A newly developed tetraplex real‐time RT‐PCR for simultaneous screening of influenza virus types A, B, C and D

Background Human‐ or avian‐to‐swine transmissions have founded several autonomously circulating influenza A virus (IAV) lineages in swine populations that cause economically important respiratory disease. Little is known on other human influenza virus types, like B (IBV) and C (ICV) in European swine, and of the recently detected novel animal influenza virus type D (IDV). Objectives Development of a cost‐effective diagnostic tool for large‐scale surveillance programmes targeting all four influenza virus types. Methods An influenza ABCD tetraplex real‐time RT‐PCR (RT‐qPCR) was developed in the frame of this study. A selection of reference virus strains and more than 4000 porcine samples from a passive IAV surveillance programme in European swine with acute respiratory disease were examined. Results Two IBV, a single IDV but no ICV infections were identified by tetraplex RT‐qPCR. IBV and IDV results were confirmed by conventional RT‐PCR and partial sequence analysis. Conclusions The tetraplex RT‐qPCR proved fit for purpose as a sensitive, specific and high‐throughput tool to study influenza virus transmission at the human‐animal interface. Complementing close‐meshed active virological and serological surveillance is required to better understand the true incidence and prevalence of influenza virus type B, C and D infections in swine.


| INTRODUC TI ON
Infections with influenza A virus (IAV) are widespread among domestic swine populations worldwide. Due to acute respiratory diseases caused by IAV, clinically often enhanced by further viral and bacterial co-infections, the pork-producing industry faces substantial economic losses. [1][2][3] Influenza A virus belongs to the Orthomyxoviridae family and has a genome with eight segments. Subtypes are categorized by the genetic and antigenic properties of the hemagglutinin (HA) and neuraminidase (NA) membrane glycoproteins: 18 HA types (H1-H18) and 11 NA (N1-N11) types are distinguished today. 4,5 Swine are susceptible to different IAV types, of both avian and human origins, and currently perpetuate autonomous porcine-adapted lineages of subtypes H1, H3, N1 and N2. 6 In addition, sporadic and dead-end infections in swine with avian influenza viruses of various subtypes have been reported. 6 All human pandemic IAVs of the 20th century (with the exception of the 1958 H2N2 viruses) and the current century established autonomously circulating lineages in swine populations following human-to-swine transmission. 6 The most recent human pandemic virus (H1N1/2009, the so-called swine flu) was likely derived from Mesoamerican swine populations. This emphasizes the zoonotic risk of swine influenza A viruses (SIV). 7,8 Zoonotic (swineto-human) and reverse zoonotic (human-to-swine) transmissions of IAV seem to be occurring regularly. [9][10][11][12] Aside from IAV, there are two further human influenza virus types, influenza B virus (IBV) and influenza C virus (ICV). [13][14][15] Similar to IAV, influenza B viruses have eight genome segments including genes encoding HA and NA glycoproteins which, however, have evolved only two distinguishable HA lineages so far. [16][17][18] Like IAV, IBV is regularly involved in seasonal influenza outbreaks in humans but until to date, only three cases of florid IBV infections have been detected. 14,19 In Europe, serological studies from the 1960s showed the sporadic presence of antibodies against IBV in domestic pigs in Hungary. 20 More recently, a surveillance for IBV in U.S. Midwest swine farms showed a seroprevalence of 7.3% at the sample and 38.5% at farm level, respectively. 21 Furthermore, the susceptibility of swine for IBV has been demonstrated under experimental conditions with pigs developing influenza-like symptoms as well as lung lesions. The transmission of the virus to sentinel pigs was also seen in this challenge trial. 21 Influenza C virus is the third human influenza virus type and is composed of seven genome segments. Instead of HA and NA, this virus expresses a hemagglutinin-esterase-fusion (HEF) protein on its surface, conveying both receptor-binding and -destroying functions. 22 It causes only mild diseases in humans, particularly in very young children, and is therefore not included in vaccination schemes. 23,24 The only report of an infection in swine came from China in 1981, where ICV was found in 15 samples during a yearlong monitoring at a slaughterhouse. Detections were confined to the winter and spring months, in pigs with no signs of illness. 25 In 2011, a novel C-like influenza virus has been described in swine in the United States. 26 Like ICV, it is composed of seven genome segments and presents a HEF protein at the membrane surface. However, the low genetic homology and lack of antibody cross-reactivity to ICV led to its designation as a new influenza virus type, tentatively named influenza D virus (IDV). 27,28 To this day, it was found in swine but more frequently in cattle showing mild respiratory disease suggesting cattle may be the domestic reservoir species for IDV. [29][30][31] No human cases were discovered thus far though antibodies for IDV were found in people with close contact to cattle and swine, and in vitro studies demonstrated the ability of the virus to grow in human-derived cells. [32][33][34] In the wake of the H1N1pdm/2009 human pandemic, European surveillance projects for swine IAV revealed that four main IAV lineages were in circulation in domestic swine populations in Europe, one of which was of avian and three of human origin. 35 Very little is known concerning the occurrence of IBV, ICV and IDV in European swine populations and other animal hosts. 20,21,[25][26][27][28][29][30][32][33][34]36 In the light of the fact that most of the porcine IAV strains established in swine are of human origin, it is conceivable that this could also be possible for human IBV and ICV. 2,3,10,11,35 Similar to IBV and ICV, there is no systematic monitoring of IDV in swine, although a few reports carried out in European countries revealed anecdotical presence of IDV in Luxemburg and Italy in swine with influenza-like disease. 31,32,36 For a better understanding of the epidemiology of IBV, ICV and IDV in swine, a time-efficient and low-cost diagnostic tool for largescale screening of four influenza virus types was developed here.
The influenza ABCD tetraplex reverse transcription real-time PCR (RT-qPCR) generically targeting IAV, IBV, ICV and IDV was validated and used to screen 4033 samples from pigs with respiratory disease obtained from 707 farms in twelve European countries.  (Table 1A). Collection sites are illustrated in

