A common partitivirus infection in United States and Czech Republic isolates of bat white-nose syndrome fungal pathogen Pseudogymnoascus destructans

The psychrophilic (cold-loving) fungus Pseudogymnoascus destructans was discovered more than a decade ago to be the pathogen responsible for white-nose syndrome, an emerging disease of North American bats causing unprecedented population declines. The same species of fungus is found in Europe but without associated mortality in bats. We found P. destructans was infected with a mycovirus [named Pseudogymnoascus destructans partitivirus 1 (PdPV-1)]. The virus is bipartite, containing two double-stranded RNA (dsRNA) segments designated as dsRNA1 and dsRNA2. The cDNA sequences revealed that dsRNA1 dsRNA is 1,683 bp in length with an open reading frame (ORF) that encodes 539 amino acids (molecular mass of 62.7 kDa); dsRNA2 dsRNA is 1,524 bp in length with an ORF that encodes 434 amino acids (molecular mass of 46.9 kDa). The dsRNA1 ORF contains motifs representative of RNA-dependent RNA polymerase (RdRp), whereas the dsRNA2 ORF sequence showed homology with the putative capsid proteins (CPs) of mycoviruses. Phylogenetic analyses with PdPV-1 RdRp and CP sequences indicated that both segments constitute the genome of a novel virus in the family Partitiviridae. The purified virions were isometric with an estimated diameter of 33 nm. Reverse transcription PCR (RT-PCR) and sequencing revealed that all US isolates and a subset of Czech Republic isolates of P. destructans were infected with PdPV-1. However, PdPV-1 appears to be not widely dispersed in the fungal genus Pseudogymnoascus, as non-pathogenic fungi P. appendiculatus (1 isolate) and P. roseus (6 isolates) tested negative. P. destructans PdPV-1 could be a valuable tool to investigate fungal biogeography and the host–pathogen interactions in bat WNS.


Materials and methods
Fungal isolates. P. destructans isolate (MYC80251) has been described previously 3 . P. destructans isolates and other Pseudogymnoascus species listed in Table 1, were maintained on SAB agar. dsRNA extraction. P. destructans was grown in a stationary culture in potato dextrose broth (PDB) at 15 °C for 1-2 months in 500-ml flasks. About 1 g wet weight mycelium was harvested and ground to powder under liquid nitrogen with a mortar and pestle. The powder was collected and suspended in 1 ml extraction buffer (150 mM sodium acetate, pH 5.0, 100 mM LiCl, 4% sodium dodecyl sulfate, 10 mM EDTA, pH 8.0, and 20 mM β-mercaptoethanol) and incubated on ice for 10 min. Total nucleic acids were obtained by standard phenolchloroform extraction followed by precipitation with LiCl and isopropanol. ssRNA and DNA were removed by treatment with S1 nuclease (Life Technologies; Carlsbad, CA) in buffer containing 30 mM sodium acetate (pH 4.6), 50 mM NaCl, 1 mM zinc acetate, 0.5 mg/ml heat-denatured DNA, and 5% (v/v) glycerol at 37 °C for 10 min, followed by incubation with DNase I (Epicentre; Madison, WI) at 37 °C for 30 min. The reaction was extracted with an equal volume of Tris-EDTA-saturated phenol-chloroform-isoamyl alcohol (25:24:1), followed by extraction with an equal volume of chloroform-isoamyl alcohol (24:1). Undigested nucleic acid was precipitated with 2 volumes of absolute ethanol at − 20 °C overnight and recovered by centrifugation for 15 min at 10,000×g. Then the pellet was rinsed with 70% ethanol, air-dried and resuspended in 40 µl of RNase-free water. The extracted material was electrophoresed on a 1% agarose gel, stained with ethidium bromide and visualized by UV transillumination. To confirm the dsRNA nature of the remaining material, a 15-µg sample was treated with RNase III and RNase A (Life Technologies) at 37 °C for 1 h. Denaturing polyacrylamide gel electrophoresis of the purified dsRNA was used to determine the size of the two RNA bands. cDNA synthesis and sequence analysis. Purified dsRNA fractions (1-5 µg) containing 