Cystic Echinococcoses in Mongolia: Molecular Identification, Serology and Risk Factors

Background Cystic echinococcosis (CE) is a globally distributed cestode zoonosis that causes hepatic cysts. Although Echinococcus granulosus sensu stricto (s.s.) is the major causative agent of CE worldwide, recent molecular epidemiological studies have revealed that E. canadensis is common in countries where camels are present. One such country is Mongolia. Methodology/Principal Findings Forty-three human hepatic CE cases that were confirmed histopathologically at the National Center of Pathology (NCP) in Ulaanbaatar (UB) were identified by analysis of mitochondrial cox 1 gene as being caused by either E. canadensis (n = 31, 72.1%) or E. granulosus s.s. (n = 12, 27.9%). The majority of the E. canadensis cases were strain G6/7 (29/31, 93.5%). Twenty three haplotypes were identified. Sixteen of 39 CE cases with data on age, sex and province of residence were citizens of UB (41.0%), with 13 of the 16 cases from UB caused by E. canadensis (G6/7) (81.3%). Among these 13 cases, nine were children (69.2%). All pediatric cases (n  =  18) were due to E. canadensis with 17 of the 18 cases (94.4%) due to strain G6/7. Serum samples were available for 31 of the 43 CE cases, with 22 (71.0%) samples positive by ELISA to recombinant Antigen B8/1 (rAgB). Nine of 10 CE cases caused by E. granulosus s.s. (90.0%) and 13 of 20 CE cases by E. canadensis (G6/7) (65.0%) were seropositive. The one CE case caused by E. canadensis (G10) was seronegative. CE cases caused by E. granulosus s.s. showed higher absorbance values (median value 1.131) than those caused by E. canadensis (G6/7) (median value 0.106) (p  =  0.0137). Conclusion/Significance The main species/strains in the study population were E. canadenis and E. granulossus s.s. with E. canadensis the predominant species identified in children. The reason why E. canadensis appears to be so common in children is unknown.

In Mongolia, more than 50% of the population lives in the capital city of Ulaanbaatar (UB), with the remainder largely following the traditional nomadic lifestyle [23][24][25][26][27][28]. For many years, CE has been recognized as a common disease in Mongolia even though there is very little published data. Since the collapse of the Soviet Union in 1991, two meetings have been held in UB on the topic of CE. The first meeting was held at the National Center of Communicable Diseases (NCCD) in May 1995 and Tsoodol, Narantuya and Goosh from Mongolia provided overviews of human cases to date [29]. During this meeting, it was reported that in 1950, 7.8% of all surgical patients were diagnosed CE, whereas this value was only 1.9% in 1990. CE was determined to be the cause of 18% of the surgical cases seen at State Central First Hospital (SCFH) in 1993 [25][26][27][28]. The second meeting was held at the Health Science University of Mongolia (HSUM) in June 2009 for the purpose of establishing a network of CE experts [27]. During this meeting, Ayushkhuu from the National Center for Maternal and Child Health (NCMCH) summarized 25 pediatric CE cases from 2008 and 2009 (19 cases in 2008 and 6 cases from Jan to May 2009). These cases consisted of 15 boys and 10 girls, including 4 children under the age of 5-years [27]. During this same meeting, Bazarragchaa from the SCFH summarized a total of 144 (63 males and 81 females) CE cases (1989 [27]. None of the reported cases differentiate between infection with E. granulosus s.s. and E. canadensis. There have been several sero-epidemiology studies of CE in Mongolia [30][31][32][33]. These studies all have used hydatid cyst fluid (HCF) for screening with an ELISA. However, none of the studies used a confirmatory test to verify their findings. Most surgical cases in Mongolia are diagnosed by histopathological identification of resected lesions at the National Center of Pathology (NCP) in UB [27,28]. The present study is the first to identify the primary species/strains of CE present in Mongolian patients that had been histopathologically confirmed as having CE as well as evaluate antibody responses to the recombinant Antigen B8/1 (rAgB) by ELISA [34][35][36][37][38][39][40][41][42][43].

