Current Status of Antibiotic Resistance in the Six Major Foodborne Bacteria

Foodborne pathogens have gained notoriety because, despite sustained quality practices and regulations regarding bacterial load in food production, foods are still found to be contaminated with bacteria. Pathogens are usually poor competitors and do not survive in the presence of existing food microflora [1] or after food processing or unit operations for food preservation (e.g., vacuum, pasteurization, nano/microfiltration) [2]. However, this ecological defense is not a guarantee of safety; food extended, erroneous manipulation, or bacterial survival mechanisms favor pathogen persistence. FDA food recalls are targeted actions to remove or correct food products that violate U.S. food safety laws. They are primarily driven by microbiological contamination, undeclared major allergens (such as milk or peanuts in mislabeled packages), and foreign objects. For example, of the 233 total food and beverage recalls from 2023, 70 (30.0%) were associated with possible or confirmed foodborne illness/bacterial contamination. Understanding the relevance of these recalls involves knowing how the FDA classifies and communicates the associated health risks. The most common contaminants were found to be 28 (41.2%), 32 (47.0%), 3 (4.4%), 4 (5.9%), and 1 (1.5%) for Salmonella spp., Listeria monocytogenes, E. coli, Cronobacter sakazakii, and Clostridium botulinum, respectively (Table 1). C. sakazakii has been the subject of contamination and product recalls of powdered and dairy formulas in Chile as well [3]. More recently, the most serious food recalls have involved Salmonella contamination in foods such as ice cream [4], cereals [5], peanuts [6], soft cheese [7], shelled walnuts [8], crispy fried onions [9], and hummus (Table 1) [10]. Recent French data indicate that, with some yearly decline in the number of cases, meats, dairy, and fishery products are among the most recalled products [11].

The presence of pathogens, especially in RTE foods, and the evidence of antibiotic resistance among these strains have evolved into a worldwide public health challenge [13,14]. Hence, detection and constant surveillance programs of these pathogens in foods are paramount. A primer on novel approaches to analyzing these bacteria in food can be found in recent work [15,16]. From the regulatory standpoint, the European Commission decision 2020/1729 includes the surveillance and antibiotic resistance report of the specific pathogens Salmonella spp., E. coli (especially those strains with extended spectrum β-lactamases (ESBL), AmpC β-lactamases, or carbapenemases), C. coli, C. jenuni, Enterobacter faecium, E. faecalis in foods such as broilers, laying hens, fattening turkeys and pigs, bovine animals under a year of age, fresh meat from broilers, turkeys, pigs, and bovine animals. Analogously, it has regulations regarding microbiological criteria for foodstuffs, specifically, 2073/2005/EC.

Official methods for detection of pathogens include ISO 6579-1:2017, ISO 11290-1:2017, ISO 10272-1:2017/Amd 1:2023, ISO 6888-1:2021, ISO 22964:2017 [17], ISO 16649-1:2018, DIS 13136-1, for Salmonella spp., and for Listeria spp., Campylobacter spp., Staphylococcus spp., Cronobacter spp., and E. coli, respectively. Relevant methods for food analysis also comprise ISO 20976-1:2019, 23418:2022, and 20976-2:2022. These methods have also evolved with the advent of new technologies for detecting and controlling foodborne pathogens [18,19,20,21,22,23,24]. Rapid, reliable, and cost-effective methods for bacteria detection are of paramount importance in food safety monitoring [25].

In the case of regulations for antimicrobial susceptibility testing, the European Committee on Antimicrobial Susceptibility Testing (EUCAST) sets epidemiological cut-off values (ECOFFs) and tentative ECOFFs (TECOFFs). For the USA, the Clinical & Laboratory Standards Institute CLSI sets these guidelines, and a workbook is available [26]. Additional guidelines for monitoring and reporting in zoonotic and commensal bacteria include 2013/652/EU (repealed by 2020/1729/EU) and 2003/99 [27]. Finally, ISO 20776-1/-2:2022 sets standards for micro broth dilution methods and other in vitro diagnostic tools for testing minimal inhibitory concentration (MIC).

Official methods for detecting pathogens and performing Antimicrobial Susceptibility Testing (AST) combine traditional phenotypic cultures with rapid molecular diagnostics. Standardized by global authorities, these tools are vital for guiding life-saving therapies, managing antimicrobial resistance (surveillance, policy, and stewardship), and ensuring optimal patient outcomes [28].

Only phenotypic results from recognized, standardized in vitro antimicrobial susceptibility testing of isolates, along with interpretive criteria, can successfully help classify multiple, extended, pan, and resistant bacteria [29].

Phenotypic testing is the recognized gold standard; it works by exposing isolated bacteria to specific antimicrobial concentrations and observing their actual growth response. While molecular/genotypic tests (such as PCR or sequencing) are excellent for identifying specific resistance genes, they cannot determine the cumulative effect of multiple resistance mechanisms. Because of this, standardizing definitions of drug resistance relies exclusively on phenotypic measurements [28].

The long-term, overuse and misuse of antibiotics in food animals and humans create ideal conditions for the rising threat of antibiotic resistance. Resistant bacteria in animals may reach humans directly or indirectly through food, water, mud, and manure used as fertilizer [30]. Some types of bacteria (high-burden resistant pathogens such as Salmonella, Shigella, Neisseria gonorrhoeae, Pseudomonas aeruginosa, and Staphylococcus aureus) [31,32] that cause serious infections in humans have already developed resistance to most or all the available treatments, and there are very few promising options in the research pipeline [33,34].

Interventions that restrict antibiotic use in food-producing animals, such as biosecurity, improved animal management, vaccination, improved hygiene, and changes in animal housing and husbandry practices, have reduced antibiotic-resistant bacteria in these animals [35]. Hygiene practice alone can attenuate the effect of antibiotic selection pressure [36]. These practices, in conjunction with One Health approaches, should be implemented to counter antibiotic resistance [37,38]. In this regard, the One Health approach is a collaborative strategy recognizing that human, animal, plant, and environmental health are deeply interconnected. Because antibiotic resistance easily spreads across these domains, containing it requires integrated monitoring, responsible prescribing, and environmental stewardship across all sectors [39].

