Genomic diversity and antimicrobial resistance profiles of Shigella species in under-five children with acute diarrhea in Addis Ababa, Ethiopia

Abstract

Background: Shigella species exhibit considerable genetic diversity, reflecting distinct genomic traits and epidemiological trends. In Ethiopia, infections caused by Shigella species, along with rising antimicrobial resistance (AMR), pose a major public health threat, especially among children under five. To our knowledge, there are few studies on the genetic diversity and antimicrobial susceptibility of Shigella species in this vulnerable age group. Figuring this out could help treat cases better and control outbreaks more effectively. Objective: This study aimed to characterize Shigella species' genomic diversity and AMR profiles in under five children with acute diarrhea in Addis Ababa, Ethiopia. Methods: Between June 2021 and April 2022, we conducted a cross-sectional study. Freshly passed stool specimens were collected and transported in Cary Blair media to the Ethiopian Public Health Institute (EPHI) laboratory for Shigella isolation and identification using standard bacteriological methods. Serogrouping was performed using polyvalent antisera, and antimicrobial susceptibility testing (AST) was conducted using the disk diffusion method. Additionally, a systematic review and meta-analysis were conducted to assess the prevalence and AMR patterns of Shigella species in East Africa. Molecular characterization was performed using whole-genome sequencing (WGS) analysis. Sequencing was conducted using the Illumina NextSeq550 (Illumina, Singapore) with a 300-cycle kit, generating paired-end reads of 149bp. Raw reads were quality-filtered and trimmed to a minimum length of 50 bp before being taxonomically classified using MiniKraken version 1. The whole genome data were aligned with Antibiotic Resistance Gene (ARG) sequences from the Comprehensive Antibiotic Resistance Database (CARD) using the Resistance Gene Identifier (RGI). Plasmid analysis was performed using the Mykrobe PlasmidFinder tool. Additionally, AMR and virulence genes were screened using the Centre for Genomic Epidemiology (CGE) web-based platform. Data were analyzed using descriptive statistical tools. The association of independent and dependent variables was evaluated with logistic regression. A P-value < 0.05 was considered statistically significant. Results: Among the 534 stool-cultured specimens in the prospective study, 47 (8.8%) were positive for Shigella species. Of these, 31 were serologically identified as S. sonnei and 16 as S. flexneri. Of 37 isolates analyzed by WGS, all 28 S. sonnei strains identified xiii serologically were confirmed as S. sonnei. However, all nine isolates initially identified as S. flexneri were found to be E. coli O37:H10. The phylogenetic tree showed that both S. sonnei and E. coli O37:H10 isolates had multiple evolutionary origins, suggesting that their phenotypic features evolved convergently. Plasmids Col156 and Col (BS512) were identified in all S. sonnei isolates, while IncFII and Col (MG828) plasmids were found only in one isolate. In contrast to S. sonnei, the most common plasmid type in E. coli O37:H10 isolates was IncFII, and virulence genes such as gad were frequently detected. Discrepancies were observed between phenotypic and genotypic AMR results. The systematic review estimated the pooled prevalence of Shigella species in East Africa was 6.2%. Despite variations in study sites and periods, the meta-analysis study further revealed an increased rate of resistance of Shigella species to tetracycline, ampicillin, amoxicillin, chloramphenicol, and co-trimoxazole. In the prospective study, AST showed that 100, 93.6, 80.9, 72.3, and 57.5% were sensitive to norfloxacin, nalidixic acid, ciprofloxacin, gentamicin, and cefoxitin, respectively. However, 100% of the isolates were resistant to amoxicillin. All isolates were resistant to three or more antimicrobials that exhibited multidrug resistance (MDR). None of the risk factors assessed showed a statistically significant association with Shigella infection. All S. sonnei isolates in our study contained genes encoding blaEC-8 and blaZEG-1. About 60.7% of the isolates were phenotypically sensitive to cefoxitin among the blaEC-8 genes detected in the genotyping analysis, whereas all isolates were completely resistant to amoxicillin phenotypically. The study also identified genes that conferred resistance to trimethoprim (dfrA1). Extended-spectrum beta lactamase (ESBL) blaEC-15 for cephalosporins, blaMIX-2, and blaMIX-6 for penicillins were detected in E. coli O37:H10. All E. coli O37:H10 isolates possessed a gene associated with trimethoprim resistance, and eight E. coli O37:H10 isolates exhibited consistent results for trimethoprim when comparing phenotype and genotype. The dominant AMR mechanism among the identified ARGs was antibiotic efflux, followed by antibiotic target alteration. Conclusion: There was no significant heterogeneity among East African studies, the majority of which were conducted in Ethiopia. Both the systematic review and the prospective phenotypic study revealed alarmingly increased levels of AMR to commonly administered antibiotics. The genotyping study revealed that the most prevalent resistant genes were associated with beta-lactam and trimethoprim drugs. The IncFII plasmid, which primarily encodes ESBL, was more frequently identified in E. coli O37:H10 isolates than xiv in S. sonnei. The study also highlighted a significant discrepancy between phenotypic and genotypic drug resistance, as well as variations in serotypes and phylogenetic relationships with global isolates. Recommendation: The study highlighted significant discrepancies between phenotypic and genotypic results. Molecular studies integrating WGS for AMR determination and strain identification into active surveillance could enhance monitoring of AMR spread and detection of potential emerging variations. Therefore, public health and clinical laboratories in Ethiopia should implement WGS to address inconsistencies in conventional analyses, enhance treatment efficacy, and inform targeted interventions.