A novel quantitative real-time polymerase chain reaction method for detecting toxigenic Pasteurella multocida in nasal swabs from swine

Background Progressive atrophic rhinitis (PAR) in pigs is caused by toxigenic Pasteurella multocida. In Switzerland, PAR is monitored by selective culture of nasal swabs and subsequent polymerase chain reaction (PCR) screening of bacterial colonies for the P. multocida toxA gene. A panel of 203 nasal swabs from a recent PAR outbreak were used to evaluate a novel quantitative real-time PCR for toxigenic P. multocida in porcine nasal swabs. Results In comparison to the conventional PCR with a limit of detection of 100 genome equivalents per PCR reaction, the real-time PCR had a limit of detection of 10 genome equivalents. The real-time PCR detected toxA-positive P. multocida in 101 samples (49.8%), whereas the conventional PCR was less sensitive with 90 toxA-positive samples (44.3%). In comparison to the real-time PCR, 5.4% of the toxA-positive samples revealed unevaluable results by conventional PCR. Conclusions The approach of culture-coupled toxA PCR for the monitoring of PAR in pigs is substantially improved by a novel quantitative real-time PCR. Electronic supplementary material The online version of this article (doi:10.1186/s13028-016-0267-7) contains supplementary material, which is available to authorized users.


Findings
Progressive atrophic rhinitis (PAR) in pigs is caused by toxigenic Pasteurella multocida (capsule types A and D) which synthesize a 145-kDa toxin (dermonecrotic toxin) [1] encoded by the chromosomal toxA gene [2,3]. PAR is of worldwide economic importance to swine production due to a reduction in growth rate in fattening pigs. In Switzerland, PAR is monitored in about 90% of the Swiss primary nucleus and secondary multiplying herds by annual bacteriological examination of 10 nasal swabs per herd for toxigenic P. multocida. Swabs are analysed by culturing the swabs on selective agar [4] and subsequent screening of bacterial colonies for toxigenic P. multocida by polymerase chain reaction (PCR) [5]. During a recent outbreak of PAR in Swiss multiplying herds [6], nasal swabs from feeder pigs were tested for toxigenic P. multocida by the method outlined above. A panel of 203 nasal swabs from this PAR outbreak were further used to evaluate a novel quantitative real-time PCR (qRT-PCR) for toxigenic P. multocida in porcine nasal swabs. The bacterial strains used in this study are given in Table 1. Strains were grown on 7% sheep blood agar aerobically at 37 °C for 24 h (PB5008A, Oxoid, Pratteln, Switzerland).
Nasal swabs were collected from a total of 203 feeder pigs on farms affected by PAR [6]. Cotton swabs were transported in Amies medium in screw cap plastic vials (VWR, Dietikon, Switzerland) to the lab. Swabs were streaked onto selective blood agar plates according to Rutter et al. [4]  Genomic DNA was released by heat lysis of bacterial colonies resuspended in water on a thermal shaker for 10 min at 99 °C. After centrifugation at 17,000g for 3 min the supernatant was transferred into a fresh tube and stored at -20 °C until further use. Bacteria from nasal swab cultures were processed accordingly.
Gene specificity of both primers and the probe were confirmed by BLAST searches. Primers were synthesized by Microsynth (Balgach, Switzerland). DNA probes were supplied by Eurogentec S.A. (Seraing, Belgium). The probes were quenched by black-hole non-fluorescent quenchers at the 3′-end. Rox dye (Life Technologies, Darmstadt, Germany) was used as an internal reference for normalization and data analyses. An internal amplification control (IAC) was introduced for monitoring each reaction, since bacterial lysates could contain inhibitory substances. Five femtogram (fg) of the pEGFP-1 strandard vector (BD Bioscience Clonetech, USA) was used as IAC template, and a 177 bp amplicon [7] was amplified with primers and the probe given in Table 2. DNA (1 pg) from a toxA-positive and a-negative P. multocida reference strain (ATCC 12948, ATCC 43137) were used as controls in each qRT-PCR run.
Cycling conditions were 10 min at 95 °C, followed by a two-step cycling stage of 40 cycles of 15 s at 95 °C and 60 s at 60 °C. Data analysis was handled by the 7500 Software version 2.0.4 (Life Technologies). Samples were considered positive when presenting a typical amplification curve with a Ct value of ≤38 for toxA and a Ct value of ≤32 for the IAC. Analyses of samples with IAC Ct values >32 were repeated after reduction of PCR-inhibitory substances by 1:2 and 1:10 dilution.
