Liver Transudate, a Potential Alternative to Detect Anti-Hepatitis E Virus Antibodies in Pigs and Wild Boars (Sus scrofa)

In recent years, cases of hepatitis E virus (HEV) infection have increased in Europe in association with the consumption of contaminated food, mainly from pork products but also from wild boars. The animal’s serum is usually tested for the presence of anti-HEV antibodies and viral RNA but, in many cases such as during hunting, an adequate serum sample cannot be obtained. In the present study, liver transudate was evaluated as an alternative matrix to serum for HEV detection. A total of 125 sera and liver transudates were tested by enzyme-linked immunosorbent assay at different dilutions (1:2, 1:10, 1:20), while 58 samples of serum and liver transudate were checked for the presence of HEV RNA by RT-qPCR. Anti- HEV antibodies were detected by ELISA in 68.0% of the serum samples, and in 61.6% of the undiluted transudate, and in 70.4%, 56.8%, and 44.8% of 1:2, 1:10, or 1:20 diluted transudate, respectively. The best results were obtained for the liver transudate at 1:10 dilution, based on the Kappa statistic (0.630) and intraclass correlation coefficient (0.841). HEV RNA was detected by RT-qPCR in 22.4% of the serum samples and 6.9% of the transudate samples, all samples used for RT-qPCR were positive by ELISA. Our results indicate that liver transudate may be an alternative matrix to serum for the detection of anti-HEV antibodies.


Introduction
Hepatitis E virus is the main cause of acute hepatitis worldwide, causing more than 20 million infections and more than 60,000 deaths per year [1]. HEV was first characterized in 1955 from an outbreak in New Delhi, India, when it was initially named non-A/non-B hepatitis. In 1983, the virus was named HEV to reflect its tendency to cause epidemics of enteric infections [2]. Hepatitis E infections in Europe have historically been caused by travelers returning from endemic countries, but in the last decade, autochthonous cases have greatly increased and have been associated with zoonotic transmission, mainly through the consumption of poorly cooked pork meat or liver-derived meat products [3][4][5].

Animal Samples
A total of 125 paired serum and liver transudate samples from 44 white pigs, 46 Iberian pigs, and 35 wild boars were included in the present study. Blood samples and liver samples (25 g) were collected at the different slaughterhouses in the case of pigs from the Spanish provinces of Cuenca, Pontevedra, Barcelona, Zaragoza, Burgos, Málaga, Córdoba, Girona, Toledo, Murcia, and Badajoz. Wild boar samples were recovered in the field after approximately 6 hours after the animal's death by hunting, in the province of Madrid, Spain. All samples were taken by authorized veterinary personnel and transported to the laboratory in refrigerated containers. Once in the laboratory, blood was preserved at 4 • C until the next day, when serum was obtained and stored at −40 • C for the ELISA and at −80 • C for the RT-qPCR. Liver samples were stored at −80 • C. In order to obtain the transudate, liver samples were frozen and thawed once at room temperature. The resulting liquid was collected between four and eight hours later and then stored at −40 • C for ELISA and at −80 • C for RT-qPCR. All animal experiments in this study were conducted according to Spanish regulations (RD 53/2013) and European regulations (EU Directive 2010/63/EU).

Detection of Anti-HEV Antibodies by ELISA
All sera and liver transudate samples were analyzed for the presence of anti-HEV antibodies using the commercial ELISA kit ID Screen Hepatitis E Indirect Multi-species (IDvet, Montepellier, France) according to the manufacturer's specifications, with a cut-off value set at 70 (S/P%). This is a duplicate-well test in which even-numbered wells are coated with a recombinant antigen from the capsid of HEV genotype 3, and odd-numbered wells are uncoated. Serum was diluted at 1:20 in the dilution buffer provided by the kit, while transudate was tested using the original undiluted sample as well as the dilutions 1:2, 1:10, and 1:20 in the dilution buffer. All plates were read at 450 nm with a Zenyth 3100 microplate multimode detector (Anthos Labtec Instruments GmbH, Salzburg, Austria).