Ethical statement
Molecular analysis of human specimens and serological analysis of antibody responses, and junior researchers from several institutes in UB, Mongolia, to do laboratory analysis of these all specimens were approved by the Asahikawa Medical University (AMU) Institutional Review Board (AMU-IRB-1435).

Patient samples
A total of 43 CE cases were evaluated, consisting of 18 pediatric cases from the NCMCH and 27 cases from the SCFH. In total, 43 hepatic CE cysts (1 cyst per patient) were obtained and confirmed to be CE histopathologically at the NCP. Pre-surgical serum samples were available for 31 of the 43 study patients (Table 1). De-identified data on patient age, sex, and province of residence were also obtained.

DNA analysis
Genomic DNA was extracted from 43 ethanol-fixed samples from hepatic CE patients (Table 1) using a DNeasy tissue kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. The extracted DNA was kept at 220uC until further analysis could be performed. DNA obtained was used as templates for polymerase chain reaction (PCR). The complete or partial mitochondrial cytochrome c oxidase subunit I (cox 1) gene was amplified by PCR as reported previously [44][45][46]. PCR products were treated with illustra ExoStar (GE Healthcare) to remove excess primers and dNTPs, and directly sequenced with a BigDye Terminator v3.1 and a 3500 DNA sequencer (Life Technologies). Obtained sequences were edited using Geneious Pro version 7.0.4 (created by Biomatters, available from http://www.geneious.com), and multiple alignments of each cox1 haplotype were made by the program MAFFT [47] with the homologous sequences of other Echinococcus species available in the GenBank database. A phylogenetic tree was constructed using the neighbour-joining method and Kimura's two-parameter model [48] in Phylogenetic Analysis Using Parsimony (PAUP) version 4.0b [49]. The robustness of the phylogenetic tree was tested by bootstrapping with 1000 replicates. For tree construction, Versteria mustelae was used as an outgroup because it is a sister to all members of the genus Echinococcus [50].

Serology
Recombinant Antigen B8/1 (rAgB) produced from E. multilocularis (rEmAgB) was applied for ELISA and immunoblot tests as reported previously [36]. Several specimens with optical density (OD) values around the cut-off were re-checked by immunoblot using the same antigen for confirmation (figure not shown) [36,40,41]. ELISA was carried out in flat-bottom 96-well microplates (Nunc, Maxisorp, Roskilde, Denmark) as previously described [36]. The microplates were coated with 100 ng/ml of rEmAgB diluted in phosphate-buffered saline (PBS) and incubated at 4 uC overnight. Excess antigen was removed by washing with PBS. Blocking was performed with blocking solution [1% casein in 20 mM Tris-HCl (pH 7.6) containing 150 mM NaCl] and the plates were incubated at 37 uC for 1 hr. The plates were washed twice with PBS containing 0.05% Tween 20 (PBST). 100 ml of diluted sera (1/100 dilution in blocking solution) was added and plates were incubated at 37 uC for 1.5 hrs. After washing five times with PBST, 100 ml of protein G-peroxidase conjugate (1/4000 dilution in blocking solution) (Invitrogen, Camarillo, CA) was added to each well and the plates were incubated at 37 uC for 1.5 hrs. Plates were washed six times with PBST and one time with PBS and incubated with 100 ml of substrate solution [0.4 mM 2,2-azinodi (3-ethyl-benzthiazoline-6-sulfonate) (ABTS) in 0.2 M citric acid buffer (pH 4.7)] at room temperature for 30 min. The color reaction was stopped by addition of 1% SDS. The optical

