Genomic characteristics of Salmonella enterica serovar Blockley

ABSTRACT

Non-typhoidal Salmonella (NTS) is a significant cause of foodborne illness worldwide, with increasing antimicrobial resistance posing a public health concern. Salmonella enterica serovar Blockley (S. Blockley) is relatively uncommon, and its antimicrobial resistance profile and population structure have been understudied. This study presents a comprehensive genomic analysis of 264 S. Blockley isolates from diverse geographical regions to elucidate antimicrobial resistance patterns and population structure. Bayesian analysis classified these genomes into 10 distinct groups (BAPS A to BAPS J), further categorized into two lineages, R and S. Lineage R comprised six BAPS clusters (BAPSs A–F), predominantly found in Asia and Africa, all of which harbored the azithromycin resistance gene mph(A) and other resistance determinants. In contrast, lineage S, lacking mph(A), comprised the remaining four BAPS clusters, which were primarily found in Europe and the Americas. Several types of mutations in gyrA were found in lineage R, which were specific to BAPS clusters. These BAPS clusters exhibited distinct geographic distributions, with BAPS B, BAPS D, and BAPS E unique to China, Taiwan, and Japan, respectively, while BAPS H and BAPS I were predominantly found in the United States. Temporal phylogenetic analysis suggested that lineage R diverged in the 1980s, with notable microevolutionary changes. The presence of a genomic island with mph(A), aph(3’)-Ia, aph(3”)-Ib, aph(6)-Id, and tet(A) in lineage R underscores the public health threat, highlighting a need for continuous surveillance.

IMPORTANCE

Antimicrobial resistance in Salmonella is a global public health concern. In this study, we focused on serovar Blockley, and a whole-genome analysis revealed its global population structure. The results revealed the existence of azithromycin-resistant strains, which were characterized both phylogenetically and geographically. The resistance genes were transmitted via genomic islands, and their micro-scale evolution was also revealed. Our findings are the first to reveal the dissemination of antimicrobial resistance genes, including azithromycin, in serovar Blockley, and provide valuable insights into understanding the spread of antimicrobial resistance.

INTRODUCTION

Salmonella is a Gram-negative rod that causes major foodborne infections worldwide. Nontyphoidal Salmonella (NTS) is a major contributor to the global burden of diarrheal disease in humans, with an estimated 197.35 million episodes and 84,799 deaths annually (1).

In the European Union, salmonellosis is the second most common zoonotic disease in humans, with 65,208 cases reported in 2022 (2). Serovar Enteritidis was the most prevalent, accounting for 67.3% of cases, followed by serovar Typhimurium (13.1%) and monophasic Typhimurium variant I, 1,4,[5],12:i:- (4.3%). NTS typically causes gastroenteritis and is self-limiting, but antimicrobial therapy is necessary for extraintestinal infections. Antimicrobial resistance in Salmonella is a significant public health concern in one-health settings. The multidrug-resistant (MDR) S. Typhimurium definitive phage type DT104, which caused a pandemic in the 1990s, is a notable example. This clone carries the Salmonella genomic island-1 (SGI-1) in the chromosome, which confers resistance to ampicillin, chloramphenicol, streptomycin, sulfonamide, and tetracycline (ACSSuT) (3, 4). More recently, the emergence of azithromycin resistance has become a new public health issue (5–7).

S. Blockley is a relatively uncommon serovar, with less than 0.05 cases per 100,000 population in the United States in recent years, compared with approximately 3 cases of S. Enteritidis (8). In Taiwan, S. Blockley accounted for 0.3% (106/40,595) of Salmonella isolates collected from 2004 to 2022 (9). In Japan, S. Blockley accounted for 1.1% (10/934) of Salmonella isolates reported in 2015–2016 (https://www.niid.go.jp/niid/images/iasr/archive/2016/bac/salm1516.pdf). Despite its low incidence, S. Blockley has garnered attention due to its association with antimicrobial resistance. Azithromycin resistance has been sporadically reported in S. Blockley, with Salmonella azithromycin resistance genomic island (SARGI) believed to play a significant role in the transmission of azithromycin resistance (10–12). However, comprehensive genomic analyses of S. Blockley isolates on a global scale are limited. Therefore, this study aims to perform a genomic analysis of S. Blockley isolates from diverse geographic regions to elucidate their phylogenetic relationships and genomic characteristics including antimicrobial resistance with a specific emphasis on azithromycin resistance.

