Development of a multiplex PCR assay for the simultaneous and rapid detection of six pathogenic bacteria in poultry

Escherichia coli, Pasteurella multocida, Proteus mirabilis, Pseudomonas aeruginosa, Salmonella spp. and Staphylococcus aureus are six bacterial pathogens of avian. However, these pathogens may cause many similar pathological changes, resulting in clinical isolates that are difficult to quickly and simultaneously detect and identify. Here, a multiplex polymerase chain reaction (m-PCR) assay is reported to rapidly identify targets genes (phoA, KMT1, ureR, toxA, invA, and nuc) of these six pathogens in clinical samples. Six pairs of specific primers were designed. The optimal reaction conditions, specificity, and sensitivity of the m-PCR assay were investigated. The results showed that betaine remarkably improved amplification of the target genes. Specific test results showed that all six pathogens were detected by the proposed m-PCR protocol without cross-amplification with viruses or parasites. Sensitivity test results showed that the m-PCR system could amplify the six target genes from bacterial genomes or cultures with template amounts of 500 pg or 2.8–8.6 × 103 colony forming units, respectively. Furthermore, the six bacterial pathogens isolated from the infected tissue samples were successfully identified. The proposed m-PCR assay is a useful tool to monitor and diagnose bacterial infection in birds with high specificity, sensitivity and throughput.


Introduction
Several factors have been linked to the spread of pathogenic bacteria to poultry, including the expansion of the poultry industry, the increased mobility of humans and animals, water pollution, environmental climate change (Rodriguez-siek et al. 2005;Benskin et al. 2009). Furthermore, antibiotic administration is conventionally used for the control of bacterial diseases in poultry. However, the failure to diagnose the bacterial diseases of poultry, which may result in the misuse of antibiotic regimens and subsequent severe economic losses to the poultry industry and potential public health risks due to the consumption of contaminated poultry products (Van Den Bogaard et al. 2002).
A variety of methods have been established for the effective diagnosis of avian bacterial diseases, which include antigen-specific enzyme-linked immunosorbent assays (ELISAs), immunogold labeling and various other molecular biology techniques (Kotetishvili et al. 2002;Yano et al. 2007;Reischl 1996), especially polymerase chain reaction (PCR) technologies. However, the failure of multi-pathogen detection still was one of major deficiencies to these detection methods. For example, Park et al. (2011) established a triple PCR method for analysis of Campylobacter spp., Escherichia coli O157:H7 and Salmonella serotypes. Hu et al. (2011) Belgrader et al. (1999) developed a rapid PCR assay that detected bacteria in 7 min and Han et al. (2011) established a loopmediated isothermal amplification technique based on the GroEL gene for rapid detection of Riemerella anatipestifer.
Furthermore, although the most important diseases are viral in poultry, the bacterial diseases are also important, some studies have shown that the main bacterial pathogens of poultry (including avian pathogenic Escherichia coli, Pasteurella multocida, Salmonella spp. and Staphylococcus aureus) also caused severe economic losses and restricted the development of the poultry industry (Bisgaard 1993). In addition, although Proteus mirabilis and Pseudomonas aeruginosa are not considered among major bacterial pathogens for chickens, which still may spontaneously cause infection for chickens (Walker et al. 2002). More importantly, bacterial and viral infections often occur simultaneously, but the similarity of clinical signs of infected animals and the lack of high-throughput methods for the detection of pathogens, especially opportunistic species, such as Proteus mirabilis and Pseudomonas aeruginosa, have seriously hampered the control of epidemic diseases (Salmon and Watts 2000;Tanaka et al. 1995). Additionally, the sensitivity and specificity of colloidal gold detection technologies and ELISA techniques are relatively low, but yet these tests are costly. In contrast, multiplex PCR (m-PCR) can detect multiple pathogens with only one reaction with high sensitivity and specificity to distinguish between very closely related organisms, which greatly reduce costs. Hence, m-PCR is a promising tool for the efficient and accurate identification of pathogenic microbes.
To address these problems, a m-PCR assay for the simultaneous and rapid detection of six bacterial pathogens of poultry was developed in this study. The m-PCR assay showed high specificity, sensitivity and throughput, which should facilitate the prevention and rapid diagnosis of avian bacterial diseases.

