INTRODUCTION
Opisthorchis viverrini is a food-borne trematode endemic in South-East Asia (e.g., Thailand, Lao PDR, Cambodia and Vietnam) (WHO, 1995). This parasite requires 3 different hosts to complete its life cycle, with snails (Bithynia spp.) and cyprinid fish acting as the first and second intermediate hosts, respectively, while the definitive hosts are carnivores (e.g., dogs, cats and humans) (Kaewkes, Reference Kaewkes2003; Andrews et al. Reference Andrews, Sithithaworn and Petney2008). The definitive host becomes infected after consuming uncooked cyprinid fish containing metacercaria.
Recently, Saijuntha et al. (Reference Saijuntha, Sithithaworn, Wongkham, Laha, Pipitgool, Tesana, Chilton, Petney and Andrews2007) reported that O. viverrini represents a species complex consisting at least 2 sibling species, one that occurs in Thailand and the other in Lao PDR. The results of this study also revealed a correlation between 6 genetically distinct clusters of O. viverrini sensu lato and 5 different wetland systems (Saijuntha et al. Reference Saijuntha, Sithithaworn, Wongkham, Laha, Pipitgool, Tesana, Chilton, Petney and Andrews2007).
O. viverrini is an important carcinogenic agent because of its role in bile duct cancer (cholangiocarcinoma) (Honjo et al. Reference Honjo, Srivatanakul, Sriplung, Kikukawa, Hanai, Uchida, Todoroki and Jedpiyawongse2005; Sripa et al. Reference Sripa, Kaewkes, Sithithaworn, Mairiang, Laha, Smout, Pairojkul, Bhudhisawasdi, Tesana, Thinkamrop, Bethony, Loukas and Brindley2007). Endemic areas of opisthorchiasis in Thailand are mainly in the northeast and north, whereas the parasite is absent from peninsular Thailand (Sithithaworn and Haswell-Elkins, Reference Sithithaworn and Haswell-Elkins2003). The prevalence of O. viverrini infection in the different districts of Khon Kaen Province, situated in North-Eastern Thailand, varies from 2 to 71% (Sriamporn et al. Reference Sriamporn, Pisani, Pipiitgool, Suwanrungruang, Kamsa-ard and Parkin2004). More recently, Sripa and Pairojkul (Reference Sripa and Pairojkul2008) have provided data showing that there has been a decrease in the prevalence of opisthorchiasis in different parts of the country, while the incidence of cholangiocarcinoma appears to have remained unchanged. This decrease in prevalence has also been recorded in Khon Kaen Province, where the very high levels of infection (80%) reported during the 1980s (Upatham et al. Reference Upatham, Viyanant, Kurathong, Brockelman, Menaruchi, Saowakontha, Intarakhao, Vajrasthira and Warren1982) has declined to 25% (Sriamporn et al. Reference Sriamporn, Pisani, Pipiitgool, Suwanrungruang, Kamsa-ard and Parkin2004). Declines in the prevalence of parasitic infection may be a direct effect of ongoing praziquantel treatment programs (Jongsuksuntigul and Imsomboon, Reference Jongsuksuntigul and Imsomboon2003). The number of genotypes that make up the population may be reduced following praziquantel treatment, resulting in fewer genotypes subsequently being available to infect susceptible intermediate hosts. This, in turn, could affect the population structure of O. viverrini.
