1 Vitamin B as a virulence factor in Pseudogymnoascus destructans skin infectionMiroslav Flieger, Hana Bandouchova, Jan Cerny, Milada Miroslav Kolarik, Veronika Kovacova Petr ebesta, Eva & Jiri Pikula caused by Pseudogymnoascus destructans g ml cell detachment, loss of mitochondrial membrane potential, polymerization of cortical actin, and Hyperaccumulation of vitamin B Microorganisms produce metabolites to support their physiological functions and biotic interactions1. ese may induce deleterious toxicity and contribute to microbial pathogenicity or play crucial roles as compounds vital for animal metabolism. Among these metabolites, riboavin (vitamin B2) enables microbial intracellular survival and promotes virulence2. Although vertebrates cannot synthesize it, riboavin is required for oxidoreduction metabolic processes5. It may also be toxic to cells exposed to light6; however, healthy, euthermic vertebrates avoid these adverse eects by excretion of its excess in urine. Multiple molecules likely enable fungal invasive infection of hibernating bats with Pseudogymnoascus destructans7, a generalist pathogen and causative agent of white-nose syndrome (WNS)10. Being restricted to the skin, WNS is not induced by a systemic fungal infection, yet it resulted in population decline in six Nearctic bat species17. Here we investigate molecular mechanisms of fungal pathogenicity that lead to tissue damage and potential detrimental physiological cascade in infected animals. We combine evidence from analytical chemistry, microbiology, pathology and cell biology to discover that accumulation of riboavin within bat skin may act as a vir ulence factor of WNS. In fact, this nding links existing data and hypotheses on WNS18 by revealing a novel mechanism of pathophysiology, operating in both torpid and euthermic animals.Resultse extent of skin damage due to invasive growth of Pseudogymnoascus destructans (Fig.1a) is visible under ultraviolet (UV) light transillumination, documenting lower disease intensity in a Palearctic bat compared to WNS lesions over extensive wing area in a Nearctic bat (Fig.1b,c). We dene WNS lesions as wing membrane Laboratory of Fungal Genetics and Metabolism, Institute of Microbiology, Academy of Sciences of the Czech Republic, Prague, Czech Republic. Department of Ecology and Diseases of Game, Fish and Bees, University of Veterinary and Pharmaceutical Sciences Brno, Czech Republic. Department of Cell Biology, Faculty of Science, Charles University in Prague, Czech Republic. Institute of Vertebrate Biology, Academy of Sciences of the Czech Republic, Brno, Czech Republic. Correspondence and requests for materials should be addressed to J.P. (email: firstname.lastname@example.org) or N.M. (email: email@example.com)R A P OPEN
2 sites that uoresced aer illumination with 360 nm UV light as yellow-orange spots (Fig.1) in cases where P. destructans was identied and conrmed with histopathology to invade living tissue of the bat. Whole-body screening of a Myotis myotis specimen using a stereo microscope revealed uorescent foci both under UV and blue excitation (Fig.2). Scanning electron microscopy revealed circumscribed dermal nodules elevated over the skins surface at sites with UV-conrmed WNS lesions (Fig.2a). While uorescent WNS spots contain 0.03.13 ng (median 0.37 ng, n 54) of P. destructans DNA per 7 mm2 of wing biopsy based on quantitative polymerase chain reaction (qPCR) (Fig.1a), ying membrane without UV uorescence had signicantly lower fungal loads (paired t-test: t53 9.972, p 0.001). In search of uorescent compounds among secondary metabolites of P. destructans we investigated crude extract (CH2Cl2) of glucose yeast liquid cultivation medium of the strain P. destructans CCF3941 (see Fig.3a for Figure 1. Ultraviolet uorescence on bat wings corresponding with WNS lesions and increased P. destructans load. ( a ) Wing punch biopsies with UV uorescent spots contain higher fungal loads than do paired spots without uorescence (y 0.294x 0.015, F1,52 20.83, p 0.001, r2 0.29). Fungal load was quantied from a wing punch biopsy using qPCR with TaqMan chemistry and dual probes50. (b ) Transillumination of a wing membrane of Palearctic Myotis dasycneme shows yellow-orange spots at wing areas indicative of WNS lesions. ( c ) Transillumination of a wing membrane of Nearctic Myotis lucifugus showing WNS-associated yellow-orange uorescence over extensive wing area. Image modied with permission from15 Scale bar 1 cm. Figure 2. Visualization of uorescent skin lesions in plantar surface of M. myotis. (a) Scanning electron microscopy shows circumscribed dermal nodules elevated over the skins surface. (b) Fluorescence microscopy, blue emission (UV excitation) showed uorescent foci at sites with dermal nodules. (c) Fluorescence microscopy, green emission (blue excitation) showed uorescent foci at sites with dermal nodules with slightly dierent pattern to (b). Fluorescence spectral variation in the WNS lesions indicates spatial distribution of secondary metabolites produced by P. destructans during progressing skin infection. (df) A control, euthermic bat captured in summer when bats heal from WNS had no nodular lesions on its plantar skin photographed with scanning electron microscopy (d), and with uorescence microscopy using blue (e) and green (f) emission. Scale bar m.
