number A30043). The raw sequencing data were processed using the default settings in the Torrent Suite software version 5.0.4. For each isolate, we obtained more than 12ϫ average depth of sequencing coverage over the P. destructans genome reference. Single-nucleotide polymorphisms (SNPs) were identified from the next-generation sequencing reads using the Snippy pipeline version 3.0 (12). Briefly, reads were aligned to the P. destructans type isolate (NWHC#20631-021) reference genome (13) (GenBank accession number GCA_ 001641265.1) using Burrows-Wheeler Aligner (BWA) version 0.7.12-r1044 (14), and variants were called using FreeBayes version 0.9.21-7-g7dd41db (15) with a minimum read coverage of 4 (for each isolate, the genome coverage was at least 98.27% at a 4ϫ threshold), minimum mapping quality set to 60, and minimum proportion for variant evidence set to 90%. A concatenated alignment of core SNPs (defined as an SNP that occurs at a genomic position present in all samples) was generated using Snippy, and an unrooted phylogenic tree was inferred from the resulting 13,379 characters using the neighbor-joining method and the Jukes-Cantor substitution model in CLC Genomics Workbench 9. As a secondary method, maximum-likelihood analysis was performed using the general time-reversible (GTR) substitution model in MEGA version 6.06.
The genome sequence of the isolate of P. destructans from Washington was statistically indistinguishable from the four North American isolates, forming a clade distinct from the four European isolates (Fig. 2). Although they were collected from a more confined geographic region (two hibernacula in the Czech Republic), the European isolates exhibited much higher genetic diversity than the North American isolates and included both known mating types; only the MAT1-1 mating type has thus far been identified among isolates from North America (Fig. 2) (16). The minimal diversity and single mating type observed among isolates of P. destructans from the Nearctic is consistent with clonal spread of the pathogen in North America. These observations also suggest that the Washington isolate of P. destructans is likely of North American origin rather than representing an independent introduction event from Eurasia.
The timing and mechanism by which P. destructans may have reached the Pacific Northwest are unclear. The nearest detection of P. destructans is over 2,100 km away in eastern Nebraska. In the decade since P. destructans was first detected in North America, the fungus has spread approximately 1,900 km from the suspected introduc- tion site. Thus, the recent detection in Washington appears inconsistent with the previously documented and predicted pattern of pathogen spread (17). Sequence analysis of the Washington bat's mitochondrial NADH dehydrogenase subunit 1, cytochrome b (cytb), and cytochrome c oxidase subunit I (COI) genes (18,19) most closely matched GenBank sequences for little brown bats from western North America. Phylogenetic analyses performed on the cytb and COI genes further demonstrated that the infected bat from Washington resided within a well-supported clade (bootstrap support values of 100 and 95 for the cytb gene and the COI gene, respectively; see Fig. S1 in the supplemental material) that included members of M. lucifugus alascensis, a subspecies restricted in distribution to western North America. This finding suggests that the animal became infected with P. destructans in the Pacific Northwest and that the bat was not a vagrant from eastern North America, where WNS is now endemic. Sampling of bats and cave environments in the western United States since the winter of 2013-2014 without detection of the fungus suggests that the pathogen is not widespread or abundant in this environment in western states (i.e., over 2,000 samples tested from Arizona, California, Colorado, Idaho, Montana, Nebraska, New Mexico, Nevada, Oklahoma, Oregon, Texas, Utah, Washington, and Wyoming; A. Ballmann, unpublished data). More intensive surveillance efforts to define the range of P. destructans in western North America should improve our understanding of the presence, distribution, and origin of this pathogen in the Pacific Northwest.
The presence of P. destructans in western North America has major implications for conservation of bat populations. The addition of a potential second disease epicenter could expose novel host species to the pathogen, accelerate the rate at which WNS spreads, and reduce the amount of time that wildlife management agencies have to develop mitigation strategies. Rapid responses to WNS in western North America may be difficult, because the locations of hibernation sites for many western bat species are unknown or inaccessible (20,21). Furthermore, the wintering strategies of many species of bats in the Pacific Northwest are not well understood, with some species known to forage throughout the winter (22). Consequently, new approaches may be necessary to facilitate pathogen surveillance, monitor disease impacts, and conduct mitigation efforts for WNS in this region. The severity, magnitude, duration, and potential ecosystem-level effects of WNS in North America rank it among the most consequential wildlife disease events ever recorded. Although much progress has been made in understanding WNS and in monitoring its spread, more work is needed to determine how the disease dynamics and impacts vary among bat populations in eastern and western North America.
Accession numbers. Nucleic acid sequences for the NADH dehydrogenase subunit 1, cytochrome b, and cytochrome c oxidase subunit I genes of the little brown bat with WNS from Washington are available in GenBank (accession numbers KX290926, KX290927, and KX463942, respectively). Next-generation sequencing data are available through the NCBI SRA database (accession number SRP075419) and the BioProject database (accession number PRJNA322173).

ACKNOWLEDGMENTS
We thank Brenda Berlowski-Zier, Elizabeth Bohuski, Stephanie Steinfeldt, and Karen Vanderwolf for assistance in preparing samples and conducting laboratory tests. We declare no conflict of interest related to this work. The use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government. This work was funded by the U.S. Geological Survey and U.S. Forest Service.