INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Group A rotaviruses are responsible for severe gastroenteritis in humans and animals (6). After replication in the gastrointestinal tract, they are excreted and may be dispersed in environmental waters (22). Rotaviruses have been implicated in waterborne gastroenteritis outbreaks in many countries. The stability of human rotaviruses in environmental waters and their resistance to physicochemical treatment processes used by sewage treatment plants may facilitate their transmission (1, 4, 12).

A number of surveys have been performed to examine the rotavirus strains in circulation (26). Rotaviruses are characterized by the outer capsid proteins that are important for virus neutralization: a glycoprotein (G) encoded by gene segment 7, 8, or 9 and the protease-cleaved protein (P), encoded by gene segment 4 (5). Furthermore, strains behave as if there is a local reservoir, since the same strains are found from year to year in the same location. It has been shown, primarily by direct typing of rotaviruses in fecal specimens with monoclonal antibodies, that the most prevalent type G rotavirus strains causing childhood diarrhea worldwide are G1 through G4 (7, 19, 24, 25, 26), although some unusual strains have been documented on rare occasions and confirmed by both G and P typing (3, 17, 22). Efforts to characterize the genetic diversity existing in rotavirus VP4 genes and to identify nontypeable strains observed in some G serotypes has stimulated the development of P and G genotyping methods as a substitute for serotyping (9, 10, 11, 14, 18).

In this study, we demonstrate the usefulness of a simple method for correctly determining the G genotypes circulating in a geographic region of interest, allowing rapid epidemiological surveys of rotaviruses in clinical and environmental samples. Our approach consists of simple concentration of rotaviruses from surface water samples, a single-step reverse transcription (RT)-PCR for amplification of the entire VP7 gene, a nested PCR (nPCR), and subsequent sequence analysis of the 189-bp amplification product.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Rotavirus-positive stool specimens. Stool specimens obtained from children who had gastroenteritis and were admitted to the hospitals in Ticino or obtained directly from pediatricians were collected over a 1-year period (from January to December 1998). A total of 36 stool specimens which had been found positive for rotavirus antigen by the latex agglutination method (Slidex Rota-Kit 2; BioMérieux, Geneva, Switzerland) were diluted to 20% in Dulbecco medium, clarified by vortexing and centrifugation, and stored at 20°C.

Collection of water samples. One-liter samples of surface water and 1-ml samples of raw sewage were collected in sterile glass bottles and tubes, respectively, and stored at 4°C until concentration of the viruses was performed in the laboratory. Water samples were collected approximately each month over a 1-year period (from January to December 1998) from the rivers where major sewage treatment plants are located, i.e., in the three major effluents of Lake Lugano as well as the Ticino River. At the treatment plants, 31 samples were collected from raw sewage, 31 samples were collected from the treated water before release, and 34 samples each were collected from the rivers before and after receiving the treated water.

Concentration of rotaviruses from surface water samples. Viruses were concentrated from surface water with SiO₂. One liter of water was acidified to pH 3.5 with acetic acid, and then 200 µl of autoclaved SiO₂ (2, 8, 23; F. Baggi et al., submitted for publication) and AlCl₃ (to a final concentration of 0.5 mM) were added. Samples were shaken at room temperature for about 30 min. The SiO₂ was sedimented by centrifugation at 7,500 × g for 10 min; alternatively, it was allowed to settle and the supernatant was decanted. After removal of the supernatant, the pellet was transferred to a 1.5-ml tube; after the removal of excess liquid by a short centrifugation step, the viruses were eluted from the SiO₂ pellet with 280 µl of glycine buffer (pH 9.5), homogenized, and incubated at 64°C for 10 min. After centrifugation at 12,000 × g for 2 min, the supernatant was collected in a new sterile tube and stored at 80°C until the extraction of nucleic acids was performed.

Concentration of rotaviruses from raw sewage samples. One milliliter of raw sewage was centrifuged at 6,800 × g for 2 min. The supernatant was collected, and 20 µl of autoclaved SiO₂ (2, 8, 23; Baggi et al., submitted) and 50 µl of acetic acid were added. The samples were shaken at room temperature for 10 min and centrifuged at 6,800 × g for 2 min. After removal of the supernatant, virus particles were eluted from the SiO₂ pellet as previously described and stored at 80°C until the extraction of nucleic acids was performed.