3 isolates details and Fig.b for their phylogenetic context) fractionated on a solid phase extraction (SPE) C 18 cartridge using a stepwise gradient of methanol in water. e fraction eluted with 80% methanol contained a pure yellowish compound. Using UV and Fourier transform mass spectrometry (FTMS), we identied the compound as lumichrome (molecular mass [M H] at m/z 243.08751 with elemental composition C12H11N4O2, Fig.4a). Lumichrome is a degradation product of riboavin. To prevent riboavin modication (formation of lumichrome as an experimental artifact), the liquid cultivation medium was directly loaded on the SPE C 18cartridge and eluted with methanol. Methanol eluate was directly analyzed by FTMS and tandem mass spectrometry (MS/MS). e determined molecular mass [M H] at m/z 377.14525 with elemental composition C17H21N4O6 signied Figure 3. Pseudogymnoascus strains used in the study. (a) Table listing fungal strains identication and origin53,54. Two nonpathogenic Pseudogymnoascus strains were isolated from swab samples from bats negative for WNS on UV transillumination test and histopathology evaluation. CZCzech Republic, RURussia. (b) Maximum likelihood phylogenetic tree based on ITS rDNA sequences showing relatedness of 12 isolates used in this study (in bold). e closest blastn matches to sequences from our study were downloaded from NCBI GenBank. Sequence alignments were obtained using the MAFFT 7 G-INS-1 algorithm55. Maximum likelihood analyses56 were performed using a K2P I substitution model chosen with an Akaike Information Criterion comparison57. Bootstrap support was obtained from 1,000 pseudoreplicates. Taxonomy was adopted from Minnis and Lindner49 and displayed with letters at the published clade representatives. Sequences with identical residues were added into the tree aer analyses. Scale bar depicts substitutions per site. Intensity x 1070 0.2 0.4 0.6 0.8 1.0 1.2 4154 15.5 4164 16.5 417 415.10149 C17H20KN4O6, [M+K] m/z Intensity x 107399.03 99.5 400.04 00.5 401.0 m/z 0.0 0.2 0.4 0.6 0.8 1.0 1.2 399.12823 C17H20N4NaO6, [M+Na] +MS +MS Intensity x 1092.0 1.5 1.0 0.5 0.0 200 400 600 800 1000 1200 1400 m/z 1+ 367.18820 1+ 591.38766 1+ 921.73190 +MS C39H62KN6O161+ 909.38553 1+ 377.14525 C17H21N4O6, [M+H]C17H20KN4O6, [M+K] 1+ 415.10124 +MS 200 400 600 800 1000 0 2 4 6 Intensity x 1081+ 265.13124 C17H20N4NaO6, [M+Na] 1+ 625.24800 C39H62KN6O15C39H63N6O16m/z C17H21N4O61+ 377.14525 C17H19N4O5C12H11N4O21+ 243.08751 +MS2 (qCID 377.14525) 100 150 200 250 300 350 0 1 2 3 400 4Intensity x 108m/z ab cd e Figure 4. FTMS and MS/MS data for riboavin. (a) MS/MS fragmentation of riboavin. Lumichrome was detected with molecular mass at m/z 243.08751. (b) FTMS spectrum of the SPE extract of 8-week-old liquid cultivation medium of P. destructans strain 20631-21T. (c) FTMS spectrum for riboavin in bat skin containing WNS lesions. Ultraviolet light guided biopsy from uropatagium of a WNS-positive M. myotis was homogenized with liquid nitrogen and prepared for the analysis in 40% MeOH/H2O. ( d,e ) Magnication of the FTMS spectrum for riboavin from bat skin biopsy. Blue circles indicate riboavin detection.
4 riboavin (Fig.4b). Likewise, riboavin was directly determined with FTMS in a 40% methanol fraction from a homogenized bat skin biopsy that contained WNS lesions (Fig.4c, and see Fig.4d,e for FTMS spectrum magnication that includes riboavin). e crude extract fraction from P. destructans liquid cultivation medium eluted with 50% MeOH also contained a cinnamon-colored compound. Analysis with FTMS and MS/MS identied the compound as the siderophore [Fe3] triacetylfusarinine C (data not shown). We used reversed-phase high-performance liquid chromatography (HPLC) to determine presence of the main extracellular metabolites (i.e., riboavin and [Fe3 ] triacetylfusarinine C) in stationary cultivated liquid cultivation medium (Fig.5a). All 12 tested Pseudogymnoascus isolates (Fig.3) produced riboavin but diered in those quantities. While P. destructans strains produced riboavin in concentrations up to 37 g ml 1 aer 12 weeks of culture (Fig.5b), nonpathogenic Pseudogymnoascus spp. strains produced a maximum of 4 g ml 1 in the same time (Fig.6c). e second main metabolite, [Fe3] triacetylfusarinine C, was produced only by pathogenic strains and its maximum concentration (26 g ml1) was reached between 6 and 8 weeks of cultivation (Fig.5c). Riboavin concentrations ranging from 25 to 200 g ml 1 did not inhibit growth of P. destructans compared to the control (Fig.6d). The secondary metabolite production of P. destructans varies in time (Fig.5b) and across tissue (Fig.2, Supplementary Video 1). To estimate riboavin concentration in a naturally infected bat wing, we used the lambda-scanning feature of a confocal microscope. UV and blue excitation colour channels revealed a nonhomogeneous distribution of uorescence patterns across the wing (Fig.7). e background emission spectrum of the bat wing membrane peaked between 460 and 510 nm (Fig.7c). e WNS lesion emission spectrum contained two characteristic peaks, with maxima at 510 nm and 633 nm. e emission spectrum of riboavin dissolved in phosphate buered saline mimicking physiological pH and osmolarity peaked at 513 nm. It constituted a major contribution to the uorescence spectrum of the rst peak in WNS lesions excited by the 405 nm laser. [Fe3 ] triacetylfusarinine C, previously identied uorescent compound associated with WNS7, exhibited an emission spectrum peak at 467 nm (Fig.7c). is diered from the two peaks of the WNS lesions, thus indicating a minor contribution of [Fe3 ] triacetylfusarinine C to the WNS lesions uorescence. e WNS lesions second peak at 633 nm corresponds to another putative molecular component. We unmixed and separated two spectral curves, wing background and WNS spot-derived channels, to estimate the local concentration of riboavin and its metabolites in the tissue (Fig.7b). e mean riboavin concentration per WNS lesion ranged from 54 to 141 g ml 1 Riboflavin 9. 484 AU 0 0. 01 0. 02 0. 03 0. 04 0. 05 0. 06 [F e3+] triacetylfusarinine C 15. 195 We eks of culture Concentration (g ml-1) 0 5 10 15 20 25 2468 10 12 Mi nute s 048 12 16 20 24 28 32 abW eeks of culture 246 81 01 2 0 5 10 15 20 25 c Figure 5. Production of riboavin in Pseudogymnoascus destructans (a) HPLC chromatogram of SPE C18 extract from liquid cultivation medium of P. destructans strain 20631-21T (see Fig.6a,b for medium and blank controls). (b) Average production curve ( standard error) of riboavin in six strains of P. destructans (Fig.3) shows continuous metabolite accumulation. (c) Average production curve of [Fe3] triacetylfusarinine C in six strains of P. destructans peaks aer 6 to 8 weeks of culture. e concentration of the metabolite was corrected for biomass content in the fermentation medium. 0 0. 04 0. 08 0. 12 0. 16 0. 2 0. 24 0. 28 0. 32 Mi nute s 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 AU 0 0. 2 0. 4 0. 6 0. 8 1 1. 2 1. 4 0 4 8 12 16 20 24 28 32 We eks of culture Concentration (g ml-1) 0 5 10 15 20 25 2468 10 12 Control5 01 00 1502 00 16 18 20 22 24 Riboflav in concentration ( g ml-1)Colon y diameter (mm) 20631-21TCCF3943 CCF4986 c a b d Figure 6. Production and tolerance of riboavin in Pseudogymnoascus fungi. (a) HPLC chromatogram of SPE C18 extract from liquid cultivation medium of Pseudogymnoascus sp. strain CCF5026. (b) Blank liquid cultivation medium HPLC chromatogram. (c) Average production curves ( standard error) of riboavin in non-pathogenic Pseudogymnoascus strains. e concentration of the metabolite was corrected for biomass content in the fermentation medium. (d) Tolerance of P. destructans strains to riboavin in concentrations ranging from 0 to 200 g ml1. Riboavin was added aer autoclaving the medium containing glucose (20 g l1), yeast extract (5 g l1), and agar (20 g l1). Colony diameter was measured aer 49 days. ree replicates were used in each experiment, and their values may overlap.
