Cryptosporidium myocastoris n. sp. (Apicomplexa: Cryptosporidiidae), the Species Adapted to the Nutria (Myocastor coypus)

Cryptosporidium spp., common parasites of vertebrates, remain poorly studied in wildlife. This study describes the novel Cryptosporidium species adapted to nutrias (Myocastor coypus). A total of 150 faecal samples of feral nutria were collected from locations in the Czech Republic and Slovakia and examined for Cryptosporidium spp. oocysts and specific DNA at the SSU, actin, HSP70, and gp60 loci. Molecular analyses revealed the presence of C. parvum (n = 1), C. ubiquitum subtype family XIId (n = 5) and Cryptosporidium myocastoris n. sp. XXIIa (n = 2), and XXIIb (n = 3). Only nutrias positive for C. myocastoris shed microscopically detectable oocysts, which measured 4.8–5.2 × 4.7–5.0 µm, and oocysts were infectious for experimentally infected nutrias with a prepatent period of 5–6 days, although not for mice, gerbils, or chickens. The infection was localised in jejunum and ileum without observable macroscopic changes. The microvilli adjacent to attached stages responded by elongating. Clinical signs were not observed in naturally or experimentally infected nutrias. Phylogenetic analyses at SSU, actin, and HSP70 loci demonstrated that C. myocastoris n. sp. is distinct from other valid Cryptosporidium species.


Introduction
Cryptosporidium is a protist genus that infects the gastrointestinal and respiratory tract of vertebrate hosts [1]. Cryptosporidiosis, the disease caused by members of this genus, frequently results in diarrhoea, which can be severe and fatal [2]. However, many species and genotypes of Cryptosporidium, particularly those infecting wild animals, do not cause clinical signs [3,4]. Genetic and biological studies have shown a high diversity within the genus Cryptosporidium, with much of this diversity observed in wildlife hosts [5][6][7][8][9][10]. To date, 47 valid species [11][12][13] and more than 100 genotypes, which are distinguished from valid species on the basis of molecular differences and probably represent separate species, have been described [2]. However, much remains to be discovered about the diversity of the genus Cryptosporidium and its host range.

Molecular Study
Total gDNA was extracted from 200 mg of faecal samples and 100 mg of tissue specimens using a GeneAll ® Exgene TM Stool DNA mini Kit (GeneAll Biotechnology, co., Ltd.; Seoul, South Korea) and DNeasy Blood & Tissue kits (QIAGEN, Hilden, Germany), respectively, followed by bead disruption for 60 s at 5.5 m/s using 0.5 mm glass beads in a FastPrep ® -24 Instrument (MP Biomedicals, Santa Ana, CA, USA). The acquired gDNA was stored at −20 • C.
Nested PCR protocols were used to amplify a partial sequence of the SSU, actin, 70 kDa heat shock protein (HSP70), and 60 kDa glycoprotein (gp60) genes using previously published protocols [24][25][26][27][28][29]. For the SSU fragment, the primers 5 TTCTAGAGCTAATACATG CG3 and 5 CCCATTTCCTTCGAAACAGGA3 were used in the primary reaction, and the primers 5 GGAAGGGTTGTATTTATTAGATAAAG3 and 5 AAGGAGTAAGGAACAACC TCCA3 were used in the secondary reaction. For the actin fragment, the primers 5 ATCRG WGAAGAAGWARYWCAAGC3 and 5 AGAARCAYTTTCTGTGKACAAT3 were used in the primary reaction, and the primers 5 CAAGCWTTRGTTGTTGAYAA3 and 5 TTTCTGT GKACAATWSWTGG3 were used in the secondary reaction. For the HSP70 fragment, the primers 5 GCTCGTGGTCCTAAAGATAA3 and 5 ACGGGTTGAACCACCTACTAAT3 were used in the primary reaction, and the primers 5 ACAGTTCCTGCCTATTTC3 and 5 GCTAATGTACCACGGAAATAATC3 were used in the secondary reaction. For the gp60 fragment, the primers 5 ATAGTCTCCGCTGTATTC3 and 5 GGAAGGAACGATGTATCT3 were used in the primary reaction and the primers 5 TCCGCTGTATTCTCAGCC3 and 5 GCAGAGGAACCAGCATC3 were used in the secondary reaction. Some PCR conditions were modified from the original publications, as previously reported [11]. Molecular grade water and DNA of C. occultus were used as negative and positive controls, respectively, for the amplification of SSU, actin, and HSP70 genes. DNA of C. hominis subtype family Ib was used as a positive control for amplification of the gp60 fragment of C. parvum and Cryptosporidium myocastoris n. sp. DNA of C. ubiquitum subtype family XIIa was used as a positive control for the amplification of C. ubiquitum. The secondary PCR products were separated by electrophoresis on an agarose gel, stained with ethidium bromide and visualised under UV illumination. All amplicons were purified using the GenElute™ Gel Extraction Kit (Sigma-Aldrich, St. Louis, MO, USA) and directly sequenced using the secondary PCR primers at Eurofins (Prague, Czech Republic).
Chromatogram analysis was performed using Chromas Pro 2.1.4 (Technelysium, Pty, Ltd., South Brisbane, Australia), and sequences were verified by BLAST analysis (https://blast.ncbi.nlm.nih.gov/Blast.cgi; accessed on 20 February 2021). The sequences obtained in this study and reference sequences obtained from GenBank were aligned using the MAFFT version 7 online server (http://mafft.cbrc.jp/alignment/server/; accessed on 20 February 2021) using the E-INS-i multiple alignment method. The alignments were manually trimmed and edited in BioEdit v.7.0.5 [30]. Phylogenetic analysis was performed using the maximum likelihood (ML) method, using evolutionary models selected by MEGAX [31]. Bootstrap supports were calculated from 1000 replications. Phylogenetic trees were produced by MEGAX and further edited for visualisation purposes with Corel DrawX7 (Corel Corporation, Ottawa, Ontario, Canada). Species-specific divergences were identified from proportional distances (%) which were calculated in the program Geneious v11.0.3 [32] based on the SSU, actin, and HSP70 datasets of all sequences under study. All nucleotide sequence data were deposited in GenBank, Accession Numbers (Acc. nos.): MW274645-MW274661 for SSU, MW280959-MW280975 for actin, MW280976-MW280992 for HSP70, and MW280993-MW281009 for gp60.

