High level aminoglycoside resistance in Enterococcus, Pediococcus and Lactobacillus species from farm animals and commercial meat products

Abstract

Inappropriate use of aminoglycosides in animal husbandry has led to the selection and emergence of high-level aminoglycoside resistance (HLAR) in lactic acid bacteria (LAB). The objective of this study was to assess the presence of aminoglycoside resistant LAB in farm animals and meat products. Gentamicin resistant LAB (n = 138) were selectively isolated from 50 different meat and farm animal sources. These native isolates of LAB were subsequently characterized for their minimum inhibitory concentration to seven different aminoglycoside antibiotics. HLAR to gentamicin, kanamycin and streptomycin was found to be 38 %, 45 % and 15 %, respectively. Selected cultures of LAB were identified by random amplified polymorphic DNA (RAPD)-PCR and 16S rDNA gene sequencing. Subsequent detection for the presence of nine aminoglycoside modifying genes [aac(6′)Ie-aph(2″)Ia, aph(3′)IIIa, aad6, ant(6)Ia, ant(9)Ia, ant(9)Ib, aph(2″)Ib, aph(2″)Ic and aph(2″)Id] was carried out by PCR. The Enterococcus spp. (n = 64) and Lactobacillus plantarum (n = 6) isolated from farm animals and chicken sausages, respectively, were positive for the bifunctional gene, aac(6′)Ie-aph(2″)Ia in PCR. In addition, Enterococcus sp. (n = 17), Lactobacillus plantarum (n = 3), and Pediococcus lolii (n = 2) possessed the kanamycin resistance gene, aph(3′)IIIa. Other LAB viz. Enterococcus faecalis (n = 2), E. faecium (n = 2) and L. plantarum (n = 1) harbored the streptomycin resistance gene, aad6. The integrase (int) gene, characteristic to Tn 916-Tn 1545 was detected in Enterococcus faecalis CS11+ and Enterococcus cecorum I40a suggesting its involvement in antibiotic resistant gene transfer. Besides, strains of L. plantarum, a species used as probiotic, isolated in this study showed the occurrence of aph(3′)IIIa as well as aac (6′)Ie-aph(2″)Ia genes that could be of concern in human health. The findings of the study highlight the spread and emergence of multi-resistance genes for aminoglycoside antibiotics among beneficial LAB.

Introduction

The spread of resistance to antibiotics among bacteria is alarming. In the global scenario, the development and spread of resistance to clinically significant drugs in commensal bacteria is associated with their improper use in animal husbandry (Mathur and Singh 2005). The situation is of serious concern when it comes to lactic acid bacteria (LAB), which are beneficial microflora that can develop resistance to life-saving aminoglycoside drugs (Jackson et al. 2010). Aminoglycosides are regarded as vital drugs for the treatment of life-threatening infections (United States Pharmacopeial Convention 2008). High level resistance to aminoglycosides in bacteria may lead to ineffective therapeutic crisis. Many countries have banned the administration of certain antibiotics in animal husbandry due to their preferred usage in human medicine (Shwarz et al. 2000). However, aminoglycosides are recommended for therapy and prophylaxis in farm animals owing to their efficient bactericidal mode of action against Gram-negative and Gram-positive bacteria (United States Pharmacopeial Convention 2008).