5 (median 86, nspots 12, nbats 4), but local deposits uoresced with maximum intensity up to 815 g ml 1 (median 200) (Supplementary Fig. S1). To examine riboavin urinary excretion in bats, we measured urine concentrations of riboavin and creatinine with HPLC. Riboavin concentration in bat urine was standardized to 1 mol of creatinine to account for density of the urine sample. Lower urinary excretion of riboavin during late hibernation (Wilcoxon rank test: W 153, p 0.001, n 100) signied metabolite hyperaccumulation in skin. Urinary excretion of riboavin did not correlate with infection intensity in late hibernation as quantied with the number of UV uorescent lesions on bat wings (Pearson's r 0.17, P 0.345, n 31; Fig.8). We treated primary bat skin broblasts with a series of riboavin concentrations (0 g ml1). Measuring lactate dehydrogenase activity indicative of cell death, riboavin proved to be cytotoxic in a dose-dependent manner under conditions simulating hibernation torpor (darkness, 8 C) from 12.5 g ml 1 and simulating periodic arousals (darkness, 37 C) from 25 g ml1 (Fig.9a,b). To mimic the consequences of bat emergence from hiber nacula prior to sunset, we approximated that bats might be exposed to reduced daylight intensities for a limited time. e cells exposed to 110 mol of photons m 2 s 1 for 30 minutes, treated with riboavin, demonstrated a rapid onset of oxidative stress from 25 g ml1 on ow cytometry (Fig.9c). We visualized integrity and dynamics of three cellular compartments: mitochondria as a reection of energy metabolism, the actin cytoskeleton as an indicator of overall plasma membrane shape and adhesion, and the nucleus as an indicator of cell cycle dynamics or apoptosis. Treatment with a riboavin concentration of 200 g ml 1 resulted in overall pathology (Supplementary Fig. S2). Fibroblasts began to detach aer approximately 12 h. Mitochondria shape was altered to vesicle-like structures. Gradual decrease in mitochondrial uorescence intensity indicated the loss of membrane potential and, therefore, depletion of adenosine triphosphate in bat broblasts. e probe for polymerized actin revealed a similar eect of high riboavin concentrations (100 and 200 g ml 1), indicating robust polymerization of cortical actin. No signs of apoptosis (fragmentation of nuclei, Figure 7. Quantication of riboavin in a WNS lesion. (a) Fluorescence of skin lesion excited with UV light (405 nm). (b) Quantication of riboavin content in skin lesion based on uorescence signal. (c) Emission spectra proles (lambda scan) of standard chemicals, wing background, and yellow-orange uorescence regions of skin lesions observed in two bat specimens. CZCzech Republic, LVLatvia, PBSphosphate-buered saline. April 0 5 10 15Riboflavin in urine (g ml-1 mol(creatinine)-1) aMonth dur ing hiber nation No vember 100.5101101.5102102.5103 10-110-0.5100100.5 Number of UV-fluorescent WNS lesions b Figure 8. Riboavin concentration in bat urine. (a) Time variation of riboavin concentration in M. myotis urine between early hibernation (November, n 36) and late hibernation (April, n 64). (b) Relationship of riboavin in bat urine and sum of WNS lesions detected during late hibernation by UV transillumination of both wings (n 31). Regression of log-transformed variables was not signicant (P 0.34) and Pearsons correlation was low (r 0.17). Riboavin concentration in bat urine was standardized to 1 mol of creatinine to account for density of the urine sample.
6 formation of apoptotic bodies) were detected visually. Instead, nuclear contraction indicating necrosis was typical for cells treated with a 200 g ml 1 concentration of riboavin (Supplementary Fig. S2), similar to necrosis in natural infections (Fig.10a). Tissue necrosis in the vicinity of a WNS lesion manifested on histopathology as a loss of tissue structure and stainability (Fig.10a,c). While no immune response was present around the necrotic tissue in some bats (Fig.10a), other cases exhibited fungal sequestration by inammatory inltration with neutrophils (Fig.10b).DiscussionDuring hibernation, bats enter a hypometabolic and hypoper-fusion state, when tissues exhibit low oxygen levels and biomolecules are slowly excreted or metabolised. Tissues of hibernators show natural resistance to the hypoperfusionreperfusion experienced in repeated cycles of torpor and arousal during hibernation23. At the site of skin infection by P. destructans hypoperfusion promotes accumulation of fungal metabolites, including riboavin. Intercontinental dierences in disease outcome are dependent on infection intensity that diers between taxa15,24 (Fig.1), not on pathogenic mechanisms exerted by the infectious agent. Pseudogymnoascus destructans strains from Europe, Asia and North America produce riboflavin, which accumulates in liquid cultivation medium and in skin lesions produced by the fungus (Fig.4). We tested a full range of riboavin concentrations to its upper solubility limit at 200 g ml 1. ese concentrations are likely encountered by cells as the fungus secretes up to 37 g ml1 of riboavin aer 12 weeks of culture (Fig.6) and its mean concentration in WNS lesions ranges between 54 to 141 g ml 1 with maxima at hundreds g ml 1 (Supplementary Fig. S1). Focal concentrations or deposits, albeit possibly underestimated due to possible eects of uorescence quenching, are beyond concentrations that are cytotoxic to cells in hibernation-like conditions (Fig.9a,b) and induce oxidative stress in conditions simulating emergence from hibernacula (Fig.9c). Spectroscopic ngerprints allow dierentiation of autouorescent microorganisms25. e photochemical quality of riboavin26 and its hyperaccumulation within the infected skin tissue is responsible for the distinctive orange-yellow uorescence in UV light used to identify bats positive for WNS15. Various dermatophytes uo resce (including representatives from Trichophyton and Microsporum ), but most do not grow at low temperatures typical for microclimate in hibernacula. Although Trichophyton and Microsporum and other geophilous dermatophytes occur frequently in the hibernacula sediment, these fungi were missing on the skin of hibernating bats, even if the selective media were used27. ere is currently no support that other microorganisms with yellow to orange uorescence co-occur on bats with P. destructans at signicant levels. Additionally, in contrast