MATERIALS AND METHODS

Genomic sequence data

The study incorporated genomic sequence data from 264 S. Blockley isolates, with 34 sequences generated for this research, 229 obtained from the public domain via Enterobase (https://enterobase.warwick.ac.uk/species/index/senterica), and 1 complete S. Blockley genome (accession number CP043662) sourced from the NCBI database (https://www.ncbi.nlm.nih.gov/). These isolates were collected across 33 countries in four regions, encompassing Africa, the Americas, Asia, and Europe. The isolates were collected from 1955 to 2023 (Table S1). Among the isolates, 60% (159/264) originated from human host, while 31% (82/264) were from non-human sources including poultry (17%) and livestock (7%). The source for 9% (23/264) of the isolates remained unidentified.

Whole-genome sequencing and analysis

Whole-genome sequencing was conducted on 34 isolates using the Illumina sequencing platform (San Diego, CA, US). DNA extraction was performed using the DNeasy Blood and Tissue Kit (Qiagen, Hilden, Germany) following the manufacturer’s protocol. Sequencing runs were executed on the Illumina MiSeq and/or the iSeq platforms. The obtained read sequences were assembled using the SPAdes v3.15.5 with the “—careful” option (13). To confirm the serovar Blockley, the assembled sequences (contigs) were subjected to in-silico serotyping using SeqSero2 v1.2.1 (14). Sequence type and eBurst Group were identified according to the scheme of Achtman (15). Antimicrobial resistance genes and plasmid replicon types were identified using ResFinder v4.1.11 (16) and PlasmidFinder (17). The rop gene, encoding the Rop family of plasmid primer RNA-binding protein found in the ColE1 plasmid, was identified based on the sequence under accession no. CP100731.1. Subsequently, assembled sequences (contigs) were annotated using DFAST v1.2.18 to identify genes such as mphR(A) and mrx(A) (18). The distribution of resistance and replicon genes were visualized using ggplot2 v3.4.2 (19), ComplexHeatmap v2.10.0 (20), and genoPlotR v0.8.11 (21).

Phylogenetic analysis

A phylogenetic analysis based on single-nucleotide polymorphisms (SNPs) was conducted using genome sequencing data. Initially, snippy v4.6.0 (https://github.com/tseemann/snippy) was employed to generate core SNP profiles, using S. Blockley strain 159838 (accession number CP043662) as the reference. Repetitive regions within the reference genome were identified and excluded using MUMmer v4.0.0rc1. Following extraction with Gubbins v3.3.0 and snp-sites v2.5.1 (https://github.com/sanger-pathogens/snp-sites) (22, 23), a total of 2,492 non-recombinant SNP sites were utilized to construct a phylogenetic tree. The maximum likelihood phylogeny was inferred using RAxML-NG v1.2.0 with 1,000 bootstrap replicates (24). The resulting tree was then visualized using ggtree v3.2.1 in R (25). Values of cumulative bases in recombination were obtained in the analysis of Gubbins.

Bayesian Analysis of Population Structure

Fastbaps v1.0.8 was employed to identify Bayesian Analysis of Population Structure (BAPS) clusters in the S. Blockley phylogeny with “optimise.baps” prior (26). Multiple-level analysis was conducted to identify level 2 clusters for BAPS F.

Phylogenetic dating

To estimate the divergence time of the lineage R isolates, a Bayesian dating analysis was performed using Bactdating v1.1.1 (27) with core genome single-nucleotide polymorphism (cgSNP) profiles of isolates.

RESULTS

Genomic characteristics and BAPS clusters

Based on the publicly available MLST scheme (15), the 264 isolates were assigned to the eBurst Group (eBG) 151 and three sequence types (STs) including ST52 (n = 259), ST4928 (n = 4), and ST3427 (n = 1). Subsequent analysis classified the isolates into 10 distinct BAPS clusters (Fig. 1). Notably, ST4928 isolates were exclusively classified under BAPS A, while ST3427 belonged to BAPS F (F.6 among the level 2 clusters). BAPSs A–F exhibited closer genetic proximity, falling within lineage R. In contrast, BAPSs G–J clustered within another lineage, lineage S (Fig. 1). All lineage R isolates were characterized by the presence of the azithromycin resistance gene mph(A). In contrast, none of the lineage S isolates harbored mph(A). Lineage R displayed a low value of cumulative bases in recombination, while lineage S exhibited a high recombination frequency. The genetic clusters displayed unique geographical distribution patterns, with BAPS B, BAPS D, and BAPS E predominantly found in China, Taiwan, and Japan, respectively, while BAPS H and BAPS I were exclusively detected in the United States (Table 1). BAPS A was identified in isolates from Asia and Europe, while BAPS G was confined to European isolates. BAPSs C and J were prevalent in the Americas and Europe. BAPS F was the most widespread; it was found in isolates from Africa, the Americas, Asia, and Europe. Three BAPS F isolates were positioned outside the main cluster of BAPS F in the phylogenetic tree (Fig. 1). The discrepancy may be due to differences in the analytical methods, which can yield varying clustering results even when using the same SNP data. This phenomenon has been observed with some frequency in BAPS clustering (26). A multiple-level analysis using fastbaps was then applied to BAPS F, resulting in six subclusters, F.1–F.6. The three outlier BAPS F isolates were assigned to BAPSs F.1 and F.6, with each subcluster forming a distinct branch in the phylogenetic tree. BAPSs F.2, F.4, and F.5 were specific to Europe, while BAPS F.3 was most prevalent in Africa (Table S2).