Animals for Transmission Studies
Six adult nutrias (nutria 0 and nutrias 1-5; Myocastor coypus), five one-week-and eight-week-old gerbils (Meriones unguiculatus), five one-week-and eight-week-old SCID mice (Mus musculus; strain C.B-17), five one-week-and eight-week-old BALB/c mice, and five one-day-old chickens (Gallus gallus f. domestica) were used for transmission studies. Three animals from each group were used as negative controls. All nutrias used for the transmission study were screened daily for the presence of specific DNA and oocysts of Cryptosporidium for two weeks prior to experimental inoculation. Mice and gerbils, which were bred under laboratory conditions, were screened for the presence of specific DNA and oocysts of Cryptosporidium one week prior to experimental inoculation. One-day-old chickens, which were hatched under laboratory conditions, were screened on the day of hatching.

Animal Care
To prevent environmental contamination with oocysts, mice and gerbils were individually housed in ventilated cages (Tecniplast, Buguggiate, Italy), and chickens were housed in plastic boxes that were disinfected at 80 • C for one hour before being used. Nutrias were kept in boxes with secured walls. The individual boxes were disinfected with pressurized steam before use. Cages and boxes were sized in accordance with European and Czech Republic regulations on the protection of animals against cruelty. An external source of heat was used in the first five days for chickens. Each animal was supplied with a sterilized diet and sterilized water ad libitum. Animal keepers wore disposable protective equipment during care of the animals. Woodchip and straw bedding and disposed protective clothing were removed from the experimental room and incinerated. All experimental procedures complied with the law of Czech Republic (Act No. 246/1992 Coll., on the protection of animals against cruelty). The study design was approved by ethical committees at the Biology Centre of CAS, the State Veterinary Administration, and the Central Commission for Animal Welfare under Protocol No. 35/2018 and 60/2019.