Micro-evolutionary changes in O. viverrini populations may occur as a consequence of selection against specific genotypes of parasite by different host species, as has been shown for acanthocephalan parasites of fish and parasitic flatworms (Steinauer et al. Reference Steinauer, Nickol and Orti2007). Such selection could occur at one or more stages of the parasite's life cycle and lead to population substructuring, and the potential for co-evolution between parasite and host species or colonization of different host species (Bartoli and Pawlowski, Reference Bartoli and Pawlowski2000; Criscione et al. Reference Criscione, Poulin and Blouin2005; Criscione and Blouin, Reference Criscione and Blouin2006). Saijuntha et al. (Reference Saijuntha, Sithithaworn, Wongkham, Laha, Pipitgool, Tesana, Chilton, Petney and Andrews2007) have shown that there are correlations between genetic clusters of O. viverrini and one of its first intermediate hosts (B. siamensis goniomphalos) in different wetland systems, suggesting possible co-evolution. The potential for population substructuring in O. viverrini may be greater in the second intermediate host than in the first intermediate host, because of a larger number of specific host species acting as second intermediate hosts. At least 18 species of cyprinid fish are recorded as hosts for O. viverrini (WHO, 1995). Differences in the ecology, behaviour, immunological responses to the invading parasite and relative susceptibility to infection of these different species of cyprinid fish may provide different selective pressures on O. viverrini. It has been reported previously that there are seasonal changes in metacercarial burden in the same fish species, and between different fish species in the same reservoir. Seasonal variation in metacercarial burden has also been observed over a short period of time (Sithithaworn et al. Reference Sithithaworn, Pipitgool, Srisawangwong, Elkins and Haswell-Elkins1997). Determining the genetic diversity of metacercariae within second intermediate hosts is essential to an understanding of the processes that ‘drive’ the evolution of trematode populations (Keeney et al. Reference Keeney, Waters and Poulin2007).
Multilocus enzyme electrophoresis (MEE) is a biochemical technique that has been used very effectively to study the population genetic structure and systematics of O. viverrini (see Saijuntha et al. Reference Saijuntha, Sithithaworn, Wongkham, Laha, Pipitgool, Petney, Chilton and Andrews2006a, Reference Saijuntha, Sithithaworn, Wongkham, Laha, Pipitgool, Petney and Andrewsb, Reference Saijuntha, Sithithaworn, Wongkham, Laha, Pipitgool, Tesana, Chilton, Petney and Andrews2007). At least 28 enzymes encoded by a presumptive 32 loci have been established to genetically characterize O. viverrini. Of these, 3 polymorphic enzymes, enolase (Enol), phosphoglucomutase (Pgm) and triose phosphate isomerase (Tpi), have been shown to be useful for examining the population genetic structure of O. viverrini (see Saijuntha et al. Reference Saijuntha, Sithithaworn, Wongkham, Laha, Satrawaha, Chilton, Petney and Andrews2008). In the present study, we investigated whether there were temporal changes in the genetic structure of O. viverrini in cyprinid fish from a single location in Khon Kaen Province in a relatively short time-span (i.e. a 4-year period) based on MEE analyses using the 3 polymorphic enzymes. In addition, we also examined whether there was genetic substructuring of the parasite population in 4 species of cyprinid fish.
MATERIALS AND METHODS
All parasites (metacercariae) used in this study were collected from 4 species of cyprinid fish (Cyclocheilichthys armatus, Hampala dispar, Puntioplites protozsron and Puntius orphoides) in the Kang Lawa Reservoir of the Ban Phai district (Khon Kaen Province, Thailand). Opisthorchis viverrini (n=234) were collected from C. armatus, each year (in July–September) during a 4-year period. The number of individual worms collected each year was 48, 65, 61 and 60 (for 2003–2006, respectively). To avoid temporal effects, O. viverrini from all 4 species of cyprinid fish were also collected at the same time in September 2005. This included 43 worms from C. armatus (which represent a subsample of the 61 worms collected from this host species during 2005), 37 from H. dispar, 30 from P. protozsron and 16 from P. orphoides. Individual parasites from each fish species were pooled for pepsin digestion. The methods used to digest fish and to generate adult O. viverrini, as well as preparing the homogenates of individual worms follow those described by Saijuntha et al. (Reference Saijuntha, Sithithaworn, Wongkham, Laha, Pipitgool, Petney and Andrews2006b).
