Genetic diversity within dominant Enterocytozoon bieneusi genotypes in pre-weaned calves

Background Cattle are commonly infected with the microsporidian parasite Enterocytozoon bieneusi. Sequence characterization of E. bieneusi in these animals at the ribosomal internal transcribed spacer (ITS) locus had identified I, J and BEB4 as the dominant genotypes. However, current studies on E. bieneusi in dairy cattle are mostly on infection rates and genotype distribution. This study aims to examine the intragenotypic diversity within dominant E. bieneusi genotypes in pre-weaned dairy calves in Shanghai, China. Methods Enterocytozoon bieneusi genotypes and subtypes were identified by PCR sequence analysis of ITS and multilocus sequence typing (MLST), based on material from farms. Chi-square test was used to examine differences in E. bieneusi infection rates between farms or age groups. Results The overall infection rate of E. bieneusi was 26.5% (214/809), ranging from 12.6% (Farm 5) to 38.5% (Farm 4). Infection rates increased with age during early life, with the peak infection rate (43.0%; 43/100) occurring at six weeks. Four genotypes were present, including J (n = 145, 67.8%), BEB4 (n = 59, 27.6%), CHN4 (n = 4, 1.9%) and CHN15 (n = 1, 0.5%), with the former two belonging to Group 2 and the latter two belonging to Group 1. Differences were detected in the distribution of the dominant genotypes J and BEB4 among five study farms. Altogether, 10 multilocus genotypes (MLGs) were identified in the two dominant ITS genotypes, including MLG-J1 to MLG-J8 of genotype J and MLG-B1 to MLG-B2 of genotype BEB4. MLG-B1 and MLG-B2 were recovered in Farms 1, 2 and 5, whereas MLG-J1 to MLG-J5 and MLG-J6 to MLG-J8 were found in Farms 3 and 4, respectively. Conclusions There is extensive genetic heterogeneity within the dominant E. bieneusi genotypes J and BEB4 in dairy calves in Shanghai, China, and MLST should be used in molecular epidemiological studies of E. bieneusi in cattle.

Based on sequence analysis of the internal transcribed spacer (ITS) of the rRNA gene (~243 bp), more than 200 E. bieneusi genotypes have been identified [1,8]. Phylogenetic analyses revealed that they belong to nine groups [9,10]. Group 1, which contains most genotypes found in humans, is considered a zoonotic group, with the remaining groups being largely host-specific. To date, over 40 E. bieneusi genotypes have been detected in cattle, most of which belong to Group 2 [11][12][13]. Among them, at least 15 genotypes, including eight genotypes in Group 1 and seven genotypes in Group 2, have been reported in humans [11,12,14], suggesting that cattle may be potential reservoirs for human infections.
Genotypes I, J and BEB4 are common E. bieneusi genotypes found in pre-weaned dairy calves worldwide [8,11,12,[15][16][17][18][19][20][21] and have been further detected in at least 13 human cases [6,22]. However, current studies on E. bieneusi in dairy calves are mostly on infection rates and genotype distribution. Little is known about the age distribution of E. bieneusi infection in pre-weaned dairy calves. In addition, genetic diversity within the dominant E. bieneusi genotypes has not been examined thoroughly using advanced molecular diagnostic tools such as multilocus sequence typing (MLST).

Specimen collection
From April 2015 to March 2016, 809 specimens, each of approximately 25 g fresh fecal material, were collected from pre-weaned Holstein calves in five farms in Shanghai, China. These farms are located in Fengxian (Farms 1, 2, 3 and 4) and Jinshan (Farm 5), two neighboring districts in suburban Shanghai. They were ranked A to E by combined farm quality score based on hygiene status, animal density, and facility condition, with A representing "excellent" and E representing "poor" [35]. Each farm was visited 2-5 times at 2-3 months intervals, for a total of five times for Farm 3, four times for Farm 1, and twice for Farms 2, 4 and 5. These fecal specimens were collected directly from the rectum by using disposable gloves into 50 ml centrifuge tubes, transported to the laboratory in coolers with ice packs, and stored in 2.5% potassium dichromate at 4°C before DNA extraction.

DNA extraction
Genomic DNA was extracted by using the Fast DNA SPIN Kit for soil (MP Biomedical, Santa Ana, CA, USA) from approximately 200 mg of each fecal specimen, which was washed three times with distilled water by centrifugation at 2000× g for 10 min. The obtained DNA was stored at -20°C until being used in PCR analysis.

PCR analysis
The occurrence and genotype distribution of E. bieneusi were determined by PCR and sequence analyses of the ITS as previously described [36]. For subtyping the dominant E. bieneusi ITS genotypes J and BEB4, the MLST technique targeting microsatellite loci MS1, MS3 and MS7 and minisatellite locus MS4 was used [24]. In a pre-study analysis, 90 of 98 E. bieneusi-positive specimens yielded the expected PCR products at the MS3 locus. Therefore, PCR analysis of the MS3 locus was used for screening of the 204 specimens positive for ITS genotypes J and BEB4. Among them, 84 MS3-positive specimens of five different MS3 subtypes were further analyzed at the MS1, MS7 and MS4 loci. Duplicate nested PCR was used in the analysis of the specimens at each genetic locus. The secondary PCR products obtained were identified by agarose gel electrophoresis.

Statistical analysis
Differences in E. bieneusi infection rates between farms or age groups were examined by using the Chi-square test implemented in SPSS Statistics v.20.0 for Windows (IBM Corp., New York, NY, USA). Differences were considered significant at P ≤ 0.05.

