Ascariasis Is a Zoonosis in Denmark
http://www.100md.com
微生物临床杂志 2005年第3期
Department of Genetics and Ecology
Institute of Medical Microbiology and Immunology, University of Aarhus, Aarhus
Danish Centre for Experimental Parasitology, The Royal Veterinary and Agricultural University
Danish Bacon and Meat Council, Copenhagen
Department of Clinical Microbiology, Viborg-Kjellerup Hospital, Viborg, Denmark
Laboratory Sciences Division, International Centre for Diarrheal Diseases Research, Dhaka, Bangladesh
ABSTRACT
A preliminary epidemiological survey indicated an association between Ascaris infections in Danish patients and contact with pigs or pig manure. In the present study, we compared Ascaris worms collected from humans and Ascaris worms collected from pigs by amplified fragment length polymorphism (AFLP) analysis, a technique for whole-genome fingerprinting, and by PCR-linked restricted fragment length polymorphism (PCR-RFLP) analysis of the internal transcribed spacer region of nuclear rDNA. The AFLP data were analyzed by distance- and model-based clustering methods. These results assigned Ascaris worms from Danish patients to a cluster different from that for worms from humans in other geographic areas. In contrast, worms from humans and pigs in Denmark were assigned to the same cluster. These results were supported by the PCR-RFLP results. Thus, all of the examined Danish patients had acquired Ascaris infections from domestic pigs; ascariasis may therefore be considered a zoonotic disease in Denmark.
INTRODUCTION
It has been estimated that 1.4 billion people throughout the world are infected with the large roundworm Ascaris lumbricoides (9, 11); this statistic represents a considerable global health burden. Infections with the closely related nematode A. suum are common in pigs. In the Nordic countries, the mean prevalence of A. suum in fatteners is 21.5% (33). The prevalence varies with management, hygiene, age of the pigs, and geographic region (32), but very few Danish swine herds are totally free of infection (31).
In recent years, several cases of child ascariasis with unexplained epidemiology were observed by the Department of Clinical Microbiology in Viborg County, Denmark (6). None of the children had been traveling, but we suspected that the cases were correlated with contact with pigs or pig manure (5).
A. lumbricoides and A. suum may constitute two different but closely related species or may represent host-associated subpopulations or races of the same species (12, 23). Ascaris populations obtained from humans and pigs both in Guatemala and in China were shown to represent sympatric populations (4, 28); i.e., there was no or very restricted gene flow between the Ascaris populations from the two different hosts living in the same areas. Furthermore, no cases of cross-infections were proven between humans and pigs living in close proximity in two Guatemalan villages (3). These observations suggest that the nematodes in the two hosts do constitute distinct taxa. On the other hand, Ascaris infections observed in humans living in areas considered to have a low prevalence of the human parasite indicated that pig Ascaris may cause zoonotic infections (10, 17, 22). Genetic analysis of worms indicated that some human Ascaris infections are zoonotic (2), and it has been shown that the parasites in the two hosts are able to cross-infect under experimental conditions (16, 35).
For the cases of ascariasis in Viborg County, it was not known for sure whether the infections were transmitted from other humans or from pigs. In order to reveal the source of the human Ascaris infections, we decided to compare Ascaris worms obtained from humans and from pigs in Denmark and in some developing countries. Our results confirm and extend Anderson's observation that pigs are the main source of human Ascaris infections in areas considered to have no or a low prevalence of the human form of this parasite (2).
MATERIALS AND METHODS
Ascaris specimens. A total of 135 Ascaris worms from pigs and humans in different countries (Fig. 1) were examined. Staff of the Department of Clinical Microbiology, Viborg, Denmark, asked 150 medical practitioners, with a catchment area population of 230,000, within hours to report whenever they were confronted with an Ascaris-infected patient. Then microbiologists immediately telephoned to ensure that existing worms were mailed and to collect epidemiological data. From 29 Danish patients were obtained 32 Ascaris worms (Table 1). In Bangladesh, 23 worms were collected from humans by one of the authors (R.H.). At the laboratory, the live worms were rinsed in lukewarm tap water, fixed in 70% ethanol, and stored at 5°C. Two worms collected from humans in Bangladesh, nine collected in Guatemala, and five collected in Nepal were provided by T. J. C. Anderson, A. Hall, and S. Williams-Blangero, respectively.
Worms from pigs were obtained from 55 fatteners at different farms throughout Denmark. A single worm was collected directly from the intestine of each pig when slaughtered at one of the five participating abattoirs (Danish Crown: Bjerringbro, Esbjerg, Odense, and Ringsted; Tican: Thisted), and the origin was registered with a farm-specific label carried by the pig. Six worms collected from pigs in Guatemala and three collected in the Philippines were provided by T. J. C. Anderson and S. Eduardo, respectively.
DNA extraction. DNA was extracted from gonads by the cetyltrimethylammonium bromide (CTAB) method (14) with the following modification. The tissue was placed in 600 μl of CTAB buffer, digested with proteinase K (250 μg per ml) overnight at 56°C, and then treated with RNase A (0.2 mg) at 37°C for 15 min. Care was taken to avoid contamination of the samples with intestinal contents and the uterus in females, as this organ may contain stored sperm or fertilized eggs (1). DNA samples were stored at –20°C.
AFLP procedure. The amplified fragment length polymorphism (AFLP) procedure was carried out as described by Vos et al. (37) with some modifications. Briefly, 100 ng of genomic DNA was digested with both EcoRI and MseI, and adaptors (EcoRI, 5'-CTCGTAGACTGCGTACC and CATCTGACGCATGGTTAA-5'; and MseI, 5'-GACGATGAGTCCTGAG and TACTCAGGACTCAT-5') were ligated to the resulting fragments by use of T4 DNA ligase. The adaptors provide priming sites for selective PCR amplification of a subset of the restriction fragments by use of primers with various 3' nucleotide extensions. Digestion and ligation were performed in one step with a total volume of 20 μl. The process was initiated by digestion for 4 h at 37°C; the temperature was subsequently reduced by 0.1°C per s to 16°C, and ligation was continued at this temperature for another 2 h. Enzymes then were denatured by raising the temperature to 70°C for 10 min.
We used 0.5 μl of the treated DNA samples as templates for PCR preamplification and primers that had one selective nucleotide (underlined): EcoRI, GACTGCGTACCAATTCC (E+C), combined with MseI, GATGAGTCCTGAGTAAC (M+C); and EcoRI, GACTGCGTACCAATTCA (E+A), combined with MseI, GATGAGTCCTGAGTAAC (M+C). The following program was used: 20 cycles of 94°C for 30 s, 56°C for 1 min, and 72°C for 1 min.
Of the preamplification PCR products, samples of 0.4 μl were used as templates for the final selective PCR amplifications. The Eco primer was labeled with fluorescein Cy5 at the 5' end. Primers had three selective nucleotides (see below), and the following program was used: 94°C for 30 s, 65°C for 30 s, and 72°C for 1 min. The annealing temperature was subsequently reduced by 0.7°C per cycle for the next 12 cycles. This step was followed by 23 cycles of 94°C for 30 s, 56°C for 30 s, and 72°C for 1 min. PCRs were carried out under standard PCR conditions with a total volume of 20 μl.
All PCR products were examined by gel electrophoresis (1% agarose) to ensure that the amplifications were successful. Three microliters of the final PCR product was analyzed on an ALFexpress sequencer (Amersham Biosciences AB, Uppsala, Sweden). External (50- to 500-bp) and internal (300-bp) standards were included for accurate calculation of the sizes of the fragments. In addition, two previously examined DNA samples were included in each run in order to evaluate the size estimates from gel to gel. Based on the presence or absence of fragments, a binary matrix was generated by using the sequence analyzer software package AFLwin version 1.0 (Amersham Biosciences AB).
The following 14 primer combinations were tested: E+CAG in combination with M+CTG, M+CTC, M+CTT, M+CGA, M+CCG, and M+CAG; and E+ACT in combination with M+CTG, M+CTC, M+CTT, M+CGA, M+CCG, M+CAG, M+CTA, and M+CAC. E and M were the respective EcoRI and MseI adaptors, and the selective nucleotides were the extensions (37). Based on the number of polymorphic bands, reproducibility, and score ability, the following four combinations were chosen for use in the study: E+CAG/M+CAG, E+ACT/M+CTT, E+ACT/M+CTG, and E+ACT/M+CTC.