Fig 1.

Phylogenetic tree of 264 S. Blockley isolates based on 2,492 core SNP sites. Tips are colored by BAPS cluster according to the inset legend. Circles indicate isolation year, geographic region, country, mph(A), and the values of Gubbins’ cumulative bases in recombination from inner to outer. Lineages R and S are highlighted in the background. Subclusters within BAPS F are indicated by black bars.

TABLE 1.

Geographic distribution of BAPS clusters^b

Region/country	BAPS										Total
	A	B	C	D	E	F	G	H	I	J
Africa						9					9
Egypt						4					4
Ethiopia						5					5
Americas			5			1		38	35	27	106
Canada			1							1	2
Dominican Republic			1								1
United States			3			1		38	35	26	103
Asia	4	4		29	11	2					50
China		4									4
Israel	1										1
Japan					11						11
Jordan	3										3
Taiwan				29		2					31
Europe	19		11			15	16			37	98
Austria						3					3
Germany						1					1
Ireland						7	1			1	9
Netherlands						1				1	2
Northern Ireland							1			1	2
United Kingdom	19		11			3	14			34	81
NAa						1					1
Total	23	4	16	29	11	28	16	38	35	64	264

^a NA, not available.

^b Subtotals by regions are shown in boldface type.

Antimicrobial resistance genetic determinants

S. Blockley isolates harbored some acquired antimicrobial resistance genes (ARGs) and mutations on gyrA and parC, conferring resistance to aminoglycosides, macrolides, tetracyclines, quinolones, phenicols, β-lactams, trimethoprim, sulfonamides, and colistin (Table 2). Notably, all isolates carried an aac(6′)-Iaa, a silent gene that may not contribute to resistance in the Salmonella isolates. Resistance determinants were predominantly observed in isolates of lineage R, while BAPS H and BAPS I isolates within lineage S lacked known ARGs but the silent gene aac(6′)-Iaa. Lineage R isolates commonly harbored aph(3″)-Ib, aph(6)-Id, aph(3′)-Ia, mph(A), and tet(A). Mutations in gyrA were identified exclusively in lineage R isolates, and the mutations exhibited variations within each BAPS cluster. BAPS A exhibited S83F and D87G mutations, BAPS B exhibited S83Y mutation, BAPSs C and D exhibited S83F mutation, and BAPS F exhibited S83F, D87Y (specific to BAPS F.2), or D87G (specific to BAPS F.4 and BAPS F.5) mutation, while BAPS E and lineage S (BAPSs G–J) had no mutation in gyrA. Certain ARGs were geographically specific, such as floR in isolates from Taiwan and bla_CTX-M-15 exclusively from Japan.

TABLE 2.

Distribution of resistance genes across BAPS clusters

Class of antimicrobials	Resistance gene	BAPS A (N = 23)	BAPS B (N = 4)	BAPS C (N = 16)	BAPS D (N = 29)	BAPS E (N = 11)	BAPS F (N = 28)	BAPS G (N = 16)	BAPS H (N = 38)	BAPS I (N = 35)	BAPS J (N = 64)	Total (N = 264)
Aminoglycosides	aac(6')-Iaa	23	4	16	29	11	28	16	38	35	64	264
	aph(3'')-Ib	19	4	16	29	11	28	2			3	112
	aph(6)-Id	19	4	16	29	11	28	2			3	112
	aph(3')-Ia	19	4	16	29	11	27					106
	aadA1							2				2
	aadA2b							2				2
Macrolides	mph(A)	23	4	16	29	11	28					111
Tetracyclines	tet(A)	23	4	16	29	9	26	4				111
Quinolones	gyrA	23	4	16	29		27					99
	S83F			16	29		12					57
	S83F+D87G	23										23
	D87Y						12					12
	S83Y		4									4
	D87G						3					3
	parC_S80R	19										19
	qnrB19			3							8	11
Phenicols	catA2		4		17	11	5					37
	floR				18							18
	cmlA1							2				2
Beta-lactams	blaCTX-M-15					10						10
	blaTEM-1B		3				3	2			1	9
	blaCMY-2			1								1
	blaSHV-12						1					1
Trimethoprim	dfrA14							2			3	5
Sulfonamides	sul2							2			3	5
	sul3							2				2
Polymyxin (colistin)	mcr-1.2						1					1

