1. Introduction
Salmonella is Gram-negative pathogenic intestinal species of bacteria of the Enterobacteriaceae family. The World Health Organization (WHO) has estimated that
Salmonella is responsible for 550 million infections among humans each year. In the USA population, approximately 1.2 million infections are caused by
Salmonella and the economic loss owing to salmonellosis is estimated to be
$3.5 billion annually [
1]. Despite the lack of official surveillance data on
Salmonella in China, it is estimated that it causes 75% of foodborne diseases [
2].
More than 2500 serovars have been identified, and these serovars are classified based on slide agglutination with O and H antigen-specific sera. Some serovars are host specific, but others can infect a variety of hosts. Pigs are considered to be one of the major environmental reservoirs of
Salmonella [
3].
Salmonella, especially
S. enterica, is one of the most common pathogens to cause diarrhea in pigs.
S. enterica can infect pigs of all ages, especially weaned pigs and young growing-finishing pigs [
4]. These pigs can be susceptible to salmonellosis as a result of their immature immune systems, the existence of concurrent diseases, and the stress that they undergo, including the stress due to weaning, transportation, and experiencing new environments. The rate of
S. enterica infection among weaned and finishing pigs is high, and salmonellosis among pigs can cause reduced weight gain and even death [
5]. Therefore, it is critical for the swine industry to implement strategies to reduce horizontal transmission and infection with
Salmonella.
Antimicrobials are important to prevent and treat bacterial infections. However, due to their misuse, drug resistance is increasing and multidrug-resistant strains have developed.
Fluoroquinolone are the drugs of choice to treat complicated cases of life-threatening salmonellosis [
6]. The emergence of fluoroquinolone resistant
Salmonella strains in several serovars poses a serious threat to public health, and the resistance rate to fluoroquinolone of
Salmonella isolates has increased dramatically around the world [
7]. The mutations of plasmid-mediated quinolone resistance (PMQR) and quinolone resistance-determining regions (QRDRs) mechanism contribute to the resistance of quinolone and fluoroquinolone antimicrobials [
8].
Colistin is being intensively used as a last-resort antimicrobial drug for treating infections with carbapenemase-producing
Enterobacteriaceae [
9]. In veterinary medicine, colistin mainly to treat Gram-negative infections of the intestinal tract, and used as a growth promoter in some countries [
10]. In 2015, the emergence of a plasmid-mediated colistin resistance mechanism involving the
mcr-1 gene was firstly reported in
Escherichia coli, and caught world-wide attention [
9].
Salmonella 4,[5],12:i:-, a monophasic variant of
Salmonella Typhimurium that lacks the second-phase flagellar antigen, was first reported in the 1980s [
11]. There has seen a rapid worldwide emergence of
S. 4,[5],12:i:- in the past two decades.
S. 4,[5],12:i:- is one of the most common serovars isolated from foodstuffs and animals in Europe, and now ranks as the top serovar among both human and veterinary isolates in some European countries [
12]. There are two major clonal lines was identified, “Spanish clone” and “European clone” [
13].
S. 4,[5],12:i:- is strongly associated with pig and pork products [
14].
China has the largest number of live pigs in the world (
http://www.fao.org/faostat/en/#data). However, there has been no specific information on the
Salmonella strains in pigs in China. In spite of many studies have been showed epidemiological and phenotypic features of
S. 4,[5],12:i:-, little is known about the ecological success of
S. 4,[5],12:i:- circulating in humans and animals in China. We investigated the frequency and distribution of
Salmonella in pigs isolated from 2014 to 2016 in various provinces in China. We also compared the antimicrobial resistance phenotypes, pathogenicity, heavy metal tolerance and genetic diversity between
S. Typhimurium and
S. 4,[5],12:i:- to investigate the possible factors that involved in the widespread of
S. 4,[5],12:i:- isolates.
3. Results
3.1. Prevalence and Distribution of Salmonella
Between 2014 and 2016, a total of 104
Salmonella isolates were obtained from the intestinal contents of pigs that died of diarrhea or from the feces of pigs with diarrhea. The geographical distribution of
Salmonella among the pigs is summarized in
Figure 1. The proportion of
Salmonella isolates was highest in Hebei (24.0%), followed by Shandong (20.2%), Hubei (12.5%), Henan (12.5%), Hunan (12.5%), Jilin (8.65%), Guangxi (4.81%) and Liaoning (4.81%). No significant difference was found in the prevalence of
Salmonella from different provinces (
p > 0.05).