PCR-RFLP procedure. For the PCR-linked restricted fragment length polymorphism (PCR-RFLP) procedure, the forward and reverse primer sequences used for amplification of the internal transcribed spacer (ITS) region were 5'-TTGAACCGGGTAAAAGTCGT-3' and 5'-TTAGTTTCTTTTCCTCCGCT-3', respectively (2). PCR amplification was carried out with a total volume of 20 μl and 20 ng of DNA from each worm as a template. The following program was used for amplification of the ITS region: 94°C for 1 min; 40 cycles consisting of 94°C for 30 s, 55°C for 40 s, and 72°C for 1 min; and a final extension at 72°C for 7 min. Five-microliter samples of the PCR products were digested with restriction endonuclease HaeIII and analyzed by gel electrophoresis (2% agarose).
Data analysis. Interpopulational relationships of the collected Ascaris worms were analyzed by distance-based cluster analysis. Ascaris worms are diploid organisms; however, due to the dominant nature of the AFLP technique, heterozygotes could not be distinguished. The presence or absence of fragments was therefore treated as being effectively in the form of haplotypes (binary data).
A distance matrix based on the binary variables was calculated by using Excel spreadsheet software (Microsoft, Redmond, Wash.) as follows. Genetic similarity estimates between pairs of worms i and j were obtained by using the classical Jaccard coefficient gsij = a/(n – d) (25). This coefficient rates the number of coincidences (a, bands present in both worms i and j) and the total number of bands (n, number of bands observed in all worms examined) without considering the negative cooccurrence (d, bands absent in both worms i and j). The latter were excluded because the absence of a band may be due to different genetic events and therefore does not necessarily imply identity. The similarities were transformed into genetic distances with the equation gdij = 1 – gsij. This procedure was evaluated and was found to be appropriate for cluster analysis with dominant markers (25).
The resulting matrix was used to construct dendrograms according to the clustering procedures unweighted pair-group method using average linkages (UPGMA), minimum evolution (ME), and neighbor joining (NJ) by using the molecular evolutionary genetic analysis (MEGA) software package (version 3; Center for Evolutionary Functional Genomics, Arizona Biodesign Institute, Arizona State University, Tempe [http://www.megasoftware.net/mega3]) (21).
In a different approach, the results from the AFLP analysis were considered to be phenotypic data obtained from diploid dominant markers. Data were analyzed by using the tools for population genetic analysis (TFPGA) software package (version 1.3; Department of Forest, Range, and Wildlife, Utah State University, Logan [http://bioweb.usu.edu/mpmbio/tfpga.asp]). The presence of a band on the gel indicated the dominant genotype (homozygote or heterozygote), while the absence of this band (blank) represented the homozygote recessive genotype. Since it was not possible to read the allele frequencies directly from the phenotypic data (see above), it was assumed that the genotype frequencies of the subpopulations were in Hardy-Weinberg equilibrium and that the genetic markers were unlinked. In this scenario, the frequency of the recessive allele could be estimated either simply as the square root of the frequency of negative cooccurrence or by the Lynch-Milligan procedure (20) included in the TFPGA program. The distance matrix was calculated by using Nei's unbiased distance (26) in the TFPGA program. The matrix was transferred to the MEGA program, and a dendrogram was generated as described above.
Ascaris worms were placed into groups based on the known origins and by assigning arbitrary cutoff points on the dendrograms. Medians and 25th to 75th percentiles for the overall pairwise genetic distances of worms belonging to the same groups (genetic distances within groups) and of worms belonging to different groups (genetic distances between groups) were calculated by using the Excel program.
The Structure computer program (version 2.1; Department of Human Genetics, University of Chicago, Chicago, Ill. [http://pritch.bsd.uchicago.edu/software.html]) (29) was used to infer population structures by a model-based method for cluster analysis. The AFLP primary data were treated as being haploid, and the model of no admixture was assumed. The program probabilistically assigns the individual worms to subpopulations without prior information about the origins of the specimens. Series of independent runs for models simulating different numbers (K) of subpopulations were performed (K values, 1 to 10; program parameters: burn-in period and collect data iterations 5 x 104). The posterior probability [Pr(K)] for each model was calculated according to the manual for the Structure program.
Analyses of molecular variance (AMOVA) were used to estimate the partitioning of AFLP genotypic variations between and within groups. Distance matrices were constructed for the groups as described above, and calculations were performed by using the Arlequin program (version 2; Genetics and Biometry Laboratory, Department of Anthropology, University of Geneva, Geneva, Switzerland [http://cmpg.unibe.ch/arlequin]).
The distributions of the ITS genotypes among worms collected from the different hosts and sources were compared, and probabilities were calculated by the chi-square test with the SigmaStat for Windows program (version 1.1; Jandel Coporation, San Rafael, Calif.). Hardy-Weinberg equilibrium was tested by using the TFPGA program.
RESULTS
Ascaris epidemiology in Viborg County. Information from 33 episodes of human ascariasis was collected during a 9-month period. Children younger than 5 years old accounted for 52% of the cases. The incidences in this age group were 3.0 per 10,000 children living in the urban area and 27.8 per 10,000 children in the rural population. This difference diminished with age. Contacts with pig manure were observed in 73% of all of the cases. There was no difference in the sex distribution of the patients. More than 80% of the patients expelled only one worm after treatment with an anthelminthic drug, and in two out of three cases a female worm was registered.
AFLP analysis. A total of 135 Ascaris worms were analyzed from five different countries worldwide (Fig. 1). Of these, 71 worms were of human origins and 64 were obtained from pigs. The 32 worms obtained from 29 Danish patients are listed in Table 1; of these, 27 were examined by AFLP.
The four primer combinations used in the AFLP analysis amplified variable numbers of bands (35 to 61). A total of 193 bands were detected, 151 (78%) of which were polymorphic. The proportions of polymorphic bands for the four primer sets varied from 60 to 89% (Table 2). Monomorphic bands were excluded from the cluster analysis. Each worm possessed a unique AFLP band pattern. Several tests of DNA prepared from the same specimens yielded identical patterns, demonstrating the reproducibility of data obtained by AFLP analysis.
Specimens were arranged in random order, and genetic distances were calculated for each pair of worms by using the Jaccard similarity coefficient based on dichotomic variables. Distance matrices were constructed from the results. Dendrograms were generated from the matrices and compared by visual inspection. Results obtained by different methods of cluster analysis (UPGMA, ME, and NJ) showed only minor discrepancies. Thus, the ME method was applied in this study. In order to test the robustness of the clustering depicted in Fig. 2, the AFLP data were also treated as diploid dominant markers and analyzed by using the TFPGA program (data not shown). The two dendrograms generated from the diploid data and from the binary data were practically identical, demonstrating that the calculation procedures yielded highly similar cluster structure formations; therefore, the patterns seemed to be reliable. Three major clusters could be distinguished in all trees examined. The worms obtained from humans in Nepal and Bangladesh were found in one separate cluster (Fig. 2). The worms obtained from humans in Guatemala were located in a second cluster, while the third cluster comprised all worms obtained from pigs. The 27 Danish human Ascaris worms included in the analysis did not form a separate cluster but clustered together with the worms obtained from pigs. This mixed cluster did not show subdivisions related to source or geographic origin, and worms were apparently randomly distributed in the cluster.
The average genetic distances within and between the group of Danish human Ascaris worms, the group of Danish pig worms, and the group of worms obtained from humans in other countries were calculated (Table 3). It was clear that the Danish and non-Danish human Ascaris worms did not belong to a single population, as the within and between distances were significantly different (P < 0.0001). In contrast, the distances within and between groups were alike (P = 0.096) when the worms from Danish human patients and from Danish pigs were compared, indicating that these worms were from the same population.
Individual worms were assigned probabilistically to groups by using the Structure program without prior information about sampling. Different models, e.g., number of groups chosen for interpretation of data, were tested (K values, 1 to 10). Several independent simulations were conducted for each K value. The results verified that the estimates were consistent. From the estimates of posterior probabilities [Pr(K)] it was clear that only the model K = 4 explained the data sufficiently [Pr(K = 4) > 0.99]. The other models tested (K = 1 to 3 and K = 5 to 10) were insufficient to model the data (P in each case, <0.001). The results from a simulation of model K = 4 are shown in Table 4. In this model, 97% of the worms obtained from humans in both Nepal and Bangladesh were assigned to a single group, while 94% of the worms obtained from humans in Guatemala were assigned to a different group. These groups closely reflected the clusters observed in the cluster analysis. All of the worms obtained from pigs and the Danish human Ascaris worms were assigned to one of the other two groups. The worms obtained from pigs in Guatemala tended to be assigned to only one of the same two groups; however, this sample contained only six worms. Of the Danish worms, 97% belonged to these two groups, and the worms obtained from humans and from pigs showed the same distributions in the groups. These two groups were therefore combined in the following analysis.