2 segments were denatured in 90% dimethyl sulfoxide (DMSO) at 65 °C for 20 min in the presence of random hexadeoxynucleotide and quickly chilled on ice. The RNA-primer mixture was precipitated with ethanol and resuspended in 11 µl RNase-free water. First-strand cDNA was synthesized using M-MLV reverse transcriptase (Life Technologies), based on the manufacturer's instructions. Briefly, the denatured dsRNA with the random hexadeoxynucleotide were mixed with dNTPs, 5 × First-Strand Buffer, 0.1 M DTT, and RNaseOUT Recombinant Ribonuclease Inhibitor (40 units/µl) and incubated at 37 °C for 2 min. After adding 1 µl (200 units) of M-MLV RT, the mixture was incubated at 25 °C for 10 min, followed by 37 °C for 50 min. The reaction was inactivated by heating at 70 °C for 15 min. The resulting cDNA was amplified with random primers and the variously sized PCR products were cloned in TOPO TA cloning vector. A series of overlapping cDNA clones were obtained from the sequencing of the positive clones (S1 Table). Determination of the ends of each dsRNA was done using FirstChoice ® RLM-RACE Kit (Life Technologies). All the sequence contigs were assembled by Sequencher 4.8 (Gene Codes Co.; Ann Arbor, MI). Conserved sequences in GenBank were identified by searches using the tblastx program. Multiple sequence alignments for the two dsRNA segments were carried out using ClustalW analysis by MacVector 7.2 (Accelrys Inc.; Cary, NC). The maximum likelihood phylogenetic trees were generated by MEGA 6.0 28 (S2 Table). www.nature.com/scientificreports/ Virus purification. Approximately 60 g wet weight fungal mycelium were collected and ground to powder as described above. The homogenate was mixed with extraction buffer (0.1 M sodium phosphate, pH 7.4, containing 0.1% β-mercaptoethanol) at a volume of 5 ml/g of wet mycelium. Following the addition of an equal volume of chloroform, the suspension was vortexed extensively, and the resulting emulsion was broken by centrifugation in a Sorvall GSA rotor at 10,000×g for 20 min. The upper aqueous layer was then mixed thoroughly with 0.5 volumes of 30% polyethylene glycol 8000 (PEG) in 0.85% NaCl and held on ice for 1 h. Negatively staining electron microscopy. Two µl of the virus solution was placed on a glow-discharged copper grid covered with a continuous carbon film. After 1 min of adsorption, the grid was washed with pure water for several seconds and stained with 3 µl of 2% (w/v) uranyl acetate solution for 1 min. The staining solution was blotted away with Whatman No. 1 filter paper. The grid was air-dried completely before it was examined in a JEOL JEM-1400 electron microscope operating at 120 keV. The micrographs were recorded at various magnifications using a 4K × 4K CMOS camera (TVIPS F-416).

Reverse transcription PCR (RT-PCR) assay.
First-strand cDNA synthesis was performed as described earlier and followed by PCR using primer pairs V2085/V2090 (40 cycles of 94 °C for 30 s, 57.5 °C for 30 s, and 72 °C for 1 min) and V2168/V2164 (40 cycles of 94 °C for 30 s, 56.5 °C for 30 s, and 72 °C for 1 min) to amplify partial dsRNA1 and dsRNA2 segments with expected fragment lengths of 828 bp and 613 bp, respectively (S1 Table). PCR products were sequenced using the same primers and aligned to the prototype PdPV-1 sequence from MYC80251 to verify the identities of the amplicons obtained.
Independent laboratory confirmatory analysis. In order to further confirm our findings, one of us (TZ) imported four of the original six positive Czech P. destructans isolates, as well as two isolates not previously tested in the US laboratory, directly from the Czech Republic to China. Additionally, the MYC80251 isolate was imported from the US to China. RNA extraction, RT-PCR, and sequencing analysis of these isolates were carried out with primers, probes, and fresh reagents in a laboratory in Beijing, China that had no prior history of P. destructans investigations.

