Host-adaptation of Rare Enterocytozoon bieneusi genotype CHN4 in Coypus (Myocastor coypus) in China

Background: Enterocytozoon bieneusi is a zoonotic gastrointestinal pathogen and can infect both humans and animals. Coypus (Myocastor coypus) are semi-aquatic rodents, in which few E. bieneusi infections have been reported and the distribution of genotypes and zoonotic potential remains unknown. Methods: A total of 308 fresh fecal samples were collected from seven coypu farms in China to determine the infection rate and the distribution of genotypes of E. bieneusi from coypus using nested-PCR amplication of the internal transcribed spacer (ITS) region of the ribosomal RNA (rRNA) gene. Results: E. bieneusi was detected with an infection rate of 41.2% (n = 127). Four genotypes were identied, including three known genotypes: CHN4 (n = 111), EbpC (n = 8) and EbpA (n = 7) and a novel genotype named CNCP1 (n = 1). Conclusions: The rare genotype CHN4 was the most common one in the present study, and the transmission dynamics of E. bieneusi in coypus were different from other rodents. This is the rst report of E. bieneusi infections in coypus in China. Our study reveals that E. bieneusi in coypus may be potential infection source to humans.

worldwide. Therefore, this study aimed to determine the genotypes and infection rate and assess the zoonotic potential of E. bieneusi from coypus in China.

Sample collection
A total of 308 fresh fecal samples were collected from asymptomatic coypus from seven farms in Anyang and Kaifeng in Henan Province, Yongzhou in Hunan Province, Laibin in Guangxi Zhuang Autonomous Region, Baoding in Hebei Province, Chengdu in Sichuan Province and Ganzhou in Jiangxi Province in China (Table 1; Fig. 1). Each farm was sampled on one occasion from August 2018 to March 2019. In each farm, about 2-4 coypus were kept in one accommodation, which was surrounded by 80 cmhigh walls to fence from each other. The ground of the accommodations was hardened with cement. An accommodation is typically composed of a piece of vacant land as the playground and a pool in which the coypus can swim. The samples were collected when the handlers nished the ground using a highpressure water gun. All the fecal samples were collected immediately after they excreted using sterile polyethylene gloves and marked with animal information. To avoid duplicate sampling of animals, only one fecal sample was collected from one location of the ground in each accommodation, and all deposits from each accommodation pooled as a single sample. All the samples were transferred to the laboratory in a cooler with ice packs within 36 hours and stored at 4°C.

DNA extraction and PCR ampli cation
Genomic DNA (gDNA) was directly extracted from 200 mg of each sample using E.Z.N.A. Stool DNA Kit (Omega Biotek Inc., Norcross, GA, USA) according to the manufacturer's protocol with minor modi cation.
All samples were tested using a nested PCR that targets ITS region (~389bp fragment) of the rRNA gene of E. bieneusi using primers described previously by Sulaiman et al [12]. Double distilled water and known positive DNA derived from Golden snub-nosed monkey (genotype D, Accession no.: KU604932) were used as negative and positive controls, respectively. The secondary PCR products were separated electrophoretically on 1% agarose (Life Technologies Corporation, CA, USA) gel stained with DNAGreen (Tiandz, Beijing, China) and visualized under UV light.

Sequencing and data analyses
Positive secondary PCR products were sequenced bidirectionally by Sangon Biotech Co. Ltd., Shanghai, China. The sequences obtained here were assembled and edited in the software Lasergene EditSeq version 7.1.0 (http://www.dnastar.com/) and multiple alignment with the reference sequences downloaded from GenBank was applied in Clustal X version 2.1 (http://www.clustal.org/). All statistical analyses were performed with IBM SPSS Statistics version 19.0 (www.ibm.com/products/spssstatistics). Difference of prevalence of E. bieneusi among different age groups were compared using Fisher's exact test, and the odds ratios (ORs) with the 95% con dence interval (CI) were also calculated. A two-sided P value of 0.05 or less was set as signi cant.
To reveal the phylogenetic relationships and zoonotic risk of E. bieneusi isolates, a phylogenetic tree was constructed by the Neighbor-Joining (NJ) method using the Kimura-2-parameter algorithm in MEGA version 7.0.26 (http://www.megasoftware.net). The robustness of the nodes was tested by a bootstrap analysis of 1,000 iterations.  Table 1). The differences in infection rates of E. bieneusi in coypus among different farms were statistically signi cant (P < 0.0001).
There was a signi cant negative correlation between the infection rate and age in this study, as an OR of 0.31 (95% CI: 0.14-0.70, P = 0.005) was associated with the 3-6-month-old group, and 0.12 (95% CI: 0.06-0.23, P < 0.0001) was associated with the > 6-month-old group.

E. bieneusi ITS genotypes
Four distinct E. bieneusi genotypes, including three previously reported genotypes [CHN4 (n = 111), EbpC (n = 8), EbpA (n = 7)], and one novel genotype (named CNCP1, n = 1) were observed. Genotype CHN4 was the most common genotype and detected in six farms except the farm in Yongzhou. Genotype EbpC was distributed in Yongzhou, Laibin and Kaifeng, while genotype EbpA and novel genotype CNCP1 were only detected in the specimens from Kaifeng.

Phylogenetic analysis of E. bieneusi
The phylogenetic relationships and zoonotic risk of E. bieneusi genotypes were analyzed by the NJ tree. Genotype CNCP1 had one single nucleotide polymorphism (SNP) at nucleotide position 274 (G to A) compared to genotype EbpA (Accession no.: MK968834). All the genotypes identi ed in this study were clustered in group 1 (Fig. 2).