Detection of HEV RNA by RT-qPCR
RT-qPCRs was performed on 58 paired serum and liver transudate samples from subsets of the animals (35 white pigs, 5 Iberian pigs, and 18 wild boars). These subsets were selected based on the amount of sample available. HEV RNA was obtained using the QIAmp Viral Mini Kit (QIAGEN, Hilden, Germany) according to the manufacturer´s instructions. The obtained product was stored at −80 • C until RT-qPCR was performed. HEV viral genome was detected using a previously described RT-qPCR assay [52] with minor modifications and a CFX96™ Real-Time Thermocycler System (Bio-Rad Laboratories, Munich, Germany). RT-qPCR was performed using the QuantiFast Pathogen RT-PCR + IC kit (QIAGEN) in a 20-µL reaction volume containing 10 µL of RNA. Reverse transcription was performed at 50 • C for 30 min, followed by denaturation at 95 • C for 5 min and 45 cycles of 95 • C for 10 s, 60 • C for 20 s, and 72 • C for 15 s. Samples showing PCR inhibition were diluted 1:10 in RNase free water.

Statistical Analysis
Animal type and the number of frozen-thaw cycles were tested in univariable logistic regression models using HEV positivity as the outcome variable. Risk factors that were significant in the univariable model with a generous p-value range < 0.2 were considered for inclusion in a multivariable model. The agreement between the qualitative results obtained in serum and liver transudate at different dilutions was measured using the Kappa statistic (κ). The corresponding optical densities (ODs) were compared using the intraclass correlation coefficient (ICC). All analyses were performed using IBM SPSS Statistics 25.0 (IBM, Armonk, NY, USA). When appropriate, results were reported together with their 95% confidence intervals (CIs). Sensitivity, specificity, and the area under the ROC curve (AUC) were calculated with the "pROC" [53] and "ROCR" [54] packages of R 3.3.3 [55].

Serological Results
A total of 85 of 125 serum samples analyzed by ELISA were positive, which indicated an overall seroprevalence of 68.0% (95%CI 59.3-75.5). The prevalence was 77.2% (95%CI 63.0-87.1%) in white pigs ( The prevalence was different among the different animal species (Table 1). The difference in HEV prevalence between transudate and serum assays ranged from 2 to 23.2 percentage points, depending on the dilution (Table 1). Specificity, sensitivity, positive predictive value (PPV), and negative predictive value (NPV) for the HEV ELISA are displayed in Table 2, together with the area under the curve (AUC) using confidence intervals of 95%, and a cut-off value of 70%. The highest overall specificity was obtained with the 1:20 transudate dilution (95.0%, 95% CI 87.5-100) followed by the 1:10 transudate dilution (90.0%, 95% CI 80.0-97.5). The highest sensitivity was obtained with the 1:2 transudate dilution (85.9%, 95% CI 77.6-95.9), followed by the undiluted transudate (82.3%, 95% CI 74.  In the univariate logistic regression analysis, no association was found between animal species and the results obtained in serum or transudate (p > 0.05). The 1:10 dilution of the liver transudate showed the best agreement with serum, based on the kappa coefficient (κ = 0.630) ( Table 1). ICCs for the comparison of ODs between serum and transudate dilutions were as follows: undiluted, 0.820; diluted 1:2, 0.803; diluted 1:10, 0.841; and diluted 1:20, 0.787 (Figure 1). The ROC curve, sensitivity, specificity, and the AUC for the undiluted and diluted liver transudates are shown in Table 2.
In the univariate logistic regression analysis, no association was found between animal species and the results obtained in serum or transudate (p > 0.05). The 1:10 dilution of the liver transudate showed the best agreement with serum, based on the kappa coefficient (κ = 0.630) ( Table 1). ICCs for the comparison of ODs between serum and transudate dilutions were as follows: undiluted, 0.820; diluted 1:2, 0.803; diluted 1:10, 0.841; and diluted 1:20, 0.787 (Figure 1). The ROC curve, sensitivity, specificity, and the AUC for the undiluted and diluted liver transudates are shown in Table 2.

