Low temperatures do not impair the bacterial plasmid conjugation on poultry meat

Abstract

Conjugation plays an important role in the dissemination of antimicrobial resistance genes. Besides, this process is influenced by many biotic and abiotic factors, especially temperature. This study aimed to investigate the effect of different conditions of temperature and storage (time and recipient) of poultry meat, intended for the final consumer, affect the plasmid transfer between pathogenic (harboring the IncB/O-plasmid) and non-pathogenic Escherichia coli organisms. The determination of minimal inhibitory concentrations (MIC) of ampicillin, cephalexin, cefotaxime, and ceftazidime was performed before and after the conjugation assay. It was possible to recover transconjugants in the poultry meat at all the treatments, also these bacteria showed a significant increase of the MIC for all antimicrobials tested. Our results show that a non-pathogenic E. coli can acquire an IncB/O-plasmid through a conjugation process in poultry meat, even stored at low temperatures. Once acquired, the resistance genes endanger public health especially when it is about critically and highly important antimicrobials to human medicine.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s42770-023-01230-9.

Introduction

Many factors drive the emergence of antimicrobial resistance (AMR), especially the indiscriminate use of these drugs in food-producing animals [1, 2]. It represents a severe problem for public health and the environment due ability of AMR to be transferred by pathogens to humans through direct contact with animals that harbor antibiotic-resistant bacteria or indirect contact like contaminated food products [2, 3].

Food safety is one of the main concerns in the poultry industry [4], which step of the poultry processing plant represents a worldwide risk for microbial contamination, such as Salmonella spp., Campylobacter spp., and Escherichia coli (E. coli), and other spoilage microorganisms as well [2, 5–7]. Furthermore, microbial contamination can persist in the environment, contaminate equipment, and consequently cross-contaminate poultry products [8–10].

The occurrence of AMR can happen through point mutations [11] or via horizontal gene transfer (HGT) by mobile genetic elements (MGEs) such as transposons and conjugative plasmids, which are accountable for the dissemination of AMR as well [12]. This last one is the major HGT mechanism, which is mediated by a type IV secretion system (T4SS) in Gram-negative bacteria [13], through the transference of MGEs across direct contact from a donor to a recipient bacterium [14]. This process can be influenced by biotic and abiotic factors such as temperature, pairing time, bacterial growth phase, and pH [15].

Despite the limitation factors, conjugation has been identified in many different environments such as soil, sewage, seawater, biofilms, human gut, pig feces, and poultry litter [16–19], and in domestic stuff as well, like cotton hand towels, and minced meat [16]. The bacterial conjugation mechanism confers advantageous metabolic properties, for example, antimicrobial and heavy metal resistance, and virulence [20], but in some cases, the acquisition of a plasmid imposes a fitness cost to hosting bacteria [21].

Based on the multifactorial conditions of conjugation, especially when their occurrence is in the domestic human environment, and between food-borne pathogens as well, we hypothesized if a pathogenic E. coli harboring a plasmid with antimicrobial resistance genes (ARGs) could conjugate with a non-pathogenic E. coli in poultry meat storage in low temperatures (4 °C) to simulate what happens during the handling of food intended for the final consumer. Furthermore, it was analyzed if the recipient used for storage of the aliment could work as a carrier/enabler of transconjugant bacteria if the process occurs.

Materials and methods

Bacterial strains and growth conditions

The donor bacteria was the pathogenic human-isolate E. coli H2332 harboring the plasmids pH2332-166 (accession no. KJ484626), which has genes conferring resistance against aminoglycosides (aadA1b; strAB), amphenicols (catA1), β-lactams (bla_TEM-1), macrolides (mph-B), tetracyclines (tetR; tetA), trimethoprim (dfrA1) and sulfonamides (sul1; sul2), and pH2332-107 (accession no. KJ484627) which has the bla_CTX-M-1 gene related to β-lactam resistance; this bacterium being classified in the pathogenic phylogroup D-O24:H26-ST57/ST350 complex [22]. Both plasmids have Integrative and Conjugative Elements (ICEs): rve, virb4, Phage_integrase, and TIGR02249 were found in pH2332-166 plasmid (51 protein sequences into 46,230 bp), while 116 protein sequences have been found into 103,578 bp of plasmid pH2332-107. All the ICEs found in both plasmids were identified through ICEFinder tool [23], and they are detailed in Supplementary Material S1.