Given the results reported above, the subsequent analyses focused on the model with the seven populations (Fig. 1) of worms assigned to three groups. The worms obtained from humans in Bangladesh and Nepal were included in one group. The second group contained the worms obtained from humans in Guatemala. The final group comprised the worms obtained from pigs in the Philippines, Guatemala, and Denmark plus the Danish human Ascaris worms. AMOVA revealed that a small but significant part of the genetic variance was distributed among the populations within the three groups (2.5%), while most of the variance was distributed within the populations (76%) or among the groups (22%) (Table 5). Thus, the results confirmed the clustering depicted in Fig. 2.
The two populations of worms obtained from pigs and from humans in Denmark were examined in a separate analysis. AMOVA (Table 6) revealed that nearly all of the genetic variance was distributed within the two populations (99.4%) and that only a small and insignificant part of the variance was distributed between the two groups (0.6%) (P = 0.08). Thus, it was concluded that all of the examined Danish Ascaris worms belonged to a single population.
ITS region polymorphisms. Amplification of the ITS region produced a fragment of approximately 1,000 bp. A two-, three-, or four-band pattern was found after digestion of the products with restriction enzyme HaeIII (Table 7), corresponding to three unique genotypes. The sizes of the fragments were approximately 580 and 420 bp (two fragments), 580, 230, and 190 bp (three fragments), and 580, 420, 230, and 190 bp (four fragments). The predominant patterns among the worms obtained from Danish pigs were the three- and four-band types, whereas the two-band type was dominant among the non-Danish human Ascaris worms. An overall chi-square test showed that worms obtained from Danish pigs and humans and worms obtained from humans in developing countries (Table 7) did not belong to one homogeneous population (2 = 88.0, P < 0.001). Pairwise chi-square tests showed that the distribution of band patterns in Danish human Ascaris worms did not differ from the distribution in worms obtained from Danish pigs (2 = 1.53, P = 0.46). These results support the assumption that all Danish worms, irrespective of host origin, were drawn from the same population.
Although the ITS DNA region exists in multiple copies (see below), it was assumed that the two- and three-band patterns represented homozygotic genotypes (two codominant alleles at one locus, i.e., 1,1 and 2,2) and that the four-band pattern represented the heterozygotic genotype (i.e., 1,2). By use of the TFPGA program, it was found that the genotype frequencies in the total population of worms obtained from humans (including worms from Danish patients) were significantly different from the expected frequencies at Hardy-Weinberg equilibrium (2 = 37.3, P < 0.0001), indicating that population subdivisions existed for these worms. In contrast, the genotype frequencies in the total population of worms obtained from Danish pigs and humans did not differ significantly from the expected frequencies at Hardy-Weinberg equilibrium (2 = 3.56, P = 0.06), indicating that worms obtained from both humans and pigs belonged to a random mating population in Denmark.
DISCUSSION
Ascaris is a common intestinal parasite in pigs raised in Denmark, with a maximum prevalence of 25 to 35% in large fattening pigs and gilts (33). In Denmark, infections caused by Ascaris worms are relatively rare in humans of all ages (2 per 10,000 per year) but are most common in small children living in or visiting rural areas and with a risk of contact with pig manure (36 per 10,000 per year). It has been suggested that up to 1% of Danes are infected and that transmission occurs abroad or through imported food (19, 27). Large roundworms obtained from humans have been designated with the species name A. lumbricoides, although it is known that cross-infections from pigs may occur (10, 17, 22), and no definite morphological criteria for discrimination between the two Ascaris species are known (12). However, based on epidemiological evidence, it has been suggested that Danish children may acquire Ascaris infections by exposure to pig manure (5, 6). This hypothesis was confirmed in the present study by the epidemiological data and by detailed genetic analysis of an international collection of Ascaris worms obtained from humans and from pigs.
The Ascaris worms were compared by AFLP analysis, a technique that detects DNA polymorphisms at multiple specific restriction sites simultaneously. AFLP analysis has been successfully used to reveal a high degree of genetic information from other nematode species (15, 24, 34). The number of polymorphic loci is more important than the allelic diversity at each locus for the statistical power of analysis used for population assignment (7). At present, AFLP analysis seems to have the highest potential for screening a maximum number of loci and is therefore useful for analysis of population diversity; however, due to the dominant nature of the AFLP data, analysis of the results was not straightforward. The results obtained from the AFLP analysis of the Ascaris worms were therefore analyzed by several different clustering methods.
The ME dendrogram shown in Fig. 2 depicts the results. All the of human Ascaris worms collected in the three developing countries, Bangladesh, Guatemala, and Nepal, were found in two clusters containing no worms from pigs. These two clusters may represent evolutionary lines of the human form of the parasite (i.e., A. lumbricoides). Worms obtained from pigs in both of the two developing countries, Guatemala and the Philippines, and from Denmark were located in a single mixed cluster. This cluster may represents the pig parasite A. suum. All worms obtained from Danes were assigned to this cluster of pig Ascaris worms without any subdivisions. The close relationship between worms obtained from Danish humans and pigs was confirmed by structure analysis (Table 4) and by AMOVA (Tables 5 and 6).
This is the first time that AFLP analysis has been used for the examination of Ascaris; thus, it was impossible to compare the results with published observations. Therefore, the worms were also examined by scoring of HaeIII restriction sites in the ITS region of the rRNA gene by PCR-RFLP analysis, a technique previously used by others (2, 28, 38). One or two HaeIII restriction sites are present in the ITS region of Ascaris; hence, two or three fragment patterns will be generated after digestion of the ITS region PCR product by this enzyme. The differences are due to a transversional change (C to G) that results in the loss of an HaeIII restriction site. The rRNA gene of which ITS is a part exists in multiple copies and, in general, only a single repeat type is found within an individual genome (38). Unequal crossing over and gene conversion are believed to cause this homogenization of the rRNA gene repeats (13). However, in Ascaris, the two types can be found together in one individual (2, 28). Samples from these "heterozygous" individuals therefore produce a four-band pattern on the gel. The occurrence of the three different band patterns in worms from the inferred clusters (Table 7) was in agreement with the results of an analysis of Ascaris worms collected in China (28), except that the distribution of the band patterns in the Ascaris worms obtained from Danish patients matched the distribution in the Ascaris worms obtained from pigs. We also found some instances of the two fragment types in worms obtained from Danish pigs. Anderson (2) found that most worms obtained from people living in North America had ITS regions characteristic of worms obtained from pigs; from a comparison with human and pig parasites obtained worldwide, it was concluded that the ITS region similarity to A. suum was not due to regional genetic variations in an A. lumbricoides population. The fragment pattern for the ITS region provided further evidence for the pig origin of Danish human Ascaris worms; however, the presence of the two-band type in worms obtained from Danish pigs demonstrated that, unfortunately, the number of bands cannot be used for distinguishing between A. suum and A. lumbricoides (38).
By using the two independent techniques, we demonstrated that all 27 worms obtained from Danish patients in the present study belonged to the same population as the worms obtained from pigs, and more than 80% of the patients had known contact with pigs or pig manure. The above results suggest that cross-infections from pigs are a common transmission route for human ascariasis in Denmark. An epidemiological study showed that 8% of Dutch primary schoolchildren and 7% of Swedish adults had antibodies to Ascaris (36), but ascariasis was seldom reported. It was shown that humans developed a smaller number of adult worms in the intestine when experimentally infected with eggs from pig Ascaris versus eggs from human Ascaris (35), and often no large worms were established at all (8, 18). These data, together with small inocula, may explain why more than 80% of our patients and the North American patients of Anderson (2) carried only a single worm. In Denmark, there is no registration system for ascariasis; therefore, it is normally a "hidden" disease. Only a few cases with multiple worms may be recorded. In Scotland, 53 cases were reported within a 5-year period (11); of these, 66% were considered not to be travel related and 79% were among adults. Ten out of 12 cases (83%) of ascariasis reported in Finland occurred in patients younger than 10 years old (30), and children under this age accounted for 76% of the cases in our survey. During our 9-month survey, approximately 0.3% of the children younger than 5 years old and living in rural areas in Viborg County were reported to be expelling Ascaris (5).