Design of Transmission Studies
Nutria 0 was inoculated orally with 5,000 purified oocysts of C. myocastoris n. sp. recovered from naturally infected nutria (described above) and suspended in 500 µL of distilled water. Oocysts of C. myocastoris n. sp., recovered from experimentally infected nutria 0, were molecularly characterised and used to infect other experimental animals (above). Other animals were inoculated with 20,000 purified oocysts of C. myocastoris n. sp. suspended in 500 µL (nutria), 200 µL (mice, gerbils, and chickens) of distilled water. Animals serving as negative controls were inoculated orally with 500 µL (nutria), 200 µL (mice and gerbils) or 20 µL (chickens) of distilled water. Faecal samples from each animal were screened daily for the presence of Cryptosporidium oocysts (ACMV staining) and Cryptosporidium-specific DNA (SSU gene amplification and sequencing). All experiments were terminated 30 days post-infection (DPI). The course of infection was evaluated based on the presence of C. myocastoris n. sp. specific DNA and the number of oocyst per gram of faeces as previously described by Kváč et al. [23]. Consistency and colour of faeces and health status were determined daily for each sample and animal, respectively. From each experimentally infected animal, PCR-positive samples from the beginning, middle, and end of the infection were additionally examined at the actin, HSP70 and gp60 genes to verify the identity of the C. myocastoris n. sp. with the inoculum and the original isolate.

Morphological Evaluation
The oocysts of Cryptosporidium myocastoris n. sp., which originated from naturally infected nutrias (isolates 31132 and 31459, Table 1) and experimentally infected nutrias (nutria N0 and nutria N1), were purified using sucrose and caesium chloride gradient centrifugation and examined using differential interference contrast (DIC) microscopy, bright field microscopy following ACMV and modified Ziehl-Neelsen (ZN) staining [35], and fluorescence microscopy following labelling of the Cryptosporidium oocyst wall with genus-specific FITC-conjugated antibodies (IFA; Cryptosporidium IF Test, Cryptocel, Cellabs Pty Ltd., Brookvale, Australia). Images of oocysts were collected using an Olympus Digital Colour camera (DP73) and Olympus cell SensEntry 2.1 software (Olympus Corporation, Shinjuku, Tokyo, Japan). Length and width of oocysts (n = 30) from naturally and experimentally infected nutrias were measured under DIC at 1000× magnification. The length-to-width ratio was calculated for each oocyst. As a control, the morphometry of C. parvum from naturally infected calfs (Bos taurus; n = 30) and C. ratti from experimentally infected rats (Rattus novegicus; n = 30) was used. Photomicrographs of Cryptosporidium myocastoris n. sp. oocysts under DIC, ACMV, ZN and IFA are part of this publication, and have been deposited as a phototype at the Institute of Parasitology, Biology Centre of the Czech Academy of Sciences, Czech Republic.

Clinical and Pathomorphological Examinations
A complete necropsy of two animals from each experimental group was performed at 10 and 20 DPI. Tissue specimens (oesophagus, stomach, duodenum, proximal, central and distal jejunum, ileum, caecum, colon, liver, spleen, kidney, bladder, and lung) from each animal were obtained using different sterile dissection tools for each location. Specimens from each organ were collected for PCR/sequencing, histology, and scanning electron microscopy (SEM). Histology sections were processed as reported by Kváč and Vítovec [36], and SEM sections were processed as described by Holubová et al. [11]. Histology sections were stained with haematoxylin and eosin (HE) and periodic acid-Schiff (PAS), examined at 100-400× magnification and documented using Olympus cell Sens Entry 2.1 (Olympus Corporation, Shinjuku, Tokyo, Japan) equipped with a digital camera (Olympus DP73). Specimens for SEM were examined using a JEOL JSM-7401F-FE scanning electron microscope equipped with a digital camera ETD Detector A PRED (Termo Fisher Scientific, Waltham, MA, USA).

Statistical Analysis
Differences in oocyst sizes were tested using Hotelling's multivariate version of the 2 sample t-test, package ICSNP: Tools for Multivariate Nonparametrics [37] in R 4.0.0. [38]. The hypothesis tested was that two-dimensional mean vectors of measurement are the same in the two populations being compared.