Extraction of nucleic acids. RNA was extracted from 140 µl of clarified stool samples or glycine buffer eluates by using the QIAgen viral RNA extraction kit (QIAGEN AG, Basel, Switzerland) and strictly following product manufacturer instructions. The 50-µl RNA eluates were stored at 20°C until the amplification of nucleic acids was performed.

Primers. Primers END9, R1, R3, and Rp were used for the amplification of sequences from the VP7 gene of group A rotaviruses (16). The oligoprobe Rp corresponds to a conserved region located within the VP7 gene. The sequences, locations, and amplification products of the chosen primers within the rotavirus VP7 gene are given in Table 1.

Single-step RT-PCR. In order to simplify the amplification procedure, a single-step RT-PCR was developed. RT and PCR were performed with a 100-µl reaction volume. For efficient denaturation of the double-stranded RNA (dsRNA), 7 µl of dimethyl sulfoxide was added to 5 µl of extracted nucleic acid. After denaturation at 97°C for 5 min, the samples were immediately cooled on ice. Following a short centrifugation, the samples were added to a mixture consisting of 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl₂, 0.2 mM each deoxyribonucleotide triphosphate (Boehringer), 1 µM primer R1, 1 µM primer END9, 5 U of Superscript-RT^II reverse transcriptase (Gibco BRL), 2.5 U of Taq DNA polymerase (Boehringer Mannheim AG, Rotkreuz, Switzerland), and pyrocarbonic acid diethyl ester (DEPC)-treated H₂O (Merck Schweiz AG, Dietikon, Switzerland) to a final volume of 100 µl. After incubation for 30 min at 42°C, the following cycling conditions were chosen: 1 min at 94°C, 2 min at 55°C, and 1 min at 72°C, for a total of 25 cycles. A final extension at 72°C for 7 min followed by storage at 4°C concluded the RT-PCR amplification step. A fragment of 1,062 bp was expected.

nPCR. nPCR was performed with 5 µl of the RT-PCR amplification product under the following conditions: Taq Buffer(+MgCl₂) (Boehringer), 0.1 mM each deoxyribonucleotide triphosphate (Boehringer), 1 U of Taq DNA polymerase, 1 µM primer R3, 1 µM primer Rp, and H₂O-DEPC in a final volume of 50 µl. The cycling conditions were as follows: 30 s at 94°C, 30 s at 50°C, and 30 s at 72°C for 3 cycles, followed by 27 cycles of 15 s at 94°C, 15 s at 50°C, and 20 s at 72°C. A final extension at 72°C for 2 min followed by storage at 4°C concluded the nPCR amplification step. A fragment of 189 bp was expected.

Amplification control reactions. Simian rotavirus SA11 dsRNA (kindly provided by M. Weitz, Institute of Clinical Microbiology and Immunology, St. Gallen, Switzerland) was extracted by using the QIAgen viral RNA extraction kit and following product manufacturer instructions. The extracted dsRNA was diluted 100-fold with H₂O-DEPC and stored at 20°C until use. In order to detect false negatives due to RT-PCR inhibition, all samples that were found negative by nPCR were retested by RT-PCR after the addition of 5 µl of 100-fold-diluted simian rotavirus SA11 dsRNA as a control.

Visualization of RT-PCR and nPCR products. RT-PCR fragments (1,062 bp) and nPCR fragments (189 bp) were analyzed by gel electrophoresis. Ten microliters of the amplification products was electrophoresed on 1.5% agarose gels in Tris-borate-EDTA buffer along with a 100-bp DNA ladder (Boehringer). Fragments were visualized after ethidium bromide staining.

Sequencing and analysis of PCR products. Templates for sequencing were prepared by purification of the amplified products (1,062 or 189 bp) by using a QIAquick PCR purification kit (QIAgen) and following product manufacturer instructions. Sequencing reactions were performed with an ABI Prism dRhodamine dye terminator cycle sequencing ready reaction kit (Perkin-Elmer Applied Biosystems International, Inc.) on a Perkin-Elmer Cycler GeneAmp 9600 system. The cycling conditions followed product manufacturer instructions. Purification of sequencing products, necessary to remove excess dye terminator, was performed using columns from CENTRI SEP Princeton Separations, Inc., Adelphia, N.J., and 10 µl of the purified sequencing products was resuspended in 10 µl of template suppressor reagent. After denaturation at 94°C for 2 min, samples were rapidly cooled on ice, briefly vortexed, centrifuged, and electrophoresed on an ABI Prism 310 instrument (Perkin-Elmer Applied Biosystems) by using the Seq POP6 (1ml)E module. DNA sequences were aligned with the Lasergene program Megalign (DNAstar).