Our results demonstrate that all of the cases of ascariasis in Danish patients examined in this study can be ascribed to cross-infections with Ascaris from pigs. This conclusion is supported by the fact that none of the patients had a traveling history. Thus, human ascariasis is a zoonosis in Denmark. We cannot exclude the possibility that some cases of A. lumbricoides infections are imported to Denmark; however, we observed the human form of the parasite only among worms collected in developing countries and not among worms collected in Denmark. Zoonotic ascariasis may also be a problem in other industrialized countries having high hygienic standards for human waste (2), and precautions to control this parasitic disease deserve renewed attention.
ACKNOWLEDGMENTS
We thank the medical practitioners and ascariasis patients in Viborg County for the human Ascaris specimens; Erik Vestergrd, Freddy Juhl Nielsen, Henni Lybye, and Niels Kristiansen (Danish Crown) and Anne Grethe Hermann (Tican) for excellent cooperation and for assistance in collecting worms from the Danish pigs; and T. J. C. Anderson, S. Eduardo, A. Hall, and S. Williams-Blangero for providing the non-Danish worms.
This work was funded by the National Research Foundation under the auspices of The Danish Centre for Experimental Parasitology, The Royal Veterinary and Agricultural University, and the Viborg County Research Fund for Health Science.
REFERENCES
Anderson, J. D., S. Williams-Blangero, and T. J. Anderson. 2003. Spurious genotypes in female nematodes resulting from contamination with male DNA. J. Parasitol. 89:1232-1234.
Anderson, T. J. C. 1995. Ascaris infections in humans from North America: molecular evidence for cross-infection. Parasitology 110:215-219.
Anderson, T. J. C., and J. Jaenike. 1997. Host specificity, evolutionary relationships and macrogeographic differentiation among Ascaris populations from humans and pigs. Parasitology 115:325-342.
Anderson, T. J. C., M. E. Romero-Abal, and J. Jaenike. 1993. Genetic structure and epidemiology of Ascaris populations: patterns of host affiliation in Guatemala. Parasitology 107:319-334.
Astrup, I., and J. Prag. 2001. Ascariasis in Denmark. Ugeskr. Laeg. 80:1741-1747. (In Danish.)
Astrup, I., and J. Prag. 2001. Five outbreaks of human ascariasis caused by pig ascaris. Ugeskr. Laeg. 163:452-453. (In Danish.)
Bernatchez, L., and P. Duchesne. 2000. Individual-based genotype analysis in studies of perentage and population assignment: how many loci, how many alleles Can. J. Fish. Aquat. Sci. 57:1-12.
Buckley, J. J. C. 1931. An observation on human resistance to infection with Ascaris from the pig. Helminthology 9:45-46.
Chan, M. S., G. F. Medley, D. Jamison, and D. A. Bundy. 1994. The evaluation of potential global morbidity attributable to intestinal nematode infections. Parasitology 109:373-387.
Crewe, W., and D. H. Smith. 1971. Human infection with pig Ascaris (A. suum). Ann. Trop. Med. Parasitol. 65:85.
Crompton, D. W. T. 2001. Ascaris and ascariasis. Adv. Parasitol. 48:285-375.
Crompton, D. W. T. 1989. Biology of Ascaris lumbricoides, p. 9-44. In D. W. T. Crompton, M. C. Nesheim, and Z. S. Pawlowski (ed.), Ascariasis and its prevention and control. Taylor & Francis, London, England.
Dover, G. 1982. Molecular drive: a cohesive mode of species evolution. Nature 299:111-117.
Doyle, J. J., and J. L. Doyle. 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochem. Bull. 19:11-15.
Folkertsma, R. T., Rouppe van der Voort, J. N., K. E. de Groot, P. M. van Zandvoort, A. Schots, F. J. Gommers, J. Helder, and J. Bakker. 1996. Gene pool similarities of potato cyst nematode populations assessed by AFLP analysis. Mol. Plant-Microbe Interact. 9:47-54.
Galvin, T. J. 1968. Development of human and pig Ascaris in the pig and rabbit. J. Parasitol. 54:1085-1091.
Jaskoski, B. J. 1961. An apparent swine Ascaris infection of man. J. Am. Vet. Med. Assoc. 138:504-505.
Koino, S. 1922. Experimental infections on human body with ascarides. Jpn. Med. World 15:317-320.
Kolmos, H. J., and C. Pedersen. 2000. Parasitologic diagnosis. Ascaris lumbricoides. Ugeskr. Laeg. 162:3748.
Krauss, S. L. 2000. Accurate gene diversity estimates from amplified fragment length polymorphism (AFLP) markers. Mol. Ecol. 9:1241-1245.
Kumar, S., K. Tamura, and M. Nei. 2004. MEGA3: integrated software for molecular evolutionary genetics analyses and sequence alignment. Brief. Bioinformatics 5:150-163.
Lord, W. D., and W. L. Bullock. 1982. Swine Ascaris in humans. N. Engl. J. Med. 306:1113.
Macko, J. K., and P. Dubinsk. 1997. Taxonomic deliberations on human and pigs ascarids. Helminthologia 34:167-171.
Marche, L., S. Valette, E. Grenier, and D. Mugniery. 2001. Intra-species DNA polymorphism in the tobacco cyst-nematode complex (Globodera tabacum) using AFLP. Genome 44:941-946.
Meyer, A. D. S., A. A. F. Garcia, and A. P. D. Souza. 2004. Comparison of similarity coefficients used for cluster analysis with dominant markers in maize (Zea mays L). Genet. Mol. Biol. 27:83-91.
Nei, M. 1978. Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 89:583-590.
Nielsen, J. O. 2004. Worm and protozoan diseases in Nordic countries, p. 678-683. In N. E. Hansen, S. Hauns, and O. B. Schaffalitzky de Muckadell (ed.), Compendium in medicine, 14th ed. Nyt Nordisk Forlag, Copenhagen, Denmark. (In Danish.)
Peng, W., T. J. C. Anderson, X. Zhou, and M. W. Kennedy. 1998. Genetic variation in sympatric Ascaris populations from humans and pigs in China. Parasitology 117:355-361.
Pritchard, J. K., M. Stephens, and P. Donnelly. 2000. Inference of population structure using multilocus genotype data. Genetics 155:945-959.
Raisanen, S., L. Ruuskanen, and S. Nyman. 1985. Epidemic ascariasis—evidence of transmission by imported vegetables. Scand. J. Prim. Health Care 3:189-191.
Roepstorff, A. 2003. Ascaris suum in pigs: population biology and epidemiology. Danish Centre for Experimental Parasitology, Royal Veterinary and Agricultural University, Copenhagen, Denmark.
Roepstorff, A., O. Nilsson, C. J. O'Callaghan, A. Oksanen, B. Gjerde, S. H. Richter, E. O. Ortenberg, D. Christensson, P. Nansen, L. Eriksen, and G. F. Medley. 1999. Intestinal parasites in swine in the Nordic countries: multilevel modelling of Ascaris suum infections in relation to production factors. Parasitology 119:521-534.
Roepstorff, A., O. Nilsson, A. Oksanen, B. Gjerde, S. H. Richter, E. Ortenberg, D. Christensson, K. B. Martinsson, P. C. Bartlett, P. Nansen, L. Eriksen, O. Helle, S. Nikander, and K. Larsen. 1998. Intestinal parasites in swine in the Nordic countries: prevalence and geographical distribution. Vet. Parasitol. 76:305-319.
Semblat, J. P., E. Wajnberg, A. Dalmasso, P. Abad, and P. Castagnone-Sereno. 1998. High-resolution DNA fingerprinting of parthenogenetic root-knot nematodes using AFLP analysis. Mol. Ecol. 7:119-125.
Takata, I. 1951. Experimental infection of man with Ascaris of man and the pig. Kitasato Arch. Exp. Med. 23:49-59.
van Knapen, F., J. Buijs, L. M. Kortbeek, and I. Ljungstrom. 1992. Larva migrans syndrome: toxocara, ascaris, or both Lancet 340:550-551.
Vos, P., R. Hogers, M. Bleeker, M. Reijans, T. van de Lee, M. Hornes, A. Frijters, J. Pot, J. Peleman, M. Kuiper, et al. 1995. AFLP: a new technique for DNA fingerprinting. Nucleic Acids Res. 23:4407-4414.