Distribution of plasmids

Seven types of replicons were identified. IncN was the most notable as it was the most abundant and found in only lineage R isolates, specifically in BAPSs A, D, E, and F (Table S3). ColE1/rop was the second most frequently identified in this study with 94% (48/51) in lineage R isolates. Of these, 75% (36/48) were detected in isolates from Taiwan (n = 25) and Japan (n = 11). As shown in Fig. S1, the qnrB19 gene was exclusively linked to the Col(pHAD28) replicon, and the floR gene was strongly associated with the IncX4 replicon, found primarily in Taiwanese isolates. The genes aph(3′)-Ia, mph(A), aph(3″)-Ib, aph(6)-Id, and tet(A) were linked to IS6100, which is a component of the SARGI.

SARGI core unit

We included mrx(A), mphR(A), and IS6100 in the analysis, as these elements, along with mph(A), comprise the SARGI core unit (IS26-mph(A)-mrx(A)-mphR(A)-IS6100) (10). The heatmap shown in Fig. S1 demonstrated a strong link between IS6100 and mrx(A), mph(A), and mphR(A), suggesting the presence of the SARGI. Genomic analysis indicated that the SARGI was inserted into the rbsK gene, which encodes a protein with two functional domains, a DeoR family transcriptional regulator and a PkfB family carbohydrate kinase (ribokinase) (6, 10, 12). As shown in Fig. S2a, the inserted segments in rbsK comprised SARGI and an additional IS26 composite transposon carrying aph(3′)-Ia, aph(6)-Id, and aph(3″)-Ib. The junction point of IS6100 in rbsK was conserved, while the junction points adjacent to IS26 were varied (Fig. S2b). Most IS26 junction points were identified at position 953 (bp) of rbsK, distributed in isolates from all BAPS clusters involved. The remaining junction points were predominantly identified in BAPS F isolates; the junction at 806 was specific to BAPS F.3, while those at 929 and 939 were specific to BAPS F.4 (Table S4).

Evolution of BAPS clusters

The emergence of the most recent ancestor of lineage R was estimated using BactDating (Fig. 2). Initial inference using a root-to-tip analysis of BactDating revealed a discrepancy in the statistical evaluation between lineages S and R. Specifically, the R² values were 0.09 and 0.62 for lineages S and R, respectively (Fig. S3), indicating that lineage R isolates could fit the analysis, whereas lineage S did not. Consequently, we applied BactDating to lineage R and a subset of lineage S, finding that the combination of lineage R and BAPS I isolates produced a moderate R² value of 0.53. Therefore, we subjected the combination of lineage R (BAPS A–F) and BAPS I to the analysis, using 10⁶ MCMC iterations. As a result, the average point mutation rate (μ) was estimated at 2.71 (95% CI [CI], 2.38–3.09) substitutions per year, and the root date for the most likely ancestor root was 1979.48 [1973.14–1984.57]. All branches of the BAPS clusters of lineage R were inferred to have diverged between the 1980s and mid-1990s. Within lineage R, five types of gyrA mutations were identified (Table 2), estimated to have diverged between the 1990s and mid-2000s.

Fig 2.

The dated phylogenetic tree of 111 lineage R isolates. The tree was inferred using BactDating. The red bars indicate the 95% CI of dating.

DISCUSSION

S. Blockley, although a minor serovar, presents significant public health concerns due to its antimicrobial resistance, particularly to azithromycin (10, 11, 28, 29). This study provides a comprehensive genomic analysis of 264 S. Blockley isolates from diverse geographical regions, showing the population structure and antimicrobial resistance patterns with a specific emphasis on azithromycin resistance.

Our findings reveal that S. Blockley isolates in this study belong to eBG151, comprising 10 distinct BAPS clusters. These clusters can be divided into two major lineages, lineage R, which includes six BAPSs (A–F), and lineage S, which includes four BAPSs (G–J). Lineage R is characterized by the carriage of the azithromycin resistance gene mph(A), a key determinant of azithromycin resistance (30), whereas lineage S lacks this gene. This genetic distinction suggests that lineages R and S have different evolutionary paths and antimicrobial resistance profiles.

