The presence of Giardia intestinalis in donkeys, Equus asinus, in China.

BACKGROUND: Giardia intestinalis is one of the most important zoonotic enteric parasites. As no information regarding prevalence and genotype of G. intestinalis in donkeys (Equus asinus) in China is available, 181 faecal samples from 48 donkeys from Jilin Province, from 104 from Shandong Province and from 29 from Liaoning Province were examined between May and December 2015.
FINDINGS: Twenty-eight (15.47%) out of 181 donkey samples were tested G. intestinalis-positive by nested amplification of the triosephosphate isomerase (tpi) gene. The prevalence in different regional groups varied from 10.42 to 18.27%. The prevalence in adult and young donkeys was 14.29 and 22.92%, respectively. Otherwise, the prevalence was 11.69% in summer and 18.27% in winter. However, no statistically significant differences were found in relation to region or age group. Sequence analysis of the tpi, glutamate dehydrogenase (gdh) and beta giardin (bg) loci identified 4, 1 and 3 subtypes of assemblage B, respectively. Moreover, four novel multilocus genotypes (MLGs novel-1 to novel-4) were identified in assemblage B.
CONCLUSIONS: This first report of G. intestinalis in donkeys in China indicates that further studies of nation-wide molecular epidemiology and geographical distribution of Giardia in donkeys are warranted. Effective strategies should be implemented to control G. intestinalis infection in donkeys, other animals and humans.

Background

Giardia intestinalis is the only species of Giardia which is found in human beings [1-4]. It is not only distributed worldwide, but also can infect many vertebrates [5, 6]. Eight assemblages (A-H) have been identified within G. intestinalis [6-8]. Of these, assemblages C-H seem to be animal-specific [9], but assemblages A and B can infect humans and a wide range of non-human hosts [6, 10]. Diarrhea is the main symptom of giardiasis [11] and transmission is mainly through ingestion of Giardia cysts in contaminated food or water [12]; approximately 2.8 × 108 cases of human giardiasis are reported world-wide per year, and the majority of them are reported in developing countries [6]. In view of such a serious situation, giardiasis has attracted considerable attention around the world. Although G. intestinalis infections have been reported in humans and a variety of animal species [13-16], there is little information in donkeys (Equus asinus).

Giardia intestinalis infection in horses has been reported in many countries around the world including China [17-21]. The donkey belongs to the genus Equus. It is an important edible animal species and used in Chinese traditional medicine, and is closely related to horses. Because they are maintained in a close association with their owners and veterinary personnel, donkeys are the important reservoirs for transmission of pathogens (such as Cryptosporidium hominis and Toxoplasma gondii) to humans and other animals [22, 23]. To determine whether donkeys are hosts of G. intestinalis, we conducted a study on the prevalence and genotypes of G. intestinalis in donkeys in Jilin, Liaoning and Shandong Provinces, China.

Methods

Collection and preparation of faecal samples

A total of 181 donkey faecal samples (48 from Jilin, 27 from Liaoning and 104 from Shandong) were collected from three provinces, in northeastern and eastern China, between March and December 2015. Each of the fresh faecal samples was collected into sterile gloves separately after its defecation onto the ground, placed into box with ice and transported to the laboratory immediately. Genomic DNA was extracted directly from each faecal sample using the Stool DNA kit (OMEGA, Norcross, Georgia, USA) according to the manufacturer’s instructions. Genomic DNA was stored at -20 °C until PCR amplification.

PCR amplification

The prevalence and genotypes of G. intestinalis were determined by the nested PCR amplification of the triosephosphate isomerase (tpi) gene, beta giardin (bg) gene and glutamate dehydrogenase (gdh) genes as described by Zhao et al. [24]. The primers and their annealing temperatures are listed in Table 1. Positive and negative controls were included in each amplification. Amplification products were observed under UV light after electrophoresis in 1.5% agarose gels containing GoldView™ (Solarbio, Beijing, China).