The rifampicin/nalidixic acid-resistant non-lactose fermenting and non-pathogenic E. coli K12 J62 strain (accession no. GCA_017681865.1) was used as the recipient [24]. Both bacteria were cultured in 10 mL of Lysogeny broth (LB) (BD Difco™, 240,230, EUA) overnight, then transferred to a new 10 mL LB, and cultured for three hours. After the incubation period, the bacterial population was adjusted using peptone water to a concentration of 1 × 10⁷ and 3 × 10⁷ colony-forming units (CFU) mL⁻¹ for the donor and the recipient strains, respectively.

Bacterial conjugation assay

Poultry meat defrosted at 4 °C for 3 h was used in the experiment, and then a swab over the entire surface of meat was tested in MacConkey agar (Oxoid®, CM0115, UK) supplement with 75 µg Rifampicin (Sigma-Aldrich®, code 13,292–46-1, R0700000, Brazil), and MacConkey agar supplement with 75 µg Rifampicin and 40 µg Ampicillin (Sigma-Aldrich®, code 69–52-3, A9518, USA) to analyze the presence of antimicrobial resistant bacteria in the aliment previously the experiment initiate, avoiding bias and possible contamination with undesirable bacteria in the future steps.

For the conjugation assay, first, 10 g of poultry meat was placed into a sterile plastic bag, and 5 mL of peptone water containing the donor strain (1 × 10⁷ CFU mL⁻¹) was added. After one hour at room temperature, liquid excess was withdrawn, and 5 mL of peptone water with the recipient strain (3 × 10⁷ CFU mL⁻¹) was added. After one hour at room temperature, liquid excess was withdrawn. Then, four different treatments mimicking the packaging of food in a refrigerator, after handling were prepared, plus a control group of conjugation: two treatments (T2 and T3) placed on sterile plastic and aluminum dishes, respectively, and kept at 4 ºC for 24 h; two treatments (T4 and T5) placed on sterile plastic and aluminum dishes and incubated at 4 ºC for 48 h; a control treatment (T1), with poultry meat incubated at 25 ºC for 24 h.

Frequency of conjugation

The colonies were enumerated and logarithmically transformed (CFU g⁻¹). The determination of the frequency of conjugation was performed according to Saraiva et al. [19] with modifications by dividing the number of transconjugants (CFU g⁻¹) by the number of total recipient bacteria (CFU g⁻¹). After incubation, the poultry meat was put inside a sterile bag, and a swabbing technique was performed on the dish's surface on which the food was in contact. Bag contents were homogenized in Phosphate-Buffered Saline (PBS 1 ×) (90 mL), and swabs were incubated in 3 mL E. coli broth (Sigma-Aldrich®, 110,765, Brazil). Right away, both were serially diluted 1:10 (v:v), and aliquots (100 µL) were plated in triplicates onto a MacConkey agar dish supplemented with 75 µg Rifampicin for recipient bacteria recovery, and MacConkey agar dish supplemented with 75 µg Rifampicin + 40 µg Ampicillin for transconjugant bacteria recovery. Bag contents and swabs with no bacteria enumeration results were incubated, and the value of 10² CFUg^-1 was assumed for positive samples obtained by enrichment for further analysis as described by Saraiva et al. [25]. Bacterial counts less than 100 CFU g⁻¹ (detection limit) of each assay were considered negative results. The CFU g⁻¹ values above the detection limit were normalized by log₁₀ for statistical analysis.

