Prevalence and antimicrobial susceptibilities of bacterial pathogens in Chinese pig farms from 2013 to 2017

Bacterial diseases of swine are a kind of multifactorial and uncontrollable diseases that commonly exist in pig farms all over the world and will lead to huge economic losses every year. In this study, a detailed and overall survey was carried out to better understand the prevalence and antimicrobial susceptibilities of bacterial diseases from 2013 to 2017 in China. A total of 19673 bacterial strains were isolated from 44175 samples collected from 9661 pig farms that distributed in 16 Chinese major pig breeding provinces. The results showed that the average isolation rates of Streptococcus suis (SS), Haemophilus parasuis (HPS), Escherichia coli (E. coli), Pasteurella multocida (Pm), Actinobacillus pleuropneumoniae (APP), Brodetella bronchiseptica (Bb), Salmonella enteria (SE), Erysipelothrix rhusiopathiae (E. rhusiopathiae) were 16.9%, 9.7%, 6.3%, 3.4%, 0.3%, 1.5%, 2.3% and 0.9%, respectively. The isolate rates of E. coli, APP and SE showed an increasing trend from 2013 to 2017. The seasonal prevalence characteristics of SS, HPS and Pm were obviously higher from April to August for first two bacteria and higher at February, March, April, and October for Pm. The dominant serotypes for SS, HPS were serotype 2 and serotype 5 (changed from serotype 4), respectively. The SS, HPS, and Pm showed very high antibiotic resistance rates to almost 8 common antibiotics (β-lactam, aminoglycoside, macrolides, lincomycin, tetracycline, quinolone, polymyxin, and sulfonamide) and an obvious increasing trend of antibiotic resistance rates from 2013 to 2017. In conclusion, the study provides detailed information on the prevalence and antimicrobial susceptibilities of different bacterial pathogens of swine from 2013 to 2017 in China. These data can provide a foundation for monitoring epidemiological patterns of bacterial diseases in the Chinese swine herds, as well as provide insight into potential antibiotic resistance profiles in these pathogens.


Materials and Methods
samples collection. All pig tissue samples were collected from 16 Chinese major pig breeding provinces from 2013 to 2017 and delivered to the Animal Disease Diagnostic Center of Huazhong Agricultural University for identification of bacterial pathogens, which is a reference lab in China where people send pig samples for diagnosis (Fig. 1). All samples that were sent to our reference lab with a total of 44175 samples from 9661 pig farms were collected and chosen for bacterial isolation and identification ( Table 1). The pig farms had a wide variety of management types and herd size with small-, medium-and large-scale commercial pig farms as well as various types of backyard farms. The tissues came from suspected sick pigs and were transported to our reference lab. The collected samples included lung, heart, spleen, joint fluid, intestine, brain, liver, trachea, effusion and so on (Fig. 2). Then, sterile operation had been taken to avoid cross contamination and all samples were processed for bacterial isolation immediately. The research was approved by the Ethics Committee of the Faculty of Veterinary Medicine of the Huazhong Agricultural University. All procedures regarding the animal care and testing were carried out according to the recommendation of Hubei provincial public service facilities.  www.nature.com/scientificreports www.nature.com/scientificreports/ Culture conditions, primers, and identification methods. Common bacterial pathogens were isolated and identified from Chinese pig farms, such as SS, HPS, Pm, E. coli, APP, Bb, SE, E. rhusiopathiae, and others. Tryptic Soy Broth (TSB), Tryptic Soy Agar (TSA) (Difco Laboratories, Detroit, USA), Shigella and Salmonella Agar and MacConkey Agar (HuanKai Microbial, Guangdong, China) medium were used and 10 μg/mL of nicotinamide adenine dinucleotide (NAD) and 5% (v/v) inactivated cattle serum (Zhejiang Tianhang Biotechnology, Zhejiang, China) were added if necessary. Primers used for identification and serotyping were listed in Table 2  and Supplementary Table 1, respectively.
the seasonal prevalence characteristics of ss, Hps, and pm. To investigate the seasonal prevalence characteristics of SS, HPS, and Pm, the number of samples and isolation rates of every month were calculated when samples were collected from 2013 to 2017. Results showed that the rates of isolation across the whole year of SS ranged from 13.6% to 23.5% with the highest isolation rate recorded in June (Fig. 4). While the monthly www.nature.com/scientificreports www.nature.com/scientificreports/ isolation rates of HPS ranged from 7.7% to 11.7% with the highest isolation rate recorded in July. By comparing the seasonal prevalence characteristics of SS and HPS, a similar phenomenon was found that they all had higher isolation rates from April to August. However, Pm showed higher isolation rates in February, March, April, and October ranging from 2.6% to 5.2% within a whole year (Fig. 5).  www.nature.com/scientificreports www.nature.com/scientificreports/ the serotype of ss and Hps. To better understand the most common bacterial pathogens in Chinese pig farms, a part of isolated SS and HPS were chosen to do serotyping experiments every year. The results showed that the major SS existed in Chinese pig farms were serotype 2, 9, 7, 3, 1, 5 from 2013 to 2017 and the highest isolated rate of SS was still serotype 2, but it declined sharply from 2013 (52.4%) to 2017 (33.6%). However, there were several other serotypes of SS increased over time. Such as serotype 7 (from 7.3% to 19.5%) and serotype 9 (from 12.1% to 25.8%) (Fig. 6A). Meanwhile, the major HPS existed in Chinese pig farms were serotype 4, 5, 13, 2, 1, 14, 10 and 12 from 2013 to 2017. Nevertheless, the proportion of different serotype of HPS changed obviously and the biggest change of HPS was serotype 4 (from 30.8% to 20.8%) and 5 (from 12.1% to 31.8%) which were the most popular serotypes of HPS in 2013 and 2017, respectively. At the same time, the proportion of HPS serotype 1 (from 7.9% to 1.3%), 2 (from 9.1% to 5.0%) and 10 (from 4.4% to 1.3%) declined, serotype 13 (from 10.9% to 17.5%) increased and serotype 12 and 14 did not show any change trends (Fig. 6B).