Author Summary
Cystic echinococcosis (CE) is a parasitic zoonosis with a cosmopolitan distribution. Molecular analysis was carried out on 43 hepatic CE cysts from 43 cases confirmed histopathologically at the NCP, Mongolia. Molecular analysis revealed two species, Echinococcus canadensis and Echinococcus granulosus s.s. Twenty three haplotypes of the cox1 gene were identified. All pediatric cases (n = 18) were by E. canadensis. Sixteen of 39 CE cases with data on age, sex and province of residence were from UB (41.0%), and 13 of these 16 cases were caused by E. canadensis (81.3%). Among the 13 cases from UB, nine were children (69.2%). A total of 31 serum samples from these 43 cases were analyzed for antibody response to rAgB with 22 (71.0%) samples positive by ELISA to rAgB. Thirteen of 20 E. canadensis (G6/7) (65%) and nine of 10 E. granulosus s.s. (90%) were seropositive. CE cases by E. granulosus s.s. showed a higher absorbance value than cases by E. canadensis (p = 0.0137). This is the first study to evaluate age distribution of and antibody responses to rAgB in CE cases caused by the two species in Mongolia. It remains unknown why E. canadensis appears to be more common in pediatric cases.

Statistical analyses
ELISA OD value results and positive ratios in CE cases caused by E. granulosus s.s. (G1) and E. canadensis (G6/7) were assessed by the Wilcoxon rank sum test and Fisher's exact test, respectively. All analyses were two-tailed and a p-value , 0.05 was considered statistically significant.