with conventional dermatophytes, UV uorescence of P. destructans infection is not associated with fungal growth restricted to the skin surface. It is elicited only aer invasive growth through and replacement of living tissues with cup-like erosions pathognomonic for WNS that are packed with fungal hyphae of P. destructans is characteristic makes UV-guided biopsy a highly successful diagnostic tool with sensitivity from 95.5 to 98.8% and 100% specicity for WNS15. Infections of hibernating bats similar in appearance to WNS caused by a dermatophyte Trichophyton redellii were not observed to uoresce like P. destructans -produced lesions31. Trichophyton redellii is easily cultivable and identiable and was not found during our surveys. Fluorescence spectral pattern that appears aer P. destructans skin invasion indicates that other molecules are responsible for breaching the protective skin barrier and utilizing resources from living tissues. Pseudogymnoascus destructans secretes proteolytic and hydrolytic enzymes8 ,32,33, putatively facilitating skin invasion. Requirement for iron in microorganisms invading host tissues34 might explain P. destructans co-secretion of riboavin and [Fe3] triacetylfusarinine C. Microorganismal iron uptake, that siderophores provide, requires reduction of Fe3 to Fe2 for which avins donate an electron35. Our results show that while riboavin linearly accumulates in liquid cultivation medium, [Fe3 ] triacetylfusarinine C production peaks aer six weeks, which might reect in a WNS lesion as minor contribution of the siderophore to UV uorescence. Cytoto xicity (%) 12.5 25 50 100 200 0 10 20 30 40 50 60 Riboflavin concentration (g ml )a 0 c b 12.5 25 50 100 200 0 10 20 30 40 50 Riboflavin concentration (g ml ) 0 0 12.52 55 0 100 200 2000 4000 6000 8000 Riboflavin concentration (g ml ) Mean fluorescence intensity (A U) Figure 9. Cytotoxicity of riboavin on primary bat broblasts, mimicking skin exposure to dierent temperature and light conditions during hibernation. (a) Cytotoxicity aer 24 h of exposure to riboavin at 8 C, representing temperature during torpor. Experiment performed in septuplicate with 15,000 cells per replicate. (b) Cytotoxicity aer 24 h of exposure to riboavin at 37 C, i.e. bat temperature during arousal from torpor. Experiment performed in septuplicate with 15,000 cells per replicate. (c) Reactive oxygen species (ROS) measured by uorescence activated cell sorting for ROS sensor positivity. Cells were grown in biological triplicates to 50% conuency, incubated with riboavin for 2 h, then illuminated in the presence of ROS sensor for 30 min at approximately 1/5 of daylight intensity. Number of cells evaluated using ow cytometry per each experiment ranged from 11,992 to 18,218.
7 Cell cultures derived from tissues of bat species at risk may provide a cutting-edge tool to study innate responses to WNS36. To evaluate biological activity of riboavin concentrations observed in WNS skin lesions, we considered the fact that riboavin produces free radicals upon contact with oxygen and exposure to light5 ,6. Such chemical reactions lead to oxidative injury of tissues, to modulation of cell signalling to apoptosis, or to necrosis37,38. Derangement of connective tissue cells characterizes a virulent P. destructans infection10 (Fig.10c). Figure 10. Histopathology of WNS lesions. (a) M. daubentonii, necrotic wing membrane tissue next to a cupping erosion packed with P. destructans hyphae (periodic acidSchi stain). (b) M. dasycneme, sequestration of WNS lesion in skin by intense neutrophilic inltration (periodic acidSchi stain). (c) M. myotis, invasive skin infection by P. destructans hyphae and tissue necrotic derangement (transmission electron micrograph). Arrow: cupping erosion; lightning bolt: loss of staining pattern in necrotic tissue; checkmark: intact tissue; star: inammatory inltration with neutrophils sequestering the cupping erosion; sun: epidermal surface colonization by the fungus; block arrow: dermal invasion by the fungus; scissors: coarsely granular amorphous material and dermal edema, indistinct outlines of individual bat tissue cells, and derangement of elastic bers. Scale bar m.
8 Gradual decrease in mitochondrial uorescence intensity indicated the loss of membrane potential and, therefore, depletion of adenosine triphosphate in bat broblasts. Natural, invasive P. destructans infection of hibernating bats showed variability of histopathological ndings (Fig.10). The differential pathology likely represents a time series, with necrosis during hibernation torpor-arousal cycles and inammation in early post-hibernation period19,39. Hibernation affects both innate and adaptive immune responses to pathogens and increases host infection risk39,40. Evolutionarily conserved innate-like mucosal-associated invariant T cells use metabolites of the riboavin biosynthetic pathway to detect microbial infection41, produce pro-inammatory cytokines, and are cytotoxic42. e riboavin molecular signature of infection by P. destructans associated with extensive wing area damage may thus trigger the intense skin pathology (Fig.10) observed in bats during the early euthermic post-emergent season19. We conclude that with long duration of hibernation there is increased riboavin accumulation, driven by suitable nutritional sources from infected skin and hypoperfusion of host tissues during torpor. High riboavin concentrations damage skin, likely leading to an intensied arousal pattern, fat reserves depletion, and death. Photosensitizing properties of riboavin might further contribute to the intense skin pathology observed in bats emerging from hibernacula during daylight15,43. e nal disease outcome might be dependent on the total wing area damaged by P. destructans infection, which is larger in the Nearctic15,44 (Fig.1c), where bats with WNS experience mass mortality45, than in the Palearctic (Fig.1b), where they show tolerance to the disease24. Tissue damage from the hyperaccumulation of riboavin and its oxidation at reper fusion proposed here illuminates disease progression in the multi-stage WNS model21 with a novel molecular mechanism. Considering the pathological eects of high concentrations of riboavin within infected tissues, we suggest silencing or downregulation of genes from the endogenous biosynthetic pathway of riboavin as potential targets for disease