Bacterial resistance to aminoglycosides occurs due to mutations, impaired transport and acquired resistance (Ramirez and Tolmasky 2010). The most common mode of aminoglycoside resistance in Gram-positive bacteria is the acquisition of aminoglycoside-modifying genes (Bismuth and Courvalin 2010). Clinically, the bifunctional gene aac(6′)Ie-aph(2″)Ia confers resistance to almost all aminoglycosides except streptomycin. It has been associated with high-level gentamicin resistance (HLGR) and high level kanamycin resistance (HLKR) with minimum inhibitory concentration (MIC) values >500 μg/mL (Murray 1990; Del Campo et al. 2000; Donabedian et al. 2003; Jackson et al. 2004; Bismuth and Courvalin 2010). This bifunctional gene is highly prevalent among clinical strains of Staphylococcus and Enterococcus spp. (Culebras and Martinez 1999), and its frequent spread among Gram-positive organisms has been attributed to its lower G+C content (Byrne et al. 1989). Of late, the selective advantage of this fused gene has been related to the structure and rigidity of the two encoded functional enzymes (Boehr et al. 2004). The monofunctional genes [aph(2″)Ib, aph(2″)Ic and aph(2″)Id] depicted lower MIC values (≤500 μg/mL) for gentamicin, while aph(2″)Ie was observed with HLGR (≥1024 μg/mL) in enterococci (Chow et al. 1997; Tsai et al. 1998; Kao et al. 2000; Chen et al. 2006). The aminoglycoside resistance genes mentioned above are known to modify 4,6 deoxystreptamines, especially clinically significant antibiotics such as amikacin, gentamicin, kanamycin and tobramycin (Chen et al. 2006; Bismuth and Courvalin 2010). Other resistance-conferring genes found in Gram-positive organisms are kanamycin modifying aph(3′)IIIa, streptomycin modifying aad6, ant(6)Ia and spectinomycin modifying ant(9′)Ia, ant(9′)Ib (Bismuth and Courvalin 2010). A high-level streptomycin resistance (HLSR) in enterococci, despite the absence of aad6 or ant(6)Ia genes has been explained due to a mutation in the 30S ribosome (Simjee and Gill 1997). Therefore, high MIC levels may not necessarily correlate with the presence of the above-mentioned resistance genes.

In the past decade, the rise and spread of aminoglycoside resistance in Enterococcus sp. from domestic and other farm animals has been documented (Donabedian et al. 2003; Jackson et al. 2004, 2009, 2010; Ramos et al. 2012; Klibi et al. 2014). Enterococci from clinical settings and farm animals from the same region were positive for aminoglycoside resistance, and the genetic relatedness signified their rapid spread among the human population (Donabedian et al. 2003). Earlier, Tenorio et al. (2001) described the emergence of aac(6′)Ie-aph(2″)Ia in Lactobacillus acidophilus, Lactobacillus salivarus and Pediococcus acidilactici from fecal samples of healthy pigs and pets, which is the only report of its kind to date to have described the gene in Lactobacillus and Pediococcus sp. Documented studies on the occurrence of aminoglycoside resistance in LAB from food origin have been very limited (Hammerum et al. 2010).

LAB are in demand industrially, because of their beneficial aspects in fermented food and pharma-based therapeutics (Popova et al. 2012). The quality presumption safety (QPS) status, which ensures the absence of acquired resistance, has to be attained in order to use LAB in food or probiotic product development (EFSA 2012). However, this norm becomes diluted in the case of food products, in view of the scale of operation, wherein such stringent tests to avoid the inclusion of resistant LAB are ignored. Most countries in Asia and Africa lack proper guidelines on the judicious use of antibiotics and there is much use of low cost antibiotics in animal husbandry that is not being monitored. The persistent use of aminoglycoside antibiotics such as apramycin, neomycin, streptomycin in farms may co-select for antibiotic resistance in LAB and pathogenic species via the food chain (Guardabassi et al. 2004). In such instances, antibiotics deliver a competitive advantage, thereby triggering the acquisition and rapid spread of resistance genes via horizontal gene transfer (Nielsen et al. 2014). The hazards of acquired resistance and its rampant spread in nosocomial enterococci is of major concern (Kuhn et al. 2005; Hammerum et al. 2010). Such findings illustrate that there is a high risk of acquisition and transfer of aminoglycoside resistance by LAB via the food chain. Thus, extreme care should be taken to avoid the emergence of aminoglycoside-resistant LAB, as they can be a potent source of antibiotic resistance genes (Mathur and Singh 2005). As a result, protocols to monitor the development of aminoglycoside resistance among LAB from farm animals are required.

