Clonal serotype 1c multidrug-resistant Shigella flexneri detected in multiple institutions by sentinel-site sequencing

Abstract

Shigella flexneri is a major diarrhoeal pathogen, and the emergence of multidrug-resistant S. flexneri is of public health concern. We report the detection of a clonal cluster of multidrug-resistant serotype 1c (7a) S. flexneri in Singapore in April 2022. Long-read whole-genome sequence analysis found five S. flexneri isolates to be clonal and harboring the extended-spectrum β-lactamases bla _CTX-M-15 and bla _TEM-1. The isolates were phenotypically resistant to ceftriaxone and had intermediate susceptibility to ciprofloxacin. The S. flexneri clonal cluster was first detected in a tertiary hospital diagnostic laboratory (sentinel-site), to which the S. flexneri isolates were sent from other hospitals for routine serogrouping. Long-read whole-genome sequence analysis was performed in the sentinel-site near real-time in view of the unusually high number of S. flexneri isolates received within a short time frame. This study demonstrates that near real-time sentinel-site sequence-based surveillance of convenience samples can detect possible clonal outbreak clusters and may provide alerts useful for public health mitigations at the earliest possible opportunity.

Keywords: Shigella flexneri; multidrug-resistant (MDR); outbreak; surveillance; whole-genome sequence (WGS).

Figure 1

Summary of Shigella flexneri detection timeline. There were eight known laboratory-proven S. flexneri cases (Cases A–H) in four hospitals within a 5-week period. Five isolates were sequenced in this study (Cases A–E, dark blue circle), among which 3 (^*) were sequenced near real-time. Two additional isolates were sequenced subsequently (^∧). Isolates of three cases from Hospitals III and V (Cases F, G, H) had been discarded and were not available for sequence analysis (gray circle). BioRender was used to create this figure.

Figure 2

Summary of laboratory-based surveillance models. (A) For diseases and isolates not routinely notifiable and transferred to central reference laboratories, delays in outbreak detection and laboratory investigations may occur due to logistical challenges. (B) Sentinel surveillance of convenience samples, implemented in the current study, enabled the streamlined detection of clonal isolates of public health importance and provided early alerts for the public health agency. BioRender was used to create this figure.

Figure 3

Shigella flexneri isolates under investigation are genetically closely related and form a distinct phylogenetic cluster. (A) Core genome single nucleotide polymorphisms (SNP) phylogenetic tree generated by Parsnp illustrates phylogenetic relatedness of 16 complete reference genomes (yellow circle) and five query genomes (blue circles). The scale represents substitution rate per site, and the number at each branch indicates the bootstrap support value. (B) in silico serotyping with ShigaTyper suggests all query isolates are of serotype 1c (7a). The horizontal axis lists ShigaTyper determinants that contributed to the eventual in silico serotype (vertical axis). The term “ipaH” in this figure represents a conserved 780 bp region of ipaH encoding for the C-terminal catalytic domain, which was originally designated as “ipaH_c” in the ShigaTyper database. (C) Parsnp phylogenetic tree constructed with five Singapore query genomes and 200 most relevant genomes out of 9,255 publicly available Shigella flexneri genomes. The scale represents substitution rate per site, and the size of the gray circle at each branch indicates the bootstrap support value. Each genome sequence is represented by a solid circle (∙), and the color of the circle represents the country from which the genome sequence was reported. Genomes that are phylogenetically more closely related to the Singaporean isolates (black circles) are labeled with basic metadata in the following format (sample source | date of sample collection in YYYY-MM). For optimal visualization, genomes that were phylogenetically more distant from the Singaporean isolates were represented by colored solid circles without associated metadata. Oac, S. flexneri-specific O-acetyltransferase gene marker; Oac1b, S. flexneri-specific O-acetyltransferase gene marker specific for serotype 1b and 7b; Sf_wzx, S. flexneri-specific O-antigen flippase gene marker; Sf_wzy, S. flexneri-specific O-antigen polymerase gene marker; Xv, S. flexneri-specific gene marker encoding for protein homolog of LTA synthase family protein; gtrI, glucosyltransferase mediating addition of first glucosyl group to the O-antigen backbone in S. flexneri serotypes 1 and 7; gtrIC, glucosyltransferase mediating addition of second glucosyl group to the O-antigen backbone in S. flexneri serotypes 1c (7a); grtII, glucosyltransferase of S. flexneri serotype 2; gtrIV, glucosyltransferase of Shigella flexneri serotype 4; gtrV, glucosyltransferase of S. flexneri serotype 5; gtrX, glucosyltransferase of serotypes 2 or X; ipaB, invasive plasmid antigen B; IpaH, a 780 bp region of the ipaH genes encoding the highly conserved C-terminal catalytic domain.

Figure 4

Summary of antimicrobial resistance and virulence markers. (A) Cases A–E had identical phenotypic antimicrobial susceptibility profiles. All isolates were resistant to ceftriaxone and had intermediate resistance to ciprofloxacin. The antibiograms for Case F–H were retrieved from the laboratory information systems of Hospital III (Case F) and Hospital V (Case G and H), as the isolates had been discarded and were not available for sequencing. (B) Cases A–E had identical genotypic antimicrobial susceptibility profiles and contained blaCTX-M-15 and blaTEM-1, consistent with the extended-spectrum beta-lactamases (ESBL) phenotype observed. (C) Global distribution of blaTEM-1 and/or blaCTX-M-15 containing Shigella flexneri are based on publicly available data (NCBI Pathogen Detection as of 24 April 2022). 0.85% of all available S. flexneri genomes (n = 11134) contain both blaCTX-M-15 and blaTEM-1. (D) Cases A–E had identical genetic markers for a panel of virulence factors, of which virF, sepA, astA, and ipaH9.8 were encoded on plasmids. Sixteen complete S. flexneri reference genomes were included for comparison.