|Home | About | Journals | Submit | Contact Us | Français|
The food-borne trematodes, Opisthorchis viverrini, O. felineus and Clonorchis sinensis, have long been recognized as the cause of major human health problems, with an estimated 40 million infected persons. Of the three species of liver fluke, only O. viverrini is classified as a type 1 carcinogen because of its role as an initiator of chronic inflammation and the subsequent development of cholangiocarcinoma. At present, there are no techniques for the early diagnosis of cholangiocarcinoma and it is fatal for most patients. There is considerable variation in parasite prevalence and disease presentation in different geographical areas, the latter of which may be associated with genetic differences among parasites. In the present study, multilocus enzyme electrophoresis was used to provide a comprehensive genetic characterization of O. viverrini from different geographical localities in Thailand and the Peoples’ Democratic Republic of Laos. Parasites from different localities were compared genetically at 32 enzyme loci. The results of the genetic analyses are sufficient to reject the null hypothesis that O. viverrini represents a single species. Therefore, O. viverrini consists of at least two genetically distinct, yet morphologically similar (i.e. cryptic) species. Moreover, there was also separation of the different populations of snails (i.e. the first intermediate hosts) into two distinct genetic groups that corresponded with the delineation of O. viverrini into two species. This suggests that there may be a history of co-evolution in this host–parasite lineage. Additionally, five distinct genetic groups of parasites were detected, each of which occurred within a different and independent river wetland system. Our findings have major implications for the implementation of effective control and surveillance programs targeted to these medically important food-borne parasites.
The food-borne trematode Opisthorchis viverrini occurs in Thailand, the Peoples’ Democratic Republic of Laos (Laos PDR), Cambodia and Vietnam (WHO, 1995). It has a three-host life cycle. The first intermediate hosts are snails of three subspecies, Bithynia siamensis siamensis, B. siamensis goniomphalos and B. funiculata (see Kaewkes, 2003). At least 18 species of cyprinid fish act as the second intermediate host, whereas carnivores (e.g., cats and dogs), but more specifically humans, are definitive hosts (WHO, 1995). Humans become infected with this parasite following the consumption of raw or partially cooked cyprinid fish containing metacercariae. Infection in humans causes hepatobiliary disease with the subsequent development, in most cases, of cholangiocarcinoma (CHCA) (IARC, 1994; Sithithaworn et al., 1994; Honjo et al., 2005).
This liver fluke is endemic in Southeast Asia, with an estimated nine million people infected, predominately in Thailand (Sithithaworn and Haswell-Elkins, 2003). In the highest endemic region of northeast Thailand, it has been estimated that opisthorchiasis costs US $120 million annually in lost wages and medical care (Muller, 2002). Its prevalence in snails is generally low (<1%), whereas in cyprinid fish prevalence is extremely high (90–95%) (Vichasri et al., 1982; Brockelman et al., 1986). Additionally, there are significant differences in the prevalence and intensity of infection by O. viverrini in humans between geographical areas of Thailand (Jongsuksuntigul et al., 1992; Jongsuksuntigul and Imsomboon, 1998). Endemic areas of opisthorchiasis in Thailand are mainly in the northeast and north but the parasite is absent in the south (Jongsuksuntigul and Imsomboon, 2003). The prevalence of O. viverrini is at its highest in the northeast, but the rate of infection in humans varies considerably among villages, districts and provinces (Sithithaworn and Haswell-Elkins, 2003). For instance, prevalence in the northeast was reported to range from an average of 14% in Khon Kaen Province to 34% in Nakhon Phanom Province (Jongsuksuntigul, 2002), some 300 km apart. However, a more recent report (Sriamporn et al., 2004) indicates that O. viverrini infection ranges from 2% to 71% (mean 25%) in different districts within Khon Kaen Province. The prevalence of O. viverrini infection in the neighboring country of Laos PDR has been estimated to be as high as 60% (Sithithaworn and Haswell-Elkins, 2003). There is also variation in the incidence of CHCA in different communities in Thailand, however, there is no apparent correlation with the prevalence of O. viverrini (see Sriamporn et al., 2004). To date, the role of genetic variation of O. viverrini on this variability in prevalence, transmission and associated disease is not known.
Multilocus enzyme electrophoresis (MEE) has been used very effectively over many years to study the population genetics and systematics of a variety of parasitic helminths (Andrews and Chilton, 1999). However, this technique has not been applied to examine the magnitude of genetic variation in O. viverrini from different geographical regions. The aim of the present study was to use MEE to provide significant new insights on genetic variability of this parasite of socio-economic importance. In addition, genetic variation among parasites from different river wetland systems was compared with the genetic variation of their first intermediate hosts (snails) collected from the same geographic localities.
