Genetic Variation of Echinococcus canadensis (G7) in Mexico |
| |
| Authors: | Ulises Rodriguez-Prado Diego Emiliano Jimenez-Gonzalez Guillermina Avila Armando E. Gonzalez Williams Arony Martinez-Flores Carmen Mondragon de la Pe?a Rigoberto Hernandez-Castro Mirza Romero-Valdovinos Ana Flisser Fernando Martinez-Hernandez Pablo Maravilla Jose Juan Martinez-Maya |
| |
| Abstract: | We evaluated the genetic variation of Echinococcus G7 strain in larval and adult stages using a fragment of the mitochondrial cox1 gen. Viscera of pigs, bovines, and sheep and fecal samples of dogs were inspected for cystic and canine echinococcosis, respectively; only pigs had hydatid cysts. Bayesian inferences grouped the sequences in an E. canadensis G7 cluster, suggesting that, in Mexico, this strain might be mainly present. Additionally, the population genetic and network analysis showed that E. canadensis in Mexico is very diverse and has probably been introduced several times from different sources. Finally, a scarce genetic differentiation between G6 (camel strain) and G7 (pig strain) populations was identified.Echinococcus granulosus sensu lato (s.l.) includes species that cause cystic echinococcosis (CE), one of the most important and widespread parasitic zoonoses. Recent phylogenetic studies based on both mitochondrial and nuclear DNA genes show that E. granulosus s.l. consists of at least four valid species: E. granulosus sensu stricto (s.s.; genotypes G1–G3), E. equinus (G4), E. ortleppi (G5), and E. canadensis (G6–G10). Genotypes G6/G7 are closely related and referred to as camel and pig strains, respectively.1–3 The pig–dog cycle is mainly present in Mexico and maintains the G7 strain.4,5 Although there are isolated reports of E. oligarthrus in a wild cat,6E. ortleppi (E. granulosus s.l.; G5) in a patient,7 and E. granulosus s.s. (G1) in a rural pig, there is no evidence that these species are maintained in Mexico.8 No data of CE caused by G7 have been documented in Mexican patients, although there is a high number of E. canadensis G7-infected patients in central Europe, pointing to the importance of this strain as a cause of human CE.9,10 There are only two genetic studies performed in samples from Mexico. Cruz-Reyes and others5 documented that G7 parasites of Mexican and Polish pig isolates showed similar patterns by restriction fragment length polymorphism (RFLP) of ribosomal DNA (rDNA) internal transcribed spacer 1 (ITS1) and random amplified polymorphic DNA (RAPD) techniques, and although polymerase chain reaction (PCR) -sequencing analysis of mitochondrial cox1 gen fragment was performed, no polymorphism data were reported. Sharma and others11 identified two variants (A and B) inside of the G6/G7 group consisting of samples from Mexico and Argentina using five nuclear markers (elongation factor 1α, transforming growth factor-β receptor kinase, thioredoxin peroxidase, calreticulin, and ezrin-radixin-moesin-like protein). Because some local slaughter records from northern Mexico indicate the presence of Echinococcus spp. in livestock animals,5 the objective of this study was to investigate if parasites in pigs and dogs correspond to G7 and if so, describe its genetic variation.Infected animals were identified in the municipal slaughterhouse of Calera, Zacatecas (north central Mexico), where farm and backyard livestock animals coming from the whole state and other surrounding states were included. For this purpose, viscera from 387 pigs, 243 bovines, and 32 sheep were inspected for the larval stage of Echinococcus. Nine pigs (six pigs from Zacatecas, two pigs from Aguascalientes, and one pig from Morelos) were found infected, and hydatid cysts were obtained under aseptic conditions. After cyst contents were aspirated and centrifuged, aliquots were examined under microscopy to confirm the presence of protoscolices, and pellets were kept in 70% ethanol at −20°C until DNA extraction. Each cyst from each animal was considered as an isolate.Based on the presence of the parasites previously identified in Calera''s slaughterhouse, a rural community located in the central area of Zacatecas at 22°55′ N, 102°48′ W was selected to look for the adult stage of this parasite. For this search, all dogs (60) present in the community were sampled one time for feces after obtaining verbal consent from the owner; samples were used to identify taeniid eggs by the Faust technique, antigens in stool samples (copro-antigens) by enzyme-linked immunosorbent assay (ELISA; CpAg ELISA), and DNA by Copro-PCR. The CpAg ELISA was performed as described by Allan and others12 and Moro and others.13 For Copro-PCR, only positive samples by CpAg ELISA were