Detection and Typing of Highly Pathogenic Porcine Reproductive and Respiratory Syndrome Virus by Multiplex Real-Time RT-PCR

Porcine reproductive and respiratory syndrome (PRRS) causes economic losses in the pig industry worldwide, and PRRS viruses (PRRSV) are classified into the two distinct genotypes “North American (NA, type 2)” and “European (EU, type 1)”. In 2006, a highly pathogenic NA strain of PRRSV (HP-PRRSV), characterized by high fever as well as high morbidity and mortality, emerged in swine farms in China. Therefore, a real-time reverse transcription polymerase chain reaction (RT-qPCR) assay specific for HP-PRRSV was developed and combined with type 1- and type 2-specific RT-qPCR systems. Furthermore, an internal control, based on a heterologous RNA, was successfully introduced. This final multiplex PRRSV RT-qPCR, detecting and typing PRRSV, had an analytical sensitivity of less than 200 copies per µl for the type 1-assay and 20 copies per µl for the type 2- and HP assays and a high diagnostic sensitivity. A panel of reference strains and field isolates was reliably detected and samples from an animal trial with a Chinese HP-PRRS strain were used for test validation. The new multiplex PRRSV RT-qPCR system allows for the first time the highly sensitive detection and rapid differentiation of PRRSV of both genotypes as well as the direct detection of HP-PRRSV.


Introduction
Porcine reproductive and respiratory syndrome (PRRS), one of the economically most important diseases of swine worldwide, is characterized by reproductive failure in pregnant sows and respiratory disease in piglets [1]. The causative agent is the antigenically, genetically and clinically heterogeneous PRRS virus (PRRSV) [2][3][4], an enveloped positive-strand RNA virus that belongs to the order Nidovirales, family Arteriviridae [5]. PRRSV isolates are generally classified into genotype 1 (EU, type 1) and genotype 2 (NA, type 2) [6].
In 2006, a highly pathogenic strain of type 2-PRRSV (HP), causing high fever and severe morbidity and mortality in pigs of all ages, emerged in swine farms all over China [7]. HP-PRRSV, characterized by a unique discontinuous deletion of 30 amino acids (aa) in the non-structural protein 2 (Nsp 2) [7], affected a large number of pigs and causes enormous economic losses [8,9].
Apart from China, HP-PRRSV has already been detected in other Asian countries such as Vietnam, where it caused a serious epidemic [9,10]. Therefore it is essential to monitor the spread of these highly pathogenic PRRSV strains. For the detection of HP-PRRSV several real-time reverse transcription polymerase chain reaction (RT-qPCR) assays have been developed [8,11].
In this study a HP-PRRSV specific RT-qPCR assay was developed and combined with systems specific for type 1 and 2. An internal control (IC) using heterologous RNA [12] was successfully included to verify efficient RNA extraction and uninhibited amplification. Furthermore, an animal trial was conducted with a Chinese HP-PRRSV strain in pigs and the multiplex assay was further validated with samples from the experimental infection.

Viruses and Diagnostic Samples
The PRRSV strains used in this study (Table 1) were provided by the virus collection of the Friedrich-Loeffler-Institut, Isle of Riems, Germany. The Eastern European isolate ''Lena'' [13] was kindly provided by U. Karniychuk (Ghent University, Belgium).

RNA Extraction
RNA was extracted with the QIAampH Viral RNA Mini Kit (Qiagen GmbH, Hilden, Germany) according to the manufacturer's recommendations, modified by addition of an internal control RNA (IC2) as described by Hoffmann et al. [12], and finally eluted in 50 ml kit elution buffer. RNA from diagnostic serum samples was extracted with the MagNA Pure LC Total Nucleic Acid Isolation Kit for automated extraction (Roche Diagnostics Deutschland GmbH, Mannheim, Germany) and finally eluted in 100 ml elution buffer.

Primers, Probes and Real-time PCR
Published sequence information of highly virulent, Chinesetype, isolates of PRRSV was used for the selection of primers and probes. Based on sequence alignments, primers and probes were selected with Beacon Designer 7.01 (PREMIER Biosoft International, Palo Alto, CA, USA). Primers and probes of the type 1 and 2 assays have been described previously [14]. In contrast to the type 2 assay by Kleiboeker et al. (2005), which uses two forward primers, a single modified forward primer was applied in this study. Furthermore the type 1 system was adapted using the available sequence information about European strains. Primers and probe of the EGFP internal control assay have been published by Hoffmann et al. [12]. Sequences of all primers and probes are shown in Table 2.
The oligonucleotides were transcribed in vitro with the SP6/T7 Transcription Kit (Roche Diagnostics Deutschland GmBH, Mannheim, Germany) according to the manufacturer's instructions. The T7-transcribed positive controls were digested with DNaseI and purified using the RNeasy Kit (Qiagen GmbH, Hilden, Germany). The RNA concentration was determined by spectrophotometry and the exact number of RNA molecules was calculated.