Recovery of transconjugant bacteria and PCR analysis

Separately, according to which treatment, the meat was incubated in 10 mL Escherichia coli bacterial suspension with 40 µg ampicillin at 37 °C for 16 h (T1-T5). Moreover, a swabbing technique was performed on the surface on which the food was in contact and incubated in 3 mL E. coli broth supplemented with 40 µg ampicillin at 37 °C for 16 h (T2-T5). After incubation, all samples were plated on MacConkey agar supplement with 75 µg of rifampicin and 40 µg of ampicillin for transconjugant bacteria recovery. Poultry meat is a product that harbors significant diverse microorganisms on its surface [26], therefore this protocol allowed the recovery of transconjugants only (ampicillin-resistant E. coli J62). A polymerase Chain Reaction (PCR) assay was performed for transconjugant confirmation, especially to define which plasmid the recipient bacteria received. We used a GoTaq Master Mix (Promega^©, M7123, Brazil) reaction containing 10 µL of master mix, 2,5 µM of each primer (Invitrogen^©, Brazil), 2 µL of DNA template, and 6 µL of HyPure™ Molecular Biology Grade Water (Cytiva^©, Brazil). Cycling conditions were initial denaturation at 95 °C for 3 min, followed by 30 cycles at 95 °C for 1 min, 60 °C for 40 s, and 72 °C for 1 min, with a final step of 72 °C for 7 min. The PCR results were analyzed by electrophoresis at 6 V/cm for 80 min in a 2% (w/v) agarose gel stained with SYBR® Safe DNA gel stain (Invitrogen^©, USA). Primer’s information for PCR reaction is detailed in Table S1.

Minimum inhibitory concentration (MIC)

The MIC test of ampicillin (Amp), cefotaxime (Ctx), ceftazidime (Caz), and cephalexin (Lex) in broth microdilution technique [27] was performed in technical triplicates using the recipient strain E. coli J62 before and after the conjugative assay of the plasmid. The strains E. coli ATCC 25922 [28] and E. coli H2332 (the plasmid donor) were used as reference strains.

Statistical analysis

Bacterial enumeration of both total recipient and transconjugant, as well as data on the frequency of conjugation, were analyzed by two-way ANOVA followed by Bonferroni’s multiple comparison test (p < 0.05), while MIC results were analyzed by one-way ANOVA (p < 0.05). The analyses were performed by means of GraphPad Prism® software version 9.5 for Mac OS.

Results

No difference in the enumeration of recipient bacteria (p > 0.05) was observed between the treatment groups (T2-T5) in both poultry meat and different surfaces; an exception was observed in the control group (T1) from poultry meat, which presented the highest recipient bacterial enumeration (p < 0.0001). No recipient bacteria were recovered from T1 above the detection limit (10² CFU g⁻¹) from different surfaces. In general, the highest recipient bacterial count was recovered from poultry meat compared with the counts from the surface (p < 0.005) (Table 1). Bacterial enumeration was not recovered above the detection limit for transconjugant bacteria, except from poultry meat in all treatment groups, by performing the recount, in which samples were considered 10² CFU g⁻¹. Therefore, conjugation frequencies were exclusively obtained from poultry meat in all treatments (Table 2). No bacteria have been recovered from swabs for all treatments (Tables 1 and 2). After PCR analysis we identified that all transconjugants had the pH2332-107 (Figure S1).

Table 1.

Recipient bacterial enumeration from poultry meat (treatments 1–5) and surface (treatments 2–5)

Recipient bacterial counts (Log of CFU g−1)
Treatments	Poultry Meat	Surface	Treatment Means
T1	8.564A	-	-
T2	7.245Ba	5.927b	6.586
T3	7.084Ba	5.765b	6.425
T4	7.361Ba	6.042b	6.701
T5	6.937Ba	5.619b	6.278
Source Means	7.438	5.838	6.638

-: No bacteria enumeration recovered above detection limit (10² UFC g⁻¹)

T1: control treatment; T2: plasmid conjugation assay placed on sterile plastic and kept at 4 ºC for 24 h; T3: plasmid conjugation assay placed on aluminum dishes and kept at 4 ºC for 24 h; T4: plasmid conjugation assay placed on sterile plastic and incubated at 4 ºC for 48 h; T5: plasmid conjugation assay placed on aluminum dishes and incubated at 4 ºC for 48 h. Means with different uppercase letters in the columns or different lowercase letters in the lines differ significantly by the two-way ANOVA with Bonferroni multiple comparison test at p < 0.05

Table 2.

