Expanded host and geographic range of tadpole associations with the Severe Perkinsea Infection group

Severe Perkinsea infection is an emerging disease of amphibians, specifically tadpoles. Disease presentation correlates with liver infections of a subclade of Perkinsea (Alveolata) protists, named Pathogenic Perkinsea Clade (PPC). Tadpole mortality events associated with PPC infections have been reported across North America, from Alaska to Florida. Here, we investigate the geographic and host range of PPC associations in seemingly healthy tadpoles sampled from Panama, a biogeographic provenance critically affected by amphibian decline. To complement this work, we also investigate a mortality event among Hyla arborea tadpoles in captive-bred UK specimens. PPC SSU rDNA was detected in 10 of 81 Panama tadpoles tested, and H. arborea tadpoles from the UK. Phylogenies of the Perkinsea SSU rDNA sequences demonstrate they are highly similar to PPC sequences sampled from mortality events in the USA, and phylogenetic analysis of tadpole mitochondrial SSU rDNA demonstrates, for the first time, PPC associations in diverse hylids. These data provide further understanding of the biogeography and host range of this putative pathogenic group, factors likely to be important for conservation planning.


Introduction
Severe Perkinsea infection (SPI) has been associated with tadpole mass mortality events (MMEs) in the USA [1] and likely represents the third most common infectious disease of amphibians in North America [1]. The disease pathology has been associated with a specific group of Perkinsea protists (syn. Perkinsids, Perkinsozoa) called Pathogenic Perkinsea Clade (PPC) based on small-subunit ribosomal DNA (SSU rDNA) sequencing [1,2]. This group is part of a wider clade of the freshwater Perkinsea named Novel-Alveolate-Group-01 (NAG01) [3]. Tadpole associations with specific clades of NAG01-Perkinsea have been detected across multiple continents [3], yet the PPC clade associated with SPI has only been detected in the USA, and mainly from SPI-diseased tadpoles [1,2]. Formal identification of the disease relationship satisfying Koch's postulates is absent [4]. Yet, in localities with SPI outbreaks, one study demonstrated that 100% of tadpoles exhibiting SPI were PPC-positive, whereas a cohort of asymptomatic tadpoles sampled nearby demonstrated only 2.5% (2/81) PPC prevalence [1]. This pattern provides circumstantial evidence for a relationship between the disease (SPI) and the protist infection (PPC) [5,6].
The first documented MMEs associated with SPI were recorded in New Hampshire in 1999 [7] and, since then, SPI-associated mortality events have been reported in seven other states across a broad geographic range [1,2,7,8]. Most of these SPI-associated MMEs have occurred in Lithobates spp. tadpoles [1,2,8,9], but have also been detected in larval hylids (Hylidae), including Pseudacris crucifer (spring peeper) and Acris gryllus (southern cricket frog) [1]. Here, we investigate the wider geographic and host range of PPC by screening tadpoles from Panama and a candidate SPI outbreak in UK captive-bred specimens. Using SSU rDNA sequencing, we demonstrate a hitherto unsampled biogeographic and host-associated taxonomic provenance of this putative pathogen, suggesting that PPC may be more biogeographically widespread than previously assumed, a finding that is crucial in managing animal trade and directing future conservation efforts.