Metacercariae of O. viverrini were obtained from infected cyprinid fish collected from five different major river wetland systems, including dams, natural reservoirs and secondary branches, in Thailand (11 localities) and Laos PDR (four localities) (Table 1 and Fig. 1). To ensure a high recovery of metacercariae, several hundred fish from each collection locality were pooled and their combined tissues (≈weight of 1–10 kg) digested with 0.3% pepsin solution. Pooled fish samples were used because infected fish contain a small number of metacercariae. It was also important to combine metacercariae from a large number of fish from each locality for the electrophoretic analyses because single fish may contain clones of a single fluke and therefore would not provide a reliable indication of the genetic diversity of O. viverrini at that locality. Metacercariae were identified using a light microscope, 50–100 metacercariae were used to infect hamsters, with at least five hamsters used per collection locality. After 4–6 months, adult worms were collected from the bile duct, identified using morphological characters (Kaewkes, 2003), then washed extensively in physiological saline, placed in pools (20 worms), into microcentrifuge tubes, frozen live and stored at −80 °C for subsequent electrophoretic analyses. Pools of worms were used to provide sufficient sample volume to enable comprehensive electrophoretic analyses at the largest number of loci possible. Although the approach of using pools of individuals, rather than screening large numbers of single individuals, has some disadvantages (i.e. unable to determine the frequency of different alleles in a population), one advantage is that alleles occurring at a low frequency in the population are more likely to be detected in pools of 10 or more individuals rather than in a smaller number of individuals (Chilton et al., 1993).
Snails (B. s. goniomphalos) were also collected for MEE from most localities from which O. viverrini metacercariae were obtained. The one exception was that a different species of snail (B. funiculata) was collected from Chiang Mai, in the Mae Ping River wetland (Table 1 and Fig. 1).
The project protocol was been approved by the Animal Ethics committee, Faculty of Medicine, Khon Kaen University. Procedures of care and use of animals recommended in the guidelines of the National Ethical Committee on animal experimentation, National Research Council of Thailand, are strictly followed.
Frozen homogenates of O. viverrini and uninfected hamster liver (i.e. the host enzyme control on gels) were prepared for electrophoretic examination by adding an equal volume of lysing solution (100 ml distilled H2O, 100 μl β-mercaptoethanol, 10 mg nicotinamide adenine dinucleotide phosphate) to the thawed sample, sonicating and centrifuging at 8000g for 10 min at 4 °C. Supernatants were stored in capillary tubes as about 5 μl aliquots at −20 °C until used. MEE was conducted on a pool of worms from each geographical area and host liver tissue, using cellulose acetate (Cellogel; Milan) as the support medium. Each gel was stained histochemically for a specific enzyme (Richardson et al., 1986). Of the 46 enzymes examined, 28 enzymes encoding a presumptive 32 loci gave sufficient staining intensity and resolution to enable reliable genetic interpretation of pooled samples of at least 20 O. viverrini (Table 2). The 28 enzymes used and their Enzyme Commission (E.C.) numbers were: aconitate dehydrogenase (ACON; E.C. 184.108.40.206), adenylate kinase (AK, E.C. 220.127.116.11), aldolase (ALD, E.C. 18.104.22.168), creatine kinase (CK, E.C. 22.214.171.124), enolase (ENOL, E.C. 126.96.36.199), esterase (EST, E.C. 188.8.131.52), fructose-1,6-diphosphatase (FDP, E.C. 184.108.40.206), fumarate hydratase (FUM, E.C. 220.127.116.11), glyceraldehyde-3-phosphate dehydrogenase (GAPD, E.C. 18.104.22.168), glucose-6-phosphate dehydrogenase (G6PD, E.C. 22.214.171.124), aspartate amino transferase (GOT, E.C. 126.96.36.199), glucose-phosphate isomerase (GPI, E.C. 188.8.131.52), alanine amino transferase (GPT, E.C. 184.108.40.206), hexokinase (HK, E.C. 220.127.116.11), isocitrate dehydrogenase (IDH, E.C. 18.104.22.168), lactate dehydrogenase (LDH, E.C. 22.214.171.124), malate dehydrogenase (MDH, E.C. 126.96.36.199), malic enzyme (ME, E.C. 188.8.131.52), nucleotide diphosphate kinase (NDPK, E.C. 184.108.40.206), peptidase valine–leucine (PEPA, E.C. 220.127.116.11), peptidase leucine–glycine–glycine (PEPB, E.C. 18.104.22.168), peptidase phenylalanine–proline (PEPD, E.C. 3.4.13), phosphoglycerate mutase (PGAM, E.C. 22.214.171.124), phosphoglucomuatase (PGM, E.C. 126.96.36.199), 6-phosphogluconate dehydrogenase (6PGD, E.C. 188.8.131.52), pyruvate kinase (PK, E.C. 184.108.40.206), triose phosphate isomerase (TPI, E.C. 220.127.116.11) and uridine monophosphate kinase (UMPK, E.C. 18.104.22.168). Five individual snails from each locality were also subjected to MEE using the same methodology as for the parasites. Each snail was genetically characterized at 19 enzymes encoding a presumptive 26 loci (Table 3). Two of the enzymes used to compare genetic variation among snails, glycerol-3-phosphate dehydrogenase (αGPD, E.C. 22.214.171.124) and purine nucleoside phosphorylase (NP, E.C. 126.96.36.199), were not used in the comparison of the parasites.
