Introduction

Enterocytozoon bieneusi and Giardia duodenalis are two common intestinal protists in humans, farm animals, domestic pets, and wild mammals (Ryan et al. 2019; Wang et al. 2018). Humans are infected by the two pathogens through a similar fecal–oral transmission route, after exposures to infected persons (anthroponotic transmission) or animals (zoonotic transmission), or consumption of contaminated food or water (food-borne or water-borne transmission). Recently, concerns have been raised on the potential of cattle as a source for zoonotic transmission of the two pathogens (Abeywardena et al. 2015; Feng et al. 2019).

Enterocytozoon bieneusi is the most prevalent human-pathogenic species of microsporidia (Didier and Weiss 2011). It forms at least 11 genotype groups with different host preferences, with Group 1 containing mostly zoonotic genotypes and the remaining groups being largely host adapted (Li et al. 2019; Li and Xiao 2019). The detection of E. bieneusi in cattle was first reported in 2000, with genotypes I and J being identified initially (Rinder et al. 2000). Since then, more than 50 E. bieneusi genotypes have been detected in cattle, mostly belonging to Group 2 (Zhang et al. 2018). Among them, some genotypes in Groups 1 (D, Type IV, Peru 6, CHN4, and EbpA) and 2 (J, I, BEB4, BEB6, CHN3, CS-4, and EbpC) have been identified in humans, indicating the zoonotic potential of bovine E. bieneusi genotypes (Jiang et al. 2015; Yu et al. 2019).

Similarly, molecular analyses have identified at least eight assemblages of A–H within G. duodenalis. Among them, zoonotic assemblages A and B infect humans and various animals, whereas host-adapted assemblages C–H each infects a specific group of animals (Feng and Xiao 2011). In cattle, assemblage E is the predominant G. duodenalis genotype found in most areas worldwide (Abdel-Moein and Saeed 2016; Hogan et al. 2014; Matsuura et al. 2017; Nguyen et al. 2016; Zhong et al. 2018), although zoonotic assemblages A and B were occasionally detected in some studies (Bartley et al. 2018; Lee et al. 2018; Wegayehu et al. 2016). Recently, a multilocus genotyping approach has been used in characterizations of G. duodenalis infections in cattle (Wang et al. 2014). This has led to the identification of numerous multilocus genotypes (MLGs) within assemblage E (Cui et al. 2018; Feng et al. 2019; Zhong et al. 2018). In some recent studies, assemblage E was found in a few human cases (Abdel-Moein and Saeed 2016; Fantinatti et al. 2016; Zahedi et al. 2017).

Despite the prevalence of E. bieneusi and G. duodenalis in dairy cattle, the clinical significance of two organisms in dairy cattle is not entirely clear. In a few studies, G. duodenalis infections in dairy cattle have been associated with the occurrence of diarrhea (Di Piazza et al. 2013; Toledo et al. 2017). This was supported by results of a case–control study in which a higher infection rate was seen in pre-weaned calves with diarrhea (Mahato et al. 2018). Healthy pre- and post-weaned calves, however, are also frequently infected with G. duodenalis (Feng et al. 2019; Kakandelwa et al. 2016). The association of E. bieneusi infection with the occurrence of diarrhea is even weaker, as most infected dairy cattle do not have diarrhea (Fayer et al. 2007; Tang et al. 2018).

In the present study, we examined the infection patterns, clinical significance, and genetic characteristics of E. bieneusi and G. duodenalis in dairy cattle on a large dairy farm in Jiangsu, China using molecular diagnostic tools.

Materials and methods

Ethics statement

This study was approved by the Ethics Committee of the East China University of Science and Technology. Fecal specimens from dairy cattle were collected with the permission of the farm manager. During the specimen collection, cattle were handled in accordance with the Animal Ethics Procedures and Guidelines of the People’s Republic of China.