Material and methods (a) Sample collection and preparation
Fieldwork was conducted in Panama (November 2018). Eightyone tadpoles, without apparent pathologies and with unknown species identification, were collected from seven sites (electronic supplementary material, tables S1 and S2). New dissection consumables were used for processing each distinct sample set and flame-sterilized between processing each individual tadpole; each tadpole was individually washed in H 2 O, euthanized using MS-222 and washed in physiological saline before dissection. Liver tissues were stored in LifeGuard™ (Qiagen) at −20°C for 1-2 weeks and then at −80°C. All livers appeared normal upon gross visual inspection. Due to permit regulations, all DNA samples were extracted in Panama.
Captive-bred Hyla arborea tadpoles were collected from a single-aquarium MME (AME1) in Surrey, UK, in July 2019 (electronic supplementary material, table S1) and stored in LifeGuard™ at −80°C. All the livers from tadpoles collected in the UK showed signs of SPI, specifically enlargement and yellow discoloration (identified using a dissecting microscope at 10× magnification). Subsamples of UK tadpoles were preserved for histopathological microscopy (electronic supplementary material, figure S1). Due to the small size of these tadpoles, two groups of five dissected livers from the same MME were pooled for DNA analysis.
DNA was extracted using the DNeasy Blood and Tissue Kit (Qiagen) with an overnight lysis at 37°C and disruption with 425-600 µm acid-washed glass beads (Sigma-Aldrich) on a FastPrep-24™ tissue homogenizer (MP Biomedicals). Each batch of DNA extractions included a 'blank' replicate, processed in the same manner as the tissue samples, but without tissue samples. New sterile tools, collection tubes and gloves were used between each sample recovery and individual sample preparation. However, we cannot exclude the possibility that positive PPC detection may be due to contamination from the surface of the tadpole and not directly from the liver tissue, specifically in the case of the tadpoles sampled in Panama. Indeed, such errors can arise in PCR screens [10]. Nonetheless, such a result would still infer that a putative parasite within the SPI-associated PPC-group was in close environmental association with the tadpoles recovered.
(b) NAG01 screening Two hundred and ninety base pairs of SSU rDNA sequence was amplified using 300F-B [3] and NAG01R_1 primers (electronic supplementary material, table S3). Each 25 µl reaction comprised 1× PCR Master Mix (Promega, containing Taq DNA polymerase), 500 nM of each primer and 5 µl DNA. Each assay included a negative (no-template) and positive control (1 µl of DNA from PPCinfected L. sylvaticus tadpole liver (KNA_DNA [11])). Extraction blanks (5 µl) were screened alongside liver DNA samples; these extraction controls were all negative. Cycling conditions were 2 min at 95°C, followed by 35 cycles of 30 s at 95°C, 30 s at 56°C and 60 s at 72°C, with an additional 10 min 72°C extension. PCR products were checked on a 2% agarose gel and purified using the GeneJET PCR Purification Kit (Thermo Fisher).
PCR products were cloned using the StrataClone PCR Cloning Kit (Agilent Technologies) following the manufacturer's protocol. Plasmid DNA was extracted from transformants using the GeneJET Plasmid Miniprep Kit (Thermo Fisher) and Sanger sequenced externally using standard T3-and T7-primers (Eurofins Genomics). Contigs were assembled using Sequencher (v. 5.4.6). To determine if the DNA samples contained additional NAG01 diversity, we triplicated the PCR, cloning and sequencing (electronic supplementary material, table S4). One of these samples (UK_HA01-5) was subjected to a second triplicate PCR with a lower annealing temperature (54°C) to further test for clade diversity by decreasing primer specificity.

(d) Host identification
A long-term goal of this research is to develop protocols for largescale screening technologies to facilitate amphibian conservation, preferably using at-bench and/or in-the-field technologies (e.g. Nanopore sequencing). As such, we adapted and tested a frog rDNA taxonomic barcode sequencing protocol for parallel-MinION™ amplicon sequencing: see electronic supplementary material, table S1 legend for methodological details. This approach allowed rapid identification of tadpole taxonomy, allowing tadpole collections to remain within permit limits, and without the need to export biological materials, or for complex/cumbersome laboratory-based instrumentation [19].

(e) Host species phylogeny
The amphibian mitochondrial rDNA sequences were searched against the NCBI 'nr' database using BLASTn (June-2019), allowing preliminary taxonomic identification. The sequences were aligned using the methodology described above (alignment: [13]). For the ML analysis, a GTR + F + I + G4 model was selected.

Results
Liver tissues from 91 tadpoles (81 from Panama and 10 from the UK) were screened for NAG01-Perkinsea (electronic supplementary material, The 72 NAG01-Perkinsea SSU rDNA sequences generated from both the Panama and the UK samples were highly similar to PPC sequences from the USA [1,2]. Two sequences from PA_T135 grouped with a subset of USA-derived sequences (figure 1). To exclude the possibility that the UK-Panama similarities were the product of contamination, we replicated a subset of the PCR and cloning steps, confirming this shared royalsocietypublishing.org/journal/rsbl Biol. Lett. 17: 20210166   figure S1) similar to other histopathological observations of SPI [2].

