Previous Article | Next Article ![]()
Journal of Clinical Microbiology, December 2005, p. 5940-5944, Vol. 43, No. 12
0095-1137/05/$08.00+0 doi:10.1128/JCM.43.12.5940-5944.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.
LCPME-UMR 7564 CNRS-UHP, Equipe Microbiologie et Physique, Faculté de Pharmacie, 5 rue Albert Lebrun, BP 80 403, 54001 Nancy Cedex, France
Received 16 June 2005/ Returned for modification 2 August 2005/ Accepted 28 September 2005
|
|
|---|
|
|
|---|
Among the six species identified in the Giardia genus, only G. lamblia infects humans and numerous other mammals as well (1, 25). Moreover, isolates of G. lamblia are classified into seven assemblages, based on the characterization of the glutamate dehydrogenase (gdh), small-subunit (SSU) rRNA, and triosephosphate isomerase (tpi) genes (12, 18, 20, 21). Assemblages A and B infect humans and a broad range of other hosts, including livestock, cats, dogs, and wild mammals. The assemblage A isolates have been further grouped into subgroups I and II. The assemblage B isolates have been separated into subgroups III and IV (17, 24). Genetic assemblages C, D, E, F, and G appear to be host restricted to domestic animals, livestock, and wild animals (19, 21).
At present, antigen detection immunoassays for Giardia are used as the routine diagnostic procedure of choice in many hospitals and public health laboratories (8, 13, 27). However, these methods are unable to differentiate between the genetic asemblages of Giardia lamblia. Molecular detection methods based on PCR have been developed to detect G. lamblia cysts in feces. These techniques have numerous advantages in terms of sensitivity, speed, and specificity in comparison to conventional methodologies (3, 16). Moreover, these molecular techniques may allow the genotyping of Giardia lamblia cysts (3, 4, 20).
We previously described a method for the successful extraction and detection of giardial DNA from naturally contaminated wastewater (5). In the present study, we evaluated primers for rapid and sensitive classification of G. lamblia cysts from human feces into assemblages A and B. These primers were previously designed for the detection and quantification of G. lamblia assemblages A and B in environmental samples by real-time PCR (data not shown). In this study, we evaluated the distribution of these major assemblages in sporadic human giardiasis in France and also in samples from 15 French soldiers coming back from the Ivory Coast. Moreover, the simultaneous use of our assay and two previously described PCR-restriction fragment length polymorphism (RFLP) assays allowed comparison of the tpi and gdh genes for detection and genotyping of G. lamblia.
|
|
|---|
DNA samples. Three DNA samples of Giardia lamblia from cattle, sheep, and pigs were included in this study. A sequencing analysis had showed 100% matches between the amplified products (165 bp) obtained with these samples and the sequence with GenBank accession number AF069559 corresponding to G. lamblia assemblage E strain P-15 (data not shown). DNA recovered from Cryptosporidium parvum (Institut National de la Recherche Agronomique, Nouzilly, France), Entamoeba histolytica, Entamoeba dispar (Institut de Parasitologie, Strasbourg, France), Campylobacter jejuni (ATCC 29428; Central Hospital, Nancy, France), Salmonella enterica Serovar Typhimurium WG49, and Escherichia coli K12 Hfr was also included in this study.
Fecal samples. Forty-one fecal samples from patients in which Giardia cysts had been detected by conventional techniques were collected during this study; 26 fecal samples originated from patients with sporadic cases of giardiasis diagnosed between October 2000 and September 2004. Of these, 18 samples were provided by three civilian hospitals (Centre Hospitalier Universitaire Brabois, Nancy; Pitié-Salpêtrière Hospital, Paris; and Bichat Hospital, Paris, France) and eight samples from adults were sent by a military hospital (Legouest Army Hospital, Metz, France). Of the 18 patients in civilian hospitals, 61% were adults aged between 21 to 61 years and 78% were males (Table 1). No information about the immune status of these patients was provided by the laboratories. Fifteen samples from French soldiers (males aged between 23 and 50 years) were also collected from Legouest Army Hospital. These 15 cases of giardiasis were diagnosed in April 2004 after a 4-month stay in the same city in Ivory Coast (West Africa).
|
View this table: [in a new window] |
TABLE 1. Results of PCR and PCR-RFLP analyses of the Giardia lamblia tpi and gdh genes amplified from 26 samples corresponding to sporadic cases
|
Microscopy. The staining procedure was carried out as previously described (5).
