Genomic characterization of Salmonella enterica serovar Choleraesuis from Brazil reveals a swine gallbladder isolate harboring colistin resistance gene mcr-1.1

Abstract

Salmonella enterica serovar Choleraesuis (S. Choleraesuis) is a swine-adapted serovar associated to invasive infections in humans. In Brazil, data of strains of this serovar are scarce. In the present study, six S. Choleraesuis strains of animal (n = 5) and human (n = 1) origin from Brazil were screened for phenotypic antimicrobial resistance using disk-diffusion assay and using whole-genome sequencing data to search for antimicrobial resistance genes, plasmids, prophages, and Salmonella pathogenicity islands (SPIs). Its genetic relatedness was evaluated by MLST and SNP analysis. A single isolate from swine gallbladder harbored the colistin resistance gene mcr-1.1 into a IncX4 plasmid. In the six strains analyzed, resistance was found to tetracycline, nalidixic acid, ciprofloxacin, ampicillin, piperacillin, streptomycin, cefazoline, gentamycin, sulfamethoxazole-trimethoprim, and choloramphenicol, along with resistance genes aac(6')-Iaa, aac(3)-IV, aph(3'')-Ib, aph(6)-Id, aph(4)-Ia, aadA1, aph(3')-IIa, bla_TEM-1A, floR, sul1, sul2, tet(B), drfA1, erm(B), mph(B), lnu(G), qacE, and gyrA point mutation Serine83 → Tyrosine and parC Threonine57 → Serine. Furthermore, IncF and IncH plasmids, ten SPIs, and seven prophage types were detected. All strains were assigned to ST145 and five belonged to a common SNP cluster of S. Choleraesuis strains from Brazil. The presence of S. Choleraesuis isolated from animals harboring relevant antimicrobial resistance profiles and virulence determinants reinforced the urge for enhanced surveillance to avoid its transmission to humans through food items.

Supplementary Information

The online version contains supplementary material available at 10.1007/s42770-022-00812-3.

Introduction

Non-typhoid Salmonella are among the four main pathogens associated to human gastroenteritis caused by the consumption of contaminated meat and related products from food-producing animals [1]. It is estimated that, each year, 93.8 million cases of gastroenteritis and 155 thousand deaths occur due to non-typhoid Salmonella infections [2].

Salmonella enterica subspecies enterica serovar Choleraesuis (S. Choleraesuis) is a strongly swine-adapted serovar and responsible for the paratyphoid fever in these animals, whose main clinical manifestations are enterocolitis and sepsis in swine [3, 4]. Although other serovars such as S. Typhimurium have been currently more prevalent in swine and more associated to human infections caused by the consumption of contaminated pork meat, S. Choleraesuis may also rarely infect humans and lead to severe infections, such as acute gastroenteritis, sepsis, and extra-intestinal infections, especially among immunocompromised patients [3–5].

The U.S. Center for Disease Control and Prevention (CDC) have classified drug-resistant non-typhoid Salmonella as a “serious threat” due to the increasing rates of antimicrobial resistance [6]. This fact may not only be a result of the overuse of antibiotics in human and animal therapy in past years, but also due to its possible illegal use as growth promoters in the production of food-producing animals, such as swine, poultry, and bovines [1, 7]. The development of resistance in bacteria is largely favored by mobile genetic elements, such as plasmids, that have been demonstrated to have an important role into the easy acquisition and dissemination of antimicrobial resistance genes [8].

Due to the large dissemination risk of drug-resistant strains to humans through the contamination of food, food-producing animals, and farm and industry settings, an active and constant monitoring of pathogens such as S. Choleraesuis is necessary, especially in Brazil, one of the largest pork meat exporters in the world [9]. In recent years, whole-genome sequencing (WGS) have fostered significant advances through the development and easy access to methods for the detection of antimicrobial resistance genes, plasmids, virulence factors, and the determination of the genetic relationships of bacterial pathogens, for example [8, 10]. However, in regard to the S. Choleraesuis strains circulating in the swine production chain, food products, or humans in Brazil, little information based in genomics is currently available [11].

In this way, the aims of the present study were to characterize genomic aspects related to antimicrobial resistance, virulence, and genetic relatedness of some S. Choleraesuis strains isolated in Brazil by using WGS data. In addition, the phenotypic antimicrobial resistance profiles of the strains were provided.

