https://orcid.org/0000-0002-1534-7838
Figures
Abstract
Salmonella enterica serovar Agona (S. Agona) has been increasingly recognised as a prominent cause of gastroenteritis. This serovar is a strong biofilm former that can undergo genome rearrangement and enter a viable but non-culturable state whilst remaining metabolically active. Similar strategies are employed by S. Typhi, the cause of typhoid fever, during human infection, which are believed to assist with the transition from acute infection to chronic carriage. Here we report S. Agona’s ability to persist in people and examine factors that might be contributing to chronic carriage. A review of 2233 S. Agona isolates from UK infections (2004–2020) and associated carriage was undertaken, in which 1155 had short-read sequencing data available. A subset of 207 isolates was selected from different stages of acute and persistent infections within individual patients. The subset underwent long-read sequencing and genome structure (GS) analysis, as well as phenotyping assays including carbon source utilisation and biofilm formation. Associations between genotypes and phenotypes were investigated to compare acute infections to those which progress to chronic. GS analysis revealed the conserved arrangement GS1.0 in 195 isolates, and 8 additional GSs in 12 isolates. These rearranged isolates were typically associated with early, convalescent carriage (3 weeks– 3 months). We also identified an increase in SNP variation during this period of infection. We believe this increase in genome-scale and SNP variation reflects a population expansion after acute S. Agona infection, potentially reflecting an immune evasion mechanism which enables persistent infection to become established.
Author summary
The persistence of non-typhoidal Salmonella following acute gastroenteritis is a significant challenge for public health but the causes that lead to such chronic infections are understudied. This study aimed to investigate the ability of Salmonella Agona (S. Agona), the fourth most common non-typhoidal Salmonella serovar in the UK, to persist in humans and transition from acute illness to chronic infection. A phylogenomic analysis of 2233 S. Agona isolates (from all reported UK infections during 2004 to 2020) revealed there were no specific lineages that were responsible for persistent infection. A subset of 207 isolates underwent genotype and phenotypic analysis to explore within-patient changes and compare acute infections to those which progress to chronic. Key findings revealed increases in single nucleotide polymorphism and genomic structure variation during the early stages of persistent infection, indicating either population expansion after acute infection or a population bottleneck during transition to persistent infection.
Citation: Waters EV, Lee WWY, Ismail Ahmed A, Chattaway M-A, Langridge GC (2024) From acute to persistent infection: revealing phylogenomic variations in Salmonella Agona. PLoS Pathog 20(10): e1012679. https://doi.org/10.1371/journal.ppat.1012679
Editor: Andreas J. Baumler, University of California Davis School of Medicine, UNITED STATES OF AMERICA
Received: July 22, 2024; Accepted: October 17, 2024; Published: October 31, 2024
Copyright: © 2024 Waters et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All sequencing data generated in this study are available in the European Nucleotide Archive, with individual accession numbers for isolates analysed in this study are given in S1 Table. Illumina sequencing data are available under the Project accession PRJNA248792 and Nanopore sequencing data and hybrid assemblies have been deposited under Project accession PRJEB76791.
Funding: WWYL received a scholarship from the Medical Research Foundation in the National PhD Training Programme in Antimicrobial Resistance Research (MRF-145-0004-TGP-AVISO). EVW and GCL gratefully acknowledge the support of the Biotechnology and Biological Sciences Research Council (BBSRC); this research was funded by the BBSRC Institute Strategic Programmes Microbes in the Food Chain BB/R012504/1 and Microbes in Food Safety BB/X011011/1 and their constituent projects BBS/E/F/000PR10349 and BBS/E/QU/230002B. This study is funded by the National Institute for Health Research (NIHR) Health Protection Research Unit (NIHR200892) in Genomics and Enabling Data at University of Warwick in partnership with UKHSA, in collaboration with University of Cambridge and Oxford. MAC is based at UKHSA. The views expressed are those of the author(s) and not necessarily those of the NIHR, the Department of Health and Social Care or UKHSA. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Funder websites: https://www.medicalresearchfoundation.org.uk/ https://www.ukri.org/councils/bbsrc/ https://warwick.ac.uk/fac/cross_fac/hpruged/.
Competing interests: I have read the journal’s policy and the authors of this manuscript have the following competing interests:GCL has previously consulted for RevoluGen Ltd on bioinformatic analyses. Fire Monkey DNA extraction kits were provided free of charge by RevoluGen in this project.
Introduction
Salmonella, a ubiquitous Gram-negative organism, is responsible for salmonellosis, manifesting in humans as either enteric fever or gastroenteritis, caused by typhoidal and non-typhoidal Salmonella (NTS), respectively [1]. NTS cause an estimated 93.8 million global cases per year, of which 86% are attributed to foodborne gastroenteritis [2]. While NTS infections are typically self-limiting, the potential necessity for antimicrobial treatment has formidable healthcare financial implications [3].
Salmonella enterica serovar Agona (S. Agona) has been increasingly recognised as a prominent cause of foodborne gastroenteritis. In the European Union (EU), S. Agona has consistently featured in the top 20 serovars causing salmonellosis since reporting began in 2008 [4–10]. The UK, in particular, experiences an annual average of ~8500 NTS cases, of which S. Agona has always been among the top 10 most reported serovars [11–15]. A surge in cases in 2017 raised S. Agona from a historical ~1.7% prevalence [11,12] to 2.4%, sustaining this heightened level until 2019 and moving S. Agona into the top 5 most reported serovars in the UK [13–15]. Similar trends were seen across the EU, with S. Agona reaching its highest-ever ranking as the sixth most common serovar in 2017 [8].
This spike in S. Agona cases coincided with two multi-country outbreaks under investigation in 2017 [7]. One, encompassing 147 cases across five EU countries, was tentatively linked to ready-to-eat food products containing cucumbers; however, attempts to pinpoint precisely where contamination occurred in the production chain remained inconclusive [16]. The other, involving 39 cases across three EU countries, was traced back to infant formula produced in a French facility, which had previously been associated with a 141-case outbreak in 2005 [17–19]. Due to France’s lack of genomic surveillance pre-2017, it is unknown if these two outbreaks were caused by S. Agona that had persisted in the same facility for over a decade undetected. The ability of S. Agona to persist in dry food production environments has been documented previously. A US cereal facility suffered outbreaks separated by a decade was due to the persistence of an identical strain which was hypothesised to persist in wall cavities prior to reintroduced into the facility by a construction project [20,21], whilst Norwegian fish feed factories have seen evidence of persistence lasting at least three years due to ineffective decontamination procedures [22,23].
The ability of Salmonella to persist and remain viable for extended periods of time in low moisture environments is likely to be caused by many factors, which may vary between serovars and strains [24]. It is well documented that switching to low moisture environments is a trigger for increased levels of osmoprotectants, rRNA degradation, filament formation, outer membrane porin transcription and alternative sigma factors [24]. Salmonella are known for their capacity to form biofilms [25] and rearrange their genome [26], which both have the potential to enhance pathogenicity and alter phenotypic and metabolic capabilities. These survival mechanisms may also contribute to inducing a viable but non-culturable (VBNC) state, allowing persistence in food processing environments before more favourable conditions facilitate resuscitation [27].