| Design of primers and probes
Primers and probes for detection of IAV, IBV, ICV and IDV for use in a tetraplex RT-qPCR were selected from previously published assays Primer and probe sets (Table 2A) were evaluated using strains of the reference collections (Table 3). Based on these results and on comparisons with published assays, further rounds of optimizing oligonucleotide sequences were initiated. Table 2A,B lists final sets of primers and probes for RT-qPCRs (A) and conventional RT-PCRs (B).

| One-step RT-qPCR
The AG-Path-ID ™ One-Step RT-PCR Kit (Ambion) was used throughout. Thermocycling conditions on a Bio-Rad CFX96 real-time PCR detection system were optimized by adapting annealing time and temperature. These cycling conditions were found to be optimal for the generic ABCD-specific RT-qPCR: • 10 minutes 45°C, 10 minutes 95°C, 42 cycles each of 15 seconds 95°C -20 seconds 55°C -30 seconds 72°C.

| Copy-based standard for ABCD-Flu-specific RT-qPCRs
For the production of RNA run-off transcripts, products of the above-  copies/reaction down to one copy were used in the ABCD tetraplex RT-qPCR to determine the limit of detection (LOD). Samples with a Cq-value below 40 were considered positive.

| Virus isolation on cell culture
Madin

| Mapping of farm locations
Maps of the collection sites of the field samples were calculated based on latitude and longitude data with the online tool Microreact (http://www.edge.microreact.org/). 41

| Assembly and analytical performance of an influenza virus type-specific ABCD tetraplex RT-qPCR for the simultaneous detection and differentiation of IAV, IBV, ICV and IDV in porcine samples
Various primer/probe sets, either published or newly developed, were tested in monoplex RT-qPCRs for their sensitivity and  Copy-based standards were used to determine the limit of detection of the tetraplex RT-qPCR ( Figure 2). An LOD of 10 target copies was evident for all four targets when only a single of the four targets was present.
To prove the sensitivity of the tetraplex RT-qPCR for potential  Table S2). In reverse fashion, further mixtures of IBV, ICV and IDV with initial cq-values of 20, 25 and 30 were spiked with IAV at cq 30 and were tested . In addition, triple and 4-fold mixtures were examined as well. All yielded results close to the initial cq-values (Table S2).

| Screening for IBV, ICV and IDV in European domestic swine populations April 2015-March 2017
Only three IAV-negative samples from three different farms yielded a positive RT-qPCR result for either IBV or IDV, albeit with high cqvalues, (Table 4, Figure 1B). To exclude the possibility of false-positive results due to sample or PCR contamination, re-extracted RNA of these samples was examined in conventional IBV-or IDV-specific RT-PCRs as listed in Table 1B. Sequence analysis of these amplicons confirmed the findings of the ABCD tetraplex RT-qPCR (sequences listed in Table S3). The two confirmed IBV-positive cases were detected in samples from the Netherlands (age of swine unknown) and Germany (adult sow), respectively. Temporally, the detection was made in

| D ISCUSS I ON
The zoonotic propensity of porcine IAV is well established and has been demonstrated impressively during the human H1N1 pandemic of 2009. 2,35 In this respect, it is astonishing to note that, in Europe, after the discontinuation of the ESNIP3 monitoring programme, 37 so far only France runs a sustained government-administrated surveillance programme targeting IAV in swine populations. [44][45][46] No ef- F I G U R E 4 Comparison of the IAV cq-values from the initial (monoplex-) IAV screening and the cq-values of the IAV (BCD)tetraplex RT-qPCR. First screening was done within a week of sample receipt; thereafter, RNA was stored at −80°C up to 26 months. Cut-off for the ABCD tetraplex RT-qPCR was set at cq 40 Samples analysed in this study were not representative for the swine populations in the different European countries; it is therefore not possible to draw conclusions regarding the true incidence of influenza virus infections. Also, no data from healthy swine have been collected, since the samples originated from passive surveillance targeting only swine with acute respiratory disease; a similar knowledge gap also exists for IAV infections in healthy pigs.
A broader approach by an active surveillance for influenza virus types would be required to better understand endemic circulation at least of IAV in larger swine holdings. The newly developed tetraplex RT-qPCR is an appropriate tool for this purpose which may also cast further light on potential reverse zoonotic transmission events of IBV and ICV and the presence of IDV in swine populations. In addition, the ABCD tetraplex RT-qPCR should also be useful in screening samples of other host species including man.
Therefore, projects aiming at the examination of influenza virus transmission at the swine-human interface might benefit from this tool.

ACK N OWLED G EM ENTS
We The study has been funded by grants from IDT Biologika, Dessau, Germany (project Ri-0402).
We thank all practicing veterinarians who keep on submitting swine samples for influenza A virus monitoring.