Results
In total, 72 and 78 faecal samples were examined from the Czech Republic and Slovakia, respectively ( Table 1). None of the faecal samples had a consistency that would indicate diarrhoea. Examination of faecal smears revealed the presence of Cryptosporidium sp. oocysts in three samples, and infection intensity ranged from 10,000 to 25,000 OPG (Table 1). Cryptosporidium-specific DNA was detected in 11 samples by nested PCR targeting the SSU gene ( Figure 2). From these positive samples, partial sequences of the genes encoding actin, HSP70 and gp60 were amplified/sequenced. ML trees constructed from SSU, actin and HSP70 sequences showed the presence of C. parvum (n = 1), C. ubiquitum (n = 5) and Cryptosporidium myocastoris n. sp. (n = 5). Cryptosporidium myocastoris n. sp., previously known as the Cryptosporidium coypu genotype, is described in detail as a new species later in this publication (Figures 2-4). There was no intraspecies variability in SSU, actin, and HSP70 sequences from this study. Subsequent subtyping of C. parvum and C. ubiquitum at the gp60 gene showed the presence of subtype families IIaA16G1R1 and XIId, respectively (Table 1, Figure 5). All gp60 sequences from C. ubiquitum were identical ( Figure 5). Two novel subtype families, which we have named XXIIa (n = 2) and XXIIb (n = 3), were detected within C. myocastoris n. sp. (Table 2, Figure 5).
Oocysts of C. myocastoris n. sp. recovered from naturally infected nutria (isolate 31132) were infectious for Cryptosporidium-free farmed nutria (nutria N0), which shed oocysts that were genetically and morphometrically identical to the inoculum, from five DPI ( Table 2;     Note: Length and width of 30 oocysts from each isolate were measured under differential interference contrast at 1000× magnification, and out of these the length-to-width ratio of each oocyst was used to calculate. Oocysts of C. myocastoris n. sp. from nutria N0 were infectious for five farmed nutrias (nutrias N1-N5), but not for one-week-and eight-week-old BALB/c and SCID mice, gerbils, or one-day-old chickens. All groups of nutrias, mice, gerbils, and chickens used as negative controls remained uninfected. Experimentally infected nutrias N1-N5 started to shed oocysts of C. myocastoris n. sp., detectable by light microscopy and PCR, at 5-6 DPI, and all animals remained infected until the end of the experiment ( Figure 6). The infection intensity ranged from 2000 to 62,000 OPG. The highest infection intensity, 20,000-62,000 OPG, was observed from 6 to 11 DPI. Beginning on day 12 post-infection, all animals shed fewer than 10,000-15,000 OPG ( Figure 6). None of the animals showed signs of cryptosporidiosis, and faecal consistency was appropriate to the age of the animal and the food intake. Cryptosporidium myocastoris n. sp. DNA and Cryptosporidium developmental stages were detected exclusively in posterior jejunum and ileum (Figures 7 and 8). Histology and SEM showed low infection intensity, with one or two developmental stages typically observed on an isolated villus in the posterior jejunum and ileum (Figures 7 and 8). This low infection intensity was consistent throughout the posterior jejunum and ileum. The brush border microvilli adjacent to attached developmental stages responded by elongation (Figure 8). The area of elongated microvilli increased with the size of the developmental stage, and the microvilli were elongated by up to 2 µm (Figure 8). The lamina propria in the jejunum and ileum was occasionally observed to be slightly oedematous, but these changes were probably not related to the Cryptosporidium infection.   Oocysts of C. myocastoris n. sp. recovered from experimentally infected nutrias (nutria N0 and N1) were morphometrically identical to those recovered from naturally infected nutrias (isolates 31132 and 31459; Table 2).   (Figure 9). The oocyst wall is smooth and colourless (Figure 9a). The oocyst residuum is composed of numerous small granules and one spherical globule is clearly visible; a suture is not noticeable. Four sporozoites are clearly visible within oocysts. The morphology and morphometry of other developmental stages is unknown. Remarks. Oocysts of C. myocastoris n. sp. are well stained by ACMV and ZN staining methods, similarly to other Cryptosporidium spp. (Figure 9b,c), and their oocyst walls crossreact with immunofluorescence reagents developed primarily for C. parvum (Figure 9d). Oocysts from naturally and experimentally infected nutrias did not differ in size (T 2 = 0.16, df 1 = 2, df 2 = 121, p = 0.8506; Table 2). Oocysts of C. myocastoris n. sp. are smaller than those of C. parvum (T 2 = 33.11, df 1 = 2, df 2 = 48.15, p = p < 0.001) and C. ratti (T 2 = 33.22, df 1 = 2, df 2 = 45.96, p < 0.001), but these differences are not of practical significance for identification (Table A1 in Appendix A). Cryptosporidium myocastoris n. sp. can be differentiated genetically from other Cryptosporidium species based on sequences of SSU, actin, and HSP70 genes. At the gp60 locus, C. myocastoris n. sp. develops two wellsupported clades. Pairwise distances between Cryptosporidium myocastoris n. sp. and the selected closest and furthest Cryptosporidium species at SSU, actin, and HSP70 genes are shown in Table A2.

