Galleria mellonella experimental model for bat fungal pathogen Pseudogymnoascus destructans and human fungal pathogen Pseudogymnoascus pannorum

ABSTRACT Laboratory investigations of the pathogenesis of Pseudogymnoascus destructans, the fungal causal agent of bat White Nose Syndrome (WNS), presents unique challenges due to its growth requirements (4°-15°C) and a lack of infectivity in the current disease models. Pseudogymnoascus pannorum is the nearest fungal relative of P. destructans with wider psychrophilic – physiological growth range, and ability to cause rare skin infections in humans. Our broad objectives are to create the molecular toolkit for comparative study of P. destructans and P. pannorum pathogenesis. Towards these goals, we report the successful development of an invertebrate model in the greater wax moth Galleria mellonella. Both P. destructans and P. pannorum caused fatal disease in G. mellonella and elicited immune responses and histopathological changes consistent with the experimental disease.


Introduction
White-nose syndrome (WNS) is an epizootic disease responsible for the deaths of more than 5 million bats in North America since its first discovery in 2006 [1,2]. The latest surveillance and modeling data suggest that WNS is likely to lead to the regional extinction of some bat species in the United States while the disease remains well-tolerated among the European bats [3,4]. Pseudogymnoascus destructans is the fungal causal agent of WNS [5]. Pseudogymnoascus destructans is a psychrophile, well-adapted to growing at 4°-15°C, and widely distributed in the caves and mines that serve as bat hibernacula [5][6][7][8]. Bats infected with P. destructans exhibit histopathological lesions on wings and other body parts, are frequently aroused from hibernationinduced torpor, and possibly succumb to WNS due to fat depletion and starvation [9][10][11]. Some North American bat species and bats elsewhere in the world do not suffer from WNS-associated fatality likely due to a protective immune response against P. destructans [4,[12][13][14] P. destructans is unique among known human and animal fungal pathogens for the manifestation of its virulence attributes at temperatures lower than the expected physiological range, which is a common trait of pathogenic fungi as detailed by Kohler et al. [15]. P. pannorum is the nearest fungal relative of P. destructans, which grows over psychrophilicphysiological temperatures, and causes rare skin infections in humans [5,[16][17][18]. We have devoted efforts to create the molecular toolkit for the comparative studies of P. destructans and P. pannorum as there is no framework to investigate fungal pathogenesis at low temperatures. Towards this end, we earlier reported a comparison of the draft genomes of the two pathogens and created a transformation system for targeted gene disruption and live-cell imaging [19][20][21] . In the current report, an in vivo model is described using the wax moth Galleria mellonella to facilitate comparative pathogenesis studies. Both P. destructans and P. pannorum caused fatal disease in G. mellonella and elicited immune responses and histopathological changes that are the hallmark of experimental disease in this model system.

G. mellonella survival studies
Inocula were prepared by collecting P. destructans and P. pannorum conidia from cultures grown in PDA agar at 15°C for one-week, harvested by gently scraping the fungal growth, suspending in phosphate buffered saline (PBS) and passing the suspension through a 27G needle to separate conidia. Conidia were washed twice with PBS by centrifugation and counted with a hemocytometer to determine the fungal cell concentration, and plated on culture plates to determine colony forming units (CFU). G. mellonella larvae (Vanderhorst Wholesale, St. Marys, OH) at their final instar stage were inoculated with 5 × 10 5 or 1 × 10 6 CFU of P. destructans or P. pannorum suspended in PBS. Each infection group contained 16 randomly chosen larvae of the appropriate weight (330 ± 25 mg). The inoculum was injected in a 10 μl volume directly to the last left pro-leg using a Hamilton syringe [22]. Larvae were incubated at 5°C and 15°C, and the number of dead larvae was scored daily. Killing curves were plotted, and statistical analysis was performed by the Kaplan-Meier method using STATA 6 statistical software (Stata). Killing curves were performed in duplicate, and representative graphs were reported.

G. mellonella hemocyte density
In an experiment subsequent to survival studies, the larvae were injected with the fungi at concentrations of 5 × 10 5 CFU/larvae or 1 × 10 6 CFU/larvae at the last left pro-leg then incubated at 15°C. Hemocytes were collected from the hemocoel at 24 h post-injection. Larvae were punctured with a number 11 blade scalpel and bled into tubes containing cold, sterile insect physiologic saline (IPS) (150 mM sodium chloride; 5 mM potassium chloride; 100 mM Tris-hydrochloride, pH 6.9 with 10 mM EDTA, and 30 mM sodium citrate), pooling four larvae for each group. The hemocytes were enumerated with the aid of a hemocytometer. However, we did not differentiate between the six types of hemocytes (prohemocytes, coagulocytes, spherulocytes, oenocytoids, plasmatocytes, and granulocytes). Results were averaged from four replicates, and a Student's t-test was used to compare groups, P < 0.05 was considered significant.

Histopathology
Infected larvae were fixed in buffered formalin at 4°C for 48-hours, and embedded in paraffin. Paraffin blocks were sectioned and stained with H&E and Periodic acid-Schiff (PAS). Blind histopathological evaluation was performed on 8-9 full-length sagittal sections of larvae in each group.