The overall reduction in the use of antibiotics in food-producing animals, including the complete restriction of these substances for growth promotion and disease prevention, without proper diagnosis. Malaise animals should be assessed to determine the most effective and prudent antibiotic to treat their specific infection. Meanwhile, healthy individuals should only receive antibiotics to prevent disease if, within a particular population, an animal has already been diagnosed [40].

Antibiotic use is further exemplified by the application of streptomycin (STR), oxytetracycline (OXY), copper-based products, and certain fungicides (e.g., triazoles) in plant agriculture. Their use is correlated with increased resistance among plant pathogens [41,42,43]. Horizontal gene transfer is responsible for genes being exchanged among a variety of bacteria in the plant production setting, including phytopathogens, soil bacteria, and zoonotic bacteria that are occasionally present in that environment and the food chain [30,42]. Through gene transfer, co-resistance, cross-resistance, and gene up-regulation, resistance to one compound may confer resistance and multi-drug resistance to other similar, or even very dissimilar, compounds [42].

Genetic information encoding bacterial resistance mechanisms can be chromosomal or carried by mobile elements called plasmids. Some bacteria are also inherently resistant to a particular antibiotic if they lack the structure on which the antibiotic acts. Several mechanisms of resistance have continuously evolved, including target modification or mutation, permeability reduction, expression of efflux pump genes, production of hydrolases or inactivating enzymes, metabolic enhancement or auxotrophy, synthesis of target-protective proteins, self-repair systems, changes in cell morphology, and biofilm protection (Figure 1) [44,45]. For a thorough primer on specific mechanisms of resistance by antibiotic class, the work by van Hoeck et al. [46] is suggested.

Plasmids are major vectors of bacterial antibiotic resistance [47]. Recent evidence demonstrates that antibiotic resistance-related plasmids are differentially dispersed among humans and livestock, reflecting known antibiotic prescribing practices [48]. For example, using a large-scale multivariate approach, the authors found that only 0.42% of livestock plasmids carried carbapenem resistance (compared with 12% of human plasmids); conversely, tetracycline resistance was enriched in livestock vs. human plasmids [48]. However, they also found that the prevalence of carbapenem and colistin plasmid-related resistant genes is on the rise.

Despite the long list of antibiotics commercially and clinically available (Table 2), resistance has made some of these drugs therapeutically useless against multidrug-resistant bacteria. Hence, the discussion of alternatives to antibiotics has resulted in natural product-led research [49,50,51]. Several primers regarding natural products with antimicrobial capabilities are advised [52,53,54,55,56,57]. Interestingly, recent research suggests that constant bacterial exposure to even these alternatives can affect their resistance profile (that is, decreased sensitivity). Exposure to biocides, pesticides, common home cleaners, heavy metals, food preservatives, and even natural products can lead to a cross-adaptation to antibiotics induced by exposure to non-antibiotic antimicrobials [58,59,60]. This co-resistance occurs when bacterial exposure to sub-lethal concentrations of non-antibiotic agents, used within the clinical, agricultural, and domestic settings, inadvertently selects strains that are resistant to therapeutic antibiotics (i.e., the physiological changes often double as defense mechanisms against antibiotics) [61]. More worrisomely, evidence of this behavior in pathogenic foodborne bacteria such as S. aureus, E. coli, S. typhimurium, P. aeruginosa, K. pneumoniae, L. monocytogenes, A. baumanii, C. sakazakii, and other species is clear [59].

Salmonella is a genus of rod-shaped (bacillus), motile, non-spore-forming, chemotrophic, facultative anaerobe, and Gram-negative bacteria of the family Enterobacteriaceae. It is known for its two species, S. enterica and S. bongori. S. enterica is further divided into six subspecies that include over 2600 serotypes [63,64].

Nontyphoidal Salmonella serotypes are zoonotic and can be transferred from animals to humans. Infection by this pathogen is usually due to ingestion of contaminated, undercooked food. Salmonella is one of the most common causative agents of foodborne illness in humans, causing approximately 1.35 million infections, 26,500 hospitalizations, and 420 deaths in the United States alone every year [65].

Treatment of salmonellosis is further complicated by the fact that isolates may also carry resistance genes conferring resistance to antimicrobials. For example, Salmonella species carrying antimicrobial resistance genes have been isolated from animal- and plant-derived foods, including poultry, eggs, fish, pork, beef, vegetables, fruits, nuts, and flour [56]. Outbreak studies have recognized food commodities, including fresh produce, vegetables, cantaloupes, cereals, alfalfa sprouts, fruit/fruit pulp, pistachios, poultry meat, turkeys, tuna, ground beef, pork, dried/shredded coconut, tomatoes, and mangos, as sources of foodborne Salmonella-related outbreaks in the past decade [66,67,68,69].

Soubeiga et al. [70] found Salmonella-related virulence genes associated with vegetable/fruit sources. Multidrug resistance profiles were consistent determinants in the 148 isolates, as genes bla_TEM, temAB (β-lactams), StrA, aadA (aminoglycosides), sul1, sul2 (sulfonamides), and tetAB (tetracyclines) were consistently present (Figure 2).

Similar gene profiles [e.g., aadA, aac, aph, bla_CARB, bla_PSE, bla_TEM, bla_CTX-M, bla_CMY, floR (phenicols), folate pathway inhibitors sul and dfrA (folate pathway inhibitors), qnrAB, oqxAB (fluoroquinolones), tetABM, and mph(A) and lnu(F) (macrolides)] were found recently, but in retail American poultry [71], Mexican ground beef samples [72], and Chilean poultry production line samples [45]. Analogous gene profiles were also found in food chain links in Poland [73]. On the other hand, recent Italian isolates from swine mostly exhibited resistance determinants parC, gyrA, catA, nfsAB, bla_TEM, and tetAB. For a detailed primer regarding mechanisms of resistance for Salmonella, the work by Monte et al. [74] and Wang et al. [75] is recommended.

Furthermore, as global trends indicate that more people are basing their diets on fresh produce and vegetables [76], Salmonella contamination is relevant, as an increase in produce-related outbreaks has been documented [77]. Additionally, multidrug-resistant bacteria can disseminate more quickly due to an ever-expanding regional and international food trade [77]. More problematic still, it is easy to forget that foods, in general, and especially fruit and vegetables [78], can carry more than one pathogen, and these pathogens can carry resistance genes and even share them amongst themselves and other bacteria. In a very thorough Cuban study, 14 different food groups were analyzed for the presence of Salmonella spp., E. coli, Staphylococcus spp., and V. cholerae [79]. Meat and dairy products and RTE foods showed not only multiple pathogen contamination but also multiple drug resistance profiles [79]. A recent study saw lettuce, cabbage, spinach, carrot, tomato, and green peppers contaminated with other antibiotic-resistant species alongside Salmonella (e.g., E. coli, S. aureus, S. epidermis) [78].

