Porcine blood used as ingredient in meat productions may serve as a vehicle for hepatitis E virus transmission

Accepted Manuscript Porcine blood used as ingredient in meat productions may serve as a vehicle for hepatitis E virus transmission

Ingeborg L.A. Boxman, Claudia C.C. Jansen, Geke Hägele, Ans Zwartkruis-Nahuis, Jeroen Cremer, Harry Vennema, Aloys S.L. Tijsma PII: DOI: Reference:

S0168-1605(17)30297-0 doi: 10.1016/j.ijfoodmicro.2017.06.029 FOOD 7626

To appear in:

International Journal of Food Microbiology

Received date: Revised date: Accepted date:

28 March 2017 6 June 2017 26 June 2017

Please cite this article as: Ingeborg L.A. Boxman, Claudia C.C. Jansen, Geke Hägele, Ans Zwartkruis-Nahuis, Jeroen Cremer, Harry Vennema, Aloys S.L. Tijsma , Porcine blood used as ingredient in meat productions may serve as a vehicle for hepatitis E virus transmission, International Journal of Food Microbiology (2016), doi: 10.1016/ j.ijfoodmicro.2017.06.029

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Porcine blood used as ingredient in meat productions may serve as a vehicle for hepatitis E virus transmission Ingeborg L.A. Boxmana#, Claudia C.C. Jansena, Geke Hägelea, Ans Zwartkruis-Nahuisa, Jeroen

Food and Consumer Product Safety Authority (NVWA), Akkermaalsbos 4, 6708 WB, Wageningen,

IP

a

T

Cremerb, Harry Vennemab, Aloys S.L. Tijsmacd

CR

the Netherlands; [email protected]; [email protected]; [email protected]; [email protected];

National Institute of Public Health and the Environment (RIVM), Anthonie van Leeuwenhoeklaan 9,

US

b

c

AN

3721 MA, Bilthoven, the Netherlands; [email protected]; [email protected] Food and Consumer Product Safety Authority (NVWA), Catharijnesingel 59 3511 GG, Utrecht

ED

M

Utrecht, the Netherlands; [email protected]

#

CE

PT

Running title: HEV RNA in porcine blood used in food

Corresponding author:

AC

Ingeborg Boxman

[email protected]; tel. +31 (0)88 223 0447 d

Current address of Aloys Tijsma is Viroclinics Biosciences B.V., Marconistraat 16, 3029 AK

Rotterdam, The Netherlands

1

ACCEPTED MANUSCRIPT Abstract The aim of the present study was to investigate whether the use of porcine blood(products) in food could be a risk for a hepatitis E virus (HEV) infection. HEV RNA was detected in 33/36 batches of (non-heated) liquid products and in 7/24 spray dried powder products. Contamination levels varied

T

among the products, but were highest in liquid whole blood, plasma and fibrinogen reaching levels of

IP

2.2 x 102 to 2.8 x 102 HEV genome copies per 0.2 gram. Sequence analyses revealed genotype 3 strains,

CR

of which two were 100% (493 nt) identical to recently diagnosed HEV cases, although no direct epidemiological link was established. The industry provided information on processing of blood

US

products in (ready-to-eat)-meat. From this, it was concluded that blood products as an ingredient of

AN

processed meat may not be sufficiently heated prior to consumption, and therefore could be a vehicle

M

for transmission.

AC

CE

PT

ED

Keywords: HEV, meat products, blood, food, plasma, fibrinogen

2

ACCEPTED MANUSCRIPT 1.

Introduction

The incidence of hepatitis E virus (HEV) infection in the Netherlands has increased recently and is high as compared to other European countries (Adlhoch et al., 2016). In 2013 and 2014, HEV RNA was detected in 1 per 762 blood donations (Hogema et al., 2016) and the HEV strains detected belong to

T

genotype 3 (gt3). HEV gt3 is abundantly present in domestic pigs, which indicates that these animals

IP

probably are a major zoonotic HEV reservoir. The main transmission route(s) still need to be

CR

determined, but consumption of raw or undercooked porcine meat might be of great importance. Domestic pigs and wild boars show high infection rates in Europe with viral sequences that are closely

US

related to those detected in human hepatitis E patients (van der Poel et al., 2001, 2014). Apart from a

AN

high seroprevalence of HEV antibodies in pigs, also viremia has been demonstrated in pigs at slaughter age (Grierson et al., 2015; Rutjes et al., 2014). It has long been considered plausible that the persistence

M

of viremia in pigs up to the time of slaughter could provide a potential vehicle for zoonotic transmission

ED

to humans in relation to meat products. Food-borne transmission of HEV via consumption of raw and undercooked liver, meat, or sausages from domestic pigs has been documented in several studies and

PT

HEV RNA has been detected in porcine liver, pork and pork products by several groups as recently

CE

been reviewed (Pavio et al., 2015). The presence of infectious HEV was demonstrated in pork liver sausage and livers (Berto et al., 2013; Feagins et al., 2007).