management46,47.MethodsNonlethal sampling of wing membrane biopsies was performed in accordance with Czech Law No. 114/1992 on Nature and Landscape Protection, based on permits 01662/MK/2012S/00775/MK/2012, 866/JS/2012 and 00356/KK/2008/AOPK issued by the Agency for Nature Conservation and Landscape Protection of the Czech Republic. Experimental procedures were approved by the Ethical Committee of the Academy of Sciences of the Czech Republic (No. 169/2011). Sampling in Latvia was approved by the Nature Conservation Agency (Permit No. 3.15/146/2014-N) and in Russia by the Institute of Plant and Animal Ecology, Ural Division of the Russian Academy of Sciences (No. 16353/325). e authors are authorized to handle free-living bats according to Certicate of Competence No. CZ01341 (, Act No. 246/1992 Coll.) and a permit approved by the Latvian Nature Conservation Agency (No. 05/2014). Animals used in this study were sampled during 2014 and 2015 in accordance with the approved guidelines. To derive primary bat skin broblasts, a healthy bat (M. myotis) was captured during autumn swarming in the Czech Republic. Two 4-mm punch biopsies (Kruuse, Langeskov, Denmark) from each wing membrane were collected prior to releasing the bat. Samples for transmission or scanning electron (TEM or SEM) and confocal microscopy were cadavers without gross and histopathological signs of decomposition (e.g. epidermal/dermal separation and hair detachment, dermal degeneration and disintegration) and biopsies collected in hibernacula in the Czech Republic (M. myotis, n 3) and Latvia (M. dasycneme, n 1). We used samples where antemortem origin of fungal-induced tissue pathology was witnessed by neighbouring areas of normal skin histology within the same sample (cf. Fig.10a,b). e cadavers were processed for experiments directly aer transport to the laboratory (SEM and confocal microscopy), and UV uorescent spots were recorded and biopsied for further analyses (confocal microscopy, FTMS, MS/MS, histopathology, TEM). Biopsies were stored at 80 C (confocal microscopy, FTMS, MS/MS) or in liquid medium (histopathology, TEM). Live bats for fungal load quantication, urine collection and histopathology were handled so as to minimize stress and duration of sampling procedures between capture and release at the site. Bats right wing was extended on a sterilized glass slide over a UV lamp and two areas were biopsied: an area without discernible yellow-orange uorescence (designated as UV-negative throughout the paper) and an area with UV uorescence characteristic of WNS lesions (called UV-positive in the text). Biopsy punches were taken from bats in the Czech Republic and Latvia prior to their emergence24. Urine samples were collected with a pipette into amber light protection tubes from bats urinating during arousal initiated by handling.Six strains of Pseudogymnoascus destructans representing various countries of origin (USA, Czech Republic, Russia) and mating types (MAT1-1-1, MAT1-1-2) were selected to study fungal extracellular metabolites. Six additional strains of four other Pseudogymnoascus spp. nonpathogenic to bats but living in their hibernacula were used for comparison (Fig.3). All strains were characterized based on internal transcribed spacer (ITS) rDNA sequences and mating type idiomorphs according to Palmer et al.48. Pseudogymnoascus spp. isolates not belonging to P. destructans were compared to those published previously49, and their relatedness was demonstrated using a maximum likelihood phylogeny (Fig.3).Stock cultures of all monosporic strains were maintained on malt agar slants (malt extract 20 g l 1, agar 20 g l 1, pH adjusted to 6.5 by NaOH). Cultivation was performed on a glucose (20 g l 1) yeast (5 g l 1) extract liquid medium adjusted to pH 6.5. Stationary surface cultivations were carried out in 500 ml Erlenmeyer asks containing 50 ml of the medium for 3 months at 8 C in darkness. e bat wing membrane was extended over a Woods lamp (366 nm; BLAK-RAY Model UVL-56, San Gabriel, CA, USA) and photographed with a Nikon D80 digital SLR camera (in cave: ISO
9 1000, f-stop 18, shutter speed 1.6 s; in dark room: ISO 400, f-stop 4, shutter speed 0.25.4 s) mounted on a tripod. e camera lens was directed perpendicular to the wing. Whole-body imaging was performed on a M. myotis specimen using a Zeiss Axio Zoom.V16 motorized uorescent stereo microscope (Carl Zeiss Microscopy GmbH, Jena, Germany; UV, blue and green excitation light). Tissues from a dead animal with strong uorescence signal corresponding to the presence of WNS lesions were excised (namely the ying membrane and pelvic limb), xed (3.7% paraformaldehyde in PBS, 24 h, 4 C), dehydrated via ethanol solutions, transferred to acetone, and processed for SEM applications (Bal-Tec CPD 030 critical point drier and Bal-Tec SCD 050 sputter coater). Visualization was performed using a JEOL JSM-6390 LV scanning electron microscope (JEOL USA, Peabody, MA, USA).Wing punch biopsy samples were collected directly into tissue lysis buer with proteinase K (DNeasy Blood & Tissue Kit, Qiagen, Halden, Germany) and isolated within 10 h of sampling according to the manufacturers protocol. Fungal load in the samples was estimated using a qPCR method based on TaqMan chemistry (Life Technologies, Foster City, CA, USA) and using dual probes for P. destructans and Pseudogymnoascus sp. detection50.Fermentation broth of the strain P. destructans CCF3941 was centrifuged and extracted three times with an equal volume of CH2Cl2. Pooled extracts were dried over anhydrous Na2SO4, ltered, then evaporated to dryness under reduced pressure. e crude extract (23.1 mg) diluted in MeOH (0.2 ml) was loaded into a 20 g SPE C18 cartridge (Phenomenex, Chromservis, Prague, Czech Republic), rinsed with water, then eluted stepwise with H2O/MeOH gradient (20%, 30%, 40%, 50%, 80%, 100% [V/V]). e fractions were evaporated to dryness and reconstituted in MeOH. e content of the fractions was checked using HPLC. e fraction eluted with 80% MeOH contained a pure yellowish compound and the fraction with 50% MeOH contained a cinnamon-colored compound, which were further analyzed using UV and FTMS. Fermentation broths of all tested strains (2 ml) and riboavin standard solutions (Sigma, Steinheim, Germany) in water were loaded into a 0.5 g SPE C18 cartridge (Phenomenex, Chromservis, Prague, Czech Republic), conditioned