First detection and molecular identification of Borrelia species in Bactrian camel (Camelus bactrianus) from Northwest China.

Comprehensive epidemiological surveys for Lyme disease have not been conducted for the Bactrian camel in China. In this study, a total of 138 blood specimens collected from Bactrian camels from Zhangye City in Gansu Province and Yili and Aksu in Xinjiang Province, China, were examined for the presence of Borrelia spp. Species-specificity nested PCR based on the 5S-23S rRNA, OspA, flaB and 16S rRNA genes revealed that the total positive rate of Borrelia spp. was 3.6% (5/138, 95% CI = 0.2-17.9). These results were confirmed by sequence analysis of the positive PCR products or positive colonies. This is the first report of Borrelia pathogens in camels in China. Two Borrelia species that cause Lyme disease and one that causes relapsing fever were identified in the camel blood samples by sequencing. The findings of this study indicate that the Bactrian camel may serve as a potential natural host of Lyme disease and/or relapsing fever in China.

Introduction

Borrelia species are distributed throughout the world and are maintained in nature within various arthropod vectors and mammalian, avian or reptilian hosts (Brisson et al., 2012; Vollmer et al., 2011). In humans, Borrelia spp. are the causative agents of a major disease: Lyme borreliosis (LB) (mainly caused by B. garinii, B. afzelii, and B. burgdorferi sensu stricto). LB is also an important disease of domestic animals and wildlife worldwide. LB-group spirochetes, commonly known as B. burgdorferi s.l., cause one of the most significant natural zoonosis diseases that is carried and transmitted by Ixodes spp. ticks (Wodecka et al., 2010). There are at least 20 genospecies of B. burgdorferi s.l., which are classified based on their genetic differences. More than six of these genospecies have been reported in China (B. burgdorferi s.s., B. garinii, B. afzelii, B. valaisiana, B. sinica and B. yangtze) (Yu et al., 2016). B. garinii is the main genospecies and is distributed mainly in northern China, while B. burgdorferi s.s. is widely distributed in south China (Chen et al., 2013; Wan et al., 1998). B. burgdorferi s.l. has been detected in more than 20 mammalian species and seven genera of birds (Li, 2009). Studies have shown that in addition to humans, at least six other taxa of mammals (sheep, cattle, horses, dogs, cats and mice) and two types of birds (seabirds and migratory birds) can be infected by Borrelia spirochetes in China (Keesing et al., 2009). Previous research used antibodies to detect the Borrelia spp. antigen within camel sera in Egypt and reported a positive rate of 47.8% (Helmy, 2000).

The genus Camelus contains three species: Camelus dromedaries (one-hump dromedary), Camelus bactrianus (two-hump Bactrian camel), and Camelus bactrianus ferus (two-hump Bactrian camel). Dromedaries are mainly found in the Arabian Peninsula, the Middle East, and parts of Africa, whereas Bactrian camels are mainly located in Central and Northeast Asia, Northern China, and Mongolia (Liu et al., 2015). Camelus bactrianus ferus is a new species of camel, and it is mainly distributed in China and Mongolia (Guo, 2009). Currently, there are an estimated 250,000 camels, which are mainly distributed in Xinjiang and the Inner Mongolian Autonomous Region in China (Feng, 2016). The Bactrian camel is an important economic livestock animal in northwest China.

To the best of our knowledge, little molecular information is available on the presence of B. burgdorferi s.l. complex in camels in China. The investigation of known Borrelia and new Borrelia spp. associations is important for animal health and the livestock industry. The aim of this study is to understand the role of camels as carriers and potential spreaders of Lyme disease in China. The exposure of camels to B. burgdorferi s.l. complex was investigated using PCR assay and sequencing. The sequences of the 5S-23S rRNA, OspA, flaB and 16S rRNA genes obtained from positive DNA samples were analysed, and the ticks that potentially act as vectors were discussed.