On the other hand, Cao et al. [80] found a high prevalence of Salmonella (isolating the bacteria at a rate of 29.4% per sample tested) contamination in cow’s milk. A similar account was reported by Ejo et al. [81], who found both a higher prevalence of contamination and a drug resistance profile in raw meat, eggs, and milk.

Quesada et al. [82] reviewed Salmonella reports from 2003–2014 from Latin America, as well as the antibiotic-resistant profiles of these isolates (especially recovered from the Brazilian and Mexican poultry industry). Among their findings, S. Enteritidis, S. Typhimurium, S. Paratyphi, and S. Heidelberg were the most commonly encountered serovars. Similar to other reports [83,84] most strains were multidrug-resistant and showed a high tolerance to STR, NAL, AMP, TET, CHL, and STX (Table 3). Relatedly, a 22-year historical analysis of antibiotic-resistant Salmonella from Bangladesh found 81% tetracycline resistance [85], rendering this antibiotic effectively without any therapeutic or clinical use. However, the continued use of tetracyclines in other agricultural settings (e.g., crop production, apiculture, and aquaculture) still provides a selective pressure on bacteria [40].

As macrolides (the first-choice drugs) for Salmonella infections in humans have failed, cephalosporins are now the paramount therapeutics. However, ESBL enzymes encoded by bla_CMY2, bla_SHV, and bla_TEM emerged, thus reducing the efficacy of cephalosporins. Also, generalized mcr genes have been described in Salmonella and encode colistin resistance, an antibiotic regarded as critical. In both cases, resistance genes were carried by conjugative plasmids (e.g., IncI2, IncHI2, IncX3, IncX4) [86].

Interestingly, Siddique et al. [87] not only found Salmonella strains carrying β-lactam, cephalosporins, and carbapenems resistance genes (e.g., bla_CMY-2, bla_OXA-1, bla_PSE-1, bla_TEM-1, bla_NDM-1, and ampC), and the zoonotic potential virulence gene sopE, but also noted that the strains recovered had moderate-to-high abilities to aggregate into biofilms at 30 and 37 °C. Other virulence factors (e.g., fimA, invA, spvC) co-occurring with similar resistance gene patterns have been recently reported in Poland [73], Italy [88] and China [89,90,91,92].

For Brazilian poultry farms, Rau et al. [93] found, in two different sampling moments (i.e., 3 years apart), that traditional Salmonella serovars Typhimurium and Enteritidis have been supplanted by other serovars like Heidelberg or Kentucky. In the same period, Brazil saw an increase in the number of isolates exhibiting antimicrobial resistance by almost double. Similar phenomena seem to be occurring of late in countries like the USA and Costa Rica [94,95].

As the debate about rendering animals without antibiotics rages on, evidence suggests that alternative meat production practices reduce the frequency of bacteria exhibiting antibiotic resistance, which favors “organic” practices [71]

Salmonella enterica serovar Infantis (or S. Infantis) is an emergent serotype worldwide (Table 3) [96]. This serotype has been reported as one of the most prevalent non-typhoidal Salmonella (NTS) in humans in Europe and the USA [97,98,99]. Several studies describe S. Infantis as the most prevalent serotype in poultry [98,99]. Recent data from over 10,000 isolates demonstrated that S. Infantis and S. Seftenberg with pESI plasmid carrying a bla_CTX-M-65 gene constitute one of the biggest clusters of Salmonella along the United States, both originating from turkeys [100].

Moreover, multidrug-resistant phenotypes of this serotype are of public health concern [101]. For example, Mattock et al. [102] recently related the presence of pESI megaplasmid with 73% of the poultry isolates from 74 different countries showing multidrug resistance. Hence, the presence of Salmonella in poultry production has attracted serious attention, and several studies have found in poultry a vector for several species of this bacterium. Whole genome sequencing from broiler integrations determined that the most prevalent non-typhoid Salmonella serotype was S. Infantis [103]. Not only do most isolates show potential for resistance against medically important antibiotics [especially aminoglycosides (mutations in genes aac, ant, aph) and quinolones (parC, gyrA)] and disinfectants, but also almost all isolates harbored the pESI-like megaplasmid, which plays an important role in the global dissemination of antimicrobial resistance. Among analogous Latin American studies, the bla_CTX-M-65 ESBL antibiotic resistance gene is also prevalent in isolates. This was more recently confirmed by Costa Rican-led research, in which ampicillin resistance in the more prevalent S. Infantis reached 81.6% [95]. In terms of plasmid prevalence, a similar trait has been observed in S. Muenchen, S. Schwarzengrund, S. Agona, and S. Senftenberg in a large-scale broiler and chicken isolate study [104]. Previous studies, again in the Andean Region (mostly Ecuador, Chile, and Peru), seem to support similar findings [105,106,107,108,109]. A more recent Chilean study [47] has reported mobile elements carrying multiple drug resistance determinants in the poultry production line, predominantly related to S. Infantis and S. Heidelberg (IncFIB and Incl1-Iα plasmids, respectively).

Furthermore, the ESI clone carrying an extended spectrum β-lactamase bla_CTX-M-65 on a pESI-like plasmid and a mutation in the gyrA gene has also been found in poultry meat in regions such as Russia, USA, Italy, Germany, Hungary, Poland, and Slovenia [110,111,112,113,114]. All these data suggest that pESI is not only widespread but is acquired by Salmonella spp. with relative ease. Finally, recent evidence has described pESI-like megaplasmids as carriers of additional antimicrobial resistance genes, such as tet, dfr, and sul, and has demonstrated that these structures are genetically plastic and can also harbor virulence and fitness genes, as well as toxin/antitoxin systems that enhance their persistence in the S. Infantis host [97].

Salmonella monitoring and surveillance in animal feed are not only related to production costs and animal welfare but are also instrumental in understanding the behavior of foodborne pathogens at the start of the food chain and how this relates (if any) to prevalence, drug-resistant profiles, and specific species found downstream in human or clinical isolates.