DNA extraction. All buffers and reagents used in this step are provided in the QIAamp DNA stool kit (QIAGEN, Courtaboeuf, France), except buffer AVL and RNA carrier (Qiagen). The DNA extraction was carried out on a 200-µl purified sample. Three modifications of this protocol were employed as previously described (5). (i) Cyst wall lysis and adsorption of impurities were improved by increasing the time of incubation at 95°C to 10 min and the time of incubation with InhibitEx to 3 min; (ii) 5 µl RNA carrier (1 µg/µl in buffer AVL) was added during the protein digestion step; and (iii) protein digestion and column purification were applied twice per sample. After the first elution of DNA extract from the column, the protocol was started again at the protein digestion step. The DNA was then stored at 80°C.
PCR amplification and restriction fragment length polymorphism analysis. (i) Oligonucleotide primers. Five sets of oligonucleotide primers obtained from Proligo France were used for the analysis of fecal samples. Two sets of primers for detection of Giardia lamblia assemblages A and B were designed against the coding region of the tpi gene using the Primer Express Oligo Design software (v. 1.5; Applied Biosystems). ClustalX (v. 1.8) was used for the determination of the variations of sequence between the following Giardia species and genotypes (GenBank accession number): G. lamblia assemblage A: WB (L02120), JH (U57897), Ad-1 (AF069556), Ad-2 (AF069557), isolate 2907 (AY228647), isolate from wild deer (AY302562); assemblage B: GS/M (L02116), BAH-12 (AF069561), Ad-19 (AF069560), isolate 2924 (AY228628), isolate 2582 (AY228629), isolate 2506 (AY228630), isolate 2887 (AY228631), isolate 2902 (AY228632), isolate 2877 (AY228633), isolate 2900 (AY228634), isolate 2901 (AY228635), isolate 3470 (AY228636), isolate 3565 (AY228637), isolate 3577 (AY228638), isolate 1758 (AY228639); assemblage C: Ad-136 (AF069563), isolate 2643 (AY228641), isolate 2669 (AY228642), isolate 2674 (AY228643), isolate 2665 (AY228644); assemblage E: P-15 (AF069559), isolate 109 (AY228645), isolate 15 (AY228646); assemblage F: Ad-23 (AF069558); and assemblage G: Ad-157 (AF069562), isolate 2135 (AY228640); G. ardeae (AF069562); G. microti: isolate 3463 (AY228648), isolate 3460 (AY228649); G. muris (AF069565).
The primers used for assemblage A amplification were forward (A-for) 5'-GGAGACCGACGAGCAAAGC-3' (positions 839 to 857 on the WB sequence, GenBank no. L02120), and reverse (A-rev), 5'-CTTGCCAAGCGCCTCAA-3' (positions 970 to 986 on the WB sequence). A 148-bp fragment of the assemblage A gene was amplified with primers A-for and A-rev (A-PCR). The primers used for assemblage B amplification were forward (B-for), 5'-AATAGCAGCACARAACGTGTATCTG-3' (positions 126 to 150 on the BAH-12 sequence, GenBank no. AF069561), and reverse (B-rev), 5'-CCCATGTCCAGCAGCATCT-3' (positions 188 to 206 on the BAH-12 sequence). An 81-bp fragment of assemblage B gene was obtained with primers B-for and B-rev (B-PCR). Primers sets TPIAF-TPIAR (TPIA-PCR) and TPIBF-TPIBR (TPIB-PCR) were used as previously described (2) for amplification of the G. lamblia tpi gene from assemblages A and B, respectively. Finally, primer set GDHiF-GDHiR was used for amplification of the G. lamblia gdh gene, as previously described (20).
(ii) PCR amplification and RFLP analysis. Amplification of the tpi and gdh genes was performed as a single PCR. Amplification reactions (50 µl) contained 5 µl of DNA, 1x PCR buffer corresponding to a final concentration of 1.5 mM MgCl2 (Qiagen), each deoxynucleotide triphosphate at a concentration of 200 µM (Applied Biosystems), each primer at a concentration of 0.5 µM, and 2.5 U of HotStarTaq DNA polymerase (Qiagen). Cycling parameters were 15 min at 95°C (initial heat activation step), followed by 50 cycles of 30 s at 94°C, 30 s at 62°C, and 30 s at 72°C, with a final extension of 7 min at 72°C. Both positive and negative controls were included in each PCR to validate results. Quantities of DNA equivalent to 600 G. lamblia assemblage A (purified from feces) cysts and 4,500 G. lamblia assemblage B (strain H-3) cysts were used as the templates for the positive controls, and distilled water was used as the template for negative controls throughout.