Discussion
This is the first report on antibody responses to rAgB in patients with CE caused by E. canadensis. As shown in Figure 3, nine of 10 E. granulosus s.s. (G1) (90%) and 13 of 20 E. canadensis (G6/7) (65%) were antibody positive to rEmAgB. While there was no statistically significant difference in the antibody positive ratio between the two species, CE cases caused by E. granulosus s.s. (G1) had higher absorbance values (median ratio = 1.131) than CE cases caused by E. canadensis (G6/7) (median ratio = 0.106) (p = 0.0137). Therefore, the question remains if there is a difference in antigenicity of rAgBs produced by different species of the genus Echinococcus or in the antibody response to rAgB in CE patients infected with different Echinococcus species. The potential influences of other factors such as pathological stages of CE or differences in health conditions not examined in the present study remain unknown [1][2][3].
The Antigen B8/1 gene has high homology in amino acid sequences (92.6%) for E. granulosus s.s. (G1) (EgAgB) and E. multilocularis (EmAgB) and has been shown to have the same sensitivity and specificity for detecting CE cases caused by E. granulosus s.s. (G1) [34][35][36][37][38][39][40][41][42][43]51,52]. For example, a study in Italy [53] showed that there was no difference in sensitivity or specificity using rEgAgB versus rEmAgB to detect CE cases [36]. The merit to use rEmAgB over rEgAgB is much higher yield performance (Sako et al. unpublished). More critical comparative biochemical and molecular studies [54][55][56][57][58][59][60][61] have shown that the amino acid sequence of the epitope region of AgB is well preserved between E. granulosus s.s. (G1) and E. canadensis (G6/7), with minor difference between E. granulosus s.s. (G1) and E. multilocularis. Minor differences in amino acid sequences at the N-terminal do not result in any conformational change in the rAgB epitope itself [36], leading to the belief that rAgB prepared from any Echinococcus species is useful for detection of antibody response in CE. Serological studies from Australia, France, China, Jordan, Turkey, the United States, Nepal, Poland, Italy, and Iran using rEmAgB have all successfully identified cases of CE [34,36,41,53]. In retrospect, CE cases provided from different countries should have  Table 1). doi:10.1371/journal.pntd.0002937.g002 been evaluated for both E. granulosus s.s. and E. canadensis. Additional comparative studies using both rEgAgB and rEcAgB (E. canadensis), rEc6/7AgB or rEc10AgB, to detect CE caused by E. canadensis may help confirmation that no difference in sensitivity and specificity exists between rAgBs from different Echinococcus species [56].
If we assume that the differences in absorbance values between E. granlosus s.s. and E. canadensis were not due to a difference in the epitope of rAgBs, one possible cause is a difference in the expression rate of AgB in CE metacestodes [58], which then influences the production of antibodies. As summarized previously [1][2][3][34][35][36][37][38][39][40][41][42][43][51][52][53], antibody response to rAgB differs by the stage of the larval parasite. The response is poor for the cystic lesion (CL) stage as well as inactive cysts (CE4-5), but higher in active and transitional cysts (CE1-3) [51,53]. Sufficient data were not available from the current study to identify cyst stage for the included patients. However, the present study suggests that abdominal ultrasound alone may not be sufficient to differentiate the Echinococcus species responsible for CE [1][2][3] and additional blinded studies are needed to determine if differences can be detected on ultrasound examination [62].
The two species of E. granulosus s.l. found in the present study have already been reported from Mongolia by Jabbar et al. [17], however, with opposite results on the predominant species. Jabbar et al. [17] found that 34 of 50 CE cases (68%) were due to infection with E. granulosus s.s. (G1-G3) and identified one haplotype of E. granulosus s.s. and three haplotypes of E. canadensis (G6-G10). The CE cases from the Jabbar et al. [17] were primarily from the eastern part of Mongolia, with no cases from UB. In contrast, CE cases evaluated in the present study were from the central and western parts of Mongolia ( Figure 2). Among 39 CE cases with age, sex and province of residence, in the current study, 16 cases (41.0%) were from UB. As more than 50% of the population of Mongolia lives in UB, this was not unexpected. Recent studies have revealed that wolves act as a definitive host of E. canadensis (G6/7 and G10) in Mongolia [63]. Wolves are common in the mountainous areas around UB [63]. These mountainous areas are popular holiday destinations for the population of UB. It is, therefore, possible that these individuals are at higher risk for contracting the condition.
The population of Mongolia has close interactions with dogs due, in part, to their nomadic lifestyle. Each household typically has at least one dog, which is usually not tied and is often left to hunt for its own food during the summer months. In these situations, dogs will hunt small mammals and scavenge offal from domestic livestock. It is also common for herdsmen to train dogs to hunt marmots (Marmota sibirica), which could potentially be an intermediate host for Echinococcus spp. in this region [63]. There is no data from domestic dogs in Mongolia. Therefore, studies on both stray and domestic dogs are needed to better understanding risk factors for human infection. Antibody responses to rEmAgB have been confirmed in cattle, goats, and sheep in Mongolia [64]. Additional studies on antibody responses in camels are needed as is molecular identification of CE cysts in domesticated animals (for example, sheep, goats, cattle, camels) and wild herbivores and omnivores with the potential to act as intermediate hosts.
It is thoroughly unknown why children were over-represented as cases of E. canadensis in the current study. If it is assumed that children have access to the same contamination source as adults, one hypothesis is that CE due to E. canadensis may become symptomatic much faster than CE due to E. granulosus s.s. Further studies, with incorporate sociology and animal ethology components are needed to better understand the transmission ecology of CE caused by E. granulosus s.s. and E. canadensis in Mongolia.

Conclusion
Molecular identification of the causative species is important for epidemiological studies on CE. Use of rAgB serology for preoperative and postoperative diagnoses of CE may be a good complement to diagnostic imaging. Further studies are needed to explore the age variation in CE cases caused by E. granulosus s.s. versus E. canadensis in Mongolia and all other countries where both species are co-distributed. Additional studies on human CE cases, intermediate hosts, and definitive hosts are needed to better evaluate the epidemiology of the circulating Echinococcus species as well as risk factors for infection. Accession numbers of cox1 gene sequcences of Echinococcus granulosus s.s. and Echinococcus canadensis. The nucleotide sequences of all haplotypes reported in the present study are available in DDBJ/EMBL/GenBank database under the accession numbers AB893242-AB893264.