Genotypes and public health potential of Enterocytozoon bieneusi and Giardia duodenalis in crab-eating macaques

Background Enterocytozoon bieneusi and Giardia duodenalis are common human and animal pathogens. Studies have increasingly shown that non-human primates (NHPs) are common hosts of these two zoonotic parasites. However, few studies have explored the genetic diversity and public health potential of these pathogens in laboratory monkeys. In this study, we examined the genetic diversity of the two pathogens in crab-eating macaques (Macaca fascicularis) in a commercial facility in Hainan, China. Results Enterocytozoon bieneusi and G. duodenalis were detected by PCR analysis in 461/1452 (31.7%) and 469/1452 (32.3%) fecal specimens from the animals, respectively. Significantly higher detection rates of E. bieneusi were detected in males (36.5%, 258/706) than in females (26.7%, 160/599; χ2 = 14.391, P = 0.0001), in animals with loose stools (41.4%, 151/365) than those with normal stool (28.5%, 310/1087; χ2 = 20.83, P < 0.0001), and in animals of over 3 years of age (38.6%, 135/350) than those of 1–3 years (29.6%, 326/1,102; χ2 = 9.90, P = 0.0016). For G. duodenalis, the detection rate in males (33.4%, 236/706) was higher than in females but not statistically significant (30.2%, 181/599; χ2 = 1.54, P = 0.2152), in monkeys with loose stools (41.1%, 150/365) than those with normal stools (29.3%, 319/1087; χ2 = 17.25, P < 0.0001), and in monkeys of 1–3 years of age (36.6%, 403/1102) than those over 3 years (18.9%, 66/350; χ2 = 38.11, P < 0.0001). Nine E. bieneusi genotypes were detected in this study by DNA sequence analysis of the internal transcribed spacer of the rRNA gene, namely Type IV (236/461), Peru8 (42/461), Pongo2 (27/461), Peru11 (12/461), D (4/461) and PigEbITS7 (1/461) previously seen in NHPs as well as humans, and CM1 (119/461), CM2 (17/461) and CM3 (3/461) that had been only detected in NHPs. DNA sequence analyses of the tpi, gdh and bg loci identified all G. duodenalis specimens as having assemblage B. Altogether, eight (4 known and 4 new), seven (6 known and 1 new) and seven (4 known and 3 new) subtypes were seen at the tpi, gdh and bg loci, leading to the detection of 53 multi-locus genotypes (MLG-B-hn01 to MLG-B-hn53). Most of them were genetically related to those previously seen in common Old-World monkeys. Conclusions Data from this study indicate a common occurrence of zoonotic genotypes of E. bieneusi and assemblage B of G. duodenalis in farmed crab-eating macaques in Hainan, China.