Materials and methods

In total, six S. Choleraesuis strains were analyzed in the present study (Table 1). Specifically, five strains were isolated from animal (swine and chicken) sources in the state of Minas Gerais in 2017, and one was isolated from human newborn feces in 2015 in the state of Paraná (Table 1). These strains are part of a WGS characterization project that comprised multiple strains of Salmonella serovars of less frequent occurrence in Brazil, isolated from 2013 to 2018 from diverse sources. All isolates were provided by the Brazilian Salmonella reference laboratory of the Oswaldo Cruz Foundation (FIOCRUZ-RJ), which receives Salmonella isolates from diverse isolation sources and regions of the country for classical serotyping and antimicrobial susceptibility testing.

Table 1.

Isolation data and results obtained through the phenotypic and genomic characterization of Salmonella Choleraesuis isolates from Brazil studied

Strain	Isolation data			Phenotypic resistance profile	Acquired resistance genes	Chromosomal point mutations	Plasmids	Salmonella Pathogenicity Islands (SPIs)	Prophage sequences
	Year	State	Material
2357	2015	Paraná	Human newborn feces	TET; NA; CIP(i)	aac(6')-Iaa, aac(3)-IV, aph(3'')-Ib, aph(6)-Id, aph(4)-Ia, tet(B)	gyrA (Ser83 → Tyr), parC (Thr57 → Ser)	IncFIB(S), IncFII(S), IncHI2, IncHI2A	1, 2, 3, 4, 5, 9, 11, 13, 14, CS54	Gifsy_2, Salmon_118970_sal3, Salmon_ST64T, Salmon_SEN34
957	2017	Minas Gerais	Swine gallbladder	AMP; PRL; S; TET; NA; CIP (i); C	aph(3'')-Ib, aac(3)-IV, aac(6')-Iaa, aph(4)-Ia, aph(6)-Id, floR, blaTEM-1A, sul2, tet(B), mcr-1.1	gyrA (Ser83 → Tyr), parC (Thr57 → Ser)	IncX4; IncFIB(S), IncFII(S), IncHI2, IncHI2A	1, 2, 3, 4, 5, 9, 11, 13, 14, CS54	Gifsy_2, Salmon_118970_sal3, Salmon_SEN34, Salmon_epsilon34
959	2017	Minas Gerais	Chicken illeum	AMP; PRL; KZ(i); S; TET; NA; CIP (i); C	aph(6)-Id, aph(4)-Ia, aph(3'')-Ib, aac(6')-Iaa, floR, blaTEM-1A, sul2, tet(B)	gyrA (Ser83 → Tyr), parC (Thr57 → Ser)	IncFIB(S), IncFII(S), IncHI2, IncHI2A	1, 2, 3, 4, 5, 9, 13, 14, CS54	Salmon_118970_sal3, Salmon_epsilon34
2143	2017	Minas Gerais	Chicken spleen and gallbladder	AMP; PRL; S; TET; SXT; NA; CIP (i); C	aph(6)-Id, aadA1, aac(6')-Iaa, aph(3')-IIa, aph(3'')-Ib, floR, blaTEM-1A, sul1, sul2, tet(B), dfrA1, erm(B), mph(B), qacE	gyrA (Ser83 → Tyr), parC (Thr57 → Ser)	IncFIA(HI1), IncHI1A, IncHI1B(R27), IncFIB(S), IncFII(S)	1, 2, 3, 4, 5, 9, 11, 13, 14, CS54	Gifsy_2, Salmon_118970_sal3, Salmon_SEN34, Salmon_epsilon34
2335	2017	Minas Gerais	Swine gallbladder	AMP; PRL; S; CN (i); TET; NA; CIP (i)	aph(3'')-Ib, aac(6')-Iaa, aph(6)-Id, aph(4)-Ia, aac(3)-IV, blaTEM-1A, sul2, tet(B)	gyrA (Ser83 → Tyr), parC (Thr57 → Ser)	IncFIB(S), IncFII(S), IncHI2, IncHI2A	1, 2, 3, 4, 5, 9, 13, 14	Gifsy_1, Gifsy_2, Salmon_SEN34, Entero_lambda, Salmon_ST64T
2529	2017	Minas Gerais	Swine gallbladder	S; TET	aph(3'')-Ib, aac(6')-Iaa, aph(6)-Id, lnu(G), sul2, tet(B)	parC (Thr57 → Ser)	IncFIB(S), IncFII(S)	1, 2, 3, 4, 5, 9, 11, 13, 14, CS54	Gifsy_2, Salmon_118970_sal3, Salmon_SEN34, Salmon_epsilon34