RFLP analysis was performed by digesting 2.5 µl of PCR product with 2 U of RsaI (Promega) or 2 U of NlaIV (New England Biolabs) in 1x enzyme buffer in a final volume of 20 µl for 3 h at 37°C. The RsaI digestion allowed the distinction between assemblage A group I and group II after amplification with the TPIAF and TPIAR primers (2). The NlaIV digestion was used for the distinction between assemblage A group I, assemblage A group II, and assemblage B after amplification with the GDHiF and GDHiR primers. RsaI digestion distinguished between assemblage B group III and assemblage B group IV after use of the GDHiF and GDHiR primers (20).
(iii) PCR product and restriction fragment detection. PCR products and restriction fragments were separated by horizontal electrophoresis in 2 and 3.2% agarose gels, respectively, with ethidium bromide (0.6 µg/ml) staining. A 100-bp DNA ladder (Promega) was included as a size marker. PCR products and restriction fragments were recorded by UV transillumination.
|
|
|---|
The specificity of our PCR assay was first evaluated by subjecting the primers to a BLAST test (http://www.ncbi.nlm.nih.gov/BLAST/). For the A-for and A-rev primers, the BLAST tests returned sequences other than G. lamblia assemblage A, and only one sequence showed a 100% match with the A-rev primer. However, the sequences other than G. lamblia assemblage A returned by these two BLAST tests were different. For the B-for and B-rev primers, the BLAST tests returned sequences other than G. lamblia assemblage B with one to eight mismatches. Furthermore, the mismatched sequences returned by these two BLAST tests were different.
The specificity was then examined by performing PCR assays. By using a quantity of DNA equivalent to 600 Giardia lamblia assemblage A cysts per reaction, the predicted 148-bp product was obtained by A-PCR, but no product was amplified by B-PCR. Conversely, by using a quantity of DNA equivalent to 4,500 Giardia lamblia assemblage B (strain H3) cysts per reaction, the predicted 81-bp product was observed with B-PCR, but no product was obtained with A-PCR. No product was amplified by performing either A-PCR or B-PCR with DNA extracted from purified Giardia lamblia assemblage E from cattle, sheep, and pigs, Giardia muris, Cryptosporidium parvum, Entamoeba histolytica, Entamoeba dispar, Campylobacter jejuni, Salmonella enterica, or Escherichia coli.
To estimate the sensitivities of our PCR assays, 10-fold dilutions of purified cysts were performed in PBS (pH 7.2) prior to DNA extraction. The amplification step was then performed in triplicate on 10-fold serial dilutions of cysts. For G. lamblia assemblage A, the predicted 148-bp product was obtained for each well of the triplicates for concentrations ranging from 600 to 6 cysts per reaction mixture. For G. lamblia assemblage B, the 81-bp product was observed for each well of the triplicates for concentrations ranging from 4,500 to 4.5 cysts per reaction mixture.
Among the 26 fecal samples corresponding to sporadic giardiasis diagnosed in hospital laboratories, the presence of Giardia cysts was reconfirmed by immunofluorescence (IF) microscopy for all of these during our study. On the other hand, the number of positive samples by PCR varied according to the PCR assay employed and the target gene amplified. Indeed, the tpi gene was amplified from 25 samples (96.1%) with A-PCR and B-PCR developed in our laboratory. The same target gene was amplified from 24 samples (92.3%) with TPIA-PCR and TPIB-PCR. However, only 18 positive samples (69.2%) were obtained with a single PCR. A double PCR allowed the achievement of six additional positive samples. For the gdh gene, 21 samples (80.7%) were positive with the GDHiF and GDHiR primers. Moreover, a single PCR resulted in 19 positive samples (73.1%) and a double PCR was needed for two other samples.
Regarding classification into assemblages A and B (Table 1), the results of the various PCR and PCR- RFLP methods were in agreement for the sporadic cases. Assemblage A was detected in nine (34.6%) and seven (26.9%) samples by using the tpi and gdh genes as the target, respectively. When targeting the tpi gene, assemblage B was identified in 16 (61.5%) and 15 (57.6%) samples with B-PCR and TPIB-PCR, respectively. However, when using the gdh gene as the target, assemblage B was detected in only 13 (50%) fecal samples.
A more specific analysis of these sporadic cases was then conducted with the distinction between assemblage A groups I and II and assemblage B groups III and IV on the basis of their fragment patterns obtained by RFLP analysis (Table 1). The results obtained with the two published methods were in agreement for sporadic cases. Whatever the gene used, subgroup I was never observed in these fecal samples. On the other hand, six samples were genotyped as assemblage A group II with both RFLP techniques and two samples were classified in this group with only one of these methods. For one sample classified in assemblage A by PCR assays targeting the tpi gene, the group could not be specified. Indeed, the quantity of PCR product obtained with A-PCR was insufficient for RFLP analysis, and no product was amplified with the GDHiF and GDHiR primers.