The geographical distribution of the genetic lineages is noteworthy. Lineage R isolates were predominantly found in Asia and Africa, with specific BAPS clusters showing country-specific distributions, for instance, BAPS B in China, BAPS D in Taiwan, and BAPS E in Japan. Lineage S, on the other hand, was primarily found in isolates from Europe and the Americas, with BAPS H and BAPS I being exclusively found in those from the United States. This geographical skew highlights the regional dissemination patterns of the two lineages.

The SARGI core unit (IS26-mph(A)-mrx(A)-mphR(A)-IS6100), a key element conferring azithromycin resistance, was consistently identified in lineage R (Fig. S2). We searched the core unit in the NCBI database and found that it has been identified in numerous Enterobacteriaceae species, including Citrobacter spp., Cronobacter sp., Edwardsiella spp., Enterobacter spp., Escherichia spp., Klebsiella spp., Kluyvera cryocrescens, Leclercia adecarboxylata, Morganella morganii, Proteus mirabilis, Providencia rettgeri, Raoultella spp., Serratia marcescens, and Shigella spp., as well as in species from other families, such as Aeromonas spp., Bordetella trematum, Shewanella spp., and Vibrio spp. (data not shown). However, the genomic islands containing the core SARGI unit, tet(A), aph(3′)-Ia, aph(6)-Id, and aph(3″)-Ib, as shown in Fig. S2a, are found only in S. Haardt and S. Blockley.

Genomic analysis indicated that the SARGI core unit, along with IS26–flanking segments carrying tet(A), aph(3′)-Ia, aph(6)-Id, and aph(3″)-Ib, was inserted in the rbsK gene on the chromosome, forming a more extensive SARGI-containing genomic island (Fig. S2a). This region, comprising IS26, tet, aph genes, and IS26, is currently unique to S. Blockley. Chromosomal integration is an effective strategy for the dissemination of ARGs. The SARGI-containing genomic island likely contributes to the stable maintenance of resistance traits in lineage R.

Intriguingly, the junction point of IS6100 in rbsK is conserved, but multiple junction points are observed for IS26. At the junction of IS6100 in the rbsK gene, a “GG” sequence, the conserved end of the IS6 family (31), and a target-like inverted repeat sequence “GGCTCTGTTGCAAA” (32) are found. On the other hand, the multiple junction points of IS26 suggest that IS26 is active in mediating intramolecular transposition that leads to DNA deletion (33).

In addition, multiple IS26 are found in SARGI-containing genomic islands, which can mediate combinatorial variation of ARGs such as tet(A) and aph(3′)-Ia–aph(6)-Id–aph(3″)-Ib as well as mobile gene elements like IncN rep within the genomic island. Such variant islands have been identified in S. Typhimurium, S. Agona, and S. Concord (3, 34–36).

The dating analysis indicates that lineage R may have diverged around the 1980s, with most BAPS clusters within this lineage diverging before the mid-1990s. Chinese isolates (BAPS B) could have diverged earliest from lineage S. This temporal divergence correlates with the emergence of multiple gyrA mutation types, which occur in the chromosome and confer resistance to quinolones. These mutations were acquired in a BAPS cluster-specific manner, further illustrating the microevolutionary dynamics within lineage R.

Conclusion

This study provides a detailed genomic analysis of S. Blockley, highlighting significant antimicrobial resistance patterns and population structure. S. Blockley falls into two main lineages, R and S, with lineage R predominantly found in Asia and Africa and characterized by the carriage of mph(A), whereas lineage S, primarily found in Europe and the Americas, lacked this resistance gene. In addition to mph(A), the lineage R isolates harbor genomic islands containing multiple resistance genes, including aph(3′)-Ia, aph(3″)-Ib, aph(6)-Id, and tet(A), posing a serious public health threat. Phylogenetic analysis suggests that the lineage R diverged in the 1980s, with notable microevolutionary changes since then. Although uncommon, S. Blockley shows significant antimicrobial resistance, particularly to azithromycin, necessitating continuous surveillance and evaluation of its impact on public health.

DATA AVAILABILITY

The short-read sequence data from this study have been submitted to NCBI/ENA/DDBJ under BioProject PRJDB18357 (BioSample SAMD00797488–SAMD00797521, accession numbers DRR576566–DRR576599). The accession numbers for all isolates used in this study are listed in Table S1.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/spectrum.02048-24.

Supplemental figures. spectrum.02048-24-s0001.pdf.

Fig. S1 to S3.

Supplemental tables. spectrum.02048-24-s0002.xlsx.

Tables S1 to S4.

OPEN PEER REVIEW. reviewer-comments.pdf.

An accounting of the reviewer comments and feedback.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.