Dynamics of the Gut Microbiome in Shigella-Infected Children during the First Two Years of Life

Abstract

Shigella continues to be a major contributor to diarrheal illness and dysentery in children younger than 5 years of age in low- and middle-income countries. Strategies for the prevention of shigellosis have focused on enhancing adaptive immunity. The interaction between Shigella and intrinsic host factors, such as the microbiome, remains unknown. We hypothesized that Shigella infection would impact the developing microbial community in infancy and, conversely, that changes in the gastrointestinal microbiome may predispose infections. To test this hypothesis, we characterized the gastrointestinal microbiota in a longitudinal birth cohort from Malawi that was monitored for Shigella infection using 16S rRNA amplicon sequencing. Children with at least one Shigella quantitative polymerase chain reaction (qPCR) positive sample during the first 2 years of life (cases) were compared to uninfected controls that were matched for sex and age. Overall, the microbial species diversity, as measured by the Shannon diversity index, increased over time, regardless of case status. At early time points, the microbial community was dominated by Bifidobacterium longum and Escherichia/Shigella. A greater abundance of Prevotella 9 and Bifidobacterium kashiwanohense was observed at 2 years of age. While no single species was associated with susceptibility to Shigella infection, significant increases in Lachnospiraceae NK4A136 and Fusicatenibacter saccharivorans were observed following Shigella infection. Both taxa are in the family Lachnospiraceae, which are known short-chain fatty acid producers that may improve gut health. Our findings identified temporal changes in the gastrointestinal microbiota associated with Shigella infection in Malawian children and highlight the need to further elucidate the microbial communities associated with disease susceptibility and resolution. IMPORTANCE Shigella causes more than 180 million cases of diarrhea globally, mostly in children living in poor regions. Infection can lead to severe health impairments that reduce quality of life. There is increasing evidence that disruptions in the gut microbiome early in life can influence susceptibility to illnesses. A delayed or impaired reconstitution of the microbiota following infection can further impact overall health. Aiming to improve our understanding of the interaction between Shigella and the developing infant microbiome, we investigated changes in the gut microbiome of Shigella-infected and uninfected children over the course of their first 2 years of life. We identified species that may be involved in recovery from Shigella infection and in driving the microbiota back to homeostasis. These findings support future studies into the elucidation of the interaction between the microbiota and enteric pathogens in young children and into the identification of potential targets for prevention or treatment.

Keywords: Shigella; gut microbiome; infant microbiome.

FIG 1

Cohort sample summary. (A) Number of samples collected (right y axis) and infant age at sample collection (left y axis) in each month of the year (x axis). The orange symbols represent samples that were Shigella quantitative polymerase chain reaction (qPCR) positive. (B) Sample matching strategy: samples from matched cases and controls from 0 to 24 months of age. Each row represents samples collected at each time point from each matched pair of cases (right panel) and controls (left panel). The triangles represent samples collected with diarrhea, and the circles represent samples with no diarrhea. The orange symbols represent samples that were Shigella qPCR positive.

FIG 2

Alpha diversity comparisons in infant microbiomes. (A) Shannon diversity indices at different ages at the time of sample collection, from 2 months to 24 months of age. The symbols represent individual values. The R² and P values are from a simple linear regression model. (B) Shannon diversity indices of samples from case (orange) versus control (purple) individuals at the time of sample collection. The symbols represent individual values. The R² values from simple linear regression models are shown separately for the cases and the controls within the graph.

FIG 3

Taxonomic compositions of infant gastrointestinal microbiomes by age. (A) Mean relative abundance of the 10 most abundant taxa by infant age. n represents the number of samples at each time point (month). (B) Bar graphs showing the mean relative abundances (%) of the top 10 most abundant taxa at the infant ages of 6, 12, 18, and 24 months. The symbols represent individual values. The mean relative abundances at different ages were compared via a one-way analysis of variance (ANOVA). *, P < 0.05.

FIG 4

Taxa differentially abundant in cases after Shigella infection. (A) Chart representing samples that were included in the analysis. All samples after the collection of a Shigella qPCR positive sample in the case individuals (right column) and their matching controls (left column). Samples in orange represent those that were found to be Shigella qPCR positive. (B) Taxa identified to be significantly (adjusted P < 0.05) abundant in cases versus controls after Shigella infection. The x axis indicates the estimated difference (log₂-fold change, logFc) between the abundance of taxa in the cases compared to the controls. These estimates were obtained from logistic regression models. A positive logFc indicates a greater abundance in the cases than in the controls. The adjusted P values were determined using a mixed effects linear regression model controlled for match, sample age, infant ID, diarrhea, and short-term antibiotic use.

FIG 5

Taxa differentially abundant in cases after Shigella-driven diarrhea versus diarrhea from other causes. (A) Chart representing samples that were included in the analysis. Case samples are those that were collected from case individuals after a Shigella qPCR positive and diarrhea positive sample (right column, orange triangles). Control samples are those from matched control individuals after a diarrhea-positive sample was collected within a month of the case sample (left column, purple triangles). (B) Taxa identified to be significantly (adjusted P < 0.05) abundant in cases after Shigella positive diarrheal infection versus the controls after a diarrhea positive sample was collected. The x axis indicates the estimated difference (log₂-fold change, logFc) between the abundance of taxa in the cases compared to the controls. These estimates were obtained from logistic regression models. A positive logFc indicates a greater abundance in the cases than in the controls. The adjusted P values were determined using a mixed effects logistic regression model controlled for infant sex, birth month, sample age, and infant ID.