Table 1

Primers used in the study, annealing temperatures used in the PCRs and expected sizes of the PCR products

Gene	Primer	Sequence (5'-3')	Annealing temperature (°C)	Fragment length (bp)	Reference
tpi	F1	AAATIATGCCTGCTCGTCG	55	530	[24]
	R1	CAAACCTTITCCGCAAACC
	F2	CCCTTCATCGGIGGTAACTT	55
	R2	GTGGCCACCACICCCGTGCC
gdh	F1	TTCCGTRTYCAGTACAACTC	50	530	[24]
	R1	ACCTCGTTCTGRGTGGCGCA
	F2	ATGACYGAGCTYCAGAGGCACGT	65
	R2	GTGGCGCARGGCATGATGCA
bg	F1	AAGCCCGACGACCTCACCCGCAGTGC	50	511	[24]
	R1	GAGGCCGCCCTGGATCTTCGAGACGAC
	F2	GAACGAACGAGATCGAGGTCCG	60
	R2	CTCGACGAGCTTCGTGTT

Sequence and phylogenetic analyses

Positive secondary PCR products were sequenced by Genscript Company (Nanjing, China). Bidirectional sequencing was used to confirm the accuracy of the sequences. Sequences with mutations were considered as novel genotypes when confirmed from independent two PCR reactions on the same sample. To identify the assemblages and subtypes, nucleotide sequences were aligned with known reference tpi, gdh and bg gene sequences of G. intestinalis available in GenBank using the BLAST (http://www.ncbi.nlm.nih.gov/BLAST/) and computer program Clustal X 1.83.

Statistical analysis

The relationship between prevalence of G. intestinalis-infected donkeys and different variables including age, geographic origin and seasons were analyzed by Chi-square test using SPSS version 17.0 (SPSS Inc., Chicago, IL, USA) [25]. The results were considered significant statistically if P < 0.05. Odds ratios (ORs) and their 95% confidence intervals (CI) are also given.

Results and discussion

In this study, a total of 28 (15.47%, 95% CI: 10.20–20.74) out of 181 donkey samples were PCR-positive for G. intestinalis (Table 2). The prevalence was 22.92% (95% CI: 11.03–34.81) in young donkeys and 14.29% (95% CI: 8.34–20.23) in adults; no significant difference was observed (χ 2 = 1.90, df = 1, P = 0.17) (Table 2). The prevalence in summer and winter was 11.69% (95% CI: 4.51–18.87) and 18.27% (95% CI: 10.84–25.70), respectively (χ 2 = 1.47, df = 1, P = 0.23) (Table 2). Donkeys from Jilin Province (5/48, 10.42%, 95% CI: 1.78–19.06) had a lower prevalence than those from Shandong Province (19/104, 18.27%, 95% CI: 10.84–25.70) and Liaoning Province (4/29, 13.79%, 95% CI: 1.24–26.34); however, these differences were not significant (χ 2 = 1.62, df = 2, P = 0.44) (Table 2). Moreover, G. intestinalis prevalence in different farms ranged from 6.12 to 29.09% (Table 3). Sequences analysis of the tpi, gdh and bg loci indicated only assemblage B was found in the present study (Additional file 1: Figure S1).

Table 2

Factors associated with prevalence of Giardia intestinalis in donkeys in northern China

Factor	Category	No. tested	No. positive	% (95% CI)	P-value	OR (95% CI)
Region	Jilin Province	48	5	10.42 (1.78–19.06)	0.44	Reference
	Liaoning Province	29	4	13.79 (1.24–26.34)		1.38 (0.34–5.60)
	Shandong Province	104	19	18.27 (10.84–25.70)		1.92 (0.67–5.50)
Age	Adult	133	19	14.29 (8.34–20.23)	0.17	Reference
	Young	48	11	22.92 (11.03–34.81)		1.33 (0.58–3.07)
Season	Summer	77	9	11.69 (4.51–18.87)	0.23	Reference
	Winter	104	19	18.27 (10.84–25.70)		1.69 (0.72–3.97)
Total		181	28	15.47 (10.20–20.74)