Zhu, X., N. B. Chilton, D. E. Jacobs, J. Boes, and R. B. Gasser. 1999. Characterisation of Ascaris from human and pig hosts by nuclear ribosomal DNA sequences. Int. J. Parasitol. 29:469-478.(Peter Nejsum, E. Davis Pa)
Institute of Medical Microbiology and Immunology, University of Aarhus, Aarhus
Danish Centre for Experimental Parasitology, The Royal Veterinary and Agricultural University
Danish Bacon and Meat Council, Copenhagen
Department of Clinical Microbiology, Viborg-Kjellerup Hospital, Viborg, Denmark
Laboratory Sciences Division, International Centre for Diarrheal Diseases Research, Dhaka, Bangladesh
ABSTRACT
A preliminary epidemiological survey indicated an association between Ascaris infections in Danish patients and contact with pigs or pig manure. In the present study, we compared Ascaris worms collected from humans and Ascaris worms collected from pigs by amplified fragment length polymorphism (AFLP) analysis, a technique for whole-genome fingerprinting, and by PCR-linked restricted fragment length polymorphism (PCR-RFLP) analysis of the internal transcribed spacer region of nuclear rDNA. The AFLP data were analyzed by distance- and model-based clustering methods. These results assigned Ascaris worms from Danish patients to a cluster different from that for worms from humans in other geographic areas. In contrast, worms from humans and pigs in Denmark were assigned to the same cluster. These results were supported by the PCR-RFLP results. Thus, all of the examined Danish patients had acquired Ascaris infections from domestic pigs; ascariasis may therefore be considered a zoonotic disease in Denmark.
INTRODUCTION
It has been estimated that 1.4 billion people throughout the world are infected with the large roundworm Ascaris lumbricoides (9, 11); this statistic represents a considerable global health burden. Infections with the closely related nematode A. suum are common in pigs. In the Nordic countries, the mean prevalence of A. suum in fatteners is 21.5% (33). The prevalence varies with management, hygiene, age of the pigs, and geographic region (32), but very few Danish swine herds are totally free of infection (31).
In recent years, several cases of child ascariasis with unexplained epidemiology were observed by the Department of Clinical Microbiology in Viborg County, Denmark (6). None of the children had been traveling, but we suspected that the cases were correlated with contact with pigs or pig manure (5).
A. lumbricoides and A. suum may constitute two different but closely related species or may represent host-associated subpopulations or races of the same species (12, 23). Ascaris populations obtained from humans and pigs both in Guatemala and in China were shown to represent sympatric populations (4, 28); i.e., there was no or very restricted gene flow between the Ascaris populations from the two different hosts living in the same areas. Furthermore, no cases of cross-infections were proven between humans and pigs living in close proximity in two Guatemalan villages (3). These observations suggest that the nematodes in the two hosts do constitute distinct taxa. On the other hand, Ascaris infections observed in humans living in areas considered to have a low prevalence of the human parasite indicated that pig Ascaris may cause zoonotic infections (10, 17, 22). Genetic analysis of worms indicated that some human Ascaris infections are zoonotic (2), and it has been shown that the parasites in the two hosts are able to cross-infect under experimental conditions (16, 35).
For the cases of ascariasis in Viborg County, it was not known for sure whether the infections were transmitted from other humans or from pigs. In order to reveal the source of the human Ascaris infections, we decided to compare Ascaris worms obtained from humans and from pigs in Denmark and in some developing countries. Our results confirm and extend Anderson's observation that pigs are the main source of human Ascaris infections in areas considered to have no or a low prevalence of the human form of this parasite (2).
MATERIALS AND METHODS
Ascaris specimens. A total of 135 Ascaris worms from pigs and humans in different countries (Fig. 1) were examined. Staff of the Department of Clinical Microbiology, Viborg, Denmark, asked 150 medical practitioners, with a catchment area population of 230,000, within hours to report whenever they were confronted with an Ascaris-infected patient. Then microbiologists immediately telephoned to ensure that existing worms were mailed and to collect epidemiological data. From 29 Danish patients were obtained 32 Ascaris worms (Table 1). In Bangladesh, 23 worms were collected from humans by one of the authors (R.H.). At the laboratory, the live worms were rinsed in lukewarm tap water, fixed in 70% ethanol, and stored at 5°C. Two worms collected from humans in Bangladesh, nine collected in Guatemala, and five collected in Nepal were provided by T. J. C. Anderson, A. Hall, and S. Williams-Blangero, respectively.
Worms from pigs were obtained from 55 fatteners at different farms throughout Denmark. A single worm was collected directly from the intestine of each pig when slaughtered at one of the five participating abattoirs (Danish Crown: Bjerringbro, Esbjerg, Odense, and Ringsted; Tican: Thisted), and the origin was registered with a farm-specific label carried by the pig. Six worms collected from pigs in Guatemala and three collected in the Philippines were provided by T. J. C. Anderson and S. Eduardo, respectively.
DNA extraction. DNA was extracted from gonads by the cetyltrimethylammonium bromide (CTAB) method (14) with the following modification. The tissue was placed in 600 μl of CTAB buffer, digested with proteinase K (250 μg per ml) overnight at 56°C, and then treated with RNase A (0.2 mg) at 37°C for 15 min. Care was taken to avoid contamination of the samples with intestinal contents and the uterus in females, as this organ may contain stored sperm or fertilized eggs (1). DNA samples were stored at –20°C.
AFLP procedure. The amplified fragment length polymorphism (AFLP) procedure was carried out as described by Vos et al. (37) with some modifications. Briefly, 100 ng of genomic DNA was digested with both EcoRI and MseI, and adaptors (EcoRI, 5'-CTCGTAGACTGCGTACC and CATCTGACGCATGGTTAA-5'; and MseI, 5'-GACGATGAGTCCTGAG and TACTCAGGACTCAT-5') were ligated to the resulting fragments by use of T4 DNA ligase. The adaptors provide priming sites for selective PCR amplification of a subset of the restriction fragments by use of primers with various 3' nucleotide extensions. Digestion and ligation were performed in one step with a total volume of 20 μl. The process was initiated by digestion for 4 h at 37°C; the temperature was subsequently reduced by 0.1°C per s to 16°C, and ligation was continued at this temperature for another 2 h. Enzymes then were denatured by raising the temperature to 70°C for 10 min.
We used 0.5 μl of the treated DNA samples as templates for PCR preamplification and primers that had one selective nucleotide (underlined): EcoRI, GACTGCGTACCAATTCC (E+C), combined with MseI, GATGAGTCCTGAGTAAC (M+C); and EcoRI, GACTGCGTACCAATTCA (E+A), combined with MseI, GATGAGTCCTGAGTAAC (M+C). The following program was used: 20 cycles of 94°C for 30 s, 56°C for 1 min, and 72°C for 1 min.
Of the preamplification PCR products, samples of 0.4 μl were used as templates for the final selective PCR amplifications. The Eco primer was labeled with fluorescein Cy5 at the 5' end. Primers had three selective nucleotides (see below), and the following program was used: 94°C for 30 s, 65°C for 30 s, and 72°C for 1 min. The annealing temperature was subsequently reduced by 0.7°C per cycle for the next 12 cycles. This step was followed by 23 cycles of 94°C for 30 s, 56°C for 30 s, and 72°C for 1 min. PCRs were carried out under standard PCR conditions with a total volume of 20 μl.
All PCR products were examined by gel electrophoresis (1% agarose) to ensure that the amplifications were successful. Three microliters of the final PCR product was analyzed on an ALFexpress sequencer (Amersham Biosciences AB, Uppsala, Sweden). External (50- to 500-bp) and internal (300-bp) standards were included for accurate calculation of the sizes of the fragments. In addition, two previously examined DNA samples were included in each run in order to evaluate the size estimates from gel to gel. Based on the presence or absence of fragments, a binary matrix was generated by using the sequence analyzer software package AFLwin version 1.0 (Amersham Biosciences AB).
The following 14 primer combinations were tested: E+CAG in combination with M+CTG, M+CTC, M+CTT, M+CGA, M+CCG, and M+CAG; and E+ACT in combination with M+CTG, M+CTC, M+CTT, M+CGA, M+CCG, M+CAG, M+CTA, and M+CAC. E and M were the respective EcoRI and MseI adaptors, and the selective nucleotides were the extensions (37). Based on the number of polymorphic bands, reproducibility, and score ability, the following four combinations were chosen for use in the study: E+CAG/M+CAG, E+ACT/M+CTT, E+ACT/M+CTG, and E+ACT/M+CTC.