Results
isolation and characterization of dsRnA from P. destructans. Total nucleic acids extracted from P.
destructans isolate MYC80251 migrated as multiple bands when analyzed by native agarose gel electrophoresis (Fig. 1, lane 1). The most slowly migrating set of bands was found to be removed by treatment with DNase I, showing that these corresponded to fungal genomic DNA (Fig. 1, lane 2). Conversely, only the high molecular weight fungal genomic DNA bands remained following treatment of total P. destructans nucleic acids with a combination of RNase III, a dsRNA-specific endoribonuclease, and RNase A, a pancreatic ribonuclease that cleaves ssRNA (Fig. 1, lane 3). This demonstrated that all of the other bands were RNA species. The two dominant RNA bands were most likely fungal 28S and 18S ribosomal RNAs. Migrating more slowly than the rRNAs were two prominent discrete bands of unknown identities (labeled by an asterisk in lane 1). To determine the nature of these bands, total extracted nucleic acids were digested with a combination of S1 nuclease (a single-strand-specific endonuclease) and DNase I. This treatment abolished all RNA and DNA species except the unknown doublet of bands (Fig. 1, lane 4). Collectively, these results strongly indicated that the unknown bands were double-stranded RNA (dsRNA), most likely derived from a fungal virus. Because the analysis in Fig. 1 was carried out in native agarose gels, the mobilities of RNA species were compared to those of dsDNA size markers. To more accurately gauge the sizes of the two dsRNA bands, they were analyzed by denaturing polyacrylamide gel electrophoresis in comparison to ssRNA size markers. In this type of gel, the denatured dsRNAs were almost totally masked by the presence of the 18S rRNA band in a sample of total P. destructans RNA (Fig. 2, lane 1). However, removal of rRNAs by RNase A treatment of total RNA prior to reisolation, denaturation and electrophoresis allowed us to estimate that the dsRNA species fell in the size range of 1.5-2.0 kb (Fig. 2, lane 2, labeled by asterisk).

Electron microscopy (EM).
The presence of viral particles in P. destructans was confirmed by transmission EM. Viral particles (named as PdPV-1) were purified from mycelia of P. destructans MYC80251 by homogenization, PEG precipitation, and sucrose density gradient centrifugation. Large amounts of isometric viral particles were observed under EM. The diameter of the virions was estimated to be about 33 nm (Fig. 3).
Sequence and phylogenetic analysis. The  www.nature.com/scientificreports/ Figure 1. Analysis of nucleic acids extracted from P. destructans by differential endonuclease sensitivity or resistance. Samples were digested with the indicated enzymes, and undigested species were separated by electrophoresis in 1% native agarose gels. Lane 1: total P. destructans nucleic acids; the 28 S and 18 S ribosomal RNAs are indicated, and the two unknown, more slowly migrating bands are labeled with an asterisk. Lane 2: total nucleic acids following digestion with DNase I. Lane 3: total nucleic acids following digestion with both RNase III and RNase A. Lane 4: total nucleic acids following digestion with both S1 nuclease and DNase I. Size markers: lanes M1, lambda DNA-HindIII digest; lane M2, 2-log DNA ladder (New England BioLabs).  Fig. S1). Specifically, the RdRp of PdPV-1 showed 76% identity with PsV-S RdRp, whereas the CP of PdPV-1 showed 67% identity with that of PsV-S. Homologies (amino acid identity in red and consensus match in blue) were much higher within the core motif regions of the RdRp (amino acids 174-491) of PdPV-1, than elsewhere in the molecule (Fig. 4). Phylogenetic analyses were performed by using the RdRp and CP sequences of the various mycoviruses such as alphapartitiviruses, betapartitivirues, deltapartitiviruses, gammapartitiviruses, totiviruses, and chrysoviruses. Phylogenetic trees derived from both RdRp and CP sequences exhibited three major branches and suggested that PdPV-1 is a member of the genus Gammapartitivirus in the family Partitiviridae ( Fig. 5 and Fig. S2). In addition, phylogenetic analyses of the putative RdRp and CP of PdPV-1 showed that PdPV-1 was most closely related to the gammapartitivirus PsV-S. We found that the 5′ untranslated regions (UTRs) of the plus strands of segments dsRNA1 and dsRNA2 were 8 and 60 nucleotides, respectively; whereas the 3′ UTRs of the plus strands of segments dsRNA1 and dsRNA2 were 54 and 159 nucleotides, respectively. In contrast, a separate study of PdPV-1 33 obtained longer RLM-RACE reads than ours and was able to identify conserved 5′-terminal nucleotide sequence motifs in both genome segments, a feature that is common to the majority of partitiviruses 29,[34][35][36] . Additionally, as a consequence of the observed longer 5′ end of the dsRNA2 sequence, the deduced CP amino-acid sequence in that study 33 had a 35-residue amino-terminal extension relative to the CP ORF that we determined. If confirmed, such an extension would make the PdPV-1 CP unique among the partitiviruses (Fig. S1). This apparent discrepancy remains to be resolved in the future by the direct characterization of the CP. Very recently, a PdPV-1 sequence was also assembled in a metagenomic analysis of fungal transcriptomic datasets 37 . Notably, other than the amino-terminal CP extension 33 , the PdPV-1 sequences obtained in all three studies were 100% identical in the dsRNA1 and dsRNA2 coding regions.

