Recent emergence of cephalosporin resistant Salmonella Typhi carrying IncFIB(K) plasmid encoding bla CTX-M-15 gene in India

The emergence and spread of Salmonella Typhi resistant to third generation cephalosporins are a serious global health concern. In this study, we have genomically characterized twelve cephalosporin resistant S . Typhi strains isolated from India. Comparative genome analysis of study isolates revealed the emergence of a new clone of ceftriaxone-resistant S . Typhi containing three plasmids of the incompatibility group IncFIB(K), IncX1 and IncFIB(pHCM2). Among the three, IncFIB(K) plasmid confers resistance to third-generation cephalosporins by means of bla CTX-M-15 gene, as well as other resistance determinants such as aph(3") , aph(6' ), sul2 , dfrA14 and tetA . Phylogenetic analysis of strains revealed that a single isolate belongs to a clade corresponding to genotype 4.3.1 and isolates from Ahmedabad ( n=11 ) belong to a distinct subclade within genotype 4.3.1.2 (H58 lineage II). SNP-based phylogenetic analysis of the core genes in IncFIB(K) revealed the plasmid backbone is closely related to that of IncFIB(K) from other Enterobacterales. The findings suggest that H58 lineage II can acquire MDR plasmids from other Enterobacteriales if compensatory evolution balances the cost of carrying the plasmids. Though, like previously reported, exposure to the third generation cephalosporins during the treatment may have selected these variants, this could indicate the beginning of a new wave of ceftriaxone resistant S . Typhi in India. The implementation of control measures such as vaccination, improved water, sanitation, etc., could be undertaken in areas where MDR or XDR S Typhi strains are prevalent.


Introduction
Enteric fever is a systemic febrile illness caused by human-restricted pathogens Salmonella enterica serovar Typhi and Paratyphi A, B, and C (Crump and Mintz, 2010). It continues to be a major cause of morbidity and mortality in low and middle-income countries in South 3 Asia, Southeast Asia, and Africa (Antillón et al., 2017). In most cases, typhoid infection is contracted from contaminated food or water, rather than through direct contact with an infected person. Globally typhoid fever is estimated to cause approximately 11 to 20 million cases and between 28,000 and 161,000 deaths annually ( The introduction of genome-based surveillance for typhoid has greatly improved our knowledge of the origins, transmission, and antibiotic resistance of S. Typhi isolates. The current population structure of S. Typhi strains identifies 85 haplotypes (H1 to H85), of which haplotype 58 (genotype 4.3.1) is highly pathogenic as well as multidrug resistant (Wong et al., 2015;. During the past few decades, H58 S. Typhi has become the global dominant lineage due to its efficient transmission and persistence within human populations (Pragasam et al., 2020;Carey et al., 2022 India, however, its long term persistence and stability was not completely understood (Jacob 5 et al., 2021). In this study, we aimed to analyze the genomic epidemiology and evolutionary origin of twelve cases of ESBL-positive S. Typhi in India. We also explored the recent plasmid transfer events and gene flow that led to the emergence of IncFIB(K) carrying cephalosporin resistant S. Typhi in India.

Ethical Statement
This study was approved by the Institutional Review Board (IRB) of Christian Medical College, Vellore (IRB Min No: 13489 dated 28.10.2020).

Study settings
The isolates used in this study were obtained from Neuberg Supratech Reference Laboratory A subset of three isolates were subjected to sequencing using Nanopore Technology (ONT).
The libraries for sequencing were prepared using Ligation Sequencing Kit (SQK-LSK114) (Oxford Nanopore Technologies, UK) and sequenced on an ONT MinION device using R10.4.1 flow cell and Q20 chemistry according to the instructions of the manufacturer.

QC, Assembly
Short reads generated from Illumina platform were checked for quality using FastQC v0.12.1 and adapters and indexes were trimmed using Trimmomatic, v0.39. Contamination detection and filtering has been performed by Kraken v2. and subsequent reads were used to estimate sequence coverage. The high quality reads hence generated were assembled using SPAdes (https://github.com/ablab/spades) with default settings.
Long read data from ONT sequencer were subsampled using Rasusa v0.7.1 to depths of 100x coverage (average) and reads shorter than 5000 bp were also removed using filtlong v0.2.1.
The reads were subjected to hybrid assembly using Unicycler and short reads were then used to polish further using polypolish v0.5.0. The complete genome hence generated were assessed by QUAST v5.2.0 and seqkit v2.4.0. Default parameters were used unless otherwise mentioned.

Comparative genomics
From the global collection of S. Typhi genomes (n=412), reads representing all major genotypes were obtained from the European Nucleotide Archive (ENA; http://www.ebi.ac.uk/ena) or the Sequence Read Archive (SRA). Genomes were classified according to their sequence types using the Multilocus Sequence Typing (MLST) pipeline offered by the Center for Genomic Epidemiology (https://cge.cbs.dtu.dk/services/MLST/). Genome collection of S. Typhi genotypes were further characterized using the genotyphi scheme as described on the GitHub repository (https://github.com/katholt/genotyphi).

Phylogenetic analysis
The assembled contigs (n=424) were mapped to the reference genome of S. Typhi CT18 (GenBank: AL513382.1) using Snippy v4.6.0. Single nucleotide polymorphisms (SNPs) were called from the core multiple-sequence alignment file using snp-sites v2.5.1 and maximumlikelihood phylogeny were inferred from the alignment using RAxML-NG v1. The plasmid sequences were annotated with Prokka (https://github.com/tseemann/prokka) using the default parameters. Annotated assemblies in the GFF3 format were used as input for pan-genome analysis using Panaroo (https://github.com/gtonkinhill/panaroo) in its 9 "Strict" mode with core threshold 80. The core genome alignment hence generated were used to construct plasmid core gene phylogeny using IQ-Tree v2.2.2.6 with parameters -m GTR+F+I+G4. We performed RhierBAPS to define the IncFIB(K) population structure and phylogenetic cluster carrying IncFIB(K) were further redefined to fully understand plasmid evolution. The gene presence or absence in each genome obtained were grouped according to the phylogenetic lineages using twilight scripts (https://github.com/ghoresh11/twilight) with default parameters.