S. Agona has been identified as a strong biofilm forming serovar [28] that can undergo genome rearrangement [29] and enter a VBNC state with ~1% of the population remaining metabolically active [27]. These strategies of persistence are employed by S. Typhi, the causative agent of typhoid fever, during human infection where it is hypothesised that genome rearrangement and biofilm establishment are required for the transition from acute infection to chronic carriage [30]. The frequency and mechanism by which acute S. Agona infections develop into chronic carriage in humans remain largely unknown. In this study we have reviewed the surveillance data for S. Agona, to assess the demographic changes in the number of cases and the use of HierCC as a potential method for sentinel surveillance with respect to geographical origin per isolate. We have also investigated the genomic diversity of S. Agona strains that persist in the human host and assessed relevant phenotypes that might contribute towards clinical persistence.
Results
Demographic overview of S. Agona infections in the UK
A total of 2233 isolates of S. Agona were received by UK Health Security Agency’s (UKHSA) Gastrointestinal Bacteria Reference Unit (GBRU) from England (2139) and Wales (94), between January 2004 and December 2020 inclusive. These were from 2098 patients with 105 patients having additional isolates referred after initial isolation. Patients reported to general practitioners or local hospitals with diarrhoea, abdominal pain and/or symptoms consistent with gastroenteritis. Most common specimen source as expected was faeces (2055, 92.0%), followed by urine (40, 1.8%), blood (16, 0.7%) and other sources (7, 0.3%) in which one isolate was found in the gallbladder. There were 115 isolates (5.2%) where the specimen type was not stated.
Of the 2098 patients, 1035 of the patients were male (49.3 %), 1054 were female (50.2%) and 9 cases were unknown, with the most common age range for infection being between 0–5 years (Fig 1A). Analysis of S. Agona cases referred to UKHSA’s GBRU from 2004 to 2020 demonstrates fluctuating trends in incidence rates, with notable increases observed in certain years (e.g. 2008, 2017 and 2018 with over 200 cases) and periods of decline (e.g. 2006 and 2014 with ~80 cases, the lowest being in 2020 with 39 cases during the COVID pandemic) (Fig 1B).
(A) Population pyramid of age and sex distribution of 2098 cases of infections. (B) Annual case numbers. (C) Neighbour-joining phylogeny categorised by year of isolation (inner ring), travel (second ring), most common HC5 clusters (third ring) and carriage status (outer ring) as denoted by colour. S. Agona strains associated with carriage are distributed throughout the population structure with domestic and travel associated clusters differentiated at the HC5 level.
Population structure analysis of the post genomic era (2014 onwards) showed clonal expansion of strains at the HC5 level. HC5 describes strains within 5 cgMLST alleles of one another, indicating clustered strains or closely related clones. Such clusters were identified in 2017 (HC5_70668, HC5_2186) and 2018 (HC5_2186) indicating that potential outbreaks may account for the increase of cases in these years though further epidemiological investigation would be needed to confirm (Figs 1C and S1). The majority of travel associated cases with S. Agona infection was from Southeast Asia with the top travel destinations being Pakistan (144, 6.9%), India (106, 5.1%), Thailand (66, 3.1%), Cuba (45, 2.1%), Egypt (38, 1.8%) and Afghanistan (34, 1.6%). Population structure analysis indicates certain clusters at the hierarchical cluster 5 level corresponds with travel to certain countries such as Pakistan (HC5_71609, 70763, 20641, 126490 & 2079), India (HC5_71231), Southeast Asia including Thailand or Vietnam (HC5_4181), Cuba (HC5_67370) and Japan (HC5_108552). There are also clonal groups not associated with travel indicating a potential domestic source in England (HC5_2186 & 7392) (Figs 1C and S1). Both domestic and international clusters have been around for several years showing that these strains have adapted well in their local niches and continue to thrive (Figs 1C and S1).
Persistent infection exists in S. Agona
Patients were classed as having a persistent infection if they provided an isolate at least 3 weeks after their initial infection. In total, 144/2233 isolates (6.4%) were attributed to persistent infection in 59 patients, which ranged from 22 days to over 6 years from first identified infection. Patients were classified into three bands of carriage depending on the time frame over which their isolates were obtained: convalescent carriage (>3 weeks–≤3 months), temporary carriage (>3 months– 12 months) and chronic carriage (>12 months) [31,32]. Across the dataset, we identified 86 isolates (3.9% of total collection) from 41 convalescent carriers, 36 isolates (1.6%) from 12 temporary carriers, and 22 isolates (1.0%) from 6 chronic carriers (S1 Table). An additional 34 isolates (1.5%) were collated from 17 patients who suffered from the same episode of acute infection, i.e. ≤3 weeks from the first submitted isolate (S1 Table). All subsequently viable isolates associated with persistent infection (133/144) and same episode (32/34), alongside an independent representative set from one-off acute infections (42), were long-read sequenced to generate hybrid assemblies with whole genome sequencing (WGS) data (S2 Table).
Strains associated with carriage are distributed throughout the population
To understand if carriage strains had any clonality, the population structure was assessed with the different carriage status of individual cases. To investigate the potential clonality of carriage strains, the population structure was evaluated based on the varying carriage statuses of individual cases. This analysis revealed that strains linked to all carriage categories were dispersed sporadically throughout the population structure. This suggests the absence of distinct clonal strains that have achieved long-term persistence and clonally expanded within the human population. Instead, it implies that numerous strains possess the capability for persistence without a single dominant clonal lineage (Fig 1C).
Genome analysis reveals increase in variation at transition to carriage
In accordance with the wider dataset, phylogenetic relationships between fully sequenced isolates from acute and persistent infection (S2 Fig) suggested that specific lineages were not responsible for persistent infection.
Where persistent infection did develop in a patient, we calculated the variation observed between the original isolate (first isolation) and all other isolates taken during the infection (S3 Table). By adjusting for time since first isolation, we were able to determine the average number of SNPs/year identified in isolates falling within the convalescent, temporary or chronic stage of infection (S3 Table). This revealed a much higher rate in the convalescent and temporary phases, averaged at 17–18 and 11–12 SNPs per year respectively. For the chronic phase the rate was 3–4 SNPs per year, which is a frequency comparable to that of 1–2 SNPs per year previously reported in S. Agona and other NTS serovars [33,34] though we caution that this is calculated from a small number of chronic isolates. Should this lower rate be supported by data from additional chronic isolates, this raises the possibility in convalescent and temporary phases of either hypermutation over short time span, or more likely, a larger bacterial population size being sampled from for diagnosis.
Investigating the 266 genes affected by the identified SNPs showed that 28 were observed across multiple patients. However, most of these were identified in a maximum of two patients (S3 Table). Seven genes were affected in 3/4 patients: fimD which fimbrial outer membrane usher protein, transcriptional regulation gene tdcA, formate hydrogenlyase transcriptional activator flhA, DNA-binding transcriptional regulation gene glpR which responds to sugar metabolism, response regulation gene arcA, and dacA and gyrB which encodes for D-alanine carboxypeptidase and DNA topoisomerase respectively. Additionally, SNPs in btuB which encodes for TonB-dependent vitamin B12 transporter was seen in 2/6 chronic carriage patients.
In addition to changes at the nucleotide level we also investigated larger changes at the chromosomal level. These genome rearrangements occur when homologous recombination around long-repeat sequences of ribosomal operons cause large genome fragments to shift position and/or orientation in the genome to ultimately produce different unique genome structures (GSs) without affecting the underlying nucleotide sequence. We have previously shown that S. Agona strains typically harbour the conserved genome structure (GS1.0) but are also capable of producing extremely unbalanced GSs [29]. In line with this, 94.2% (195/207) of isolates possessed the conserved structural genotype of GS1.0, which is a frequency comparable to that seen in other NTS serovars [29]. However, a total of 8 additional GSs were also identified across the collection (Fig 2A and 2B) and were observed only during the first 66 days of infection. Patients were considered carriers once they had harboured infecting bacteria for at least 21 days. In all bar one patient who submitted 2 samples with varied GSs, all subsequent isolates only displayed GS1.0.