Relatively few studies have tackled feed contamination despite being considered a main source of contamination on farm grounds (Table 3) [65]. In Thailand, the prevalence of Salmonella contamination in compound feeds (a specially formulated, balanced mixture of feed components designed for feeding livestock and poultry) was registered at 12.5% [115]. Similar data have been recovered from isolates from American and Australian feeds [71,116,117,118]. The authors found more prevalence in the compound feed than in any of its respective ingredients. However, Parker et al. [118] recently related contamination hotspots with the milling equipment during feed manufacturing. Hsieh et al. [117] found an 84% contamination in fish meals and 48% in animal products. From these samples, the acquisition of the bla_CMY-2 gene, which confers resistance to cephalosporins, was most commonly observed. Interestingly, multidrug-resistant isolates were recovered from monensin (an ionophore/Rumensin^®)-medicated dry beef cattle feeds. A possible association between the use of ionophores in livestock and antibiotic resistance has recently been raised [119].

Of all the bacterial pathogens discussed, Salmonella seems to have received the most attention in the feed sector (both in surveillance and research). Antibiotic resistance in Salmonella often stems from agricultural practices, where antimicrobial usage in animals selects resistant strains that spread into the human food supply. This poses a major public health issue, as these pathogens can trigger severe infections that are difficult to treat with standard medications. The cycle of how antibiotic resistance moves from animal feed to human consumers operates as follows: i. Contaminated Feed, ii. Livestock Exposure, iii. Indirect Transmission to Humans [120].

As stated, however, there is a very close link between livestock and human pathogen contamination. For instance, the main source of Salmonella infections in humans is contaminated animal-derived foodstuffs, with pork products being among the most important contributors [65].

Listeria is a genus of Gram-positive, rod-shaped, facultatively anaerobic, and non-endospore-producing bacteria that acts as an intracellular parasite in mammals. Although in food analysis L. monocytogenes is the most targeted species, early reports have focused on other strains, such as L. innocua, L. seeligeri, and L. welshimeri. Though no resistance was found in the latter, 19.3% of the L. innocua isolates, mostly from frozen food items, were resistant to one or more antibiotics [168]. A more recent study showed that 29% of samples from butcher shops were positive for Listeria spp. (48, 28, and 7% for L. innocua, L. seelligeri, and L. monocytogenes, respectively) [169].

L. monocytogenes resistance has been attributed to conjugation (e.g., plasmid horizontal transfer from other Gram-positive bacteria such as Enterococcus, Streptococcus, and Staphylococcus), efflux pumps [170,171] genetic alteration, and biofilm formation (especially contamination during food packaging) [172]. An excellent primer on Listeria-specific mechanisms of resistance can be found in the work of Luque-Sastre and collaborators [173].

Listeria isolates are differentiated by probing for immunoreactivity to two cell-surface structures, i.e., the somatic O and flagellar H antigens. For L. monocytogenes, four O-serogroups (1/2, 3, 4, 7) have been recognized, which are further differentiated into serotypes 1/2a, 1/2b, 1/2c, 3a, 3b, 3c, 4a, 4b, 4c, 4d, 4e, and 7. Three of these (1/2a, 1/2b, and 4b) cause most of the human illness cases. Specifically, serotype 4b is most associated with outbreaks [174,175]. For example, in Mexico, from 300 food samples, 5.6% of recovered strains were identified as L. monocytogenes, of which 68.4% belonged to serogroups 4b, 4d, and 4e, and 92.1% exhibited at least four virulence genes (i.e., actA, hly, and plcB) [176]. A similar case was found in Romania, where L. monocytogenes isolates from RTE, meat, and dairy foods included 47.1 and 29.4% of the 1/2a-3a and 4b-4d-4c serotypes, respectively [177]. An additional study in this country found that 75% of L. monocytogenes strains recovered from RTE foods were serotype 1/2a (with regulatory factor prfA present in all isolates). Analysis of antimicrobial resistant genes identified tetracycline-resistant genes, particularly tet(C), tet(M), and tet(K). The presence of ampC and dfrD was also noticeable [178].

From a total of 50 Listeria strains isolated from meat and dairy products, consisting of seven L. monocytogenes and 43 L. innocua strains, which showed the presence of antimicrobial resistance genes: tet M, tet L, mef A, msr A, erm A, erm B, lnu A, and lnu B. Multidrug resistance was identified in 27 Listeria strains, four belonging to L. monocytogenes. Resistance to CDA (a lincosamide) was the most common resistance phenotype, identified in 45 Listeria strains. Hence, the preferred first-line therapy remains AMP or PEN, often combined with GEN or RIF for synergistic effects [170]. Worryingly, recent data suggest L. monocytogenes strains recovered from foods exhibited 87.3% of multidrug resistance and 92.4% to β-lactams [179]. Similarly, South African isolates exhibit resistance determinants for sulfonamides (sul1 and sul2, 64.4%), aminoglycosides [strA, aadA, aph (3)-IIA, 47.9%], and β-lactams (bla_TEM, and blaz, 44.3%). For patients allergic to β-lactams, STX is the standard alternative, as Listeria remains broadly susceptible to it (Table 4) [170]. Additionally, in some cases, high resistance to FOS has been documented as a natural feature of the pathogen (together with NAL, FMQ, and cephalosporins) (Table 4) [179,180,181].

Interestingly, a longitudinal study found that significant resistance genes in L. monocytogenes included fosX, lin, abc-f (macrolide), and tet(M). No increase in antimicrobial resistance determinants was observed over 11 years [182]. There is additional evidence suggesting co-selection of resistance traits in food production systems and bacterial responses to stress and disinfectants (e.g., bcrABC) [181]. Over eight years, acquired antimicrobial phenotypes were shown to confer resistance to tetracyclines (tetM), trimethoprim (dfrD), lincosamides (lnuG), macrolides (ermB and mphB), and phenicols (fexA). A similar trend was recently described in isolates recovered from Chile, in which 100% of the L. monocytogenes strains carried the fosX, lin, norB (fluoroquinolones), mprF (lipopeptides), tetA, and tetC resistance genes. In addition, the arsBC, bcrBC, and clpL genes were detected, which conferred resistance to stress and disinfectants [183]. Recent research also supports these findings, with next-generation sequencing of L. monocytogenes showing that fosX determinants are ubiquitous across 58 studies and, in up to 60.38% of reports, the presence of the lin and norB genes [184].

