Antimicrobial resistance of Salmonella spp . strains isolated from hamburgers

Susceptibility of 45 Salmonella spp. strains isolated from beef and beef and chicken hamburgers was evaluated. Twelve antimicrobial drugs were used in the disc diffusion method. A cladogram was constructed and similarity across strains was assessed. Strains were susceptible to co-trimoxazole, chloramphenicol, amikacin, tetracycline and gentamicin (91.11, 91.11, 84.44, 80.0, 77.78%), though resistance was detected against cefotaxime, ampicilin, cephalothin, ceftriaxone, cefotixin, ceftazidime, and aztreonam (88.89, 71.11, 68.89, 53.33, 48.89, 42.22, 35.56%). The least effective antimicrobial drug was cefotaxime (88.89%). A high percentage of multiresistant Salmonella spp. strains was observed. This raises concerns as to the high risk for salmonellosis in humans, when these food items are consumed underdone, or by cross-contamination during preparation. Another important aspect in terms of public health is that these infections may become even more severe due to the lower efficacy of antimicrobials in treatments, both in humans and animals.


INTRODUCTION
Antimicrobial agents are commonly employed as growth promoters and in the prevention or treatment of infections in meat and milk producing animals.Nevertheless, the indiscriminate, unsupervised use of these drugs in feeds and water destined to animals represents a serious public health hazard due to the emergence of multiresistant strains responsible for the high prevalence of infectious diseases in humans as well as animals.
The use of antimicrobial agents both in humans and animals may select resistant bacteria populations (White et al., 2002).These drugs are used in animal feeds to treat and prevent bacterial infectious diseases, and as growth promoters.An undesirable outcome of this practice is the emergence of resistant microorganisms and the consequent transmission to humans due to the contamination of animal-derived foods.Estimates say that over 50% of all commercially available antimicrobial *Corresponding author.E-mail: jfortuna@uneb.br.agents are used in the animal-derived food industry (Aarestrup, 1999).Worldwide, considerable amounts of antimicrobial agents are used to promote growth and in prophylaxis, though the use of such drugs in therapy strategies is also significant.In this scenario, bacterial species normally present in animal products are exposed to a substantial and oftentimes constant selective pressure.According to Witte (2004), the selective pressure exerted by antimicrobials as growth promoters in animals used for food production may be behind expressive resistance transfer pools in the environments these animals are reared.Additionally, it has been described that antimicrobial agents used in treatment or prevention of bacterial infections in animals, either in therapy, metaphylaxis, prophylaxis and in growth promotion, belong essentially to the same class of compounds used in human medicine (Witte, 2004;Schwarz et al., 2001).
More specifically, in therapy antimicrobials are used to control an existing bacterial infection in one single individual.In turn, metaphylaxis is the use of antimicrobial agents in a whole group of animals when signs of an infection are detected in one or more individuals, if the whole group is considered to be at risk.Prophylaxis with antimicrobial drugs may be adopted for individual animals as well as groups thereof, via water and food.This strategy may also act as growth promoter (Schwarz et al., 2001).
Resistant Salmonella spp.strains have been increasingly isolated in human salmonellosis outbreaks due to the inclusion of antimicrobial agents in animal production feeds, mainly as growth promoters.This poses a public health hazard, due to the transfer of resistant Salmonella spp.strains to humans through the consumption of contaminated foods and infectious diseases that are not properly treated.
Salmonella spp.strains resistant to antimicrobials represent a risk to public health.These strains are responsible for salmonellosis outbreaks in humans and animals, which are difficult to treat efficiently.The individuals infected with resistant strains are more prone to prolonged, more severe diseases that require treatment in hospital -or that even lead to death -as compared to subjects infected with susceptible strains.Additionally, the indiscriminate use of antimicrobials may select resistant genes in pathogenic bacteria that later can transfer the acquired resistance to different nonpathogenic bacteria species (Aarestrup, 1999).For this reason, monitoring antimicrobial resistance plays an essential role as a means to obtain information on the span and trends of resistance, aiming to plan and observe the effects of treatment (Álvarez-Fernández et al., 2012).