Antibiotic resistance rates of SS, HPS, and Pm. A total of 8 kinds of antibiotics have been chosen to
do antimicrobial susceptibility tests including β-lactam (cephradine, ceftriaxone, amoxicillin and ampicillin), aminoglycoside (streptomycin, gentamicin, spectinomycin, kanamycin, amikacin and neomycin), macrolides (spiramycin and azithromycin), lincomycin (lincomycin and clindamycin), tetracycline (doxycycline), quinolone (ofloxacin, ciprofloxacin and enrofloxacin), polymyxin (polymyxin B) and sulfonamide (trimethoprim) antibiotics. Antibiotic resistance rates of SS, HPS, and Pm were presented in Table 3. Results showed that antibiotic resistance rates of SS to aminoglycoside, macrolides, lincomycin, tetracycline polymyxin, and sulfonamide antibiotics were all over 60% and it exhibited increasing antibiotic resistance rates in β-lactam (from 11.6% to 19.7%) and quinolone (from 34.7% to 49.9%) antibiotics (Fig. 7A). Meanwhile, the antibiotic resistance rates of HPS and Pm to different antibiotics are obviously lower than SS. But, HPS showed significant increasing antibiotic resistance rates to all kinds of antibiotics except sulfonamide which showed a strong antibiotic resistance rate to HPS (Fig. 7B). Just like HPS, Pm also had increasing antibiotic resistance rates to all kinds of antibiotics except lincomycin and quinolone antibiotics (Fig. 7C).