Bacterial strains and growth conditions
Six pathogenic bacteria were isolated from diseased birds ( Table 1). The bacterial, viruses and parasites were preserved in our laboratory (Table 1). Pasteurella multocida was cultured in sterile Martin broth medium (Qingdao Hope Bio-Technology Co., Ltd., Qingdao, Shandong, China) overnight at 37 °C. Streptococcus All bacteria were cultured until the mid-log phase, and then the bacterial genomes were extracted according to the previous methods (Velegraki et al. 1999) with some modifications. The genomes of the parasites and viruses were preserved in our laboratory.

Numbers of colony forming units (CFU) of the six pathogens
After culturing of the six pathogenic bacteria overnight on agar plates, the cells were collected and washed twice with phosphate-buffered saline. The optical density at 600 nm (OD 600 ) of the bacterial suspensions was adjusted to 1, then 10 4 -, 10 5 -, 10 6 -, and 10 7 -fold dilutions were prepared. Aliquots (2 μL) of the bacteria solution were placed in agar plates, which were cultured overnight at 37 °C. After 12 h, the CFUs of six pathogens (OD 600 = 1.0) were counted, respectively.
The sequences of the phoA, KMT1, ureR, toxA, invA, and nuc genes were obtained from the GenBank database (http://www.ncbi.nlm.nih.gov/Genba nk/). All oligonucleotide primers used in this study were synthesized by Shanghai Sunny Biotechnology Co., Ltd. (Shanghai, China). The sequences of the PCR primers are shown in Table 2.
In order to evaluate and verify the specificity of the primers, PCR analysis was performed using the genomes of the six pathogens as DNA templates.

Optimization of m-PCR primers
Optimization of the primer combinations was based on the orthogonal experimental method. In the 15 double combinations, the optimal combination was selected as the initial double PCR and the remaining four primer pairs (the initial concentration of each primer was 0.4 µM) were added to the double combination to form a triple PCR. An optimal triple PCR was then selected and the remaining three primer pairs were added to form a quadruple PCR, until completion of the m-PCR.
After the addition of a new primer pair to an optimal PCR, if the combination was not optimal, the primers were redesigned, and then the concentration of each primer was adjusted from initial concentration of 0.4 µM to achieve the best results.

Optimization of m-PCR conditions
The PCR reaction is affected by many factors. Therefore, the parameters of the m-PCR assay were optimized by varying concentration of deoxyribonucleotide triphosphate (dNTPs; 0.1-0.4 mM), Mg 2+ (0.2-0.5 mM), Taq DNA polymerase (1.0, 1.5, 2.0, and 2.5 U), and betaine (0.05-0.4 mM) in a 25-µL reaction volume. A mixture of the genomic DNA, which contained same amount of genomic DNA of the six types of bacteria, was used as a template to amplify the corresponding target genes. The total volume of each reaction system (recommended system) was 25 μL, which included 1 μL of template DNA (about 150 ng of genomic DNA).
PCR cycles were as follows: pre-denaturation at 94 °C for 4 min, denaturation at 94 °C for 40 s, annealing at 58 °C for 30 s, extension at 72 °C for 1 min, for 25-35 cycles, extension at 72 °C for 10 min, and preservation at 16 °C. After the reaction, 5 μL of the reaction solution was mixed with 1 μL of loading buffer (6×; TaKaRa Biotechnology) for 1.5% agarose gel electrophoresis.

Specificity of the m-PCR assay
In order to confirm the specificity of the m-PCR established in this experiment, the genomes of seven species of bacteria (including Klebsiella pneumoniae, Shigella spp., Bacillus subtilis, Bacillus cereus, Enterococcus faecalis, Listeria monocytogenes and Streptococcus suis), two avian parasites (Cryptosporidium baileyi and Eimeria tenella) and three viruses (NDV, IBDV and AIV) were selected as the DNA template for m-PCR under optimized conditions (Table 3). The m-PCR test was performed with the genomes of six bacteria as the template DNA, which served as a positive control.
Furthermore, different serotypes of bacterial species were selected to verify the specificity of the multiplex PCR detection system: including O 1 , O 2 and O 78 serotype of avian pathogenic Escherichia coli, Salmonella typhimurium, Salmonella enteritidis and Salmonella pullorum of Salmonella spp.

Sensitivity of the m-PCR assay
The sensitivity of the m-PCR assay was evaluated using a tenfold serial dilution method. Briefly, the six strains were cultured to OD 600 = 1 and then diluted to 0.1, 0.01, and 0.001, and 2 µL of the above diluents were used as PCR templates.