Multiresistance to antimicrobials is particularly problematic in pathogenic bacteria, and may be considered a possible result of the use of different antimicrobial agents (Aarestrup, 1999).In this sense, the widespread emergence of multiresistant Salmonella spp. in foods reveals the need for more careful antimicrobial application strategies by farmers, veterinarians and doctors.Apart from this, efforts should be made towards reducing the prevalence of resistant strains of bacteria of this genus in foods, which include the adoption of guidelines for a more careful use of antimicrobial agents in animals used in food production, the enforcement of new food safety regulations, and efforts to reduce pathogen populations in abattoirs (Thakur and Bajaj, 2006).
The increased resistance of microorganisms to antimicrobials may be a consequence of vertical or horizontal resistance gene transfer (Witte, 2004).However, the resistance to antimicrobials may also result from mutations and the acquisition of genes that code for resistance (Fluit, 2005).The emergence of multiresistant Salmonella spp.serotypes has become a worldwide concern, since antimicrobial resistance worsens these pathogens' virulence.Resistance and virulence determinants may be localized in the bacterial chromosome, in transposons or plasmids, grouped in resistance or pathogenicity islands and transferred by movable genetic elements of phagi.In this sense, the presence of both determinants in the transposon or in the same plasmid may be selected by antimicrobial pressure, resulting in Salmonella spp.strains that are more resistant to antimicrobials (Wannaprasat et al., 2011).Moreover, Salmonella spp.resistance to antimicrobial agents is caused by several mechanisms, such as the production of enzymes that inactivate antimicrobial drugs based on structural degradation or modification, the reduction of antimicrobial penetration in bacterial cells, the activation of antimicrobial efflux pumps, and the emergence of different targets of antimicrobial on bacterial cells (Hur et al., 2012).
Determining the susceptibility and resistance profile of Salmonella spp.strains isolated from foods in a given region, apart from being an important epidemiological marker, may be useful in therapeutic procedures in human and veterinary medicine (Baú et al., 2001).Also, due to the ubiquitous character of the microorganism, its pathogenicity and the alarming increase in resistance to antimicrobials, identifying resistant strains is essential in empidemiological studies (Thakur and Bajaj, 2006).In this sense, the objective of the present study was to assess the susceptibility of Salmonella spp.strains isolated from hamburgers made with beef and a mixture of beef and chicken sold in the city of Niterói, state of Rio de Janeiro, Brazil.

MATERIALS AND METHODS
Samples of beef and beef and chicken hamburgers purchased in local supermarkets and butcher's shops in the city of Niterói, Brazil, were analyzed in the Laboratory of Microbiological Control of Animal Products, Department of Food Technology, Veterinary School, Universidade Federal Fluminense.
The microbiological isolation and identification of Salmonella spp.were carried out according to Ordinance No. 62, issued by the Ministry of Agriculture, Livestock and Supplies, Office for the Protection of Agriculture and Livestock, which publishes the official analytical methods used in the microbiological analyses of animal products and water (Brasil, 2003).Strains were isolated and identified as Salmonella spp.according to the conventional microbiological growth method.Serological and biochemical confirmation was carried out.
After microbiological confirmation, 45 isolated Salmonella spp.strains were tested for susceptibility against 12 different antimicrobial compounds according to the diffusion methodology described by Bauer et al (1966) and interpreted based on standards defined by The National Committee for Clinical Laboratory Standards (NCCLS, 2003).The antimicrobial agents used were: amikacin, gentamicin, ampicilin, aztreonam, cephalothin, cefotaxime, cefotixin, ceftazidime, ceftriaxone, chloramphenicol, cotrimoxazole and tetracycline.