Transconjugant bacterial enumeration (Log of CFU g⁻¹) and frequency of conjugation in poultry meat after swab and E. coli broth incubation

Poultry meat
Treatments	Transconjugant bacterial counts (Log CFU g−1)	Frequency of conjugation
T1	2	-6.5645
T2	2	-5.2269
T3	2	-5.1281
T4	2	-5.2575
T5	2	-5.0141

T1: control treatment; T2: plasmid conjugation assay placed on sterile plastic and kept at 4 ºC for 24 h; T3: plasmid conjugation assay placed on aluminum dishes and kept at 4 ºC for 24 h; T4: plasmid conjugation assay placed on sterile plastic and incubated at 4 ºC for 48 h; T5: plasmid conjugation assay placed on aluminum dishes and incubated at 4 ºC for 48 h

For the microdilution analysis, aiming to avoid any bias in the results, only transconjugants recovered from the poultry meat were used (treatments 1–5). In general, all transconjugants presented a significant increase in the MIC, as represented in Fig. 1.

Fig. 1.

Results of minimum inhibitory concentrations (MIC) of ampicillin, cephalexin, cefotaxime, and ceftazidime. H2332—Escherichia coli H2332; J62 – E. coli J62; EC1, EC4, EC8, EC9 and EC13—E. coli J62 transconjugant from treatments 1–5, respectively

After the plasmid acquisition, bacteria recovered from all treatments significantly decreased their susceptibility to tested beta-lactams when compared to the recipient strain: ampicillin, cephalexin, cefotaxime (p < 0.0001), and ceftazidime (p = 0.0154) (Fig. 1). Transconjugants had an improvement in the MIC, reaching the donor bacteria threshold for Amp with an increase of approximately 38-fold, Lex at least 19-fold, and Caz with an increase of at least fivefold. On the other hand, despite Ctx not having reached similar MICs of donor bacteria, it had a 298-fold increase in inhibitory concentration in comparison with the recipient strain, according to the CLSI breakpoint [28].

Discussion

The control of pathogens represents a constant challenge for the food industry, for instance, antimicrobial-resistant E. coli has already been isolated in diverse chicken products and critical points of poultry production systems [29, 30] endangering consumers. Additionally, these bacteria have been isolated from houseflies, and cosmopolitan pests, which represent a greater risk to consumer health [31] and could act vector-borne and contaminate food during its preparation. Therefore, we evaluated the influence of both temperature and different dish materials in the conjugative plasmid transfer. We obtained the recipient bacterial counts in the highest numbers from poultry meat at 25 °C, similar to a previous study in which the reduction of temperature was associated with the reduction of the growth rate of E. coli O157:H7 in ground beef [32]. Moreover, no difference was observed in the recipient bacteria counts when compared time of refrigeration, neither from poultry meat nor from dish materials.

At a temperature of 4 °C, we did not obtain a transconjugant bacterial count in the surfaces, which was expected since a temperature of 9 °C leads to a lower frequency of conjugation with an IncP-1-plasmid [33], in spite of here focused on an IncB/O-plasmid, we strongly believe that it can be the same for other plasmids harboring antimicrobial resistant determinants, once the conjugation phenomenon is similar to other plasmid-types [16, 18]. Also, at 25 °C there was no transconjugant bacterial count in poultry meat, and that is possibly related to competition between bacteria. The acquisition of plasmids usually creates a fitness cost for bacteria, leading to a decrease in their competitive ability among bacterial communities [34], beyond what an increase in temperature can favor the multiplication of other microorganisms present in poultry meat [35].

