Abattoirs selection and cross-sectional study
Piedmont is located in Northwestern Italy and is the third Italian region important by the number of swine farms (11% of Italian production) . Italian pig production is different from the other European countries due to its long production cycle. Weaners between 8 and 25 kg, growers between 25 and 70 kg and finishers, between 70 and 110 kg are slaughtered for fresh meat consumption (from 1 to 8 months). The adult pigs, that produce Protected Designation of Origin ham (Prosciutto di Parma, Prosciutto di Cuneo), are usually slaughtered at nine months with an average weight of 160 kg.
The present investigation was designed as a prevalence cross-sectional study. Three different abattoirs located in the Piedmont region, which slaughter only pigs, with a body weight ranging from 13 to 110 kg, were selected for the study. 26 animals were sampled from the first abattoir, 63 from the second abattoir and 67 from the third abattoir. Therefore, a total of 156 pigs, belonging to 20 different farms, ranging from 1 to 8 months of age, were analysed.
The sample size was determined considering a 95% confidence interval, an expected prevalence of 50%, and an error of 8%. No correction for the finite population was used as the population target consist of 248,076 animals, within the study area . The number of animals to analyse, for each abattoir, was determined on the different numbers of weekly-slaughtered pigs.
The blood, the liver, the bile, and the faeces were collected from the target population of the three abattoirs. The abattoirs workers were located at different points along the slaughter chain, to collect the five biological matrices (blood, liver, bile, cecum and muscle) analysed for each sampled animal. Blood was collected during the animal's bleeding using special test tubes (with and without anticoagulant) previously marked with the animal's identification number and then stored at 4 °C. A lobe of the liver and the entire gallbladder, together with the cecum, were collected during the evisceration phase. From the cecum, the faeces were subsequently sampled. Finally, a portion of neck muscles was removed from the carcass in the phase before refrigeration. All the biological matrices collected were then placed in a sterile disposable plastic bag. The bile of two animals was not available and the blood samples were collected in EDTA tubes and stored at 4 °C. The muscle samples were collected to investigate the co-presence of HEV in animals that previously tested positive in the other biological matrices. The traceability of the biological samples was ensured during the entire slaughter chain and all the biological samples were stored at -80 °C after the collection in the abattoirs.
RNA extraction and real-time RT-PCR conditions
The 10% weight/volume suspensions were prepared in PBS for liver, bile and faeces, whereas the blood samples were analysed undiluted. Then 350 µl of the suspensions were used for the RNA extraction thanks to the TRI Reagent® (Sigma-Aldrich, Darmstadt, Germany) and the QIAamp® Viral RNA Mini Kit (Qiagen, Hilden, Germany) following the manufacturer’s protocol. Instead, the RNA extraction from the muscle samples was performed according to Chelli et al. .
The detection of HEV in the liver, faeces, bile, and blood was performed by real-time RT-PCR with the CFX96 Touch™ Real-Time PCR Detection System (Biorad, Hercules, USA). Detection employed the TaqMan approach, targeting the ORF3 genomic region. A real-time RT-PCR protocol was adopted from Jothikumar et al.  with primers and TaqMan probe manufactured by Thermo Fisher Scientific (Waltham, USA). The amplification of HEV extracted from the muscles was performed according to Chelli et al. . In detail, the reaction volume of 25 μl consisted of 5 μl of template, 5 µl of RNA Ultrasense™ reaction mix (Invitrogen, Waltham, USA), 500 nM of primer Forward JVHEVF (5′-GGTGGTTTCTGGGGTGAC-3′), 900 nM of primer Reverse JVHEVR (5′-AGGGGTTGGTTGGATGAA -3′), 250 nM of Taqman Probe JVHEVP-MGB (5′-FAM-TGATTCTCAGCCCTTCGC-MGB-3′), 1,25 µl of RNA Ultrasense enzyme mix (Invitrogen, Waltham, USA) and nuclease-free water up to the final volume. The assay was carried out with the CFX96 Touch™ Real-Time PCR Detection System (Biorad, Hercules, USA), using the following PCR cycling conditions: 1 cycle of reverse transcription at 50 °C for 1 h, 1 cycle of PCR initial activation step at 95 °C for 5 min followed by 45 cycles of 95 °C for 15 s, 60 °C for 1 min and 65 °C for 1 min.
