REPEATED SAMPLING OF WILD INDIVIDUALS REVEALS OPHIDIOMYCES OPHIDIICOLA INFECTION DYNAMICS IN A PENNSYLVANIA SNAKE ASSEMBLAGE

Ophidiomyces ophidiicola (Oo) is the causative agent of ophidiomycosis (formerly known as snake fungal disease) and is an emerging fungal infection in snakes (Allender et al. 2011; Lorch et al. 2016), that was first described by Rajeev et al. (2009) as Chrysosporium ophiodiicola. In 2013, it was determined to be a unique genus that was only found in snakes (Sigler et al. 2013; Kirk 2020). Ophidiomyces ophidiicola has since been found in many snake species in North America, Europe, and Australia (Sigler et al. 2013; Allender et al. 2015; Lorch et al. 2016; Franklinos et al. 2017; Haynes et al. 2020; Davy et al. 2021). In North America, fungal skin infections have been associated with population declines in snake species such as the timber rattlesnake, Crotalus horridus (Clark et al. 2011; McBride et al. 2015; Bittel 2021).

Skin lesions caused by Oo infections are typically conserved across species and may include, but are not limited to, scale flaking and crusting, displaced and/or discolored scales, increased rapidity in shedding cycles, and swelling and disfiguration of affected tissue (Baker at al. 2019). Infection with Oo and resultant ophidiomycosis are not always fatal. There is evidence that some snakes can clear the clinical signs of Oo infection (Lind et al. 2018), but this finding does not indicate the actual clearing of the infection, nor has it been shown how Oo DNA detection relates to changes in clinical signs over time. Additionally, Oo can be transmitted vertically and cause mortality in offspring (Stengle et al. 2019). The ability of Oo to be transmitted horizontally (McKenzie et al. 2020; Durante et al. 2021) and vertically (Stengle et al. 2019) makes it a pathogen that may drive its hosts to extirpation with relative ease (de Castro and Bolker 2005).

In the Northeastern US, Oo has been detected in multiple snake species (Allender et al. 2016; Guthrie et al. 2016; Ravesi et al. 2016; Licitra et al. 2019) and in Pennsylvania, US, it is known to affect two pitviper species: Sistrurus catenatus (Regester et al. 2017) and C. horridus (Januszkiewicz et al. 2019). It was also recently detected from wild snakes at Fort Indiantown Gap, a military installation in Lebanon County, Pennsylvania, US, but the infected species were not disclosed (Allender et al. 2020). In addition to infections in wild snakes, Oo has also been found in captive animals in the state (Ohkura et al. 2016). We used a scoring system for suspected ophidiomycosis and employed two DNA-based detection methods to determine the prevalence of Oo infection in five species of a snake assemblage subject to long-term ecological study in southwestern Pennsylvania.

Powdermill Nature Reserve (PNR) is owned by the Carnegie Museum of Natural History. The 856.7-ha site is located in the Laurel Ridge of the Allegheny Mountains in Westmoreland County, Pennsylvania (Meshaka et al. 2008). The property is comprised mainly of mixed forest (89.5%), with primary and secondary grasslands now making up less than 5% of the habitats (Meshaka 2010). There are 13 species of snakes reported from PNR, which have been monitored by approximately monthly checks of artificial cover boards during the active season (May–September) using mark-recapture methods with passive integrated transponders (PIT tags) since 2002 (Meshaka 2010), and the population sizes of some of them have experienced declines in recent years (W.E.M. data not shown).

All free-ranging snakes encountered during mark-recapture surveys at PNR in the spring and summer of 2020 were scored for the presence of clinical signs of ophidiomycosis (Fig. 1 and Table 1). The scoring system was adapted from McCoy et al. (2017). All captured snakes were swabbed twice. Depending on the presence or absence of clinical signs of disease, one of two approaches was taken. If the snake did not show lesions, the snake was swabbed between the neck and vent five times, with the swab being rotated on each stroke, with two separate swabs that had been wetted with sterile water. If the snake showed skin lesions, the lesions were swabbed five times while rotating the swab. If there were multiple skin lesions, each was swabbed five times with the same swab, using the same method. The swab applicators were then broken off with gloved hands (without touching the tip) and the tips placed into labeled 1.5-mL centrifuge tubes for storage. New gloves were used for handling every new snake. Swabs were then stored in a freezer at –20 C until DNA extraction for either traditional PCR or quantitative (q)PCR.