Since L. monocytogenes is ubiquitous in agricultural environments, the cultivation of fresh produce predisposes it to contamination. Cold-stored dairy, meat products, fresh produce, and novel foods made thereof (e.g., hummus or guacamole) are considered high-risk foods [185].

In this scenario, the contamination of RTE foods originates either from raw ingredients (of both plant and animal origin) or, more commonly, from cross-contamination within food-processing facilities [163]. Interestingly, recent research has also demonstrated that stressors in food-processing environments (e.g., heat, acidic and alkaline pH, UV irradiation, and high osmolarity) can induce tolerance that may contribute to the predominance of Listeria and the development of antimicrobial resistance, even without antibiotic selection pressure [186]. Said stressors induce bacterial tolerance by triggering global stress responses that alter cellular physiology. When bacteria encounter stressors such as sublethal antimicrobials, temperature shifts, or nutrient starvation, they temporarily slow their metabolism, deploy protective chaperones, and alter membrane permeability, thereby effectively blocking antibiotics from reaching their targets [187].

For example, in carcasses obtained from poultry slaughterhouses in Northern Greece, an 80 and 99% of samples were positive for Listeria spp. and L. monocytogenes, respectively [166]. All isolates were resistant to NAL (a naphthyridone) and OXO (both used to treat Gram-negative bacteria), most were resistant to CDA, and only a few were resistant to OXY and TET [188].

Finally, although relatively recent attention has been drawn to the resistance of Listeria species in food, some countries, possibly compelled by listeriosis outbreaks, have already amassed considerable research on the matter [189,190]. Notably, some other African countries have also identified human breast milk as a reservoir for Listeria [191].

Campylobacter spp. are Gram-negative spiral, non-spore-forming, chemoorganotrophic, rod-shaped, or curved bacteria. These bacteria are a leading cause of foodborne gastroenteritis [201,202]. The advent of molecular detection of genes responsible for antibiotic resistance (e.g., gyr for fluoroquinolones, erm for macrolides, tet for tetracyclines, or aph for aminoglycosides) has facilitated clinical therapy and surveillance of the emergence and dissemination of antibiotic-resistant bacteria [203]. For an excellent primer regarding Campylobacter resistance mechanisms (including fluoroquinolones, macrolides, FLO, β-lactams, aminoglycosides, and tetracyclines, the work by Iovine, Shen et al., Tang et al., and, more recently, Aleksic et al., are proposed [204,205,206,207].

In a very thorough work regarding contamination in a slaughterhouse, Schreyer et al. [208] found not only Campylobacter spp. in chicken cecum but also evisceration knives and workers’ hands, and samples before and after the chiller. Interestingly, C. coli contamination was significantly higher than that of C. jenuni. This trait has been observed in analogous (poultry-related) studies in other latitudes [209,210]. Not only was the contamination within the meat chain relatively high, but the strains also exhibited several genes responsible for resistance, such as cmeA/B, tet(O), and bla_OXA-6.

Efflux pump CmeABC expels substances such as antimicrobials, toxic heavy metals, dyes, and detergents from the interior of the bacterial cell [211]. Interestingly, it has been correlated with surface polysaccharides, which serve as a resistance mechanism against phenolic compounds [212]. It is also largely responsible for macrolide [213] and multidrug resistance in Gram-negative bacteria [211,214], and its role in the clinical context is well defined, as it contributes significantly to high-level resistance to CIP and TET in C. jejuni and C. coli [215].

Additionally, the tet(O) determinant is a common trait and has been described in similar studies [216,217]. Such evidence adds to the previous body of research, which heavily linked fluoroquinolone-resistant Campylobacter spread from poultry (since Campylobacter is a normal enteric flora in many avian species) to humans [209,218,219]. However, as a zoonotic infection, it can be present in domestic and wild birds and other animals [220]. Hence, campylobacteriosis warrants a One Health approach to tackling this problem [221].

Alike results have also been obtained from pig slaughterhouses. Campylobacter spp. have been isolated from pig guts (200 samples, 47.6% contaminated), feces, and surfaces. 90.8% of strains exhibited multidrug resistance [222]. A general increase in Campylobacter resistance toward fluoroquinolones (e.g., NAL/CIP) and TET has been reported worldwide [220,223,224,225]. EFSA reported that rates of resistance to CIP were very high in C. jejuni and C. coli isolates from humans and very high to extremely high in C. jejuni and C. coli isolates from poultry, pigs, and calves [224]. A recent review that reported South American Campylobacter isolates indicated that although the bacteria showed an overall low resistance to ERY (a drug of choice to treat campylobacteriosis), it was widespread [225], a trait that persists in Europe [224]. On the other hand, in general, cattle and chickens seem to be important hosts in the Sub-Saharan African region, and may play an important role in transmitting to humans. Most Campylobacter isolates were resistant to ERY (44%), AMP (39%), TET (33%), NAL (31%), and CIP (30%) [226].

A recent study [227] found multidrug resistance in almost all isolates from cattle, chicken, and water samples, and also determined that the most resistance determinant within the strains was the tet(O) gene. Despite several restrictions in some countries, tetracyclines are still among the most systematically used as growth promoters in animal feed, which maintains selective pressure over gut microbiota [228,229,230,231]. A recent Canadian study [232] found that, even in 2021, the most used antibiotic in turkey farming includes bacitracin (at 472.7 mg·kg⁻¹ per biomass), followed by β-lactams, orthomycin (a combination of isoniazid and a streptomycin derivative) [233], and tetracyclines (at 58.8 mg·kg⁻¹ per biomass). Other genes coding for resistance were those for β-lactams (bla_OXA-61), aminoglycosides (aph-3-1), (fluoro)quinolones (gyrA), and multidrug efflux pump (cmeB). Tunisian laying hens and eggs have been recently found to harbor similar resistance determinants [234]. The authors also found the erm(B) gene in almost all the isolates from these samples. Byambajav et al. also found the erm(B) gene or a nucleotide substitution at nt 2075 of 23S rDNA in all C. coli isolates from Mongolia [235].

