Genomic Analysis of Two MDR Isolates of Salmonella enterica Serovar Infantis from a Spanish Hospital Bearing the blaCTX-M-65 Gene with or without fosA3 in pESI-like Plasmids

Salmonella enterica serovar Infantis (S. Infantis) is a broiler-associated pathogen which ranks in the fourth position as a cause of human salmonellosis in the European Union. Here, we report a comparative genomic analysis of two clinical S. Infantis isolates recovered in Spain from children who just returned from Peru. The isolates were selected on the basis of resistance to cefotaxime, one of the antibiotics of choice for treatment of S. enterica infections. Antimicrobial susceptibility testing demonstrated that they were resistant to eight classes of antimicrobial agents: penicillins, cephalosporins, phenicols, aminoglycosides, tetracyclines, inhibitors of folate synthesis, (fluoro)quinolones and nitrofurans, and one of them was also resistant to fosfomycin. As shown by whole-genome sequence analysis, each isolate carried a pESI-like megaplasmid of ca. 300 kb harboring multiple resistance genes [blaCTX-M-65, aph(4)-Ia, aac(3)-IVa, aph(3′)-Ia, floR, dfrA14, sul1, tet(A), aadA1 ± fosA3], as well as genes for resistance to heavy metals and disinfectants (mer, ars and qacEΔ1). These genes were distributed in two complex regions, separated by DNA belonging to the plasmid backbone, and associated with a wealth of transposable elements. The two isolates had a D87Y amino acid substitution in the GyrA protein, and truncated variants of the nitroreductase genes nfsA and nsfB, accounting for chromosomally encoded resistances to nalidixic acid and nitrofurantoin, respectively. The two S. Infantis isolates were assigned to sequence type ST32 by in silico multilocus sequence typing (MLST). Phylogenetic analysis revealed that they were closely related, differing only by 12 SNPs, although they were recovered from different children two years apart. They were also genetically similar to blaCTX-M-65-positive ± fosA3 isolates obtained from humans and along the poultry production chain in the USA, South America, as well as from humans in several European countries, usually associated with a travel history to America. However, this is the first time that the S. Infantis blaCTX-M-65 ± fosA3 MDR clone has been reported in Spain.


Introduction
Salmonella enterica serovar Infantis (S. Infantis) is one of the main non-typhoid serovars of S. enterica. In 2020, it ranked in the fourth position as a cause of human salmonellosis in the European Union, only preceded by S. Enteritidis, S. Typhimurium and the monophasic 1,4, [5],12:i:-variant of the latter serovar. In addition, it was the most common serovar in broiler flocks and broiler meat [1]. Apart from the European Union, S. Infantis has increasingly been found in clinical, animal and food samples from many other geographical regions [2][3][4][5][6]. The epidemiological success of S. Infantis is strongly associated with the spread of isolates harboring a large conjugative plasmid which confers resistance, virulence and fitness properties. This plasmid, termed pESI (plasmid of Emerging S. Infantis), was first reported in a multidrug-resistant (MDR) clone of S. Infantis detected since 2006 in Israel [7,8]. pESI was identified as a chimeric megaplasmid of ca. 280 kb, found to contain the origin of replication (oriV) and incompatibility region of IncP-1α plasmids, and the IncI leading and transfer region. Later on, pESI-like plasmids were also detected in S. Infantis from many other countries, including Italy, Switzerland, Denmark, England and Wales, the USA and South America, and they were shown to contain the IncFIB replicon [2,[9][10][11][12][13][14][15]. All pSE1-like plasmids confer resistance against multiple antimicrobial agents, including clinically relevant antibiotics such as third-generation cephalosporins. So far, two distinct types of pESI-like plasmids bearing genes encoding extended spectrum β-lactamases (ESBL) have been reported: bla CTX-M-1 -positive and bla CTX-M-65 -positive plasmids, found in European and American isolates, respectively [16].