Table 3

Giardia intestinalis genotypes identified in donkeys in different farms

Region	Farm ID	Age category (n)	No. positive/No. tested (%)	Genotype (n)
Jilin Province	Farm 1	Young (10); Adult (38)	5/48 (10.42)	BIV-1 (n = 2); BIV-novel-2 (n = 2); BIV-novel-3 (n = 1)
Liaoning Province	Farm 2	Young (6); Adult (23)	4/29 (13.79)	BIV-1 (n = 2); BIV-novel-3 (n = 1); BIV-novel-4 (n = 1)
Shandong Province	Farm 3	Young (14); Adult (35)	3/49 (6.12)	BIV-1 (n = 1); BIV-novel-4 (n = 2)
	Farm 4	Young (18); Adult (37)	16/55 (29.09)	BIV-1 (n = 12); BIV-novel-3 (n = 4)
Total		Young (48); Adult (133)	28/181 (15.47)	BIV-1 (n = 17); BIV-novel-2 (n = 2); BIV-novel-3 (n = 6); BIV-novel-4 (n = 3)

The overall prevalence of G. intestinalis infection in donkeys was 15.47% (28/181, 95% CI: 10.20–20.74), which was higher than that in horses in Xinjiang (1.5%) [17], Brazil (0.5%) [26], foals in Belgium (14.2%) [20], Germany (10%) [20], Greece (11.6%) [20], the Netherlands (11.4%) [20] and Italy (8.6%) [18], but lower than that in horses in Colombia (17.4%) [19]. Previous studies demonstrated that survival of G. intestinalis is more likely to be affected by climate (temperature and relative humidity) [6, 27], so the difference in G. intestinalis prevalence in different regions may be due to different local climatic conditions, as well as the detection methods, sampling time and sample sizes. Moreover, probably because of the smaller sample sizes, seasonal and age-related correlates previously found in cattle [6] and horses [20] were not found in this study.

Assemblages A and B, responsible for the vast majority of human giardiasis [6, 28], and assemblage E, a common assemblage of G. intestinalis in cattle [9], have also been reported in horses [20]. However, perhaps due to the smaller sample sizes, only assemblage B was identified in donkeys in the present study. Assemblage B has a broad host range worldwide [20, 28]. In China, isolates of assemblage B have also been found in non-human primates [2], rabbits [3], horses [17], cattle [29], golden takins (Budorcas taxicolor bedfordi) [24], pet chinchillas (Chinchilla lanigera) [30], captive wildlife [13], sheep [31] and goats [31], suggesting interspecies transmission of G. intestinalis may be commonly occurring in China, and we should pay enough attention to. More importantly, assemblage B was also identified in raw urban wastewater in northern China [32]. Therefore, our results also suggest that donkeys could be a source of giardiasis outbreaks.

Mixed infections of G. intestinalis genotypes have been recorded from a wide range of hosts worldwide [29, 33]. In cases of co-infections, some assemblages may be detected preferentially using a single locus primers [34]; thus PCR amplification of a single locus may not reflect the accurate information to G. intestinalis infection [34, 35]. A multilocus genotype (MLG) method (tpi, gdh and bg loci) has been developed and widely used for detection of G. intestinalis infection [6, 33, 34]. In the present study, 28 G. intestinalis-positive samples were also genotyped based on bg and gdh loci. A total of 28 tpi, 22 bg and 16 gdh gene sequences were obtained, and analysis of these genes revealed only one assemblage (B); however, high genetic polymorphism was observed at these loci within this assemblage (Table 4), implying high genetic diversity of G. intestinalis in donkeys in the investigation regions. At the tpi locus, seven polymorphic sites were found compared with the GenBank reference sequence AY368169 (Table 4), and four different assemblage B subtypes were identified (Table 4, Additional file 1: Figure S1c). These sequences all represented new subtypes (KU892519–KU892522), and showed a 99% similarity with the reference sequence (accession no. AY368169, from wastewater in USA [36]). Only one subtype was found at the gdh locus (Additional file 1: Figure S1b), and the sequence (KU892523) had 99% similarity with the reference sequence available in GenBank with accession number of KR048463 (from a takin in China [24]). Moreover, a total of five SNPs were observed at the bg locus (Table 4), and these sequences (KU892516–KU892518) represented three subtypes (Additional file 1: Figure S1a). Moreover, these subtypes were closely clustered with sub-assemblage BIV (Additional file 1: Figure S1), suggesting that sub-assemblage BIV was the most common subtype in donkeys in the investigated regions. Furthermore, phylogenetic analysis of these isolates showed that these isolates exhibit close relationship with isolates from horses, humans and chinchillas, suggesting that transmission of G. intestinalis may be occurring among these hosts.