An interesting finding that supports the (over)use of antibiotics in livestock as a major driver of the selection of resistant bacteria is reported by Paintsil et al. [210]. The authors found that fecal samples from animals on smallholder farms exhibited not only significantly lower Campylobacter contamination but also reduced antibiotic resistance.

Notably, a Belgian study [236] showed that human isolates (from diarrheal cases) had resistance profiles similar to those found in animals, with a tendency toward increased resistance to NAL, CIP, and TET. Whole genome sequencing of human isolates from the USA [237] found antimicrobial resistance genes encoding mostly resistance to the β-lactam (bla_OXA-193, bla_OXA-460), tetracyclines [tet(O), tet(O/32/O)], fluoroquinolones (gyrA), and aminoglycosides (aph(3′)-III). Another African study reports similar results in pediatric cases with clinical diarrhea, supporting the transfer of genetic determinants of antibiotic resistance from animals to humans [238].

In conclusion, evidence indicates that the relatively high rates of fluoroquinolone resistance (Table 5) have limited their usefulness in treating Campylobacter infections. Macrolides like azithromycin are the current drugs of choice when antibiotic treatment is indicated. Resistance to macrolides in Campylobacter has remained stable. The increased prevalence of antibiotic-resistant Campylobacter cannot be solely attributed to antibiotic overuse. Interspecies transmission and subsequent clonal expansion also contribute to the dissemination [221].

There seems to be a certain plasticity in Campylobacter species, and genetic material from C. jejuni to C. coli is a possibility, which then increases virulence [234]. The result is a higher number of antibiotic-resistant infections, which could become a health burden. Furthermore, genetic exchange between Campylobacter and Gram-positive bacteria occurs in the natural environment and is more frequent than previously realized. These antibiotic-resistance genes present in Campylobacter species can further disseminate by clonal expansion or horizontal gene transfer [221].

Staphylococci are clump-forming Gram-positive cocci of 1 μm in diameter. They are salt-tolerant and often hemolytic. S. aureus and S. intermedius are coagulase-positive. All other staphylococci are coagulase negative. Staphylococcal food poisoning is a rapid-onset illness caused by eating foods contaminated with S. aureus, often transmitted by food handlers or through dairy/meat products.

An early report on whey and cheese showed that staphylococci coagulase-negative (CoNS, 54% of the isolates) included species such as S. lentus. Coagulase-positive staphylococci identified were divided into 3 groups: S. aureus (40%), S. intermedius (2%), and S. hyicus (4%), indicating a high incidence of PEN resistance. S. lentus isolates exhibited resistance to OXA, ERY, and LIN [250].

More recently, Osman et al. [251] found S. hyicus, S. aureus, S. schleiferi subsp. coagulans, S. intermedius, and S. lentus in beef meat. These strains were characterized by possessing genes gyrA, gyrB, and grlA (encoding essential type II/IV topoisomerases in bacteria that regulate DNA supercoiling during replication and transcription). Though phenotypically already resistant to OXA, only 40.7 and 7.4% carried the mecA and cfr genes, respectively. The chloramphenicol-florfenicol resistance gene is a mobile genetic element that encodes a 23S rRNA methyltransferase, conferring broad-spectrum, multidrug resistance to Phenicols, Lincosamides, Oxazolidinones, Pleuromutilins, and Streptogramin A [251].

Although S. aureus is mostly considered a clinically relevant pathogen [252], several reports have contributed to establishing its incidence in food samples, food handlers, food-producing animals, and food processing environments worldwide (e.g., Brazil [253,254], Portugal [255], China [256], Italy[257]). Of special interest are reports that include the presence of Methicillin-Resistant Staphylococcus aureus (or MRSA) in foods. MRSA (resistance primarily driven by the mecA gene, which encodes a low-affinity penicillin-binding protein [PBP2a]) is one of the leading causes of infections and a major threat to human health worldwide due to its numerous toxins [252,258].

One particularity among food Staphylococcal contamination is the specific attention given to direct handling (Table 6) [259]. S. aureus is known to easily colonize the hands and nose. In fact, among health professionals, carriage of S. aureus on hands and nose was found to be 8.9 and 39.6% (4.7 and 17.2% for MRSA), respectively. In addition to β-lactams, the majority of nasal MRSA were also resistant to ERY, and CIP [260].

Mekhloufi et al. [261] found that the S. aureus strains recovered from foods belonged to seven different staphylococcal enterotoxin gene (SEg)-types harboring the following combination of se and sel genes: (1) selW, selX; (2) egc (seG, seI, seM, seN, seO), selW, selX; (3) seA, seH, seK, seQ, selW, selX; (4) seB, selW, selX; (5) seD, selJ, seR, selW, selX; (6) seH, selW, selX, selY; and (7) seA, egc (seG, seI, seM, seN, seO), selW, selX. Meanwhile, 2.1% and 4.2% tested positive for the tst- (toxic shock syndrome toxin-1) and mecA- (staphylococcal chromosomal cassette mec-type IV) genes, respectively.

Interestingly, Elsalkh et al. [262] found that 48.6 and 8.6% of isolates recovered from pizza were identified as methicillin- and vancomycin-resistant S. aureus through detection of mecA and vanA genes, respectively.

In yet another report [263], the hla (encoding an α-hemolysin) gene was detected in 87.1% of the isolated strains. Antimicrobial treatments heavily influence the hla gene; while some antibiotics (like β-lactams) can inadvertently trigger its expression, others (like CDA) actively suppress toxin production to help limit tissue damage [263].

Analogously, do Nascimento Pereira et al. [264] evaluated S. aureus strains isolated from food handlers of pilot kitchens. In this case, 83.8 and 18.9% of isolates exhibited seA and tst genes, respectively. Also, agr (the accessory gene regulator locus is a critical quorum-sensing system in S. aureus that controls virulence factor expression based on cell density). Typing classified 58.3% of the strains as type I, and 18.9% were classified as MRSA. The SCCmec typing characterized strains as type II, III, IV, and V. For a thorough description of the staphylococcal cassette chromosome mec (SCCmec) typing, please refer to Uehara [265].