Materials and methods

Sample collection and DNA extraction

The Bactrian camel is free-choice grazing animal that inhabits desert regions. A total of 138 blood specimens from Bactrian camels were collected during May 2014 and November 2015 in two LB-endemic localities at three sites in northwest China (Gansu (Zhangye) and Xinjiang (Ili and Aksu)) (Fig. 1 ). Each blood sample was collected from the jugular vein of the camel into a sterile tube containing an anticoagulant (ethylene diamine tetraacetic acid, EDTA). Genomic DNA was extracted from individual specimens using a commercial QIAamp DNA Blood Kit (QIAGEN, Maryland, USA) and a QIAamp DNA Mini Kit (QIAGEN, Hilden, Germany) according to the manufacturer's instructions. The extracted genomic DNA samples were then stored at −20 °C until use.

Fig. 1

Bactrian camel blood specimens were harvested in Northwest China (Gansu (Zhangye) and Xinjiang (Ili and Aksu)).

PCR amplification and sequence analysis

A nested PCR for the detection of B. burgdorferi s.l. was carried out using four independent sets of species-specific primers to amplify the 5S-23S rRNA, OspA, flaB and 16S rRNA genes, as described in previous studies (Zhang et al., 2014; Postic et al., 1994; Guy and Stanek, 1991; Wodecka, 2007; Ni et al., 2014; Zhai et al., 2017). The primer sequences are shown in Table 1 . First-round PCR reactions were performed in a thermocycler (BioRad, Hercules, CA, USA) with a total volume of 25 μl containing 2.5 μl of 10× PCR buffer (Mg2+ Plus), 2.0 μl of each dNTP at 2.5 mM, 1.25 U of Taq DNA polymerase (TaKaRa, Dalian, China), 1.0 μl of template DNA, 1.0 μl of each primer (10 pmol), and 17.25 μl of distilled water. The PCR conditions were as follows: 4 min of denaturation at 94 °C; 35 cycles of 94 °C for 30 s, annealing for 30 s (annealing temperatures of primers are listed in Table 1), and 72 °C for 30–48 s (depending on amplicon size); with a final extension step at 72 °C for 10 min. Nested PCR reactions included 1 μl of the first-round PCR product as template for another 30 cycles with the same parameters and annealing temperature profile as described above and in Table 1. To avoid cross-contamination and sample carryover, pre- and post-PCR processing and PCR amplification was performed in separate rooms. B. garinii SZ genomic DNA (from Lanzhou Veterinary Research Institute) was used as a positive control, while distilled water was used as a negative control. PCR products were separated by 1.5% agarose gel electrophoresis, and some positive amplicons from the second round of PCR amplification were directly used for sequencing (GENEWIZ, Inc. Beijing China) or cloned into pGEM-T Easy vector (Promega, Madison, WI, USA). Positive colonies were then selected for PCR amplification and sequencing.

Table 1

Sequences of the primers used in this study.

B. burgdorferi sensu lata target gene	Primer	Nucleotide sequence (5′-3′)	Annealing temperature (°C)	Amplicon size(bp)	Reference
16S	OuterF	GCGAACGGGTGAGTAACG	50	1360	Ni et al. (2014)
	OuterR	CCTCCCTTACGGGTTAGAA
	InnerF	GAGGCGAAGGCGAACTTCTG	60.2	622	Zhai et al. (2017)
	InnerR	CTAGCGATTCCAACTTCATGAAG
5S–23S	5S rRNA (rrf)	CGACCTTCTTCGCCTTAAAGC	57.6	412	Zhang et al. (2014)
	23S rRNA (rrl)	TAAGCTGACTAATACTAATTACCC
	5S rRNA (rrf)IN	CTGCGAGTTCGCGGGAGA	57.3	253	Postic et al. (1994)
	23S rRNA (rrl)IN	TCCTAGGCATTCACCATA
OspA	N1	GAGCTTAAAGGAACTTCTGATAA	52.5	561	Guy and Stanek (1991)
	CI	GTATTGTTGTACTGTAATTGT
	N2	ATGGATCTGGAGTACTTGAA	53	352
	C2	CTTAAAGTAACAGTTCCTTCT
FlaB	OuterF	TGGTATGGGAGTTTCTGG	53.3	774	Wodecka (2007)
	OuterR	TCTGTCATTGTAGCATCTTT
	InnerF	CAGACAACAGAGGGAAAT	54.7	604
	InnerR	TCAAGTCTATTTTGGAAAGCACC