(A) Number and distribution of genome structures (GSs) identified from acute/same episode infection (orange) or carriage (blue). (B) All GSs identified in the collection where GS designations are given as X.Y which represent the order (X) and orientation (Y) of the seven genome fragments (coloured blocks labelled 1–7) which are located between the seven ribosomal operons (open arrows) that split Salmonella genomes [29]. GS fragments are labelled numerically clockwise in respect to the Salmonella enterica reference LT2 (GS1.0) and individual inverted fragment orientations are denoted prime (′) with striped colours. Isolate numbers displaying each GS are shown. Dashed and solid black lines show the origin and terminus of replication, respectively.
Gains and losses in metabolic capacity during infection
To identify any changes in metabolic capacity during persistent infection, carbon source utilisation was investigated by comparing the first and last isolates of chronic carriers P1, P4 and P5, spanning a timeframe of 15 to 30 months per patient (S4 Table). For chronic carrier P6, carbon source utilisation was assessed longitudinally across 4 isolates spanning 16 months. Overall, metabolic capacity of isolates remained largely unchanged within each patient. However, it was striking to note that both gains and losses of metabolic ability were observed in later isolates (Figs 3 and S3). The final isolates of P1 and P5 exhibited a gain in ability to utilise glyoxylic acid (P1), and L-lyxose and L-threonine (P5). Conversely in P4, the final isolate showed a gain in ability to use lactulose but a loss in the ability to use 1,2-propanediol and alpha-keto-glutaric acid. In several cases, where the initial isolate (#127) of P6 was able to utilise the substrate, utilisation was much reduced in both the second (#129) and third isolates (#131) from the convalescent and temporary carriage phases of infection (e.g. 1,2-propanediol, alpha-keto-glutaric acid). This was even more pronounced for the third isolate (#131) alone, where utilisation of many carbon sources was reduced in comparison to all other isolates, including the most recent isolate from the chronic phase of infection (Figs 3 and S3). Interestingly, a comparative analysis of the SNP variation per isolate (S3 Table) with the accompanying metabolic capacity observed from the Biolog revealed that for patients 4 and 5 (Fig 3B, 3C, 3D, 3E, and 3F), one or no SNPs were identified, respectively. Therefore, nucleotide level variation alone cannot explain these metabolic changes. Similarly for patients 1 and 6 (Fig 3A, 3G, and 3H), a limited number of SNPs were observed whose relationships to the metabolic changes were assessed using EcoCyc (www.ecocyc.org), with only a connection to 1,2-propanediol utilisation identified (patient 1, SNP in btuB). These results raise the possibility of short-term adaptive changes in response to the host environment, enabling transcriptional modulation of substrate utilisation. Such modulation can therefore include a return of metabolic capacity as demonstrated by the final isolate taken from patient 6 (Fig 3G and 3H).
Highlighted comparisons of carbon utilisation of original and final isolates of chronic carriers P1 (A), P4 (B-D), P5 (E and F) and P6 (G and H). For P6, carbon source utilisation was also assessed for another two intermediate isolates. Keys are included with each graph showing isolates in chronical order from the original at the top to last. Highlighted carbon source is named below each graph.
Reduced biofilm ability associated with carriage under 12 months
The ability to form a biofilm is a key factor in S. Agona’s persistence both in and around food and has been cited in relation to disease outbreaks. We therefore assessed biofilm capacity across our collection using a crystal violet assay following growth in rich media (S4 Fig). Biofilm formation was observed across the collection, with only 3 isolates producing no discernible biofilm at all (S5 Table). We then investigated whether the biofilm ability of isolates varied in accordance with patient carriage status (S4 Fig). Our data revealed that isolates from patients with convalescent (p = 0.004) and temporary carriage (p = 0.002) of S. Agona had a significantly poorer ability to biofilm than isolates from patients with acute illness. To determine if this had a genetic basis, we focused on a targeted set of genes that have been previously identified as important for Salmonella biofilm formation in food production environments (S5 Table) [35,36]. The regulatory gene rpoS, the type III secretion system invasion gene invA and the attachment-related genes fliC and wcaA have been identified as essential/100% conserved genes in biofilm-forming Salmonella [35]. These genes were identified in all isolates in this study, except in isolate 10 which harboured an 80kb deletion encompassing the SPI-1 locus and surrounding genes including rpoS.
Other genes which have been identified as having a high association with biofilm-forming Salmonella include: the additional attachment-related genes adrA, csgB, csgD, fimH and glyA; the additional regulation genes csrA, ompR and sirA; quorum sensing-related genes luxS, pfs, sdiA; and genes sipB and sipC located on Salmonella pathogenicity island 1 (SPI-1) which encode for the needle tip of the type III protein secretion system complex. The majority of these genes were detected across the entire collection. However, the large deletion in isolate 10 likely explains why it did not produce a biofilm. We also looked for any amino acid changes in this gene set, identifying 10/17 as being fully unaffected (S5 Table). A range of SNPs were detected that lead to synonymous substitution or changes that were predicted to have no effect on protein function (S5 Table). Only 3 isolates harboured nonsynonymous changes predicted to impact function: filC in isolate 61, sirA in isolate 111 and wcaA in isolate 10 (S5 Table).
Plasmid presence associated with majority of resistance genes
A total of 38 antimicrobial resistance genes (ARGs), in addition to point mutations in genes encoding DNA gyrase and topoisomerase IV (gyrA/parC) were identified across 92 isolates in the collection (S5 Table), associated with resistance to ampicillin, ceftriaxone, chloramphenicol, fluoroquinolones, aminoglycosides, lincomycin, rifampicin, sulfonamide, tetracycline and trimethoprim. The distribution of these resistance determinants across patients with acute and persistent infection was similar: eighteen isolates related to 38 patients with acute illness, twenty isolates to 10 patients from the same episode and fifty-four isolates from 24 patients with persistent infection. Multidrug resistant (MDR) profiles (defined as resistance to three or more antimicrobial classes) were observed in sixty-seven isolates in the collection from 37 patients, thirteen were associated with acute illness, eighteen isolates related to 9 patients with isolates from the same episode and the remaining thirty-six isolates were from 15 patients with persistent infection. AMR profiles were largely stable within patients with persistent infection (21/24 patients), with only 3 patient AMR profiles showing variable presence between isolates (S5 Table).
A total of 15 different plasmid replicons were identified across 62 isolates in the collection (S5 Table), including small cryptic plasmids, Salmonella virulence plasmids IncFII(S) and IncFIB(S) and other Inc types. Individual isolates harboured between 1 and 3 replicons each. Eleven isolates related to 11 patients with acute illness, fourteen to 7 patients with isolates from the same episode and the remainder was distributed across repeat isolates from 17 patients with persistent infection. This allowed us to investigate whether plasmid replicon presence was stable or varied within patients over time. The closest time points were from patients with isolates taken in the same episode of illness, spanning a maximum of 20 days. Within these patients, plasmid presence was largely stable (7/10 replicons), with only 3 replicons showing variable presence between isolates. In patients with persistent infection, variation was more pronounced, with 12 stable replicons within patients, and 15 showing variable presence. The variable replicons were either Col or Inc types, including the virulence plasmids IncFII(S) and IncFIB(S).