Salmonella spp.strains were inoculated in slanted test tubes containing nutrient agar and incubated at 37ºC/24 h.The bacterial inoculum was suspended with saline 0.85% in a 0.5 McFarland scale.Then, the inoculum was seeded on Petri dishes containing Muller Hinton Agar (MHA).A set of previously chosen antimicrobial discs (Polisensidisc™) was placed on dishes, which were then that similar to usually attainable blood and tissue levels and for which response rates may be lower than that of susceptible isolates.The intermediate category implies clinical efficacy in body sites where the drugs are physiologically concentrated or when a higher than normal dosage of a drug can be used.This means that the microorganism may be inhibited by achievable concentrations of some drugs, when higher doses may be given or when the infection is in a site where the drug attains higher physiological doses.Resistant isolates are not inhibited by concentration of an antimicrobial drug in a site, or present specific resistance mechanisms.
Percentages of resistant Salmonella spp.strains in beef and beef and chicken hamburgers were compared using the Student's t test in the software Bioestat™ 3.0 (Ayres et al., 2003).Similarity analysis, to assess the formation of clusters or clades or microorganisms, was carried out using the software WinClada™ 1.00.08 (Nixon, 2002).The cladogram obtained revealed the similarity across strains, in terms of the categories susceptible, intermediate and resistant.The clade analysis included 12 different characteristics based on the 12 drug used.A binary matrix was constructed (45 strains X 12 antimicrobial drugs), with the values zero, 1 and 2 assigned to susceptible, resistant and intermediate strains, respectively, in the nodes of lines and columns.The dendrograms were obtained by parsimony analysis through heuristic search and constructed according to the single linkage method or the neighbor-joining method using the software Nona™ 2 (Goloboff, 1999).
The similarity analysis enables the formation and identification of significant clusters of individuals or objects of a sample population.Individuals are grouped according to the degree of similarity of variables (characteristics), while elements in distinct groups present different characteristics.
The Student's t test revealed no statistically significant differences in percentage of resistant strains isolated from beef and beef and chicken hamburgers (absolute value calculated: 0.9461; p=0.3644, below the significance levels α=0.01 (3.11) and α=0.05 (2.20); df: 11).
Of the 45 Salmonella spp.strains tested against the 12 antimicrobial drugs, 42 (93.33%)exhibited resistance to at least one drug.However, if intermediate strains are considered potentially resistant and if multiresistant strains are those that present resistance to two or more antimicrobials, it is then possible to conclude that only three (6.67%)strains have low resistance, with one (2.22%)being susceptible to all drugs tested (strain 19) and two (4.44%) resistant to one antimicrobial only (strains 12 and 21).Consequently, 42 strains (93.33%) were considered multiresistant (Table 3), of which three   (6.67%) were strains resistant to two drugs, three (6.67%) were resistant to three drugs, five (11.11%) to four, seven (15.56%) to five, six (13.33%) to six, six (13.33%) to seven, seven (15.56%) to eight, three (6.67%) to nine, and two (4.44%) to ten different antimicrobial drugs.No strain was resistant to 11 or 12 drugs.The cladogram constructed revealed some discrepancies between strains.However, it was possible to identify several similar strains.Nine different clusters were formed (A, B, C, D, E, F, G, H and I; Figure 2).Cluster A was formed only by susceptible strains, while the other clusters were composed of multiresistant strains.Some strains (04 and 18, cluster B; strains 28, 43, 01 and 34, cluster F), in spite of being isolated from the two kinds of hamburger (beef and beef and chicken) were similar.Oppositely, it was also possible to observe that strains placed in different clusters were present in one same hamburger sample (strains isolated from sample 27, Figure 2).