Results and discussion
Pseudogymnoascus destructans and P. pannorum caused temperature and dose-dependent mortality in G.
mellonella. P. destructansand P. pannorum-injected G. mellonella suffered significant mortality at 15°C compared to larvae injected with PBS, based on assessment with a   Kaplan-Meier survival plot (P < 0.001). This mortality pattern was consistent with 5 × 10 5 CFU and 1 × 10 6 CFU within the 144-hour evaluation period (Figures 1-2). Mortality of 50% was reached as early as 75-hours postinoculation for the infected groups. At 5°C, there was no impact on the larvae survival by different dosage of inoculum of both pathogens (data not shown). Upon visual inspection of larvae, we observed melanization induced within larvae infected with either Pseudogymnoascus species (Figure 3), suggesting recognition of the pathogen by the host immune system. The larvae exhibited melanization at the injection site; the pigmentation spread throughout the body, and it was retained during the infection until death. Infection with P. destructans and P. pannorum resulted in reduced hemocyte density within the hemolymph compared to larvae injected with PBS ( Figure 4) (P < 0.01). The reduction in hemocyte density was significant with both 5 × 10 5 CFU (P < 0.05) and 1 × 10 6 CFU  (P < 0.05) of P. destructans. A significant reduction in hemocyte density was also observed with 5 × 10 5 CFU (P < 0.05) of P. pannorum while injection of higher dosage (1x10 6 CFU) did not result in significant reduction (P > 0.05).
The histopathological examination of the hemocoel of uninfected sham inoculated (PBS) larvae showed normal organelles and fat bodies, and lack of melanization in the hemocoel cavity ( Figure 5). All larvae inoculated with P. detructans or P. pannorum showed evidence of fungal dissemination throughout the hemocoel. Variable numbers and sizes of melanized spots were observed, which represented immune-competent hemocytes surrounding entrapped fungal spores or hyphae. The larvae inoculated with either fungus and housed at 5°C, had numerous clusters of hemocytes admixed with hemolymph, melanin, conidia, and hyphae throughout the hemocoel (Figures 6 -7). These clusters were less organized in the larva inoculated with P. destructans and mostly expanding in the fat body and coelomic space ( Figure 6). In the P. pannorum inoculated larvae, there were numerous loosely and well-defined hemocyte nodules throughout the hemocoel (Figure 7). In both groups, the hemocyte nodules consisted of conidia and hyphae surrounded by prominent melanin deposition and scattered hemocytes (Figures 6-7). The histopathological changes were more severe in the larvae inoculated with either P. destructans or P. pannorum and housed at 15°C (Figures 8 -9). Additionally, there were subtle differences in the severity of the immune response and dissemination. The P. destructans inoculated larvae had numerous granuloma-like hemocyte nodules throughout the hemocoel, fat body, and expanding within the wall of organelles containing round-to-spindloid hemocytes encapsulating fungal elements and abundant melanin (Figure 8). Similar immune response and distribution were observed in the larva inoculated with P. pannorum with wider dissemination of the fungal elements into the body wall and cuticle (Figure 9).
Our novel findings establish a tractable experimental system to study host-pathogen interactions of psychrophilic fungi P. destructans and P. pannorum. Many earlier publications have described the wax moth model as a preferred in vivo system for the study of human, insect, and fish pathogenic fungi [23][24][25][26][27]. Our findings are novel as we took advantage of the G. mellonella ability to survive in a wide range of temperature to conduct the experiments at low temperature. In contrast, other studies maintained the infected G. mellonella larvae at temperatures ranging from 25°to 37°C. The present study provided proof that the G. mellonella experimental system is amenable to an extended temperature range from 5°C to 37°C. It is well-known that temperature affects G. mellonella immune response to pathogens; however, the relevant previous studies focused on the effects of pre-incubation periods as they related to the immune responses [28,29]. Importantly, the current study differs from earlier publication in that the larvae were maintained at room temperature until infection and were not placed at 5°C and 15°C until after infection. It is likely that in this case, the lower temperature was conducive both for the growth of the fungal pathogens and G. mellonella. The fact that robust immune response is invoked in G. mellonella kept at 4°C might have contributed to the lack of P. destructansand P. pannorum-induced mortality in this study [27,29] The decrease in hemocyte density after P. destructans and P. pannorum infections also held true at low temperatures as reported for the higher temperatures in other studies. The hemocyte decline fulfilled an additional correlation of experimental disease as postulated by Bergin and colleagues [30]. The decline in hemocyte count was not dose-dependent likely because the experiments were performed with different batches of larvae. The observed melanin in the infected larvae suggested that the insect host mounted a robust albeit ineffective immune response in the form of phenoloxidase activation and resulting melanin production [24].
Galleria mellonella models have become invaluable to study cellular and humoral immune responses against pathogens [31][32][33], epigenetic influences on the evolution of virulence [32,34,35], and for development and testing antimicrobials [36][37][38][39]. It is relevant to add that the complete genomes of P. destructans and G. mellonella have become available recently [40,41]. Thus, G. mellonella -Pseudogymnoascus pathosystem is now ready as a tractable model for the fungal gene-phenotype studies as well as for the high-throughput screens in search for effective antifungal agents.

Acknowledgments
We thank Helen Johnson of the Wadsworth Center Histopathology Core for help with worm sections. We thank the three anonymous reviewers for various suggestions for the improvements in the original manuscript.

Disclosure statement
No potential conflict of interest was reported by the authors.