From the above understanding of the risk related to aminoglycoside resistance and its spread via the food chain, an attempt was made in the present study to evaluate the presence of aminoglycoside resistance in native LAB from farm and meat origin in the Mysore region of Karnataka, India. We report the widespread presence of the aminoglycoside resistant bifunctional gene along with the kanamycin and streptomycin modifying genes in different native isolates of Enterococcus sp., Lactobacillus plantarum, and Pediococcus lolii.

Material and methods

Media, antibiotics, chemicals and control strains

The Iso-sensitest broth was purchased from Thermo Fisher Scientific (Basingstoke, UK). de Mann, Rogosa and Sharpe (MRS) medium and antibiotics such as amikacin, apramycin, gentamicin, kanamycin, neomycin, streptomycin and spectinomycin were procured from Hi-Media Laboratories (Mumbai, India). Antibiotic powder was weighed as required and dissolved in water. The dNTPs mix, 25 mM MgCl₂, lysozyme and 10 Kb DNA ladder were procured from GeNei (Merck, Bangalore, India). Proteinase K, Trizma base, Taq DNA polymerase, primer M13, aminoglycoside resistance gene specific primers, 16S rDNA primer and PCR purification kit were purchased from Sigma-Aldrich (St. Louis, MO). The primers used in this study are detailed in Table 1. Sodium hydroxide, sodium chloride, bromophenol blue, sodium dodecyl sulphate, EDTA disodium salt and phenol were obtained from Sisco Research Laboratories (Mumbai, India). The control strains Enterococcus faecalis JH2-2 and E. faecalis RE-25 were obtained from C.M.A.P Franz, Federal Research Centre for Nutrition and Food, Institute of Hygiene and Toxicology, Karlsruhe, Germany.

Table 1 Primers used in this study

Selective screening of gentamicin-resistant LAB

Aminoglycoside resistant LAB from the intestines of slaughtered animals [chicken (n = 25), sheep (n = 7), beef (n = 3), pigs (n = 5)] and meat products [raw meat (n = 4) and chicken sausages (n = 3), hams (n = 2), salami (n = 1)] were isolated by selective screening on MRS agar containing gentamicin (64 μg/mL). Briefly, to describe the isolation of LAB, homogenized samples in 10 g quantities were emulsified in 90 mL 0.1 % peptone water and aliquots of appropriate serial dilutions were plated on gentamicin-containing MRS agar. Plates were incubated for 48 h at 37 °C and typical gentamicin resistant colonies were selected at random and isolated. In addition, the viable count of LAB was recorded as CFU/g. The isolates were confirmed to be LAB by Gram-staining and catalase tests. Purified colonies were stored at −40 °C in MRS broth containing 20 % glycerol. Further, LAB isolates were tested for carbohydrate fermentation, growth at different temperatures (15, 45 and 50 °C), pH (4.5 and 5.5) and 6.5 % NaCl concentration and tentatively identified to genus level as described in Bergey’s manual of Determinative Bacteriology (Holt et al. 1994).

Determination of minimum inhibitory concentration

The minimum inhibitory concentration (MIC) of individual antibiotics like amikacin, apramycin, gentamicin, kanamycin, neomycin, streptomycin and spectinomycin was determined by the broth microdilution method using 96-well microtiter plates in LAB susceptibility test medium (LSM-90 % Iso-sensitest broth supplemented with 10 % MRS broth) according to the CLSI (2012) standards. Quantification was performed using an ELISA reader (Model No:1510, Thermofischer Scientific, Vantaa, Finland) after incubation for 24 h at 37 °C. The lowest concentration of antibiotic at which no growth was observed was taken as the MIC values of different aminoglycosides (CLSI 2012). The strain Enterococcus faecalis JH2-2 was used a negative control in MIC tests. E. faecalis RE-25 was used as a positive control for kanamycin and streptomycin resistance.