ITS genotypes of E. bieneusi by farm
DNA sequencing of ITS PCR products was successful for 209 of 214 PCR-positive specimens. The remaining five specimens generated ITS sequences with underlying signals because of the presence of mixed E. bieneusi genotypes. Four E. bieneusi genotypes were identified among the 209 successfully genotyped specimens, including J, BEB4, CHN4 and CHN15, with the latter being identical to an unnamed genotype (JF909995) in wastewater from Tunisia [37]. The dominant genotypes in calves were J (n = 145, 67.8%) and BEB4 (n = 59, 27.6%), which both belong to Group 2. The remaining two genotypes,  (1)  Farm ranks A-E were ranking scores used to evaluate the hygiene status, animal density and facility condition, with A representing "excellent" and E representing "very poor" (see [35] for details) CHN4 and CHN15, were seen in only four (1.9%) and one (0.5%) E. bieneusi-positive calves, respectively ( Altogether, there were eight, five, two and four subtypes at the MS1, MS3, MS4 and MS7 loci, respectively. The diversity of subtype was different between genotypes J and BEB4, with seven, five, two, and three subtypes being detected in genotype J, compared to one, one, one, and two subtypes in BEB4, respectively. As expected, the dominant subtype at each locus was different between genotypes J and BEB4 (Table 3). At the MS1 locus, the dominant subtype was MS1-3 in genotype J (28/53), while it was MS1-1 in BEB4 (29/ 31). At the MS3 locus, the dominant subtype was MS3-1 in genotype J (29/53), while it was MS3-2 in BEB4 (31/31). At the MS4 locus, MS4-1 (24/53) and MS4-2 (24/53) were the dominant subtypes in genotype J, while MS4-2 was the only subtype in BEB4 (30/31). At the MS7 locus, the dominant subtype was MS7-1 both in genotypes J (18/53) and BEB4 (12/31).  The distribution of subtypes in genotype J differed between Farms 3 and 4. The main subtypes on Farm 3 were MS1-3, MS3-1, MS4-2 and MS7-1 at the four loci, while the main subtypes on Farm 4 were MS1-8, no MS3 amplification, MS4-1 and MS7-1. In contrast, the main subtypes in genotype BEB4 at the four loci were the same (MS1-1, MS3-2, MS4-2 and MS7-1) on all BEB4-positive farms (Farms 1, 2 and 5).

Discussion
In the present study, E. bieneusi was found in 26.5% ( [8,11,[18][19][20]38]. Variations in E. bieneusi infection rates in preweaned dairy calves among studies could be due to differences in detection methods, age and management of animals, and climate. Pre-weaned dairy calves appear to have peak E. bieneusi infection around 4-7 weeks of age. In this study, although calves were infected by E. bieneusi from one to nine weeks of age, E. bieneusi occurrence in newborn animals increased gradually with age, with the peak infection rate (43.0%) being reached at six weeks of age. Thus, the mean infection rate at 4-7 weeks of age was significantly higher than at 1-3 weeks. This agrees with the result of the only other study of the age pattern of E. bieneusi infection in pre-weaned dairy calves in the USA [8]. Previously in China, only slightly higher E. bieneusi infection rates were reported in pre-weaned dairy calves than in post-weaned dairy calves: 17.7% and 15.5% in Xinjiang [16], 29.3 and 23.9% in Henan and Ningxia [15], 10.0 and 7.3% in Henan and Shandong [21], and 7.4 and 4.3% in Northeast China [12], respectively. Lumping all pre-weaned calves into one group could be responsible for the small differences in E. bieneusi infection rates between pre-weaned and postweaned calves.
Results of the MLST analysis support the existence of differences in the transmission of the two dominant E. Abbreviation: n novel subtype bieneusi ITS genotypes. All dominant subtypes in genotype BEB4 at the four individual loci, such as MS1-1, MS3-2, MS4-2 and MS7-1, and the most common MLG (MLG-B1) in genotype BEB4 were present on all farms with ITS genotype BEB4 (Farms 1, 2 and 5). In contrast, the dominant subtypes of ITS genotype J at each locus were different between Farms 3 and 4. In fact, genotype J on Farm 4 was so divergent from the one on Farm 3 at the MS3 locus that none of the 15 genotype J-positive specimens from Farm 4 generated the expected MS3 PCR product whereas 85 of the 129 genotype J-positive specimens from Farm 3 generated it. The most common genotype J MLG (MLG-J2) was only seen on Farm 3, and all other genotype J MLGs identified in this study, were exclusively present on Farm 3 or 4. Therefore, although all five farms are owned by the same dairy enterprise and located in two neighboring districts of suburban Shanghai, there is extensive genetic heterogeneity within the dominant E. bieneusi genotypes, especially ITS genotype J.

Conclusions
Results of this study indicate that E. bieneusi infection is common in pre-weaned dairy calves in suburban Shanghai, China, with animals of 4-7 weeks of age having the highest occurrence of the pathogen. Data of MLGs among farms suggest that there are apparent differences in the distribution of dominant E. bieneusi genotypes among farms and extensive genetic heterogeneity within ITS genotypes. Molecular epidemiologic studies involving advanced pathogen characterization should be conducted to improve understanding of the population genetics of E. bieneusi in cattle and relationship among infection rates, age-associated infection patterns, genotype distribution, farm management, and transmission of the pathogen.