The epidemiological investigation of co‐infection of major respiratory bacteria with pseudorabies virus in intensive pig farms in China

Abstract Porcine respiratory disease complex (PRDC), a respiratory disease caused by a variety of factors, is one of the most common problems in the intensive pig farms. To investigate the mixed infection incidence of wild‐type pseudorabies virus (WT PRV) and respiratory bacteria, a total of 1,293 clinical samples were collected from pigs with typical respiratory signs from 14 different provinces of China from September 2016 to February 2018. The WT PRV was detected by ELISA targeting gE antibody while the bacteria were detected by bacterial isolation and serotyping by PCR. The results revealed that the detection rate of A. pleuropneumoniae and B. bronchiseptica infection associated with WT PRV infection were 6.30% and 15.99%, respectively, which were significantly higher than those without WT PRV infection (3.41% and 4.41%) at the farm level (p < .05). There were no significant differences in the detection rate of H. parasuis, S. suis or P. multocida between WT PRV positive and negative farms (p > .05). However, the detection rate of attenuated H. parasuis and S. suis strains were 68.19% and 64.75%, respectively, in WT PRV infected farms, which were significantly higher than those (41.56% and 52.25%) in WT PRV free farms (p < .05). The prevalent serotypes of H. parasuis‐5/12 and S. suis‐2 were also investigated by multiplex PCR. These results indicated that the presence of WT PRV increased the chance of bacterial infection and the number of pathogenic strains in the respiratory system of pigs. Therefore, the eradication of pseudorabies is an effective approach to prevent and control the bacterial respiratory diseases in the intensive pig farms in China.

Traditionally, PRV is an etiological agent causing reproductive failure in sows, nervous disorder in nursery and growing pigs, respiratory problem in growing and finishing pigs. PRV can inhibit the synthesis of chemokines (Viejo-Borbolla, Ana, Enrique, & Alcamí, 2009), the transcription of interferon (Brukman & Enquist, 2006), the expression of MHC I molecules by shutting off host protein synthesis (Mellencamp, O'Brien, & Stevenson, 1991), and subsequently cause immune-suppression (Chinsakchai & Molitor, 1992). It was noted that PRV infection could increase the severity of bacterial pneumonia (Opriessnig et al., 2011). Lesions, such as polyarthritis and fibrinous pericarditis, are more abundant and acute in pigs with mixed challenge exposure, compared with pigs infected only with S. suis (Iglesias, Pijoan, & Molitor, 1992); PRV infection can allow H. parasuis to proliferate in the lung by destroying the respiratory epithelial cells of pigs (Narita, Kawashima, Matsuura, Uchimura, & Miura, 1994). The clinical symptoms of A. pleuropneumoniae became more severe with concomitant infection with PRV (Sakano et al., 1993); PRV and P. multocida mixed infection also produce more severe pneumonia than P. multocida infection alone, and lead to a significant decrease in the average daily weight gain (Fuentes & Pijoan, 1987). However, since 2011, outbreaks of PRV caused by novel variant strains has been documented in lots of swine farms in China, causing serious economic losses to the swine industry. Therefore, it is urgent to understand the co-infection status of PRV and main bacteria in PRDC in the pig farms.
In this study, bacterial isolation and serotyping were performed from WT PRV negative and positive pigs, and then we analysed the impact of WT PRV infection on bacterial respiratory diseases in intensive pig farms in China. It might pave the way to control bacterial diseases of the porcine respiratory system to be more precisely and efficiently.

| Sample collection
To assess the mixed infection of WT PRV and respiratory bacteria in intensive pig farms (≥ 1,000 pigs), a total of 1,293 clinical samples were collected from pigs with typical respiratory signs from 14 different provinces, such as Hubei, Henan, Hunan, Hebei and others from September 2016 to February 2018. The clinical samples were collected from suspected pigs and shifted to our diagnostic laboratory.
The collected samples included nasal swabs (total of 574), lungs (total of 334), spleens (total of 105), joint fluids (total of 89), brains (total of 110), tracheal fluids (total of 81). Then, under complete sterile measures to avoid cross-contamination, bacterial isolation was performed immediately. Blood samples were collected from the jugular vein of pigs and kept in 5 ml blood-collecting tubes without anticoagulant.
The research was approved by the Ethics Committee of the Faculty of Veterinary Medicine of the Huazhong Agricultural University. All procedures regarding animal care and testing were carried out according to the recommendation of Hubei provincial public service facilities.

| Serological detection of gE antibody against WT PRV
Commercially available PRV/AD gE Ab ELISA kit with sensitivity and specificity 96.7% and 99.8%, respectively (IDEXX, USA) was used to detect gE antibody, which differentiates between vaccinated and infected pigs. In this study, a herd was considered to be a positive herd if at least one WT PRV positive sow was detected. Otherwise, the herd was considered to be a negative herd. Gram-staining characteristics and oxidase (Gram-negative bacilli)

| PCR primer sequences
According to sequences published in previous literatures, primers were synthesized by Sangon Biotech Co., Ltd (Shanghai). The primers for amplifying target genes of S. suis, H. parasuis, P. multocida, B.
bronchiseptica and A. pleuropneumoniae are listed in Table 1.