Random amplified polymorphic DNA polymerase chain reaction (RAPD-PCR)

The total DNA extracted (Mora et al. 2000) from individual LAB isolates was employed as a template for initial grouping of LAB by RAPD-PCR. The PCR reactions were carried out according to the method described by Schillinger et al. (2003). The 25 μL reaction mixture for all the PCR amplification carried out in this study consisted of 0.4 μM M13 primer, 25 ng/μL total DNA, 1.5 mM MgCl₂, 0.5 mM dNTPs, 1X PCR buffer, 0.3 U Taq DNA polymerase and DNase-free sterile water. The amplified products were examined on a 1.8 % agarose gel in TAE buffer at a voltage of 65 V for 4–5 h. The RAPD profiles were evaluated using the Quantity one® software (Bio-Rad, Milan, Italy).

16S rDNA amplification

To identify the LAB isolates, 16S rDNA gene amplification were performed according to Kimprasit et al. (2013). Representative cultures from each group with similar RAPD banding patterns as well as tentatively identified Enterococcus and Lactobacillus isolates that showed varied MIC levels for gentamicin, kanamycin and streptomycin were also considered for 16S rDNA gene sequencing. The sequences obtained were further subjected to BLAST search (Altschul et al. 1997) analysis for comparison with known sequences from the public database.

Detection of aminoglycoside resistance genes

The presence of aminoglycoside resistance genes in gentamicin resistant LAB was evaluated by PCR employing total DNA as a template. The primer sequences for aac(6′)Ie-aph(2″)Ia, aph(2″)Ic, aph(2″)Id and int were taken from literature (Doherty et al. 2000; Donabedian et al. 2003). As detailed in Table 1 the primers were designed based on the GenBank accession numbers of aph(3′)IIIa (X92945.2), aad6 (X92945.2), ant(6)Ia (GQ900487.1), ant(9)Ia (AY764268.1), ant(9)Ib (M69221.1), aph(2″)Ib (AF207840.1). The isolates positive for the presence of aminoglycoside resistance genes were further assessed for transposon integrase (int) gene pertaining to Tn 916-Tn 1545 family. The amplified product was checked on 1–1.2 % agarose gel in 1X TAE buffer as described previously.

Gene sequencing, nucleotide accession numbers, and phylogenetic tree construction

The amplified products of the aminoglycoside resistance genes and 16S rDNA were purified using a PCR purification kit and sequenced at Amnion Biosciences (Bangalore, India). The GenBank accession numbers obtained for 16S rDNA nucleotide sequences of the LAB cultures [I40a, CS31+, S9, chl, chl3, C11(5)], [J12G1, B20G3, P1, CS32+, 39-1, CHB1], [33-1, CS12+, CS13+, C26a, CSG-8, L32] and J10G3 are [KJ420396–401], [KM016947–952], [KM016954–959] and KJ995522, respectively. Likewise, accession numbers for the sequences of the bifunctional gene in isolates I40a, CS31+, 5, S9 and C11(5) are KJ420402–406. The accession numbers for kanamycin and streptomycin resistance genes aph(3′)IIIa and aad6 in cultures CS11+ and 33-1 are KJ995521 and KM016953, respectively. The int gene detected in the culture I40a was also deposited with GenBank under the accession number KJ420407. To construct the phylogenetic tree, DNA fingerprinting by RAPD-PCR was compared and a dendrogram created by the cluster analysis method using dice coefficient values by the NTYSIS software [Version 2.02e] (Jamshidi and Jamshidi 2011). Patterns were grouped with unweighted pair group algorithm averages (UPGMA).