Achek et al. [266] found in both food and clinical isolates, the tetM/K, blaZ, aacA-aphD, ermC, and mecA genes. More recently, Anas et al. [267] found that the prevalence of Staphylococcus was very high (75 and 50%) in street kebab and raw buffalo meat, respectively. These isolates (82%) showed not only multidrug resistance (MDR) patterns but also produced β-lactamases or extended-spectrum β-lactamases, at rates of 48% and 24%, respectively. The most detected resistance genes were mecA, sul1, and qnrS/qnrB (29.2, 25, 20.8%, respectively). Similarly, Urimi et al. [268] found that 75% of S. aureus isolates from fast food restaurants were MDR, but their indices of multiple antibiotic resistance (MAR) ranged from 0.2 to 0.6. For a comprehensive list of resistance genes found in Staphylococcus, food, and food-related environments, the work by Chieffi et al. [257] is suggested.

Some researchers have used specific antibiotics against Staphylococci species, such as MCP (effective against topical infections caused by MRSA and S. pyogenes) [268], PRI (to treat severe Gram-positive infections, including MRSA) [250,261], or CRN (used to treat acute MRSA skin and skin structure infections) to test sensitivity [267]. Notably, from these three antimicrobials, resistance has already been proven towards MCP [269,270,271].

Finally, several approaches have been suggested as alternatives to antimicrobials for controlling S. aureus disease (e.g., bacteriophages, plant-based antimicrobials, nanoparticle- or light-based therapeutics, and probiotics) [272].

E. coli is a gram-negative, facultative anaerobic, rod-shaped, coliform bacterium. As a fecal indicator, it may suggest fecal contamination or poor, unhygienic food handling. As an indicator organism of contamination, E. coli’s resistance profile is very relevant and has been documented (see discussion below and Table 7).

Babines-Orozco et al. [279] concluded that in Latin America, most antibiotic resistance genes present in E. coli strains are carried through IncF-type plasmids or class 1 integrons, including qnr (fluoroquinolones), dfrA1 (diaminopyrimidines), bla_SHV, bla_TEM-1, bla_CTX-M (β-lactams), tetA, tetB (tetracyclines), aac (6)-Ib, aadA1 (aminoglycosides), sul (sulfonamides), and cat-1, cmlA (phenicols), being the most commonly found in diarrheagenic strains, food, water, and some livestock animals [279,280,281,282]. Similar findings have been documented elsewhere, e.g., China, Italy, Korea, or Nigeria in foods (including rabbits, poultry, pigs, dairy/beef cattle) [280].

Pathogenic E. coli utilizes a combination of specific virulence factors (molecules that enable it to cause disease) and antimicrobial resistance mechanisms to evade host immune responses and survive antibiotic treatments. These traits allow the bacteria to colonize tissues and cause severe intestinal or extraintestinal infections [281]. In this regard, at least 11 distinct pathotypes have been described in regional research [279].

For example, a recent study found that E. coli was present in 100% of food samples, with higher contamination rates in vegetables and fruits. Enterotoxigenic (ETEC: elt) and Uropathogenic E. coli (UPEC: vat, cnf1, hylA) genes were also identified in all samples; meanwhile, Diffusely Adhering E. coli (DAEC: afa) was found in 27% of the assay objects. The most effective antibiotics against the recovered strains were TET, STX, polymyxins, and quinolones [282]. Other virulence factors of pathogenic E coli, tested elsewhere, include Enteropathogenic (EPEC: eaeA, bfpA), Enterohemorrhagic (EHEC: hylA), and Enteroinvasive E. coli (EIEC: ial) [283].

Li et al. [284] found a detection rate of E. coli isolated from foods as follows: pork (59.6%), mutton (52.6%), retail fresh milk (52.4%), duck (36.4%), beef (35.3%), chicken (33.3%), RTE food (12.9%), and fish and vegetables were <11%. The detection rate of resistance genes was as follows: tetA (38%), tetB (27%), bla_OXA (40%), bla_TEM (20%), floR (20%), sul1 (16%), sul2 (27%), aad_Ala (19%), aadB (11%), strA (28%), and strB (24%); tetC and bla_PSE were not detected. Virulence genes fimC, agg, stx2, fimA, fyuA, papA, stx1, and eaeA were found. Finally, they also found an interesting association between fimC and resistance toward CIP, GEN, AMK, LFX, and STR.

Likewise, Ronald et al. [285] isolated 30 E. coli with multiple antibiotic resistance (MAR) index of greater than 0.2 from meat samples; 88.9% of the samples tested positive. Isolates showed that the tetA (for TET resistance) and sul1 (for STX resistance) genes were present in 100.0% of isolates, while the bla_TEM (for AMP resistance) and strA (for STR) genes were present in 2 isolates. O157 lipopolysaccharide O-antigen synthesis gene rfbE-O157 was absent in all strains. Similar findings (i.e., high resistance to TET and STX and high homology in genes bla_TEM and gryA) have been further corroborated in food/clinical E. coli isolates from China [286].

A recent review of 61 studies from 18 countries showed that most food samples (i.e., chicken meat and eggs) had prevalence rates of 8.9% and 80.5% for Shiga toxin-producing E. coli (STEC) and serotype O157, respectively. Additionally, the isolates were resistant to the following antibiotics: AMC, CHL, TET, CIP, GEN, AMP, NEO, and AMO [287]. An independent study found similar trends in eggs and poultry sausage, with prevalence of 59.3% and 14.4% for E. coli and 2.2% and 5.6% for Salmonella, respectively [288].

Finally, some reports demonstrate a close link between agriculture [289] and aquaculture [290] based human activities and the dissemination of bacteria with antimicrobial resistance traits. For example, van Hamelsveld et al. [291] showed that polluted water (with normal contamination levels of E. coli) in continuous contact with shellfish magnifies their microbial contamination levels, rendering the food above the threshold apt for human consumption. The authors also demonstrated that the bacteria found within the shellfish show a regular and increased rate of plasmid horizontal transfer. Osman et al. [292] demonstrated that β-lactam resistance genes (bla_OXA, bla_MOX, bla_CIT-like, bla_SHV, bla_FOX) in E. coli were abundant both in poultry hatchlings and in the hatchery environment. Isolates were also able to form biofilms, thereby increasing their tolerance to antibiotics [293].

Furthermore, at least eight studies, published from 1940–2016, representing 18% of all revised literature, could directly link the transmission of antimicrobial resistance in E. coli from food animals to humans, 25 studies (56%) suggested transmission between animals and humans with no direction specified, and 12 studies (26%) did not support transmission [294].