Across the 91 isolates with AMR determinants, 32 isolates were confirmed to have some or all ARGs located on plasmids. Changes in AMR profiles between isolates from persistent infection were associated with the acquisition (3/4 patients) or loss (1/4) of Inc type plasmids containing AMR determinants. Limited evidence of plasmid integration into chromosome was found in 2 acute patient isolates (#63, #64) where plasmid replicons were identified in the chromosome and in the later isolate of convalescent patient P54 (#158) where ARGs originally located on a IncX1 plasmid had integrated into the chromosome.
All isolates have a similar genetic capacity for virulence
A total of 154 virulence factors were identified in 207 isolates, alongside 4 additional genes that have been identified as major virulence factors responsible for salmonellosis [37–39] including: the regulation gene hilA, the type I secretion system gene siiD, the enterotoxin gene stn, and fimA which encodes for a fimbrial protein (S5 Table). Demonstrating the virulence capacity of S. Agona, 90/158 of these genes were detected across the entire collection. Isolate 10 again lacked a high number of virulence genes in comparison to the rest of the collection, namely hilA, sptP and avrA and operons inv, org, prg, sic, sip and spa. Salmonella plasmid virulence spv genes and pefA were only found in isolates 71 and 96, corresponding to the presence of the IncFIB(S) plasmid in these isolates. No specific association between virulence factor presence and length of infection was identified.
Discussion
This research has employed routinely generated genomic data at UKHSA to gain deeper insights into the trends, prevalence and phylogenomics of S. Agona isolated from domestic and returning travellers in England and Wales with the view to identify and understand long term persistence in clinical infection.
While these findings can serve as a proxy for sentinel surveillance to monitor strains in different geographical regions [32] continuous real-time comparison with global data remains crucial to validate this approach and identify novel clones. In this study, HierCC at the HC5 level [40] has proven to be a valuable tool and typing scheme for assessing clonal groups across S. Agona strains and establishing links between population structure and case demographics (Figs 1C and S1). Although the study concentrates on S. Agona, it underscores the potential application of this approach across non-typhoidal Salmonella strains. Whilst S. Agona’s ability to persist in food processing environments and products is well documented [18,21,23], the development of S. Agona infections from acute into carriage is less understood. In this study we utilised a comprehensive set of 2233 S. Agona isolates gathered over a 16-year period from routine public health surveillance. From this collection, 6.4% of isolates were attributed to persistent infection in 59 patients, which accounted for 2.8% of the total patients in this study. This conversion percentage from acute infection to chronic carriage is similar to the 2–5% previously seen in NTS serovars and the more well studied S. Typhi [31,41]. The ability of S. Agona strains to cause persistent or acute infections does not appear to be hereditary; isolates from all carriage states were located across the phylogenetic tree demonstrating no connection between lineage and infection period. This supports similar findings by Grote and colleagues [42] who demonstrated that persistent salmonellosis can be caused by a range of NTS serovars, again indicating that persistence can occur across diverse lineages rather than being limited to specific ones. For S. Typhi it is hypothesised that genome rearrangement and biofilm formation are required for the transition to chronic carriage [30]; however, it was unknown if this would be the case for S. Agona. Our genomic analysis of a 207 isolate subset revealed that genomic rearrangement was relatively uncommon, but when it did happen, it was associated with early stages of infection. Alongside the evidence of greater SNP variation early in infection, this suggests an increase in the bacterial population during this period, plausibly required to support the transition from acute to persistent infection. An alternative explanation could be that a mixed S. Agona population caused these infections, which could have been present on the original food source and are now in a flux within the patient [33]. It appears clear that some type of bottleneck occurs in persistent infection as variation at later stages (e.g. in temporary and chronic carriage) was lower, with only one genome arrangement observed, and very little SNP variation. Our observed SNP variation can be interpreted as in-host evolution, as either a cause or consequence of persistent infection. If causal, S. Agona strains may have developed immune evasion strategies though this is dependent upon their niche during persistent infection. Perhaps more likely, persistent infection may be driven by an immunocompromised host or a dysbiotic gut microbiome, with S. Agona able to adapt and thrive in this environment.
It is well established that Salmonella which are highly or specifically host-adapted have reduced metabolic capacity intrinsically linked to the ability to cause systemic and persistent infections. Whilst S. Agona is not considered to be adapted to the human host, we wanted to understand whether S. Agona genomes of persistent isolates displayed any hallmarks of these adaptive traits. A prime example is the utilisation of 1,2-propanediol, which requires the presence of vitamin B12 via BtuB-dependent transport. The inability to utilise 1,2-propanediol is highly associated with the host restricted serovars such as S. Typhi, S. Gallinarum and S. Pullorum [43,44] which cause systemic, typhoid or typhoid-like disease. It has also been associated with the evolution of invasive NTS including S. Typhimurium ST313 [45] and S. Enteritidis in Africa [46]. We observed this loss of function three times in our test set of 4 patients (patients 1, 4 and 6), all in isolates from carriage stages of infection. In patient 1 this was accompanied by a SNP in btuB, which we also observed in patient 3, whose metabolic capacity was not assessed. Together, this raises the possibility that the loss of 1,2-propanediol utilisation could represent a marker of long-term persistent S. Agona infections.
The reduced biofilm ability associated with persistent infections of under 12 months did not appear to have a genetic basis, as almost all genes associated with Salmonella biofilm formation were found to be present across all isolates excluding one non-biofilm former, and the amino acid analysis indicated these were fully intact across all but 2 biofilm-forming isolates. This suggests that while these S. Agona strains have the capacity to form biofilms, transcriptional rewiring likely occurs during infection that reduces biofilm production, and these changes are maintained during subsequent laboratory growth. It is important to note that the biofilm assays presented here are unlikely to accurately reflect the biological conditions under which S. Agona persists within the human host. Once more is reported and understood about the specific bodily sites associated with persistent NTS infection, it will be of value to assess biofilm forming ability in such conditions.
Overall, the distribution of AMR genes was largely stable within persistently infected patients with, though plasmid presence was more varied. There was also little evidence for plasmids integrating into the chromosome during the infection timeline. These findings are comparable to a previous study which reported comparably stable AMR and plasmid profiles during persistent NTS infection [42].
S. Agona does not typically harbour IncF virulence plasmids which carry the spv operon and are commonly seen in Salmonella, but a version, spvC, has occasionally been found [47,48] and a megaplasmid pESI, which is predicted to provide MDR, increase virulence, and improve fitness and persistence, has started to circulate [49]. Our results support that plasmid association is not a key component of S. Agona genomes, with only 2 acute isolates harbouring the IncFIB(S) plasmid and associated spv genes, but the plasmids present do attribute a considerable proportion of the AMR (42.7%) seen in this collection.
S. Agona are capable of causing persistent clinical infections that can take weeks to years to resolve. Here we have identified SNP variation and changes in genome structure during early infection, suggestive of a role in the transition from acute to persistent infection.
Methods
Ethics statement
A collection of 2233 S. Agona isolates referred to UKHSA’s GBRU, from local diagnostic laboratories in England and Wales, between 1 January 2004 and 31 December 2020 were included in this study. Ethical approval for the detection, identification, typing and antimicrobial resistance characterisation of gastrointestinal bacterial pathogens from infected patients submitted to the GBRU is not required as this is covered by UKHSA’s surveillance mandate. Patient consent was not required, as UKHSA has authority to handle patient data for public health monitoring and infection control under section 251 of the UK NHS Act of 2006.
