Antimicrobial Resistance, Virulence Properties and Genetic Diversity of Salmonella Typhimurium Recovered from Domestic and Imported Seafood

The quantity of seafood imported and produced by domestic aquaculture farming has increased. Recently, it has been reported that multidrug-resistant (MDR) Salmonella Typhimurium may be associated with seafood. However, information is limited to the antimicrobial resistance, virulence properties, and genetic diversity of S. Typhimurium recovered from imported and domestic seafood. This study investigated the antimicrobial resistance, virulence properties, and genetic diversity of S. Typhimurium isolated from domestic and imported catfish, shrimp, and tilapia. A total of 127 isolates were tested for the presence of multidrug-resistance (MDR), virulence genes (invA, pagC, spvC, spvR), and genetic diversity using the Sensititre micro-broth dilution method, PCR, and pulsed-field gel electrophoresis (PFGE), respectively. All isolates were uniformly susceptible to six (amoxicillin/clavulanic acid, ceftiofur, ceftriaxone, imipenem, nitrofurantoin, and trimethoprim/sulfamethoxazole) of the 17 tested antimicrobials and genetically diverse. Fifty-three percent of the Salmonella isolates were resistant to at least one antimicrobial and 49% were multidrug resistant. Ninety-five percent of the isolates possessed the invA gene, 67% pagC, and 43% for both spvC, and spvR. The results suggest that S. Typhimurium recovered from seafood is frequently MDR, virulent, and have the ability to cause salmonellosis.


Introduction
Salmonella belongs to the family Enterobacteriaceae and includes more than 2500 serovars. The occurrence of Salmonella in seafood is mainly due to cross-contamination associated with zoonotic and/or human involvement [1,2]. Salmonellosis is a global issue and one of the leading causes of foodborne illness in the U.S. that causes an estimated~1.4 million non-typhoid cases of salmonellosis and 1.8 million cases of Typhoid salmonellosis per year resulting in 26,500 hospitalization and 420 fatalities [3][4][5][6]. Salmonella-contaminated food including seafood was linked to non-typhoidal human salmonellosis [7][8][9][10]. Approximately 3-10% of salmonellosis cases result in bacteremia requiring treatment with antimicrobials [5]. Serovar, S. Typhimurium is the most common cause of illnesses in the United States and accounts for most human infections [3][4][5][6].

Salmonella Isolates
One hundred twenty-seven confirmed S. Typhimurium var 5 previously isolated from 440 domestic and imported frozen shrimp (domestic: Pandalus jordani, Litopenaeus setiferus, and Crangon franciscorum; imported: Litopenaeus vannamei, Penaeus merguensis, and Metapenaeus spp.-total n = 156), catfish (domestic: Ictalurus punctatus; imported: Pangasius swai-total n = 142), and tilapia (domestic: Oreochromis aureus; imported: Oreochromis niloticus-total n = 142) were used in this study. Of them, 62 isolates were from domestic, and 65 isolates were from imported seafood. All samples were purchased from four retail stores on the Eastern Shore of Maryland [34]. One isolate from each positive sample was selected for characterization using antimicrobial susceptibility testing, virulence genes, and PFGE.

Antimicrobial Susceptibility Testing
Minimal Inhibitory Concentrations (MIC) tests for Salmonella isolates were determined by broth microdilution and interpreted according to the Clinical and Laboratory Standards Institute's guidelines [35,36]. In brief, different microdilutions of tested antimicrobials were made in Mueller Hinton broth (Thermo Fisher Scientific, Wilmington, DE, USA) considering their breakpoints. Each Salmonella isolate was added to all antimicrobial microdilutions. The minimal concentration that inhibited the growth of the pathogen was determined and compared to the breakpoints of each antimicrobial to determine whether the pathogen was susceptible, intermediate resistant, or resistant to the antimicrobial. Seventeen antimicrobials were tested (Trek Diagnostic Systems Inc., Cleveland, OH, USA) ( Table 1). All antimicrobials were chosen based on the type of bacteria, their clinical usage for Salmonella in both humans and animals, as well as aquaculture practices. Escherichia coli ATCC 25922 and Pseudomonas aeruginosa ATCC 27853 were used as control strains.