For each sample, DNA was extracted from one swab using DNA Mini Kits (Qiagen, Hilden, Germany) as described by Allender et al. (2015). The extracted DNA was screened for the presence of Oo using traditional PCR with the following primers from Allender et al. (2015): ChrysoITS-F 5′-TGATCCGAGGTCAACCG-G A A G AAA-3′ and ChrysoITS - R5′-TGGAACCGTCAACGAACTCTGTGA-3′. The traditional PCR program was as follows: 94 C for 5 min followed by 40 cycles of (94 C for 30 s; 59 C for 30 s; 72 C for 30 s) then 72 C for 4 min and then held at 4 C until removed for use (adapted from Allender et al. 2015). The ChrysoITS primers amplify a gene fragment from the internal transcriber spacer 1 (ITS1) region that spans between the 18S and 5.8S rRNA and is approximately 400 base pairs (bp) in length (as per Allender et al. 2015). All PCRs were run twice for each sample, and the PCR products were run out on a 1% agarose gel for measurement of fragment size. Samples where both PCRs each had bands at the 400-bp level were considered to be positive for Oo DNA for that sample. Each PCR included a negative control and a positive control for Oo DNA—a cloned segment of the ITS1 region developed by Allender et al. (2015). In cases where one of the PCRs had a band indicating positive and the other did not, the sample was rerun in duplicate. If the second PCR produced duplicate 400-bp bands, the sample was considered positive for Oo.

For the second set of swabs, DNA extraction and qPCR amplification were performed in the Wildlife Epidemiology Laboratory at the University of Illinois, College of Veterinary Medicine (Urbana, Illinois, USA) as described (Allender et al. 2015). Briefly, qPCR reactions targeting a 68-bp segment of the ITS1 region were run in triplicate on a QuantStudio 3 real-time thermocycler (Thermo Fisher Scientific, Waltham, Massachusetts, USA). The quantity of DNA (nanograms per microliter) and quality (absorbance at 260:280 nm) were determined by spectrophotometry. We considered samples positive if all replicates had a lower mean cycle threshold value than the lowest detected standard dilution. Mean fungal quantities (copies per reaction) were standardized to the total quantity of DNA (copies per nanogram of DNA). Copies per reaction were standardized to the total quantity of DNA in the sample by dividing the mean number of gene copies/microliter for each sample by the DNA concentration.

We tested 34 samples from five species for Oo DNA in this study: northern ring-necked snake, Diadophis punctatus edwardsii (n=1), eastern milksnake, Lampropeltis triangulum (n=15), northern watersnake, Nerodia sipedon sipedon (n=2), eastern ratsnake, Pantherophis alleghaniensis (n=4), eastern gartersnake, Thamnophis sirtalis sirtalis (n=12; Table 2). Of these, 19 snakes exhibited no clinical signs of ophidiomycosis and 16 were recaptures from previous years' monitoring efforts. We found skin lesions in 19 snakes (44%), but traditional PCR detected Oo DNA in only two snakes, both L. triangulum; one of these did not present with skin lesions. When qPCR analyses were performed, 19 samples from two species (L. triangulum and T. s. sirtalis) were positive for Oo DNA (Table 2): 80% of L. triangulum were positive with mean fungal quantities ranging from 0.42 to 6,836.11 gene copies/ng of total DNA, and 50% of T. s. sirtalis were positive with mean fungal quantities ranging from 0.17 to 2,966.89 gene copies/ng of total DNA. Six of the L. triangulum samples that were positive for Oo DNA using qPCR analysis did not exhibit clinical signs of ophydiomycosis, whereas only one T. s. sirtalis that did not show clinical signs of ophydiomycosis was positive. Both samples from L. triangulum that tested positive using traditional PCR also tested positive using qPCR, and they exhibited two of the highest fungal quantities among all samples (2,062.11 and 6,836.11 gene copies/ng of total DNA).

Two individuals of L. triangulum were sampled for O. ophidiicola at multiple times throughout the season: a female (Snake ID: LT-774) was sampled three times (once each in May, June, and July) and a male (Snake ID: LT-062) two times (once in May and June). Both snakes exhibited minor signs of ophidiomycosis in May (field scores of 1) and severe clinical signs in June (field scores of 3), but neither tested positive for Oo DNA at any sampling interval using traditional PCR. The female L. triangulum later in July exhibited no clinical signs of ophidiomycosis, despite visible skin lesions in May and more-extensive lesions in June. This female tested negative using qPCR in May and tested positive for Oo DNA using qPCR in both June (3,402.99 gene copies/ng of total DNA) and July (71.22 gene copies/ng of total DNA). The male tested positive for Oo DNA using qPCR in both May (30.9 gene copies/ng of total DNA) and June (1,301.2 gene copies/ng of total DNA). From six instances where multiple snakes were sampled at the same time from under the same cover board, we did not detect any discordance across test results for Oo DNA (PCR and qPCR), such that either all snakes found under the same board tested positive or all tested negative (Table 2). Five of these instances were of the same species sharing the same board, but on one occasion a L. triangulum was found under a board with a T. s. sirtalis and both snakes tested positive for Oo DNA.