Reverse transcription pcR (Rt-pcR) screening of P. destructans isolates for the presence of
PdPV-1. RT-PCR was performed to examine whether PdPV-1 exists in various P. destructans isolates from diverse sources. P. destructans isolates listed in Table 1 include those collected from fungus-infected bats in the caves from various geographic locations in the US and the Czech Republic, as well as those that were purified from environmental samples collected from caves or mines where bats with WNS hibernated. Closely related strains of Pseudogymnoascus roseus that were isolated from soil samples collected in bat hibernacula were also tested. P. roseus WSF-3629 was collected from Amorphus peat in Wisconsin 27 . P. roseus UAMH1658 and P. appendiculatus UAMH10510 were obtained from the University of Alberta Microfungus Collection and Herbar- www.nature.com/scientificreports/ www.nature.com/scientificreports/ ium. All these bat tissue samples and environmental isolates were subjected to total RNA extraction, followed by virus-specific RT-PCR detection and sequencing of RT-PCR product fragments. Both dsRNA1 and dsRNA2 segments of PdPV-1 were detected in RNA extracts of all P. destructans isolates from the US, irrespective of where they were collected, and their partial sequences (GenBank accession numbers: MN990689-MN990708) were 100% identical to those of our original PdPV-1 strain from MYC80251. Six out of sixteen P. destructans isolates collected in the Czech Republic were also found to be PdPV-1 positive. dsRNA1 and dsRNA2 segments of these six PdPV-1 positive isolates were partially sequenced (GenBank accession numbers: KY609328-KY609339), and we found that some of these PdPV-1 strains had polymorphisms in both their dsRNA1 and dsRNA2 segments, while others were identical to the sequences of PdPV-1 from the US P. destructans isolates. Additionally, exami-  www.nature.com/scientificreports/ nation of the closely related fungi P. roseus and P. appendiculatus showed no detectable dsRNA1 or dsRNA2 segment of PdPV-1 dsRNA (Table 1).
independent analysis of P. destructans isolates from the Czech Republic. To obtain independent verification of some of these findings, four of the Czech Republic isolates (CCF3938, CCF3941, CCF4127, and CCF4130) that had been found to be positive for PdPV-1 by testing in the US were imported directly from the Czech Republic to China and were tested in a laboratory in Beijing that had no prior history of work with P. destructans. These isolates, plus two previously untested Czech isolates (CCF3939 and CCF4129), were confirmed by RT-PCR and sequencing to harbor PdPV-1 (GenBank accession numbers: MK789668-MK789673 and MK789675-MK789680) ( Table 1).The laboratory in Beijing also confirmed the presence of PdPV-1 in the US P. destructans isolate MYC80251 (GenBank accession numbers: MK789667 and MK789674) ( Table 1).