PCR-RFLP procedure. For the PCR-linked restricted fragment length polymorphism (PCR-RFLP) procedure, the forward and reverse primer sequences used for amplification of the internal transcribed spacer (ITS) region were 5'-TTGAACCGGGTAAAAGTCGT-3' and 5'-TTAGTTTCTTTTCCTCCGCT-3', respectively (2). PCR amplification was carried out with a total volume of 20 μl and 20 ng of DNA from each worm as a template. The following program was used for amplification of the ITS region: 94°C for 1 min; 40 cycles consisting of 94°C for 30 s, 55°C for 40 s, and 72°C for 1 min; and a final extension at 72°C for 7 min. Five-microliter samples of the PCR products were digested with restriction endonuclease HaeIII and analyzed by gel electrophoresis (2% agarose).
Data analysis. Interpopulational relationships of the collected Ascaris worms were analyzed by distance-based cluster analysis. Ascaris worms are diploid organisms; however, due to the dominant nature of the AFLP technique, heterozygotes could not be distinguished. The presence or absence of fragments was therefore treated as being effectively in the form of haplotypes (binary data).
A distance matrix based on the binary variables was calculated by using Excel spreadsheet software (Microsoft, Redmond, Wash.) as follows. Genetic similarity estimates between pairs of worms i and j were obtained by using the classical Jaccard coefficient gsij = a/(n – d) (25). This coefficient rates the number of coincidences (a, bands present in both worms i and j) and the total number of bands (n, number of bands observed in all worms examined) without considering the negative cooccurrence (d, bands absent in both worms i and j). The latter were excluded because the absence of a band may be due to different genetic events and therefore does not necessarily imply identity. The similarities were transformed into genetic distances with the equation gdij = 1 – gsij. This procedure was evaluated and was found to be appropriate for cluster analysis with dominant markers (25).
The resulting matrix was used to construct dendrograms according to the clustering procedures unweighted pair-group method using average linkages (UPGMA), minimum evolution (ME), and neighbor joining (NJ) by using the molecular evolutionary genetic analysis (MEGA) software package (version 3; Center for Evolutionary Functional Genomics, Arizona Biodesign Institute, Arizona State University, Tempe [http://www.megasoftware.net/mega3]) (21).
In a different approach, the results from the AFLP analysis were considered to be phenotypic data obtained from diploid dominant markers. Data were analyzed by using the tools for population genetic analysis (TFPGA) software package (version 1.3; Department of Forest, Range, and Wildlife, Utah State University, Logan [http://bioweb.usu.edu/mpmbio/tfpga.asp]). The presence of a band on the gel indicated the dominant genotype (homozygote or heterozygote), while the absence of this band (blank) represented the homozygote recessive genotype. Since it was not possible to read the allele frequencies directly from the phenotypic data (see above), it was assumed that the genotype frequencies of the subpopulations were in Hardy-Weinberg equilibrium and that the genetic markers were unlinked. In this scenario, the frequency of the recessive allele could be estimated either simply as the square root of the frequency of negative cooccurrence or by the Lynch-Milligan procedure (20) included in the TFPGA program. The distance matrix was calculated by using Nei's unbiased distance (26) in the TFPGA program. The matrix was transferred to the MEGA program, and a dendrogram was generated as described above.
Ascaris worms were placed into groups based on the known origins and by assigning arbitrary cutoff points on the dendrograms. Medians and 25th to 75th percentiles for the overall pairwise genetic distances of worms belonging to the same groups (genetic distances within groups) and of worms belonging to different groups (genetic distances between groups) were calculated by using the Excel program.
The Structure computer program (version 2.1; Department of Human Genetics, University of Chicago, Chicago, Ill. [http://pritch.bsd.uchicago.edu/software.html]) (29) was used to infer population structures by a model-based method for cluster analysis. The AFLP primary data were treated as being haploid, and the model of no admixture was assumed. The program probabilistically assigns the individual worms to subpopulations without prior information about the origins of the specimens. Series of independent runs for models simulating different numbers (K) of subpopulations were performed (K values, 1 to 10; program parameters: burn-in period and collect data iterations 5 x 104). The posterior probability [Pr(K)] for each model was calculated according to the manual for the Structure program.
Analyses of molecular variance (AMOVA) were used to estimate the partitioning of AFLP genotypic variations between and within groups. Distance matrices were constructed for the groups as described above, and calculations were performed by using the Arlequin program (version 2; Genetics and Biometry Laboratory, Department of Anthropology, University of Geneva, Geneva, Switzerland [http://cmpg.unibe.ch/arlequin]).
The distributions of the ITS genotypes among worms collected from the different hosts and sources were compared, and probabilities were calculated by the chi-square test with the SigmaStat for Windows program (version 1.1; Jandel Coporation, San Rafael, Calif.). Hardy-Weinberg equilibrium was tested by using the TFPGA program.
RESULTS
Ascaris epidemiology in Viborg County. Information from 33 episodes of human ascariasis was collected during a 9-month period. Children younger than 5 years old accounted for 52% of the cases. The incidences in this age group were 3.0 per 10,000 children living in the urban area and 27.8 per 10,000 children in the rural population. This difference diminished with age. Contacts with pig manure were observed in 73% of all of the cases. There was no difference in the sex distribution of the patients. More than 80% of the patients expelled only one worm after treatment with an anthelminthic drug, and in two out of three cases a female worm was registered.
AFLP analysis. A total of 135 Ascaris worms were analyzed from five different countries worldwide (Fig. 1). Of these, 71 worms were of human origins and 64 were obtained from pigs. The 32 worms obtained from 29 Danish patients are listed in Table 1; of these, 27 were examined by AFLP.
The four primer combinations used in the AFLP analysis amplified variable numbers of bands (35 to 61). A total of 193 bands were detected, 151 (78%) of which were polymorphic. The proportions of polymorphic bands for the four primer sets varied from 60 to 89% (Table 2). Monomorphic bands were excluded from the cluster analysis. Each worm possessed a unique AFLP band pattern. Several tests of DNA prepared from the same specimens yielded identical patterns, demonstrating the reproducibility of data obtained by AFLP analysis.
Specimens were arranged in random order, and genetic distances were calculated for each pair of worms by using the Jaccard similarity coefficient based on dichotomic variables. Distance matrices were constructed from the results. Dendrograms were generated from the matrices and compared by visual inspection. Results obtained by different methods of cluster analysis (UPGMA, ME, and NJ) showed only minor discrepancies. Thus, the ME method was applied in this study. In order to test the robustness of the clustering depicted in Fig. 2, the AFLP data were also treated as diploid dominant markers and analyzed by using the TFPGA program (data not shown). The two dendrograms generated from the diploid data and from the binary data were practically identical, demonstrating that the calculation procedures yielded highly similar cluster structure formations; therefore, the patterns seemed to be reliable. Three major clusters could be distinguished in all trees examined. The worms obtained from humans in Nepal and Bangladesh were found in one separate cluster (Fig. 2). The worms obtained from humans in Guatemala were located in a second cluster, while the third cluster comprised all worms obtained from pigs. The 27 Danish human Ascaris worms included in the analysis did not form a separate cluster but clustered together with the worms obtained from pigs. This mixed cluster did not show subdivisions related to source or geographic origin, and worms were apparently randomly distributed in the cluster.
The average genetic distances within and between the group of Danish human Ascaris worms, the group of Danish pig worms, and the group of worms obtained from humans in other countries were calculated (Table 3). It was clear that the Danish and non-Danish human Ascaris worms did not belong to a single population, as the within and between distances were significantly different (P < 0.0001). In contrast, the distances within and between groups were alike (P = 0.096) when the worms from Danish human patients and from Danish pigs were compared, indicating that these worms were from the same population.
Individual worms were assigned probabilistically to groups by using the Structure program without prior information about sampling. Different models, e.g., number of groups chosen for interpretation of data, were tested (K values, 1 to 10). Several independent simulations were conducted for each K value. The results verified that the estimates were consistent. From the estimates of posterior probabilities [Pr(K)] it was clear that only the model K = 4 explained the data sufficiently [Pr(K = 4) > 0.99]. The other models tested (K = 1 to 3 and K = 5 to 10) were insufficient to model the data (P in each case, <0.001). The results from a simulation of model K = 4 are shown in Table 4. In this model, 97% of the worms obtained from humans in both Nepal and Bangladesh were assigned to a single group, while 94% of the worms obtained from humans in Guatemala were assigned to a different group. These groups closely reflected the clusters observed in the cluster analysis. All of the worms obtained from pigs and the Danish human Ascaris worms were assigned to one of the other two groups. The worms obtained from pigs in Guatemala tended to be assigned to only one of the same two groups; however, this sample contained only six worms. Of the Danish worms, 97% belonged to these two groups, and the worms obtained from humans and from pigs showed the same distributions in the groups. These two groups were therefore combined in the following analysis.