analyzed using JB3 and JB4 primers to amplify a cox1 gen fragment.14 Coprological analysis of dogs showed that 11 samples were positive by CpAg ELISA (18.3%); only 2 of these samples had taeniid tapeworms (3.4%), and 3 of 11 samples yielded products of approximately 450 bp. All amplicons obtained of hydatid cysts and fecal samples were purified, sequenced on both strands, submitted to GenBank (accession numbers {"type":"entrez-nucleotide-range","attrs":{"text":"KF734649-KF734660","start_term":"KF734649","end_term":"KF734660","start_term_id":"570963772","end_term_id":"570963794"}}KF734649-KF734660), and compared with several mitochondrial DNA sequences of cox1. Dogs positive for taeniid eggs or antigens were purged and treated with praziquantel at 30 mg/kg and arecoline bromide at 2 mg/kg. The protocol was previously approved by the Ethics and Research Committees of the General Hospital “Dr. Manuel Gea Gonzalez”; government and health authorities of the municipality and community also authorized our study.All sequences were subjected to the Basic Local Alignment Search Tool (BLAST) search in the GenBank database; multiple alignments were performed with the CLUSTAL W and MUSCLE programs,15,16 with manual adjusted in MEGA program v517 to determine the appropriate model of molecular evolution in the Modeltest 3.7 program.18 The phylogenetic reconstruction using Bayesian inference was performed with Mr Bayes 3.2.1 program.19 Unrooted haplotype networks were created using NETWORK 4.611 software and nested according to the rules in median-joining networks.20 An analysis of genetic diversity within and between populations was performed using DnaSPv421 and included nucleotide diversity (π), haplotype polymorphism (θ), genetic differentiation index (FST), and Tajima''s D test. Analysis of molecular variance (AMOVA) was used to examine the population genetic structure between populations by ΦST as the genetic fixation index (analogous to FST) obtained by ARLEQUIN software.22After multiple alignments, all sequences of larval and adult stages showed 98% or higher identity with E. canadensis, whereas the Bayesian phylogenetic tree and the haplotype network inference grouped these sequences in the E. canadensis G7 cluster. Sequences for cox1 of E. canadensis from Africa, Asia, Europe, Latin America, and North America deposited in the GenBank databases (N = 58) as well as our sequences (accession numbers {"type":"entrez-nucleotide-range","attrs":{"text":"KF734649-KF734660","start_term":"KF734649","end_term":"KF734660","start_term_id":"570963772","end_term_id":"570963794"}}KF734649-KF734660) were analyzed. The results for π and θ were 0.0118 and 0.718, and the result of Tajima''s D test was −2.1885 (P < 0.01). Genetic differentiation indexes between different paired sequences of E. canadensis genotypes are shown in | Open in a separate windowEurope (G7) includes G7 sequences from Italy, Poland, and Romania. Latin America (G7) includes G7 sequences from Mexico and Peru. Africa (G6) includes G6 sequences from Algeria, Ethiopia, Mauritania, and Sudan. Asia (G6) includes G6 sequences from Iran and Kazakhstan. Africa–Asia (G6) includes G6 sequences from China, Iran, Mauritania, Mongolia, and Russia. SS = sum of squared; VC = variance of components.For the network analysis, haplotypes of E. canadensis (G6, G7, G8, and G10), according to their hosts and country of origin, were included and exhibited three relevant dispersion centers (clustering more than nine haplotypes in each one of them): one for G10 from North America with elk/wolf, one for G6/G7 from Iran, Mauritania, and Peru with camel and sheep, and one for G6/G7 from Africa, Asia, and Latin America with cattle, camel, dog, elk, goat, and human. Interestingly, some G7 pig haplotypes from Mexico are displayed around the third dispersion center; in contrast, other G7 haplotypes from European and Asian countries are clustered around the second dispersion center ().Open in a separate windowHaplotype network for E. canadensis using cox1 sequences of different countries and hosts. Numbers on branches refer to mutational changes. Sizes of circles are proportional to haplotype frequencies (numbers of haplotypes are shown inside circles). Thus, major circles represent ancestral haplotypes, and small circles represent missing haplotypes. Hosts are shown on a side of the haplotypes, and the three big ellipses with discontinuous lines containing G6/G7, G8, and G10.The sequences obtained from three dogs and nine infected pigs showed that E. canadensis (G7) was the only strain identified, indicating that it is the main genotype present in Mexico, which had been previously reported.4,5 This study