AC

Blood is a rich source of iron and proteins of high nutritional value and functional quality. The maximal utilization of animal blood, coupled with recent advances in blood collection and processing techniques, have led to a myriad of blood protein ingredients becoming available for use in human foods and dietary supplements (Ofari and Hsieh, 2012). Blood proteins are used as ingredients in meat industry, mainly as a binder of water and fat, but also as natural color enhancers and emulsifiers. Products like fibrinogen are used as a fresh meat cold-set binder to produce restructured fresh meat, whereas spray-dried plasma

3

ACCEPTED MANUSCRIPT powder is used as a hot-set binder due to its ability to form gels upon heating to bind water and fat from meat. Blood taken from a healthy animal is essentially sterile, and both manufacturers and processors have instituted measures concerning bacteria to guarantee the safety of these blood proteins to be used in

T

food industry. Porcine blood has been evaluated for its microbial quality (Ramos-Clamont et al., 2003)

IP

and the use of fibrinogen and thrombin in food has been evaluated (EFSA, 2005, 2015), however, viral

CR

zoonotic threats, like hepatitis E virus (HEV), might have been overlooked. The aim of the present study was to investigate whether the use of porcine blood (products) in food can

US

be an additional risk. Unlike the situation for HEV gt1, pregnant women are not considered a risk group

AN

for HEV gt3 infection with serious consequences. Older men with no international travel history are considered the most at risk for clinically overt HEV gt3 infection. In addition, HEV gt 3 infections are

M

increasingly recognized to cause persistent hepatitis E infections in immunocompromised patients, with

ED

an increased risk of progression to cirrhosis (Clemente-Casares et al. 2016; Tedder et al., 2017). For this, blood products were analyzed by quantitative RT-qPCR for the presence of HEV RNA.

PT

Subsequently, the obtained data were combined with an inventory of applications and processing

CE

characteristics of blood products in meat industry to provide data for risk assessment. In addition,

2.

AC

sequences detected in blood products were compared to those recently identified in Dutch patients.

Materials and Methods

2.1. Virus preparations A porcine fecal sample was selected as source for virus preparation after testing positive for HEV RNA in RT-qPCR and being typed as genotype 3c (Acc No. MF185108) using methods as described in this paper. The fecal sample was suspended in 10 mM phosphate buffer saline, pH 7.4, to obtain a final of

4

ACCEPTED MANUSCRIPT 50% suspension (w/v), then vortexed and centrifuged at 4,000 g for 20 minutes at 4ºC. The supernatant was added to two volumes of 30% (w/v) polyethylene glycol and 0.9 M NaCl, mixed, and stored at 4ºC for overnight virus precipitation. After centrifugation (4,000 g for 20 minutes at 4ºC), the obtained pellet was resuspended PBS, aliquoted and stored in -80 ºC. The titer of the virus stock was estimated

T

by RT-qPCR using serial-dilutions of the WHO HEV genotype 3a standard (PEI code 6329/10) at the

IP

ViroScience group of the ErasmusMC Rotterdam (Dr. S. Pas)(Pas et al., 2012). Based on this method

CR

the titer of the HEV virus was estimated at 1.2 x 107 IU/ml. Murine norovirus (MuNoV) was kindly provided by Dr. H. Virgin IV, Washington University, St. Louis, Missouri and estimated by cell culture

AN

US

method at 4x107 /ml TCID50 by Dr. E. Duizer, RIVM, the Netherlands (Tuladhar et al., 2012).

2.2. Sampling

M

A selection was made for porcine blood products intended to be used in food, including both liquid and

ED

spray dried powder products. Sampling occurred on three separate points in time, October 2015, April 2016 and August 2016. Six batches of whole blood and stabilized hemoglobin and eight batches of

PT

hemoglobin, plasma, and fibrinogen were sampled, as well as eight batches of plasma, hemoglobin and

CE

stabilized hemoglobin spray dried powder products. The size of the batches varied between 25 kg and 1000 kg. One sample of about 500 grams was taken from each batch. Frozen whole blood, frozen

AC

fibrinogen and plasma powder of bovine origin were collected and served as control material for the inhouse-validation study after they had tested negative for the presence of HEV RNA in the RT-qPCR.

2.3. Viral and RNA extraction 2.3.1. Liquid blood products

5

ACCEPTED MANUSCRIPT For virus extraction from liquid blood products, 0.2 gram blood product was mixed with 1.8 ml TGBE buffer (100 mM Tris, 50 mM Glycine, 1% (w/v) beef extract, pH 9.5 buffer) and 10 µl of MuNoV (4x104 TCID50), and incubated at room temperature for 20 min. Subsequently, the mixture was clarified by centrifugation (10,000 g for 20 min at 4 ºC) and the supernatant was transferred into a new tube. For

T

RNA extraction, 2 ml of Nuclisense lysis buffer (BioMérieux) was added and mixed with the

IP

supernatant by rotation for 10 min at room temperature. Subsequently, 50 µl of magnetic silica

CR

(Nuclisense Magnetic Extraction Reagents kit, BioMérieux) was added to the buffer and the buffer was mixed well by vortexing briefly. After an incubation period of 10 min at room temperature, the total

US

mixture was used as input for the extraction of nucleic acids using reagents from the Nuclisense

AN

Magnetic Extraction Reagents kit (BioMérieux) according to the manufacturer’s instruction. RNA was

M

tested directly or stored frozen at -80 ºC until testing.

ED

2.3.2. Blood product powders

Virus extraction from hemoglobin containing powders was performed in the same way as described for

PT

the liquid blood products, except that 0.2 gram of powder was pre-wetted with 300 µl ethanol (100%)

CE

prior to mixing with 1.8 ml TGBE buffer. Virus extraction from plasma powder was performed as described above, except that 0.2 gram of powder was pre-wetted with 300 µl ethanol (100%) prior to

AC

mixing with 1.8 ml Nuclisense lysis buffer (BioMérieux). All solubilized powders were clarified by centrifugation (10,000 g for 20 min at 4ºC) and supernatants were added to 2 ml of Nuclisense lysis buffer (BioMérieux) for RNA extraction. RNA was tested directly or stored frozen at -80 ºC until testing.

6

ACCEPTED MANUSCRIPT 2.4. Detection of MuNoV and HEV RNA All RT-qPCR reactions were performed in a CFX96 real time PCR detection system (BioRad). RNA of MuNoV and HEV was detected by RT-qPCRs using oligonucleotides as described previously (Baert et al., 2008; Jothikumar et al., 2006) after in-house optimization for the CFX96 platform. The RNA

T

Ultrasense one-step qRT-PCR system kit (ThermoFisher) was used with 5 µl of nucleic acid preparation

IP

in a total reaction volume of 25 µl.