Three polymorphic enzyme loci (i.e. Enol, Pgm and Tpi) were used to compare the genetic variation among O. vivierrini collected in different years and/or different second intermediate hosts. The histochemical staining methods and methods used for MEE have been described previously by Saijuntha et al. (Reference Saijuntha, Sithithaworn, Wongkham, Laha, Pipitgool, Petney and Andrews2006b). 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 at individual worms of each locus were consistent with the expectations of heterozygous individuals for enzymes with either a monomeric (Pgm) or dimeric (Enol and Tpi) quaternary structure.
The GENEPOP software version 3.3 (Raymond and Rousset, Reference Raymond and Rousset1995) was used to calculate allele and genotype frequencies for each locus. Fixation index within subpopulation (F IS) based on Weir and Cockerham (Reference Weir and Cockerham1984) estimation, genetic differentiation among populations (F ST), and also to test compliance with Hardy-Weinberg equilibrium (HWE) expectations in the sample populations, using Fisher's exact test. Estimation without bias of exact P-values was made using the Markov-chain method, following the algorithm of Guo and Thompson (Reference Guo and Thompson1992). Values of F IS range from −1 to 1; negative values indicate an excess of heterozygotes; positive values indicate heterozygote deficiency compared with values predicted under the HWE. Values of F ST vary between 0 and 1, so that if all the populations have the same allele frequencies, then F ST=0, whereas F ST=1 when all populations are fixed for different alleles.
RESULTS
The allelic profiles of 317 individuals of O. viverrini collected in this study were determined for 3 polymorphic enzyme loci (i.e. Enol, Pgm and Tpi). The allelic and genotypic frequencies of samples collected in different years and from different species of cyprinid fish are shown in Tables 1 and 2, respectively. Two alleles were detected among O. viverrini individuals for Enol. For this locus, allele a occurred at a relatively high frequency (0·077–0·092) each year and in individuals from all species of intermediate host. The frequencies of the 3 genotypes for Enol were relatively consistent irrespective of year or host species sampled. Four alleles were detected within samples for both Pgm and Tpi. For Pgm, allele a was the most common (frequencies of 0·072–0·076), and the frequencies of each of the 4 alleles were very similar between years and host species. The frequency of the Pgm genotypes was also very similar between years and host species. For Tpi, the allele frequencies of the most common allele (b) varied from 0·049 to 0·063 between years and among O. viverrini individuals from different host species. The allele occurring at the lowest frequency (d) was only detected in O. viverrini individuals collected from C. armatus in 2003 and 2004. Differences in frequency of alleles for Tpi between years and host species are also reflected in the variation in genotype frequencies for this locus.
Table 1. Allele frequencies of Opisthorchis viverrini samples collected in different years from Cyclocheilichthys armatus (N=234) and from different hosts in 2005 (N=126)
* Sample size.
** Samples were collected from Cyclocheilichthys armatus only.
*** Ca=Cyclocheilichthys armatus, Hd=Hampala dispar, Pp=Puntioplites protozsron and Po=Puntius orphoides.
Table 2. Genotype frequencies of Opisthorchis viverrini samples collected in different years from Cyclocheilichthys armatus and from different host species
* Sample size.
** Samples were collected from Cyclocheilichthys armatus only.
*** Ca=Cyclocheilichthys armatus, Hd=Hampala dispar, Pp=Puntioplites protozsron and Po=Puntius orphoides.
Significant heterozygote deficiencies compared with the predictions under HWE were detected in 3 of the 4 years (2004–2006) for Pgm (Table 3). Similarly, heterozygote deficiencies for Tpi were detected in 2 years (2003 and 2004). No heterozygote deficiency was detected for Enol in the 4 years. No significant heterozygous deficiencies were detected among O. virerrini collected from different species of second intermediate host in 2005 (Table 4), except for the individuals at Pgm and Tpi that were obtained from P. protozsron. However, there were no significant difference in the pairwise F ST values between O. virerrini collected from C. armatus in different years (Table 5) or from different species of second intermediate host in 2005 (Table 6).