3.2. Serotyping
We performed biochemical tests for each of the isolates. The serovars characterized by the biochemical tests were
Salmonella spp. and
S. enterica subsp. arizonae. These results were confirmed by serotyping. The H-antigens (phase 2) could not be detected in 56 (53.85%) isolates, but these isolates were identified as
S. Typhimurium based on the MLST analysis. As a result, the 104 isolates were assigned to the following six serovars (
Figure 1):
S. Agona (2, 1.92%),
S. Infantis (4, 3.85%),
S. Weltevreden (5, 4.81%),
S. Derby (8, 7.69%),
S. London (11, 10.6%),
S. Typhimurium (18, 17.3%) and
S. 4,[5],12:i:- (56, 53.9%).
3.3. Antimicrobial Resistance
Antimicrobial susceptibility testing of the 104 isolates showed that they were all sensitive to amikacin and meropenem, and resistant to trimethoprim-sulfamethoxazole (
Figure 2 and
Figure 3). Furthermore, all the isolates showed sensitivity or intermediate resistance to nitrofurantoin. Most of the resistance among the isolates involved trimethoprim-sulfamethoxazole (n = 104, 100%), ampicillin (n = 84, 80.8%) and tetracycline (n = 80, 76.9%). They also showed resistance to the antimicrobials used only in veterinary medicine, such as florfenicol (63, 60.6%), ceftiofur (48, 46.2%), enrofloxacin (15, 14.4%).
When antimicrobial resistance was analyzed in region, the prevalence of resistance to the major tested antimicrobials among the Salmonella isolates were different. The prevalence of resistance to the major tested antimicrobials among the Salmonella isolates from Jilin and Liaoning provinces was lower than in the other provinces (p < 0.05). S. Agona from Henan province were only resistant to trimethoprim-sulfamethoxazole, intermediate susceptibility to florfenicol and ceftiofur, and sensitive to the other 16 antimicrobials.
Salmonella isolates resistant to at least three different classes of antimicrobials were defined as multidrug resistant (MDR) isolates. Among the 104 isolates, 73 (70.2%) showed varying degrees of MDR, and Shandong was the most MDR province (
Table 1). Fortunately,
S. Agona and
S. Infantis do not exhibited MDR. The most common resistance profiles in
S. Typhimurium (11, 14.9%) were AMP-TE-SXT-CHL-FFC-OLA. Among 18
S. Typhimurium isolates, 12 isolates with the ASSuT genotype. Among 56
S. 4,[5],12:i:- isolates, the ASSuT genotype was observed in 34 isolates.
Among 104 Salmonella isolates, qnrB, qnrS, aac(6′)-Ib-cr, oqxA and oqxB were observed, aac(6′)-Ib-cr was the dominant PMQR gene (84, 80.8%), followed by qnrS (49, 47.1%), qnrB (32, 30.8%), oqxA (22, 21.2%) and oqxB (22, 21.2%). Phenotypic resistance to nalidixic acid, ciprofloxacin and enrofloxacin was found with the presence of qnrB, qnrS and aac(6′)-Ib-cr. The oqxAB gene was observed in S. 4,[5],12:i:-, and S. Typhimurium without oqxAB gene. All Salmonella isolates were tested for the presence of mcr-1 gene, mcr-1 gene was observed in 12 isolates. Isolates of Salmonella strains that carried mcr-1 genes only showed S. Typhimurium and S. 4,[5],12:i:-. All mcr-1-positive isolates were colistin-resistant with MIC of 8–16 µg/mL.
Similar antimicrobial resistance patters were observed in
S. Typhimurium and
S. 4,[5],12:i:-. Of 18
S. Typhimurium isolates, high antimicrobial resistance rate were found for ampicillin, tetracycline, and trimethoprim-sulfamethoxazole; and 18 of 18
S. Typhimurium were MDR isolates. The
aac(6′)-Ib-cr gene was the dominant PMQR genes (12, 66.7%), followed by
qnrS (10, 55.6%). Among 56
S. 4,[5],12:i:- isolates, high antimicrobial resistance rate was found for tetracycline, trimethoprim-sulfamethoxazole, and florfenicol; and 56 out of 56 were MDR isolates. The
aac(6′)-Ib-cr gene was the dominant PMQR genes (32, 57.1%), followed by
qnrS (25, 44.6%). No significant difference (
p > 0.05) has been showed in the presence of PMQR genes between
S. Typhimurium and
S. 4,[5],12:i:-. We determined the ciprofloxacin and enrofloxacin MIC for intracellular
S. Typhimurium and
S. 4,[5],12:i:- (
Figure 4). We found that the ciprofloxacin and enrofloxacin MIC for intracellular
