Forty-three swine faecal samples were gathered as already described in a previous study . Briefly, stools were sampled from young and adult pigs from different swine premises in Flanders (Belgium), where the animals showed clinical/necropsy signs of enteritis. The samples were gathered over a period of three months (June, July, August 2007). Other swine stool samples (n=111) originated from the Netherlands. A first series of 18 samples, tested by the Central Veterinary Institute (CVI, Lelystad, the Netherlands), had a previously known (but not communicated before testing) status for PoSaV and PoNoV infection, while the remaining 93, sampled from finishing pigs at slaughterhouse, had an unknown status. Positive controls for PoSaV strains belonging to GGVII and VIII were kindly provided by the CVI.
Bovine stool samples positive for bovine norovirus (BoNoV) and bovine kobuvirus (BoKoV) sequences had been gathered during previous studies [24–26]. Human stool samples positive for norovirus (HuNoV) or HuSaV were either gathered during studies in the laboratory (confirmed after RT-PCR amplification either by NoV-specific or pan-calicivirus primers and sequencing reactions), or were kindly provided by Dr P. Huynen (ULg) and the CVI (HuSaV GI.2). Feline calicivirus (FCV)-positive samples were sampled with nasal swabs on cats showing clinical signs of upper respiratory tract disease or in stool of dogs and cats with enteritis. These samples were confirmed as positive by RT-PCR with FCV-specific primers and sequencing reactions.
RNA extractions and conventional RT-PCR
The QIAamp Viral RNA Mini kit (Qiagen GmbH, Hilden, Germany) was used to perform RNA extractions on the Belgian samples and samples with known status from the Netherlands, using ten grams of stool ten-fold diluted in PBS complemented with 0.01% sodium azide. RNA extractions on the remaining 93 samples from the Netherlands were performed after a ten-fold dilution in PBS using the High Pure RNA Isolation kit (Roche, Mannheim, Germany).
Two microlitres of each viral RNA extraction was subjected to a one-step RT-PCR reaction using the Quick Access kit (Promega, Madison, WI, USA) in a mix containing the p289-290 primers  (400 nM final concentration; p290: GATTACTCCAAGTGGGACTCCAC; p289: TGACAATGTAATCATCACCATA) and bovine serum albumin (BSA, 400 ng/μl final concentration). Reverse transcription was performed at 48°C for 10 min and the PCR parameters were as follows: 3 min at 95°C, 45 cycles of 1 min at 95°C, 45 sec at 51°C and 45 sec at 68°C, and 7 min at 68°C for the final extension step.
RT-PCR amplicons were purified from agarose gel with the QIAquick purification kit (Qiagen GmbH, Hilden, Germany) or by precipitation, and were directly sequenced twice or cloned (PC45, 46, 47, 48, 49 and 63) into pGEM-Teasy (Promega, Madison, WI, USA) before sequencing. Plasmid DNA was purified on three clones for each previously mentioned amplicon with the Miniprep kit (Invitrogen, Carlsbad, CA, USA). Sequencing reactions were carried out with BigDye terminator kit version 3.1 and resolved with an ABI 3730 automatic capillary sequencer (AppliedBiosystem, Foster City, CA, USA).
Sequences were analysed with the BioEdit Sequence Editor version 7.0 software . Nucleotidic similarity with the NCBI genetic database was assessed using the BLAST tool (available at http://www.ncbi.nlm.nih.gov/blast/Blast.cgi). Phylogenetic inference was performed with the MEGA version 4 software package . Phylogenetic trees were constructed by neighbor-joining analysis where evolutionary distances were computed using the Maximum Composite Likelihood method. The confidence values of the internal nodes were calculated by performing bootstrap analyses with 1,000 replicates.