AMP, ampicillin; C, chloramphenicol; CIP, ciprofloxacin; CN, gentamycin; KZ, cefazoline; NA, nalidixic acid; PRL, piperacillin; S, streptomycin; SXT, trimethoprim-sulfamethoxazole; TE, tetracycline

Genomic DNA of the strains was extracted by the phenol–chloroform-isoamyl alcohol method [12]. WGS was performed on the Illumina MiSeq sequencer using the 2 X 150-bp paired-end MiSeq Reagent Kit version 3 (Illumina, San Diego, CA) according to the manufacturer’s recommendations and libraries were prepared using 1 ng of genomic DNA with the Nextera XT DNA library preparation kit (Illumina, San Diego, CA). Quality control was performed in MicroRunQC workflow on the Galaxy platform, de novo assemblies were generated using SKESA 2.2 assembler, and genome drafts were annotated with NCBI’s Prokaryotic Genome Annotation Pipeline (PGAP). Sequencing information and accession numbers of the strains analyzed are displayed in Supplementary Table 1.

The search for acquired antimicrobial resistance genes and point mutations conferring antimicrobial resistance was performed by using ResFinder 4.1 (https://cge.cbs.dtu.dk/services/ResFinder/). Salmonella Pathogenicity Islands (SPIs) and phages were respectively searched with SPIFinder 2.0 (https://cge.cbs.dtu.dk/services/SPIFinder/) and PHASTER (https://phaster.ca/). Default parameters were applied for all the abovementioned tools. The sequence types (STs) of the S. Choleraesuis studied were determined in silico using multi-locus sequence typing (MLST) (https://cge.cbs.dtu.dk/services/MLST/). In order to verify the possible genetic relationship of the six S. Choleraesuis strains studied to additional isolates deposited in NCBI’s Pathogen Detection platform (https://www.ncbi.nlm.nih.gov/pathogens/isolates), the single-nucleotide polymorphism (SNP) clusters of the strains were searched using the accession numbers provided in Supplementary Table 1.

PlasmidFinder 2.1 (https://cge.cbs.dtu.dk/services/PlasmidFinder/) was employed to search for plasmids and its respective incompatibility (Inc) groups in the six S. Choleraesuis strains analyzed. In the strain harboring the plasmid associated to colistin resistance genes, additional genomic and phenotypic analyses were performed. To evaluate the genetic structure and content of this plasmid, the BLAST Atlas tool, included in the GView server (https://server.gview.ca/), was employed to align the plasmid found in the genomes studied to reference plasmid sequences. In addition, the transferability of this plasmid was determined by conjugation [13] with azide-resistant recipient Escherichia coli strain J53. Transconjugants were selected in MacConkey agar containing 2 μg/mL of polymixin B and 100 μg/mL of sodium azide.

The six S. Choleraesuis strains had the phenotypic antimicrobial resistance determined by disk-diffusion assay for amoxicillin-clavulanic acid (30 µg), piperacillin (10 µg), ampicillin (10 µg), cefazolin (30 µg), cefoxitin (30 µg), ceftriaxone (30 µg), cefotaxime (30 µg), ceftazidime (30 µg), cefepime (30 µg), imipenem (10 µg), amikacin (30 µg), gentamycin (30 µg), streptomycin (10 µg), trimethoprim-sulfamethoxazole (25 µg), tetracycline (30 µg), nalidixic acid (30 µg), ciprofloxacin (5 µg), and chloramphenicol (30 µg). Broth microdilution has also been performed in the strain harboring colistin resistance gene to determine its minimum inhibitory concentration (MIC). Clinical and Laboratory Standards Institute (CLSI) guidelines were followed in the performing and interpretation of results in the assays [14].