Background
Giardia duodenalis and Enterocytozoon bieneusi are common human pathogens. At present, there are more than 200 million of annual giardiasis cases in humans, while microsporidiosis is a common cause of diarrhea [1,2]. The incidence of giardiasis has been reported to be 5.5 per 100,000 people in the UK and 7.5 per 100,000 people in the USA [3]. In China, almost 30 million people are infected with G. duodenalis every year [2]. For E. bieneusi, the detection rates range between 2-78% in industrialized and developing countries [4][5][6]. Over 200 giardiasis outbreaks have been reported in the world during the period 2004-2016, while E. bieneusi also caused an outbreak in France [6][7][8][9].
Non-human primates (NHPs) are important experimental animals in public health research because of their high genetic similarity to humans [10]. A growing number of studies have found that NHPs are the hosts of many parasites, including gastrointestinal protists E. bieneusi and G. duodenalis, which are transmitted in similar fecal-oral routes [2,11]. In addition to affecting the health and wellbeing of these laboratory animals, both pathogens are potentially zoonotic, causing diseases in humans [12,13].
Of the 17 known human-pathogenic microsporidian species, E. bieneusi is the most common [14]. Based on sequence analysis of the internal transcribed spacer (ITS) of the rRNA gene, more than 200 E. bieneusi genotypes have been identified [15]. In phylogenetic analysis, these E. bieneusi genotypes are divided into at least 11 groups [16]. Among them, Group 1 contains most genotypes found in humans and many genotypes in animals, thus is considered to be the zoonotic group. In contrast, Groups 2-11 include genotypes found in specific groups of hosts, including humans, thus are considered more host-specific. There is also a so-called outlier group in dogs [17][18][19][20][21][22]. Thus, genotyping E. bieneusi in NHPs can help us understand the zoonotic potential of E. bieneusi in these animals.
Similarly, eight distinct G. duodenalis assemblages (A-H) have been identified by genetic analysis of triosephosphate isomerase (tpi), ssrRNA, β-giardin (bg), glutamate dehydrogenase (gdh) and other genes [2,32,33]. Among them, assemblages A and B are most commonly found in humans and NHPs, assemblages C and D are mainly detected in canines, assemblage E mainly infects ruminants and other hoofed animals, whereas assemblages F, G and H usually infect cats, rodents and seals, respectively [2]. Similar to E. bieneusi, genotyping G. duodenalis also can help us to understand the transmission of this pathogen.
NHPs are also potential reservoir hosts for zoonotic transmission of G. duodenalis [2,12]. In previous studies, assemblages A, B and E have been identified in humans and NHPs [34]. Among them, assemblage B appears to be most common, while assemblage E is only occasionally detected [34][35][36][37][38]. Although assemblage A has been further classified into three major sub-assemblages (AI-AIII) by sequence analysis of several genetic loci, consistent secondary classification of assemblage B has not been established [2,12,39,40]. Multilocus genotyping (MLG) has been used in several studies to understand the host specificity and zoonotic potential of assemblage B in human and NHPs [41][42][43][44]. Controversies exist on the differences in virulence between assemblages A and B in humans [45]. There are no such studies on G. duodenalis in NHPs.
In the present study, we examined the prevalence of E. bieneusi and G. duodenalis in fecal specimens from commercial crab-eating macaques (Macaca fascicularis) in Hainan. The genetic diversity of the two pathogen species was assessed using sequence analysis of the ITS (E. bieneusi) and using MLG analysis of the tpi, gdh and bg gene (G. duodenalis). The data generated were used to explore the human-infective potential of these two common gastrointestinal parasites in NHPs.

Specimen collection
A total of 1452 fecal specimens were collected in April 2016, June 2017, October 2017 and January 2018 from laboratory crab-eating macaques kept on a commercial farm. The farm was founded in 2003 and has been awarded full accreditation from the International Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) since 2008, with over 20,000 animals at the sampling time. All animals were kept in separate cages (1 × 1 × 1 m) after they were born on the farm, with approximately 30 animals kept in each room of ~ 60 m 2 . The cages were elevated 1 m from the ground so that the feces could fall onto the ground. Every monkey had contact with animals in the neighboring cages. The rooms were cleaned every morning and afternoon to ensure a clean living environment.
Feed, fruits (apple, banana and peach) and drinking water were regularly distributed by farm staff to each cage every day.
The sampling plan took into consideration the number, age and sex of animals on the farm, and the needed number of positive specimens to generate data for a meaningful assessment of the distribution and human-infective potential of E. bieneusi and G. duodenalis genotypes in these animals. Among the sampled animals, 706 were male, 599 were female and 147 sampled animals had missing information on the sex. The sampled animals belonged to two age groups: 1102 were 1-3 years-old and 350 were adult monkeys older than 3 years. Regarding the latter, as monkeys over 4 years-old were often sold, the oldest animals sampled in the study were 5 years-old. At the time of sampling, 365 monkeys had loose stools, as defined by runny fecal consistency, and 1087 monkeys were apparently normal. The specimens were stored in 2.5% potassium dichromate solution at 4 °C prior to DNA extraction.