with MeOH (2 ml), and equilibrated with H2O (2 ml). e cartridge was rinsed with water (5 ml) and eluted with MeOH (2 ml). Prior to HPLC analyses, the collected MeOH fractions were stored at 20 C in darkness. All experiments were done in triplicates under conditions of reduced light. e HPLC system consisted of a pump equipped with a 600E system controller, 717 autosampler, and 2487 dual UV detector (Waters, Milford, MA, USA). e data were processed using Empower 2 soware. Water containing mobile phases had been ltered through a 0.22 m GS lter (Millipore, Billerica, MA, USA) and degassed in an ultrasonic bath for 10 min before use. A Gemini 5 m C18 column (250 4.6 mm, Phenomenex, Torrance, CA, USA) with a guard column was used for the analysis. e mobile phase consisted of 5% MeOH in H2O and MeOH. Gradient elution started at 30% MeOH (0 min), increasing linearly to 100% MeOH within 20 min, at a ow rate of 1.0 ml/min. UV detection was performed at 260 and 350 nm. For calibration experiments and quantitative determination of riboavin, standard solutions of riboavin were prepared in H2O at nal concentrations of 1.25, 2.5, 5, 10, and 25 g ml 1 (injected in triplicate). e calibration graph was constructed by plotting the integrated peak areas of individual compounds versus concentration. e linear regression equation y 39081x with correlation (r) 0.999 and determination (r2) 0.999 coecients were obtained for the HPLC procedure. Samples of bat urine were diluted in water 20 times and injected directly. Riboavin and creatinine were quantied using equation y 43518x 3292 (r 0.999, r2 0.999) for riboavin and y 29803x 50742 (r 0.999, r2 0.999) for creatinine. Riboavin recovery was determined at three dierent concentrations (6.25, 12.5, and 25 g ml 1) and calculated by comparing the peak areas of the analytes in the extract from the spiked liquid cultivation medium and corresponding standard solutions in H2O (three replicates for each concentration). e recovery value of 98.54 2.09% was determined for all concentrations tested. Samples were measured on a commercial 12T solariX FTMS instrument (Bruker Daltonics, Billerica, MA, USA) equipped with an ESI/MALDI ion source and ParaCell. e analysis was per formed using electrospray ionization (ESI) and the spectra were acquired in positive ion mode. e cell was opened for 0.75 ms, accumulation time was set at 0.1 s for the MS experiment (0.6 s for the MS/MS experiment), and one experiment (consisting of the average of four spectra) was collected per sample. Aer the MS experiment, one MS/MS experiment was performed for the ion of interest. e isolation window was set at 3 a.m.u. and the collision energy was kept at 16V. e size of the acquisition data set was set to 2M points with the mass range starting at m/z 100 a.m.u., resulting in a resolution of 350,000 at m/z 400. One l of the sample was dissolved in MeOH/H2O (50%) and introduced into the mass spectrometer by direct infusion into the electrospray ion source. e instrument was externally calibrated using singly charged NaTFA clusters, resulting in sub-ppm accuracy. e spectra were apodized using sine apodization with one zero lling. Data were processed using Data Analysis 4.2 soware (Bruker Daltonics), and possible elemental compositions were calculated using Smart Formula calculator.
10 Flying membrane samples were analyzed using a Leica TCS SP2 laser scanning confocal microscope (Leica Microsystems GmbH, Wetzlar, Germany) with a 405 nm diode laser for excitation and a 10x/0.4 objective. Lambda scans were performed using a 20 nm spectral window shied continuously from 400 nm to 750 nm through the spectrum in 40 steps, thus yielding an 8.25 nm step. e obtained lambda stacks were analyzed with Leica LCS (Leica Microsystems GmbH) soware and FIJI51. e calibration curve of riboavin solutions (100, 80, and 50 g ml 1) was analyzed in a focused drop on a cover slip under the same conditions as those of the lambda scans of the ying membranes. Linear curve ttings were computed in FIJI through the highest spectral points, yielding the linear regression equation y 2.80777x 4.68651, r2 0.999. Spectral unmixing of the natural background and fungal lesion spot spectrum were done in LCS soware. e spectrally unmixed channel of the fungal lesion spot was then compared to the obtained calibration curve and the maximum and mean riboavin concentrations were estimated. ree-dimensional lambda stacks were used for 3D reconstruction. Primary bat skin broblasts were derived from the ying membrane of a M. myotis spec imen. Tissue was loosened mechanically with scissors without using proteases. Migratory adherent cells were cultivated for 1 month together with the remaining tissue sample. Aerwards, cells were passaged to conuency twice followed by cryopreservation of the stock in liquid nitrogen. Skin-derived cells were identied as broblasts based on their spindle shape combined with positive staining for the mesenchymal marker vimentin (EXBIO Praha, Vestec, Czech Republic) and the presence of typical stress-ber organization of the actin cytoskeleton. Cell culture was negative for mycoplasma infection (MycoAlert mycoplasma detection kit, Lonza, Walkersville, MD, USA). Cells for the experiments were grown on glass coverslips in 24 well-plates (Nunc, USA) for 48 h approximately to 50% conuency and passaged no more than 10 times. Treatment with concentrations of riboavin (dissolved directly in the cultivation medium followed by ltration with a 0.2 m lter; Nunc) ranging from 25 to 200 g ml 1 and cell culture medium control lasted for 24 h. Treated cells and controls were incubated at the end of the experiment for 20 min with MitoTracker Red CMXRos (Invitrogen, Carlsbad, CA, USA), xed (3.7% paraformaldehyde in PBS, 20 min, room temperature), permeabilized (0.1% Triton X-100 in PBS, room temperature, 3 min), blocked (1% BSA in PBS, room temperature, 20 min) and stained with Phalloidin-Alexa Fluor 488 conjugate (Invitrogen). Staining of nuclei was performed by mounting specimens in Mowiol-DAPI (Invitrogen). Visualization was done using an inverted Olympus IX51 microscope under a 40x objective. All procedures with living cells exposed to riboavin were performed in a dark room to prevent photooxidation damage.Quantication of cytotoxicity was based on lactate dehydrogenase activity released from damaged cells (Cytotoxicity Detection