Discussion
Based on samples collection information of geographical distribution, tissue sources and bacterial isolation rates, results had been concluded that SS, HPS, E. coli, Pm, APP, Bb, SE, and E. rhusiopathiae were ubiquitous in almost all Chinese pig farms (Figs 1 and 2). From the previous reports, we know that SS, HPS, Pm, APP, and Bb are the major bacteria which seriously influence the respiratory system of pigs 26 , E. coli and SE are the major pathogens of digestive system 4,5 , and E. rhusiopathiae will lead to endocarditis, cutaneous necrosis and arthritis 7 . They are all important pathogens that influence pig growth and productivity, and will lead to huge economic losses to the pig industry 27,28 . Though significant progresses have been made in the last few decades in reducing the prevalence of animal diseases, there is still an increasing concern over the economic losses associated with diseases that cause   www.nature.com/scientificreports www.nature.com/scientificreports/ a reduction in production efficiency. Therefore, it is very important to know the prevalence characteristics and geographical distribution of these bacteria to control bacterial diseases efficiently.
This study fully evaluates the occurrence of common bacterial pathogens in Chinese pig farms. The results showed that the proportion of SS was highest among all isolated bacteria, followed by HPS, E. coli, Pm, APP, Bb, SE, and E. rhusiopathiae. Meanwhile, the order of the isolation rates of these bacterial pathogens was the same as the proportion of isolated bacteria (Fig. 3). Compared with the previous reports, higher isolation rates of SS, HPS, E. coli, and Pm were observed from pigs with respiratory diseases in agreement with some previous studies [29][30][31] . Whereas, lower isolation rates of Bb, APP, and E. rhusiopathiae were obtained 20,29,32 (Fig. 4). On the other hand, although Pm and Bb did not show any obvious epidemic characteristics from 2013 to 2017, the changing trend of their isolation rates was almost the same which may associate with the characteristic that they always exist side by side and they are all the common pathogens of swine atrophic rhinitis 33 . for antimicrobial susceptibility tests included β-lactam (cephradine, ceftriaxone, amoxicillin and ampicillin), aminoglycoside (streptomycin, gentamicin, spectinomycin, kanamycin, amikacin and neomycin), macrolides (spiramycin and azithromycin), lincomycin (lincomycin and clindamycin), tetracycline (doxycycline), quinolone (ofloxacin, ciprofloxacin and enrofloxacin), polymyxin (polymyxin B), and sulfonamide (trimethoprim) antibiotics. As a whole, the differences in resistance rates of SS, HPS, and Pm to all antibiotics from 2013 to 2017 were significant (p < 0.05), except resistance rates of SS to β-lactam and Pm to quinolone. Statistical analysis was performed using the χ2 test.
www.nature.com/scientificreports www.nature.com/scientificreports/ By analyzing the isolation rates of SS, HPS, and Pm among different months, the characteristics of their seasonal distribution were shown clearly. The isolation rates of SS and HPS were obviously higher in the warm season (from April to August) than the cold season because hot and wet climate is beneficial to the growth and transmission of SS and HPS 34 . However, the seasonal distribution of Pm was different with SS and HPS, Pm showed a significantly higher isolation rate from February to April (Fig. 5) which is almost the same with Bb as the report 20 . The reasons that lead to the phenomenon may be due to the discovery that Pm is always commonly existed with Bb which also has a high isolation rate from January to April 20,33 . On the other hand, because of cold and wet weather, the ventilation installations are always closed in pig farms to keep warm which contributes to the spread of Pm. Therefore, the climate is an important factor to monitor and control bacterial diseases.
It is reported that SS can be divided into at least 33 kinds of serotypes 35 and HPS can also be divided into 15 kinds of serotypes 36 . However, their pathogenicity is different between different serotypes and they lack of cross-protection 37,38 . So, it is very important to know the clinical serotypes of SS and HPS for controlling bacterial diseases. Our serotyping results showed that the major serotypes SS existed in China were 2, 9, 7, 3, 1, 5 (Fig. 6A). But, it was different with other reports, especially, the decline of serotypes 3, 4, 8 and 1/2 and the increase of serotypes 7 and 9, which suggested the changing trend of different serotypes of SS in China 39 . On the other hand, SS of serotype 2 declined sharply from 2013 (52.4%) to 2017 (33.6%) which may be associated with the use of SS vaccine of serotype 2 in Chinese pig farms (Fig. 6A). A similar changing trend also appeared between different serotype of HPS. Based on serotyping results, serovars 4, 5 and 13 were still the major serotype of HPS which were the same as the reports before in China, North America, and Spanish [40][41][42] . Nevertheless, the percentage of different serotype of HPS varied largely and the main serotype of HPS changed from serotype 4 in 2013 (30.8%) to serotype 5 in 2017 (31.8%) which were all the main serotype of HPS existed in Chinese pig farms. There was also an obvious change in other serotypes (Fig. 6B). The main reasons that lead to the difference in serotype of SS and HPS may be due to the difference of environment, antibiotic, and the use of bacterial vaccines. Therefore, it is very important to change the serotype of the inactivated vaccines in controlling these diseases.
Because of the good effect of antibiotics in promoting growth and preventing infection, large amounts of antimicrobial agents are still being used in modern swine production around the world. This, in turn, would facilitate the emergence and development of antimicrobial resistance. So, it is very important to monitor antibiotic resistance rates of clinical pathogens 43 . In the experiments, we analyzed antibiotic resistance rates of SS, HPS, and Pm from 2013 to 2017. The results indicated that SS, HPS, and Pm all showed similar and very high antibiotic resistance rates to 8 kinds of detected antibiotics with the reports 15,44,45 . Thereinto, the antibiotic resistance rates of SS to these 8 kinds of antibiotics were all over 60% except β-lactam and quinolone. Meanwhile, these pathogens all displayed a rapid increase in antibiotic resistance rates of the common used antibiotics in China from 2013 to 2017, which maybe indicate irregular and excessive use of antibiotics. Based on the previous reports, the ways that some bacteria produce antibiotic resistance to aminoglycoside, macrolides, lincomycin, tetracycline, polymyxin, and sulfonamide mainly due to obtainment of exogenous resistance genes. However, the main reasons that lead to antibiotic resistance to β-lactam and quinolone are because of the mutation of drug targets, which is obviously more difficult than the first one 46 . Therefore, the difference of resistance mechanism maybe an important reason that the resistance rates of aminoglycoside, macrolides, lincomycin, and sulfonamide is obviously higher than that to β-lactam and quinolone in SS, HPS, and Pm. In total, reasonable application of antibiotics has become an urgent issue for the control of bacterial diseases.

Conclusions
In summary, the identification and analysis results of 44175 collected samples showed that the main bacteria in Chinese pig farms were still SS, HPS, E. coli, and Pm. However, the isolation rates of different strains and serotypes of SS and HPS have obviously changed from 2013 to 2017. For example, SS, E. coli, APP, and SE displayed an increasing isolation rates. Nevertheless, HPS and E. rhusiopathiae displayed reversed results. The main serotypes of HPS had changed from serotype 4 to serotype 5, and the proportion of serotype 2 SS also decreased sharply from 2013 to 2017. Meanwhile, SS and HPS had an higher isolation rate in hot season. But, February, March, April, and October were the main seasons for the isolation of Pm. In addition, Antimicrobial susceptibility test indicated that SS, HPS, and Pm presented very high and increasing resistance rates to 8 kinds of common antibiotics. In conclusion, the study provides us very detailed information on the prevalence and antimicrobial susceptibilities of several main bacteria in China from 2013 to 2017, which help us to understand, prevent and control bacterial diseases of Chinese pig farms.