Bacterial isolates included in this study
For patients presenting with persistent infections (i.e. with multiple isolates), the isolation date was assessed, and carriage status was allocated as one of four categories: same episode (≤3 weeks from the first submitted isolate), convalescent carriage (>3 weeks–≤3 months), temporary carriage (>3 months– 12 months) and chronic carriage (>12 months) [31,32].
Whole genome sequencing
WGS was conducted as part of routine sequencing of all Salmonella isolates referred to UKHSA’s GBRU from 2014 onwards, as previously described [50]. Prior to 2014, Salmonella isolates were speciated and sub-speciated using PCR [51–53]. Selected strains isolated before 2014 were cultured from archived stocks prior to sequencing. Briefly, DNA was extracted using QIAsymphony DSP DNA midi kits (Qiagen) before being fragmented and tagged for multiplexing with NexteraXT DNA Preparation Kits (Illumina) and sequenced on an Illumina HiSeq 2500, yielding 100 bp paired end reads. Trimmomatic [54] was used to trim sequence data with leading and trailing at <Q30. All subsequent analysis was carried out on trimmed sequences to identify Salmonella species, subspecies, serovar and antimicrobial resistant determinants as previously described [52,53]. FASTQ data is available from the NCBI Short Read Archive, BioProject accession PRJNA248792. Individual accession numbers for isolates analysed in this study are given in S1 Table.
Population structure of S. Agona
Raw WGS data files of 1125 isolates from cases in England were uploaded to EnteroBase (https://enterobase.warwick.ac.uk/) and were assembled using the then current backend pipelines (v3.61–4.1) including cgMLST analysis to produce a cgST using cgMLST (v2) and HierCC (v1) algorithms as previously described [55,56]. There were 1078 isolates that met the cgMLST quality parameters for Salmonella (size >4000 - ≤5800 kbp, N50 ≤20 kbp, number contigs ≤600, low-quality sites ≤5%, minimum taxonomic purity ≥70%) [40] and were included for analysis (S1 Table). The minimum spanning tree was created in EnteroBase using the MSTree (v2) algorithm and visualizing on GrapeTree [56]. Previous studies have shown that analysing strains at the 5 SNP threshold might be appropriate to detect clusters or closely related clones, and that cgMLST is equivalent to SNP when detecting clusters [32,53,57–59]. Therefore, HierCC was analysed at the five allelic level (HC5 –strains linked within five cgMLST alleles) for trend analysis in association with travel, year of isolation and carriage status patterns. Phylogenetic analysis was undertaken using the Ninja Neighbour Joining method [60] and visualized on iTOL (v5) [61]. The historical isolates (2004–2013) were not included in the interpretation of results.
Nanopore sequencing and hybrid assembly
A subset of 207 isolates (S2 Table), comprising all viable persistent infection isolates (133) and an independent representative set from solely acute infection (74), were long-read sequenced at the Quadram Institute to generate hybrid assemblies with WGS data. High-molecular weight genomic DNA was extracted using Fire Monkey kits (RevoluGen) following manufacturer’s instructions. Long-read sequencing libraries, containing 48 DNA samples, were prepared using the Ligation Sequencing Kit (SQK-LSK109, Oxford Nanopore Technologies (ONT)) in conjunction with the Native Barcoding Expansion 96 Kit (EXP-NBD196, ONT). Full details of library prep are provided in supplementary information (S1 Data). These libraries were loaded and sequenced on MinION R9.4 flow cells (FLOMIN106, ONT) with run times of up to 120 hours. Raw sequencing data was collected using ONT MinKNOW software (v4.0.5) and subjected to local base-calling, de-multiplexing and barcode trimming using ONT Guppy (v5.0.11).
The open platform Galaxy (v19.05) was used to create hybrid assemblies and perform subsequent bioinformatic analysis: long-read sequences were filtered for high-quality using Filtlong (v0.2.0) (https://github.com/rrwick/Filtlong) assembled using Flye (v2.5) [62], and polished with two rounds of Racon (v1.3.1.1) [63] and one round of Medaka (v0.11.5, ONT). Hybrid assemblies were generated by using WGS data to polish the long-read assemblies with two rounds of Pilon (v1.20.1) [64]. Minimap2 (v2.17) [65] was used to provide the alignment of reads to the Flye and long-read assemblies required by Racon and Pilon, respectively.
Genomic analysis
AMR and plasmid profiles of hybrid assemblies were determined using staramr (v.0.10.0) [66] with the ResFinder [67], PlasmidFinder [68] and Salmonella PointFinder [69] databases, downloaded on 24 May 2022, 18 January 2023 and 1 February 2021, respectively. AMR profiles generated from these databases were checked for consistency with the UKHSA pipeline output. All discrepancies were investigated using short-read assemblies generated with Shovill (v1.1.0) (https://github.com/tseemann/shovill) and were found to be due to low coverage regions in hybrid assemblies. Virulence factor profiles of all hybrid assemblies were determined using ABRicate (v.0.9.7) (https://github.com/tseemann/abricate) with the virulence factor database [70] downloaded on 5 November 2021. If any differences in AMR profiles were observed between isolates from the same patient, or if a hybrid assembly had multiple genomic contigs, profiles were confirmed using short-read Shovill assemblies. Variation in biofilm and additional virulence genes responsible for salmonellosis was determined using pangenome analyses conducted with Roary (v.3.13.0) [71] from annotated files generated with Bakta (v1.6.1) [72]. Additionally, amino acid substitutions within biofilm genes were identified by aligning in SnapGene software (www.snapgene.com) and their effects were predicted using the PredictProtein web server [73].
Genome structures were identified either by socru (v2.2.4) [29] or by manual alignment and visualisation in Artemis Comparison Tool (v18.0.2) [74]. Nucleotide variation was assessed in the 133 isolates that were associated with persistent infection using Snippy (v4.4.3) (https://github.com/tseemann/snippy) where short reads were aligned to the hybrid assembly of the earliest isolate for each patient.
Phenotypic analysis
The biofilm abilities of the 207 isolate subset were determined using a crystal violet (CV) assay. Isolates were grown overnight in LB broth containing no additional NaCl [75] (denoted LB-NaCl). Optical density was measured with a Benchmark plus microplate spectrometer system (Bio-Rad) at 600 nm (OD600) to ascertain an inoculation factor. Cultures were then 1:1000 diluted into 200 μL LB-NaCl to give an OD600 of ~0.001 in 96-well polystyrene plates. After 48 hours incubation at 30°C covered in gas-permeable seals, the wells were emptied and rinsed with water before the residual biofilms were stained for 10 mins with 200 μL 0.1% CV. The excess, unbound CV was then removed, and the wells were rinsed with water. The CV bound biofilms were then solubilised in 200 μL 70% ethanol and the absorbance was measured at 590 nm. Variations in inoculation amounts were accounted for using the inoculation factor, which were then used to correct measured CV amounts. This measurement was transformed to give a relative value in comparison to the control S. Typhimurium strain 14028S (accession: CP001363), which was set at 100% [76–78]. The biofilm dataset was tested and found to be approximately normally distributed by the Kolmogorov-Smirnov test and through graphical visualisation. A two-sample t-test was performed to compare each carriage status group (same episode, convalescent, temporary and chronic) to the acute illness group.