KitPLUS, Roche, Mannheim, Germany). Optimum cell concentration for the assay was determined in a preliminary experiment at 15,000 cells per well. e broblasts were grown in Dulbeccos modied Eagles high glucose medium with L-Glutamine (Biosera, Boussens, France) supplemented with 1% fetal bovine serum and grown at 37 C in 5% CO2 humidied environment. Riboavin solutions (200, 100, 50, 25, 12.5 and 0 g ml1 concentrations) were prepared as described above immediately prior to the exper iment, and cells were incubated in dark at 8 C or 37 C for 24 h. Cytotoxicity was estimated according to the manufacturer's recommendation and recorded using ELISA reader ELx808 (BioTek, VT, USA). Primary bat broblasts were grown in biological triplicates to 50% conuency in 24 well plates (Nunc), incubated with riboavin (2 h; 200, 100, 50, 25, 12.5 and 0 g ml 1 concentrations), then illuminated in the presence of ROS sensor CellROX Green Reagent (1 M; Invitrogen) in a botanical climabox (30 min, light intensity 110 mol of photons m 2 s 1). Cells were detached from cultivation wells with trypsin (0.1%, 200 l per well, Sigma-Aldrich, USA). Trypsinization was stopped with complete cultivation medium (200 l) and Hoechst 33258 solution was added (1 M, 10 min; Invitrogen). Intensity of CellROX Green Reagent uorescence was quantied in cell population gated as viable singlets by uorescence-activated cell sorter LSR II (Becton Dickinson, USA). Wing-membrane biopsies for standard histopathology and TEM guided by UV transillumination were xed in 10% neutral buered formalin and 2% glutaraldehyde, respectively. Formalin-xed samples were embedded in paran, cut into 5 m serial tissue sections and stained with periodic acidSchi stain. Glutaraldehyde-preserved biopsies were post-xed in 1% OsO4, dehydrated in acetone, then embedded in EponDurcupan mixture (Epon 812, Serva, Germany; Durcupan, ACM Fluka, Switzerland). e samples were then stained with 2% uranyl acetate and 2% lead citrate and observed at 80 kV under a Philips EM 208 TEM microscope (FEI, Czech Republic). WNS diagnosis followed guidelines issued by the USGS National Wildlife Health Center52. A case was positive for WNS when histologic lesions of bat WNS10,13,14 were present and P. destructans was detected.References1. Scharf, D. H., Heineamp, T. & Brahage, A. A. Human and plant fungal pathogens: the role of secondary metabolites. PLoS Pathog. 10, e1003859 (2014). 2. Bonomi, H. et al. An atypical riboavin pathway is essential for Brucella abortus virulence. PLoS ONE 5, e9435 (2010). 3. Garfoot, A. L., Zemsa, O. & appleye, C. A. Histoplasma capsulatum depends on de novo vitamin biosynthesis for intraphagosomal proliferation. Infect. Immun. 82, 393 (2014). 4. Zylberman, V. et al. Evolution of vitamin B2 biosynthesis: 6, 7-dimethyl-8-ribityllumazine synthases of Brucella J. Bacteriol. 188, 6135 (2006). 5. Massey, V. e chemical and biological versatility of riboavin. Biochem. Soc. Trans. 28, 283 (2000). 6. Wondra, G. T., Jacobson, M. & Jacobson, E. L. Endogenous UVA-photosensitizers: mediators of sin photodamage and novel targets for sin photoprotection. Photochem. Photobiol. Sci. 5, 215 (2006).
11 7. Mascuch, S. J. et al. Direct detection of fungal siderophores on bats with white-nose syndrome via uorescence microscopy-guided ambient ionization mass spectrometry. PLoS ONE 10, e0119668 (2015). 8. ODonoghue, A. J. et al. Destructin-1 is a collagen-degrading endopeptidase secreted by Pseudogymnoascus destructans, the causative agent of white-nose syndrome. Proc. Natl. Acad. Sci. 112, 7478 (2015). 9. Pannu, E. L., isch, T. S. & Savary, B. J. Isolation and Identication of an Extracellular Subtilisin-Lie Serine Protease Secreted by the Bat Pathogen Pseudogymnoascus destructans. PLoS ONE 10, e0120508 (2015). 10. Bandouchova, H. et al. Pseudogymnoascus destructans: evidence of virulent sin invasion for bats under natural conditions, Europe. Transbound. Emerg. Dis. 62, 1 (2015). 11. Blehert, D. S. et al. Bat White-Nose Syndrome: An Emerging Fungal Pathogen? Science 323, 227 (2009). 12. Lorch, J. M. et al. Experimental infection of bats with Geomyces destructans causes white-nose syndrome. Nature 480, 376 (2011). 13. Meteyer, C. U. et al. Histopathologic criteria to conrm white-nose syndrome in bats. J. Vet. Diagn. Invest. 21, 411 (2009). 14. Piula, J. et al. Histopathology Conrms White-Nose Syndrome in Bats in Europe. J. Wildl. Dis. 48, 207 (2012). 15. Turner, G. G. et al. Nonlethal screening of bat-wing sin with ultraviolet uorescence to detect lesions indicative of white-nose syndrome. J. Wildl. Dis. 50, 566 (2014). 16. Zual, J. et al. White-Nose Syndrome Fungus: A Generalist Pathogen of Hibernating Bats. PLoS ONE 9, e97224 (2014). 17. Fric, W. F. et al. Disease alters macroecological patterns of North American bats. Glob. Ecol. Biogeogr. 24, 741 (2015). 18. Cryan, P., Meteyer, C., Boyles, J. & Blehert, D. Wing pathology of white-nose syndrome in bats suggests life-threatening disruption of physiology. BMC Biol. 8, 135 (2010). 19. Meteyer, C. U., Barber, D. & Mandl, J. N. Pathology in euthermic bats with white nose syndrome suggests a natural manifestation of immune reconstitution inammatory syndrome. Virulence 3, 583 (2012). 20. eeder, D. M. et al. Frequent arousal from hibernation lined to severity of infection and mortality in bats with white-nose syndrome. PLoS ONE 7, e38920 (2012). 21. Verant, M. L. et al. White-nose syndrome initiates a cascade of physiologic disturbances in the hibernating bat host. BMC Physiol 14, 10 (2014). 22. Warnece, L. et al. Pathophysiology of white-nose syndrome in bats: a mechanistic model lining wing damage to mortality. Biol. Lett. 9, 20130177 (2013). 23. Xu, Y. et al. Molecular signatures of mammalian hibernation: comparisons with alternative phenotypes. BMC Genomics 14, 567 (2013). 24. Zual, J. et al. White-nose syndrome without borders: Pseudogymnoascus destructans infection tolerated in Europe and Palearctic Asia but not in North America. Sci. ep. 6, 19829 (2016). 25. Bhatta, H., Goldys, E. & Learmonth, Use of uorescence spectroscopy to dierentiate yeast and bacterial cells. Appl. Microbiol. Biotechnol. 71, 121 (2006). 26. Insisa-a, M., Golcza, A. & Siorsi, M. Photochemistry of riboavin derivatives in methanolic solutions. J. Phys. Chem. A 116, 1199 (2012). 27. Crous, P. W. et al. Fungal Planet description sheets: 371. Persoonia 35, 264 (2015). 