Phylogenetic VP7 sequence analysis and accession numbers. Selected sequence data were aligned with the Lasergene program Megalign, and statistical and phylogenetic analyses were performed using Molecular Evolutionary Genetics Analysis version 1.01 (15). Phylogenetic trees were reconstructed by the neighbour joining method. The number of nucleotide substitutions per site was estimated by several distance methods, and the robustness of each node was tested by bootstrap analysis with 500 replications. Phylogenetic trees were drawn using TreeExplorer 2.01 (Koichiro Tamura).

Sensitivity of rotavirus concentration from water samples. The sensitivity of virus concentration from water samples was determined as follows: six 1-liter sterile water samples were spiked with serial dilutions of simian rotavirus SA11 (final concentrations, 3.2 × 10⁴, 3.2 × 10², 3.2 × 10¹, 3.2, 3.2 × 10¹, and 3.2 × 10² 50% tissue culture infective doses [TCID₅₀]/liter, respectively). Concentration, elution, extraction, and amplification were performed as previously described.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Evaluation of the sensitivity of rotavirus concentration from water samples. Only the sample seeded with 3.2 × 10⁴ TCID₅₀/liter displayed the 1,062-bp RT-PCR fragment; nPCR increased the sensitivity to 3.2 TCID₅₀/liter (Fig. 1).

View larger version (43K): [in this window] [in a new window]   FIG. 1.   Sensitivity of rotavirus concentration from 1 liter of water, as analyzed by electrophoresis of amplification products. Lanes: 1, 100-bp DNA ladder; 2, 3.2 × 104 TCID50/liter; 3, 3.2 × 102 TCID50/liter; 4, 3.2 × 101 TCID50/liter; 5, 3.2 TCID50/liter; 6, 3.2 × 101 TCID50/liter; 7, 3.2 × 102 TCID50/liter; 8, RT-PCR and nPCR negative controls.

Phylogenetic analysis of rotavirus VP7 complete (1,062-bp) and partial (189-bp) sequences. Sixty-two human VP7 sequences of group A rotavirus deposited in GenBank were selected to represent all the major VP7 serotypes (serotype G1, 12 sequences; serotype G2, 17 sequences; serotype G3, 20 sequences; serotype G4, 11 sequences; and serotype G8, 2 sequences). Two published human VP7 sequences of group C rotavirus were also included. Alignment and phylogenetic analysis of both the whole, 1,062-bp VP7 gene (RT-PCR product) and the 189-bp fragment (nPCR product) were performed as previously described in Materials and Methods. Results of both analyses were in agreement, allowing correct G genotyping of all sequences by examination of only the 189-bp nPCR product. In order to determine whether we could distinguish between animal VP7 sequences which could be found in environmental water samples and sequences of human origin, phylogenetic analysis of the 189-bp fragment included different published animal rotavirus VP7 sequences as well. All sequences of nonhuman animal origin generated distinct clusters clearly distinguishable from the human group A rotavirus sequences; therefore, it is possible to correctly discriminate between sequences of human and animal origins. Part of the phylogenetic sequence analysis is displayed in Fig. 2.

View larger version (100K): [in this window] [in a new window]   FIG. 2.   Phylogenetic sequence analysis of rotavirus VP7 sequences (189 bp) of human and animal origins (neighbor joining tree): human group A rotavirus sequences from different serotypes (REF strains); RG81-09 (rotavirus type 8 isolated from a river); HGC 1-39 and HGC 1-40 (human group C rotavirus); AVI 7-10 (avian rotavirus); CAN 2-37 (canine rotavirus); CHI 5-01 (chicken rotavirus); EQU 2-18 (equine rotavirus); LAM 7-30 (lamb rotavirus); LAP 1-35 (lapin rotavirus); MUR 1-10, MUR 1-33, and MUR 6-01 (murine rotavirus); POR 3-31 and POR 3-32 (porcine rotavirus); and POR 7-02 (porcine group C rotavirus). Numbers indicate percent similarity.