CR

Each series of virus extractions consisted of a negative extraction control sample in between each set of three samples that was run through all stages of the analytical process. Water controls and positive

US

target RNA template controls were included in each PCR run. Each sample was also tested for

AN

inhibition of the reaction in separate reaction well using a HEV reaction mix with the ssRNA HEV-

ED

the control (Diez-Valcarce et al., 2011).

M

standard used as external control (Diez-Valcarce et al., 2011) and using the PrfAP probe for detection of

2.5. Calculation for process control extraction efficiency and efficiency of RT-PCR detection

PT

An equal amount of MuNoV as spiked to each sample (4 x 104 TCID50) was also subjected to RNA

CE

extraction as a 100% control. Viral extraction efficiency for each sample was calculated using the process control virus RNA standard curve (ISO 15216-1, 2017), setting the minimal recovery to be ≥

AC

1%. Inhibition of RT-PCR detection was determined by subtracting the Cq value for the external control (EC) ssRNA HEV standard when added to water from the Cq value for the EC- ssRNA HEV standard when added to a RNA sample, allowing the result of the subtraction to be maximally 2.

2.6. Quantification of HEV RNA

7

ACCEPTED MANUSCRIPT To develop the dsDNA standard a synthetic oligo was ordered at Metabion, Germany, stretching from 5261 to 5330 based on GenBank accession no. M73218, having a BamHI restriction site added at the 3’of the probe binding site. This oligo was amplified with JHEVF and JHEVR primers (Jothikumar et al., 2006), subsequently ligated into pGEM-T-easy. After Midiprep, DNA was linearized using ApaI,

T

quantified by A260 measurement, diluted in TE and stored in single use aliquots. Each RNA sample

IP

was run against serial dilution of dsDNA HEV standard (101 to 105 genome copies/µl). Results were

CR

expressed as genome copies/0.2 gram. A theoretical limit of quantitation of 100 genome copies/0.2

AN

2.7. Determination of limit of detection (LOD)

US

gram was applied corresponding to 5 genome copies per reaction.

Bovine whole blood, fibrinogen and plasma powder were artificially inoculated with 10-fold serial

M

dilutions of HEV and MuNoV. For each inoculation level, four extractions were performed, resulting in

ED

four cycle threshold (Cq) values. Recovery (%) per sample was calculated using 10(ΔCq/m) x 100%, in which ΔCq = Cq (RNA inoculated sample) – Cq (RNA of spike dilution) and m= slope of the RT-qPCR

CE

PT

standard curve obtained from ten-fold dilution of the HEV or MuNoV virus stock.

2.8. Typing of HEV RNA positive samples and comparison to human HEV strains

AC

HEV presumptive positive samples, as judged by the presence of a typical S-curve in the amplification plot, were re-amplified using a nested RT-PCR using codehop primers targeting ORF3 with a final 493 nucleotide fragment for sequence analyses (Table 1). In brief, cDNA synthesis was performed for 60 min at 42 ºC and ended at 95 ºC for 5 min. The Superscript III reverse transcriptase (10 U/ul, Thermo Fisher Scientific) was used with 5 µl of nucleic acid preparation in a total cDNA reaction volume of 10 µl. The reaction mixture further consisted of 1X RT buffer (Thermo Fisher Scientific), dNTPs (1 mM

8

ACCEPTED MANUSCRIPT each, Roche); reverse primer HEV-orf2-ro-ch (5 µM), DTT (5 mM, Thermo Fisher Scientific), and RNaseOUT (2 U/µl, Thermo Fisher Scientific). The first RT-PCR program was 35 cycles of 30 s at 95 ºC, 30 s at 42 ºC (with ramp rate of 0.4 ºC/sec) and 45 s at 60 ºC, and was run on a CFX96 platform (BioRad). The reaction mix with a total volume of

T

50 µl consisted of 10 µl of cDNA reaction, 1x Taq PCR buffer with MgCl2, pH 8.3 (Roche), dNTP mix

IP

(0.2 mM, Roche), forward prime HEV-orf2-fo-ch (1 µM), Taq DNA polymerase (0.05 U/µl) (Roche).

CR

After 6 minutes of denaturation at 95 ºC, the nested PCR program was 40 cycles of 30 s at 95 ºC, 20 s at 60 ºC and 15 s at 72 ºC, and was run on a CFX96 platform (BioRad). The reaction mix with a total

US

volume of 50 µl consisted of 1 µl of the First PCR template, 1x FastStart PCR buffer with MgCl2

AN

(Roche), dNTP mix (0.2 mM, Roche), forward primer HEV-orf2-fi-ch and reverse primer HEV-orf2-rich (0.4 µM both), FastStart DNA polymerase Sigma (0.05 U/µl) (Roche).

M

Nested products obtained were purified using Qiaquick PCR Purification Kit (Qiagen), processed with

ED

the BigDye Terminator v1.1 Cycle Sequencing Kit (Thermo Fisher Scientific) with sequence primers HEV-orf2-fs or HEV-orf2-rs (Table 1) and subsequently processed with DyeEx 2.0 Spin Kit (Qiagen),

PT

all according manufacturer’s instructions and subsequently sequenced on an ABI 3500 Genetic

CE

Analyzer. Genotypes were assigned on the basis of their similarity to reference strains representing known genotypes and variants (Smith et al., 2014).

AC

Phylogenetic analyses on 493 nt was performed on HEV sequences derived from blood products and results were subsequently compared with human HEV strains as detected in Dutch cases with sampling dates between January 1st, 2015 and December 31st 2016. A Maximum Parsimony tree was calculated with Bionumerics software. Sequences have been submitted to GenBank with accession numbers KY781989-KY782010 and KY774987-KY775031.

9

ACCEPTED MANUSCRIPT 3.