Animal Experiment with HP-PRRSV
On day 0 of the study, three serologically PRRSV negative swine (6-8 weeks of age) were inoculated intranasally with 10 5.8 Table 3. Sequences of positive standards used in this study, primer and probe binding regions are depicted in capitals.

Sensitivity
The analytical sensitivity of the multiplex PRRSV RT-qPCR was determined with a series of 10-fold dilutions of the positive standards. The type 2 and HP-assays showed an amplification efficiency of more than 98% with a linear dynamic range from 10 6 copies down to 20 copies per ml (Fig. 1). For the type 1-assay, the linear dynamic range reached down to 200 copies per ml with a PCR efficiency similar to the type 2-assay (Fig. 1).
The sensitivity of each PRRSV assay in the presence of the other genotype was tested in a checkerboard titration where the EU strain ''Stendal V953'' was titrated against the HP strain ''China''. Internal control RNA was added to the RT-qPCR master mix. Since the HP-strains belong to type 2, amplification was observed in both RT-qPCR systems, PRRSV-HP and type 2-PRRSV. The sensitivity of the type 2 and HP-assay was hardly affected even by the co-amplification of high amounts of a type 1-PRRSV (Fig. 2a, 2b). Both assays showed comparable quantification cycle (Cq; [15]) values. The 10 27 dilution of the HP-strain scored positive in three samples out of six in the type2-assay, and four out of six in the HP system. In the samples which were tested negative in one or both assays, the concentration of the type 1strain, also present in the reaction, ranged from 10 22 to 10 27 . Independent of the HP-and type 2 strain amplification, type 1 PRRSV was reliably detected even in low concentrations (Fig. 2c). In addition, the IC-assay was obviously not affected by the coamplification of PRRSV (Fig. 2d).
The multiplex RT-qPCR was also compared to the ADIA-VETH PRRS EU/NA Kit. Both systems achieved equal results for diverse virus strains ( Table 1).
The diagnostic sensitivity was evaluated by testing different sample types, including serum and lung samples ( Table 4). All samples showed comparable results when tested with the novel multiplex PRRSV RT-qPCR and the ADIAVETH PRRS EU/ NA Kit. The multiplex PRRSV RT-qPCR assay failed to detect the IC in one case, the ADIAVETH PRRS EU/NA kit with two samples.

Specificity
The sequences of primers and probes included in the PRRSV multiplex RT-qPCR were compared to published sequence information (NCBI GenBank) with a special focus on porcine viruses. There was no indication of possible cross-reactions.
The diagnostic specificity of the multiplex RT-qPCR assay was validated using 132 serum samples from animals previously tested negative for PRRSV by a commercial RT-qPCR kit. All scored clearly negative in the type 1 and 2, as well as PRRSV-HP-specific assay within the multiplex approach with positive results in the ICspecific RT-PCR.
When testing the virus strains (Table 1) as well as the diagnostic samples (Table 4), the multiplex PRRSV RT-qPCR and ADIAVETH PRRS EU/NA Kit correctly determined the genotype for all samples.
The inter-assay repeatability was determined for 10-fold dilution series on four separate days. Mean values and standard deviations of the PRRSV specific assays are shown in Fig. 3. The IC-specific PCR showed a mean Cq value of 33.86 with a standard deviation of 0.72.

Animal Experiment with HP-PRRSV
The body temperature of the infected pigs reached 40uC at day 3 post infection (dpi) and 41-42uC at 8 dpi. The fever then lasted until the end of the study. In-contact animals reached 40uC 3 days after first exposure and more than 41uC after 8 days. Monitoring of clinical signs revealed symptoms like red to purple dermal discoloration and purple and blue ears 5 days after infection or 7 days after first exposure, respectively. The discolorations persisted until the end of the study. Additionally, one of the infected and one of the in-contact animals showed diarrhoea and nasal discharge. Coughing was observed in one of the infected animals, while sneezing was seen in one of the in-contact animals.
Serum samples of infected and in-contact animals were investigated using the multiplex PRRSV RT-qPCR (Fig. 4). Samples from day 0 of the study scored negative for all animals. On 3 dpi the infected animals showed positive results, while the incontact pigs remained negative. On 5 dpi PRRSV-RNA could be detected in samples of all animals.
In serum samples of the infected animals, Cq values declined until 7 dpi, and subsequently increased till 14 dpi. The in-contact animals showed decreasing Cq values up to day 10, followed by increasing values also until the end of the study. All samples showed equal results in the type 2 as well as the HP-specific system ( Figure 4). However, one weak positive sample scored negative with the type 2-specific assay, whereas the HP-specific assay gave a positive result.
The WBC count of infected animals dropped from 0 dpi until 3 dpi, and subsequently increased again. In-contact animals showed a decreasing number of WBC from day 3 until day 7 ( Figure 5).