M-PCR for the detection of six pathogenic bacteria from experimentally or naturally infected tissue samples
The ability of the m-PCR assay to detect six pathogens in liver, spleen, and blood samples from experimentally infected chicks was evaluated. The 7-day-old San Huang chicks were obtained from Songjiang Chicken Farm (Shanghai, China) and were housed in cages under a controlled temperature of 28-30 °C and a 12 h light/dark cycle with free access to food and water during the study period. Briefly, 7-day-old San Huang chicks were injected with 5 × 10 5 CFU of Escherichia coli, 5 × 10 3 CFU of Pasteurella multocida, 2 × 10 8 CFU of Proteus mirabilis, 1 × 10 8 CFU of Pseudomonas aeruginosa, 5 × 10 4 CFU of Salmonella spp. and 2 × 10 7 CFU of Staphylococcus aureus in the leg muscle, respectively. Then, the liver, spleen, and blood samples were aseptically collected 24 h after injection in accordance with the guidelines of the Animal Management and Use Committee of the Shanghai Veterinary Research Institute (Chinese Academy of Agricultural Sciences). The liver, spleen, and blood samples were homogenized in phosphate-buffered saline, then cultured for 4 h in LB broth, and boiled for 5 min to extract the genomic DNA for m-PCR detection. Genomic DNA was extracted using a DNA extraction mini kit (Tiangen Biotech Co., Ltd., BeiJing, China) according to the manufacturer's instructions. Furthermore, the 6 bacterial genome mixtures (including Escherichia coli, Pasteurella multocida, Proteus mirabilis, Pseudomonas aeruginosa, Salmonella spp. and Staphylococcus aureus) were prepared in advance as DNA template, which was used as a positive control for the m-PCR assay. Moreover, for evaluation of the potential application of this PCR in clinical investigation, some tissue samples from diseased chicks from different poultry farms were processed during 2018-2019. The tissue samples were tested as described above.

Optimization of the m-PCR primers
As shown in Fig. 2, of the 15 double combinations, the optimal combination of Proteus mirabilis and Salmonella spp. (Fig. 2, lane 1) was selected for the initial double PCR assay. Subsequently, a third primer pair was added to form a triple PCR. For the triple m-PCR assay, the combination of Proteus mirabilis, Salmonella spp. and Pasteurella multocida (Fig. 2, lane 2) was optimal. According to the orthogonal experiments, quadruple, quintuple, and sextuple m-PCR assays were successively established (Fig. 2, lanes 3, 4, and 5, respectively).

Optimization of the m-PCR conditions
The results showed that the optimal annealing temperature of the m-PCR reaction was 54 to 58 °C, while the optimal dNTP and Mg 2+ concentrations were 0.1 mM and 2.5 mM, respectively (data not shown). The optimum betaine concentration was 0.4 M (data not shown).
In addition, the concentrations of each pair of primers were optimized based on the orthogonal experimental method, the results showed that the optimal concentrations of each pair of oligonucleotide primers were 0.2 µM (Pseudomonas aeruginosa and Pasteurella multocida), 0.4 µM (Proteus mirabilis, Salmonella spp. and Escherichia coli) and 0.8 µM (Staphylococcus aureus), respectively (data not shown).
Furthermore, the number of cycles largely determines the required total duration of the m-PCR assay. The optimal number of m-PCR cycles was 25, which is considerably shorter the normally required 30-35 cycles (Fig. 3).