Results and discussion

A total of 138 LAB cultures were isolated by selective screening against gentamicin from the samples analyzed. The LAB isolates were tentatively identified by initial biochemical characterization and classified into enterococci (n = 101), lactobacilli (n = 32) and pediococci (n = 5) species. The LAB isolates were assessed by determination of MIC to seven different aminoglycosides. The MIC values obtained for gentamicin, kanamycin and streptomycin were compared with established breakpoint levels (EFSA 2012); As stated by EFSA (2008), LAB tolerant to these three antibiotics lowers the risk of acquired resistance to other aminoglycoside. In this study, LAB were subjected to MIC determination to other aminoglycosides to investigate the level of resistance. Likewise, the percentage of native LAB isolates resistant to apramycin, streptomycin, neomycin, amikacin, spectinomycin, kanamycin and gentamicin were evaluated. A MIC value of ≤64 μg/mL for apramycin, streptomycin, neomycin, amikacin, spectinomycin, kanamycin and gentamicin was observed in 25 %, 22 %, 37 %, 42 %, 18 %, 31 % and 41 % of LAB isolates, respectively. Similarly, 68 %, 63 %, 48 %, 39 %, 77 %, 24 % and 21 %, respectively, of LAB isolates displayed MIC values of ≥128–≥512 μg/mL and 7 %, 15 %, 15 %, 19 %, 5 %, 45 % and 38 %, respectively, showed MIC in the range ≥1024–≥4096 μg/mL. The MIC results inferred that LAB obtained in this study exhibited low (≤64 μg/mL), moderate (≥128–≥512 μg/mL) and high (≥1024–≥2048 μg/mL) level resistance to aminoglycosides, as also documented in earlier studies (Murray 1990; Simjee and Gill 1997; Donabedian et al. 2003; Bismuth and Courvalin 2010). In comparison to the MIC levels, the majority of isolates showed resistance in the range of ≥128–≥512 μg/mL to all the aminoglycosides, except amikacin, kanamycin and gentamicin. However, a few LAB cultures exhibited high level resistance to all the aminoglycoside antibiotics. The low level resistance among a few of the LAB isolates observed in our study could be attributed to its anaerobic nature (Lopes et al. 2003). However, these isolates were susceptible to higher doses and do not exhibit acquired resistance. Enterococci from our study showed HLGR (38 %), HLKR (45 %) and HLSR (15 %), which was consistent with earlier studies (Lopes et al. 2003; Del Campo et al. 2005; Ramos et al. 2012).

The resistance pattern to clinically relevant aminoglycosides of enterococci, lactobacilli and pediococci is shown in Table 2. Enterococci (n = 101) was the most prevalent, followed by lactobacilli (n = 32) and pediococci (n = 5) isolates. A significant number of LAB isolates from poultry were resistant to gentamicin (n = 67), kanamycin (n = 55) and streptomycin (n = 69), followed by sheep, pork and beef. Earlier studies indicated a higher prevalence of HLAR in enterococci from pig, cattle and sheep (Ramos et al. 2012; Klibi et al. 2014). The higher frequency of aminoglycoside resistance in enterococci from poultry indicates the significant use of these antibiotics in poultry farms. Pediococci isolated from poultry and pork were resistant to gentamicin (n = 4), kanamycin (n = 4) and streptomycin (n = 3). In contrast, meat products such as chicken sausages were found to be contaminated with aminoglycoside-resistant enterococci and lactobacilli. Lactobacilli derived from meat products only were resistant to gentamicin (13 %) and kanamycin (19 %). Eleven lactobacilli were observed with MIC values of ≥128 μg/mL for streptomycin.

Table 2 Prevalence of aminoglycoside-resistant lactic acid bacteria (LAB) obtained from different sources. MIC Minimal inhibitory concentration. n Number of isolates

