Temporal bacterial and metabolic development of the preterm gut reveals specific signatures in health and disease

Abstract

Background: The preterm microbiome is crucial to gut health and may contribute to necrotising enterocolitis (NEC), which represents the most significant pathology affecting preterm infants. From a cohort of 318 infants, <32 weeks gestation, we selected 7 infants who developed NEC (defined rigorously) and 28 matched controls. We performed detailed temporal bacterial (n = 641) and metabolomic (n = 75) profiling of the gut microbiome throughout the disease.

Results: A core community of Klebsiella, Escherichia, Staphyloccocus, and Enterococcus was present in all samples. Gut microbiota profiles grouped into six distinct clusters, termed preterm gut community types (PGCTs). Each PGCT reflected dominance by the core operational taxonomic units (OTUs), except of PGCT 6, which had high diversity and was dominant in bifidobacteria. While PGCTs 1-5 were present in infants prior to NEC diagnosis, PGCT 6 was comprised exclusively of healthy samples. NEC infants had significantly more PGCT transitions prior to diagnosis. Metabolomic profiling identified significant pathways associated with NEC onset, with metabolites involved in linoleate metabolism significantly associated with NEC diagnosis. Notably, metabolites associated with NEC were the lowest in PGCT 6.

Conclusions: This is the first study to integrate sequence and metabolomic stool analysis in preterm neonates, demonstrating that NEC does not have a uniform microbial signature. However, a diverse gut microbiome with a high abundance of bifidobacteria may protect preterm infants from disease. These results may inform biomarker development and improve understanding of gut-mediated mechanisms of NEC.

Keywords: Gut microbiome; Metabolomics; Necrotising enterocolitis; Preterm infant.

Fig. 1

Analysis of each of the six preterm gut community types (PGCT). a Heatmap showing the six PGCTs and the status. Status based on disease or control; Control = pink, DControl = green, PreNEC = blue, and PreLOS = purple. Top 25 most abundant OTUs shown. b Boxplot analysis of the observed OTUs and Shannon diversity showing each PGCT. P values based on the Kruskal-Wallis test show a significant increase in number of OTUs and Shannon diversity in PGST 2. c Boxplot analysis of the eight most significantly distinct OTUs in each PGST. ***Represents a Kruskal-Wallis P value ≤ 0.001

Fig. 2

Dynamics of the microbiome through each preterm gut community type (PGCT) in patients diagnosed with NEC compared to matched controls over the initial weeks of life. a Transition network analysis showing PGCTs in the PreNEC samples. Approximated as a Markov chain with subject-independent transition probabilities. Arrow weights reflect the transition probabilities from the existing PGCT to the subsequent PGCT in next sample. Size of circle reflects the relative number of samples associated with that PGCT. Increasing fractions represent PGCT that have relatively larger number of predisease diagnosis samples. b Visualisation of the PGCTs in each individual patient overtime. Red dotted lines represent day of NEC diagnosis. Only samples up to day 50 of life are included. Patient 180 died during the study

Fig. 3

LCMS metabolomic features associated with NEC diagnosis. a OPLS-DA of NEC and matched control samples at time point 3 (diagnosis). b–f Box plot analysis showing the temporal development of the five most significant features associated with NEC diagnosis. Lettering represents Tukey’s pairwise comparison results, where groups that do not share a letter are significantly different. NEC diagnosis is significantly different from control samples. All time points included

Fig. 4

Comparison of significant LCMS metabolites associated with NEC and the PGCT determined by 16S bacterial profiling of the respective sample. Samples from all time points included in the analysis. a, b C21-steroid hormone biosyntehesis and linoleate metabolism pathways. c, d Linoleate metabolism pathways. e Leukotriene metabolism and prostagladlin formation from arachidonate pathway. PGCT 6 represents the community type associated with exclusively healthy samples