DNA extraction
The stored fecal specimens were washed three times with distilled water by centrifugation at 2000×g for 10 min. Genomic DNA was extracted from the washed fecal material using the FastDNA SPIN Kit for soil (MP Biomedicals, Santa Ana, CA) [46]. DNA was stored at − 20 °C before being used in PCR analysis within one year.

Detection and genotyping of Enterocytozoon bieneusi and Giardia duodenalis
Enterocytozoon bieneusi was detected by nested PCR amplification of a 392-bp fragment of the rRNA gene containing the entire ITS sequence [47]. The genotypes of E. bieneusi found in this study were determined by sequencing the PCR products and comparing the sequences obtained from the specimens with the reference sequences from known genotypes. The established genotype nomenclature was used in naming E. bieneusi genotypes identified in this study [15].
Giardia duodenalis was detected by nested PCR amplification of a 530-bp fragment of the tpi gene, a 511-bp fragment of the bg gene and a 599-bp fragment of the gdh gene [48][49][50]. The specimen was considered G. duodenalis-positive if any of the PCRs generated the expected PCR product. The genotypes of G. duodenalis found in this study were identified by sequencing PCR products from G. duodenalis-positive specimens and comparing sequences obtained with the reference sequences from the known genotypes at each genetic locus.

Sequence analysis
The secondary PCR products of the ITS, tpi, bg and gdh genes were sequenced in both directions on an ABI 3730 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA). Nucleotide sequences generated were assembled and edited with software ChromasPro v.1.32 (http://techn elysi um.com.au/Chrom asPro .html), and aligned with reference sequences from GenBank by using ClustalX (http://clust al.org).

Phylogenetic analysis
To explore genetic diversity within the assemblage B of G. duodenalis, tpi, bg and gdh sequences from specimens with complete data at the three genetic loci were concatenated to form multi-locus sequences. They were compared with the reference sequences MLG1-MLG15 from NHPs, and Sweh001, Sweh059, Sweh074, Sweh107, Sweh136, Sweh158, ECUST5414, ECUST1710, ECUST4064 and ECUST981 from humans [28,34,51]. A maximum likelihood (ML) tree was constructed in MEGA v.6 (https ://www.megas oftwa re.net) using evolutionary distances calculated by the commonly used general time reversible model. The reliability of clusters formed was assessed by bootstrap analysis using 1000 replicates.

Statistical analysis
Differences in E. bieneusi and G. duodenalis detection rates between groups of different sex, age, or fecal consistency were assessed by using the Chi-square test implemented in SPSS Statistics v.20.0 (IBM Corp., Armonk, NY, USA). The difference was considered significant when P < 0.05.

Phylogenetic relationship of G. duodenalis assemblage B
Phylogenetic analysis of concatenated sequences of the 53 assemblage B MLGs in this study, and those from previous studies [28,34,51] showed that most MLGs from this study were related to MLGs previously found in Old World monkeys (MLG-3, MLG-4, MLG-7, MLG-8, MLG-14 and MLG-15). However, one of the MLGs, MLG-B-hn31, seen in one animal, clustered together with MLGs in humans. In addition, MLG-B-hn42 and MLG-B-hn43 were genetically separated from Old World monkeys, ring-tailed lemurs and humans (Fig. 1).

