1. Introduction
Salmonellosis, caused by
Salmonella enterica (
S. enterica), is among the most frequently reported foodborne bacterial diseases [
1]. Contaminated poultry and its products are a major source of motile Salmonellae causing salmonellosis in human worldwide [
2,
3]. In South Korea,
Salmonella is the leading (23%) cause of bacterial foodborne poisoning. The annual production of chicken meat, the second-largest source of animal protein, was 957,000 metric tons in 2019, indicating an increase of about 1.6% since 2018 [
4,
5]. Among them, broilers represented 77% of slaughtered chickens in 2018. Contamination of poultry may occur throughout the broiler production chain, and the potential risk for contamination at each stage has been identified [
6].
Humans are likely to be exposed to antimicrobial-resistant (AMR)
Salmonella, which results from the use of antimicrobials in animals, through contaminated food, thus, leading in a health threat [
7]. In recent years, increases in the emergence and spread of AMR
Salmonella, particularly multidrug-resistant (MDR)
Salmonella, in humans and animals have been reported worldwide, making it a global challenge [
8,
9,
10]. Therefore, fluoroquinolones (FQs) and third generation cephalosporins (3GC) have become critically important for treating salmonellosis in humans [
11]. Thus,
Salmonella resistant to FQs and 3GC frequently arise in animals with easy dissemination across the food chain [
12]. The dissemination of AMR
Salmonella through the food chain, particularly through chickens, has important implications for the failure of salmonellosis treatment, thus, creating an increased risk to public health by the spread of AMR
Salmonella via chickens [
13].
It is essential to inhibit microorganisms in the broiler chicken supply chain to produce hygienic chicken meat. Epidemiological studies have clearly shown the transmission pathway of
S. enterica and its serovars associated with poultry and the challenges posed by their virulence and antimicrobial resistance profiles [
14]. Consequently, this pathogen has become one of the main targets for the implementation of control strategies along the poultry production chain. Breeding flocks, hatcheries, rearing farms, and slaughter plants are all recognized as critical focal points for managing the risk associated with salmonellae. A hatchery plays an important role in collecting hatching eggs from the upper breeder farm and selling newly hatched chicks to a commercial broiler farm. Some
Salmonella serovars can persist in hatcheries longer than others, probably due to their ability to form biofilms [
15]. Hazard Analysis and Critical Control Point (HACCP) has been applied to poultry farms (including broiler and breeder farms) and chicken slaughterhouses in Korea. However, it has not been applied to hatcheries, thus, warranting a systematic investigation and evaluation of hatchery hygiene, which has not yet been performed [
16].
Salmonella can be introduced into hatcheries by horizontal and vertical transmission routes. The newly hatched chicks are more susceptible to
Salmonella infection than older birds; as their intestinal flora and immune system are immature, they may become infected with a challenge of 10–100
Salmonella cells [
17]. A high prevalence of
Salmonella in one-day-old chickens from hatcheries has been previously reported [
18].
Salmonella contamination in a hatchery can produce poor-quality chicks, resulting in a decreased feed conversion rate, increased mortality, and poor flock uniformity [
16]. Moreover, the prevalence of
Salmonella in hatcheries is related to
Salmonella prevalence in derived meat products during processing [
19]. Prevention of
Salmonella contamination in chicken products requires detailed knowledge of the major sources of contamination. The critical role of a hatchery in disseminating
Salmonella to commercial broiler farms and possibly exposing breeder flocks to contamination on egg trays, trolleys, and vehicles has also been previously reported [
20,
21,
22]. Most of these works have focused on the potential for cross-contamination and infection caused by
Salmonella in chicks during incubation. Considering that most integrated companies show vertical integration in Korea with numerous potential sources of
Salmonella contaminants in this system,
Salmonella control in integrated broiler chicken operations is complicated [
23].
It is necessary to investigate the occurrence and antimicrobial resistance of
Salmonella in the poultry production chain as it may aid the optimization of HACCP strategies and reduce the incidence of salmonellosis in humans. In recent years, several reports describing
Salmonella prevalence in the integrated broiler supply chain in Korea have been published [
19,
24,
25,
26]. However, studies focusing on the dissemination or tracing of the AMR
Salmonella along an integrated broiler chicken operation are limited. The dissemination of AMR
Salmonella in the broiler farm, slaughterhouse, and its downstream retail markets has been previously described, possibly contributing to the original dissemination of AMR
Salmonella to retail chicken meat [
6,
27]. The high prevalence of AMR
Salmonella colonized in newly hatched chicks in broiler farms has emphasized upstream breeder farm and hatchery as the sources of AMR
Salmonella in broiler chickens. The purposes of the present study were to identify AMR
Salmonella occurrence in hatcheries and their upstream breeder farms and reveal the dissemination in a vertically integrated broiler chicken operation in South Korea.