All five species of snakes examined in our study are known to harbor O. ophidiicola (Oo) infections in the wild across the Nearctic region (Fuchs et al. 2020; Davy et al. 2021). Therefore, it was unsurprising that we found clinical signs (skin lesions) of ophydiomycosis in 44% of our samples; however, the disparity between the number of positive traditional PCR results and the number of positive qPCR results (two versus 19) was quite unexpected, especially given that Allender et al. (2015) found a much smaller discrepancy between these methods (three positive snakes with traditional PCR compared to seven positives with qPCR). While our traditional PCR results agreed with the qPCR results, we found that the detection of Oo DNA could be nearly 10× greater when using qPCR compared to traditional PCR (6% prevalence for traditional PCR and 56% prevalence for qPCR). Furthermore, the traditional PCR method frequently returned negative results even when snakes exhibited a clinical score of 3 (Table 2). This result suggests that extreme caution is warranted when interpreting the results of studies that only use traditional PCR, as it likely significantly underestimates the prevalence of Oo DNA present in samples. However, it is unclear the extent to which qPCR may overestimate the presence of Oo DNA or return false positives. Seven samples that did not show signs of ophydiomycosis (field score of 0) screened positive for the presence of Oo DNA using qPCR, whereas only one snake without signs of disease tested positive using traditional PCR. Interestingly, three samples exhibiting minor skin lesions (field score of 1) tested negative for Oo DNA with both qPCR and traditional PCR, suggesting that visible signs of disease, probably via healing lesions, may persist although snakes have nondetectable levels of Oo DNA.

The widespread nature of ophydiomycosis signs and possible infections at PNR is of concern. All species, except for D. p. edwardsii (only one individual sampled), had at least one individual that showed clinical signs of ophydiomycosis according to our classification system (Tables 1, 2). Frequency and severity of disease varied between the two species with the highest numbers of individuals that tested positive for the presence of Oo DNA. In T. s. sirtalis, whose incidence of infection was 50% with qPCR, no snakes exhibited field scores of more than 1, with slightly more than half with a score of 0. On the other hand, L. triangulum appeared to be particularly susceptible to Oo infection and ophidiomycosis at PNR, with 53% of our samples exhibiting skin lesions, 86% testing positive with Oo DNA (qPCR and PCR), and this was the only species to have individuals with clinical scores of 3 for signs of ophydiomycosis (Fig. 1). Repeated testing through mark-recapture techniques was particularly informative, as it provided insight about seasonal variation in disease dynamics, particularly in the female L. triangulum that progressed from 1 to 3 on our clinical signs scale from May to June but was free of clinical signs in July, while the qPCR changed from negative in May to positive in June (3,402.99 gene copies/ng of total DNA) and still remained qPCR-positive in July, if at a lower concentration (71.22 gene copies/ng). This suggests a lag time in which clinical signs of Oo infections may precede DNA detection and that the individual may then clear the clinical signs before the presence of Oo DNA is no longer detectable. A small sample size precludes us from drawing a wide-reaching conclusion about disease dynamics, but the results from the two resampled individuals suggest that apparent infection and subsequent recovery over 3 mo in the summer from putative disease may be very important in maintaining perennial presence of Oo in individuals and perhaps in the snake assemblage.

With the apparent decline of several snake species at PNR (W.E.M. data not shown), it is important to understand the true extent of Oo infection in case it may be contributing to or is the cause of these declines. In addition to the value of repeated sampling of individuals, our finding that snakes from the same cover board exhibited the same qPCR test results (all positive or all negative) suggests that communally used retreats may play a role in the spread of Oo because seasonal aggregations under cover, especially during courtship, could increase the opportunity for transmission within and among species. Alternatively, these results could suggest similar behavioral responses to infection across species, as at least one instance occurred where two species (L. triangulum and T. s. sirtalis) that were found under the same cover both tested positive. Our results further demonstrated the striking difference in sensitivity of detecting Oo DNA between two screening techniques (qPCR versus PCR) and indicate that caution is warranted when interpreting the prevalence of Oo from traditional PCR analyses. Sampling of historical museum specimens from this assemblage at PNR might help determine the starting point in the timeline of this epizootic outbreak. Determining the prevalence of Oo in museum specimens and continued monitoring will be necessary to evaluate the response of PNR's snake assemblage, which occupies a fully protected environment, to the long-term exposure of a readily transmissible and potentially lethal fungal pathogen.

We would like to thank M. C. Allender and the Wildlife Epidemiology Lab at the University of Illinois for providing us with swab extraction methods and the positive PCR control, and for performing the qPCR analysis. Funding was provided by startup funds from Coe College to D.F.H., Powdermill Nature Reserve to A.K., and the Health of Herpetofauna Research Group led by A.L.J.D.