Results

3.1. Preliminary experiment outcomes In preliminary experiments different methods were compared for their ability to extract HEV, as the target virus, and MuNoV, as the process control virus, from spiked blood products. For both HEV and

T

MuNoV the highest extraction efficiency was obtained when 0.2 gram of blood products was mixed

IP

with 1.8 ml TGBE buffer and subsequent nucleic acids were extracted. For the blood powders, pre-

CR

wetting with 300 µl ethanol (100%) improved extraction efficiency. For plasma products, both in liquid and powder form, the extraction efficiency was highest when 1.8 ml lysis buffer was added to 0.2 grams

AN

US

of product (data not shown).

3.2. Limit of detection of HEV RNA in blood products

M

As preliminary experiments had shown that HEV RNA was present in porcine blood products, bovine

ED

blood products were tested for HEV RNA to determine whether they could be used for validation instead. None of the batches of bovine origin, whole blood, fibrinogen, and plasma powder used in the

PT

validation study tested positive for HEV RNA. The recovery rates for HEV and MuNoV obtained in

CE

artificially contaminated bovine blood samples ranged from 28 to 81% and 27 to 125%, respectively (Table 2). The limit of detection (LOD), defined as the lowest amount of HEV detected in four

AC

extractions, was 9.3 x 101 IU HEV per 0.2 g for whole blood and plasma powder and 1.2 x 102 IU HEV per 0.2 g in fibrinogen. Additional experiments demonstrated detection of HEV RNA did not improve after dilution of RNA extracts, indicating no or only low levels of inhibitory substances were present as we also concluded from data obtained with the external amplification control (ssRNA EC-HEV standard) (data not shown).

10

ACCEPTED MANUSCRIPT 3.3. Monitoring porcine blood products for the presence of HEV RNA In total 60 batches of porcine blood products were tested for the presence of HEV RNA (Table 3). Samples were considered positive if amplification plots of the RT-qPCR signals showed an S curve and had a threshold cycle (Cq) value below 40. HEV RNA was detected in all samples of whole blood

T

(n=6), liquid plasma, hemoglobin, fibrinogen (n=8 each) and in three out of six samples of stabilized

IP

hemoglobin product (Table 3). Liquid product batches tested positive in both RT-qPCR duplicates, in

CR

contrast to most of the stabilized hemoglobin products, which tested positive only in one of the two duplicates. When the HEV dsDNA standard became available, all positive RNA samples, stored at -80

US

ºC for one-to nine months, were re-run for quantification. The quantitative results showed differences in

AN

the level of HEV contamination between products and batches, revealing the highest average copy number in whole blood, plasma and fibrinogen samples (Table 3). Some of the weakly positive samples

M

had become negative during storage of the RNA until the HEV dsDNA standard became available.

ED

Similarly, the spray-dried products were analyzed, showing HEV RNA in four out of eight plasma powder samples and in three out of eight hemoglobin-based powder product samples but only at low

PT

contamination levels, too low to quantify, and often not in both duplicates. None of the eight samples of

CE

pure hemoglobin powder samples tested positive.

AC

3.4. Typing of HEV RNA detected in blood products HEV positive RNA extracts were subjected to nested RT-PCR targeting another part of the HEV genome for sequence analyses. Sequences of an open reading frame 3 fragment (493 nt, positions 59416472 of reference sequence NC_001434) were obtained for 27/32 samples, all being liquid products, all with Cq levels below 40.5 (Table 3), and all identified as HEV genotype 3c. Although the batches were composed of thousands of animals likely originating from different farms, reads derived from single

11

ACCEPTED MANUSCRIPT strains were obtained for 24 of 27 samples, of which two were incomplete. Clearly mixed sequences were detected in three fibrinogen samples. Sequences of the 22 samples with a single strain were aligned with reference strains as well as selected representative genotype 3 c strains with close similarity obtained in diagnostic samples from patients in 2015 and 2016 (Figure 1). Sequences detected

T

in porcine blood were closely related, and in two cases identical over 493 nt, to humans strains. One of

IP

the 100% matches was detected in a fibrinogen batch that was produced in June 2016, whereas the

CR

patient had an onset of disease in November 2015. The other strain was detected in stabilized hemoglobin, produced in January 2016, whereas the patient had an onset of disease in June 2015. For

AN

US

some samples sequence analyses was not successful due to a weak signal.

3.5. Inventory of industrial processes

M

Inquiries were made at the blood product production plant to gain insight into production processes.

ED

From this it was concluded that none of the liquid blood products are subjected to any thermal treatment or any other treatment that may partly or sufficiently inactivate HEV. Spray dried powders were said to

PT

be heated at least 80 ºC for 30-90 s.

CE

Subsequently, meat processing plants, identified as recipients of porcine blood products, were requested to provide information on the types of (ready-to-eat) food in which blood products are being used. If

AC

applicable, cooking times and core-temperatures were inventoried (Table 4). The food products identified were, among others, pâté, liver sausages, Cordon Bleu’s, Schnitzels, bacon and smoked raw ham. Cooking times and core temperatures varied between producers and between meat products, of which the least stringent treatment is presented in table 4. Some meat products identified underwent heat treatment with core temperatures below 71 ºC. Other products had core temperatures above the 71 ºC, but for less than 20 minutes, a heat-treatment described to be required for complete HEV

12

ACCEPTED MANUSCRIPT inactivation in a pate-like product (Barnaud et al., 2012). Only a part of the products met this treatment. Besides ready-to-eat products, meat preparations were identified that require to be cooked by either consumer or catering services, for which cooking times and core-temperatures may vary and may be poorly controlled (Table 4). Liquid hemoglobine appeared not to be used in meat products, but to be

Discussion

CR

4.

IP

T

used in vitamin additives.