Standard of quantification
A p289-290 amplicon, amplified on a canine sample, was cloned into pGEMt-Easy (pSTD) (Promega, Madison, WI, USA) and sequenced at the GIGA facilities of the University of Liège. Sequencing was carried out with the BigDye terminator kit version 3.1 and resolved with an ABI 3730 automatic capillary sequencer (AppliedBiosystem, Foster City, CA, USA). The p289-290 amplicon was then in vitro transcribed with the Ribomax kit (Promega, Madison, WI, USA) following the manufacturer’s instructions. Briefly, 4 μg of pSTD were linearised with the SpeI restriction endonuclease and purified with phenol:chloroform:isoamyl alcohol (25:24:1). DNA was precipitated with 3M Na acetate and ethanol purified, and the concentration was then measured on a Nanodrop 1000 spectrophotometer (Technilab). The transcription was performed on 10 μg of DNA with T7 RNA polymerase and the reaction mix was then treated with 5U DNAse for 1 h at 37°C. The transcribed RNA was controlled for quality and purity (DNA absence) by both RT-PCR (Quick Access kit, Promega, Madison, WI, USA) and PCR (Taq polymerase from Westburg, Leusden, the Netherlands) with the p289-290 primers. RNA concentration was measured on a Nanodrop 1000 spectrophotometer (Technilab). Genomic copies were deduced and serial ten-fold dilutions were prepared with ultrapure RNAse free H2O (Invitrogen, Carlsbad, USA). Aliquots of the master stock were stored at −80°C and measured once again before dilution and use.
SYBR green qRT-PCR
The following mixture was constituted using the iScript One-Step RT-PCR kit for SYBR Green assay (Biorad, Nazareth, Belgium): p289-290 primers (300 nM each), 0.5 μl enzyme mix, 12.5 μl reaction buffer, BSA (400 ng/μl final concentration), 2 μl of RNA extraction, ultrapure RNAse free H2O (Invitrogen, Carlsbad, USA) to 25 μl. The protocol included a reverse transcription step of 18 min at 48°C, an initial denaturation step of 5 min at 95°C, 45 cycles of 10 sec at 95°C, 20 sec at 51°C, 45 sec at 60°C and a final extension step of 3 min at 95°C. Data were obtained during the elongation period. Melting curve analysis was performed after RT-PCR reaction at a start- and an end-temperature of 65°C and 95°C respectively. The real time RT-PCR reactions were carried out on an iCycler thermocycler (Biorad, Nazareth, Belgium).
Evaluation of the assay parameters
RNA extractions from three PoSaV-positive samples (named PC34, PC29 and PC42) were selected on the basis of phylogenetic analysis performed on partial RdRp sequences during a previous study , RNA extractions from three PoSaV-positive samples were selected: PC34, PC29 and PC42. Their sequences were genetically related to GGIII, and to the proposed GGVI and proposed GGVII respectively. RNA extractions from two PoNoV-positive swine samples (PC23, PC26) , two HuSaV- (Be1, Be2) and two HuNoV-positive human samples (ISP475, CrH2) were also incorporated into the melting curve study.
Assay repeatability and reproducibility
Assay repeatability and reproducibility were established by determination of the mean melting temperature, standard deviation and coefficient of variation on the previously described 3 PoSaV-, 2 HuSaV-, 2 HuNoV- and 2 PoNoV-positive samples, in three different assays, three times in each assay.
The specificity of the test was determined by comparison of the melting temperature obtained on 7 HuNoV-, 6 feline calicivirus-, 6 BoNoV-, 1 HuSaV-, 6 PoNoV- and 3 BoKoV-positive samples with those obtained on GGIII, GGVII and GGVIII PoSaV-positive samples.
Relative sensitivity and specificity of the assay
Relative sensitivity and relative specificity of the SYBR Green real-time RT-PCR were tested first in a blind test, on 18 samples from Dutch swines (PC44 to PC63) where 6 had been previously diagnosed as PoSaV-positive, 7 PoNoV-positive and the five remaining samples as negative for both viruses.
Genomic quantification and detection limit
Genomic quantification was carried out by running samples of the standard curve with unknown samples. Positive samples from the previously tested 18 Dutch samples were also quantified. The detection limit was established by both conventional and real-time analysis on a serial ten-fold dilution of sample PC29 in RNAse free water. The same assay was also performed on serial dilutions of PC29 in PC16 (PoSaV-negative sample, ).
Assay application to field samples
The newly developed test was subsequently performed on 93 Dutch porcine samples with an unknown status for calicivirus infection. These samples were also tested in parallel using conventional RT-PCR (the gold standard) with the p289-290 primers.