also shows that E. canadensis (G6, G7, G8, and G10) is lightly more polymorphic than other species of the genus Echinococcus (π = 0.0118), and the negative value of Tajima''s D test suggests a recent expansion for the populations. Haag and others23 reported π = 0.0005 for E. multilocularis and π = 0.0090 for E. granulosus using mitochondrial (nad) and nuclear (ActII, Hbx2, and AgB) sequences; in addition, Sharma and others11 performed a population genetic analysis of E. granulosus s.s. using cox1 sequences and found that π ranged from 0.0039 to 0.0093 for E. granulosus s.s. isolates from India, and they also found a negative value for Tajima''s D test. Small sample sizes and lengths of the nucleotide sequences might affect the π values, showing a tendency toward underestimation.24 In addition, most studies of genetic variation in Echinococcus have used around a dozen sequences; therefore, π results might not be directly comparable among them. However, even under these considerations, this comparison allows us to highlight the genetic diversity among populations of E. canadensis. Furthermore, we found that, in E. canadensis populations, G6 and G7 have a scarce differentiation (FST and ΦST close to 0.1), whereas it is high for E. canadensis G8 and G10 (FST and ΦST > 0.6). In contrast, in a study focused on the genetic diversity of E. granulosus s.s., hydatid cysts from four European countries (Bulgaria, Hungary, Romania, and Italy) were evaluated by sequences of cox1 and showed FST values up 0.187.25 In this study, when G6 and G7 were divided in geographic areas, a similar genetic differentiation was observed with FST and ΦST < 0.1, except when Latin America (G7) was matched with Europe, Africa, or Asia (FST and ΦST = 0.15–0.2), suggesting that the former population reflects a great genetic differentiation regarding the latter populations. This is strengthened by the network analysis, in which some haplotypes of pigs from Mexico are clustered in different branches from those from pigs of European countries.Based on the network analysis, we might deduce the following inferences. (1) E. canadensis G7 in Mexico is very diverse and has probably been introduced from abroad several times from different sources (i.e., shows that six Mexican isolates have from 4 to 14 mutational changes between the isolate and the main haplotype). (2) Haplotypes grouped in the North American wildlife cluster (G10) are closer within them (with one or two mutational changes), and they are placed far away from Mexican isolates; thus, they might be ruled out as sources of introduction to Mexico. (3) Differentiation between G6 and G7 would not make any sense based on the differentiation of genetic indexes found for both genotypes (FST and ΦST close to 0.1). Additionally, one of the main ancestral dispersion centers in the network analysis clustered identical haplotypes of G6 and G7 from China, Mexico, Peru, Sudan, and Russia. The species status of E. canadensis is still controversial,1–3,5,25 because biologically different strains (G6–G10) have been unified. The camel (G6) and pig (G7) strains (both maintained primarily by dog-mediated domestic lifecycles from tropical to temperate zones) are ecologically and geographically segregated from G8 to G102,26; therefore, some works have suggested that G6 and G7 should be treated as a single species: E. intermedius.5,27 However, in recent taxonomic revisions, this proposal has been considered inappropriate,2,26 and the specific name of E. canadensis seems to be the most suitable for handling the closely related genotypes. Thompson and Lymbety28 have argued that knowledge of the genetic structure of cestodes can be applied to the epidemiology and the control of these parasites, because genetic variation within and between populations determines future evolutionary changes, genetic differentiation, and speciation. According to our results, it is probable that E. canadensis G7 has been accidentally introduced from abroad several times through different sources, except from North America (where G10 is more prevalent). This knowledge may have important implications for control of the zoonosis, mainly in areas that lack adequate veterinary control, which could prompt an important health problem. Although presently there are few cases of human cystic echinococcosis in Mexico, interestingly, a study performed in a rural community where an autochthonous human case of CE was detected in 2006 showed that, although some risk practices (such as feeding dogs with infected viscera) were observed, no data of CE in livestock and canine echinococcosis were found, suggesting that CE in Mexico has an unclear pattern.29 |