4. Discussions
In the present study, the prevalence of
S. enterica (16.4%) was higher in hatcheries than in its upstream breeder farm (3.0%), despite fumigation being routinely used during hatching (
Table S2) [
16]. The prevalence of
Salmonella in hatcheries varied widely from operation to operation (6.78–44.9%) and may have been associated with differences in hygiene and sanitation levels of each operation and the different detection methods used in each study [
25,
36,
37]. Herein, the
Salmonella isolation rate from breeder farms was relatively lower (3.0%) than in previous studies in Korea (14.7–19.0%) and China (10.53–18.15%); however, even infected breeder flocks have been shown to cause widespread
Salmonella contamination [
18,
25]. The comparison of our
Salmonella isolation rates in cloacal swabs (1/125, 0.8%) and litter (5/75, 6.7%) samples from breeder farms with previous studies (0% and 40%, respectively) indicates that different sample types may also be a factor influencing the prevalence of
Salmonella [
24,
38].
In total,
S.
enterica ser. Albany was the dominant serovar in hatcheries (
Table 1). These results are consistent with our previous study on isolating AMR
Salmonella isolates from chicken slaughterhouses and retail markets [
6]. Similarly, studies conducted on the prevalence of
Salmonella in poultries in Vietnam, Malaysia, and Myanmar showed that the predominant serovar was
S.
enterica ser. Albany (34.1%, 35.4%, and 38%, respectively) [
39,
40,
41]. Contrary to our results,
S.
enterica ser. Hadar was the most frequently reported serovar in integrated broiler operations in Korea; the
Salmonella serotype most often isolated from the hatchery was
S.
enterica ser. Senftenberg [
16,
19,
25].
S.
enterica ser. Albany, one of the prominent serovars in poultry that has been infecting animals and humans for several decades, may be an emerging serotype in Korea in the future [
42,
43]. Consistent with our results,
S.
enterica ser. Montevideo,
S.
enterica ser. Senftenberg, and
S. enterica ser. Virchow were the most frequently reported serovars in the poultry industry in Korea [
25,
26,
44].
S. Enteritidis has typically been the most common serotype responsible for
Salmonella infections in poultry in Korea for several decades; however, this prevalence has decreased, and this serovar has been gradually replaced by other emerging serovars. Therefore, the predominant
Salmonella serovar varies from company to company and time to time.
Antimicrobial resistance of
Salmonella is a globally emerging problem of public health concern. In the present study, no susceptible isolate was found in any of the 42 isolates. Among the 16 antimicrobial agents tested, the highest resistance rate observed in the hatchery was to NAL (97.2%), followed by SXT (55.6%), AMP (52.8%), TET (50.0%), and CHL (50.0%), which is consistent with previous reports (
Table 2) [
6,
45]. Antimicrobial resistance was detected even in isolates from hatcheries that were not treated with antimicrobials. One potential explanation is that those AMR isolates came from upstream breeder farms. Quinolones, ampicillin, and tetracyclines have been widely used for therapy, prophylaxis, and growth promotion by breeders, while sulfonamides have been used in human and veterinary medicine for 40 years [
46,
47]. Another potential explanation is that the hatcheries are contaminated with AMR
Salmonella in the internal environment of the hatchery or the external natural environment [
48]. Contrary to the increasing incidence of FQ-resistant
Salmonella reported worldwide, we did not find FQ resistance in the present study. However, a marked resistance to NAL reported herein could be a matter of concern because NAL resistance has been associated with a decrease in susceptibility to FQs, which are used to treat salmonellosis in humans. The present study indicated that MDR
Salmonella contamination was more widespread in the hatchery (19/36, 52.8%) (
Table 2) than suggested by another report [
49]. Among MDR isolates, we observed resistance to 3GC and COL, which are critically important in treating salmonellosis in humans [
11]. For example, two
S.
enterica ser. Virchow isolates from breeder farms were resistant to XNL, with the antimicrobial resistance profiles NAL-NEO-STR-TET-AMP-XNL and NAL-NEO-TET-AMP-XNL (
Table 3); one isolate (untypable) from the hatchery was resistant to COL, with the antimicrobial resistance profile SXT-AMP-FOX-COL. Colistin, as an antimicrobial substance, was used against Gram-negative bacteria. The use of colistin has been limited due to systemic toxicity. However, it has been re-introduced as a last-line option in the treatment of human infections [
50]. The resistance can be transmitted to humans through the food chain. Eventually, it can lead to microbial cross-resistance and pose a threat to human health. Therefore, it is mandatory to monitor the dissemination of resistance to colistin [
51].
Although PMQR has been studied and increasingly reported, QRDR mutations seem to represent the main mechanism of quinolone resistance in animal isolates [
52]. Moreover, PMQR was commonly detected in
Enterobacteriaceae, particularly in
E. coli, and the prevalence of PMQRs in
Salmonella remains extremely low [
53]. This finding was consistent with our observations that the global level of FQ non-susceptibility may be mainly due to QRDR chromosomal mutations (
Table 4). Our results indicated that missense mutations frequently occurred in the QRDR of
gyrA and
parC, which are considered to be the major quinolone resistance determinants in
Salmonella [
54]. In the present study, we identified QRDR point mutations in the
gyrA and
parC genes in all selected
Salmonella isolates from hatcheries, a finding inconsistent with previous studies showing one-point mutation only in the
gyrA gene as the main pattern [
55,
56,
57]. The results imply that resistance to FQs is continuously evolving with time. The results also show that antimicrobial pressure, rather than the horizontal transmission of antimicrobial resistance genes in chickens, leads to the appearance of antimicrobial resistance; clonal dissemination seems to be a key contributing factor for increasing resistance to FQs among
Salmonella in hatcheries, where antimicrobials are not applicable [
58]. This finding indicates the potential risk that
Salmonella isolates with mutations in
gyrA and
parC could naturally be maintained during hatching, even with no antimicrobial pressure. The
Salmonella isolates with QRDR may be directly transmitted to the downstream broiler farms through their carrying by day-old chicks [
59]. Our results also showed that one resistance gene (
blaCTX-M-15), which encodes resistance to ESBL, was identified in
S.