Discussion
The work presented here represents the most comprehensive study to date on Cryptosporidium infecting feral nutrias, and a novel, nutria-adapted Cryptosporidium species is described. The prevalence of Cryptosporidium in nutrias was relatively low (7.3%). In a study of 108 wild nutrias in Argentina, the prevalence of Cryptosporidium infection was similarly low (3.7%), although oocysts were detected by microscopy, which is less sensitive than PCR [15]. The prevalence of Cryptosporidium in other aquatic rodents also appears to be low. A study of 145 capybaras (Hydrochoerus hydrochaeris) in Brazil reported a prevalence of 5.5% [41]. Zhou et al. [42] did not detect Cryptosporidium in any of the 84 North American beavers (Castor canadensis) they tested in the United States, but they did report that almost 12% of 237 muskrats (Ondatra zibethicus) were positive. Paziewska et al. [43] reported that 7.7% of 22 European beavers (Castor fiber), sampled in Poland, had Cryptosporidium antigen in their faeces. Considering that Cryptosporidium infections are frequently associated with transmission through contaminated water [44,45], the low prevalence in nutrias and other water rodents might be surprising. Future studies should address how Cryptosporidium spp. are transmitted in aquatic mammals: is waterborne transmission important, or is it some other route such as contaminated food or direct contact among individuals? In this context, it is noteworthy that C. myocastoris n. sp. has not been reported in any of the studies on the occurrence of Cryptosporidium spp. in surface water [46]. This absence may be explained by (i) the low prevalence within the population of nutrias together with the low infection intensity; or (ii) the limited number of studies reporting water contamination in the areas where nutrias occur. This is the first study to genotype Cryptosporidium from nutrias. Martino et al. [15] found Cryptosporidium sp. oocysts (4 to 4.5 × 4.0 µm) in faecal samples from the colon and rectum of wild nutrias in Argentina. An earlier study reported C. parvum in nutrias based on oocyst morphology [15,20]. Given that C. parvum has a broad host specificity, is reported infrequently in wildlife species [5,6,[47][48][49][50], and was found in a single nutria in the present study, it is possible that Pavlásek and Kozakiewicz [20] correctly identified the species. In support, they found that oocysts from naturally infected nutria were infectious for four-day-old laboratory mice under experimental conditions, a characteristic of C. parvum but not C. myocastoris (as we have shown in the present study). However, based on our current knowledge, oocyst morphology cannot reliably distinguish among intestinal species of Cryptosporidium [51], and the size of oocysts reported by Pavlásek and Kozakiewicz [20] (5.0 × 4.75 µm) is similar to C. myocastoris n. sp. (5.02 × 4.85 µm) and C. ubiquitum (5.04 × 4.66 µm), two species that we found to be more prevalent than C. parvum (5.19 × 4.9 µm) in nutrias.
The finding of C. ubiquitum and C. myocastoris n. sp. in several animals at different localities suggests that these species are common in feral nutrias. Cryptosporidium ubiquitum, a species with broad host specificity, has been reported in domestic and wild ruminants, rodents, carnivores, and human and non-human primates [52][53][54][55][56]. The C. ubiquitum gp60 subtype family XIId in nutrias has been reported in humans, macaques, red deer, raccoons, woodchucks, chinchillas, and mink [18,29,[56][57][58][59], suggesting that it is broadly specific. In contrast, C. myocastoris n. sp. appears to have a narrow host specificity. The origin of C. myocastoris n. sp. in nutria in central Europe is difficult to elucidate without further studies, but it may have been introduced into Europe with imported nutrias, similarly to the Cryptosporidium skunk genotype, which was likely introduced to Europe with eastern grey squirrels [60]. The specificity of C. myocastoris n. sp. for nutrias is supported by its presence in geographically isolated nutrias, its infectivity for nutrias under experimental conditions, and the absence of any record of this species in any of the thousands of molecular epidemiological studies published in last two decades [46,51,61,62]. For these reasons, it is most likely that nutrias are the major host of C. myocastoris n. sp., although we cannot exclude the possibility that other host species have a role as major or minor hosts.