Detection of Virulence Genes
Salmonella isolates were tested for virulence genes (invA, pagC, spvC, and spvR) by PCR using the methods previously described [24,37,38]. In brief, genomic DNA for each isolate was extracted using the InstaGene matrix according to the manufacturer's instructions (Bio-Rad Laboratories, Hercules, CA, USA). The DNA concentrations of Salmonella were determined using NanoDrop Lite, according to the manufacturer's instructions (Thermo Fisher Scientific, Wilmington, DE, USA). Primers were used for Salmonella characterization following the manufacturer (Invitrogen by Life Technologies, Carlsbad, CA, USA) instructions (Table 2). GoTaq ® Green Master Mix (Promega Corporation, Madison, WI, USA) was used according to the manufacturer's guidelines. Twenty-five microliter reaction volumes were used; in which 12.5 µL of GoTaq Master Mix (2X) was added to 0.5 µL upstream primer and the same amount of downstream primer of each gene, 3 µL of DNA template and 8.5 µL nuclease-free water. Samples were loaded into the cycler and cycling conditions were adjusted for each targeted virulence gene as the denaturation, annealing, extension temperature and time varied depending on the testing genes [39,40] (Table 3). Salmonella Typhimurium strain Lt-2x3324 containing a recombinant plasmid with invA [41], S. Typhimurium containing a recombinant plasmid with spvC and spvR, and S. Typhimurium SR 11 x3337 containing a recombinant plasmid with pagC [42] were used as the positive controls and E. coli DH5-α (Invitrogen, Carlsbad, CA, USA) was used as the negative control for all reactions. Electrophoresis was carried out in a 1% agarose gel to separate the PCR products and the gels were stained with ethidium bromide.

Pulsed-Field Gel Electrophoresis of Salmonella
Pulsed-field Gel Electrophoresis (PFGE) technique was used to determine the change in the genomic profiles and was evaluated using XbaI (Boehringer Mannheim, Indianapolis, IN, USA) according to the guidelines developed by the Centers for Disease Control and Prevention (CDC) The CHEF-DR III SYSTEM (Bio-Rad Laboratories, Hercules, CA, USA) was used to perform electrophoresis using 1% SeaKem Gold agarose gel in 0.5X Tris-borate-EDTA (TBE) buffer at 14 • C for 19 h. The conditions of electrophoresis were as follows: initial switch time value of 2.16 s, final switch time of 63.8 s at a gradient of 6 V/cm, and an included angle of 120 • . After electrophoresis, the gel was stained with ethidium bromide solution (40 mg/mL) and then de-stained with deionized water. DNA bands were visualized with a UV transilluminator and a digital image of PFGE patterns was acquired. To analyze the data BioNumerics software (Version 7.6, Applied Math, Sint-Martens-Latem, Belgium) was used. The relatedness between the strains was determined using the Dice Coefficient of similarity, and strains were grouped by hierarchal clustering of inter-strain similarities based on the Unweighted Pair Group Method with Arithmetic Averages (UPGMA) [37].

Statistical Analysis
Descriptive summaries of the frequency of antimicrobial resistance and prevalence of virulence genes were determined, as well as the number and frequency of antimicrobial resistance profiles among isolates that were resistant to one or more antimicrobials. Statistical significance of differences in the frequency of antimicrobial resistance of isolates by source (domestic vs. import) was determined for each antimicrobial agent using Fisher's exact test. Fisher's exact test was also used to determine the significance of differences in the presence of virulence genes in isolates from different types of seafood. All statistical calculations were performed using R version 4.1.2 [43]. Differences were considered significant when the p-value was less than 0.05.