The molecular epidemiology of S. Infantis and its pESI-like plasmids in Europe has been investigated by whole-genome sequencing and bioinformatics analysis, using isolates from America as an outgroup [16]. This study revealed the heterogeneity of the European S. Infantis population, which is composed by multiple clusters defined at the core-genome level, confirming previous proposals of the polyphyletic nature of this serovar [11,17]. However, pESI-like variants, which were present in 64.1% of the isolates analyzed (a total of 362; [16]), were genetically more homogeneous and were distributed among different clonal lineages in most countries. It has been proposed that once acquired by an S. Infantis isolate, the conjugative megaplasmid would rapidly become established in the local population because of the selective advantages it confers to the host bacteria [16]. Such advantages are not only provided by the presence of antibiotics resistance genes, but also by genes coding for resistances to quaternary ammonium compounds and heavy metals, which are likewise widespread in other successful clones of non-typhoid serovars of S. enterica [13,[18][19][20][21]. Moreover, pESI-like megaplasmids also provide genes involved in colonization, virulence and fitness, including those encoding two types of fimbriae (Fim and K88-like) and the siderophore yersiniabactin, as well as toxin-antitoxin systems (like CcdBA, PemKI) [7,16].
In a screen for S. enterica resistant to third-generation cephalosporins in Spanish hospitals, two isolates of S. Infantis resistant to cefotaxime were identified. The present study aimed to combine epidemiological information with experimental research and whole-genome sequencing (WGS) analysis for in-depth characterization of these isolates. This provided evidence for the migration of the S. Infantis clone carrying pESI-like plasmids of the "American type" from Peru to Spain.

Origin of the Isolates, Genome Sequencing and Resistance Phenotypes
The two isolates, HUMV 13-6278 and HUMV 15-5476, were recovered in 2013 and 2015 at the "Hospital Universitario Marqués de Valdecilla" (HUMV) from fecal samples of small children with gastroenteritis ( Table 1). The children did not require hospitalization and were attended to at Primary Care Centers associated with the hospital. In both cases, the onset of the disease coincided with their return to Spain after a trip to Peru.
As shown in Tables 1 and 2, the latter compiling MIC values for relevant compounds, the two isolates were resistant to 14 out of the 21 antimicrobials tested, including cefotaxime (MIC of 32 µg/mL; used as the bases for selection of the isolates), nalidixic acid (MIC of 128 and >256 µg/mL for HUMV 13-6278 and HUMV 15-5476, respectively), pefloxacin (selected by EUCAST as a surrogate of ciprofloxacin to disclose clinical resistance to fluoroquinolones; https://eucast.org/clinical_breakpoints; accessed on 8 April 2022), ciprofloxacin (MIC of 0.125 µg/mL, resistant according to EUCAST criteria), and nitrofurantoin (MIC of 512 µg/mL). HUMV 15-5476 but not HUMV 13-6278 was also resistant to fosfomycin, displaying MIC values of 512 and 0.19 µg/mL, respectively. With regard to heavy metals, the two S. Infantis isolates were resistant to mercury as well as to inorganic and organic arsenic compounds (Table 2).  The draft genomes of HUMV 13-6278 and HUMV 15-5476 contained a total of 88 and 86 contigs, respectively, with an assembly size of about 4700 kb (Table 1). They shared the antigenic formula 7:r:1,5 and the sequence type ST32, both associated with S. Infantis. Each isolate harbored a single pESI-like plasmid of more than 310 kb (Table 1), termed pHUMV 13-6278 and pHUMV 15-5476, respectively. The two plasmids carried the IncFIB replicon, the IncP origin of replication (oriV) and a gene for a MOBP-family relaxase. Additionally, plasmid typing identified the ardA_2, pilL_3, sogS_9, trbA_21 alleles of the pMLST scheme designed for the IncI1 incompatibility group, but lacked the repI1 gene. Based on the detected alleles, the closest related sequence type within the IncI pMLST scheme was ST71.