Adaptation to one host species is not unique within the genus Cryptosporidium, and several mammalian Cryptosporidium spp. have been reported almost exclusively in a single host. Examples include C. wrairi and C. homai in guinea pigs (Cavia porcellus), C. scrofarum and C. suis in pigs (Sus scrofa), C. bovis in cattle (Bos taurus), C. occultus in rats (Rattus spp.), and C. macropodum in kangaroos [63][64][65][66][67]. The findings of these species in other hosts represent rare cases or mechanical passage rather than host adaptation [68][69][70].
Nutrias positive for C. parvum or C. ubiquitum did not have detectable oocysts in their faeces, suggesting a low level of infection, which is consistent with our previous finding [39]. Three out of five nutrias naturally infected with C. myocastoris n. sp. shed oocysts at levels between 10,000 and 20,000 OPG, and similar numbers were detected in experimentally infected nutrias. No macroscopic changes in the intestinal mucosa were observed at necropsy of C. myocastoris positive nutrias. The low level of oocyst shedding was consistent with the intensity of developmental stages detected in the intestinal epithelium and the absence of clinical symptoms. A similar relationship between intestinal involvement and oocyst shedding has been observed in other Cryptosporidium spp. infecting the small intestine [6,12,[73][74][75][76]. Examination by scanning electron microscopy shows the elongation of the microvilli around the C. myocastoris developmental stages, which has also been previously observed in SCID mice infected with C. parvum [77]. Borowski et al. [78] reported the elongation of microvilli on the gliding trails of C. parvum sporozoites between an excysted oocyst and newly formed trophozoites. While the extending of microvilli in Borrowski's study was up to 15 µm, we observed much less extension. Clinical symptoms are rarely, if ever, observed in wild animals infected with host-specific Cryptosporidium species [79][80][81].
Equivalently to the host specificity, most Cryptosporidium species are characterized by adaptation to a different part of the digestive tract. Gastric Cryptosporidium of mammals exclusively infect the glandular part of the glandular stomach, whereas intestinal species are adapted to different parts of the small or large intestine. Similar organ adaptation has been reported in Eimeria spp. [82]. AS with C. ratti or C. scrofarum, the life cycle of C. myocastoris n. sp. is located in the posterior jejunum and ileum [12,66]. No developmental stages were found in the large intestine where the life cycle of, e.g., C. occultus, C. suis or C. ornithophilus is located [63,65,81].
Our previous work has shown that only using sequences of SSU to infer evolutionary relationships of Cryptosporidium may lead to erroneous conclusions [5,29,89]. Therefore, it is necessary to use other polymorphic loci, such as HSP70, actin, or COWP genes, in phylogenetic analyses [5,85]. Although bootstrap support for the SSU tree was lower than for the actin and HSP70 trees in this study, C. myocastoris n. sp. formed a separate clade in SSU, actin and HSP70 ML trees, with the most closely related group comprising species such as C. parvum, C. cuniculi, and C. wrairi. At the SSU locus, the pairwise distance between C. myocastoris n. sp. and C. parvum (0.014) or C. cuniculi (0.018) is similar to that between C. apodemi and C. occultus (0.016) or C. andersoni and C. serpentis (0.019). At the actin locus, the distance between C. myocastoris n. sp. and C. parvum/C. cuniculi (0.039) is similar to that between C. meleagridis and C. erinacei (0.034). Comparably, at the HSP70 locus, the distance between C. myocastoris n. sp. and C. parvum/C. cuniculi (0.035/0.031) is similar to the distance between C. meleagridis and C. erinacei (0.041) or C. ornithophilus and C. avium (0.035). These results support the genetic uniqueness of C. myocastoris n. sp. and their status as a separate species of genus Cryptosporidium.

Conclusions
In summary, the findings that Cryptosporidium sp. coypu genotype is genetically distinct from all described Cryptosporidium species and specific for nutrias under natural and experimental conditions support its description as a new species, and we propose that it be named Cryptosporidium myocastoris n. sp.

Conflicts of Interest:
None of the authors has any competing interests in the manuscript.