Prevalence of the Antimicrobial Resistance Phenotypes
All isolates were evaluated for antimicrobial resistance phenotypes using 17 different antimicrobials (Table 1). Table 4 showed the antimicrobial resistance phenotypes of Salmonella recovered from all domestic and imported seafood. All tested Salmonella isolates were uniformly susceptible to six (amoxicillin/clavulanic acid, ceftiofur, ceftriaxone, imipenem, nitrofurantoin, and trimethoprim/sulfamethoxazole) of the 17 tested antimicrobials. Fifty-three percent of the Salmonella isolates (n = 67) were resistant to at least one antimicrobial and 49% of them (n = 63) were multidrug resistant. Forty-seven percent (n = 60) of all Salmonella isolates (23 isolates were domestic and 37 were imported) were susceptible to at least one of eleven tested antimicrobials. Isolates recovered from domestic seafood were less resistant to some antimicrobials compared to those recovered from imported seafood. Salmonella-resistant isolates show high resistance against tetracycline (TET) (domestic 35% and imported 40%), nalidixic acid (NAL) (domestic 23% and imported 38%), gentamicin (Gen) (23% for the domestic and 34% for imported) and ciprofloxacin (CIP) (domestic showed low resistance (8%) in contrast to the imported (32%)). Differences in resistance to Ciprofloxacin, Chloramphenicol, Ampicillin, and Cefoxitin among isolates recovered from domestic versus imported seafood were statistically significant (p < 0.05).
The antimicrobial profiles of Salmonella recovered from shrimp were varied and displayed nine profiles. Domestic isolates (n = 21) possessed three profiles including 15 isolates that were susceptible to all antimicrobials while imported (n = 20) exhibited six profiles; 6 were susceptible to all antimicrobials. The predominant profile was ampicillin-cefoxitin-gentamicintetracycline (AMP-FOX-GEN-TET) exhibited by four isolated recovered from imported seafood. The antimicrobial profiles for Salmonella recovered from tilapia were varied and displayed six profiles. The predominant profile for the isolates recovered from imported tilapia was cefoxitin-ciprofloxacin-gentamicin-nalidixic acid-tetracycline (FOX-CIP-GEN-NAL-TET). The antimicrobial profiles for Salmonella recovered from catfish were varied and displayed 12 profiles. Domestic isolates (n = 26) possessed eight profiles while imported (n = 15) exhibited four. The most frequent profiles within the isolates recovered from domestic catfish were amikacinampicillin-cefoxitin-doxycycline-florfenicol-gentamicin-kanamycin-nalidixic acid-tetracycline (AMI-AMP-FOX-DOX-FFN-GEN-KAN-NAL-TET). No apparent differences were observed among antibiotic resistance profiles of isolates from different types of seafood.   Table 6 showed the presence of virulence genes invA, pagC, spvC, and spvR. The amplicon's sizes for invA, pagC, and spvC genes were 284 bp, 318 bp, 571 bp, and 310 bp, respectively. Overall, 95% of the isolates were positive for the invA gene (91% of the isolates recovered from domestic seafood, and 100% of the isolates were recovered from imported seafood). Sixty-seven percent of the isolates possessed pagC, and 43% for both spvC and spvR. With respect to the seafood type and source, the invA gene was detected in 90% of the isolates recovered from domestic shrimp and 100% of the isolates recovered from imported shrimp. It was also detected in 86% and 100% of the isolates recovered from domestic and imported catfish, respectively. The gene was also detected in all isolates (100%) recovered from tilapia of both sources. The pagC gene was detected in 67% (n = 85) of the isolates. The gene was detected in 54% (n = 35) and 81% (n = 50) of the domestic and imported isolates, respectively. Fifty-two percent of the isolates recovered from the domestic and 95% of the isolates recovered from imported shrimp possessed the pagC gene. The gene was also detected in 46% and 43% of the isolates recovered from domestic and imported catfish, respectively. On the other hand, it was detected in 69% and 89% of isolates recovered from both domestic and imported tilapia, respectively. Similar prevalence of spvC and spvR genes were observed in all domestic and imported seafood. Forty-three percent of all isolates had both genes. The two genes were detected in 29% (n = 6) and 50% of domestic and imported shrimp 36% (n = 10) and 57% (n = 8) of the domestic and imported catfish, respectively. Both genes were also detected in 63% (n = 10) and 39% (n = 11) of the domestic and imported tilapia, respectively ( Table 6). Except for the pagC gene, no statistically significant differences in the frequency of detection of the gene targets by seafood type were identified. For the pagC gene, the lower frequency of detection in isolates from catfish versus shrimp and tilapia was statistically significant (p < 0.05). There was no apparent association between the presence of the tested virulence genes and antimicrobial resistance profiles of Salmonella isolates.

Molecular Characterization of Salmonella Using Pulsed-Field Gel Electrophoresis-PFGE
Pulsed-field Gel Electrophoresis (PFGE) was used to appraise the genetic relatedness among the strains of the same pathogen recovered from the three types of seafood (shrimp, catfish, and tilapia). Salmonella isolates were digested with the restriction enzyme XbaI. Digestion of DNA resulted in 12 to 20 bands and the molecular weights of bands ranged from 50 to 1000 kbp.
The PFGE profiles of all Salmonella recovered from domestic (n = 62) and imported (n = 65) seafood combined (shrimp, catfish, and tilapia,) revealed 18 banding patterns which have been grouped into 14 clusters (from A to N) with 73% pattern of similarity. The similarity index of clusters ranged from 73% to 74% and the number of isolates in clusters was 1-30 ( Figure 1).