Table 3. The expected (H e) and observed heterozygosity (H o) at three polymorphic loci for Opisthorchis viverrini collected in different years from Cyclocheilichthys armatus in Kang Lawa Reservoir, Khon Kaen Province, Thailand
* Sample size.
** Significant heterozygote deficiency (P<0·005).
Table 4. The expected (H e) and observed heterozygosities (H o) at three polymorphic loci for Opisthorchis viverrini collected in 2005 but from different intermediate host species
* Sample size.
** Significant heterozygote deficiency (P<0·005).
*** Ca=Cyclocheilichthys armatus, Hd=Hampala dispar, Pp=Puntioplites protozsron and Po=Puntius orphoides.
Table 5. Pairwise F ST values (below diagonal) and P values (above diagonal) for Opisthorchis viverrini collected in different years from Cyclocheilichthys armatus in Kang Lawa Reservoir, Khon Kaen Province, Thailand
* Sample size.
Table 6. Pairwise F ST values (below diagonal) and P values (above diagonal) for Opisthorchis viverrini collected from four different host species
* Sample size.
** Ca=Cyclocheilichthys armatus, Hd=Hampala dispar, Pp=Puntioplites protozsron and Po=Puntius orphoides.
DISCUSSION
The genetic diversity of O. viverrini sensu lato within a variety of cyprinid fish species (i.e. second intermediate hosts) may have important implications for microevolutionary and macro-evolutionary change in this medically important parasite. Metacercariae in the second intermediate hosts represent the transition from a phase of asexual reproduction by the parasite in snails (i.e. the production of clones) to sexual reproduction, involving cross- and self-fertilization of adults, in the definitive host. There is, however, limited information available concerning the population genetics of O. viverrini. The results of a recent study by Saijuntha et al. (Reference Saijuntha, Sithithaworn, Wongkham, Laha, Pipitgool, Tesana, Chilton, Petney and Andrews2007) revealed varying levels of genetic variation among populations for each of the 2 species of what is currently recognized as O. viverrini. An accurate assessment of the genetic structure within populations was not possible in the study of Saijuntha et al. (Reference Saijuntha, Sithithaworn, Wongkham, Laha, Pipitgool, Tesana, Chilton, Petney and Andrews2007) because of the relatively small sample sizes used for each population (i.e. sampling locality). In the present study, the population genetic structure of O. viverrini from Kang Lawa Reservoir in the province of Khon Kaen (Thailand) was compared temporally (i.e. from 1 species of second intermediate host but sampled in different years; 2003–2006) and spatially (i.e. parasites originating from 4 species of second intermediate host in 2005) using MEE data for 3 polymorphic loci.
The genetic data for Enol revealed no significant departures from Hardy-Weinberg expectations, irrespective of the year or the species of second intermediate host from which O. viverrini were collected. In contrast, significant departures from Hardy-Weinberg expectations were detected in O. viverrini from C. armatus for Pgm in 3 of the 4 years sampled (2004–2006) and for Tpi in 2 years (2003 and 2004). These represented a deficit of heterozygotes. However, in 2005, 61 O. viverrini were collected from C. armatus at different times of the year. A subset of 43 individuals was collected in September, the remainder in July and August and they showed no significant departure from Hardy-Weinberg expectations for Pgm. This suggests either a seasonal (i.e. within-year) variation in population genetic structure of O. viverrini or an effect caused by the sampling strategy used in our study. It has been demonstrated previously, that there is seasonal variation in the prevalence and intensity of O. viverrini infection in C. armatus (Sithithaworn et al. Reference Sithithaworn, Pipitgool, Srisawangwong, Elkins and Haswell-Elkins1997). This may account for the genetic differences detected among subpopulations of metacercariae in this host species. Departures from Hardy-Weinberg expectations for Pgm were also detected for O. viverrini from H. dispar and P. protozsron, which were collected at the same time (September 2005). Significantly fewer heterozygotes at Tpi were detected for O. viverrini from P. protozsron. A reduction in the number of heterozygotes in natural populations can be caused by factors, such as natural selection, non-random mating, the Wahlund effect (i.e. consequence of pooling data from