Specimen collection

From April 2016 to November 2017, a total of 1366 fecal specimens were collected from dairy cattle on one farm in Jiangsu Province, East China. The farm had ~ 5500 Holstein (> 99%) and Jersey (< 1%) cattle at the time of sampling. The dairy cattle were divided into four age groups: pre-weaned (< 3 months in age, n = 615) calves, post-weaned (3–12 months in age, n = 508) calves, adults (> 24 months in age, n = 192), and those without age information (n = 51). Diarrhea was observed in 240 sampled animals, with the majority in pre-weaned (n = 158) and post-weaned (n = 74) calves, while the remaining eight specimens were from cattle of unknown age. Fecal specimens were collected directly from the rectum of cattle into 50 mL centrifuge tubes by using disposable gloves, transported to the laboratory with ice packs, and stored in 2.5% potassium dichromate at 4 °C before DNA extraction. Each animal was sampled only once during the study period.

DNA extraction and PCR analysis

After being washed three times with distilled water by centrifugation at 2000×g for 10 min, 200 mg of fecal specimen was used in DNA extraction by using the FastDNA SPIN Kit for Soil (MP Biomedical, Santa Ana, CA, USA). The DNA was stored at − 80 °C before PCR analyses. For the detection and genotyping of E. bieneusi, a rRNA gene fragment covering the entire internal transcribed spacer (ITS) was amplified by nested PCR (Sulaiman et al. 2003). In contrast, three nested PCR assays targeting the β-giardin (bg), glutamate dehydrogenase (gdh), and triosephosphate isomerase (tpi) genes were used in detecting and genotyping G. duodenalis (Caccio et al. 2008). Each specimen was analyzed at each genetic locus at least twice. DNA of genotype PtEb IX from dogs was used as the positive control in PCR analysis of E. bieneusi, while DNA of assemblage C from dogs was used as the positive control in PCR analysis of G. duodenalis, with reagent-grade water being used as the negative control. The secondary PCR products generated were examined by electrophoresis in 1.5% agarose gels stained with ethidium bromide.

Sequence analysis

All positive secondary PCR products were sequenced in both directions on an ABI 3730 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA) at the BioSune Biotechnology Company (Shanghai, China). The obtained sequences were assembled using ChromasPro 2.1.6 (http://technelysium.com.au/ChromasPro.html), edited using BioEdit 7.1.3 (http://www.mbio.ncsu.edu/BioEdit/bioedit.html), and aligned with each other and reference sequences in the GenBank database using ClustalX 1.81 (http://clustal.org) to determine the genotype identity of E. bieneusi and G. duodenalis. For G. duodenalis, a previous nomenclature system was used in naming G. duodenalis assemblage E subtypes at each locus (Wang et al. 2017). Genetic relationships among multilocus genotypes (MLGs) of assemblage E were accessed by using eBURST v3 (http://eburst.mlst.net).

Representative nucleotide sequences generated in this study were deposited in GenBank under the accession numbers MK890212 for E. bieneusi and MK890213–MK890222 for G. duodenalis.

Statistical analysis

The χ2 test implemented in SPSS Statistics 21.0 (IBM Corp., New York, NY, USA) was used to compare differences in infection rates among age groups or clinical signs. Differences were considered significant at P < 0.05. Odds ratios with 95% confidence intervals (CIs) were calculated for the occurrence of each pathogen in animals with diarrhea.

Results

Infection patterns of E. bieneusi and G. duodenalis in dairy cattle

Of the 1366 fecal specimens collected from dairy cattle on the study farm, 177 (13.0%) were positive for E. bieneusi in ITS PCR. By age, the infection rate in pre-weaned calves (21.0% or 129/615; P < 0.0001) was significantly higher than rates in post-weaned calves (6.3% or 32/508) and adult cattle (2.6% or 5/192) (Table 1). Among pre-weaned calves (Fig. 1a), animals of 7–12 weeks of age had a much higher E. bieneusi infection rate than younger calves (27.4–53.5% vs. 0–9.8%); the overall infection rates were highly significant between the two age groups (χ2 = 124.39, P < 0.0001). In post-weaned calves, the E. bieneusi infection rate was the highest at 3 months of age (18.8% or 9/48). Afterward, the infection rate declined gradually with increased age (Fig. 1b).