Molecular Characterization of Salmonella Using Pulsed-Field Gel Electrophoresis-PFGE
Pulsed-field Gel Electrophoresis (PFGE) was used to appraise the genetic relatedness among the strains of the same pathogen recovered from the three types of seafood (shrimp, catfish, and tilapia). Salmonella isolates were digested with the restriction enzyme XbaI. Digestion of DNA resulted in 12 to 20 bands and the molecular weights of bands ranged from 50 to 1000 kbp.
The PFGE profiles of all Salmonella recovered from domestic (n = 62) and imported (n = 65) seafood combined (shrimp, catfish, and tilapia,) revealed 18 banding patterns which have been grouped into 14 clusters (from A to N) with 73% pattern of similarity. The similarity index of clusters ranged from 73% to 74% and the number of isolates in clusters was 1-30 ( Figure 1).  The PFGE profiles of all Salmonella (n = 62) recovered from domestic seafood (shrimp, catfish, and tilapia) revealed 18 banding patterns which have been grouped into four clusters (from A to D) with a 74% similarity index ( Figure 2).
The PFGE profiles of all Salmonella (n = 62) recovered from domestic seafood (shrimp, catfish, and tilapia) revealed 18 banding patterns which have been grouped into four clusters (from A to D) with a 74% similarity index (Figure 2). The PFGE of all Salmonella (n = 65) recovered from all types of imported seafood (shrimp, catfish, and tilapia) revealed nine banding patterns which have been grouped into six clusters (from A to F) with an 82% similarity index (Figure 3). The PFGE of all Salmonella (n = 65) recovered from all types of imported seafood (shrimp, catfish, and tilapia) revealed nine banding patterns which have been grouped into six clusters (from A to F) with an 82% similarity index (Figure 3).
High genetic diversity was also observed, as evidenced by the dendrogram using both domestic and imported Salmonella isolates. Moreover, we did not observe any apparent association between antimicrobial resistance and PFGE profiles. There was also no association among PFGE profile, type of seafood, and country of origin through a few isolates displayed a tendency to cluster based on their source, type of seafood, and/or antimicrobial profiles. Although some isolates showed similar antimicrobial resistance profiles, they were genetically diverse. Each cluster showed a different similarity index (74-92%), isolates (1-36 isolate antimicrobial resistance profiles. In addition, clusters contained isolates from di seafood types/sources (Figures 1-3).
High genetic diversity was also observed, as evidenced by the dendrogram both domestic and imported Salmonella isolates. Moreover, we did not observe any ent association between antimicrobial resistance and PFGE profiles. There was also sociation among PFGE profile, type of seafood, and country of origin through a fe lates displayed a tendency to cluster based on their source, type of seafood, and/o microbial profiles. Although some isolates showed similar antimicrobial resistanc files, they were genetically diverse.

Discussion
This study characterized 127 Salmonella Typhimurium isolates from importe domestic catfish, shrimp, and tilapia for antimicrobial resistance profiles, virulence