DISCUSSION
In a study that investigated the same antimicrobial agents, Cortez et al. (2006) observed a different behavior as compared to the results reported herein.The authors isolated a total of 29 Salmonella spp.strains from samples collected in broilers in an abattoir, 25 (86.2%) of which were resistant to aztreonam, 25 (86.2%) to ampicilin, 21 (72.4%) to tetracycline and 16 (55.2%)to co-trimoxazole.However, similar results were observed for susceptibility, since gentamicin was the drug to which the least strains were resistant (1 strain, 3.45%), followed by amikacin (2 strains, 6.9%).The authors also observed that strains were susceptible to gentamicin (96.5%) and chloramphenicol (48.3%).None of the strains tested was resistant or susceptible to all antimicrobial agents tested.
Different results were reported by Szych et al. (2001) in a study that investigated 320 Salmonella Enterica strains isolated from feces of diarrhea patients and carriers of the pathogen between 1998 and 1999.The authors observed that 91.4 and 21.5% of strains were resistant to tetracycline and ampicilin, respectively.Yet, low resistance was also seen for chloramphenicol and gentamicin.
The susceptibility of strains isolated from beef and beef and chicken hamburgers in the present investigation are similar to those reported in the study of Cardoso et al. (2006) in a study that analyzed resistance of 80 Salmonella Entiriditis strains isolated from broiler chicken carcasses between 1995 and 1996 resistant to gentamicin and co-trimoxazole.
Strains presenting intermediate resistance levels may be considered resistant.According to the study of Cortez et al. (2006), the prescription of antimicrobial agents against bacteria with intermediate resistance helps select resistant strains.Additionally, intermediate strains may promote the emergence of resistant strains, if antimicrobials are prescribed at concentrations below recommended values or if the infection site is not reached effectively by drugs at appropriate concentrations.
Multiresistance has not as yet been universally defined.Several studies have categorized multiresistant strains as those that are not affected by two or more antimicrobial drugs (Szych et al., 2001;Dargatz et al., 2003).However, recent papers define multiresistance as resistance Fortuna et al. 7531 against three or more drugs (Wannaprasat et al., 2011).Multiresistant strains have also been considered to be resistant to four or more agents (Castagna et al., 2001).This controversy around a precise, reliable definition of multiresistance makes it difficult to appropriately compare results.Szynch et al. (2001) reported that 49.9% of Salmonella Enterica strains isolated from feces of diarrhea patients and carriers of the pathogen resisted against two or more antimicrobials.However, in a study that tested 702 strains, 11.7% were shown to be multiresistant, most of which to tetracycline (35.9%), while 10.4% of samples were resistant to ampicilin and to chloramphenicol and no sample was resistant to amikacin and gentamicin (Dargatz et al., 2003).
Comparatively, these prevalence values for multiresistance are lower than those reported for 99 strains of Salmonella spp.isolated from pigs, where 24 (24.24%) were shown to be multiresistant in a study that stipulated multiresistance as resistance to four or more drugs (Castagna et al., 2001).In the present study, multiresistance was considered to be resistance against two or more drugs.
In a study that considered multiresistance as resistance against at least three drugs, Wannaprasat et al. (2011) evaluated 162 Salmonella spp.strains and observed that 89% were resistant to at least one antimicrobial drug, while 63% were multiresistant.The authors also observed that 74% of strains were resistant to tetracycline, and that no strain resisted to ceftriaxone, as opposed to the results of the present study.Also, a study compared the resistance of 40 strains isolated in 1993 to 19 strains isolated in 2006.The authors observed that 100% of strains of the first group presented multiple resistance to at least three drugs (25%), as well as four (52.5%) or five (22.5%) antimicrobials.The strains in the second group were resistant to three (26.3%),four (26.3%), five (10.5%), six (26.3%), seven (5.3%) and 13 (5.3%)different antimicrobial drugs (Álvarez-Fernández et al., 2012).Similarly, Cortez et al. (2006) reported that seven (24.14%) of the 29 Salmonella spp.strains isolated in a broiler chicken abattoir were resistant to seven antimicrobial agents, six (20.7%) to eight and four (13.8%) to five.