| Serotype identification of S. suis and H. parasuis
According to the reports, S. suis can be classified into 33 serotypes based on the difference of capsular polysaccharide (Liu et al., 2013), and H. parasuis can be classified into 14 serotypes based on the difference of capsular loci (Howell et al., 2015). S. suis and H. parasuis strains were randomly chosen for further serotyping by typing PCR according to the previously described methods.

| Statistical analysis
At the farm level and the individual sample level, all research data were analysed to identify the statistical differences of bacterial respiratory diseases between the WT PRV free or positive farms.
To avoid the confusion from the presence of maternal derived gE antibody, only data from breeding and fattening pigs were used.
Statistical analyses were undertaken with SAS version 9.0 (SAS Institute Inc.). Univariate association between variables and isolation rates of different bacteria were determined by using univariate ordinary logistic regression analysis and Chi-square test. p < .05 and p < .01 were considered to be significant and highly significant, respectively.

| Bacterial test results of various samples
The results showed that, among the tested bacteria, the detection rate of S. suis was the highest in all types of samples, and for S. suis, the nasal swabs were the most suitable samples with the highest detection rate. The detection rate of H. parasuis, P. multocida, B. bronchiseptica and A. pleuropneumoniae was high in tracheal fluid samples. It can be noticed that the highest rate of respiratory bacteria could be observed in the tracheal fluid samples while the lowest rate in the joint fluid samples. Hence, the most suitable sample for respiratory bacterial isolation is the tracheal fluid samples. The detailed bacterial detection rates are shown in Table 2.

| WT PRV gE antibody test results
Together with tissue samples, a total of 1,293 serum samples from 94 intensive pig farms were tested for presence of WT PRV infection by gE-ELISA, of which 499 and 794 samples were from 45 and 49 WT PRV negative or positive pig farms, as shown in Table 3. The detection rates of gE antibody were further categorized corresponding to the growth stage of pigs from which sera were collected. In this study, only fattening and breeding pigs were selected to eliminate the effect of maternal derived antibodies,

| Detection rate of respiratory bacteria at WT PRV negative and positive farms "at the farm level"
To analyse the difference in bacterial respiratory diseases between the WT PRV positive and negative farms, the detection rate of different respiratory bacteria was listed and compared ( Figure 1).
There was no significant differences in the detection rate of H. parasuis, S. suis and P. multocida between WT PRV negative and positive farms. However, the detection rate of B. bronchiseptica and A. pleuropneumoniae were significantly higher in WT PRV positive farms than in PRV negative farms (p < .01).

| Detection rate of respiratory bacterial infection with and without co-infection with WT PRV "at the individual level"
To

| The distribution of S. suis serotypes in WT PRV positive or negative pigs
Totally, 317 S. suis were identified from the samples from finishing and breeding pigs. From them, 139 and 178 strains were isolated from WT PRV positive and negative pigs, respectively.
As shown in Table 5, the S. suis serotypes 1, 2, 7 and 9 were regarded as virulent serotypes. The detection rate of avirulent strains was 52.25% in PRV non-infected pigs, which was also significantly lower than that in PRV infected pigs (64.75%) (p < .05).

| The prevalent serotyes of respiratory diseaserelated bacteria
In this study, the prevalent serotype of S. suis and H. parasuis were Notably, there is also high proportion of isolation rate of non-typeable serotypes in both S. suis and H. parasuis.
Traditionally, the viral infection may plan a key role for secondary infection. Through serotyping, it was found that there was no difference in detection rates for virulent serotypes of the tested bacteria from samples collected from both WT PRV positive or negative

| The level of biosecurity and husbandry management
Although this study only considers the relationship between pseudorabies and bacteria, CSF, PRRS and PCV2 also cause immune-suppression and may induce similar results. It has been reported that PRRSV accelerates S. suis or H. parasuis infection in vivo and in vitro, and also causes more severe respiratory symptoms (Yu et al., 2012;Huong et al., 2016;Li et al., 2017Li et al., , 2018 among pigs with clinical symptoms of respiratory diseases, the highest percentage of PCV2 infection was with P. multocida in all cases (Kim et al., 2003), pigs infected with PCV2 and PRRSV suffer from severe immune-suppression, so it's easy for occurrence of secondary bacterial infection (Chang, Peng, Chang, Chaung, & Chung, 2008;Opriessnig et al., 2011;Wang et al., 2019).
Also, the poor biosecurity and husbandry can cause more bacterial infection. Moreover, it can also cause outbreaks such as PR, CSF, PRRS and PCV2. Therefore, for better control of bacterial respiratory diseases, the first aim is to improve the level of biosecurity and husbandry management. Then carrying out the prevention and control of PR, CSF, PRRS, PCV2 and bacterial respiratory diseases can achieve better results in pig farms.

| CON CLUS ION
The prevalent serotypes of H. parasuis and S. suis were H. parasuis-5/12 and S. suis-2, accounting for 24.79% and 19.87%, respectively. When pigs were infected with WT PRV, it's more likely

CO N FLI C T S O F I NTE R E S T
We have no other conflicts to declare.