M-PCR analysis of experimentally or naturally infected tissue samples
As shown by the results of experimentally infected tissue samples in Table 4, all pathogens can be detected in the liver samples, while Escherichia coli, Pasteurella multocida and Salmonella spp. were detected in the blood samples, and all, except for Proteus mirabilis, were detected in the kidney samples.
The results indicated that the assay can provide specific detection of six pathogenic bacteria in experimentally or naturally infected tissue samples.
Primer specificity is a critical determinant of the success of an m-PCR assay. In this study, an m-PCR assay was developed to target specific genes of six pathogens (invA, phoA, KMT1, toxA, ureR, and nuc) based on the following previous studies. Rahn et al. (1992) reported the use of the invA gene for specific detection of Salmonella spp. Thong et al. (2011) established an m-PCR assay for detection of the phoA gene of Escherichia coli. Townsend et al. (1998) and Blackall and Miflin (2000) developed PCR assays for identification of the KMT1 gene of Pasteurella multocida. Song et al. (2000) designed specific primers for the rapid identification of the Pseudomonas aeruginosa toxA gene. The Proteus mirabilis ureR gene was identified as a transcriptional regulator of the urease enzyme (Nicholson et al. 1993) and has been used as target gene for the detection of Proteus mirabilis by PCR (Huang et al. 1999). Brakstad and Maeland (1995) established a method for the direct identification of the Staphylococcus aureus nuc gene.
Further, to test cross reaction by agents that could be found as secondary infection with avian parasites (cryptosporidium and coccidia) and viruses (NDV, IBDV and AIV) showed that the proposed m-PCR assay had very high specificity. Moreover, the specificity also was tested by the different major serotype of the avian pathogenic Escherichia coli and Salmonella spp. which were the most important bacterial pathogens of poultry according to the clinical isolation samples. In addition, in poultry infections, the major serotype of the other bacteria such as Pasteurella multocida, Proteus mirabilis, Pseudomonas aeruginosa and Staphylococcus aureus is single.
In most cases, the sensitivity of an m-PCR assay will be reduced with increased numbers of target genes in the system. However, the detection limit of the proposed m-PCR assay was 2.8-8.6 × 10 3 CFU of each bacterial species, which is in agreement with the results of previous studies. For example, the sensitivity of Escherichia coli detection with the proposed m-PCR assay was 10 3 CFU/mL, which was superior to that reported by Kong et al. (2002) of 10 4 CFU/mL. The detection limits of Proteus mirabilis and Pasteurella multocida were 8.6 × 10 3 and 2.8 × 10 3 CFU/mL, which were the same orders of magnitude as those reported by Huang et al. (1999), Takeuchi et al. (1996). On the contrary, previously reported PCR assays for the detection of these six pathogens were single, triple, or quadruple methods. In comparison, the PCR detection sensitivities established in this study were close to or even exceeded those of the cited PCR assays. Moreover, as compared with traditional detection methods, the six pathogens tested in this study can be detected at one time with high sensitivity, thereby greatly reducing the detection time, while improving the efficiency.
For optimization of the m-PCR assay, the concentrations of primers, Taq DNA polymerase, and dNTPs, as well as the addition of a PCR additive, were optimized in this study. In pre-experiments, the concentrations of the first primer pairs for Proteus mirabilis and Salmonella spp. were set at 0.4 µM. For the following orthogonal experiments, a third pair of primers was added and the concentration was adjusted from 0.1 to 0.8 µM until the specificity was judged as appropriate.
dNTPs are raw materials for the synthesis of target fragments. Hence, to synthesize larger target fragments, more dNTPs are consumed. In this study, the target fragments were all within 1000 bp, but six were synthesized. Therefore, under consideration of cost, we recommend a dNTP concentration of 0.25 mM to amplify the corresponding target fragments.
Betaine is widely used as an enhancer to optimize various PCR assays. For example, Marshall et al. (2015) used betaine to enhance the formation of long PCR products and Henke et al. (1997) reported that betaine improved the amplification of genes by reducing the formation of secondary structures caused by GC-rich regions. As compared to dithiothreitol and dimethyl sulfoxide, betaine had the best PCR enhancing properties at a concentration of 0.8 M for all primer pairs and was more effective since the PCR output was enhanced for all of the target fragments (Hengen 1997;Kang et al. 2005;Lajin et al. 2013). The major limitation of detection is a low quantity of the template. In this study, the addition of 0.4 M betaine improved the sensitivity of the PCR assay so that the detection limit of the sextuple PCR assay was similar to that of a single assay.
In addition, to determine whether the m-PCR assay was appropriate for the detection of pathogens in clinical and laboratory samples, 7-day-old chicks were inoculated with the six tested pathogens. Then the pathogens were enriched from the tissues and organs of chicks for detection by m-PCR. The results showed that all six pathogens were detected in the liver samples with the proposed m-PCR assay. Besides, more than 150 bacterial strains were identified and isolated from diseased chicken by the assay. The results indicated that the assay also can be used in clinical investigation.
In conclusion, a rapid diagnostic m-PCR assay was established for the detection of six pathogenic bacteria in a short time. Moreover, this method can effectively and rapidly detect most pathogenic bacterial infections in poultry with good specificity, accuracy, and sensitivity.