To our knowledge, this is the first time that porcine blood products used as ingredients in food have

US

been tested for the presence of HEV RNA. The study shows that HEV RNA can be detected, quantified

AN

and sequenced from the majority of the tested products. None of the identified liquid blood products appears to be heat treated prior to its processing into food. Inactivation of HEV present in these products

M

therefore depends on heat-treatment either performed during industrial processing or by the end-user

ED

e.g. consumer or catering industry. Some of the industrial heat-treatments may be inadequate to (completely) inactivate blood-derived HEV, however, this may depend on initial HEV contamination

PT

levels and the constituents of the matrix. The findings presented in this study are important as they may

CE

describe a new risk for HEV infection especially for vulnerable individuals such as immunosuppressed

AC

patients, unaware of the presence of porcine blood in meat products.

Porcine blood intended to be used in food is currently considered to be safe for consumption as it is collected at slaughter houses only from pigs that are approved for consumption by the Competent Authority. This consideration lacks however assessment for the risk of zoonotic transmission of HEV; as pigs or their blood are currently not being tested HEV RNA routinely. Batches collected at slaughter houses are composed of blood from over thousands of pigs, pooled from different farms. Therefore,

13

ACCEPTED MANUSCRIPT blood of a limited number of HEV viraemic pigs (Grierson et al., 2015) may contaminate the whole batch. This may explain the high percentages of batches that tested positive for the presence of HEV RNA and the finding of unique sequences in a large proportion of the samples.

T

In the present study, spray dried powders and stabilized hemoglobin showed low contamination levels.

IP

Recently, spray-dried porcine plasma used as animal feed has also been tested HEV RNA positive

CR

(22.4%) in another study (Pujols et al., 2014), however without any indication of seroconversion in HEV-free pigs. Therefore, the process of spray drying by heating to a temperature in excess of 80°C

US

may possibly be sufficient to ensure HEV inactivation (Pujols et al., 2014). The average contamination

AN

levels for whole blood, frozen plasma, and fibrinogen samples were at a level of a few hundred copies per 0.2 gram. Given the percentage (4-15%) in which the highest contaminated blood products are being

M

applied in meat, 100 gram meat end-product could contain 103 to 104 copies HEV (Table 4). Applying

ED

HEV positive blood products to muscle tissue, with a low expected HEV positivity (Berto et al., 2012; Di Bartolo et al., 2012), will therefore render the final meat product HEV contaminated, possibly

PT

contributing to the spread of HEV in the food chain. At present, it is unknown whether the HEV RNA

CE

detected by RT-qPCR is derived from infectious HEV particles warranting further investigation, but this is most likely the case, as the liquid blood products lack any heat treatment or any other treatment that

AC

might reduce HEV infectivity. The number of infectious particles in the end product will depend on the production process of the meat preparations and thus on cooking temperatures and time used by the industrial producer (ready-to-eat products) or consumer (Table 4). Thermal inactivation of HEV has been studied in various experimental settings (FSA, 2014). In a recently developed cell culture-based titration system using the cell culture-adapted genotype 3 strain 47832c (Johne et al., 2016), the thermal stability of HEV gt3 viral suspensions was shown to rapidly decrease at temperature > 65 ºC, however

14

ACCEPTED MANUSCRIPT leaving residual virus detectable after treatment for 90s at 70 ºC. The composition of meat products may, however, influence the effect of heat treatment on HEV infectivity as in cell cultures assays viruses within a context of matrix were more resistant to heat treatment than viruses in suspension (Yunoki et al., 2008). Heat treatment with a core temperature of at least 71 ºC for 20 minutes was

T

required to prevent HEV infection in the piglets that had been intravenously inoculated with

IP

suspensions of pate composed of infectious HEV liver (Barnaud et al., 2012). However, due to

CR

experimental design there may have been an overestimation of this stringency of heat treatment needed, warranting further work in this field (FSA, 2014). Only a minority of the ready to-eat (RTE) industrial

US

meat products containing blood in our inventory undergoes an industrial heat treatment of ‘71 ºC for 20

AN

minutes’ that would make them completely safe with regards to HEV inactivation. All other products with less stringent treatment could be safe, only when the level of reduction is enough to reduce the

M

viral load of the RTE products to below the infectious dose, this dose, however, is not (yet) known and

ED

may depend on the individual health status of the consumer, and the amount of product consumed. To control the risk of infectious HEV in RTE meat products, industrial production processes (in particular

PT

cooking times and temperatures) should be evaluated and when needed adjusted. An alternative but less

CE

practical approach would be HEV testing of blood at the slaughterhouse or to treat all raw porcine blood prior to use in food industry. In the meanwhile vulnerable persons, such as transplant patients and

AC

patients using immunosuppressive medication, should be informed that consumption of certain meat products containing porcine blood products may increase the exposure and risk of HEV gt 3 infection and therefore are best to be avoided or cooked sufficiently. In the present study, two viral strain sequences were identified that were identical (493 nt) in porcine products and clinical samples. This might indicate a zoonotic transmission. The onset of disease in the patients however preceded the production of the implicated blood products with about half a year,

15

ACCEPTED MANUSCRIPT excluding a direct epidemiological link. To date the main transmission routes for HEV gt3 infections in Western-European countries, like the Netherlands, remain unclear. In addition to the risk described in this study, the exposure may also occur after consumption of products containing porcine liver, fresh produce like salads and soft fruits, shellfish or water if contaminated with HEV (Yugo and Meng,

IP

(15-60 days), which complicates finding an HEV infection source.

T

2013). Discrimination between these transmission routes is hard as the incubation period of HEV is long

CR

Taken together, this study shows for the first time that HEV RNA can be detected in porcine blood products that are used in food. The liquid blood products are not heat-treated during production, leaving

US

the HEV detected most likely infectious. Porcine blood products as an ingredient of processed meat

AN

products, not sufficiently heated prior to consumption, might therefore be considered as a vehicle in

M

transmission of HEV in the food chain.