Nonetheless, when performing the swabbing technique, transconjugant bacteria have been recovered from the poultry meat (all treatments), indicating that these bacteria not only survive but are able to conjugate in lower frequencies, under conditions where foods are often stored [33]. The presence of this bacterium in a domestic environment endangering public health since these resistant bacteria could establish contact with human microbiota due to food manipulation. Commensal E. coli has been associated with the transfer of antimicrobial resistance genes to the human intestinal microbiota [36], and intra- and extra-intestinal pathogens [37]. This bacterium is generally found in the microbiota of different hosts and is considered one of the major reservoirs of antimicrobial resistance genes, usually acquired by horizontal gene transfer [38].

The acquisition of ARGs represents an advantage to microbial survival [39] and a threat to public health, especially, when pathogenic bacteria are involved [40]. Furthermore, HGT can contribute to the acquisition of virulence factors [41]. Many pathogenic multidrug-resistant E. coli have been associated with nosocomial infections, associated or not with intestinal infections [42, 43], mostly when immunocompromised patients are involved [44]. Besides, considering the proximity during manipulation these bacteria represent an opportunity to cross-contaminate other food, leading to an enhanced risk of outbreaks [45].

When performed broth microdilution assay, all transconjugants presented a significant increase of the MIC of Amp, Lex, Ctx (p < 0.0001), and Caz (p = 0.0154) when compared to the recipient strain. Transconjugants had the profile of resistance to ampicillin, cephalexin, and cefotaxime (Amp-Lex-Ctx), with susceptibility, observed only for ceftazidime. These drugs are among the critically and highly important antimicrobials to human medicine [46], and antimicrobial resistance to these drugs leads to treatment failure [47]. Ampicillin is considered a front-line antimicrobial in the treatment of pneumonia [48, 49], and cephalexin, a first-generation cephalosporin, is usually an option in the treatment of uncomplicated lower urinary tract infections [50]. Finally, resistance to cefotaxime and ceftazidime, third-generation cephalosporins, have been associated with longer hospital stays and an increased risk of death in patients with bloodstream infections [51].

Thus, the presence of contaminated poultry meat represents a risk to the final consumer, once those pathogenic bacteria can survive at low temperatures and conjugate with non-pathogenic ones, indicating the necessity of efficient food safety measures. Our work shows that a non-pathogenic E. coli can work as a reservoir of antimicrobial resistance genes with an increase of the minimal inhibitory concentration of important antimicrobials used in human medicine. Therefore, it creates an opportunity for these bacteria to exchange these genes with other pathogenic bacteria, contributing to the ineffectiveness of the treatment of human infections, ultimately a burden to public health.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

Conceptualization: Isabella C. Campos, Mauro M. S. Saraiva, and Angelo Berchieri Junior; Methodology: Isabella C. Campos, Mauro M. S. Saraiva, and Valdinete P. Benevides; Formal analysis and investigation: Isabella C. Campos, Mauro M. S. Saraiva, Valdinete P. Benevides, Taísa S. Ferreira, Viviane A. Ferreira, and Adriana M. Almeida; Data analyses: Isabella C. Campos, and Mauro M. S. Saraiva; Manuscript preparation: Isabella C. Campos, Mauro M. S. Saraiva, and Valdinete P. Benevides; Supervision: Mauro M. S. Saraiva, and Angelo Berchieri Junior. All authors have read and approved the final version of the manuscript.

Funding

This study was financed in part by São Paulo Research Foundation (FAPESP) [grant number 2021/04321–2 to I.C.C.] and the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brasil (CAPES) – Finance Code 001 and the National Council of Technological and Scientific Development (CNPq).

Data availability

All data generated or analyzed during this study are included in this article.

Declarations

Consent for publication

All authors consent for publication.

Conflict of interest

The authors declare no competing interests.