Salmonella was significantly higher than extracellular
Salmonella. However, no significant difference (
p > 0.05) has been showed in the intracellular MIC between
S. Typhimurium and
S. 4,[5],12:i:-.
3.4. Molecular Typing
For the PFGE analysis, the 104 isolates were divided into two groups. The PFGE typing of the first 30 isolates using
XbaI led to their categorization into three clusters and ten PFGE subtypes (
Figure 2). Strains from the same farm that belonged to the same serovar were highly genetically related. However, the similarity coefficient of strains from the same farm that belonged to different serovars was low. The PFGE typing of the 74
S. Typhimurium and
S. 4,[5],12:i:- isolates by
XbaI led to their categorization into seven clusters (
Figure 3). The resultant dendrogram of
Salmonella isolates exhibited clusters with a high level of diversity. A part of the similarity coefficient of
Salmonella isolated from the same farms was lower. The similarity coefficients between the
Salmonella isolates were relatively low whether or not the isolates were from the same province or the same year. However, we also found that the similarity coefficient for the
Salmonella isolates from different provinces was >90%. Whether or not the isolates were from the same province, the
S. Typhimurium isolates were highly similar genetically to
S. 4,[5],12:i:-.
Based on the Salmonella MLST scheme associated with the University of Warwick MLST database, seven STs were identified among the 104 Salmonella isolates investigated, namely, ST34, ST32, ST19, ST155, ST40, ST13, and ST365 (Figure 4). The majority of isolates (67 isolates, 64.4%) were assigned to ST34. Eleven isolates (10.6%) belonged to ST155, eight (7.69%) to ST40, seven (6.73%) to ST19, five (4.81%) to ST365, four (3.85%) to ST32, and two (1.92%) to ST13. One S. Typhimurium and six S. 4,[5],12:i:- isolates belonged to ST19.
3.5. Pathogenicity Analysis
We tested in vitro the adhesion and internalization abilities of
S. Typhimurium and
S. 4,[5],12:i:- (
Figure 4). Several time points of adhesion and internalization were tested to evaluate the capacities of
Salmonella adhesion to IPEC-J2 cells and internalization into IPEC-J2 cells. At 3 h after
Salmonella challenge, we found that there is no significant difference (
p > 0.05) in the capacities of adhesion and internalization between
S. Typhimurium and
S. 4,[5],12:i:-. As a comparison, adhesion and internalization capacities of 6 h, 9 h, and 12 h were also assessed. The adhesion and internalization capacities of
S. Typhimurium were the same as
S. 4,[5],12:i:- at different time points.
All the S. Typhimurium were positive for 24 virulence genes tested. Among 56 S. 4,[5],12:i:- isolates, the pefA gene was the least prevalent of all the virulence genes (85.7%), followed by spvC, invA, avrA, sipA, fliC, misL and yafD. The sopE gene was observed only in S. 4,[5],12:i:- isolates (14, 25.0%).
3.6. Heavy Metal Tolerance
Among 74 Salmonella isolates, pcoA, pcoD, silA, silE and terF were observed in all isolates. The merA gene was only found in S. 4,[5],12:i:- isolates. No significant difference in the MIC for copper sulfate and zinc chloride between S. Typhimurium and S. 4,[5],12:i:- could be seen.
3.7. Phylogenetic Analysis
Single Nucleotide polymorphism (SNP) based on phylogeny using SO4698-09 as a reference genome, which is a representative multidrug European clone revealed 2211 high-quality shared SNP positions between each strain and this reference genome [
13]. A maximum likelihood tree was constructed using
S. Typhimurium,
S. 4,[5],12:i:- and the 28
S. 4,[5],12:i:- from Europe. As shown in
Figure 5, one main clusters were identified. Cluster 1 comprises SO4698-09 and 27 strains, none of which was ST19; this cluster included 2