ED

Funding

This research did not receive any specific grant from funding agencies in the public, commercial or non-

AC

CE

PT

for-profit sectors.

16

ACCEPTED MANUSCRIPT 5.

References

Adlhoch, C., Avellon, A., Baylis, S.A., Ciccaglione, A.R., Couturier, E., de Sousa, R., Epštein, J. Ethelberg, S., Faber, M., Fehér, Á., Ijaz, S., Lange, H., Mand’áková., Z., Mellou, K, Mozalevskis, A., Rimhanen-Finne, R., Rizzi , V. Said, B., Sundqvist, L., Thornton, L., Tosti, M.E., van Pelt,

T

W., Aspinall, E., Domanovic, D., Severi, E., Takkinen, J, Dalton, HR., 2016. Hepatitis E virus:

IP

Assessment of the epidemiological situation in humans in Europe, 2014/15. J. Clin. Virol. 82, 9-

CR

16.

Baert, L., Wobus, C.E., Van Coillie, E., Thackray, L.B., Debevere, J., Uyttendaele, M., 2008. Detection

US

of murine norovirus 1 by using plaque assay, transfection assay, and real-time reverse

AN

transcription–PCR before and after heat exposure. Appl. Environ. Microbiol. 74, 543–546. Barnaud, E., Rogée, S., Garry, P., Rose, N., Pavio, N., 2012. Thermal inactivation of infectious hepatitis

M

E virus in experimentally contaminated food. Appl. Environ. Microbiol. 78, 5153–5159.

ED

Berto, A., Martelli, F., Grierson, S., Banks, M., 2012. Hepatitis E Virus in Pork Food Chain, United Kingdom, 2009–2010. Emerg. Infect. Dis. 18, 1358-1360.

PT

Berto, A., Grierson, S., Hakze-van der Honing, R., Martelli, F., Johne, R., Reetz, J., Ulrich, R-G, Pavio,

CE

N., Van der Poel, W.H., Banks, M., 2013. Hepatitis E virus in pork liver sausage, France. Emerg. Infect. Dis. 19, 264-266.

AC

Clemente-Casares, P., Ramos-Romero, C., Ramirez-Gonzalez, E., Mas, A., 2016. Hepatitis E Virus in Industrialized Countries: The Silent Threat. Biomed. Res. Int. 2016:9838041. Di Bartolo, I., Diez-Valcarce, M., Vasickova, P., Kralik, P., Hernandez, M., Angeloni, G., Ostanello, F., Bouwknegt, M., Rodríguez-Lázaro, D., Pavlik, I., Ruggeri, F.M., 2012. Hepatitis E Virus in Pork Production Chain in Czech Republic, Italy, and Spain, 2010. Emerg. Infect. Dis. 18, 1282-1289.

17

ACCEPTED MANUSCRIPT Diez-Valcarce, M., Kovač, K., Cook, N., Rodríguez-Lázaro, D., Hernández, M., 2011. Construction and analytical application of internal amplification controls (IAC) for detection of food supply chainrelevant viruses by real-time PCR-based assays. Food Anal. Methods. 4, 437–445. European Food Safety Authority (EFSA), 2005. Opinion of the Scientific Panel on Food Additives,

T

Flavourings, Processing Aids and Materials in Contact with Food. Use of an enzyme preparation

IP

based on thrombin:fibrinogen derived from cattle and/or pigs as a food additive for reconstituting

CR

food. EFSA Journal. 214, 1-8.

European Food Safety Authority (EFSA), 2015. EFSA Panel on Food Contact Materials, Enzymes,

AN

pigs blood. EFSA Journal. 13, 4018, 11 p.

US

Flavourings and Processing Aids (CEF) Scientific Opinion on thrombin from cattle (bovines) and

Feagins, A.R., Opriessnig, T., Guenette, D.K., Halbur, P.G., Meng, X-J., 2007. Detection and

M

characterization of infectious hepatitis E virus from commercial pig livers sold in local grocery

ED

stores in the USA. J. Gen. Virol. 88, 912–917. Food Safety Authority (FSA), 2014. A critical review of the effect of heat, pH and water activity on the

PT

survival of Hepatitis A and E viruses. FSA Project FS101074, 98 pp. Retrieved (18th November

CE

2016) from:

https://www.food.gov.uk/science/research/foodborneillness/b14programme/b14projlist/fs101074

AC

Grierson, S., Heaney, J., Cheney, T., Morgan, D., Wyllie, S., Powell, L., Smith, D., Ijaz. S., Steinbach, F., Choudhury, B., Tedder, R.S., 2015. Prevalence of hepatitis E virus infection in pigs at the time of slaughter, United Kingdom, 2013. Emerg. Infect. Dis. 21, 1396-1401. Hogema, B.M., Molier, M., Sjerps, M., de Waal, M., van Swieten, P., van de Laar. T. , Molenaar-de Backer, M., Zaaijer, H.L., 2016. Incidence and duration of hepatitis E virus infection in Dutch blood donors. Transfusion. 56, 722–728.

18

ACCEPTED MANUSCRIPT International Organization for Standardization (ISO), 2017. ISO 15216-1: Microbiology of food and animal feed - Horizontal method for determination of hepatitis A virus and norovirus in food using real-time RT-PCR - Part 1: Method for quantification. International Organization for Standardization, Geneva, Switzerland.

T

Johne, R., Trojnar, E., Filter, M., Hofmann, J, 2016. Thermal Stability of Hepatitis E Virus as Estimated

IP

by a Cell Culture Method. Appl. Environ. Microbiol. 30, 4225-4231.

CR

Jothikumar, N., Cromeans, T.L., Robertson, B., Meng, X-J., Hill, V.R., 2006. A broadly reactive onestep real-time RT-PCR assay for rapid and sensitive detection of hepatitis E virus. J. Virol. Meth.