Discussion
The infection rate of E. bieneusi in rodent species varies from 2.5% to 100% worldwide [13,14]. To the best of our knowledge, this is the rst report of E. bieneusi infections in coypus in China. In the present study, the infection rate of E. bieneusi was 41.2% in coypus, which is higher than the infection rate of E. bieneusi reported in brown rats (7.9%) [8], bamboo rats (5.1%) [15], experimental brown rats (4.8%) [16], commensal rodents (mouse and brown rat) (4.0%) [14], pet chinchillas (3.6%) [17] and red squirrels (19.4%) [18] in China. In addition, lower infection rates were also reported in wild house mice (10.7%) from a hybrid zone across the Czech Republic-Germany border [19], and beavers (15.3%) and muskrats (8.4%) from USA [20]. However, higher infection rates of E. bieneusi were reported in chipmunks (71.4%) and woodchucks (100%) from USA [13]. Similar infection rates of E. bieneusi have been reported in small rodents (mouse, bank vole, yellow-necked mouse and striped eld mouse) (38.9%) from southwestern Poland [21], and a laboratory prairie dog colony (37.9%) from USA [22]. The infection rates of E. bieneusi in rodents could be in uenced by many factors, such as animal immune status, age distribution, sample size, detection method, feeding environment, management system and population density [16]. Because the high infection rate detected in coypus in our study, we can draw a preliminary inference that coypus are more susceptible to E. bieneusi than many other rodent species, which should be con rmed by more investigations in the future.
A variation of positive rate of E. bieneusi in coypus was observed in the present study with the highest being detected in Anyang (72.3%, 73/101) and the lowest in Laibin (9.1%, 2/22). Geographical locationbased variation in the prevalence of E. bieneusi in rodents has been reported. Such as in brown rats in different provinces in China, which was ranged from 2.9% to 14.7% [8,14,16,23,24]. This phenomenon has also been reported in other animals, for example, in alpacas (Vicugna pacos) in China (0 -42.9%) [25] and in Asiatic black bear (Ursus thibetanus) in China (0 -50%) [26]. The difference may be related to geographical environments and feeding density.
In the present study, the dominant genotype of E. bieneusi was CHN4, which was detected in six cities except Yongzhou, indicating that genotype CHN4 is commonly found in coypus in China. This genotype has been identi ed in three human and two cattle samples [27] and four pre-weaned calf samples [28] from China, and it was rstly found in coypus here. These nding indicated that genotype CHN4 has a wide range of animal reservoirs and potential for zoonotic transmission. Genotype D was identi ed in squirrels from China [29] and USA [13], chipmunks [30], bamboo rats [15] and brown rats [8,23] from China, house mice from Czech Republic-Germany border [19] and striped eld mice from Poland [21], and genotype WL4 was observed in squirrels, chipmunks and muskrats from USA [13,20] (Table 3). EbpA, EbpC, PigEBITS7, S7, Peru16 and CHG14 have also been reported as the most common genotypes in experimental brown rat, beaver, giant rat, guinea pig, guinea pig and brown rat, respectively [10,14,16,20,23,31]. Additionally, in a more recent study of E. bieneusi in Himalayan marmots (Marmota himalayana) and Alashan ground squirrels (Spermophilus alashanicus) revealed that genotype ZY37 was the most common one [9].The rare genotype CHN4 was the dominant genotype, indicating that the transmission dynamic of E. bieneusi in coypus is different from other rodents. This may be explained by the unique life habits of coypus as aquatic rodents compared to other rodents involved in previous studies.
Genotype EbpA and EbpC have been detected in several rodent species (squirrel, house mouse, experimental brown rat, muskrat, bamboo rat and beaver) worldwide [15,16,19,20,29] (Table 3). They are two of the most common genotypes detected in both immunocompetent and immunocompromised people worldwide [1]. Meanwhile, genotype EbpA and EbpC have a vast host range, such as non-human primates (NHPs), livestock (cattle, buffalo, sheep and goat), pets (dog and horse), wild animals (deer, fox, raccoon, bear, panda and otter) and birds (pigeon, crane and parrot) [1]. These two genotypes also have been observed in lake water [32], river water [33] and wastewater treatment plants [34,35]. According to these data, the interspecies transmission of genotype EbpA and EbpC pose a zoonotic risk to human or other animals, and coypus may serve as a reservoir of EbpA and EbpC in the E. bieneusi transmission.
In the phylogenetic analysis, an NJ tree was constructed and the novel genotype CNCP1 clustered with CHN4, EbpC and EbpA in group 1. The majority of the zoonotic genotypes belongs to the group 1, and genotypes CHN4, EbpC and EbpA have been reported in humans [27,36,37], indicating that genotype CNCP1 maybe has zoonotic potential and the E. bieneusi isolates in coypus detected in this study can be transmissible from coypus to humans, especially the animal handlers, or vice versa.    Neighbor-joining tree of Enterocytozoon bieneusi ITS genotypes. Phylogenetic relationships of Enterocytozoon bieneusi genotypes of this study and other genotypes previously deposited in GenBank.
Bootstrap values >50% from 1,000 are shown on nodes. Sample names include GenBank accession number followed by host and then genotype designation. Known and novel genotypes identi ed in this study are indicated by empty and lled triangles, respectively.