Discussion
The aim of this study was to investigate if PPC-tadpole associations had a biogeographic range beyond North America. We surveyed tadpoles from natural populations in Panama, a geographic region subject to extensive amphibian decline [20], and investigated a disease outbreak in a UK captive population. These results revealed evidence for PPC-tadpole associations in both regions, and generated PPC sequences identical, or nearly identical, to published PPC sequences from USA MMEs. These data demonstrate that PPC host range is broader than previously indicated [1,2], as the PPC-positive samples include Hyloidea species, including two from the Bufonidae family (PA_T128 and PA_T169 -figure 2), one of the most widespread and speciose anuran families [21]. Given the association between PPC and MMEs [1], the identification of PPC in additional locales, and diverse hosts, this result has potentially important implications for global frog conservation.
Phylogenetic analysis identified additional diversity within the PPC clade which could be a product of DNA polymerase errors, strain variation or intranuclear variation (figure 1, alignment available at [13]), with the latter known to be a factor in many alveolates [22]. However, the sequences show little geographical variation, suggesting: (i) the rDNA sequence region sampled is not subject to sufficiently high rates of mutation to allow identification of biogeographical structure, and/or (ii) the PPC clade has rapidly spread between the USA, UK and Panama. The latter possibility would have significant conservation implications, discussed below. However, at this point, we note that the results presented in figure 1 do not allow us to identify either strain/species or biogeographical phylogenetic structure and, as such, the phylogeny serves only to confirm that all sequences detected are part of a closely related clade within the SPI/PPC radiation. Further biogeographic and strain phylogenetic structure could potentially be resolved by sequencing additional, rapidly evolving markers, allowing further insights into the spread of PPC and host/microbe coevolution, as has been investigated for marine Perkinsea parasites [23,24].
Previously published North American PCR surveys found cryptic PPC associations without evidence of disease to be rare; indeed, only two out of 81 tadpoles that appeared asymptomatic (L. sphenocephalus) were PCR-positive for PPC [1]. Our results show that cryptic PPC associations without evidence for the disease are more prevalent than previously indicated, as all 10 of the PPC-positive Panama specimens appeared normal upon gross examination, demonstrating that PPC is present in Panama, but with no evidence of a disease-causing association. Interestingly, most of the SPI-associated tadpole MMEs in North America have been documented in ranids, i.e. superfamily Ranoidea [1][2][3][7][8][9], whereas the PPC-positive samples from this study were all Hyloidea. This disparity suggests that Hyloidea spp. may be more tolerant of infection. However, we note that the PPC-positive DNA samples from the UK were derived from symptomatic tissue from H. arborea (Hyloidea), and the SPI disease phenotype has been reported in two North American hylid species, Acris gryllus and Pseudacris crucifer [1]. Alternatively, the disparity observed might simply reflect the larger ratio of hylids-to-ranids in Central America compared to North America [25], or could be a product of differing time frames of coevolutionary interaction in different geographic areas, an important factor when considering recent candidate parasite introductions.
The discovery of PPC in UK captive populations raises concerns regarding PPC transmission. It is not possible to track the origin of the infection; the H. arborea frogs are bred separately from other amphibians, and the disease-causing agent could have been introduced from adult H. arborea frogs from a UK outlet in 2018, or by unidentified tank water or equipment contamination. Aquaculture disease outbreaks pose a threat to native species [26], as captive-bred/farmed animals (and their pathogens) often spread from aquaculture to natural environments [27]. Currently, we are not aware of any other PPC-linked MMEs in wild amphibian populations in the UK or elsewhere in Europe, but further research is necessary. Pathogens can be unintentionally spread across a large geographical range when captive-bred amphibians are translocated for trade, food and other commercial purposes; a phenomenon known as 'pathogen pollution' [28,29]. This has alarming implications for conservation, as many amphibian species are currently suffering catastrophic population decline [30]. Central America is home to a large diversity of amphibian species, many of which are in decline [31] and some now persist in local captive assurance colonies [32]. Our survey for NAG01-Perkinsea associations in Panama tadpoles did not yield evidence of SPI-associated disease. However, our detection of PPC DNA associated with a diverse group of amphibian hosts in Central America is a cause for concern. In the light of the current imperilled state of Panama amphibians, and the threat that infectious disease outbreaks pose to captive assurance colonies [33], our findings encourage management strategies, including routine monitoring for PPC. manuscript. All authors reviewed and approved the manuscript, agreeing to be held accountable for the content therein.