enterica ser. Virchow, one of the most frequently identified serotypes in 2015–2016 [
6]. There was a potential risk of ESBL-producing
Salmonella isolates being transmitted to humans through contaminated poultry products [
60].
Currently, PFGE is an easy and effective method to assess relatedness among
Salmonella isolates from different sources [
6,
27]. The clonal relationship among isolates from hatcheries and their upstream breeder farms at the chromosome level was accessed using PFGE (
Figure 1). There was frequent
Salmonella cross-contamination among hatcheries and among hatcheries and upstream breeder farms. An identical PFGE type (M1) was shared between isolates from a hatchery and its upstream breeder farm isolates, suggesting that
Salmonella contamination in hatcheries could be achieved by a direct vertical top-down transmission [
36]. Isolates from different hatcheries shared the same PFGE types, indicating
Salmonella cross-transmission among the hatcheries, possibly due to the sharing of the same source of eggs or trucks within the same operation [
24].
To further determine and compare the genotypic relatedness of isolates from the integrated broiler chicken operation, selected isolates of the three most prevalent serotypes from breeder farms, hatcheries, broiler farms, slaughterhouses, and retail markets were analyzed by PFGE (
Figure S1). A highly consistent PFGE pattern (SM-7) from different sources revealed that the AMR
S. enterica clones could disseminate through the broiler chicken supply chain (
Table 6 and
Figure S2). The SM-7, not the main PFGE type in hatcheries, could be disseminated to the downstream stage (broiler farm) even throughout the broiler supply chain, suggesting that
S.
enterica ser. Montevideo could persist and is difficult to eliminate from the environment, probably due to its biofilm-producing ability [
61,
62]. Herein, 52% of the
S.
enterica ser. Montevideo and 29.6% of the
S.
enterica ser. Albany isolates from the downstream of the hatchery carried the same PFGE types as those of the hatchery, indicating that
Salmonella contamination in hatcheries was an important source of
Salmonella contamination in the integrated broiler chicken operation (
Table 6 and
Figure S2). From
Table S4, it can be concluded that biocontrol of
Salmonella contamination in hatcheries is important; and control of
S. enterica ser. Montevideo isolates from hatcheries are more necessary.
The routes of
Salmonella cross-contamination within the hatchery and between the hatchery and its upstream breeder farm were complicated.
Salmonella was detected in the hatchery and its upstream breeder farm (represented by
S.
enterica ser. Montevideo) (
Table 5). Two PFGE types (SM-3 and SM-7) were consistent with upstream breeder farms, indicating that vertical transfer of infection from breeding birds to progeny is an important aspect of the epidemiology of
S. enterica infection within the poultry industry [
24,
63]. More importantly, emerging serotypes and genotypes were detected in the hatchery. The hatchery samples were contaminated with non-original
Salmonella with a serotype different from those in the breeder farm (represented by
S.
enterica ser. Albany); the hatchery samples were contaminated with non-original
Salmonella with PFGE types different from those in the breeder farm (represented by SM-11, SM-12, SM-13, and SM-14).
Salmonella was absent in upstream breeder farms but present in the hatchery, suggesting that contamination can happen during hatching. Therefore, apart from the original
Salmonella contamination in upstream breeder farms, there is at least one more route of
Salmonella contamination in the hatchery. There are many ways in which
Salmonella can enter these extensive and integrated operations and be recirculated and amplified by various routes [
26]. In some cases, clonal horizontal transmission in the hatchery and on the farm during the rearing period is of greater importance and leads to the isolation of a greater variety of
Salmonella serovars [
64]. For instance, several risk factors for horizontal transmission have been identified, such as inadequate cleaning and disinfection of hatching houses, which leads to contamination of the downstream hatching eggs and a poor level of hygiene [
64]. Usually,
Salmonella infection does not interfere with hatchability, but during hatching, the organisms are widely spread in the hatcher via ducts and the fluff that is disseminated by forced ventilation [
63,
65]. The higher prevalence of
Salmonella in a hatchery than in a relatively clean place indicates that intervention strategies must target this stage to prevent
Salmonella from entering the downstream broiler farm. This approach requires that
Salmonella should be detected quickly and accurately at the hatching stage before entering the broiler farm. Our study had the limitation that we did not provide direct evidence to prove the horizontal transmission of
Salmonella in the hatchery. Further studies focusing on the investigation of the horizontal transmission routes of
Salmonella in hatcheries are needed.