Phylogenetic analysis of rotavirus VP7 partial (189-bp) sequences from stool and water samples. After RT-PCR amplification, all 36 clinical samples displayed the 1,062-bp fragment, and sequencing was performed with both R1 and END9 primers. After phylogenetic analysis of the corresponding 189-bp sequence, 29 patients displayed type G1 sequences, 5 patients exhibited type G2 sequences, and 2 patients had type G4 sequences. No sequences matching type G3 or G8 were found. For environmental water samples, the use of nPCR was indispensable in order to enhance the sensitivity of detection. In fact, all positive water samples were negative after RT-PCR but positive after nPCR. The absence of inhibitors during nucleic acid amplification was assayed for each nPCR-negative sample. A total of 16 positive water samples were found and were sequenced with both R3 and Rp primers. All samples were classified as G1. The positive water samples were found in January (seven positive samples), May (five positive samples), and June (four positive samples), in accordance with the observation that rotavirus infections have peak occurrences in winter and spring months. The results of the phylogenetic sequence analysis of all positive samples are displayed in Fig. 3.

View larger version (106K): [in this window] [in a new window]   FIG. 3.   Phylogenetic sequence analysis (189 bp) of rotavirus-positive samples (neighbor joining tree) (CLIN, clinical stool samples [from children]; H2O, water samples; CARRIER, stool sample from a healthy adult; and REF, human group A rotavirus sequences from different serotypes [HG1, HG2, HG3, HG4, and HG8]).

Carriers of rotaviruses. In order to assess if healthy adults could serve as reservoirs for rotaviruses found in environmental water or raw sewage, 20 stool samples from healthy, asymptomatic adults (collected for routine sanitary inquiries) were analyzed by both RT-PCR and nPCR. All samples but one were found negative by nPCR. Sequence analysis of the positive sample revealed type G1 (Fig. 3).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The present study demonstrates the efficiency of a single-step RT-PCR followed by nPCR and subsequent sequence analysis of a 189-bp product from the 5' end of the VP7 gene for characterizing rotavirus contamination of environmental water samples and the usefulness of these techniques for studying epidemiological relationships between strains involved in human infections and those found in water. The single-step RT-PCR was developed in order to avoid reopening of the tubes and therefore to reduce the danger of cross contamination. Furthermore, this strategy reduced the technical steps necessary to perform the analysis while maintaining good sensitivity and specificity for detecting low viral loads in both human and water samples. Concentration of viruses from water using SiO₂ is easy, inexpensive, and rapid and has good sensitivity and specificity. An nPCR result can be obtained 24 h after sampling, and a sequence profile can be obtained after a further 24 h. These properties allow rapid (within 48 h) detection of possible contamination sources, allowing appropriate preventive measures to be undertaken in time. Inhibitors were efficiently removed by RNA purification and extraction with the QIAgen viral RNA extraction kit.

Raw sewage and secondary effluent samples were found to be contaminated by rotaviruses despite the most efficient water treatment measures in use in sewage treatment plants (Baggi et al., submitted). Sequence analysis of the positive water samples revealed the presence of only rotavirus type G1, whereas among the clinical samples, type G1 predominated but types G2 and G4 also could be found. The fact that in healthy adults rotavirus could be detected by nPCR sustains a possible link between rotaviruses found in the environment and those found in clinical samples. In fact, our results support the presumption that healthy adults may serve as reservoirs for the spread of rotaviruses (12, 13, 20, 21). The knowledge of the sensitivity of detection of rotaviruses by RT-PCR and nPCR (3.2 × 10⁴ and 3.2 TCID₅₀/liter, respectively) makes it possible to gain useful information about the levels of rotavirus that one healthy adult may be shedding.

The phylogenetic analysis of different animal rotavirus VP7 sequences resulted in a clearly distinct cluster that did not match any of the human group A rotavirus sequences. This is a very fortuitous finding, because it allows the application of this molecular technique to environmental water samples to correctly discriminate between sequences of human and animal origins.

In conclusion, the present study demonstrates the efficiency of the nucleotide sequence analysis of 189 bp at the 5' end of the VP7 gene for studying epidemiological relationships between strains involved in human infections and those found in water. From the sequence analysis, it was possible to predict the G genotype and to evaluate possible correlations between water contamination and viral sequences found among clinical samples.