S. 4,[5],12:i:-, 5
S. Typhimurium isolates, and 20
S. 4,[5],12:i:- from Europe.
S. 4,[5],12:i:- and
S. Typhimurium isolates were interspersed with
S. 4,[5],12:i:- isolates of Europe origin.
4. Discussion
In this study we identified 104 Salmonella isolates from pigs in 66 farms in 8 provinces of China (out of the 26 provinces investigated). In order of frequency, these isolates belonged to S. 4,[5],12:i:-, S. Typhimurium, S. London, S. Derby, S. Weltevreden, S. Infantis and S. Agona. Almost all the Salmonella isolates were resistant to a large majority antimicrobials. The PFGE and MLST results indicated that the Salmonella isolates from different regions were genetically diverse. S. 4,[5],12:i:- currently circulating in pig in China are most likely part of a European clade, which are mainly ST34 and harbors ASSuT genotype, heavy metal tolerance genes, and, to an extent, PMQR genes, mcr-1 gene, sopE gene.
The
Salmonella isolation rate for pigs assessed in this study is higher than in previous studies in Shandong province (7.23%) [
23], in central China (4.09%) [
24], in Sichuan province (16.4%) [
25], in the European Union (EU) (7.44%) [
26], and in the USA (8.79%) [
27]. The differences in these isolation rates might be caused by a variety of factors. Sample types (e.g., feces and intestinal contents), collection seasons and isolation methodologies can limit the comparison of different samples of
Salmonella isolates. Therefore, it is essential to establish a global standard protocol for the isolation of
Salmonella strains, especially for long-term monitoring of
Salmonella infections.
The
Salmonella serovars differed by region. In this study, the most commonly isolated serovar in pigs with diarrhea was
S. 4,[5],12:i:-. In contrast, in previous studies, the dominant serovar in pigs was
S. Derby in Shandong province [
23]; IIIb in central China [
24];
S. Adelaide in the USA [
27]; and
S. Typhimurium in the EU [
26]. The differences in the dominant serovars in pigs may be due to differences in geographical regions. According to the WHO Global Foodborne Infections Network (GFN) Country Databank (
http://www.who.int/gfn/activities/en/), the dominant serovar in foodstuffs in China was
S. Derby. However, the main serovar in foodstuffs in the USA was
S. Typhimurium, and it was
S. Infantis in Japan. Thus, it is clear that each country or region can have different serovars, and different diversities and complexities.
Antimicrobial resistance is one of the biggest threats to global health, food security, and development. The extensive use and misuse of antimicrobials in animals exert a selection pressure on bacteria such as
Salmonella, so the antimicrobial-resistant strains are more likely to survive. Consistent with other studies carried out in China [
23,
24,
25], the
Salmonella isolates were most likely to be resistant to trimethoprim-sulfamethoxazole and less likely (but still likely) to be resistant, to ampicillin, tetracycline, nalidixic acid, chloramphenicol and florfenicol. The high frequency of resistance to trimethoprim-sulfamethoxazole, tetracycline, and florfenicol may be related to the long-term use of these antimicrobials in China, and the incorporation of antimicrobials into animal feed so that they are present in animals at therapeutic levels in order to prevent bacterial disease.
In agreement with the findings of a previous study in Sichuan province,
S. Typhimurium were more likely to be resistant to four antimicrobials or more [
25]. When resistance profiles and serovars or geographical regions were compared, we found that there appeared to be a relationship between antimicrobial resistance and particular serovars or regions.
In general, the frequency of resistance of the Salmonella isolates to each of the 19 tested antimicrobials was lower in Jilin and Liaoning provinces than in the other provinces. The isolates in these two provinces were sensitive to all the antimicrobials tested except ampicillin, cefazolin, tetracycline, trimethoprim-sulfamethoxazole and ciprofloxacin. The frequencies of antimicrobial resistance in Guangxi province were similar to in Jilin and Liaoning provinces. However, in contrast to the other Salmonella isolates from Guangxi province, there were two isolates from this province that were resistant to ampicillin, amoxicillin-clavulanate, cefazolin, ceftiofur and gentamicin. This may be partly due to the use of different kinds and doses of antimicrobials for the treatment and prevention of diseases in pigs at different farms.