Discussion
Data from this study suggests that crab-eating macaques in Hainan, China are commonly infected with E. bieneusi. In this study, the detection rate of E. bieneusi in these animals was 31.7% (461/1452). This is higher than the reported detection rates in NHPs in various countries [52][53][54][55]. Similarly, it is mostly higher than detection rates in studies of E. bieneusi in NHPs in China [18,19,35,50,[56][57][58][59]. Many of the studies reporting low detection rates of E. bieneusi in NHPs were performed using wild, captive and zoo animals [19,35,52,54,55,57]. The present report represents the first one carried out using a large number of laboratory NHPs.
Crab-eating macaques are apparently infected with zoonotic E. bieneusi genotypes. In this study, all nine E. bieneusi genotypes detected in these animals belong to the zoonotic Group 1 (Table 1). Among them, Type IV, D, Peru8, Peru11 and PigEbITS7 are known human pathogens in many countries [13, 15, 26-29, 31, 60, 61]. Others such as CM1, CM2 and CM3 have been thus far only found in NHPs in China [18,19,62], but this is probably because only a small number of studies have been performed on human E. bieneusi infection within the country. The remaining genotype, Pongo2, was reported in China for the first time in this study. This genotype was initially seen in orangutans in Indonesia, indicating that it has the capability to infect a broad range of NHPs [53]. Laboratory crab-eating macaques are also apparently common hosts of G. duodenalis. In this study, the detection rate of G. duodenalis was 32.3% (469/1452) in Hainan, China. This confirms the prevalence of this pathogen in NHPs in various countries [36,37,63,64] and different areas within China [34,35,38,44,50,57,65]. The very high detection rate of G. duodenalis as well as E. bieneusi in the present study could be attributed to the intensive farming of NHPs in this study, which congregates numerous susceptible individuals in confined areas.
To date, assemblages A, B and E of G. duodenalis have been reported in NHPs [34,38,50,54,[65][66][67]. Among them, assemblage B is the most common genotype in different species of NHPs, including various monkeys, lemurs, gibbons, chimpanzees and gorillas [34, 36-38, 50, 57, 63-65]. It is also common in humans in both developing and industrialized countries, and is more common than the other major human-pathogenic genotype, assemblage A [2,12,28]. In this study, assemblage B was the only G. duodenalis genotype in the crab-eating macaque. This could have been due to the confined nature of animals in the facility, which limits the introduction of new G. duodenalis genotypes. Nevertheless, a high genetic heterogeneity of assemblage B was seen in animals in the laboratory facility, as revealed by subtype analysis at three genetic loci.
The zoonotic potential of G. duodenalis assemblage B in crab-eating macaques was supported by subtype analysis of specimens. Of the eight subtypes detected at the tpi locus, B-sh01 (JX994245) and B6 (GU564284) have been previously found in humans [28,68]. Similarly, among the six known subtypes at the gdh locus, B-sh03 (JX994233) and B-VANC/87/UBC/8 (KM190714) have been previously found in humans [28,69]. Likewise, among the four known bg subtypes, B-Egyh8 (MG736242) has been previously found in humans [69]. Therefore, many of the known subtypes of G. duodenalis obtained in this study at individual genetic loci had been previously found in humans, supporting the human-pathogenic potential of the assemblage B in crab-eating macaques.
Nevertheless, there appears to be some host segregation within assemblage B of G. duodenalis [34]. In this study, MLG analysis has identified 53 MLGs. Phylogenetic analysis showed that only MLG-B-hn31 is genetically similar to MLGs of assemblage B isolates from humans in China and Sweden [28,51]. In contrast, most of other MLGs were genetically related to assemblage B isolates in pig-tailed macaques, rhesus macaques, golden monkeys, yellow baboons and green monkeys, all common Old-World monkeys. They were different from MLGs in ring-tailed lemurs, which are natives of the island nation Madagascar and evolved independently from monkeys and apes.

Conclusions
In this study, we have shown a frequent occurrence and high genetic diversity E. bieneusi and G. duodenalis subtypes in crab-eating macaques in one commercial laboratory animal facility in Hainan, China. Most of the E. bieneusi genotypes and G. duodenalis assemblage B subtypes are potentially zoonotic. Additional genetic characterizations of these pathogens at other genetic loci, including more conservative ones for G. duodenalis, are needed to better understand the transmission of these pathogens and possible occurrence of host segregation within G. duodenalis assemblage B. Measures should be implemented at the commercial facility to reduce the transmission of enteric parasites.