Bioinformatics analysis

All the sequences obtained in this study were subjected to BLAST search via the NCBI website (https://blast.ncbi.nlm.nih.gov/Blast.cgi) using the BLASTn program. Multiple sequence alignment was executed in Florence Corpet (http://multalin.toulouse.inra.fr/multalin/). Phylogenetic analysis was performed using MEGA 6.06 software (Tamura et al., 2013). Phylogenetic trees of Borrelia spp. strains were constructed using all the sequences generated in this study and related sequences previously deposited in GenBank to show the relationships between different strains.

Statistical analysis

A 95% confidence interval (95% CI) for the overall prevalence value of each Borrelia spp. strain was calculated using IBM SPSS Statistics version 19.0.

Nucleotide accession numbers

All 8 sequences of Borrelia spp. 16S rRNA (including 3 sequences from Ili City and 5 sequences from Aksu City) were deposited in GenBank under the following accession numbers: KY284013-KY284020 (among them, AKU1 = AKU3-12, AKU3 = AKU3-14, AKU4 = AKU3-38, and AKU5 = AKU3-45). The sequences of the Borrelia spp. 5S-23S rRNA gene, including 2 sequences from Zhangye City and 2 sequences from Aksu City, were deposited in GenBank under the following accession numbers: KU865304-KU865307. Two sequences of the Borrelia spp. flaB gene from Aksu City were deposited in GenBank with the following accession numbers: KU865318-KU865319. Two sequences of the Borrelia spp. OspA gene from Zhangye City were deposited in GenBank with the following accession numbers: KU865317 and KY328701.

Results

PCR detection of Borrelia burgdorferi spirochetes from Bactrian camel blood samples

The blood samples collected from a total of 138 Bactrian camels in the field two Chinese provinces were screened for the presence of Borrelia spp. by nested PCRs based on four gene loci. The samples were amplified, and the PCR products had lengths of 253–622 bp for the four genes. The results of the nested PCR amplification for positive sample screenings are summarized in Table 2 . Of these specimens, 4 tested positive for 5S-23S rRNA, 2 tested positive for OspA, 2 tested positive for flaB and 8 tested positive for 16S rRNA, with positive rates of 2.9% (95% CI = 0.1–16.3), 1.4% (95% CI = 0.1–14.9), 1.4% (95% CI = 0.1–10.6) and 5.8% (95% CI = 1.3–18.5), respectively. (The sample ZY53 was detected from both 5S-23S rRNA and OspA, the samples AKU3-12 and AKU3-18 were detected from both 5S-23S rRNA and 16S rRNA, and the samples AKU3-14 and AKU3-45 were detected from both flaB and 16S rRNA). At least one B. burgdorferi s.l. gene was detected in 5 of the 138 blood samples examined (3.6%, 95% CI = 0.2–17.9).

Table 2

Prevalence of Borrelia spp. in camel blood samples from two Provinces in China assessed by the nested-PCRs.