For each locus, the electrophoretic banding pattern of each sample was interpreted allozymically, that is, the allozyme with the least electrophoretic mobility from the cathode was designated as allele a. The multiple banding patterns of individual snails for two enzyme loci (Ak and Gpd) were consistent with the expectations of heterozygous individuals for enzymes with either a monomeric (AK) or dimeric (GPD) quaternary structure. As a consequence of using pools of individual O. viverrini to maximize the number of available genetic markers for analysis, a number of isolates exhibited multiple bands at nine enzyme loci (Acon, Ak, Enol, Gapd, Hk, Mdh, Ndpk, Pgm and Tpi). The banding patterns were often consistent with the expectations of heterozygous individuals within the pools of adult worms for enzymes with either a monomeric (ACON and PGM) or dimeric (ENOL, MDH and TPI) quaternary structure. However, this will be confirmed in future studies by examining the banding patterns of individual worms for these loci. Nonetheless, the presence of multiple bands in isolates does not affect interpretation of the data with respect to the proportion of loci with ‘fixed’ allelic/genetic differences between samples (i.e. where an isolate does not share any alleles in common with another isolate at a particular locus) (Richardson et al., 1986). This genetic distance measure is believed to be appropriate for applications of MEE to studies of the systematics of organisms and has been shown to correlate with other genetic distance measures, such as Nei’s D (Richardson et al., 1986).
Phenograms were constructed using Unweighed Pair–Group Method with Arithmetic averages (UPGMA) (Sneath and Sokal, 1978) analysis of pair-wise comparisons of the proportion of loci that showed fixed allelic differences between O. viverrini isolates or between snails from different localities.
Based on our previous MEE studies to define a sufficient number of enzyme loci to genetically characterise O. viverrini (Saijuntha et al., 2006a,b) 28 enzymes encoding a presumptive 32 loci were used herein for the genetic characterization of O. viverrini. This number of enzyme loci is considerably more than that employed in MEE studies on other species of food-borne trematodes (Agatsuma et al., 1986; 1994; Park et al., 2002) and provides the opportunity to assess the significance of genetic variation detected among geographical localities, and to define a sufficient number of independent genetic markers for biological and epidemiological studies.
Comparison of the allelic profiles of O. viverrini (Table 2) revealed no differences among isolates at 13 loci. At seven of these (Fum-1, Gpt-2, Pep-a, Pep-b, Pep-d, Pgam and Umpk) all O. viverrini isolates had a different single-banded pattern each (i.e. allozyme, hence allele) compared with that of the related food-borne trematode, Fasciola gigantica (Table 2). These thus represent useful genetic markers to distinguish the two species. These loci may represent potential genetic markers to distinguish O. viverrini from other sympatric species of food-borne trematode (e.g., Clonorchis sinensis) that are morphologically similar to O. viverrini and/or produce similar clinical effects (Honjo et al., 2005). Allelic variation among O. viverrini isolates was detected at 19 (60%) enzyme loci (Tables 2 and 4). Some of these polymorphic loci can be used to distinguish O. viverrini from different geographical areas. For example, all O. viverrini from Thailand possess different alleles to O. viverrini originating in Laos PDR for the enzyme loci Ck and Enol. In addition, O. viverrini from Khon Kaen Province (Thailand) have different alleles from O. viverrini in Laos PDR at a further nine enzyme loci (Acon, Ck, Enol, Est, Got-1, Gpi, Gpt-1, Me-2 and Pk), and can be distinguished from O. viverrini from the other provinces in Thailand using a single enzyme locus (Got-1). Fixed genetic differences, where samples do not share any alleles at a specific locus (Richardson et al., 1986; Andrews and Chilton, 1999), were also detected among isolates from Khon Kaen Province at three loci: Ald, G6pd and Pk (Table 2).