The genus Cronobacter comprises seven species that are opportunistic pathogenic, Gram-negative, facultative anaerobic, non-spore-forming, bacillus-shaped bacteria. They can cause serious and even fatal infections, affecting infants, immunocompromised patients, and the elderly. In infants, the typical infection can cause meningitis, necrotizing enterocolitis, and bacteremia, and is usually foodborne, whereas in adults, bacteremia and urosepsis are typical and are normally nosocomial [306].

Foodwise C. sakazakii has been isolated from various plant- and animal-derived matrices, including confectionery, dried herbs and spices, teas, nuts, pasta, rice, oats, linseed, and cereal products, fresh/frozen produce [306,307], fruits, and powdered infant, follow-up formulas and milk, cheese, and meats (usually powdered foods with low moisture and water activity values). Table 8 provides a detailed account of the prevalence of the bacteria. The work of Sani and Odeyemi, and, more recently, that of Kumar et al., is recommended [308,309].

A pivotal study on C. sakazakii in households was conducted in 2012 and demonstrated that the bacterium is not only a relevant foodborne pathogen but is also present in food-related domestic environments. From this study, 76.1% of isolates were resistant to penicillin. Most of the research regarding C. sakazakii in foods and its antibiotic resistance profile has been focused on infant formulae. Overall, strains isolated from foods have shown special resistance to β-lactams. For example, Fei et al. [310] found resistance to AMC/AMP with MIC ≤ 8 µg·mL⁻¹. Similar values were found for isolates from a milk processing facility in Serbia [311]. The EUCAST threshold for assessing AMC is 8 µg·mL⁻¹. They also found that some strains were tolerant to dryness and osmotic pressure, which has a profound impact on food processing and storage [312,313]. More worrisomely, at least one study has positively identified C. sakazakii as an ESBL-producer (e.g., class C β-lactamases) [312,314]. On the other hand, a Brazilian study demonstrated that 94.4% of isolates from infant foods were tested resistant to cefazolin. Additionally, six antibiotic-resistance genes, particularly those encoding efflux pumps, were identified. Pakbin et al. [315] found that isolates from powdered infant formula milk were highly resistant to AMC, AMP, CEX, CFP, ERY, CTX, CIP, and CHL and susceptible to GEN, TET, NOR, and ATM antibiotics. Meanwhile, the bla_CTX-M-1 gene was detected in 96% of the isolates. Previously, the same research group [316] also found a higher prevalence of C. sakazakii in bulky imported domestically packaged powdered milk infant formula samples.

Although isolated strains do seem to possess relevant resistant phenotypes, the overall prevalence is relatively low. For example, in an Iranian study of 200 samples (100 infant formula and 100 baby foods), only eight (4%) samples tested positive [317]. However, the sole presence of a few bacteria and their multidrug resistance profile suggests that this pathogen can disseminate resistance genes with relative ease within these food samples [318]. Worrisomely, most reports of C. sakazakii have demonstrated multiple drug resistance profiles and biofilm formation in isolates from infant contaminated foods; both bacterial mechanisms result in therapy failure [319]. Biofilms and drug resistance profiles are the primary drivers of chronic, recurrent, and untreatable infections. Together, they cause therapy failure by creating physical, chemical, and genetic barriers that prevent antibiotics from either reaching or eliminating target pathogens [320].

Holý et al. [321] found in their C. sakazakii isolates a type IV secretion system-related gene (found on hybrid or cointegrate plasmids, incFIB/incH1 are central components of bacterial conjugation machineries), which plays a pivotal role in bacterial evolution by facilitating the horizontal transfer of genetic material—such as antimicrobial resistance and virulence factors—between bacteria [307,318]. These genes have also been described in C. malonaticus, C. turicensis, C. muytjensii, and C. universalis [322]. Several antibiotic target site modification and efflux-related genes have also been identified in C. sakazakii isolates (e.g., mdf(A), which encodes a multidrug efflux transporter, emrR, emrB, rsmA, cRP, adeF, and marR for fluoroquinolones or fos genes encoding resistance for fosfomycin) [323,324,325]. Recently, in a large Chinese study performed by Gan et al. [318], Strains isolated from powdered infant formula and supplementary food collected from 29 provinces exhibited at least 7 antimicrobial resistance genes, which corroborates other findings supporting multidrug resistance. Similarly, Fei et al. [325] also isolated C. sakazakii from goat milk-based infant formula from Shaanxi Province, China. From the serotypes found, O2 was the most abundant (n = 15/32, 46.88%) and showed, as well, mostly resistance to antimicrobials such as CEP (87.5%) and AMO (25%). Resistance phenotyping for other Enterobacteriaceae can be found in the work by Mardaneh et al. [326]. Recently, Song et al. [327] found, using a proteomics approach, that there is a certain relation between antibiotic tolerance and the production of enterotoxin and hemolysin.

Antibiotic resistance and the production of enterotoxins and hemolysins are critical virulence characteristics in pathogenic bacteria like S. aureus, E. coli, and B. cereus. Bacteria often acquire genes for drug resistance and virulence simultaneously and therefore are often located together on the same genetic elements (e.g., plasmids, transposons, or pathogenicity islands). When bacteria share genetic material, they can spread both traits to neighboring cells [328].

When antibiotics are used, bacteria with resistance genes survive. Because these genes are linked to toxin-producing genes, antibiotic use promotes the survival and proliferation of the most virulent strains. Hence, it appears to be a correlation between multidrug resistance and an increased presence of hemolysin and enterotoxin genes. Strains exhibiting this combination are associated with severe complications and a higher mortality rate in clinical settings [328].

On the other hand, Oloninefa et al. [329] demonstrated that strains with antibiotic sensitivity (to CIP and GEN) were eliminated in culture conditions after 240 min of exposure to the antimicrobials. Hence, those sensitive strains still seem to respond quickly to antibiosis. Interestingly, Capita et al. [330] described that C. sakazakii strains increased their resistance profile after treatment with several biocides (e.g., sodium hypochlorite, peracetic acid, benzalkonium chloride) commonly used in food processing equipment cleaning procedures [331,332]. There is also compelling evidence that suggests that antibiotic resistance increases virulence in Cronobacter spp. isolates from food [321,333].

Finally, resistance in isolates from clinical cases of neonatal hemorrhagic diarrhea and meningitis has been reported in countries like Mexico and China [322,324,334,335,336].