Given the results reported above, the subsequent analyses focused on the model with the seven populations (Fig. 1) of worms assigned to three groups. The worms obtained from humans in Bangladesh and Nepal were included in one group. The second group contained the worms obtained from humans in Guatemala. The final group comprised the worms obtained from pigs in the Philippines, Guatemala, and Denmark plus the Danish human Ascaris worms. AMOVA revealed that a small but significant part of the genetic variance was distributed among the populations within the three groups (2.5%), while most of the variance was distributed within the populations (76%) or among the groups (22%) (Table 5). Thus, the results confirmed the clustering depicted in Fig. 2.
The two populations of worms obtained from pigs and from humans in Denmark were examined in a separate analysis. AMOVA (Table 6) revealed that nearly all of the genetic variance was distributed within the two populations (99.4%) and that only a small and insignificant part of the variance was distributed between the two groups (0.6%) (P = 0.08). Thus, it was concluded that all of the examined Danish Ascaris worms belonged to a single population.
ITS region polymorphisms. Amplification of the ITS region produced a fragment of approximately 1,000 bp. A two-, three-, or four-band pattern was found after digestion of the products with restriction enzyme HaeIII (Table 7), corresponding to three unique genotypes. The sizes of the fragments were approximately 580 and 420 bp (two fragments), 580, 230, and 190 bp (three fragments), and 580, 420, 230, and 190 bp (four fragments). The predominant patterns among the worms obtained from Danish pigs were the three- and four-band types, whereas the two-band type was dominant among the non-Danish human Ascaris worms. An overall chi-square test showed that worms obtained from Danish pigs and humans and worms obtained from humans in developing countries (Table 7) did not belong to one homogeneous population (2 = 88.0, P < 0.001). Pairwise chi-square tests showed that the distribution of band patterns in Danish human Ascaris worms did not differ from the distribution in worms obtained from Danish pigs (2 = 1.53, P = 0.46). These results support the assumption that all Danish worms, irrespective of host origin, were drawn from the same population.
Although the ITS DNA region exists in multiple copies (see below), it was assumed that the two- and three-band patterns represented homozygotic genotypes (two codominant alleles at one locus, i.e., 1,1 and 2,2) and that the four-band pattern represented the heterozygotic genotype (i.e., 1,2). By use of the TFPGA program, it was found that the genotype frequencies in the total population of worms obtained from humans (including worms from Danish patients) were significantly different from the expected frequencies at Hardy-Weinberg equilibrium (2 = 37.3, P < 0.0001), indicating that population subdivisions existed for these worms. In contrast, the genotype frequencies in the total population of worms obtained from Danish pigs and humans did not differ significantly from the expected frequencies at Hardy-Weinberg equilibrium (2 = 3.56, P = 0.06), indicating that worms obtained from both humans and pigs belonged to a random mating population in Denmark.
DISCUSSION
Ascaris is a common intestinal parasite in pigs raised in Denmark, with a maximum prevalence of 25 to 35% in large fattening pigs and gilts (33). In Denmark, infections caused by Ascaris worms are relatively rare in humans of all ages (2 per 10,000 per year) but are most common in small children living in or visiting rural areas and with a risk of contact with pig manure (36 per 10,000 per year). It has been suggested that up to 1% of Danes are infected and that transmission occurs abroad or through imported food (19, 27). Large roundworms obtained from humans have been designated with the species name A. lumbricoides, although it is known that cross-infections from pigs may occur (10, 17, 22), and no definite morphological criteria for discrimination between the two Ascaris species are known (12). However, based on epidemiological evidence, it has been suggested that Danish children may acquire Ascaris infections by exposure to pig manure (5, 6). This hypothesis was confirmed in the present study by the epidemiological data and by detailed genetic analysis of an international collection of Ascaris worms obtained from humans and from pigs.
The Ascaris worms were compared by AFLP analysis, a technique that detects DNA polymorphisms at multiple specific restriction sites simultaneously. AFLP analysis has been successfully used to reveal a high degree of genetic information from other nematode species (15, 24, 34). The number of polymorphic loci is more important than the allelic diversity at each locus for the statistical power of analysis used for population assignment (7). At present, AFLP analysis seems to have the highest potential for screening a maximum number of loci and is therefore useful for analysis of population diversity; however, due to the dominant nature of the AFLP data, analysis of the results was not straightforward. The results obtained from the AFLP analysis of the Ascaris worms were therefore analyzed by several different clustering methods.
The ME dendrogram shown in Fig. 2 depicts the results. All the of human Ascaris worms collected in the three developing countries, Bangladesh, Guatemala, and Nepal, were found in two clusters containing no worms from pigs. These two clusters may represent evolutionary lines of the human form of the parasite (i.e., A. lumbricoides). Worms obtained from pigs in both of the two developing countries, Guatemala and the Philippines, and from Denmark were located in a single mixed cluster. This cluster may represents the pig parasite A. suum. All worms obtained from Danes were assigned to this cluster of pig Ascaris worms without any subdivisions. The close relationship between worms obtained from Danish humans and pigs was confirmed by structure analysis (Table 4) and by AMOVA (Tables 5 and 6).
This is the first time that AFLP analysis has been used for the examination of Ascaris; thus, it was impossible to compare the results with published observations. Therefore, the worms were also examined by scoring of HaeIII restriction sites in the ITS region of the rRNA gene by PCR-RFLP analysis, a technique previously used by others (2, 28, 38). One or two HaeIII restriction sites are present in the ITS region of Ascaris; hence, two or three fragment patterns will be generated after digestion of the ITS region PCR product by this enzyme. The differences are due to a transversional change (C to G) that results in the loss of an HaeIII restriction site. The rRNA gene of which ITS is a part exists in multiple copies and, in general, only a single repeat type is found within an individual genome (38). Unequal crossing over and gene conversion are believed to cause this homogenization of the rRNA gene repeats (13). However, in Ascaris, the two types can be found together in one individual (2, 28). Samples from these "heterozygous" individuals therefore produce a four-band pattern on the gel. The occurrence of the three different band patterns in worms from the inferred clusters (Table 7) was in agreement with the results of an analysis of Ascaris worms collected in China (28), except that the distribution of the band patterns in the Ascaris worms obtained from Danish patients matched the distribution in the Ascaris worms obtained from pigs. We also found some instances of the two fragment types in worms obtained from Danish pigs. Anderson (2) found that most worms obtained from people living in North America had ITS regions characteristic of worms obtained from pigs; from a comparison with human and pig parasites obtained worldwide, it was concluded that the ITS region similarity to A. suum was not due to regional genetic variations in an A. lumbricoides population. The fragment pattern for the ITS region provided further evidence for the pig origin of Danish human Ascaris worms; however, the presence of the two-band type in worms obtained from Danish pigs demonstrated that, unfortunately, the number of bands cannot be used for distinguishing between A. suum and A. lumbricoides (38).
By using the two independent techniques, we demonstrated that all 27 worms obtained from Danish patients in the present study belonged to the same population as the worms obtained from pigs, and more than 80% of the patients had known contact with pigs or pig manure. The above results suggest that cross-infections from pigs are a common transmission route for human ascariasis in Denmark. An epidemiological study showed that 8% of Dutch primary schoolchildren and 7% of Swedish adults had antibodies to Ascaris (36), but ascariasis was seldom reported. It was shown that humans developed a smaller number of adult worms in the intestine when experimentally infected with eggs from pig Ascaris versus eggs from human Ascaris (35), and often no large worms were established at all (8, 18). These data, together with small inocula, may explain why more than 80% of our patients and the North American patients of Anderson (2) carried only a single worm. In Denmark, there is no registration system for ascariasis; therefore, it is normally a "hidden" disease. Only a few cases with multiple worms may be recorded. In Scotland, 53 cases were reported within a 5-year period (11); of these, 66% were considered not to be travel related and 79% were among adults. Ten out of 12 cases (83%) of ascariasis reported in Finland occurred in patients younger than 10 years old (30), and children under this age accounted for 76% of the cases in our survey. During our 9-month survey, approximately 0.3% of the children younger than 5 years old and living in rural areas in Viborg County were reported to be expelling Ascaris (5).