Results and discussion

In the present study, it was concerning to notice the presence of the colistin resistance gene mcr-1.1 in strain 957 isolated from swine gallbladder in 2017 in the state of Minas Gerais, in Brazil (Table 1). Colistin is a broad-spectrum polymyxin antibiotic, considered as a “last resource” agent to treat severe drug-resistant bacterial infections in humans, and the gene mcr-1.1 was first reported in 2016 in an Escherichia coli isolate originated from food-producing animals in China [15]. This gene has been largely associated to IncX4, an epidemic plasmid with easy transferability to human pathogenic bacteria such as Klebisella pneumoniae and Pseudomonas aeruginosa [15], which was also detected in the strain 957 of the present study harboring mcr-1.1 gene (Table 1).

In Brazil, a recent study has also detected a S. Choleraesuis isolate (446_18B; Pathogen Detection accession number PDT000706279.1) from a bloodstream infection case in the state of São Paulo showing polymixin B and colistin MICs of 8 mg/l and carrying mcr-1 gene into a transferable ~ 40 kb IncX4 plasmid [11]. Other studies conducted in the country also identified Salmonella strains of serovars Typhimurium and Schwarzengrund carrying the same gene and plasmid in the country [16, 17].

In the present study, the transferability of this plasmid was not achieved in the conjugation assay, and a susceptible MIC < 0.5 µg/ml to colistin was found. These results may possibly indicate that the strain 957, during the storage period from the DNA extraction previous to WGS (which occurred in 2020) until the performing of the conjugation and broth microdilution assays, may have suffered the loss of the IncX4 plasmid and, consequently, the loss of the mcr-1.1 gene, which could explain the susceptible MIC found and the absence of plasmid transferability.

However, it was possible to observe through the genomic analyses that mcr-1.1 gene was in fact present in an IncX4 plasmid in the S. Choleraesuis strain here studied (Fig. 1) of similar size and genetic content than previously reported [18]. The alignment performed with BLAST Atlas showed that strain 957 from the present study and strain 446_18B reported by Santos and collaborators [11] presented the same genetic content of pICBEC72Hmcr (GenBank accession number CP015977), the first IncX4 plasmid harboring mcr-1.1 gene detected in Brazil in an Escherichia coli isolate [18] (Fig. 1). Among the genetic features that were demonstrated to be shared by pICBEC72Hmcr and strains 957 and 446_18B, it is important to mention the presence of mcr-1.1 gene (confirming the plasmid location of this gene in the IncX4 plasmid, and not in the chromosome of the strains studied), insertion sequence 26 (IS26, which has been previously demonstrated by Poirel and collaborators to be part of the genetic environment of the mcr-1.1 cassette) [19], virB genes (associated to the codification of type IV secretion systems), and conjugation and partitioning genes (traG and parA, respectively, which may play a role in the transmissibility of the IncX4 plasmid) (Fig. 1).

Fig. 1.

BLAST Atlas diagram showing the genomic similarity between pICBEC72Hmcr (red ring; GenBank accession number CP015977), the first IncX4 plasmid harboring colistin resistance gene mcr-1.1- described in an Escherichia coli isolate identified in Brazil, in comparison to the Salmonella Choleraesuis IncX4 mcr-1.1-harboring plasmids contained in the genomes of the Brazilian strains 957 (green ring), isolated from swine gallbladder in 2017 in the state of Minas Gerais and analyzed in the present study, and the S. Choleraesuis strain 446_18B (blue ring; Pathogen Detection accession number PDT000706279.1), isolated from human bloodstream in 2018

Together, these findings should represent an alert for the necessity of continuous surveillance by Brazilian authorities to prevent the transmission of S. Choleraesuis strains from animal and food origin harboring plasmid-borne colistin resistance genes to humans, as well as the dissemination of epidemic plasmids such as IncX4 to other relevant human pathogenic bacteria.