Province	City		16S	5S–23S	flaB	OspA	At least one gene
Gansu	Zhangye	46	0 (0)	2 (4.3, 0.2–19.6)	0 (0)	2 (4.3, 0.2–19.6)	1(2.2, 0.1–13.8)
Xinjiang	Aksu	43	5 (11.6, 3.2–36.8)	2 (4.7, 0.2–19.8)	2 (4.7, 0.2–19.8)	0 (0)	4 (9.3, 2.1–28.9)
	Ili	49	3 (6.1, 1.3–20.1)	0 (0)	0 (0)	0 (0)	0 (0)
Total		138	8 (5.8, 1.1–17.2)	4 (2.9, 0.1–16.3)	2 (1.4, 0.1–10.6)	2 (1.4, 0.1–10.6)	5 (3.6, 0.2–17.9)

Sequence and phylogenetic analysis

Sequence analysis of the positive PCR products of the 4 genes assayed in this study revealed that the sequences were most similar to those of B. garinii based on the 16S rRNA, 5S-23S rRNA and flaB gene sequences (100%, 85–100%, and 96–100% identity, respectively) (GenBank accession numbers: CP007564, DQ102468 and CP007564) and were also similar to B. burgdorferi (100%, 100%, and 98–100% identity, respectively) (GenBank accession numbers: AF210138, KP400556 and CP009657). Interestingly, a novel Borrelia genospecies (GenBank accession number: KY284014) was identified from the Ili region, which had a high identity with B. theileri KAT (99.5%, GenBank accession number: KF569941), which was detected from Rhipicephalus geigyi, and with Borrelia sp., with 99.8% identity (GenBank accession number: AB897890), which was detected from Haemaphysalis ticks collected from wild sika deer (Cervus nippon yesoensis) from Hokkaido, Japan, based on 16S rRNA.

Phylogenetic trees were constructed based on the identified Borrelia spp. 16S rRNA (n = 8), 5S-23S rRNA (n = 4), flaB (n = 2) and OspA (n = 2) gene sequences by the neighbour-joining method using the software MEGA 6.06 (Fig. 2, Fig. 3, Fig. 4, Fig. 5 ). The phylogenetic tree based on 16S rRNA sequences indicated that the 16S rRNA gene sequences (5 from Aksu, 3 from Ili) from our study formed three distinct clades. The five AKU (Aksu) strains cluster within a sub-clade of one of three main clades, forming a sister clade with the B. garinii 16S rRNA sequences from China. Interestingly, the strains from the Ili area belonged to two different clades, with two sequences of B. garinii and B. burgdorferi belonging to the LD Borrelia spp. group and one sub-clade of B. theileri belonging to RF Borrelia spp. group. In general, the results show the presence of high heterogeneity among the 16S rRNA sequences of the 8 different Borrelia species strains (Fig. 2). The 5S-23S rRNA sequences formed two distinct clades: 4 strains (2 from Aksu and 2 from Zhangye) formed two distinct sub-clusters in the Borrelia 5S-23S rRNA phylogenetic tree (Fig. 3). The 5S-23S rRNA sequences from the 2 Aksu strains (AKU3–12 = AKU2 of 16S rRNA, AKU3–38 = AKU4 of 16S rRNA) and one Zhangye Borrelia spp. strain belong to one clade of the same branch, which are sister to B. garinii 16S rRNA. One Zhangye strain of the 5S-23S rRNA sequence was located within the same branch (Fig. 3). A phylogenetic tree was constructed based on all the Borrelia flaB sequences deposited in GenBank and two sequences (AKU3-14 = AKU3 of 16S rRNA and AKU3-45 = AKU5 of 16S rRNA) obtained in this study. The flaB sequences from Borrelia spp. formed two main clades. The sequences of the two Aksu strains clustered together with the B. garinii flaB sequences (Fig. 4). Two distinct clades were formed from the sequences of the Borrelia spp. OspA phylogenetic tree. The sequences of two strains from Zhangye clustered with the B. burgdorferi OspA sequences (Fig. 5).

Fig. 2

Phylogenetic tree of the 16S rRNA gene sequences of Borrelia species obtained in the present study and those deposited in GenBank from different countries; accession numbers are shown after isolate names. The 16S rRNA gene sequences obtained in this study are indicated by bold triangles. The tree was inferred using the neighbour-joining method of MEGA 6.06; bootstrap values are shown at each branch point. Numbers above the branch reflect bootstrap support from 500 replications. All sites of the alignment containing insertions-deletions or missing data were eliminated from the analysis (option “complete deletion”).

