From a collection of Salmonella isolates obtained in previous studies, 351 Salmonella isolates were randomly selected for characterization in this study. Salmonella isolates were obtained from fecal grab samples (n = 31), beef hides (n = 104), and beef carcasses (n = 215) samples. Isolates included in this study were obtained from three abattoirs in Southeastern Mexico (Veracruz, Merida, and Cancun). The Salmonella isolates from Veracruz included in this study were obtained during five separate sampling periods: July and December 2009, and April, August, and December 2010. The Merida isolates were recovered during 4 time periods: March, September, and December 2012, and March 2013. Finally, the Cancun isolates were recovered from the Cancun abattoir during three time periods in October and December 2012, and March 2013 (Table 1).

All 351 Salmonella presumptive positive isolates were streaked onto tryptic soy agar (Beckton Dickinson, Sparks, MD) containing 5% defibrinated sheep blood and incubated at 37°C for 18 to 20 h. Antibiotic susceptibility was evaluated using the Sensititre automated antimicrobial susceptibility system (Trek Diagnostic Systems, Westlake, Ohio) following the manufacturer’s instructions. The following quality control organisms were used: E. coli 25922, Enterococcus fecalis ATCC 29212, Staphylococcus aureus ATCC 29213, and Pseudomonas aeruginosa ATCC 27853. Additionally, as defined by NARMS, a single isolate exhibiting resistance to three or more antimicrobial classes was classified as multidrug-resistant (MDR; FDA, 2015).

Fourteen antibiotics were used for testing: amoxicillin/clavulanic acid (2:1 ratio), ampicillin, azithromycin, cefoxitin, ceftiofur, ceftriaxone, ciprofloxacin, chloramphenicol, gentamicin, nalidixic acid, streptomycin, sulfisoxazole, tetracycline, and trimethoprim/sulfamethohazole. The Minimum Inhibitory Concentration (MIC) breakpoints for 12 of the antimicrobials tested were interpreted using the National Committee for Clinical Laboratory Standards for microdilution broth methods, and the MIC breakpoints for streptomycin and erythromycin were interpreted with the National Antimicrobial Resistance Monitoring System (NARMS) breakpoints (Center for Disease Control, 2011; Clinical and Laboratory Standards Institute, 2015; FDA, 2015).

Statistical analyses were performed using R (R, Version 3.2.2, The R Foundation for Statistical Computing, Vienna, Austria). Fisher’s Exact Test of independence with a 95% confidence limit (CL) was used for each comparison with a 2 × 2 factorial design. Additionally, Fisher’s Exact Test with a 95% confidence limit (CL) with Bonferroni’s multiple comparison post-test was used for comparisons with a 2 × 3 factorial design. The variation in the probability of recovery of resistant isolates versus susceptible isolates within each abattoir from a specific area (Veracruz vs. Yucatan Peninsula abattoirs), and within each animal sampling location (fecal, hides, and carcass) was evaluated. Additionally, the variation in the probability of recovery of multi-drug resistant isolates versus single resistant isolates within each abattoir from a specific area, and within each animal sampling location was evaluated. P-values less than 0.05 were considered significant.

We analyzed a total 351 Salmonella isolates for antimicrobial sensitivity screening. Our findings revealed that 205 (58.4%) Salmonella enterica isolates exhibited resistance to at least one or more antimicrobial drug. Resistance to tetracycline was the most common profile, and it was exhibited by 40.2% of the isolates (82 isolates), followed by resistance to nalidixic acid in 21.1% (43 isolates). No isolates presented resistance to gentamycin, azithromycin, and sulfisoxazole.

In addition, 26.3% (54 of 205) of these Salmonella isolates had a multidrug-resistance (MDR) phenotype. The most common MDR phenotypes exhibited by these isolates shared resistance to chloramphenicol, streptomycin, tetracycline, and trimethoprim/sulfamethoxazole (11.3%), followed by resistance to ampicillin, tetracycline, and trimethoprim/sulfamethoxazole (3.4%), and resistance to ampicillin, streptomycin, and tetracycline (2.5%; Table 2). Notably, when it came to antimicrobial resistance phenotypes in each abbatoir, we determined there was no statistical difference in the frequency of resistant vs. susceptible Salmonella isolates among the three abattoirs (Table 3).

In this study, there was no observed statistical difference in the frequency of resistant vs. susceptible Salmonella isolates across the 3 abattoirs or animal sampling locations (fecal, hides and carcasses) within each abattoir. However, we were able to determine that there was an effect on the recovery of tetracycline and nalidixic acid among the abattoirs. These findings are consistent with a study by Perez-Montano et al. (2012), where they reported that resistance to tetracycline (46.2%) was the most common profile among the Salmonella isolates recovered from beef carcasses (n = 78), followed by resistance to nalidixic acid (17.9%) in the state of Jalisco, Mexico. Authors suggested that based on their data and data in other studies that antimicrobial resistance to tetracycline, streptomycin, and chloramphenicol is very common among Salmonella isolates due to a longer and frequent use of these antimicrobials in animal production (Perez-Montano et al., 2012).