RAPD-PCR was carried out to group the aminoglycoside-resistant LAB that were selected based on the breakpoint MIC values as indicated in Fig. 1. Based on the RAPD profiles, a dendrogram was constructed and species-specific clusters of E. faecalis, E. faecium, E. avium, E. hirae, E. durans, E. cecorum, L. plantarum and Pediococcus lolii were observed using the dice coefficient parameter and UPGMA. The banding patterns revealed that the majority of gentamicin- and kanamycin-resistant isolates from chicken intestine, followed by sheep and chicken sausages, matched those of E. faecalis. Representative LAB isolates with similar RAPD profiles from each cluster were subjected to taxonomic identification by 16S rDNA gene sequencing. Enterococcus, Lactobacillus and Pediococcus isolates from different sources with low, moderate and higher MIC values for gentamicin, kanamycin and streptomycin, as well as the presence of an aminoglycoside-resistant gene, was the criteria set to select them for 16S rDNA identification. Additionally, isolates of Enterococcus and Lactobacillus derived from chicken meat, chicken intestine and chicken sausage with similar banding patterns were also sequenced since they exhibited varying MIC levels. In brief, E. faecalis CS12+ showed HLGR, HLSR and a moderate MIC (512 μg/mL) for kanamycin, while two isolates, E. faecalis CHL3 and E. faecalis CHL showed moderate MIC values (≥128–≥512 μg/mL) to the above mentioned three aminoglycosides. E. faecalis 33-1 and E. faecalis CS13+ with comparable RAPD profiles were sequenced as they exhibited high level resistance to all seven aminoglycosides (≥2048–≥4096 µg/mL). Similarly, in the case of Lactobacillus species, CSG-8, CSG-21, L32, C11(5) and S9 showed different MIC levels (range 4–128 μg/mL) for gentamicin and for kanamycin (16–128 μg/mL), while L. plantarum Lb6 was sensitive to gentamicin and exhibited kanamycin resistance (≥128 μg/mL). Since it is important to identify contamination as well as the risk associated with the presence of aminoglycoside-resistant LAB in meat processing units, it was thought appropriate to identify the above-mentioned isolates. The phenotypically resistant LAB that were identified in this study include: E. faecalis (n = 52), E. faecium (n = 2), E. durans (n = 4), E. hirae (n = 6), E. avium (n = 3), E. cecorum (n = 1), P. lolii (n = 2) and L. plantarum (n = 32). Thus, E. faecalis species was prevalent in all food animals and chicken sausage, which is in line with earlier studies (Silva et al. 2012; Ramos et al. 2012). However, it was noted that E. faecalis was frequently identified from poultry sources. Aminoglycoside-resistant E. faecium, E. durans, and E. cecorum were isolated from poultry, while E. hirae was identified from chicken, beef and sheep. In similar studies conducted by Kuhn et al. (2003), the association of E. faecalis with poultry and E. hirae with pigs and cattle was described.

Fig. 1

Dendrogram showing the coefficient of similarity of the aminoglycoside-resistant lactic acid bacteria (LAB). The representative cultures shown in bold carried the aminoglycoside resistance genes and were subjected to 16S rDNA sequencing. The MIC values indicated in bold represent the isolates harboring the respective aminoglycoside resistance genes. Gm gentamicin, Km kanamycin, Str streptomycin

From our studies, it was evident that the food chain comprises diverse groups of enterococcal populations. Similar observations have been reported with Enterococcus species found in food animals (Kuhn et al. 2003; Poeta et al. 2005; Silva et al. 2012). Aminoglycoside resistant enterococcal species associated with nosocomial infections (Murray 1990) like E. avium, E. durans, E. faecalis, E. faecium and E. hirae from food sources were also observed. The 16S rDNA sequence data of the Pediococcus and Enterococcus isolates, namely P1 and I40a, exhibited 100 % homology to P. lolii (accession no. NR041640) and 98 % to E. cecorum (accession no. NR024905). It was interesting to note HLAR in lesser-known species like P. lolii and E. cecorum from farm animals because E. cecorum was regarded as a part of the avian gut microflora until it was found to be an emerging pathogen in unhealthy chickens and infected patients (Hseueh et al. 2000; Stalker et al. 2010). The occurrence of intestinal microbiota such as P. lolii and L. plantarum in pork intestine and chicken sausage, respectively, are worth mentioning. The occurrence of HLAR in rare microbiota relates significantly to selective pressure induced by the improper use of antibiotics in farms. Additionally, L. plantarum, part of the characteristic microbiota of the human gut and also considered as a potential strain for probiotic applications, can also be regarded as a possible reservoir of aminoglycoside resistance.