Genetic Bases of Antimicrobial Drug Resistance
The resistance genes detected in the genomes of the two isolates by bioinformatics analysis were in agreement with the observed phenotypes (Table 1). Both isolates contained bla CTX-M-65 (ampicillin and cefotaxime), floR (chloramphenicol), aac(3)-IVa, aph(3 )-Ia, aph(4)-Ia, aadA1 (aminoglycosides), tet(A) (tetracycline), sul1 and dfrA14 (sulfonamides and trimethoprim, respectively), together with genes conferring resistance to heavy metals: locus merRTPCADE (mercury), arsA and arsD (inorganic arsenic compounds), arsH (organic arsenicals), arsR2 (metal-responsive transcriptional regulator), and qacE∆1 (although resistance to disinfectants was not experimentally tested). In the case of HUMV 15-5476, fosfomycin resistance was mediated by fosA3. All these genes were located on the pESI-like plasmids, as part of two complex resistance regions: R1 and R2, comprising approximately 36 or 32 kb (depending on the presence or absence of fosA3) and 21 kb, respectively. R1 and R2 were separated by DNA of the pESI-like backbone, and contained a wealth of transposable elements belonging to multiple families, including several copies of IS26 ( Figure 1). The resistance regions of the two plasmids differed only by the presence of the IS26-fosA3-IS26 pseudo-compound transposon in R1 of pHUMV 15-5476, which was lacking in the corresponding region of pHUMV 13-6278, and by a copy of IS26 located at the 5 -end of R2 in the former but not in the latter. Consistent with the high conservation of the pESI-like plasmids [16], regions identical to those of the plasmids under study were found in other plasmids of this group. However, there was also considerable variation, as shown by their comparison with some key examples (Figure 1), which include the original pESI plasmid, detected in a human isolate from Israel (119944), which lacks most of the R1 region [7,8]. A higher degree of similarity was found with two bla CTX-M-65 -positive plasmids, with or without fosA3 (FSIS1502169 and CVM N18S2198, respectively; [14,24]), identified in isolates from chicken samples in the USA.
Apart from the plasmid-encoded genes, the two isolates displayed a point mutation in the chromosomal gyrA gene (G259T), resulting in D87Y exchange in the protein, which could account for nalidixic acid resistance. In addition, both strains also carried a single-nucleotide polymorphism in their chromosomal parC gene (C170G), leading to T57S substitution in the protein. Finally, nitrofurantoin resistance of the two isolates was associated with premature stop codons in the nsfA and nsfB genes, producing truncated forms of the NsfA and NsfB nitroreductases, with only 158 and 136 amino acids, as opposed to 240 and 217 amino acids in the wild type, respectively. Figure 2 shows the phylogenetic position of the two HUMV isolates in the context of other S. Infantis isolates. Since S. Infantis is a broiler-associated pathogen, we chose a selective comparison of sequences of isolates from humans, chickens or the chicken farm environment in different countries of Europe and America, and in Israel (Supplementary Materials Table S2). All isolates harbored a pESI-like plasmid, which carried the bla CTX-M-1 , the bla CTX-M-65 gene, or neither. The Spanish isolates, which differed by only 12 SNPs, belonged to a clonal lineage that groups all bla CTX-M-65 -positive isolates together with isolates lacking a bla CTX-M gene. The minimum and maximum number of SNPs between isolates of this cluster ranged from 3 to 60, and the distance with respect to the 119944 strain from Israel, which contains the original pESI plasmid, ranged from 93 to 127 SNPs.

Discussion
The present study examined two clinical isolates of S. Infantis harboring the bla CTX-M-65 gene on pESI-like megaplasmids. As other bla CTX-M genes, it confers resistance to most β-lactam antibiotics, including penicillins, broad spectrum cephalosporins and monobactams, but not to amoxicillin-clavulanic acid, cephamycins and carbapenems [25]. As such, it conveys resistance to third-generation cephalosporins, which currently are first-line drugs for the management of severe S. enterica infections. Thus, the presence of bla CTX-M-65 is a cause of concern, particularly when associated with resistance to other classes of medically relevant antimicrobials. In fact, the two isolates were resistant to antimicrobials belonging to eight classes: penicillins, cephalosporins, phenicols, aminoglycosides, tetracyclines, inhibitors of folate synthesis, (fluoro)quinolones, and nitrofurans, as confirmed by disk diffusion assays complemented with MIC determinations for selected compounds. HUMV 15-5476, but not HUMV 13-6278, was additionally resistant to fosfomycin.