Discussion
The present study is the first to demonstrate that liver transudate may be a good alternative matrix for the detection of anti-HEV antibodies in pigs and wild boars, since our results suggest a good agreement between the qualitative results and the optical densities of both matrices in the ELISA technique, especially with the 1:10 liver transudate dilution.
Although other alternatives to serum have been tested for the detection of anti-HEV antibodies, they are scarce, and the results have not always been compared with the serum results. A previous study detected anti-HEV antibodies in body cavity transudate obtained from raccoons, raccoon dogs, cats, and dogs [47]. Meat juice has also been used in some studies to determine HEV seroprevalence in hunted wild boars and slaughtered pigs with satisfactory results [48,56]. Similarly, other reports also demonstrated the utility of meat juice to determine the prevalence of different agents such as Salmonella, Trichinella, influenza virus or Toxoplasma [57,58]. Meat juice, like other body fluids, contains fewer antibodies against HEV, Toxoplasma gondii, and other pathogens than serum [56][57][58][59], so the dilution factor in serological assays should be adjusted accordingly. Therefore, liver transudate was tested undiluted and at different dilutions to determine this factor.
In this study, anti-HEV antibodies were successfully detected in the liver transudate of pigs and wild boars, and their detection rate showed an elevated agreement with their paired serum samples. The specificity and sensitivity obtained with liver transudate were high, which indicates that it could be a good alternative to serum, which is in concordance with other studies made in another matrix [56]. Both serum and liver transudates at all dilutions presented a range of very similar OD values. In general, seroprevalence in serum was slightly higher than the one obtained with the liver transudate, which could be explained by the presence of inhibitors of the ELISA in the latter type of samples. Meat or liver exudates are tricky analytical templates and, as other many food items, they contain components, such as fat or haemo group, that can produce a matrix effect in the ELISA reaction [59,60], interfering with the ELISA by inhibiting antigen-antibody binding, reacting with epitopes, or having interfering enzymatic activity [61]. These results required optimizing the dilution factor, as reported for other matrices such as meat juice [62]; we found 1:10 to be the best dilution for detecting anti-HEV antibodies using a commercial indirect ELISA test. Although this dilution did not always give the best individual values of specificity, sensitivity, PPV, or NPV, it gave overall the best results. In addition, the 1:10 transudate dilution showed the best concordance of serum results with qualitative and optical densities in ELISA results.
Only 13 serum and four liver transudate samples out of 58 were RT-qPCR positive, and only three samples were RT-qPCR-positive for both matrices from the same animal. The presence of antibodies in serum and liver transudate was substantially higher than the presence of the virus detected by RT-qPCR in both matrices, in agreement with a previous study that found high seroprevalence 73.3% but the prevalence of only 15.6% in liver respectively based on PCR [52]. In the present study, all RT-qPCR-positive samples in both matrices were also positive by ELISA. Most serum samples (43/45) and all transudates that were negative by RT-qPCR were positive by ELISA. This may be because viremia lasts for a very short time in circulating blood, which makes it quite difficult to detect viral RNA in serum [51] while circulating antibodies persist.

Conclusions
The obtained results suggest that liver transudate, used at an appropriate dilution (1:10), may be a good alternative matrix to serum for detecting anti-HEV antibodies by ELISA in pigs and wild boars. This matrix may be useful in postmortem analyses for diagnosis, surveillance, or research in field studies, particularly when obtaining blood is impractical. Although HEV could be detected by RT-qPCR in the liver transudate, further studies with more samples are needed to obtain any conclusion.