Carbon source utilisation was assessed for the first and last isolates from 4 of the chronic carriers: patients 1, 4, 5 and 6. For patient 6, two additional isolates from temporary and convalescent stages of infection were also assessed. Strains were prepared from agar plates as per the manufacturer’s instructions (Biolog). PM1 panels, used with dye A, for each strain were incubated for 48 hours at 37°C in the Omnilog (Biolog). Data were exported from the Biolog File Manager and loaded into the R package CarboLogR [79]. Using the RunAnalysis() command, QC was performed and identified outlier panels that were excluded from downstream analysis. All strains had a minimum of 3 replicates that passed QC. Visualisation was performed in R (v4.2.273), using the tidyverse packages ggplot2 (v3.4.375), dplyr (v1.1.378) and gridExtra (v2.3).
Supporting information
S1 Fig. Highlighted phylogeny of S. Agona infections in England and Wales, from 2004–2020.
Neighbour-joining phylogeny categorised by year of isolation (inner ring), travel (second ring), most common HC5 clusters (third ring) and carriage status (outer ring) as denoted by colour.
https://doi.org/10.1371/journal.ppat.1012679.s001
(PDF)
S2 Fig. Distribution of isolates from acute and persistent infection.
Maximum likelihood tree of hybrid assemblies shows patient carriage status (key), strong biofilm formers (star) & rearranged genomes (bold text) are distributed across the tree.
https://doi.org/10.1371/journal.ppat.1012679.s002
(PDF)
S3 Fig. Comparisons of carbon utilisation of original and final isolates of chronic carriers P1, P4, P5 and P6.
For P6, carbon source utilisation was also assessed for another two intermediate isolates. Keys included with each graph showing isolates in chronical order from the original at the top to last. Biolog PM1 panel carbon sources are alphabetically ordered left to right, top to bottom.
https://doi.org/10.1371/journal.ppat.1012679.s003
(PDF)
S4 Fig. Biofilm ability of isolates from acute and persistent infection.
Biofilm ability is given as percentage relative to the control S. Typhimurium strain 14028S (accession: CP001363), which was set at 100% (Prouty and Gunn 2003 [76]; García et al. 2004 [77]; Trampari et al. 2019) [78].
https://doi.org/10.1371/journal.ppat.1012679.s004
(JPG)
S1 Table. List of all isolates presented in this study.
https://doi.org/10.1371/journal.ppat.1012679.s005
(XLSX)
S2 Table. List of subset of 207 isolates, including data accessions.
https://doi.org/10.1371/journal.ppat.1012679.s006
(XLSX)
Acknowledgments
The authors would like to thank Alice Nisbet for assistance with visualising the Biolog data and Emma Holden and Mark Webber (Quadram Institute) for providing biofilm-forming positive and negative control strains S. Typhimurium 14028S and 14028S ΔtolC::cat.
References
- 1. Achtman M, Wain J, Weill F-X, Nair S, Zhou Z, Sangal V, et al. Multilocus sequence typing as a replacement for serotyping in Salmonella enterica. PLoS pathogens. 2012; 8(6): e1002776. pmid:22737074
- 2. Majowicz SE, Musto J, Scallan E, Angulo FJ, Kirk M, O’Brien SJ, et al. The global burden of nontyphoidal Salmonella gastroenteritis. Clinical infectious diseases. 2010; 50(6): 882–9. pmid:20158401
- 3. Eng S-K, Pusparajah P, Ab Mutalib N-S, Ser H-L, Chan K-G, Lee L-H. Salmonella: A review on pathogenesis, epidemiology and antibiotic resistance. Frontiers in life science. 2015; 8(3): 284–93.
- 4. EFSA, ECDC. The European Union summary report on trends and sources of zoonoses, zoonotic agents and food-borne outbreaks in 2009. EFSA journal. 2011; 9(3): 1–378.
- 5. EFSA, ECDC. The European Union summary report on trends and sources of zoonoses, zoonotic agents and food-borne outbreaks in 2011. EFSA journal. 2013; 11(4): 1–250.
- 6. EFSA, ECDC. The European Union summary report on trends and sources of zoonoses, zoonotic agents and food-borne outbreaks in 2014. EFSA journal. 2015; 13(12): 1–190.
- 7. EFSA, ECDC. The European Union summary report on trends and sources of zoonoses, zoonotic agents and food-borne outbreaks in 2017. EFSA journal. 2018; 16(12): 1–262. pmid:32625785
- 8. EFSA, ECDC. The European Union One Health 2019 Zoonoses Report. EFSA journal. 2021; 19(2): 1–286. pmid:33680134
- 9. EFSA, ECDC. The European Union One Health 2021 Zoonoses Report. EFSA journal. 2022; 20(12). pmid:36524203
- 10. EFSA, ECDC. The European Union One Health 2022 Zoonoses Report. EFSA Journal. 2023; 21(12). pmid:38089471
- 11.
Public Health England. Salmonella data 2006 to 2015. 2016. https://doi.org/https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/598401/Salmonella_2016_Data.pdf
- 12.
Public Health England. Salmonella data 2007 to 2016. 2018. https://doi.org/https://assets.publishing.service.gov.uk/media/5b0d6c13e5274a7b496e4607/salmonella_data_2007_to_2016_may_2018.pdf
- 13.
UK Health Security Agency. Non-typhoidal Salmonella data 2010 to 2019. 2021. https://doi.org/https://assets.publishing.service.gov.uk/media/616d4e5f8fa8f52986e61cd7/salmonella-annual-report-2019.pdf
- 14.
UK Health Security Agency. Non-typhoidal Salmonella data 2008 to 2017. 2021. https://doi.org/https://assets.publishing.service.gov.uk/media/6169b905e90e07197483b750/salmonella-annual-report-2017.pdf
- 15.
UK Health Security Agency. Non-typhoidal Salmonella data 2009 to 2018. 2021. https://doi.org/https://assets.publishing.service.gov.uk/media/6169b8fad3bf7f55fa926a22/salmonella-annual-report-2018.pdf
- 16. EFSA, ECDC. Multi-country outbreak of Salmonella Agona infections possibly linked to ready-to-eat food. EFSA supporting publication. 2018; EN-1465: 1–15.
- 17. EFSA, ECDC. Multi-country outbreak of Salmonella Agona infections linked to infant formula. EFSA supporting publication. 2018; EN-1365: 1–9.