Table 1 Occurrence of Enterocytozoon bieneusi and Giardia duodenalis in dairy cattle in Jiangsu, China
Full size table
Fig. 1
figure 1

Infection rates of Enterocytozoon bieneusi and Giardia duodenalis in pre-weaned (a) and post-weaned (b) dairy calves on a farm in Jiangsu, China by age

Full size image

Giardia duodenalis was detected in 281 (20.6%) of the 1366 specimens based on PCR positivity at any of the three genetic loci. The highest infection rate was 38.9% (239/615) in pre-weaned calves, followed by post-weaned calves (7.5% or 38/508) and adults (0.5% or 1/192) (Table 1). Pairwise differences in infection rates were all highly significant among the three age groups (P < 0.001). Similar to the E. bieneusi infection, increasing G. duodenalis infection rates were observed with increased age in pre-weaned calves, with infection rates at 6–12 weeks of age (43.9–86.3%) being significantly higher than 1–5 weeks (2.4–15.2%; χ2 = 233.65, P < 0.0001) (Fig. 1a). In post-weaned calves, the infection rate of G. duodenalis was the highest at 3 months of age (47.9% or 23/48). Afterward, G. duodenalis infection rates decreased rapidly (Fig. 1b).

In both pre- and post-weaned calves, infection rates of E. bieneusi or G. duodenalis in calves with diarrhea were significantly higher than those without diarrhea (P < 0.01) (Table 2). Pre-weaned calves with diarrhea were 2.5 (CI 1.7–3.8) or 2.1 (CI 1.4–3.0) times more likely to be infected with E. bieneusi or G. duodenalis, respectively. This was also the case in post-weaned calves, with animals experiencing diarrhea 2.9 (CI 1.3–6.5) or 3.0 (CI 1.5–6.3) times more likely to be E. bieneusi positive or G. duodenalis positive, respectively (Table 2).

Table 2 Association between Enterocytozoon bieneusi or Giardia duodenalis infection and occurrence of diarrhea in dairy calves in Jiangsu, China
Full size table

Co-infection with E. bieneusi and G. duodenalis was detected in 98 animals including 93 pre-weaned and 5 post-weaned calves, with the infection rate in pre-weaned calves (15.0% or 93/615; P < 0.0001) being significantly higher than in post-weaned calves (1.0% or 5/508). Co-infection was not observed in adult cattle. Among pre-weaned calves, animals with diarrhea were 2.7 (1.7, 4.3) times more likely to have co-infection of both pathogens than those without diarrhea (Table 2).

E. bieneusi genotypes

Five genotypes were detected among 177 E. bieneusi-positive specimens, including J (n = 138), I (n = 21), BEB4 (n = 10), Type IV (n = 1), and a novel genotype named CHC17 (n = 1) (Table 1). Concurrence of two genotypes was detected in six animals, including J and I (n = 5) and J and BEB4 (n = 1). The ITS sequences generated from genotypes J, I, BEB4, and Type IV were identical to the GenBank reference sequence AF135837, AF135836, AY331008, and AF242478, respectively. Genotype CHC17 (MK890212) had two nucleotide substitutions compared with the BEB4 genotype. All the genotypes except Type IV belonged to Group 2.

Genotype J was the most common one identified in all three age groups. In contrast, genotype I was only detected in pre-weaned calves, genotypes Type IV and CHC17 were only found in adult animals, while BEB4 was detected in both pre- and post-weaned calves (Table 1). Within pre-weaned calves, genotype I was only detected in calves of 6–11 weeks (Fig. 2).

Fig. 2
figure 2

Distribution of Enterocytozoon bieneusi genotypes in pre-weaned dairy calves on a farm in Jiangsu, China by age

Full size image

G. duodenalis assemblages and subtypes

Of the 281 G. duodenalis-positive specimens, 258 were positive in bg PCR, 224 in gdh PCR, and 193 in tpi PCR. Among them, only assemblages E (n = 278) and A (n = 2) were identified, with the concurrence of A and E in one animal (Table 1).