Discussion
In 2006, a highly pathogenic disease emerged in swine farms in China. The morbidity rate was 50-100% and the mortality rate was 20-100% [16]. In the affected swine herds, high and continuous fever, anorexia, red discolorations in the bodies and blue ears were observed, followed by diarrhoea and other clinical signs in the late phase of the disease. A highly pathogenic PRRSV type with a 30-aa deletion within the Nsp2 encoding genome region could be demonstrated as the causative agent. Under experimental conditions, the virus could induce clinical symptoms identical to those observed in the field [7,[17][18][19]. However, none of these trials used in-contact control animals to study virus transmission.
In our animal experiment the typical clinical symptoms of high and continuous fever and dermal discolorations could be observed in infected as well as in contact animals. The in-contact animals showed clinical symptoms 3 days later than the infected swine. As the control animals were placed in contact to the infected swine on day 3 of the study, a very rapid infection has to be assumed. This is reinforced by the decrease of WBC two days after the first exposure. Samples from the animal trial were used to validate the newly developed PRRSV multiplex assays, allowing a differentiation of type 1 and 2 as well as a direct characterization of a PRRSV-strain as the HP-type. This is possible, since the Nsp2encoding region within the 15 kb genome of PRRSV [20] shows a remarkable genetic variation [21,22]. Although the 30-aa deletion in Nsp-2 coding region of HP-PRRSV is not related to its virulence [23], the deletion can be used as a marker to distinguish the highly pathogenic Chinese virus from other NA-PRRSV strains.
For the detection of HP-PRRSV several RT-qPCR assays have been developed [8,11]. None of these assays, however, can detect other type 2-or even type 1-strains. Furthermore, RT-PCR assays for the detection and differentiation of both genotypes have been described [14,24,25], but no HP-PRRSV detection system was included in these assays.
Type 1 PRRSV isolates have been recently introduced into North American swine herds infected with type 2 strains [2]. Natural dual infection of swine with strains of both genotypes has been documented in Europe [26]. Even in two diagnostic samples used in this study, the type 1 and 2specific assays scored positive, emphasizing the necessity for simultaneous detection and differentiation of both genotype strains.
By testing dilution series of appropriate positive standards, it was demonstrated that the novel multiplex PRRSV RT-qPCR system is a highly sensitive approach for the detection and differentiation of type 1 and 2 PRRSV and allows also the direct characterization of HP-specific genome sequences. Checkerboard titration confirmed that the sensitivity of the respective assays was not affected by the co-amplification of any other target included in the multiplex assay. In the lowest detectable dilution of HP-PRRSV either the HP-or the type 2 assay or both scored positive, independently of the concentration of the type 1-strain RNA in the same sample. Consequently, the presence of HP-PRRSV genome in weak type 2-positive, and HP-negative, samples cannot be fully excluded.
When testing diverse virus strains and diagnostic samples including serum and lung samples, the new multiplex RT-qPCR showed a high diagnostic sensitivity. However, the number of used strains is limited consequently not representing the full range of the diversity of both genotypes. Especially the Eastern European type 1 isolates are highly diverse [27,28]; one of those strains (''Lena'') was tested by the multiplex RT-qPCR and correctly detected. Sequence information of Eastern Europe isolates, in particular ORF 7 sequences, which covers the primers and probes used in this study, are scarce. As a consequence, the applicability of the multiplex PCR for those strains in not completely proven. Nevertheless, the known sequences of primers and probes enable a rapid adaption to new sequence information and emerging strains, which is an important feature in PCR diagnostics of viruses exhibiting a large genetic diversity, such as PRRSV [29][30][31][32]. In addition to the PRRSV type 1, 2 and HP-specific assays, an internal control PCR based on heterologous RNA sequences was introduced. To verify efficient RNA extraction and uninhibited amplification, IC based on housekeeping genes [33][34][35][36] or heterologous internal controls [12,37] are used. Depending on the quality of the clinical sample, even housekeeping genes have unknown and variable mRNA concentrations, which could negatively influence the amplification and detection of the virusderived nucleic acid. As the concentration of the heterologous IC RNA is known, sub-optimal RNA extractions or partial inhibitions of the RT-PCR can be detected, therefore avoiding false-negative results. However, a negative result in the IC specific PCR could also be caused by the absence of detectable RNA, but this is not significant when the virus-specific PCR scored positive.
A large number of samples from PRRSV-negative pigs were also tested and scored clearly negative in the PRRSV assays, whereas all of them scored positive in the IC-specific RT-qPCR. Using the checkerboard titration, it could be shown that the ICspecific assay was not affected by the co-amplification of PRRSV.
In serum samples obtained from the animal trial, viral genome was first detectable at day 3 after infection or first exposure, demonstrating the rapid infection of the introduced contact animals. Except in a weak positive sample, where the HP-assay scored positive and the type 2 negative, both detection systems showed comparable Cq-values.
In conclusion, the newly developed 4-colour multiplex RT-qPCR is a sensitive and specific method to confirm the presence of PRRSV genomes in field samples. For the first time, this assay can also detect and distinguish type 1 and 2 PRRSV as well as the Chinese HP-strain in a single qPCR reaction.