28. Huba, V. et al. evision of Aspergillus section Flavipedes: seven new species and proposal of section Jani sect. nov. Mycologia 107, 169 (2015). 29. Garca-Fraile, P. et al. Serratia myotis sp. nov. and Serratia vespertilionis sp. nov., isolated from bats hibernating in caves. Int. J. Syst. Evol. Microbiol. 65, 90 (2015). 30. Novov, A., Savic, D. & ola, M. Two novel species of the genus Trichosporon isolated from a cave environment. Czech Mycol. 67, 233 (2015). 31. Lorch, J. M. et al. e fungus Trichophyton redellii sp. nov. causes sin infections that resemble white-nose syndrome of hibernating bats. J. Wildl. Dis. 51, 36 (2015). 32. Chaturvedi, V. et al. Morphological and molecular characterizations of psychrophilic fungus Geomyces destructans from New Yor Bats with white nose syndrome (WNS). PLoS ONE 5, e10783 (2010). 33. eynolds, H. T. & Barton, H. A. Comparison of the white-nose syndrome agent Pseudogymnoascus destructans to cave-dwelling relatives suggests reduced saprotrophic enzyme activity. PLoS ONE 9, e86437 (2014). 34. Schaible, U. E. & aufmann, S. H. Iron and microbial infection. Nat. ev. Microbiol. 2, 946 (2004). 35. Crossley, A. et al. iboavin biosynthesis is associated with assimilatory ferric reduction and iron acquisition by Campylobacter jejuni. Appl. Environ. Microbiol. 73, 7819 (2007). 36. He, X. et al. Establishment of Myotis myotis cell lines-model for investigation of host-pathogen interaction in a natural host for emerging viruses. PLoS ONE 9, e109795 (2014). 37. Moquin, D. & Chan, F. .-M. e molecular regulation of programmed necrotic cell injury. Trends Biochem. Sci. 35, 434 (2010). 38. Susin, S. A. et al. Molecular characterization of mitochondrial apoptosis-inducing factor. Nature 397, 441 (1999). 39. Field, A. et al. e white-nose syndrome transcriptome: activation of anti-fungal host responses in wing tissue of hibernating little brown myotis. PLoS Pathog. 11, e1005168 (2015). 40. Bouma, H. ., Carey, H. V. & roese, F. G. M. Hibernation: the immune system at rest? J. Leuocyte Biol. 88, 619 (2010). 41. jer-Nielsen, L. et al. M presents microbial vitamin B metabolites to MAIT cells. Nature 491, 717 (2012). 42. Gold, M. C. & Lewinsohn, D. M. Co-dependents: M-restricted MAIT cells and their antimicrobial function. Nat. ev. Microbiol. 11, 14 (2013). 43. Turner, G. G., eeder, D. M. & Coleman, J. T. H. A ve-year assessment of mortality and geographic spread of white-nose syndrome in North American bats and a loo to the future. Bat esearch News 52, 13 (2011). 44. eichard, J. D. & unz, T. H. White-nose syndrome inicts lasting injuries to the wings of little brown myotis (Myotis lucifugus ). Acta Chiropter. 11, 457 (2009). 45. Fric, W. F. et al. An emerging disease causes regional population collapse of a common North American bat species. Science 329, 679 (2010). 46. Ndvig, C. S. et al. A CISP-Cas9 System for genetic engineering of lamentous fungi. PLoS ONE 10, e0133085 (2015). 47. Zhang, T. et al. Development of an Agrobacterium -mediated transformation system for the cold-adapted fungi Pseudogymnoascus destructans and P. pannorum Fungal Genet. Biol. 81, 73 (2015). 48. Palmer, J. M. et al. Molecular characterization of a heterothallic mating system in Pseudogymnoascus destructans, the fungus causing white-nose syndrome of bats. G3-Genes Genom. Genet. 4, 1755 (2014). 49. Minnis, A. M. & Lindner, D. L. Phylogenetic evaluation of Geomyces and allies reveals no close relatives of Pseudogymnoascus destructans comb. nov., inbat hibernacula of eastern North America. Fungal Biol 117, 638 (2013). 50. Shuey, M. M., Drees, P., Lindner, D. L., eim, P. & Foster, J. T. Highly sensitive quantitative PC for the detection and dierentiation of Pseudogymnoascus destructans and other Pseudogymnoascus species. Appl. Environ. Microb. 80, 1726 (2014). 51. Schindelin, J. et al. Fiji: an open-source platform for biological-image analysis. Nat. Methods 9, 676 (2012). 52. USGS. Diagnostic categories for reporting cases of bat white-nose syndrome (WNS) summary of revisions to WNS case denitions for the 2014/2015 season. https://www.nwhc.usgs.gov/disease_information/white-nose_syndrome/Case%20Dentions%20for%20 WNS.pdf, accessed on 4 August, 2016.
12 53. Gargas, A., Trest, M. T., Christensen, M., Vol, T. J. & Blehert, D. S. Geomyces destructans sp. nov. associated with bat white-nose syndrome. Mycotaxon 108, 147 (2009). 54. Martnov, N. et al. Increasing incidence of Geomyces destructans fungus in bats from the Czech epublic and Slovaia. PLoS ONE 5, e13853 (2010). 55. atoh, & Standley, D. M. MAFFT multiple sequence alignment soware version 7: Improvements in performance and usability. Mol. Biol. Evol. 30, 772 (2013). 56. Guindon, S. & Gascuel, O. A simple, fast, and accurate algorithm to estimate large phylogenies by maximum lielihood. Syst. Biol. 52, 696 (2003). 57. eane, T. M., Creevey, C. J., Pentony, M. M., Naughton, T. J. & McInerney, J. O. Assessment of methods for amino acid matrix selection and their use on empirical data shows that ad hoc assumptions for choice of matrix are not justied. BMC Evol. Biol. 6, 29 (2006).AcknowledgementsWe thank Tom Bartonika, Matthew C. Fisher, Valika Grobrov, Xiaocui He, Ivan Horek, Kamil S. Jaron, Petr Juraka, Bernd Kllner, Pavel Kulich, Radek K. Luan, Alexandra Zahradnkov, Jr., Alexandra Zahradnkov and Jan Zukal for their help. is study was supported by the Czech Science Foundation (P506-12-1064) and the Internal Grant Agency of the University of Veterinary and Pharmaceutical Sciences Brno (229/2015/FVHE).Author ContributionsM.F., J.C., M.K. and J.P. designed the study; M.F., H.B., J.C., M.C., M.K., V.K., P.N., O.., E.S. and J.P. performed the experiments; M.F., M.K., N.M. and J.P. analyzed the data; N.M. and J.P. wrote the paper with contributions from all authors.Supplementary information accompanies this paper at http://www.nature.com/srep Competing nancial interests: e authors declare no competing nancial interests. How to cite this article: Flieger, M. et al. Vitamin B2 as a virulence factor in Pseudogymnoascus destructans skin infection. Sci. Rep. 6, 33200; doi: 10.1038/srep33200 (2016). is work is licensed under a Creative Commons Attribution 4.0 International License. e images or other third party material in this article are included in the articles Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ e Author(s) 2016