Using MLST, all six strains studied were assigned to ST145, which has also been reported for other S. Choleraesuis from Brazil [11]. The analysis of the SNP clusters showed, by the date of this study, that the mcr-1.1-positive strain 957 and other four strains studied (959, 2143, 2529, 2357) belonged to a common genetically related SNP cluster (access number PDS000059946.6; Fig. 2) that also contained seven additional S. Choleraesuis strains isolated from clinical and environmental/other sources in the Brazilian states of São Paulo, Minas Gerais, Paraná, Rio Grande do Sul, and Goiás between 2011 and 2018. It is interesting to notice that, to the date of this study, this was the only S. Choleraesuis SNP cluster of Pathogen Detection containing isolates from Brazil. Together, the 12 isolates in the PDS000059946.6 SNP cluster showed a minimum and maximum distance of eight and 48 SNPs, respectively, with an average distance of 27 SNPs. It is also important to highlight that the strain 2355, by the date of this study, was not assigned to any SNP cluster in Pathogen Detection, suggesting that it may belong to a novel and distinct SNP cluster.

Fig. 2.

Phylogenetic SNP-based dendrogram showing SNP cluster PDS000059946.6 of NCBI’s Pathogen Detection platform that contained five Salmonella Choleraesuis (S. Choleraesuis) strains analyzed in the present study (blue squares) and additional seven sequenced S. Choleraesuis isolates (red squares) from clinical and environmental sources at different Brazilian states that were publicly available in the referred database. The dendrogram was formatted using FigTree v1.4.2 software. The strain 2335 (see Table 1), which was also analyzed in the present study, was not associated to this SNP cluster and, therefore, was not included in this dendrogram. Strains 957 and 446_18B (PDT000706279.1) were the only S. Choleraesuis genomes harboring colistin resistance gene mcr-1.1

The results here obtained may suggest that five of six strains here studied possessed a high genomic correlation among each other and possibly may belong to a distinct SNP cluster composed exclusively of Brazilian S. Choleraesuis isolates that have been present in several types of human and animal sources at several states of Brazil.

Curiously, only two of the 12 strains in the SNP cluster mentioned above harbored mcr-1.1, which were strains 957 and 446_18B [11] (Fig. 2). This fact reinforces that the acquisition of mcr-1.1 is strongly associated with its presence inside the IncX4 plasmid, since that, despite the genomic relationship of the 12 isolates comprised in the SNP cluster, only strains 957 and 446_18B carried the combination of this colistin resistance gene and plasmid.

In addition to mcr-1.1 gene, other several profiles of phenotypic and genotypic resistance have also been identified among all the six S. Choleraesuis isolates studied, which have already been commonly reported for many Salmonella serovars in Brazil, including S. Choleraesuis [11, 20]. The six S. Choleraesuis strains studied harbored different combinations of antimicrobial resistance genes, conferring resistance to aminoglycosides (aac(6')-Iaa, aac(3)-IV, aph(3'')-Ib, aph(6)-Id, aph(4)-Ia, aadA1, aph(3')-IIa), beta-lactams (bla_TEM-1A), phenicols (floR), sulfonamides(sul1, sul2), tetracycline (tet(B)), trimethoprim (dfrA1), streptogramin, macrolide and/or lincosamide (erm(B), mph(B), Inu(G)), and quaternary ammonium compounds (qacE) (Table 1). Point mutations Serine83 → Tyrosine in gyrA and Threonine57 → Serine in parC, responsible for quinolone and fluoroquinolone resistance, were also detected (Table 1). Moreover, using disk-diffusion, it was also possible to verify that the six isolates displayed phenotypic resistance or intermediate profiles to tetracycline (n = 6); nalidixic acid and ciprofloxacin (n = 5), ampicillin, piperacillin, and streptomycin (n = 4); chloramphenicol (n = 3); cefazoline, gentamycin, and sulfamethoxazole-trimethoprim (n = 1) (Table 1).

It should be noticed that not only resistance to antibiotics used in human therapy, such as quinolones and beta-lactams, were detected (Table 1). Resistance to drugs commonly used in the veterinary field for animal therapy or possibly as illegal growth promoters (such as tetracycline, aminoglycosides, and phenicols), as well as genes conferring resistance to disinfectants such as quaternary ammonium compounds, were also observed (Table 1). Also, considering that some strains showed phenotypic and/or genotypic resistance profiles to three or more different antimicrobial classes, this suggests that S. Choleraesuis isolates could also present multidrug resistance profiles (Table 1). These results should reinforce the potential hazard of a zoonotic pathogen such as S. Choleraesuis showing resistance to multiple antimicrobial agents used both in human and veterinary fields in Brazil.