The phenotypically resistant LAB isolates were subjected to PCR for the presence of gentamicin, kanamycin and streptomycin resistance genes previously found in different enterococcal species of food origin. Isolates of E. faecalis (n = 44), three each of E. durans and E. avium, as well as one each of E. hirae and E. cecorum showed MIC values associated to HLAR and were found to carry the aac(6′)Ie - aph(2″)Ia gene (Table 3). Interestingly, the bifunctional gene was also identified in E. faecalis (n = 5), E. hirae (n = 5), E. avium (n = 1) and E. durans (n = 1), with moderate gentamicin MIC values in the range of 128–256 μg/mL, similar to those found in an earlier investigation (Donabedian et al. 2003). To the best of our knowledge, the presence of this gene in E. cecorum and E. avium, which are otherwise considered to be the normal microbiota of the chicken intestine, has not been documented previously. This study highlights the emergence of HLAR in normal microbiota and its selection in the farm environment, and the results portray an increasing risk of aminoglycoside gene transfer from farm animals to the human population via the food chain. The PCR detection of the kanamycin resistance gene, aph(3′)IIIa revealed its prevalence in E. faecalis (n = 11), E. faecium (n = 2) and E. avium (n = 4). The kanamycin modifying gene was found frequently in enterococci, along with aac(2″)Ie-aph(2″)Ia, which was previously associated with HLKR with MIC values of ≥1024 μg/mL (Del Campo et al. 2000; Lopes et al. 2003; Ramos et al. 2012). However, in our study it was detected in one strain of E. faecalis with MIC of 500 μg/mL. Also, a streptomycin resistance gene (aad6) was found in two isolates each of E. faecalis and E. faecium, while several isolates displayed HLSR.

Table 3 PCR detection of aminoglycoside resistance genes among native LAB isolates

In the case of L. plantarum (n = 6), varying MIC levels (4–128 μg/mL) were observed against gentamicin; this LAB was also found to possess the widespread aac(6′)Ie-aph(2″)Ia gene. The aph(3′)IIIa kanamycin resistance gene and aad6 streptomycin modifying gene were found in three and one isolate, respectively. Isolates of L. plantarum revealed the presence of aac(6′)Ie-aph(2″)Ia, aph(3′)IIIa and aad6 genes in earlier investigations (Ouoba et al. 2008; Shao et al. 2015). The high MIC values and presence of gentamicin, kanamycin and streptomycin resistance genes in L. plantarum isolated from chicken sausages point out the unanticipated risks relating to its generally regarded as safe (GRAS) status.

Pediococcus lolii (n = 2) harbored the aph(3′)IIIa gene. That no prior existence of HLKR in P. lolii was found in the literature signifies the transmission of kanamycin resistance despite strict prohibition in its usage in animal husbandry (United States Pharmacopeial Convention 2008). Furthermore, acquired resistance and display of HLKR from food animals call into question the prophylactic use of aminoglycosides in farm animals as they are used in human therapy (Guardabassi et al. 2008). All the isolates with MIC above the breakpoint values were tested for gentamicin resistance genes such as aac(6′)Ie-aph(2″)Ia, aph(2″)Ib, aph(2″)Ic and aph(2″)Id. Contrary to an earlier investigation (Donabedian et al. 2003), the bifunctional gene was the only gentamicin-modifying gene predominant in this study. However, one isolate of E. faecalis from chicken sausage showed gentamicin MIC values of 1024 μg/mL, which failed to generate amplification for any of the above genes and therefore may contain a novel resistance mechanism or even a gene that needs to be investigated. Nine percent (n = 12) of the LAB isolated from food origin were observed with both the aac(6′)Ie-aph(2″)Ia and aph(3′)IIIa genes conferring resistance to clinically significant gentamicin and kanamycin. One isolate each of E. faecalis and L. plantarum was observed to harbor all three resistance genes [aac(6′)Ie-aph(2″)Ia, aph(3′)IIIa and aad6] conferring resistance to gentamicin, kanamycin and streptomycin, respectively, which was also found to correlate with higher MIC values.