Most of the resistance genes found were located on pESI-like megaplasmids, but resistances to (fluoro)quinolones and nitrofurantoin were mediated by chromosomal loci. Thus, the two isolates had single-point mutations both in gyrA (G259T→D87Y) and parC (C170G→T57S). The former mutation, which is now widespread among S. Infantis, was originally identified as responsible for nalidixic acid resistance in isolates from Israel. This trait was already present in a limited number of pre-emergent isolates, which were obtained before 2006 and lacked pESI, while becoming fixed in the gyrA gene of the emergent pESIpositive clone (recorded after 2006), adding to its MDR phenotype [3,7]. In addition to D87Y, three other point mutations in gyrA (D87G, S83Y and S83F) were later associated with (fluoro)quinolone resistance in an extensive collection of S. Infantis recovered in Europe from animals, meat, feed and humans [16]. In that work, a single-(fluoro)quinoloneresistant isolate tested negative for known mutations in gyrA, but presented the T57S substitution in the ParC protein. The latter change has been previously found both in resistant and in susceptible strains of different serovars of S. enterica, with or without concomitant mutations in gyrA [16,[26][27][28][29]. The contribution of the T57S substitution to (fluoro)quinolone resistance in the isolates thus remains unclear. PMQR (Plasmid-Mediated Quinolone Resistance) mechanisms, which confer low level, albeit clinically relevant, resistance to fluoroquinolones, were not present in our isolates but have already been reported in S. Infantis from different countries [16].
In S. enterica, as well as in Escherichia coli, resistance to nitrofurans has been associated with alterations affecting the nfsA and nfsB genes that encode type I (i.e., oxygen insensitive) nitroreductases. Amongst other substrates, these enzymes appear to reduce nitrofurans, generating active intermediates with antimicrobial activity. Nitrofurans are used for treating uncomplicated urinary tract infections caused by a wide spectrum of Gram-positive and Gram-negative bacteria [30,31]. In both E. coli and S. Typhimurium, a stepwise increase in resistance to nitrofurantoin has been correlated with sequential inactivation first of nfsA and then of nfsB, with alterations in the two genes being required to achieve full resistance [22,32,33]. The S. Infantis isolates in the present study were highly resistant to nitrofurantoin ( Table 2). In agreement with this, both carried premature stop codons in the two nfs genes, leading to truncated forms of the proteins. The same mutation in nfsA was previously detected in S. Infantis from Israel, without reference to alterations in the nfsB gene [7]. Yet, sequence comparisons of the nsfB gene from our isolates and strain 119944, obtained in 2008 from a clinical sample in Israel [3], revealed the presence of the mutation (data not shown). Accordingly, the isolate from Israel was likely to display high-level resistance to nitrofurantoin, although the MIC of nitrofurantoin was not reported.
As previously shown for other S. Infantis isolates, the pESI-like megaplasmids played a major role in the MDR phenotype of the isolates under study. They not only harbor numerous antibiotic resistance genes, but also genes for resistance to mercury (merRTP-CADE locus) and arsenic (arsA, arsD, arsH and arsR2) compounds. In the case of arsenic, toxicity depends on the chemical form (inorganic or organic) and oxidation state (trivalent or pentavalent species) of the compound, with As(III) arsenite being more toxic than As(V) arsenate. The product of arsA is an ATPase that interacts with ArsB to form an arsenite efflux pump energized by ATP hydrolysis. It is of note that arsA, but not arsB, is present in the pESI-like plasmids of the isolates under study, which were nevertheless resistant to NaAsO 2 [As(III)]. However, besides canonical ars genes, additional genes have been discovered that broaden the spectrum of arsenic tolerance [34]. Apart from interacting with ArsB, the ArsA ATPase has been proposed to form primary arsenite transporters in association with different membrane proteins [35]. The most likely candidate would be Acr3, but the acr3 gene was not found in the genome of our isolates. The product of arsD, a gene also present in pESI-like plasmids, acts as a metallochaperone that binds arsenite and transfers it to the ArsA ATPase for export [36]. Finally, the ArsH protein, which is under the control of the ArsR2 transcriptional regulator, encodes an organoarsenical oxidase that transforms trivalent methylated and aromatic arsenicals into their less toxic pentavalent species [37]. Accordingly, the two isolates described herein were resistant to the aromatic As(III) phenylarsine oxide. Of note, although mer and ars genes are frequently found in pESI-like plasmids, their association with phenotypic resistance of S. Infantis has not been previously tested [38]. The selective pressure caused by the use of heavy metals, and also of nalidixic acid, tetracyclines, sulfonamides and nitrofurans in livestock production, could have contributed to the selection of pESI-positive isolates of S. Infantis, as it was probably the case for other successful clones of different non-typhoidal serovars of S. enterica [13,[18][19][20][21].