The antimicrobial resistance results for Shandong province and central China (Henan, Hubei, and Hunan provinces) were similar to those found in earlier studies in Shandong province [
23] and Henan, Hubei, and Hunan provinces [
24]. This similarity may be caused by the relatively similar usage in these areas (over time) of antimicrobials for disease control and prevention during the pig breeding process. Another potential factor is that the number of live pigs is significantly higher in central China than in other areas of China [
28]. As China is one of the major pig breeding countries, the demand for and use of antimicrobials are huge. Therefore, there is a higher risk of antimicrobial misuse. Measures should be taken to further regulate the use of antimicrobials in order to guarantee food safety and public health.
Consistent with the finding in previous studies, the
mcr-1-positive
Salmonella isolates were MDR strains, and it carry fluoroquinolones resistance genes [
29]. The
mcr-1-positive
Salmonella isolates from Henan and Hebei province carried
aac(6′)-Ib-cr and
oqxAB, respectively. All the
mcr-1-positive isolates were resistant to more antimicrobials than the
mcr-1-negative isolates were. Studies have shown that there is a correlation between the rises in antimicrobial resistance among pathogenic bacteria and the use of antimicrobials in animals feeds [
30]. To improve health and promote growth, colistin (together with other antimicrobials) was frequently used in feed. This increasingly heavy use of antimicrobials could have resulted in high selective pressure in environment and led to the emergence of MDR.
From January 1, 2006, the EU withdrew its approval for the use of antimicrobial growth promoters in animal feed. In 2012, the US Food and Drug Administration (FDA) banned the use of the cephalosporin class of antimicrobials in food-producing animals. Likewise, China has also taken measures to prevent the emergence of drug-resistant bacteria and promote food safety and public health. For instance, China issued an order prohibiting certain use of four fluoroquinolones (lomefloxacin, ofloxacin, norfloxacin and pefloxacin) in food-producing animals in 2015 and the use of colistin as a growth promoter in 2016. However, it is necessary to develop more strategies and draft more regulations to control the use of antimicrobials and prevent the emergence of drug-resistant bacteria.
In this study, in addition to S. Typhimurium 4,[5],12:i:1,2, a monophasic variant known as serovar 4,[5],12:i:- was detected, and this monophasic variant was more dominant than S. Typhimurium. The frequency of antimicrobial resistance genotypes was high among S. 4,[5],12:i:- isolates, and with the presence of an ASSuT genotype, typical of the European clone.
The
sopE gene was found in 14 of 56 of
S. 4,[5],12:i:-. The result was in agreement with the previous studies [
13]. The acquisition of the
sopE gene may affect the pathogenicity, but there is no significant difference between
S. Typhimurium and
S. 4,[5],12:i:- in adhesion and internalization assays. The expression of
sopE may increase the fitness of
Salmonella, and the transmission of
Salmonella by induce inflammatory diarrhea. Furthermore,
sopE expression results in increased amounts of salmonellae in the intestinal lumen and shedding in the feces, and increased production of host nitrate.
Not only encoded virtually
S. Typhimurium heavy metal tolerance genes but also
S. 4,[5],12:i:- did. Assembled data were searched for a genomic island of resistance to heavy metals (SGI-3) and all 7 strains encoded SGI-3. Supplements such as copper and zinc, are added in pig feed as micronutrients and general antimicrobials to increase feed efficiency and to promote animal growth. In the EU, heavy metals have been used increasingly since the ban of the nonspecific use of antimicrobials as growth promoters in pig feed [
31]. Heavy metals are very persistent in the environment [
32], and concentration of heavy metals in pig intestines may represent substantial selective pressure contributing to the success of
Salmonella.
The MLST analysis identified two S. Typhimurium clonal groups from different sources that differed only in terms of one of the profiled housekeeping genes. Furthermore, S. Typhimurium 4,[5],12:i:1,2 and 4,[5],12:i:- did not correspond to only one ST. The strains of ST34 were clustered into a single pattern. Therefore, we hypothesized that these strains persisted in specific regions and continually proliferated within these regions.
Currently,
S. Typhimurium 1,4,[5],12:i:- is one of the most common
Salmonella serovars isolated from humans and animals in many countries. For example, it was the third most common serovar responsible for human and animal infections in the EU [
26] and the fifth most common in USA [
27]. In China, since 2009,
S. 4,[5],12:i:- has been one of the four most common serovars to cause human salmonellosis [
33]. Pigs are considered one of the main vectors in the spread of
S. Typhimurium 1,4,[5],12:i:- [
34]. In order to better control salmonellosis, we should monitor the emergence, frequency, and variation of the various
Salmonella serovars.