Discussion
This study characterized 127 Salmonella Typhimurium isolates from imported and domestic catfish, shrimp, and tilapia for antimicrobial resistance profiles, virulence determinants, and genetic relatedness based on PFGE profiles. It is particularly notable that S. Typhimurium was the exclusive serovar isolated in this study in contrast to previous seafood surveys. Wild-caught domestic shrimp exhibited a lower incidence of resistance to individual antimicrobials and MDR than imported aquacultured shrimp consistent with previous studies (Table 7).  Other studies key: Beshiru et al. [17] in Nigeria (#S1); Broughton and Walker, [44] in China (#S2), Das et al. [45] in Bangladesh (#S3); Karp et al. [46] in Taiwan (#S4), Obaidat and Salman [47] in Jordon (#S5), Wang et al. [48] in USA (#S6), Zhao et al. [49] in USA (#S7). N/A (not applicable): the drug was not used in the study. ND: not determined. Varies: other drugs used with variable resistance. T: Sensitive to all tested antibiotics. +: resistance was identified but the percentage was not determined in the study.
Most of the countries exporting seafood to the EU and the USA have adopted the same regulations as the country to which they export [50]. Differences in mandates between products for domestic consumption and the export market may impact the antimicrobial resistance profile of pathogens [51]. Predictably, resistance to tetracyclines (oxytetracycline is FDA approved) was greater than other antimicrobials that are not allowed for USA aquaculture. However, none of the isolates from either domestic or imported seafood were resistant to trimethoprim/sulfamethoxazole, which is FDA approved. Ciprofloxacin is in the fluoroquinolone class and is not FDA approved but is used in other countries [52]. Our results showed a lower resistance to ciprofloxacin among domestic isolates than imported isolates. All isolates in which ciprofloxacin was included in MDR were from imported seafood.
Antimicrobial profiles of Salmonella revealed resistance to the following antimicrobials: amikacin, ampicillin, cefoxitin, chloramphenicol, ciprofloxacin, doxycycline, florfenicol, gentamicin, kanamycin, nalidixic acid, and tetracycline. Salmonella isolates were susceptible to amoxicillin/clavulanic acid 2:1, ceftiofur, ceftriaxone, imipenem, and nitrofurantoin, as well as trimethoprim/sulfamethoxazole (Table 4). In comparison in a Bangladesh study on raw domestic shrimp from local farms, S. Typhimurium and S. enteritidis were isolated. S. Typhimurium was found to be resistant to amikacin, colistin, and erythromycin (Table 7). Moreover, it possessed intermediate resistance to ciprofloxacin, and kanamycin [45]. Beshiru et al. [17] in Nigeria stated that S. Typhimurium showed multidrug resistance to amoxicillin/clavulanic acid, ampicillin, doxycycline, and tetracycline. In addition, the bacterium showed resistance to amoxicillin, penicillin, and erythromycin ( Table 7). The findings of this study contradicted the findings of Wang et al. [48] who stated that Salmonella isolates were susceptible to ampicillin, chloramphenicol, ciprofloxacin, gentamicin, and tetracycline (Table 7). Zhao et al. [49] demonstrated antimicrobial-resistant Salmonella serovars including Typhimurium recovered from imported seafood. They showed resistance to at least one antimicrobial: amoxicillin/clavulanic acid 2:1, ciprofloxacin, chloramphenicol, nalidixic acid, tetracycline, and trimethoprim/sulfamethoxazole. The later antimicrobial was found susceptible in our study (Table 7). Such difference may be due to the continual mutable paradigms of antimicrobials used in each country, the long-term and combined usage of antimicrobials, as well as the epidemiological surveillance objectives. Antimicrobial profiles were complex and constantly changing episodes in different environments.
Salmonella Typhimurium recovered from domestic seafood showed low resistance (8%) to Ciprofloxacin (CIP) compared to imported seafood (32%). These results suggest the use/misuse of CIP or other fluoroquinolones in or near aquaculture facilities in the countries from which these seafood products were exported. Cabello [11] concluded that excessive use of antimicrobials as prophylactic or preventive medications could be a predisposing factor to antimicrobial resistance of bacteria recovered from seafood. Moreover, antimicrobial resistance may be transferred vertically among bacteria.
The virulence of Salmonella is associated with both chromosomal and plasmid-linked genes. This study investigated four major virulence genes: an invasion gene (invA) for epithelial cells invasion, a chromosomal virulence membrane gene (pagC) for survival within the macrophage, Salmonella plasmid virulence gene spvC (often named virA), and the transcriptional regulator, spvR, which is in an operon with the other three genes. The spv group potentiates the systemic spread of the pathogen and survival within the host cell [24].
The findings of this study reported the percentage of samples in which the different virulence genes invA (95%), pagC (67%), spvC (43%), and spvR (43%) could be detected. With the exception of pagC, the prevalence of these virulence determinants was not associated with seafood source or type (Table 6). No association between antimicrobial profiles of Salmonella isolates and seafood source or type is evident. All Salmonella isolates from fish were invA positive in an Indian study [25] and similar high prevalence results appeared universally [26][27][28][29] (Table 8). The findings above suggested that invA may be considered as a target gene to detect Salmonella irrespective of the other genes and all isolates recovered from food/environmental samples may not contain other virulence genes as clinical isolates. Studies' key: S1, Akiyama et al. [27] in the USA; S2, Beshiru et al. [17] in Nigeria; S3, Bhatta et al. [28] in Nepal; S4, Chaudary et al. [24] in India; S5, Kumar et al. [26] in India; S6, Oliveira et al. [29] in Brazil; S7, Nolan et al. [54]; S8, Soto et al. [55]; S9, Tekale, et al. [25] in India. N/A (not applicable): not tested gene. P: present in all or some isolates. ND: gene was tested but not detected. P: gene was present (the percentage was not determined).
PFGE is a well-established epidemiological tool to assess diversity among Salmonella strains and suggests their relatedness regardless of their source (environment, seafood, and/or human clinical cases) [31,32,49]. Similar to previous investigations [33,34,52], the PFGE patterns of Salmonella in our study showed that Salmonella recovered from imported and domestic seafood are genetically diverse and did not show any concordance with source/type of seafood, antimicrobial resistance, or virulence properties.
The results of this study demonstrate a highly diverse population structure within the S. Typhimurium serovar in isolates from both domestic and imported seafood. Higher resistance levels to antimicrobials banned in the US such as CIP in imported seafood isolates suggest that their use is occurring around production environments in some export countries. Further studies to identify the genetic determinants linked to resistance patterns and their potential for horizontal transfer are warranted and whole genome sequencing (WGS) of the collection is currently under consideration.  Data Availability Statement: Upon request, data from this study will be made available by the corresponding author.