- 18. Jourdan-da Silva N, Fabre L, Robinson E, Fournet N, Nisavanh A, Bruyand M, et al. Ongoing nationwide outbreak of Salmonella Agona associated with internationally distributed infant milk products, France, December 2017. Euro surveillance. 2018; 23(2): 1–5. pmid:29338811
- 19. Brouard C, Espié E, Weill FX, Kérouanton A, Brisabois A, Forgue AM, et al. Two consecutive large outbreaks of Salmonella enterica serotype Agona infections in infants linked to the consumption of powdered infant formula. The pediatric infectious disease journal. 2007; 26(2): 148–52. pmid:17259878
- 20. Russo ET, Biggerstaff G, Hoekstra RM, Meyer S, Patel N, Miller B, et al. A recurrent, multistate outbreak of Salmonella serotype Agona infections associated with dry, unsweetened cereal consumption, United States, 2008. Journal of food protection. 2013; 76(2): 227–30. pmid:23433369
- 21. Hoffmann M, Miller J, Melka D, Allard MW, Brown EW, Pettengill JB. Temporal dynamics of Salmonella enterica subsp. enterica serovar Agona isolates from a recurrent multistate outbreak. Frontiers in microbiology. 2020; 11: 1–10. pmid:32265893
- 22. Nesse LL, Nordby K, Heir E, Bergsjoe B, Vardund T, Nygaard H, et al. Molecular analyses of Salmonella enterica isolates from fish feed factories and fish feed ingredients. Applied and environmental microbiology. 2003; 69(2): 1075–81. pmid:12571031
- 23. Vestby LK, Møretrø T, Langsrud S, Heir E, Nesse LL. Biofilm forming abilities of Salmonella are correlated with persistence in fish meal- and feed factories. BMC veterinary research. 2009; 5(20). pmid:19473515
- 24. Finn S, Condell O, McClure P, Amézquita A, Fanning S. Mechanisms of survival, responses, and sources of Salmonella in low-moisture environments. Frontiers in microbiology. 2013; 4: 1–15. pmid:24294212
- 25. Galié S, García-Gutiérrez C, Miguélez EM, Villar CJ, Lombó F. Biofilms in the food industry: Health aspects and control methods. Frontiers in microbiology. 2018; 9: 1–18. pmid:29867809
- 26. Waters EV, Tucker LA, Ahmed JK, Wain J, Langridge GC. Impact of Salmonella genome rearrangement on gene expression. Evolution letters. 2022; 6(6): 426–37. pmid:36579163
- 27. Habimana O, Nesse LL, Møretrø T, Berg K, Heir E, Vestby LK, et al. The persistence of Salmonella following desiccation under feed processing environmental conditions: A subject of relevance. Letters in applied microbiology. 2014; 59(5): 464–70. pmid:25046569
- 28. Díez-García M, Capita R, Alonso-Calleja C. Influence of serotype on the growth kinetics and the ability to form biofilms of Salmonella isolates from poultry. Food microbiology. 2012; 31(2): 173–80. pmid:22608221
- 29. Page AJ, Ainsworth EV, Langridge GC. socru: typing of genome-level order and orientation around ribosomal operons in bacteria. Microbial genomics. 2020; 6: 1–6. pmid:32584752
- 30. Crawford RW, Rosales-Reyes R, Ramirez-Aguilar M de L, Chapa-Azuela O, Alpuche-Aranda C, Gunn JS. Gallstones play a significant role in Salmonella spp. gallbladder colonization and carriage. Proceedings of the National Academy of Sciences. 2010; 107(9): 4353–8. pmid:20176950
- 31. Gunn JS, Marshall JM, Baker S, Dongol S, Charles RC, Ryan ET. Salmonella chronic carriage: Epidemiology, diagnosis, and gallbladder persistence. Trends in microbiology. 2014; 22(11): 648–55. pmid:25065707
- 32. Chattaway MA, Gentle A, Nair S, Tingley L, Day M, Mohamed I, et al. Phylogenomics and antimicrobial resistance of Salmonella Typhi and Paratyphi A, B and C in England, 2016–2019. Microbial genomics. 2021; 7(8): 1–15. pmid:34370659
- 33. Bloomfield SJ, Janecko N, Palau R, Alikhan N-F, Mather AE. Genomic diversity and epidemiological significance of non-typhoidal Salmonella found in retail food collected in Norfolk, UK. Microbial genomics. 2023; 9(7): 1–12. pmid:37523225
- 34. Zhou Z, McCann A, Litrup E, Murphy R, Cormican M, Fanning S, et al. Neutral genomic microevolution of a recently emerged pathogen, Salmonella enterica serovar Agona. PLoS genetics. 2013; 9(4). pmid:23637636
- 35. Yin B, Zhu L, Zhang Y, Dong P, Mao Y, Liang R, et al. The characterization of biofilm formation and detection of biofilm-related genes in Salmonella isolated from beef processing lants. Foodborne pathogens and disease. 2018; 15(10): 660–7. pmid:30199274
- 36. Sharan M, Vijay D, Dhaka P, Bedi JS, Gill JPS. Biofilms as a microbial hazard in the food industry: A scoping review. Journal of applied microbiology. 2022; 133(4): 2210–34. pmid:35945912
- 37. Boddicker JD, Knosp BM, Jones BD. Transcription of the Salmonella invasion gene activator, hilA, requires HilD activation in the absence of negative regulators. Journal of bacteriology. 2003; 185(2): 525–33. pmid:12511499
- 38. Chaudhary JH, Nayak JB, Brahmbhatt MN, Makwana PP. Virulence genes detection of Salmonella serovars isolated from pork and slaughterhouse environment in Ahmedabad, Gujarat. Veterinary world. 2015; 8(1): 121–4. pmid:27047008
- 39. Klingl S, Kordes S, Schmid B, Gerlach RG, Hensel M, Muller YA. Recombinant protein production and purification of SiiD, SiiE and SiiF—Components of the SPI4-encoded type I secretion system from Salmonella Typhimurium. Protein expression and purification. 2020; 172. pmid:32251835
- 40. Zhou Z, Alikhan N-F, Mohamed K, Fan Y, the Agama Study Group, Achtman M. The EnteroBase user’s guide, with case studies on Salmonella transmissions, Yersinia pestis phylogeny, and Escherichia core genomic diversity. Genome research. 2020; 30(1): 138–52. pmid:31809257
- 41. Marzel A, Desai PT, Goren A, Schorr YI, Nissan I, Porwollik S, et al. Persistent infections by nontyphoidal Salmonella in humans: Epidemiology and genetics. Clinical infectious diseases. 2016; 62(7): 879–86. pmid:26740515
- 42. Grote A, Piscon B, Manson AL, Adani B, Cohen H, Livny J, et al. Persistent Salmonella infections in humans are associated with mutations in the BarA/SirA regulatory pathway. Cell host and microbe. 2024; 32: 79–92. pmid:38211565
- 43. Holt KE, Thomson NR, Wain J, Langridge GC, Hasan R, Bhutta ZA, et al. Pseudogene accumulation in the evolutionary histories of Salmonella enterica serovars Paratyphi A and Typhi. BMC genomics. 2009; 10. pmid:19159446
- 44. Langridge GC, Fookes M, Connor TR, Feltwell T, Feasey N, Parsons BN, et al. Patterns of genome evolution that have accompanied host adaptation in Salmonella. Proceedings of the national academy of sciences of the United States of America. 2015; 112(3): 863–8. pmid:25535353
- 45. Okoro CK, Barquist L, Connor TR, Harris SR, Clare S, Stevens MP, et al. Signatures of adaptation in human invasive Salmonella Typhimurium ST313 populations from sub-Saharan Africa. PLoS neglected tropical diseases. 2015; 9(3): e0003611. pmid:25803844
- 46. Feasey NA, Hadfield J, Keddy KH, Dallman TJ, Jacobs J, Deng X, et al. Distinct Salmonella Enteritidis lineages associated with enterocolitis in high-income settings and invasive disease in low-income settings. Nature genetics. 2016; 48(10): 1211–7. pmid:27548315