different populations with different allele frequencies) and the presence of null alleles. Selection against heterozygous individuals is an unlikely explanation of the genetic data for O. viverrini because deviations from the Hardy-Weinberg equilibrium were not detected at all loci or for all sampling periods (i.e. different years and host species). This would also suggest that the presence of null alleles is not an explanation for the deficiency of heterozygotes in some of the samples. Although a Wahlund effect is one possible explanation, particularly given the result for the 2 different subsets of O. viverrini collected from C. armatus in 2005, the frequencies of most alleles for the 3 polymorphic loci did not vary significantly among years or among the four host species sampled in September of 2005. Another possibility for the absence of heterozygotes is non-random mating. O. viverrini is a hermaphroditic trematode with the capacity for cross- and self-fertilization (Trouve et al. Reference Trouve, Renaud, Durand and Jourdane1996). A deficiency in heterozygotes at loci for metacercariae that have been sampled at different times and/or from different species of intermediate host may be a reflection of an increase in the relative frequency of selfing by the previous parental generation. Unfortunately, little is known of the relative ratio of self- to cross-fertilization in natural populations of O. viverrini, a ratio that varies greatly among parasitic helminths (e.g., Trouve et al. Reference Trouve, Renaud, Durand and Jourdane1996; Nollen, Reference Nollen1996a, Reference Nollenb, Reference Nollen1997; Vilas et al. Reference Vilas, Sanmartin and Paniagua2004; Criscione and Blouin, Reference Criscione and Blouin2006). Further work is needed to determine the relative importance of the different modes of reproduction in O. viverrini adults; multilocus DNA analysis might be applicable.
The 4 fish species collected from Kang Lawa Reservoir differ markedly in their relative level of infection with O. viverrini metacercariae. C. armatus has the highest average infection (9·5/host), H. dispar and P. protozsron contain fewer parasites (5·6/host), while P. orphoides contains the least (1·7/host) (Saijuntha et al. unpublished observations). Differences in the intensity of infection and in the biology and ecology of these intermediate hosts could strongly influence the population genetic structure (i.e. allele and genotype frequencies; presence/absence specific genotypes) of O. viverrini. Interestingly, there were no significant differences in the population genetic structure of O. viverrini collected from the 4 species of cyprinid fish in Kang Lawa Reservoir. There were also no significant differences in the population genetic structure of O. viverrini collected from C. armatus in the reservoir over different years. The relatively low F ST values (<0·0025) suggest a high rate of gene flow within this parasite population, and that there is no population substructuring occurring at the metacercarial stage within this reservoir. Given that host dispersal is considered to be a major determinant of parasite gene flow (Blouin et al. Reference Blouin, Yowell, Courtney and Dame1995), it would be important to determine whether there is population substructuring occurring within other parts of this wetland system (i.e. in rivers where there are fewer barriers to host dispersal) or in other wetland systems, particularly as different genetic groups of O. viverrini sensu lato occur in different wetland systems (Saijuntha et al. Reference Saijuntha, Sithithaworn, Wongkham, Laha, Pipitgool, Tesana, Chilton, Petney and Andrews2007). It is also important to examine whether this is indeed a general phenomenon in O. viverrini sensu lato by examining more polymorphic loci and through other analyses such as the application of microsatellite DNA technology which has proven effective in population genetic analyses of other parasite species, including trematodes (Wang et al. Reference Wang, Shrivastava, Johansen, Zhang, Wang and Webster2006; Thiele et al. Reference Thiele, Sorensen, Gazzinelli and Minchella2008).
This research was supported by the Thailand Research Fund, a Wellcome Trust Collaborative Research Initiative Grant, the Office of the National Research Council of Thailand and Mahasarakham University. We would like to thank the Faculty of Medicine, Khon Kaen University for an Overseas Visiting Professor Programme grant.
