Characterization of Shigella flexneri serotype 6 strains from geographically diverse low- and middle-income countries

Abstract

Shigella flexneri serotype 6 (Sf6) is one of the most common serotypes recovered from surveillance studies of moderate to severe diarrhea. Despite the clinical significance of Sf6, this serotype is understudied. In this work, we have performed both serotype-specific genomic and phenotypic comparisons of Sf6 isolates to one another and non-S. flexneri serotypes. Comparative genomic analyses identified significant nucleotide homology between Sf6 strains (n = 325), despite a broad range of collection timeframes and geographic locations. We identified Sf6 specific factors, including a potential novel Shigella virulence factor (type II secretion system). Additionally, we identified established Shigella virulence genes (ospG) and metabolic genes (rutABCDEFGR) that were absent in Sf6 strains while present in the majority of 728 non-Sf6 strains. Complete sequencing of 11 clinical Sf6 strains, demonstrated that the Sf6 virulence plasmid (pINV) is ~38 kb smaller than the average non-Sf6 pINV (~228 kb). Comparisons of S. flexneri species level antibiotic susceptibility highlighted that clinical Sf6 isolates from Africa in the Global Enteric Multicenter Study (GEMS) and Vaccine Impact on Diarrhea in Africa (VIDA) study demonstrated geographic, serotype-specific susceptibility pattern. Phenotypic analyses of Sf6 identified reduced intracellular invasion and cytokine induction from HT-29 cells, as well as reduced Ipa protein effector secretion, compared with S. flexneri serotype 2a strain 2457T. Together these data highlight conserved and unique serotype-specific genotypic and phenotypic features for Sf6. This level of conservation has not been noted for other S. flexneri serotypes and is promising for vaccine and diagnostic assays to provide global Sf6-specific coverage.IMPORTANCEShigellosis is an ongoing global public health crisis with >270 million annual episodes among all age groups; however, the greatest disease burden is among children in low- and middle-income countries (LMIC). The lack of a licensed Shigella vaccine and the observed rise in antimicrobial-resistant Shigella spp. highlights the urgency for effective preventative and interventional strategies. The inclusion of S. flexneri serotype 6 (Sf6) is a necessary component of a multivalent vaccine strategies based on its clinical and epidemiological importance. Given the genomic diversity of Sf6 compared with other S. flexneri serotypes and Sf6 unique O-antigen core structure, serotype-specific characterization of Sf6 is a critical step to inform Shigella-directed vaccine and alternative therapeutic designs. Herein, we identified conserved genomic content among a large collection of temporally and geographically diverse Sf6 clinical isolates and characterized genotypic and phenotypic properties that separate Sf6 from non-Sf6 S. flexneri serotypes.

Keywords: Shigella; Shigella flexneri; Shigella flexneri serotype 6; comparative genomics; host-pathogen interactions.

Fig 1

Genomic relationships of 325 Sf6 geographically diverse genomes. (A) Diverse E. coli and Shigella genomes (26) were compared with 325 Sf6 genomes (teal) (Data set 1). A total of 158,917 SNPs were determined relative to the E. coli IA139 (NC_011750.1) reference. The phylogenomic groups are designated PG1 to PG7 (see legend), as described previously (5), and the serotypes, where known, are indicated in parentheses next to each strain name. S. flexneri lineage delineated by color; Lineage 1 (red) containing PG 1–7 and Lineage 2 (teal) containing Sf6. (B) All Sf6 strains (n = 325) were compared with Sf6 archetype strain CCH060 (CP099865, CP099866, CP09867). A total of 6,904 SNPs were determined relative to Sf6 reference. Colors were used to indicate country of origin for each isolate (see legend). The Sf6 clades are designated Sf6_I-Sf6_VII as indicated by major branch points of genomic similarity (by SNP and BAPS comparisons) as indicated by the outmost, black ring are based on the number of SNPs identified in the phylogenomic analysis (as described in the Material and Methods). The Sf6 genomes completed in this study are indicated in bold. The genomes included are listed in Supplemental Data set 1.

Fig 2

Virulence plasmid comparison of Sf6 strains. Nucleotide sequences encompassing virA- ccdA,B were aligned from Sf5 strain M90T plasmid pWR100 (AL391753; nucleotides 144615–173950) and the plasmids from Sf6 strains CCH060 (pCCH060_195 (CP099866); nucleotides 175010–160158), 702865 (p702865_118; nucleotides 154258–170617), and 401192 (p401192_196; nucleotides 0–10288, 196515–183601) using Mauve v.2025.2.5 (38). Genes labeled for pWR100: red = genes missing from Sf6 genomes, black = genes shared with Sf6 genomes, yellow = insertion elements.

Fig 3

Invasion of geographically diverse Sf6 strains in HT-29 cells. GEMS Sf6 clinical isolates from Africa (A) and Asia (B) and control archetype strain CCH060 were used to infect HT-29 monolayers (MOI of 1:100) for 90 min. Cells were lysed to enumerate intracellular bacteria at 2 and 6 hpi. Results are displayed as percent recovery of initial inoculum and shown for a representative assay of three replicates. The asterisks above indicate statistically significant differences determined using a two-way ANOVA and Tukey posttest; ****, P < 0.0001.

Fig 4

Multiplex cytokine comparison of Sf6 and Sf2a strain 2457T at 6 hpi HT-29 infection. Supernatants from infected HT-29 cells were collected at 2 and 6 hpi and assessed by multiplex-cytokine array. Cytokines tested included interleukins IL-8 (A), IL-1β (B), and tumor necrosis factor-alpha (TNF-α; C). Data are normalized by subtraction of uninfected control wells. Data are presented as pooled data from two independent experiments with technical triplicates in each experiment. The asterisks above indicate statistically significant differences between Sf6 isolates as compared to Sf2a 2457T or Sf6 CCH060 and were determined using a two-way ANOVA and Tukey posttest; ****, P < 0.0001.

Fig 5

Sf6 strains have reduced whole cell and secreted levels of Ipa protein expression compared with Sf2a 2457T. Whole cell lysates (A and B) and supernatant fractions (C and D) were collected from mid-log phase cultures grown in TSB (-) and TSB with deoxycholate (Bile Salt +) growth conditions and were probed for IpaB, IpaC, and IpaD with a mixture of specific antibodies by Western blot. GroEL was used as a loading control for whole cell lysates (Figure 3.S3 and 3.S4). Congo red (CR) induction was used as an alternative T3SS secretion stimulant for whole cell lysates (E) and supernatants (F). Densitometry (Fiji ImageJ) was utilized to compare protein levels (47).