Only one method used in our study allowed distinction between assemblage B groups III and IV (20). The RFLP analysis resulted in eight and five samples classified in assemblage B group III and group IV, respectively. For three samples identified as assemblage B with our PCR assay, the group could not be specified with the method described by Read and colleagues (20). Of these, two samples were negative after double PCR with the GDHiF and GDHiR primers. For the third sample, the quantity of PCR product obtained with the GDHiF and GDHiR primers was insufficient for RFLP analysis. The genotyping results are summarized in Table 2.
|
View this table: [in a new window] |
TABLE 2. Summary of genotyping results obtained with the three methods
|
Among the 15 cases of giardiasis diagnosed in samples from French soldiers coming back from the Ivory Coast, the presence of Giardia cysts was reconfirmed by IF microscopy for 100% of the samples. The Giardia lamblia assemblage B genotype was isolated in 93.3% of the samples with the three methods, and tpi or gdh gene fragments were not amplified from only one sample. Moreover, the PCR-RFLP analysis developed by Read and colleagues (20) allowed the detection of assemblage B group III in all of these samples (Table 2).
|
|
|---|
In the study conducted by Amar and colleagues (2), amplification of the tpi gene was performed as a two-step PCR (seminested PCR) and resulted in 94% positive samples (n = 35). The ethyl acetate procedure, which is adapted from the formalin ethyl acetate method, was tested for fecal samples. This method allowed rapid and efficient purification of Giardia cysts prior to DNA extraction. Indeed, at least 15 fecal samples were simultaneously purified with this method, whereas the gel filtration chromatography method was time-consuming and limited to four samples per purification step. The ethyl acetate procedure is well adapted for the purification of Giardia cysts in wastewater samples prior to quantification by a real-time PCR method (data not shown). Moreover, this purification method could be used for the purification of Giardia cysts in sludge samples.
The present study provides, for the first time, information on the distribution of the genotypes of G. lamblia from humans with sporadic giardiasis in France. This work was based on a relatively small group of patients, but we are still pursuing analysis of sporadic giardiasis. However, our observation that the majority of sporadic giardiasis case isolates were assemblage B genotype (61.5%) corresponds to the findings of several studies conducted in India (100%, n = 10 [21]), Peru (76%, n = 25 [21]), United States (80%, n = 15 [11]), and United Kingdom (64%, n = 35 [2]). However, an Italian study reported 80% assemblage A in 30 stool samples examined by sequencing or PCR-RFLP analysis of the ß-giardin gene (6). The predominance of assemblage B in samples collected in sewage treatment facilities was shown in one study (10). On the other hand, two other studies concluded that these was a majority of assemblage A in wastewater samples (6, 22). The use of RFLP analysis resulted in the same percentage for assemblage A group II and assemblage B group III, whereas assemblage B group IV was less common. The absence of assemblage A group I in these 26 fecal samples corresponds to the findings of the study conducted in the United Kingdom (2).
Among the 15 cases of giardiasis diagnosed in French soldiers coming back from the Ivory Coast, the detection of assemblage B group III in 14 confirmed the hypothesis of a common source of contamination. At present, only two studies have reported the genotyping results from outbreak-associated giardiasis. Among 24 samples from a nursery outbreak, Amar and colleagues (2) detected G. duodenalis assemblage B in 88% of the samples. Sulaiman and colleagues (21) detected assemblage B in two isolates from a food-borne outbreak by sequencing analysis of the tpi gene.
In our study, G. lamblia assemblage A and assemblage B were never detected together, whereas a mixture of these assemblages has been reported previously in a few studies (2, 11, 12, 14, 15). Thus, Amar and colleagues (2) observed a mixture of assemblage A group II and assemblage B in 9% of 35 samples, whereas Lu and colleagues (15) simultaneously detected assemblage A group I and assemblage B in 33% of only three samples.
In summary, the PCR assay developed in this study, combined with the ethyl acetate procedure, allowed rapid detection and genotyping of G. lamblia cysts from clinical samples. Our results show that the tpi gene is better adapted than the gdh gene for efficient discrimination between the two major assemblages. Thus, detection methods targeting loci with a high degree of polymorphism such as tpi can be extremely useful when a common source of contamination is certainly involved.
This work provides the first information about the distribution of the two major assemblages of G. lamblia in sporadic human giardiasis in France. However, further studies with a larger series of fecal or environmental samples could lead to better knowledge of the distribution of these assemblages in humans as well as the role of domestic animals and livestock as a potential source of infection for humans.
This work was supported by European Community contract number TOFPSW EVK1.2000.22080.
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2010 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»