Salmonella isolates obtained from slaughter processing plants (carcass swabs and ground product) report Tetracycline resistance as the most common antimicrobial resistance profile in the U.S. In the latest NARMS Annual Animal Report (Center for Disease Control, 2011), of the 340 Salmonella isolates tested from cattle at slaughter, tetracycline resistance was the most common with 30.6%, this is consistent with our findings and those by Perez-Montano (Perez-Montano et al., 2012). Additionally, tetracycline resistance is also among the common antimicrobial resistance profiles found in retail meats in the U.S. based on the latest NARMS retail meat report (FDA, 2015). However, when it came to nalidixic acid, only 1.8% of the isolates from retail meat in the U.S. were resistant (Center for Disease Control, 2011).

In developing countries, Tetracycline is commonly used for treatment in human infections, and as a prophylactic agent in veterinary medicine. Tetracycline is classified as a broad-spectrum agent, meaning it can affect a wide range of gram-positive and gram-negative bacteria (Pezzella et al., 2004; Stevenson et al., 2007). Our findings are similar to those of previous studies, in that resistance to tetracycline is very common among beef cattle bacterial isolates (Thaker et al., 2010; Center for Disease Control, 2011; Perez-Montano et al., 2012). However, we are not inferring that this resistance is necessarily attributed to the selective pressure of antibiotics use in the cattle, as this information was not collected.

Additional findings in our study include high resistance to fluoroquinolones and resistance to extended-spectrum cephalosporins. The high levels of resistance to nalidixic acid observed in this study may be a concern due to the fact that nalidixic acid resistance has been associated with reduced susceptibility to fluoroquinolones in bacteria belonging to the Enterobacteriaceae family (Veldman et al., 2011). Fluoroquinolones are the last line of treatment for severe Salmonella infections, and continued antimicrobial resistance is of concern due to the potential threat to global public health (Acheson and Hohmann, 2001).

Additionally, 7.84% (16 of 205) of the Salmonella isolates showed resistance to a combination of tetracycline and streptomycin. Streptomycin is an antimicrobial that has limited current usage in human medicine, but plays a critical role in veterinary medicine, where it is used for the treatment of bacterial infections in cattle, sheep and pigs. Streptomycin is also critical for bacterial disease control in plants (Pezzella et al., 2004). Our findings are consistent with the latest NARMS Annual Animal Report (2011), where the percentage of streptomycin resistance in Salmonella isolates from slaughtered cattle has steadily increased over the years from 1997 to 2011 (Center for Disease Control, 2011).

Based on the definition for multidrug-resistance, a total of 26.3% (54 of 205) of the Salmonella isolates were found to be MDR. Traditional first-line antimicrobial drugs, such as chloramphenicol, ampicillin, and trimethoprim-sulfamethoxazole, tend to have higher reports of widespread resistance, which is consistent with our findings (Sjölund-Karlsson et al., 2011). Treatment of invasive and severe Salmonella infections include the use of fluoroquinolones (e.g., ciprofloxacin) or extended-spectrum cephalosporins (e.g., ceftriaxone), due to the widespread resistance of first-line antimicrobial drugs (Su et al., 2004; Miranda et al., 2009).

Comparing our findings to those reported by NARMS in the U.S., we can see that the percentages of MDR phenotypes observed for the Mexican Salmonella isolates in our study vary in comparison to what is reported in the U.S. (Center for Disease Control, 2011). In the U.S., resistance to at least ampicillin, chloramphenicol, and trimethoprim-sulfamethoxazole was 1.5% (n = 340) in Salmonella isolates from cattle in 2011. Additionally, resistance to ampicillin, chloramphenicol, streptomycin, sulfamethoxazole/sulfisoxazole, and tetracycline was 12.6% (n = 340) in all Salmonella isolates tested (Center for Disease Control, 2011).

Interestingly, our findings of MDR phenotypes in the isolates from the Veracruz abattoir are an indication that we can have antimicrobial resistance in the bacterial isolates, despite the implementation of food safety measures in an abattoir. Evidence of this also happens in the U.S., as NARMS reports show there are resistant pathogens in U.S. beef abattoirs (Kunze et al., 2008; Center for Disease Control, 2011). A major limitation of our study was the lack of serotype information. This information is certainly warranted to determine if the serovars of these Salmonella isolates are serovars commonly associated with human illnesses. In serovars such as Salmonella Typhimurium, Newport, and Montevideo, antimicrobial resistance is more significant than in other serovars due to the fact that these serovars tend to have the phenotypic trait of MDR; hence serotype information can help us understand the epidemiology of drug-resistant Salmonella.

In conclusion, data presented in this study clearly illustrates the presence of Salmonella drug-resistant isolates in beef cattle and the variability of antimicrobial susceptible profiles present in Salmonella isolates from cattle in Mexico. Antimicrobial resistant Salmonella in cattle feces, and on hides and carcasses have the potential to reach consumers by means of cross-contamination of the carcass during the slaughter and dressing procedure.

Tetracyclines, penicillins, and sulfonamides are among the most common antimicrobials used in animal production in developing countries, although they are no longer used to treat human infections. Therefore, the presence of drug-resistant isolates in animal production could potentially lead to the development and spread of resistance in humans.

One of the best risk management practices to prevent widespread antimicrobial drug resistance in Salmonella, and other foodborne pathogens present in food-producing animals such as beef cattle, is to monitor and report antimicrobial drugs use. Additionally, mitigation of antimicrobial resistance in Salmonella, can also be achieved by educating veterinary and human medicine specialists on the proper usage of antimicrobial drugs. Finally, additional research is necessary to characterize the genetic mechanisms responsible for the resistance profiles encountered among these Salmonella isolates.