Of the multiple resistance genes harbored by the pESI-like plasmids under study, only fosA3 was differentially present in HUMV 15-5476, which was highly resistant to fosfomycin ( Table 2). This broad-spectrum bacteriolytic antibiotic was originally approved for treating uncomplicated urinary tract infections in the early 1970s. However, the major threat posed by antimicrobial resistance to human health, and the shortage of new antimicrobial agents, has led to re-evaluating older antibiotics, including fosfomycin, for the treatment of infections caused by multidrug-resistant bacteria [39]. Therefore, the increasing rate of fosfomycin resistance, which is mainly mediated by inactivating enzymes, is a cause of concern. Like other fosA genes, the fosA3 gene of the pESI-like plasmid of HUMV 15-5476 encodes a glutathione-S-transferase, which catalyzes the opening of the epoxide ring of fosfomycin by adding glutathione [40]. In S. Infantis, fosA3 has been found in the subgroup of pESI-like plasmids positive for bla CTX-M-65 , but not in those carrying bla CTX-M-1 [16].
It is interesting to note that, as previously reported for other pESI-like plasmids, those of the HUMV isolates not only contained antimicrobial-resistant genes, but also genes associated with virulence, colonization and fitness, including genes for fimbriae, for the siderophore yersiniabactin, and for toxin-antitoxin systems which are involved in the postsegregational killing of bacteria that lose the plasmid. Taken together, these determinants confer selective advantages which explain the epidemiological success of the isolates that have managed to acquire this self-transmissible plasmid [7,41].
It is finally of note that isolates harboring pESI-like plasmids with bla CTX-M-65 ± fosA3, like those reported herein, form a well-defined cluster within the phylogenetic tree of S. Infantis ( [13,16]; Figure 2). They have been reported in humans and along the poultry production chain in the USA and South America, with only a limited number of isolates being detected in other livestock animals [14,24,[42][43][44][45][46]. Isolates carrying pESI-like plasmids positive for bla CTX-M-65 ± fosA3 have also been associated with human clinical cases in Europe, specifically in patients with a travel history to America [10,12,16]. However, to the best of our knowledge, this is the first report of such isolates in Spain, with a clearly established link to Peru. In Spain, information is also lacking on the possible presence of S. Infantis carrying pESI-like plasmids of the "European type", positive for bla CTX-M-1 instead of bla CTX-M-65 , which are circulating in food-animal production systems of many other European countries [16]. Within the One Health Concept, promoted by the World Health Organization, further studies are required to investigate the occurrence of such isolates in Spanish livestock, and particularly along the poultry food chain. WGS analyses could certainly assist in the detection and detailed characterization of isolates belonging to the different MDR CTX-M-producing lineages of S. Infantis, and to establish their phylogenetic relationships with those from other countries.