With an estimated genome size of 2.26 Mb a DNA quantity of 2.5 fg was calculated for one genome equivalent (GE) of P. multocida [8]. The efficiency of the qRT-PCR was determined by plotting standard curves (Ct values against quantified GE) in 10-fold dilution from  2 × 10 7 to 2 × 10 3 GE. The slope of the linear relationship of this curve was used to calculate the amplification efficiency [9]. To determine the minimum detectable bacterial concentrations for the qRT-PCR, twelve replicates of a DNA dilution series containing 4000, 400, 40, 20, 10, 5 and 2 GE of P. multocida ATCC 12948 per 10 μl PCR reaction were analysed. As illustrated in Table 3, the limit of detection (LOD) was determined to be 10 GE, the lowest dilution for which the acceptance criteria (Ct <38 and a standard deviation of <0.5) were fulfilled. The sensitivity of both the qRT-PCR and the conventional PCR assays was confirmed by testing toxA-positive P. multocida strains. The specificity of the qRT-PCR was confirmed by analysis of 32 different bacterial species (Additional file 1).
An amplification plot of a dilution series of toxA was created (Additional file 2). The efficiency of the qRT-PCR was determined by the use of serial dilution standard curves. The linear correlation coefficient r 2 for the toxA gene target was 0.999, showing a high degree of linearity of the qRT-PCR. Based on the slope of the standard curve (-3.348), the amplification efficiency was calculated as 98.9%. At DNA template amounts >1μg, the eGFP amplification was reduced and was completely inhibited with 1.8 μg template DNA. For diagnostic application of the qRT-PCR, a DNA template of maximum 1 μg DNA template was used. To determine the LOD of the assay the acceptance criteria of Ct < 38 and a standard deviation of <0.5 were used. With a LOD of 10 GE, the qRT-PCR showed a tenfold higher sensitivity than the conventional PCR with a LOD of on average 100 GE per PCR reaction [5]. The specificity of both PCR assays was confirmed by repeated test of toxA-positive P. multocida strains. The panel of 32 heterologous bacteria scored negative (Additional file 1). Overall, the qRT-PCR assay revealed a high precision as confirmed by the inter-and intra-assay coefficient of variations (CV) of the Ct values of the positive control as well as the IAC from 52 independent PCR runs. The CV interassay values were 5.9% for toxA and 2.7% for IAC; the intra-assay CV values were 1.1% for toxA and 0.58% for IAC.
By comparative analysis of 203 nasal swabs from pigs suspected for PAR, 90 samples (44.3%) were identified as toxA gene-positive by both PCR protocols. A sample of 11 swabs (5.4%) were identified as toxA gene-positive by qRT-PCR (Ct values 16-33.5), whereas the conventional PCR revealed unevaluable results (e.g. several PCR amplifications products of different but unexpected sizes) despite the positive and negative control reactions performed as expected. Overall, the qRT-PCR detected toxA-positive P. multocida in 101 swabs (49.8%). By comparison, the conventional PCR was less sensitive with 90 samples (44.3%) identified as toxA-positive P. multocida.
Noteworthily, a Pasteurella canis strain from the skin wound of a cat was identified as toxA-positive by our novel qRT-PCR. This finding was verified by two independent PCR reactions: Species identification was confirmed by the P. canis-specific PCR of Krol et al. [10], targeting sodA, a housekeeping gene encoding manganese-dependent superoxide dismutase. In addition to the PCR species identification, strain ZH 401 was confirmed as P. canis by MALDI-TOF analysis. Moreover, a partial fragment of toxA from P. canis ZH 401 has been sequenced and shows a few point mutations in comparison to the highly conserved sequences of toxA-positive P. multocida strains (Additional file 3). The presence of toxA in our feline P. canis was confirmed with an in-house toxA-PCR (primer 34_ToxA 5′ ACTGT AAAAGGAAAAAGTGCCGATG 3′ and 35_ToxA_rev 5′ AAGAGGAGGCATGAAGAGTGC 3′) resulting in a 3792 bp toxA specific amplicon. The expression of toxA in strain ZH 401 was confirmed with the aid of a P. multocida toxin (PMT) ELISA (Oxoid, Pratteln, Switzerland). The finding of toxA in other bacteria than P. multocida is of particular interest. Since toxA is encoded within a lysogenic prophage, it may be transferred to other bacteria (as in our case to P. canis) by transduction [11].
To conclude, our novel qRT-PCR is highly efficient and robust to diagnose toxA-positive P. multocida in nasal swabs. Compared with the conventional PCR, the