Discussion
Bat WNS continues to devastate the bat populations in the United States and Canada and therefore, there remains an urgent need to understand the origin and spread of this disease. In this study, we characterized mycovirus PdPV-1 infecting the WNS etiological fungus P. destructans. The identity of PdPV-1 was confirmed by (i) demonstration of its dsRNA content by differential sensitivity or resistance to DNA and RNA endonucleases; (ii) complete viral genome sequencing; (iii) nucleotide similarities of the dsRNA viral genome segments and phylogenetic alignment of their encoded proteins with mycoviruses from the family Partitiviridae genus Gammapartitivirus; (iv) TEM confirmation of isometric viral particles, and (v) host specificity of PdPV-1 for P. destructans and no detection of this virus in closely related Pseudogymnoascus species. Overall, PdPV-1 exhibited size, morphology, and nucleotide sequences typical of the fungal viruses in the family Partitiviridae genus Gammapartitivirus 32,38 . These findings in our study (https ://www.biorx iv.org/conte nt/10.1101/05970 9v1) were obtained independently from and prior to a report by Thapa et al. 33 . In contrast to results contained in that publication, we found that some of the Czech Republic P. destructans isolates were infected with PdPV-1. To eliminate the possibility that these results were false-positives, virus testing was repeated separately using four of the original six positive isolates that were shipped directly to China from the Czech Republic. Testing in the laboratory in Beijing verified that these four isolates contained PdPV-1, as did two other isolates not previously tested in the US (Table 1). This confirmation of our results strongly supports our conclusion that PdPV-1 infection is not unique to the North American isolates of P. destructans. We do not know the basis for the discrepancy between the results of our laboratories and those of the other report 33 ; notably, dsRNA1 (GenBank Accession number KP128044 and KY207543) and dsRNA2 (GenBank Accession number KP128045 and KY207544) from the two laboratories have 100% nucleotide identity, respectively. It is possible that it may be attributable to differences in the methods used for dsRNA extraction. Also, we note that the Czech P. destructans isolates that we tested were obtained from laboratories in the Czech Republic, whereas those analyzed by Thapa et al. 33 were acquired from the Center for Forest Mycology Research in Madison, Wisconsin.
Both P. destructans and its closely related species of P. roseus have been isolated from the soil of WNS affected caves and mines 7,39 . PdPV-1 was only detected from the randomly selected samples of P. destructans, but not from P. roseus. The result showed that all US P. destructans isolates were infected with PdPV-1, no matter whether the samples were isolated from the bats with WNS or its environment in the US. Virus-free fungal isolates of P. roseus were not found on the bats with WNS although this fungal species was found in the same environmental reservoir as the WNS pathogen. The absence of PdPV-1 infection in other Pseudogymnoascus species, collected from the same bat hibernacula in the US, suggested a close viral association with P. destructans. The observed host-specificity of PdPV-1 was consistent with the narrow host ranges of mycoviruses due to severe bottlenecks in their horizontal transmission 16 . The above results raise two questions: (i) If PdPV-1 can infect P. roseus, will P. roseus with virus cause WNS? (ii) Can virus-free isolates of P. destructans infect bats to cause WNS? Additional studies are needed to address these questions.
The discovery of a dsRNA mycovirus in P. destructans is consistent with the wide occurrence of mycoviruses in fungi 15,40 . Mycovirus infection of fungi is usually asymptomatic 41 . However, mycovirus infection could alter the ability of plant-pathogenic fungi to cause diseases 23,42 . In plant-pathogenic fungi, mycovirus infections generally reduce fungal yield, attenuate mycelial growth, abolish female fertility in sexual crosses, decrease asexual sporulation, alter colony morphologies, modulate pigment production, and diminish the accumulation of specific metabolites. These effects lead to hypovirulence, an attenuation of the pathogenic outcome of fungal infection on the plant host 20 . The harmful effects of mycoviruses on fungal growth were exploited as beneficial functions in phytopathogen control 15,43,44 . On the contrary, virus infections of animal pathogenic fungi and protozoan pathogens showed hypervirulence 45 . A recent publication suggested that partitivirus infections of the dimorphic human fungal pathogen Talaromyces marneffei caused aberrant gene expression and hypervirulence in an animal model 25 . More in-depth experiments will be needed to discern the outcome of PdPV-1 infection on P. destructans phenotype and virulence. These investigations will be facilitated by the recent availability of a molecular tool kit and experimental infection model of P. destructans 46-51 . In conclusion, mycoviruses are ubiquitous in fungi, while the connection between fungal phenotype and mycovirus presence is not always straightforward. In this study, we discovered the existence of the partitivirus PdPV-1 in P. destructans isolated from various species of bats with WNS or from their living environment in both the US and the Czech Republic. P. destructans PdPV-1 could be a valuable tool to investigate fungal biogeography and the host-pathogen interactions in bat WNS.

Data availability
All nucleotide sequences generated in this study are deposited in the NCBI database.