Data availability
Raw read data were deposited in the European Nucleotide Archive (

Population structure of IncFIB(K) plasmids
Our curated dataset consist of IncFIB(K) plasmids (n=276) representing major Enterobacterales genera. The dataset includes plasmids from a diverse of bacterial hosts, with 53.9% from K. pneumoniae, 19.9% from S. Typhi, 9.6% from E. coli and 7.9% from other Klebsiella spp. Among the selected plasmids that are assigned as IncFIB(K) replicon types, MOB typing was performed and successfully classified into three MOB types, of which MOBV were the dominant MOB type. Notably twelve plasmids were assigned to multiple MOB types. Pangenome analysis IncFIB(K) plasmid sequence showed a well conserved backbone consisting of four genes (repB, sopA, umuC, umuD) which formed a typical core gene set.
After removing outliers and including our eleven newly sequenced IncFIB(K) plasmids, a final set of n=288 plasmid sequence were used for constructing a plasmid core gene phylogeny. The genetic relatedness of selected IncFIB(K) plasmids revealed a diverse population structure with six major (level 1) BAPS clusters (Fig 3). Within the plasmid phylogeny, IncFIB(K) plasmids carried by the study isolates belonged to cluster 1 represented by multiple bacterial host. To identify the near identical IncFIB(K) plasmids that were similar to those reported from study isolates, the entire plasmid sequences within BAPS cluster 1 (n=76) were examined. Detailed inspection revealed that IncFIB(K) plasmids carried by the study isolates were closer to plasmids hosted by S. Typhi isolates previously reported from Eastern Africa (LT904889) (Fig 4).

Discussion:
One of the salient features of the SEFI phase 2 surveillance was to spot the emergence of a new ceftriaxone resistant S. Typhi clone from India. Phenotypic characterization revealed that the emergent clone is also resistant to ampicillin, ciprofloxacin and trimethoprimsulfamethoxazole but susceptible to chloramphenicol ( Table 1). The resistance profile of the study isolates differs from that of the classical XDR strains, which are resistant to the first- Analysis of S. Typhi genome sequences reveal that ceftriaxone resistant strains from Ahmedabad (n=11) were of H58 lineage II (genotype 4.3.1.2) that is characterized by 12 chromosome-mediated quinolone resistance ( Fig. 1 and 2). The predominance of the same genotype in India has already been reported by multiple genome surveillance studies (Wong et al., 2015;Pragasam et al., 2020;Britto et al., 2020;Pragasam et al., 2021). However, circulating S. Typhi strains in western India, particularly in Ahmedabad, were not adequately represented in these genomic surveillance studies. Therefore, ten contextual (ceftriaxonesusceptible) Ahmedabad isolates were sequenced over a comparable period of time for the purposes of comparison. Nine of the ten contextual isolates were located closer to the ceftriaxone resistant strains and separated by 7 SNPs (Suppl Fig: 2). This suggests that the endemic H58 lineage II clone circulating locally may have acquired an ESBL-encoding AMR Typhi. Additionally, fitness costs associated with plasmid carriage have also been considered to be a contributing factor to S. Typhi's genetic stability (Jacob et al., 2021). Hence, in the absence of antibiotic selection pressure, S. Typhi has shown to be genetically conserved with minimal genomic variation due to SNPs, recombination or plasmid acquisitions ( (Dahiya et al., 2014). Notably, these QRDR double/ triple mutant strains rapidly displaced other lineages as a result of the fitness advantage gained during the evolutionary process (Baker et al., 2013). As a result, H58 lineage II isolates are likely to limit the acquisition of plasmids due to the high fitness cost imposed during the plasmid carriage (Jacob et al., 2021). Although previously H58 lineage II isolates were reported to carry plasmids that conferred ceftriaxone resistance, the plasmids were subsequently removed by purifying selection. Specifically, there have been sporadic isolates of the IncX3 and IncN plasmids carrying H58 lineage II isolates between 2015 and 2019. However, the plasmid-free isolates likely outcompeted plasmid-bearing isolates because of the fitness cost. Despite our expectations of rare plasmid carriage in H58 lineage II, a newly circulating clone of ceftriaxone-resistant isolates carrying three plasmids has been reported from Ahmedabad India.
We hypothesize that the emergence of ceftriaxone-resistant isolates from Ahmedabad India could be the result of a recent event of acquisition of multiple plasmids from other Enterobacteriaceae donor. The SNP-based phylogeny of the core genes in IncFIB(K) plasmid backbone identified six major clusters within the collection (Fig. 3). Based on a closer examination of cluster 2, where IncFIB(K) plasmids of study isolates were located, we were able to identify sequences of plasmids from S. Typhi, S. flexneri, E. coli and other Salmonella sp. with a pairwise distances of 0 to 1 core SNPs (Fig. 4). In particular, the IncFIB(K) plasmid carried by study isolates is closely related to the plasmid identified in a previously     plasmids from Ahmedabad were located. Metadata are labeled as color strips and key for each variable were mentioned. Strip 1 and 2 indicate the MOB type, and host organism.
Heatmap represents the plasmid mediated resistance genes. The ceftriaxone resistant isolates (red branches) are separated by 7 SNPs from the rest of the branch (black).