Table 4

Variations in tpi, gdh and bg nucleotide sequences among the subtypes of Giardia intestinalis assemblage B

Locus	Subtype (n)	Nucleotide at position							GenBank ID
tpi		10	11	16	182	197	384	525
	Ref. sequence	C	G	–	G	A	G	G	AY368169
	BIV-1a (n = 17)	C	G	–	A	G	A	G	KU892520
	BIV-novel-2 (n = 2)	T	C	G	A	G	A	G	KU892519
	BIV-novel-3 (n = 6)	T	C	–	A	G	A	G	KU892521
	BIV-novel-4 (n = 3)	T	G	–	A	G	A	T	KU892522
gdh		219
	Reference sequence	G							KR048463
	BIV-novel-1 (n = 16)	C							KU892523
bg		14	179	248	446	447
	Reference sequence	G	C	C	A	G			KM926514
	BIV-1b (n = 12)	G	C	C	G	A			KU892517
	BIV-2b (n = 4)	G	T	C	G	A			KU892518
	Bb-7c (n = 6)	A	C	T	G	A			KU892516

aIdentified by Qi et al. [17]

bIdentified by Coronato Nunes et al. [37]

cIdentified by Karim et al. [2]

Furthermore, 10 out of 28 positive isolates were successfully amplified at all three loci. These samples provided four novel MLGs in the assemblage B, namely MLGs novel-1 to novel-4 (Table 5, Fig. 1). Of these, MLG novel-1 (n = 5) was the most prevalent MLG, and responsible for 50% of all MLGs in the present study. These findings suggest a high genetic diversity of this prevalent genotype in donkeys in China, in agreement with previous conclusions that G. intestinalis isolates of the same assemblage may be grouped into distinct MLGs [6, 24].

Table 5

Multilocus characterization of Giardia intestinalis assemblage B isolates from donkeys at tpi, gdh and bg loci

Isolate (n)	Genotype	GenBank ID	MLGs
L7 (5)	BIV-1, BIV-novel-1, BIV-1	KU892520, KU892523, KU892516	novel-1
L13 (1)	BIV-novel-2, BIV-novel-1, BIV-2	KU892519, KU892523, KU892517	novel-2
L64 (2)	BIV-novel-3, BIV-novel-1, Bb-7	KU892521, KU892523, KU892518	novel-3
L93 (2)	BIV-1, BIV-novel-1, Bb-7	KU892520, KU892523, KU892518	novel-4

Fig. 1

Phylogenetic relationships of Giardia intestinalis inferred by the neighbour-joining analysis of concatenated β-giardin, gdh and tpi gene sequences. Multi-locus genotypes from a previous study (Lebbad et al. [38]) are indicated. The bootstrapping was performed using 1000 replicates. Isolates identified in this study are indicated by solid circles

Conclusions

The present study demonstrated the occurrence of G. intestinalis infection in donkeys in China. Sequences analysis suggested that all the G. intestinalis isolates represented assemblage B, with four, one and three subtypes of assemblage B at the tpi gdh and bg loci, respectively. Moreover, four novel MLGs (MLGs novel-1 to novel-4) were identified within assemblage B. The results of the present study not only improve the information of the distribution of G. intestinalis genotypes in China, but also provide the foundation data for preventing and controlling G. intestinalis infection in donkeys, other animals and humans.