US

131, 65–71.

AN

Ofori, J.A., Hsieh, Y-H., P., 2012. The Use of Blood and Derived Products as Food Additives. In: ElSamragy, Y. (Ed.), Food Additive. InTech, Chapter 13, 229-256. Retrieved (29th May 2017) from:

M

http://www.intechopen.com/books/food-additive/the-use-of-blood-and-derived-products-as-food-

ED

additives.

Pas, S.D., de Man, R.A., Mulders, C., Balk, A., van Hal, P., Weimar, W., Koopmans, M.P., Osterhaus,

PT

A.d., van der Eijk, A. A., 2012. Hepatitis E Virus Infection among Solid Organ Transplant

CE

Recipients, the Netherlands. Emerg. Infect. Dis. 18, 869-872. Pavio, N., Meng, X-J, Doceul, V., 2015. Zoonotic origin of hepatitis E. Curr. Opin. Virol. 10, 34–41.

AC

Pujols, J., Rodríguez, C., Navarro, N., Pina-Pedrero, S., Campbell, J.M., Crenshaw, J., Polo, J., 2014. No transmission of hepatitis E virus in pigs fed diets containing commercial spray-dried porcine plasma: a retrospective study of samples from several swine trials. Virol. J. 11, 232-239. Ramos-Clamont, G., Fernández-Michel, S., Carillo-Vargas, L., Martininez-Calderón, E., VásquezMoreno, L., 2003. Functional Properties of Protein Fractions Isolated from Porcine Blood. J. Food Sci. 68, 1196- 1200.

19

ACCEPTED MANUSCRIPT Rutjes, S.A., Bouwknegt, M., van der Giessen, J.W., de Roda Husman, A.M., Reusken, C.B., 2014. Seroprevalence of hepatitis E virus in pigs from different farming systems in The Netherlands. J. Food Prot. 77, 640-642. Smith, D.B., Simmonds, P., members of the International Committee on the Taxonomy of Viruses

T

Hepeviridae Study Group, Jameel, S., Emerson, S.U., Harrison, T.J, Meng X-J., Okamoto, H.,

IP

Van der Poel, W.H., Purdy, M.A., 2014. Consensus proposals for classification of the family

CR

Hepeviridae. J. Gen. Virol. 95, 2223–2232.

Tedder, R.S., Ijaz, S. Kitchen, A., Ushiro-Lumb, I., Tettmar, K.I., Hewitt, P., Andrews, N. 2017.

US

Hepatitis E risks: pigs or blood-that is the question. Transfusion. 57:267-272.

AN

Tuladhar, E., Terpstra, P., Koopmans, M., Duizer, E., 2012. Virucidal efficacy of hydrogen peroxide vapour disinfection. J. Hosp. Infect. 80, 110–115.

M

Van der Poel, W.H., Verschoor, F., van der Heide, R., Herrera, M.I., Vivo, A., Kooreman, M., de Roda

ED

Husman, A.M., 2001. Hepatitis E virus sequences in swine related to sequences in humans, The Netherlands. Emerg. Infect. Dis. 7, 970-976.

PT

Van der Poel, W.H.M., 2015. Food and environmental routes of hepatitis E virus transmission. Curr.

CE

Opin. Virol. 4,91–96.

Yugo, D.M., and Meng, X-J., 2013. Hepatitis E virus: foodborne, waterborne and zoonotic

AC

transmission. Int. J. Environ. Res. Public Health. 10, 4507–4533. Yunoki, M., Yamamoto, S., Tanaka, H., Nishigaki, H., Tanaka, Y., Nishida, A., Adan-Kubo, J., Tsujikawa, M., Hattori, S., Urayama, T., Yoshikawa, M., Yamamoto, I., Hagiwara, K., Ikuta, K., 2008. Extent of hepatitis E virus elimination is affected by stabilizers present in plasma products and pore size of nanofilters. Vox Sang. 95, 94-100.

20

ACCEPTED MANUSCRIPT Table 1 Primer oligonucleotides used for typing nested PCR for HEV Sequence (5’-->3’) *

Location†

HEV-orf2-fo-ch

AAY CAR GGi TGG CGY TCi GTi GAR AC

5885-5910

HEV-orf2-ro-ch

GAR AAi GGR CGi GAi GGR GCi GG

6510-6488

HEV-orf2-fi-ch

GAG GAG GAA GCT ACC TCY GGY YTi GTi ATG CTY TGY AT

5924-5961

HEV-orf2-ri-ch

GGA GAA GGA GTT GGT CGR TCY TGY TCR TGY TGR TT

6489-6455

HEV-orf2-fs

GAG GAG GAA GCT ACC TC

HEV-orf2-rs

GGA GAA GGA GTT GGT CG

T

Oligo

IP

5924-5940

CR

6489-6473

Mixed bases in degenerate primers and probes are as follows: Y, C or T; R, A or G; i, Inosine

†

Corresponding nucleotide position of HEV (accession no. NC_001434).