The extent of the fixed differences among the four O. viverrini isolates from Khon Kaen Province (3–9%) was equivalent to that among the four isolates from Laos PDR (3–9%), but significantly less than that among isolates from different provinces in Thailand (3–41%; mean = 22%) or between countries (i.e. Thailand vs Laos PDR: 17–44%; mean = 30%) (Table 4).
The phenogram (Fig. 2) of the MEE data (Tables 2 and 4) shows the delineation of O. viverrini into two major groups, each representing at least one species with fixed differences at >30% of loci, with further subdivision into two to six clusters, each of which is distinguished from others by fixed genetic differences of 15%. The first and second cluster are represented by a single isolate from Lampang (LP) and Buri Ram (BR), respectively (22%). The third cluster contained five isolates, four from Khon Kaen Province (KBp, KLp, KBs and KPv) and one isolate from Chaiya Phum Province (CP). The fourth cluster contained two isolates from Mahasarakham Province (MS) and Kalasin Province (KS), which had 17% fixed genetic differences from the third cluster. The fifth contained two isolates from Sakon Nakhon Province (SK) and Nakhon Phanom Province (NP), which differed from the sixth cluster containing four isolates from Laos (VT, NG, TH and VV) at 25% fixed genetic differences.
Fixed genetic differences at 17% of loci were detected between B. s. goniomphalos from the Songkram River + Nam Ngum River wetlands and the Chi River + Mun River wetlands (Tables 3 and 5). This separation of the intermediate hosts into two major genetic groups corresponded with the major division detected among isolates of O. viverrini.
For sexually reproducing diploid organisms, as is the case for O. viverrini, fixed genetic differences among two sympatric populations at a single locus would indicate the lack of gene flow (Richardson et al., 1986; Andrews and Chilton, 1999), and hence establish the existence of two species, while two or more fixed genetic differences provide unequivocal evidence of two sympatric species. However, in comparisons of allopatric populations, which was the case in the present study, fixed genetic differences at >15% of enzyme loci are generally indicative of the existence of distinct species, whereas fixed differences of <15% reflect genetic variation among populations of a single species (Richardson et al., 1986; Andrews and Chilton, 1999). Nonetheless, it is prudent to obtain other evidence (e.g., biological, morphological, ecological) to confirm the specific status of geographically isolated populations. This being the case, we have adopted a highly conservative interpretation of the MEE data by using 30% fixed genetic differences as an indicator of species delineation (i.e. twice that used in other studies to distinguish species complexes of geographically isolated populations (Richardson et al., 1986; Andrews and Chilton, 1999). Therefore, the magnitude of the fixed differences detected between isolates of O. viverrini from some geographical localities is sufficient to reject the null hypothesis that it represents a single species. Specifically, O. viverrini from the localities of Vang Vieng (VV), Nam Ngum (NG), Tha Heur (TH) and Vientiane (VT) in Laos PDR are all genetically distinct (with fixed differences at 33–44% of loci) from, and thus represent a different species to, O. viverrini from the Thai provinces of Khon Kaen (KBs, KLp, KBp and KPv), Chaiya Phum (CP), Mahasarakham (MS) and Kalasin (KS). The conclusion that O. viverrini does not represent a single species is also supported by independent molecular and biological evidence.
Preliminary investigations on the fecundity of O. viverrini, based on the number of eggs/gm/worm following infection of metacercariae in hamsters, have also revealed significant differences between O. viverrini from TH (Laos PDR) and CP (Thailand) (Sithithaworn et al., unpublished). In addition, the results of random amplification of polymorphic DNA (RAPD) analyses on O. viverrini from some of the localities used in the present study, revealed that individuals from Nam Ngum Province (Laos PDR) were genetically distinct from those from the Thai provinces of Chaiya Phum, Mahasarakham, Kalasin and Khon Kaen (which includes the localities of KBs, KLp, KBp and Nampong) (Sithithaworn et al., 2007). This finding is in agreement with the results of the MEE data.
Our study has revealed another finding of major biological and epidemiological significance, namely, the separation of O. viverrini into distinct genetic groups/clusters that are associated with different wetlands. The significance of the genetic divergence within clusters (i.e. within different areas of some wetlands) from a taxonomic perspective is unclear. Whether these genetic differences are indicative of population/geographical variation within a single species, or evidence of additional cryptic species, remains to be determined following comprehensive MEE analysis of multiple isolates from each wetland and surrounding geographical areas and additional biological data. Nonetheless, of major significance is the detection of considerable genetic variation among O. viverrini from different geographical areas. This now provides the basis to determine, we believe for the first time, whether parasite genetic variation is associated with disease presentation that has been recorded in different geographical areas.