The transmission of the bifunctional and other genes could be related to transposons and horizontal gene transfer among the normal microflora. Studies have revealed the involvement of a transpositional genetic element (Tn 5281) in E. faecalis similar to the staphylococcal transposons Tn 4001 and Tn 4031 (Hodel-Christian and Murray 1991). The presence of transposons in clinical strains of E. faecalis (Ferretti et al. 1986) and E. faecium (Eliopoulos et al. 1988), indicates the possibility of a direct gene exchange between species and a common ancestral origin of gentamicin resistance. Conjugative elements of the Tn 916-Tn 1545 family are responsible for the dissemination of many antimicrobial resistance genes, usually related to tetracyclines and macrolides (Thumu and Halami 2012, 2013). In the present study, the integrase (int) gene inherent to the Tn 916-Tn 1545 family was detected in E. faecalis (n = 2) and E. cecorum (n = 1) isolated from chicken sausage and chicken intestine, respectively (Table 3). Since kanamycin resistance has been found only rarely in these transposons, it was interesting to note that only one isolate E. faecalis CS11+ contained both the int and kanamycin resistance [aph(3′)IIIa] genes that can be found in the Tn 916-Tn 1545 family. Further studies are required to investigate the localization and expression of aac(6′)Ie-aph(2″)Ia, aph(3′)IIIa and aad6 genes among LAB isolates.

Overall in this study, we have observed aminoglycoside resistance genes in LAB isolated from chicken, sheep and beef, while gentamicin resistance was not found in LAB derived from pork. Spectinomycin-modifying genes ant(9)Ia and ant(9)Ib were also not found. Kanamycin has not been recommended for veterinary medicine; however, aph(3′)IIIa gene was the second most detected gene, in combination with the bifunctional gene. Combinations of aminoglycoside-modifying genes have been described in different species of enterococci (Jackson et al. 2004, 2009). Likewise, our study revealed the presence of multiple aminoglycoside resistance genes in enterococci and lactobacilli. The presence of multi-resistance genes to aminoglycoside in known nosocomial pathogenic cultures like E. faecalis, E. faecium, E. durans and E. hirae could be a great threat in developing nations. In addition, its emergence in the normal microflora (E. cecorum and E. avium) of farm animals can be compounded by the lack of proper monitoring of antibiotic use. Since enterococci are capable of acquiring and transferring genes horizontally, a high risk lies in the rapid emergence of the bifunctional gene among other genera of LAB as well as pathogenic strains in the human gut (Mathur and Singh 2005; Devirgiliis et al. 2011). Acquired resistance in L. plantarum from meat products indicates the contamination of food by resistant strains that are otherwise regarded as harmless. The use of an aminoglycoside in farms may enhance the risk of acquiring resistance to other antibiotics of the same class (Guardabassi et al. 2008). For instance, inappropriate use of apramycin was reported to trigger cross-resistance to gentamicin among beneficial microflora (Yates et al. 2004). In a recent study, L. plantarum induced with streptomycin for a period of 30 days was described to show higher MIC values for gentamicin, kanamycin and neomycin to which it was previously susceptible (Shao et al. 2015). Likewise, use of gentamicin in animal husbandry may have propagated the selection and acquisition of the bifunctional gene in LAB that exhibits cross-resistance to other aminoglycosides (except streptomycin).

Conclusion

The overuse or underuse of antibiotics in livestock, as well as improper filtration of waste water from clinical settings and drug manufacturers could be the major reasons underlying the plague of antibiotic resistance in developing economies like India. In addition, a lack of initiative to investigate and produce consistent data on the prevalence of developing resistance in these nations is a major drawback. There have been only few reports on aminoglycoside resistant LAB from different parts of the world, while these antibiotics are used routinely for prophylaxis and treatment in both humans and animals. The sudden emergence of resistance genes against broad spectrum aminoglycosides among the LAB via the food chain cannot be ignored. If the situation prevails, it could threaten the existing efficacy of broad spectrum activity aminoglycosides, thus increasing the demand for newer antibiotics.