Our results demonstrate that all of the cases of ascariasis in Danish patients examined in this study can be ascribed to cross-infections with Ascaris from pigs. This conclusion is supported by the fact that none of the patients had a traveling history. Thus, human ascariasis is a zoonosis in Denmark. We cannot exclude the possibility that some cases of A. lumbricoides infections are imported to Denmark; however, we observed the human form of the parasite only among worms collected in developing countries and not among worms collected in Denmark. Zoonotic ascariasis may also be a problem in other industrialized countries having high hygienic standards for human waste (2), and precautions to control this parasitic disease deserve renewed attention.
ACKNOWLEDGMENTS
We thank the medical practitioners and ascariasis patients in Viborg County for the human Ascaris specimens; Erik Vestergrd, Freddy Juhl Nielsen, Henni Lybye, and Niels Kristiansen (Danish Crown) and Anne Grethe Hermann (Tican) for excellent cooperation and for assistance in collecting worms from the Danish pigs; and T. J. C. Anderson, S. Eduardo, A. Hall, and S. Williams-Blangero for providing the non-Danish worms.
This work was funded by the National Research Foundation under the auspices of The Danish Centre for Experimental Parasitology, The Royal Veterinary and Agricultural University, and the Viborg County Research Fund for Health Science.
REFERENCES
Anderson, J. D., S. Williams-Blangero, and T. J. Anderson. 2003. Spurious genotypes in female nematodes resulting from contamination with male DNA. J. Parasitol. 89:1232-1234.
Anderson, T. J. C. 1995. Ascaris infections in humans from North America: molecular evidence for cross-infection. Parasitology 110:215-219.
Anderson, T. J. C., and J. Jaenike. 1997. Host specificity, evolutionary relationships and macrogeographic differentiation among Ascaris populations from humans and pigs. Parasitology 115:325-342.
Anderson, T. J. C., M. E. Romero-Abal, and J. Jaenike. 1993. Genetic structure and epidemiology of Ascaris populations: patterns of host affiliation in Guatemala. Parasitology 107:319-334.
Astrup, I., and J. Prag. 2001. Ascariasis in Denmark. Ugeskr. Laeg. 80:1741-1747. (In Danish.)
Astrup, I., and J. Prag. 2001. Five outbreaks of human ascariasis caused by pig ascaris. Ugeskr. Laeg. 163:452-453. (In Danish.)
Bernatchez, L., and P. Duchesne. 2000. Individual-based genotype analysis in studies of perentage and population assignment: how many loci, how many alleles Can. J. Fish. Aquat. Sci. 57:1-12.
Buckley, J. J. C. 1931. An observation on human resistance to infection with Ascaris from the pig. Helminthology 9:45-46.
Chan, M. S., G. F. Medley, D. Jamison, and D. A. Bundy. 1994. The evaluation of potential global morbidity attributable to intestinal nematode infections. Parasitology 109:373-387.
Crewe, W., and D. H. Smith. 1971. Human infection with pig Ascaris (A. suum). Ann. Trop. Med. Parasitol. 65:85.
Crompton, D. W. T. 2001. Ascaris and ascariasis. Adv. Parasitol. 48:285-375.
Crompton, D. W. T. 1989. Biology of Ascaris lumbricoides, p. 9-44. In D. W. T. Crompton, M. C. Nesheim, and Z. S. Pawlowski (ed.), Ascariasis and its prevention and control. Taylor & Francis, London, England.
Dover, G. 1982. Molecular drive: a cohesive mode of species evolution. Nature 299:111-117.
Doyle, J. J., and J. L. Doyle. 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochem. Bull. 19:11-15.
Folkertsma, R. T., Rouppe van der Voort, J. N., K. E. de Groot, P. M. van Zandvoort, A. Schots, F. J. Gommers, J. Helder, and J. Bakker. 1996. Gene pool similarities of potato cyst nematode populations assessed by AFLP analysis. Mol. Plant-Microbe Interact. 9:47-54.
Galvin, T. J. 1968. Development of human and pig Ascaris in the pig and rabbit. J. Parasitol. 54:1085-1091.
Jaskoski, B. J. 1961. An apparent swine Ascaris infection of man. J. Am. Vet. Med. Assoc. 138:504-505.
Koino, S. 1922. Experimental infections on human body with ascarides. Jpn. Med. World 15:317-320.
Kolmos, H. J., and C. Pedersen. 2000. Parasitologic diagnosis. Ascaris lumbricoides. Ugeskr. Laeg. 162:3748.
Krauss, S. L. 2000. Accurate gene diversity estimates from amplified fragment length polymorphism (AFLP) markers. Mol. Ecol. 9:1241-1245.
Kumar, S., K. Tamura, and M. Nei. 2004. MEGA3: integrated software for molecular evolutionary genetics analyses and sequence alignment. Brief. Bioinformatics 5:150-163.
Lord, W. D., and W. L. Bullock. 1982. Swine Ascaris in humans. N. Engl. J. Med. 306:1113.
Macko, J. K., and P. Dubinsk. 1997. Taxonomic deliberations on human and pigs ascarids. Helminthologia 34:167-171.
Marche, L., S. Valette, E. Grenier, and D. Mugniery. 2001. Intra-species DNA polymorphism in the tobacco cyst-nematode complex (Globodera tabacum) using AFLP. Genome 44:941-946.
Meyer, A. D. S., A. A. F. Garcia, and A. P. D. Souza. 2004. Comparison of similarity coefficients used for cluster analysis with dominant markers in maize (Zea mays L). Genet. Mol. Biol. 27:83-91.
Nei, M. 1978. Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 89:583-590.
Nielsen, J. O. 2004. Worm and protozoan diseases in Nordic countries, p. 678-683. In N. E. Hansen, S. Hauns, and O. B. Schaffalitzky de Muckadell (ed.), Compendium in medicine, 14th ed. Nyt Nordisk Forlag, Copenhagen, Denmark. (In Danish.)
Peng, W., T. J. C. Anderson, X. Zhou, and M. W. Kennedy. 1998. Genetic variation in sympatric Ascaris populations from humans and pigs in China. Parasitology 117:355-361.
Pritchard, J. K., M. Stephens, and P. Donnelly. 2000. Inference of population structure using multilocus genotype data. Genetics 155:945-959.
Raisanen, S., L. Ruuskanen, and S. Nyman. 1985. Epidemic ascariasis—evidence of transmission by imported vegetables. Scand. J. Prim. Health Care 3:189-191.
Roepstorff, A. 2003. Ascaris suum in pigs: population biology and epidemiology. Danish Centre for Experimental Parasitology, Royal Veterinary and Agricultural University, Copenhagen, Denmark.
Roepstorff, A., O. Nilsson, C. J. O'Callaghan, A. Oksanen, B. Gjerde, S. H. Richter, E. O. Ortenberg, D. Christensson, P. Nansen, L. Eriksen, and G. F. Medley. 1999. Intestinal parasites in swine in the Nordic countries: multilevel modelling of Ascaris suum infections in relation to production factors. Parasitology 119:521-534.
Roepstorff, A., O. Nilsson, A. Oksanen, B. Gjerde, S. H. Richter, E. Ortenberg, D. Christensson, K. B. Martinsson, P. C. Bartlett, P. Nansen, L. Eriksen, O. Helle, S. Nikander, and K. Larsen. 1998. Intestinal parasites in swine in the Nordic countries: prevalence and geographical distribution. Vet. Parasitol. 76:305-319.
Semblat, J. P., E. Wajnberg, A. Dalmasso, P. Abad, and P. Castagnone-Sereno. 1998. High-resolution DNA fingerprinting of parthenogenetic root-knot nematodes using AFLP analysis. Mol. Ecol. 7:119-125.
Takata, I. 1951. Experimental infection of man with Ascaris of man and the pig. Kitasato Arch. Exp. Med. 23:49-59.
van Knapen, F., J. Buijs, L. M. Kortbeek, and I. Ljungstrom. 1992. Larva migrans syndrome: toxocara, ascaris, or both Lancet 340:550-551.
Vos, P., R. Hogers, M. Bleeker, M. Reijans, T. van de Lee, M. Hornes, A. Frijters, J. Pot, J. Peleman, M. Kuiper, et al. 1995. AFLP: a new technique for DNA fingerprinting. Nucleic Acids Res. 23:4407-4414.
Zhu, X., N. B. Chilton, D. E. Jacobs, J. Boes, and R. B. Gasser. 1999. Characterisation of Ascaris from human and pig hosts by nuclear ribosomal DNA sequences. Int. J. Parasitol. 29:469-478.(Peter Nejsum, E. Davis Pa)