Fig. 3

Phylogenetic tree of the 5S-23S rRNA gene sequences of Borrelia species obtained in the present study and those deposited in GenBank from different countries; accession numbers are shown after isolate names. The 5S-23S rRNA gene sequences obtained in this study are indicated by bold triangles. The tree was inferred using the neighbour-joining method of MEGA 6.06; bootstrap values are shown at each branch point. Numbers above the branch reflect bootstrap support from 500 replications. All sites of the alignment containing insertions-deletions or missing data were eliminated from the analysis (option “complete deletion”).

Fig. 4

Phylogenetic tree of the flaB gene sequences of Borrelia species obtained in the present study and those deposited in GenBank from different countries; accession numbers are shown after isolate names. The flaB gene sequences obtained in this study are indicated by bold triangles. The tree was inferred using the neighbour-joining method of MEGA 6.06; bootstrap values are shown at each branch point. Numbers above the branch reflect bootstrap support from 500 replications. All sites of the alignment containing insertions-deletions or missing data were eliminated from the analysis (option “complete deletion”).

Fig. 5

Phylogenetic tree of the OspA gene sequences of Borrelia species obtained in the present study and those deposited in GenBank from different countries; accession numbers are shown after isolate names. The OspA gene sequences obtained in this study are indicated by bold triangles. The tree was inferred using the neighbour-joining method of MEGA 6.06; bootstrap values are shown at each branch point. Numbers above the branch reflect bootstrap support from 500 replications. All sites of the alignment containing insertions-deletions or missing data were eliminated from the analysis (option “complete deletion”).

Discussion

In China, Lyme disease is caused by various Borrelia spirochetes. Many of these agents are highly pathogenic to both humans and animals (Liu et al., 2000). Previous studies reported the prevalence of Borrelia in field-collected blood samples from cattle, sheep, dogs, rabbits and rats from different areas in China. These studies primarily used serological detection methods and showed that the distribution of Borrelia varied considerably in the different areas (Hou et al., 2010; Wan et al., 1998).

The areas of Zhangye City, Gansu Province, and of Ili City and Aksu City of Xinjiang Province all include desert regions and are located along the Old Silk Road, halfway between Eastern Asia and Europe, in areas where international livestock trade and travel were frequent (Takada et al., 2001). Bactrian camels were important transportation for trade and travel in the desert within these regions.

The Bactrian camel can harbour and spread many zoonoses, such as Middle East respiratory syndrome coronavirus (MERS-CoV), Anaplasma, Toxoplasma gondii, Onchocerca, Trypanosoma evansi, and Parabronema skrjabini (Liu et al., 2015; Luo, 2012; Wang et al., 2013; Yang et al., 2004; Yang et al., 2008). Ticks are one of the most significant vectors of Borrelia burgdorferi s.l. and RF (relapsing fever). Domestic animals, rodents, and many other wild animals host ticks, and animals bitten by infected ticks can acquire the pathogen and serve as natural reservoirs.

The detection of B. burgdorferi s.l. using PCR is an alternative method that can be used to improve the control and prevention of Lyme disease. According to the nested PCR results, 138 field-collected blood samples assayed with primers targeting the 16S rRNA, 5S-23S rRNA, flaB, and OspA genes revealed 8 (5.8%), 4 (2.9%), 2 (1.4%) and 2 (1.4%) positive samples, respectively, from three regions of two provinces in China where these camels live. According to our knowledge, this is the first report of Borrelia spp. infection in camels in China, indicating their reservoir role in the maintenance of this organism in the environment. The 16S rRNA gene sequences of Borrelia spp. detected from the Aksu region had the highest infection rate (11.6%), followed by the Ili region (6.1%). The genetic identity of B. burgdorferi spirochetes can be clarified by their differential reactivity with genospecies-specific PCR primers targeting the 5S-23S rRNA intergenic spacer amplicon gene. Genetic heterogeneity should be further classified by analysing longer sequence data among B. burgdorferi strains that have been previously identified as the same genospecies of atypical strains of Borrelia spirochetes (Mathiesen Jr et al., 1997; Postic et al., 1998). Moreover, the 16S rRNA gene was detected at a higher positive rate in blood examined for Borrelia spirochetes than other genes in previous research (Wodecka et al., 2010). Our study showed that the infection rate of these genes decreased in the order 16S rRNA >5S-23S rRNA > flaB > OspA.