When a pig is HEV positive for more than one biological matrix, only that with the lowest Ct, determined through real-time RT-PCR, is used for reverse transcription to cDNA and phylogenetic analysis. Reverse transcription was carried out using One Script cDNA Synthesis kit (ABM, Richmond, Canada). The reaction was performed in a total volume of 20 μl, containing 4 μl RT buffer (5 ×), 1 μl dNTPs (10 µM), 1 μl Random primers, 1 μl One Script RTase (200 U/μl), 0.5 μl Rnase Off Ribon. Inhibitor (40 U/μl), and 7 μl RNA. The thermal profile was performed as follows: 10 min at 25 °C, 50 min at 42 °C, and 5 min at 85 °C.
The phylogenetic analysis was performed comparing the 5’ region of the ORF2 gene, targeting a variable region that provides a phylogenetic signal comparable to full genome analysis . A Nested PCR assay was performed using an external primer set (3156N forward, 3157 reverse) for a first amplification round (710 bp) and an internal primer set (3158N forward, 3159 reverse) for a second ones (348 bp) . The assay was carried out in a final volume of 50 μl with 1X Maxima Hot Start Taq buffer, 1.5 mM MgCl2, 0.2 mM of each dNTP, 0.2 μM of each primer, and 1U of Maxima Hot Start Taq DNA Polymerase (Thermo Fisher Scientific, Waltham, USA). The following thermal conditions were applied: activation of Taq polymerase at 95 °C for 4 min, followed by 40 cycles of denaturation at 95 °C for 1 min; annealing at 60 °C for 1 min; extension at 72 °C for 2 min; and final elongation at 72 °C for 7 min. Amplification products were checked by electrophoresis on 2% agarose gel, purified with the Extract me DNA Clean-up (Blirt, Gdańsk, Poland) and sequenced using a BrilliantDye Terminator v3.1 cycle sequencing kit (NimaGen, Nimega, Netherlands). The amplicons were purified with the DyeEx 2.0 Spin Kit (Qiagen, Hilden, Germany) and run on a 3130xl Genetic Analyzer (Life Technologies, Carlsbad, USA). A dataset of partial sequences of 135 bp of the 5’ ORF2 region from the analyzed samples and 151 GeneBank accessions have been aligned. The phylogenetic analysis was performed using MEGA7 software with the Neighbour-Joining method and the Kimura-2 model. Statistical robustness and reliability of the branching order were confirmed by bootstrap analysis using 1,000 reiterations .
The blood samples were also collected to extract the serum through centrifugation at 3500 g for 10 min and then stored at -20 °C until the serological analysis. A commercial ELISA kit, specific for the detection of HEV-IgG in animal serum samples, was employed according to the manufacturer’s instructions (ID screen®, Grabels, France). The plate absorbance was read at a wavelength of 450 nm and the Net OD values of the samples were calculated and expressed as percentages of reactivity (pOD) of the plate positive control. The samples with percentages of reactivity (pOD) higher than 70% were classified as positive, samples with pOD between 60 and 70% as doubtful and samples with pOD lower than 60% were considered as negative. Serums that resulted as doubtful were tested in a second test session.
No commercial ELISA kit was available on the market exclusively to detect HEV-IgM in swine serums, while the tests were being carried out. Therefore, the HEV total antibody ELISA kit (Wantai Biopharm, Beijing, China), which can detect both HEV-IgG and HEV-IgM in animal sera, was used for the 114 serums tested negative for HEV-IgG, according to the manufacturer instructions. The plate was read at the wavelength of 450 nm and the absorbance of the blank well was subtracted from the absorbance values of specimens and controls. The cut-off calculation was performed by adding the value 0.12 to the average absorbance for three negative plate controls. Samples with a ratio ≥ 1 (sample OD/cut-off) were considered positive, samples with a ratio between 0.9 and 1.1 were considered doubtful and samples with a ratio lower than 0.9 were classified negative. Serums that resulted as doubtful were tested in a second test session.
The proportion of positives was calculated for each matrix. The binomial distribution was used to calculate the exact confidence limit of each proportion. The correlation index and Cohen’s Kappa index were used to assess the agreement between the biological matrices . The Kappa index indicates the proportion of agreement, excluding that expected by chance, for categorical variables. Kappa values near 1 indicate perfect agreement, while a Kappa value of 0 indicates that all the agreement is due to chance. According to Landis et al.  the value of Kappa index between 0.60 and 0.8 was considered good agreement.
The Phi index indicated correlations between matrices, i.e., a measure of association for two binary variables. Values of Phi range from − 1 to + 1, where 1 indicates perfect agreement, -1 perfect disagreement, and 0 indicates no relationship . All statistical analyses were performed by SAS System v 9.4.
To evaluate the agreement between all five matrices, Fleiss' Kappa, as an index of interrater agreement between matrices was calculated , using package irr (version 0.84. 1) .