- 47. Silva C, Puente JL, Calva E. Salmonella virulence plasmid: Pathogenesis and ecology. Pathogens and disease. 2017; 75(6): 1–5. pmid:28645187
- 48. Manning J, Gole V, Chousalkar K. Screening for Salmonella in backyard chickens. Preventive veterinary medicine. 2015; 120(2): 241–5. pmid:25899620
- 49. dos Santos AMP, Panzenhagen P, Ferrari RG, Conte-Junior CA. Large-scale genomic analysis reveals the pESI-like megaplasmid presence in Salmonella Agona, Muenchen, Schwarzengrund, and Senftenberg. Food microbiology. 2022; 108(January): 1–9. pmid:36088119
- 50. Ashton PM, Nair S, Peters TM, Bale JA, Powell DG, Painset A, et al. Identification of Salmonella for public health surveillance using whole genome sequencing. PeerJ. 2016; 4(e1752–18): 1–18. pmid:27069781
- 51. Hopkins KL, Lawson AJ, Connell S, Peters TM, de Pinna E. A novel real-time polymerase chain reaction for identification of Salmonella enterica subspecies enterica. Diagnostic microbiology and infectious disease. 2011; 70(2): 278–80. pmid:21411262
- 52. Chattaway MA, Dallman TJ, Larkin L, Nair S, McCormick J, Mikhail A, et al. The transformation of reference microbiology methods and surveillance for Salmonella with the use of whole genome sequencing in England and Wales. Frontiers in public health. 2019; 7: 1–12. pmid:31824904
- 53. Chattaway MA, Painset A, Godbole G, Gharbia S, Jenkins C. Evaluation of genomic typing methods in the Salmonella reference laboratory in Public Health, England, 2012–2020. Pathogens. 2023; 12(2): 1–20. pmid:36839496
- 54. Bolger AM, Lohse M, Usadel B. Trimmomatic: A flexible trimmer for Illumina sequence data. Bioinformatics. 2014; 30(15): 2114–20. pmid:24695404
- 55. Chattaway MA, Chandra N, Painset A, Shah V, Lamb P, Acheampong E, et al. Genomic approaches used to investigate an atypical outbreak of Salmonella Adjame. Microbial genomics. 2019; 5(1). pmid:30648934
- 56. Zhou Z, Alikhan N-F, Sergeant MJ, Luhmann N, Vaz C, Francisco AP, et al. GrapeTree: Visualization of core genomic relationships among 100,000 bacterial pathogens. Genome research. 2018; 28(9): 1395–404. pmid:30049790
- 57. Waldram A, Dolan G, Ashton PM, Jenkins C, Dallman TJ. Epidemiological analysis of Salmonella clusters identified by whole genome sequencing, England and Wales 2014. Food microbiology. 2018; 71: 39–45. pmid:29366467
- 58. Mook P, Gardiner D, Verlander NQ, McCormick J, Usdin M, Crook P, et al. Operational burden of implementing Salmonella Enteritidis and Typhimurium cluster detection using whole genome sequencing surveillance data in England: A retrospective assessment. Epidemiology and infection. 2018; 146(11): 1452–60. pmid:29961436
- 59. Pearce ME, Alikhan N-F, Dallman TJ, Zhou Z, Grant K, Maiden MCJ. Comparative analysis of core genome MLST and SNP typing within a European Salmonella serovar Enteritidis outbreak. International journal of food microbiology. 2018; 274: 1–11. pmid:29574242
- 60. Wheeler TJ. Large-Scale Neighbor-Joining with NINJA. Algorithms in bioinformatics WABI 2009 lecture notes in computer science. 2009; 5724. https://doi.org/doi.org/10.1007/978-3-642-04241-6_31
- 61. Letunic I, Bork P. Interactive Tree of Life (iTOL) v4: Recent updates and new developments. Nucleic acids research. 2019; 47: 256–9. pmid:30931475
- 62. Kolmogorov M, Yuan J, Lin Y, Pevzner PA. Assembly of long, error-prone reads using repeat graphs. Nature biotechnology. 2019; 37: 540–6. pmid:30936562
- 63. Vaser R, Sović I, Nagarajan N, Šikić M. Fast and accurate de novo genome assembly from long uncorrected reads. Genome research. 2017; 27: 737–46. pmid:28100585
- 64. Walker BJ, Abeel T, Shea T, Priest M, Abouelliel A, Sakthikumar S, et al. Pilon: An integrated tool for comprehensive microbial variant detection and genome assembly improvement. PLoS one. 2014; 9(11): e112963. pmid:25409509
- 65. Li H. Minimap2: Pairwise alignment for nucleotide sequences. Bioinformatics. 2018; 34(18): 3094–100. pmid:29750242
- 66. Bharat A, Petkau A, Avery BP, Chen J, Folster J, Carson CA, et al. Correlation between phenotypic and in silico detection of antimicrobial resistance in Salmonella enterica in Canada using Staramr. Microorganisms. 2022; 10(2): 1–10. pmid:35208747
- 67. Zankari E, Hasman H, Cosentino S, Vestergaard M, Rasmussen S, Lund O, et al. Identification of acquired antimicrobial resistance genes. Journal of antimicrobial chemotherapy. 2012; 67(11): 2640–4. pmid:22782487
- 68. Carattoli A, Zankari E, Garciá-Fernández A, Larsen MV, Lund O, Villa L, et al. In silico detection and typing of plasmids using PlasmidFinder and plasmid multilocus sequence typing. Antimicrobial agents and chemotherapy. 2014; 58(7): 3895–903. pmid:24777092
- 69. Zankari E, Allesøe R, Joensen KG, Cavaco LM, Lund O, Aarestrup FM. PointFinder: A novel web tool for WGS-based detection of antimicrobial resistance associated with chromosomal point mutations in bacterial pathogens. Journal of antimicrobial chemotherapy. 2017; 72(10): 2764–8. pmid:29091202
- 70. Chen L, Zheng D, Liu B, Yang J, Jin Q. VFDB 2016: Hierarchical and refined dataset for big data analysis—10 years on. Nucleic acids research. 2016; 44(D1): D694–7. pmid:26578559
- 71. Page AJ, Cummins CA, Hunt M, Wong VK, Reuter S, Holden MTG, et al. Roary: Rapid large-scale prokaryote pan genome analysis. Bioinformatics. 2015; 31(22): 3691–3. pmid:26198102
- 72. Schwengers O, Jelonek L, Dieckmann MA, Beyvers S, Blom J, Goesmann A. Bakta: Rapid and standardized annotation of bacterial genomes via alignment-free sequence identification. Microbial genomics. 2021; 7(11): 1–13. pmid:34739369
- 73. Bernhofer M, Dallago C, Karl T, Satagopam V, Heinzinger M, Littmann M, et al. PredictProtein—Predicting protein structure and function for 29 years. Nucleic Acids Research. 2021; 49(1): W535–40. pmid:33999203
- 74. Carver TJ, Rutherford KM, Berriman M, Rajandream MA, Barrell BG, Parkhill J. ACT: The Artemis comparison tool. Bioinformatics. 2005; 21(16): 3422–3. pmid:15976072
- 75. Trampari E, Holden ER, Wickham GJ, Ravi A, Martins L de O, Savva GM, et al. Exposure of Salmonella biofilms to antibiotic concentrations rapidly selects resistance with collateral tradeoffs. npj biofilms and microbiomes. 2021; 3(1): 1–13. pmid:33431848
- 76. Prouty AM, Gunn JS. Comparative Analysis of Salmonella enterica Serovar Typhimurium Biofilm Formation on Gallstones and on Glass. Infection and immunity. 2003; 71(12): 7154–8. pmid:14638807
- 77. García B, Latasa C, Solano C, García-del Portillo F, Gamazo C, Lasa I. Role of the GGDEF protein family in Salmonella cellulose biosynthesis and biofilm formation. Molecular microbiology. 2004; 54(1): 264–77. pmid:15458421
- 78. Trampari E, Holden ER, Wickham GJ, Ravi A, Prischi F, Martins L de O, et al. Antibiotics select for novel pathways of resistance in biofilms. bioRxiv. 2019;
- 79. Vervier K, Browne HP, Lawley TD. CarboLogR: A Shiny/R application for statistical analysis of bacterial utilisation of carbon sources. bioRxiv. 2019;