In a study that assessed the resistance of three Salmonella spp.serotypes to antimicrobials, Lynne et al. (2009) found highly resistant Salmonella Dublin strains (80% were resistant to between one and five antimicrobials, while 64% were resistant to six or more drugs).The authors also observed high resistance, with 87.5% of Salmonella Choleraesuis strains being resistant to at least one antimicrobial, and 37.5% resistant to at least four drugs.Salmonella Pullorum was the least resistant microorganism, with 55.5% of isolates resistant to one single antimicrobial drug.The overall results of that study showed high resistance to tetracycline (64.3% of isolates), and that all isolates were resistant to amikacin.These results differed from those reported by Souza et al. (2010), in a study that analyzed 44 Salmonella Typhi strains isolated in the state of Pará, Brazil, of which 10 (22.7%) were resistant to at least one antimicrobial drug.Ribeiro et al. (2008) analyzed 79 Salmonella Enteriditis strains isolated from clinical samples collected from broilers and breeders and environmental samples collected in a chicken broiler plant.The authors observed that 65 (82.3%) were resistant to one or more antimicrobial agents.Of these, 43 (66.1%)resisted to two or more and 22 (33.8%) to only one antimicrobial drug, which are higher prevalences as compared to the values observed in the present study.Oppositely, resistance to tetracycline was found in 53 (67.1%) of strains.Such high prevalence was also observed in a study that analyzed 33 Salmonella Hadar strains isolated from chicken samples (Zimermann et al., 2008).Graziani et al. (2008) analyzed susceptibility of Salmonella Typhimurium strains isolated in Italy between 2002 and 2004 from human and animal sources.The authors observed that 87% of strains isolated from humans and 81% of strains isolated from animals were resistant to at least one antimicrobial drug.High resistance was observed against tetracycline, ampicilin and chloramphenicol (73.6, 67.6 and 32.2% of strains, respectively).
One of the factors behind such high resistance values is the indiscriminate use of antimicrobial drugs with no medical supervision, apart from the fact that in Brazil antibiotics are sometimes sold over the counter, with no prescription.Additionally, the prescription of such drugs with no prior investigation to detect strains hampers an appropriate diagnosis, either due to lack of a suitable structure for the purpose or to misinformation and carelessness by health professionals (Souza et al., 2010).
In veterinary medicine, tetracyclines and sulfonamides can only be prescribed to treat systemic infections.Their application as performance additives or animal feed preservatives is forbidden.Additionally, the production, handling, fractioning, commercialization, importation and use of chloramphenicol in veterinary medicine and in animal feed production are also prohibited.These restrictions may explain the susceptibility of most strains tested against chloramphenicol, co-trimoxazole, and tetracycline, in the present study.However, it is possible to conjecture that these drugs are used indiscriminately in broiler production, since the strains isolated from hamburgers made with beef and chicken meat were resistant to these antimicrobials.

Conclusion
A high percentage of Salmonella spp.strains isolated from beef and beef and chicken hamburgers resistant to the different antimicrobials tested was observed.This raises concerns as to the high risk for salmonellosis in humans, when these food items are consumed underdone or by cross-contamination with other food items during preparation.Another important aspect in terms of public health is that these infections may become even more severe due to de lower efficacy of antimicrobials in treatments, both in humans and animals.
The results point to the need for more effective strategies to reduce the risk of proliferation and contamination of food by resistant strains, which include the monitoring and inspection concerning the use of antimicrobials in the husbandry of food producing animals.Additionally, as a means of prevention, hamburgers should be cooked and handled appropriately so as to prevent outbreaks of Salmonella spp.infections and the spread of multiresistant strains.

Table 1 .
Absolute and percentage susceptibility of the 45 Salmonella spp.strains isolated from hamburgers to 12 different antimicrobial drugs.

Table 2 .
Absolute and percentage susceptibility of the 45 Salmonella spp.strains isolated from beef and beef and chicken hamburgers to 12 different antimicrobial drugs.