Besides IncX4, plasmids of different incompatibility groups have also been identified among the S. Choleraesuis isolates studied (Table 1). IncFIB(S) and IncFII(S) plasmids, detected in all six strains analyzed, showed a high identity percentage in the PlasmidFinder output to the previously described pSLT-BT of S. Typhimurium and pSPCV of S. Paratyphi C, respectively, which have been associated with the presence of spv, pef, and tra operons, which were described to influence in the systemic phase of Salmonella infection into its host [21, 22]. IncHI2 and IncHI2A plasmids were detected in four strains and IncFIA(HI1), IncHI1A, and IncHI1(B27) plasmids were detected in individual strains. These plasmids showed high identity percentages according to the output of PlasmidFinder to S. Typhi plasmid R27 and Serratia marcescens plasmid R478, which have been broadly detected in Enterobacteriaceae and were described to possess a variable amount of antimicrobial resistance genes and metal tolerance determinants [23, 24].

Regarding the presence of pathogenicity islands, all strains harbored SPI-1, -2, -3, -4, -5. -9, -13, and -14. CS54 island was detected in five strains and SPI-11 in four strains (Table 1). SPIs are common in Salmonella isolates and are mainly composed of genes associated to virulence functions [25]. SPI-1 and SPI-2 are broadly studied and act in the invasion to intestinal epithelial cells and survival and replication within phagocytic cells, respectively, through the formation of type 3 secretion systems. SPI-5 is related to the fluid secretion and inflammatory response. Finally, SPI-3, SPI-4, SPI-11, SPI-13, SPI-14, and CS54 have also been associated to Salmonella survival to stresses and adaptation within macrophages [25].

Different phages have also been detected in the strains analyzed. Gifsy_2, Salmon_118970_sal3, and Salmon_SEN34 were detected in five strains, Salmon_epsilon34 in four strains, Salmon_ST64T in two strains, and Gifsy_1 and Entero_lambda in individual strains (Table 1). Although the influence of the presence of many phages in Salmonella serovars still remains unclear, some as Gifsy-2, Gifsy-1, ST64T, and Epsilon 34 have been detected in many serovars and reported to influence on the survival and adaptation within phagocytic cells [26]. Also, its diverse presence is commonly referred as an indicative of diversification among serovars [26]. It is important to reinforce the necessity of further studies to evaluate the phenotypic role of these phages in Salmonella serovars, especially among S. Choleraesuis in which information is scarce.

In conclusion, the presence of S. Choleraesuis isolated from animal sources in Brazil harboring diverse and relevant antimicrobial resistance determinants reinforced the urge for enhanced surveillance by Brazilian authorities to avoid its dissemination through food items and impair a possible transmission of invasive drug-resistant S. Choleraesuis to humans.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

FPV: conceptualization, formal analysis, investigation, data curation, writing—original draft, preparation, and visualization. DPR: conceptualization, investigation, resources, and data curation. JCF: formal analysis, investigation. ALCD: formal analysis, investigation. MWA: resources, data curation, writing—review and editing, supervision, project administration, and funding. JPF: resources, writing—review and editing, supervision, project administration, and funding.

Funding

This study was supported by research grants from the FDA/Center for Food Safety and Applied Nutrition (CFSAN) under the supervision of M. W. Allard and from the São Paulo Research Foundation (FAPESP; Proc. 2019/19338–8) under the supervision of J.P. Falcão. During the course of this work, F.P. Vilela was supported by Master and PhD student scholarships from FAPESP (Proc. 2019/06947–6) and the National Council for Scientific and Technological Development (CNPq; Proc. 141017/2021–0), respectively. J.P. Falcão also received a productive fellowship from CNPq (Proc. 304399/2018–3 and 304803/2021–9). This study was financed by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior—Brasil (CAPES) – Finance Code 001.

Data availability

The authors confirm that all the identification of the data analyzed is publicly available for consult and has been provided in the body of the manuscript or supplementary data file.

Code availability

Not applicable.

Declarations

Ethics approval

No ethical approval was necessary for the present study.

Consent to participate

Not applicable.

Consent for publication

All authors agreed with the final version of the manuscript.

Conflict of interest

The authors declare no competing interests.