Among the three assemblage A-positive specimens, sequence analysis of the bg, gdh, and tpi genes showed the presence of subtype A5 in one specimen, A1 + A5 in one specimen, and A1 + E in one specimen, respectively. The sequences produced were identical to the reference sequences of the bg (KR075938), gdh (KR075940), and tpi (JX845464) genes. Only one of the three assemblage A specimens was positive by PCR at all three genetic loci. Based on the A1 and A5 subtype identity at these loci, the specimen apparently had the sub-assemblage AI (Table S1).

A higher sequence diversity was found within assemblage E at all three genetic loci (Table S2). At the bg locus, 13 subtypes were identified, eight of which were reported previously: E3 (n = 117), E2 (n = 53), E1 (n = 26), E5 (n = 9), E6 (n = 8), E29 (n = 7), E8 (n = 6), and E11 (n = 6); the remaining five represented novel subtypes, including E28 (n = 14), E18 (n = 4), E30 (n = 4), E27 (n = 1), and E32 (n = 1) (MK890213-MK890217). At the gdh locus, seven subtypes were identified, including five known subtypes of E3 (n = 91), E1 (n = 75), E11 (n = 7), E2 (n = 6), and E15 (n = 2), and two novel subtypes of E32 (n = 5) and E33 (n = 2) (MK890218, MK890219). Double peaks were observed at nucleotide position 95 in gdh sequences from 34 specimens. At the tpi locus, 10 subtypes were detected including seven known subtypes and three novel ones. The former included E34 (n = 112), E17 (n = 22), E1 (n = 11), E53 (n = 11), E50 (n = 6), E54 (n = 5), and E52 (n = 2), whereas the latter included E55 (n = 16), E56 (n = 4), and E49 (n = 3) (MK890220–MK890222).

Multilocus genotypes of assemblage E

A total of 143 assemblage E-positive specimens were successfully subtyped at all three loci, forming 65 assemblage E MLGs (Table S1). A concurrent infection of A1 and E was excluded in the MLG analysis, as well as 31 additional specimens that showed the presence of double peaks at the gdh locus. All 143 specimens harboring 65 assemblage E MLGs were collected in April 2016. In pre-weaned calves, 59 assemblage E MLGs were identified with MLG-E1 (n = 26) and MLG-E2 (n = 12) as the most common MLGs. In post-weaned calves, 11 assemblage E MLGs were identified, including MLG-E1 in two specimens, and MLG-E6, MLG-E9, MLG-E17, MLG-E19, MLG-E25, MLG-E28, MLG-E43, MLG-E48, MLG-E60, and MLG-E63 each in one specimen.

The 65 assemblage E MLGs mostly belonged to two clusters in the eBURST analysis. Among them, two most common MLGs, MLG-E1 (n = 28) and MLG-E2 (n = 12), were identified as the primary and subgroup founders, respectively. They, however, were genetically related, differing only at one of the three genetic loci (Fig. 3).

Fig. 3
figure 3

Relationships among 65 multilocus genotypes (MLGs) of Giardia duodenalis assemblage E identified in dairy cattle by eBURST analysis of allelic data. Each dot represents a MLG; the size is proportional to the sample size of the MLG. Blue dot is the primary founder and yellow dots are subgroup founders. MLGs related by single-locus variants are linked by lines. The allelic profile of each MLG is indicated in Table S1

Full size image

Discussion

In this study, the overall infection rate of E. bieneusi was 13.0% in dairy cattle in Jiangsu Province, East China. This is within the range of reported E. bieneusi infection rates of 2.0–46.8% in dairy cattle in China (Qiu et al. 2019). By age, the E. bieneusi infection rate was significantly higher in pre-weaned calves than in post-weaned calves. This is similar to some previous studies in other provinces in China, which mostly reported a slightly higher rate in pre-weaned calves (Li et al. 2016b; Qi et al. 2017). In contrast, a lower E. bieneusi infection rate was observed in pre-weaned calves than in post-weaned calves in two other studies in China (Hu et al. 2017; Yu et al. 2019). Within pre-weaned calves, the E. bieneusi detection rate (21.0%) on the study farm is higher than rates reported in most other studies in China (Hu et al. 2017; Qi et al. 2017; Wang et al. 2016). In addition, E. bieneusi infection rates increased with age even in pre-weaned calves, with peak infection in calves of 7–12 weeks. This is largely in agreement with the age pattern described in two previous studies in the USA and China (Santin and Fayer 2009; Tang et al. 2018).