AC

CE

PT

ED

M

AN

US

*

21

ACCEPTED MANUSCRIPT Table 2 Detection of HEV and MuNoV RNA in inoculated blood products Whole blood

IU/0.2 g

Plasma powder

Recovery %

Cq

Recovery %

Cq

Recovery %

avg ± sd

avg ± sd

avg ± sd

avg ± sd

avg ± sd

avg ± sd

1.2 x 104

32.3 ± 0.3

39.8 ± 11.7

32.7 ±0.5

28.1 ± 8.3

32.4 ± 1.2

55.2 ± 35.2

3

1.2 x 10

35.7 ± 0,6

41.4 ± 18.9

35.9 ±0.5

33.0 ± 13.4

36.0 ± 0.7

43.8 ± 35.0

1.2 x 102

38.8 ± 0.7

63.7 ± 52.1

40.3 ±2.0

22.4 ± 26.2

39.3 ± 1.0

nd

nc

40.6 (1/4)

nc

nd

nc

1

†

T

Cq

Fibrinogen

IP

HEV

*

81.4 ± 69.9 (3/4)

40.9 (1/4)

nc

blank

nd

nc

nd

nc

MuNoV

Cq

Recovery %

Cq

Recovery %

Cq

Recovery %

TCID50/0.2 g

avg ± sd

avg ± sd

avg ± sd

45.0 ± 11,0

125.0 ± 95.7

26.7 ± 16.6

33.3 ± 0.6

55.6 ± 16.3

avg ± sd

avg ± sd

4.0 x 10

29.8 ± 0.5

54.7 ±30.7

30.6 ± 0.3

4.0 x 100

33.3 ± 0.2

41.3 ±8.9

34.8 ± 0.9

US

29.7 ± 1.0

4.0 x 10-1

36.6 ± 0.9

53.1 ±22.2

38.2 ± 1.3

30.0 ± 24.1

36.7 ± 0.4

29.0 ± 6.9

39.1 ± 0.1 (2/4)

nc

37.9 ± 0.0 (2/4)

nc

nc

nd

nc

-2

†

38.8 ± 0.1 (2/4)

nc

blank

nd

nc

*

Average and standard deviation of Cq values of four extractions, if not all four extractions tested positive, the number of positives is

PT

presented within brackets.

CE

Nd not detected; nc not calculated

AC

†

nd

ED

4.0 x 10

AN

avg ± sd

1

M

CR

1.2 x 10

22

ACCEPTED MANUSCRIPT Table 3 Monitoring of 60 batches of blood products for the presence of HEV RNA Screening

Quantification

No. positive/

No. positive/

No. tested

No. tested

Whole blood (liquid)

6/6

Copy genomic HEV per 0.2 g avg ± sd

*

No. tested 2.2 x 102 ± 1.9 x 102

6/6

2

7/8

2.8 x 10 ± 2.0 x 10

Hemoglobin (liquid) †

8/8

7/8

< 1 x 102

Fibrinogen (liquid)

8/8

8/8

2.4 x 102 ± 1.3 x 102

3/6

1/6

< 1 x 102

*

*

CR

Stabilized Hemoglobin (liquid)

33/36

29/36

Plasma powder

4/8 #

1/8 #

< 1 x 102

Hemoglobin powder

0/8

Not tested

Not tested

Stabilized Hemoglobin powder

3/8 #

2/8 #

subtotal

7/24

3/16

total

40/60

32/52

AN

M

27/32

0/2 Not tested 0/4 0/6 27/38

ED

appeared not be used in meat products

3/4

part of the samples tested positive in only one of two reactions

PT

#

*

7/8

CE

†

loo low to quantify

*

7/7

AC

*

US

subtotal

< 1 x 102

4/7

T

8/8

*

6/6

2

Plasma (liquid)

*

Typing nested PCR No. positive/

IP

Type of product

23

ACCEPTED MANUSCRIPT Table 4 Application of liquid blood products in processed meat

Type of blood (product)

Avg. copy number/ 0.2 gram blood (product)

Ready to eat product (yes/no) †

Bloodproduct is used in: *

% (v/v) # (estimated genome copies per 100 g of final product)

Industrial heat treatment: lowest temperature/time combination §

liver sausages

Yes

black pudding

Yes

15%

blood

72 ºC

**

4

Yes

Bologna sausage

Yes

US

luncheon meat

2 min

4%

smoked sausage

Yes

M

reaching

temperature.

No

roulade

No

precooked-schnitzel

Yes

4%

precooked cordon bleu

Yes

(4,8 x 103)

precooked pork steak

Yes

precooked breakfast bacon

Yes

CE

75 ºC ††

core

No

smoked raw ham

cooling after

Yes

Not-precooked schnitzel

PT

2.4 x 102

(5.6 x 103)

ED

Tenderloin

AN

2.8 x 102

Immediate

Yes

Berliner liver sausage

Fibrinogen

Length

(1.7 x 10 )

Bologna sausage Plasma

IP

2.2 x 102

No

CR

Whole

minced meat

T

Temperatu re

45 ºC ##

270 min

Yes

AC

* examples of meat products are given. The list is not non-exhaustive. † if yes, than product does not need to be heated in order to make it suitable for human consumption # according to application instructions of manufacturer of blood products § the least stringent temperature/time combination is shown for the group of meat products in which a particular blood(product) is incorporated ** least stringent heat treatment as provided for 7 meat products †† least stringent heat treatment for 40 meat products ## least stringent heat treatment for 28 meat products

24

ACCEPTED MANUSCRIPT Caption to Figure 1

Phylogeny of genotype 3 strains from porcine blood products and patients with acute hepatitis in the Netherlands.

T

Nucleotide sequences of a 493-nt open reading frame 3 fragment (positions 5941- 6472 of reference

IP

sequence NC_001434) from porcine blood products (n = 22) or from patients with acute hepatitis E in

CR

the Netherlands in 2015-2016 (n = 45) were used to produce a Maximum Parsimony tree with

AC

CE

PT

ED

M

AN

US

Bionumerics software.

25

M

AN

US

CR

IP

T

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

Fig. 1

26

ACCEPTED MANUSCRIPT Highlights

CE

PT

ED

M

AN

US

CR

IP

T

Porcine blood products are an ingredient in many (ready-to-eat) meat preparations HEV RNA was detected, quantified and typed as gt3 in most of the tested products Liquid blood products are not heat-treated before entering the food chain HEV in liquid products may be insufficient inactivated in industrial processed meat Two of the obtained genotype 3c strains matched 100% with recent HEV cases

AC

    

27