Whole-Genome Sequencing and Bioinformatics Analysis
For WGS, the genomic DNA of the isolates was purified from overnight cultures grown in Luria-Bertani (LB) broth, using the GenEluteTM Bacterial Genomic DNA Kit, according to the manufacturer's protocol (Sigma-Aldrich). WGS was performed with Illumina technology, at the "Centro de Investigación Biomédica", La Rioja (CIBIR), Spain. Paired-end reads of 100 nt, obtained from PCR-free fragment libraries of ca. 500 bp, were assembled de novo using the VelvetOptimiser.pl script implemented in the on line version of PLACNETw (https://omictools.com/placnet-tool; accessed on 6 June 2019). This tool also allows distinguishing contigs of chromosomal or plasmid origin [48]. The assembled genomes were deposited in GenBank under the accession numbers included in the "Data Availability Statement" (see below), and annotated by the NCBI Prokaryotic Genome Annotation Pipeline (PGAP; https://www.ncbi.nlm.nih.gov/genome/annotation_prok/; accessed on 4 August 2021). Relevant parameters regarding the quality of the assemblies are shown in Table S1. Bioinformatics analysis was performed with PLACNETw, MOBscan (a web application for the identification of relaxase MOB families [49]), and with different tools available at the Center for Genomic Epidemiology (CGE) of the Technical University of Denmark (DTU) (https://cge.cbs.dtu.dk/services/; accessed on 7 June 2020). These included SeqSero 1.2 [50], MLST 2.0 [51], ResFinder 4.1 [52,53], PlasmidFinder 2.1 and pMLST 2.0 [54].
Once identified by PLACNETw, plasmid contigs carrying resistance genes were joined by PCR amplification (using the primers shown in Table S4), followed by Sanger sequencing of the obtained amplicons, when required. The annotation of the assembled resistance regions were manually curated with the aid of blastn, blastp (https://blast.ncbi.nlm.nih.gov/Blast. cgi; accessed on 28 April 2022) and CLONE Manager Professional (CmSuit9). The genetic organization of these regions was represented with EasyFig BLASTn (https://mjsull.github.io/ Easyfig/; accessed on 7 May 2022).

Phylogenetic Analysis
A phylogenetic tree based on single-nucleotide polymorphisms (SNPs) was built with the CSI phylogeny 1.4 tool (https://cge.cbs.dtu.dk/services/CSIPhylogeny/; accessed on 6 May 2022 [55]). The pipeline was run with default parameters, setting a bootstrap replication value of 1,000 to generate the consensus tree [56]. Apart from the genomes of the HUMV isolates, 42 other S. Infantis genomes retrieved from databases were included in the analysis (see Table S3 for accession numbers and additional details). The genome of isolate 119944 from Israel (accession No. GCA_000506925.1) was used as the reference for SNP calling, and the resulting SNP matrix is shown in Table S4.

Conclusions
This study documents the import of the S. Infantis MDR clone which carries the bla CTX-M-65 within pESI-like megaplasmids from Peru into Spain. To the best of our knowledge, this is the first report on the presence of this clone in Spain, which has probably emerged in South America [46]. We also first experimentally demonstrated the resistance of the clone to mercury and to inorganic and organic arsenicals. Further studies will be required to establish the actual epidemiology of these and other CTX-M-producing S. Infantis lineages in Spain.

Supplementary Materials:
The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/antibiotics11060786/s1, Table S1: Accession numbers of the genomes cefotaxime resistance plasmid-mediated of Salmonella enterica serovar Infantis isolates from clinical samples and parameters related to the quality of the assemblies; Table S2: Origin and accession numbers of the genomes of Salmonella enterica serovar Infantis isolates used for phylogenetic analysis in the present study. Table S3: Pairwise SNP distance matrix calculated from the genomes of two Salmonella enterica serovar Infantis isolates from a Spanish hospital and 42 additional isolates obtained from different sources and countries. Table S4: Oligonucleotides designed for this work and used for the as-57 sembly of the resistance regions (R1 and R2) of pESI-like plasmids found 58 in Salmonella enterica serovar Infantis isolates from a Spanish hospital. References [57][58][59][60] are cited in the supplementary materials. Funding: This research was funded by project FIS PI17/00474 of the "Fondo de Investigación Sanitaria, Instituto de Salud Carlos III, Ministerio de Economía y Competitividad", Spain, cofunded by the European Regional Development Fund of the European Union: a way to making Europe. X.V. was the recipient of grant BP17-018 from the Program "Severo Ochoa" for support of Research and Teaching in the Principality of Asturias, Spain.
Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.
Data Availability Statement: The genome sequences generated in the present study were deposited in GenBank under accession numbers JAICDN000000000 and JAICDM000000000 for Salmonella enterica serovar Infantis HUMV 13-6278 and HUMV 15-5476, respectively.