As in most previous studies in China and Egypt (Cui et al. 2018; Liu et al. 2015; Naguib et al. 2018), significant declining of G. duodenalis infection rates was also observed with increased age in this study, with the highest in pre-weaned calves (38.9%). Within pre-weaned calves, G. duodenalis infection rates were extremely low during the first 5 weeks and thereafter increased rapidly at 6–12 weeks of age, which is largely in concordance with observations in some earlier studies in the USA, New Zealand, and two recent surveys in Shanghai and Guangdong, China (Feng et al. 2019; Trout et al. 2004; Wang et al. 2017; Winkworth et al. 2008).

In both pre- and post-weaned calves in this study, the odds of E. bieneusi infection in animals with diarrhea were 2.5–2.9 times higher than in those without diarrhea, which is contrary to previous findings elsewhere that E. bieneusi infection was not associated with the occurrence of any clinical signs in infected dairy cattle (Qiu et al. 2019; Santin and Fayer 2011). Similarly, the chance for G. duodenalis infection was 2.1–3.0 times higher in calves with diarrhea. This observation is consistent with the results in some previous studies in Italy, Brazil, and Nepal (Di Piazza et al. 2013; Mahato et al. 2018; Toledo et al. 2017), further supporting the association between G. duodenalis infection and occurrence of diarrhea in dairy calves.

Similar to most previous studies in China (Hu et al. 2017; Li et al. 2016b; Qi et al. 2017), E. bieneusi genotype J was the dominant one in cattle of all ages in Jiangsu. Interestingly, genotype I, the second most common genotype in this study, was only detected in pre-weaned calves of 6–11 weeks. This is different from the age-related distribution of E. bieneusi genotypes in dairy cattle elsewhere in the world, where genotype I was commonly detected in all age groups (da Silva Fiuza et al. 2016; Santin and Fayer 2009; Zhang et al. 2018). One study in the USA found that genotype I was detected primarily in post-weaned calves (Santin and Fayer 2009).

Assemblage E was found to be the dominant G. duodenalis genotype in dairy cattle in Jiangsu. This is consistent with most previous studies in China and elsewhere (Bartley et al. 2018; Cui et al. 2018; Hu et al. 2017; Naguib et al. 2018). In addition, sub-assemblage AI was detected in three specimens in this study. Although both are more commonly seen in animals (Feng and Xiao 2011), G. duodenalis infection in dairy cattle may still present a potential risk for public health, as sub-assemblage AI and assemblage E have been found in humans in some recent studies (Abdel-Moein and Saeed 2016; Fantinatti et al. 2016; Zahedi et al. 2017).

A high genetic diversity was observed within assemblage E in this study, with 65 MLG-Es forming two major clusters on a single dairy farm. Previously, multiple MLGs of assemblages E had been reported in dairy cattle in China (Cui et al. 2018; Zhong et al. 2018). Those studies, however, were mostly conducted on multiple farms. The occurrence of multiple MLGs on a single farm suggests the likely occurrence of intra-assemblage genetic recombination, which could be responsible for the high genetic heterogeneity within assemblage E (Aguiar et al. 2016; Wang et al. 2017). In addition, the most common MLGs (MLG-E1 and MLG-E2) in the study were primary and subgroup founders in eBURST analysis, supporting the likely occurrence of genetic recombination of the two dominant MLGs. This needs substantiation by using population genetic analyses.

In conclusion, E. bieneusi and G. duodenalis are common in dairy cattle in Jiangsu, China, with both more prevalent in pre-weaned calves than in post-weaned and adult cattle. The pathogenicity of E. bieneusi and G. duodenalis was also observed in both pre- and post-weaned calves, with the association between E. bieneusi infection and occurrence of diarrhea being described in dairy calves for the first time. Genetic recombination is suspected to be the cause of the high genetic heterogeneity within G. duodenalis assemblage E. Nevertheless, more advanced epidemiological studies are required to characterize the transmission and clinical significance of E. bieneusi and G. duodenalis on dairy farms.