Comparisons of the O. viverrini phenogram with that of its first intermediate host, B. s. goniomphalos (Fig. 2) revealed major similarities in the patterns of genetic divergence of both parasites and snails. Bithynia s. goniomphalos from both the Songkram River and Nam Ngum River wetlands were genetically distinct (fixed differences at 17% of loci) from individuals from the Chi River and Mun River wetlands. This separation of the B. s. goniomphalos into two major genetic groups was in total concordance with the delineation of O. viverrini into two species. MEE analyses were conducted on B. s. goniomphalos because this subspecies of snail occurs in Laos PDR and throughout northeast Thailand (Wykoff et al., 1965; Brockelman et al., 1986), which included most of the localities from where O. viverrini were collected in our study. One exception was Kil Lom Dam (LP; Lampang Province) on the Wang River wetland, where B. s. goniomphalos is replaced by B. funiculata as the first intermediate host. Specimens of B. funiculata from a neighboring province, Chiang Mai (CM) (Fig. 1), were used in the analyses (Fig. 2). The third subspecies of snail used by O. viverrini, B. s. siamensis, only occurs in central Thailand (Upatham and Sukhapanth, 1980).
Our findings raise interesting questions concerning the possibility of co-evolution between species of O. viverrini and their first intermediate hosts, and the role of wetland systems as isolating mechanisms preventing gene flow between parasite populations. This three-host parasite system may therefore represent an ideal model to examine fundamental questions relating to the evolution of parasites and their hosts. More detailed analyses, using the genetic markers established in the present study, are required to determine the extent of associations between specific genotypes of O. viverrini and snails, as well as the second intermediate hosts (fish), in different wetlands. This will provide a sound foundation within which to better understand the origin, occurrence, pattern and spread of opisthorchiasis through human populations and to examine, in detail, whether correlations exist concerning the incidence of CHCA, and the presence of specific O. viverrini genotype(s) that initiate the disease.
Given the magnitude of the genetic variation detected among O. viverrini from different wetlands and among B. s. goniomphalos from different geographical areas, and the possibility that different genotypes/species of O. viverrini may be associated with different rates of opisthorchiasis in humans, further work will be undertaken to determine whether O. viverrini can establish successful infections in snails from different geographical areas. Furthermore, the relative importance of the biotic and abiotic features of different wetlands on the ecology and genetic variation of trematodes and their intermediate hosts also needs to be explored. Such studies are of critical importance given the potential for the establishment of mixed species/genotypes as a consequence of an increase in relocation of persons to different regions within Southeast Asia (e.g., Laotian refugees and laborers relocated to Thailand), some of whom will be infected by adults worms of a different parasite genotype/species. This may also occur in years of extensive flooding of different wetlands. Such scenarios have major implications for the surveillance, control and treatment regimes for opisthorchiasis.
In conclusion, analyses of the MEE data demonstrate that distinct genetic populations and/or species of O. viverrini sensu lato are associated with different major river wetland systems, which include Chi River, Mun River, Songkram River and the Wang River in Thailand, and the Nam Ngum River in Laos PDR. The separation of the parasite into at least two sibling species is supported by preliminary evidence from other molecular (Sithithaworn et al., 2007) and biological data (Sithithaworn et al., unpublished) and coincides with the division of B. s. goniomphalos into two major genetic groups. Detection of a new species of food-borne parasite may provide an explanation for the considerable variation in the currently available epidemiological data (e.g. parasite prevalence, intensity, morbidity data, etc.) relating to O. viverrini. Given that the human populations in these wetlands are potentially infected with different species and/or distinct genetic populations of O. viverrini, it is now necessary to determine what these differences are and whether control strategies and treatment (medication) will be equally effective between areas in Southeast Asia, where an estimated 67.3 million people are at risk of infection with O. viverrini (see Keiser and Utzinger, 2005).
This research was supported by a Wellcome Trust Collaborative Research Initiative Grant, the European Commission, the Thailand Research Fund through the Royal Golden Jubilee Ph.D. Program (grant no. PHD/0221/2544 to Weerachai Saijuntha), and funding assistance provided by Pro Vice Chancellor – Research, Ian Davey University of South Australia. We would also like to acknowledge the continuing support of Olympus Australia and to Rojchai Satrawaha for valuable assistance and experience in SE Asia wetland environments.