Two Borrelia species, B. garinii and B. burgdorferi s.s., were identified, and B. garinii was found to be widely distributed in camels in China. In the present study, B. garinii and B. burgdorferi s.s. were identified in camels from Aksu and Ili in Xinjiang Province. Sequence and phylogenetic analysis revealed that those isolates were closely related to the corresponding genotypes based on 16S rRNA gene with high sequence similarities (99.4%–100%, GenBank accession numbers: CP007564; 100%, GenBank accession numbers: AY342028), although the bootstrap values of the phylogenetic tree were relatively low. This finding suggested the genetic diversity of B. garinii and B. burgdorferi s.s in different hosts and geographic locations. Interestingly, the sequencing of cloned PCR products from the 16S rRNA gene of Borrelia spp. from the Ili region showed the presence of a new Borrelia species belonging to the relapsing fever group. The 16S rRNA gene sequence of Borrelia sp. obtained from camel has a 99.5%, 99.8% and 99.8% similarity to the gene of B. theileri KAT strain, Borrelia sp. 120618D12 and Borrelia sp. 130707_13_HJF, respectively (GenBank accession numbers: KF569941, AB897890 and AB897891). B. theileri belongs to the RF Borrelia spp. group and is the causative agent of bovine borreliosis. It was initially identified in cattle and subsequently in goats, sheep and deer from Africa, South America, Mexico and Australia (L, 1903; Mathiesen Jr et al., 1997). Most of the RF Borrelia spp. are transmitted by soft-bodied ticks, but B. theileri is found in hard-bodied ticks and is transmitted by Rhipicephalus spp., including R. annulatus, R. decoloratus, R. microplus and R. evertsi (Barbour et al., 2005; Smith et al., 1985; Trees, 1978). This study provides the first report of B. theileri in camel blood samples in China.

At present, B. burgdorferi has been isolated from nine Ixodes ticks: I. acutitarsus, I. persulcatus, I. granulatus, H. longicornis, H. bispinosa, H. concinna, H. formosensis, Boophilus microplus and D. silvarum (Niu et al., 2009). A previous study reported that the Borrelia isolates were isolated from D. marginatus collected from camels in Xinjiang, China, and these isolates were genetically identified as B. burgdorferi sensu stricto (Wang et al., 2015). The blood samples from Bactrian camels in this study were donated by Dr. Li, who reported that there are ticks available to be collected from these Bactrian camels that have been identified as belonging to H. asiaticum, H. dromedarii, R. sanguineus group, and D. niveus (Li et al., 2015). Thus, these tick species might act as potential vectors to carry and transfer Borrelia spp. that cause camel borreliosis in China. Further study is required to determine whether these ticks are competent vectors for Borrelia spp.

In conclusion, we successfully identified infection with Borrelia spirochetes from camel blood samples from different geographic locations of Gansu Province and Xinjiang Province in China. B. garinii and B. burgdorferi s.s. were highly prevalent in the sampling areas of the two provinces surveyed. Further studies concerning the prevalence of Borrelia spp. groups for both Lyme disease and relapsing fever should be performed to confirm the presence of these different Borrelia species in camels within China. Our findings suggest that Borrelia infection in camels could potentially present a concern for public health. More detailed and widespread monitoring of tick populations and the screening for Borrelia in a greater variety of hosts are warranted in future studies.