Integrated analysis of serum metabolomics and fecal microbiome in infants with necrotizing enterocolitis

Abstract

Background: Necrotizing enterocolitis (NEC), a lethal gastrointestinal disorder in preterm infants, remains poorly understood in its pathology, and early diagnosis are critically limited. Multi-omics approaches present unprecedented opportunities to elucidate NEC pathogenesis and identify clinically translatable biomarkers.

Methods: Infants with Bell stage II-III NEC and gestational age-matched controls were enrolled. Serum/stool samples from NEC patients at acute (NEC-D) and recovery (NEC-R) phases, and controls (non-NEC) were collected. Fecal metagenomic sequencing and serum untargeted metabolomic profiling were performed. Clinical parameters were compared.

Results: The study comprised seven NEC and seven non-NEC infants. Baseline neonatal characteristics and maternal perinatal parameters showed no significant differences between NEC-D and non-NEC except for markedly lower leukocyte counts in NEC infants. Fecal metagenomics revealed severely diminished alpha diversity in NEC-D versus both non-NEC controls and NEC-R, characterized with lower Chao1 index. NEC-D exhibited elevated Escherichia coli relative abundance alongside reduced Staphylococcus haemolyticus, Staphylococcus aureus, Staphylococcus epidermidis, and Lactobacillus paracasei. Correspondingly, KEGG functional gene analysis demonstrated impaired metabolism in NEC-D. Serum metabolomics identified significantly decreased ornithine, DL-arginine, L-threonine, leucine, and D-proline in NEC-D versus non-NEC. NEC-D also showed lower taurodeoxycholic acid, glycocholic acid, and chenodeoxycholic acid compared to NEC-R. Integrative analysis revealed a positive correlation between the metabolites D-proline and ornithine and the Lactobacillus paracasei, Staphylococcus epidermidis, and Staphylococcus aureus abundance.

Conclusion: NEC is characterized by gut microbiota dysbiosis with reduced diversity, altered functional gene expression, and disrupted host-microbiota metabolic crosstalk. The identified serum metabolite-microbiome correlations provide mechanistic insights into NEC pathogenesis and potential diagnostic biomarkers.

Keywords: metabolomics; microbiome; multi-omics; necrotizing enterocolitis; preterm infant.

Figure 1

Different gut microbiota composition, diversity, functional genes between groups. (A,B) The alpha diversity as measured by the Chao1 index (the genus level). (C) Comparison of beta diversity between NEC-D, NEC-R, and non-NEC groups based on PCA analysis on genus level classification. (D) Comparison of beta diversity between NEC-D, NEC-R, and non-NEC groups based on PCA analysis on species level classification. (E) LEfSe analysis of specific bacteria in NEC-D and non-NEC groups. (F) Differential analysis of lefse KEGG function between NEC-D group and non-NEC group.

Figure 2

Orthogonal Partial Least Squares Discriminant Analysis (OPLSDA) was employed to perform a comprehensive statistical evaluation of the metabolite profiles within each group. (A–C) The OPLS-DA analysis was conducted in anionic mode between the two groups. (B–D) The OPLS-DA analysis was conducted in cationic model between the two groups. (E) OPLSDA permutation test comparison between NEC-D and non-NEC in anionic mode. (F) OPLSDA permutation test plot between NEC-D and non-NEC in cationic model. (G) OPLSDA permutation test plot between NEC-D and NEC-R in Anionic Mode. (H) OPLSDA permutation test plot between NEC-D and NEC-R in cationic model.

Figure 3

VIP diagram of differential metabolites. (A) Compare the differential metabolites in NEG between non-NEC and NEC-D. (B) Compare the differential metabolites in POS between non-NEC and NEC-D groups. (C) Compare the differential metabolites in NEG between NEC-R and NEC-D groups. (D) Compare the differential metabolites in NEG between NEC-R and NEC-D groups.

Figure 4

A heat map illustrating the correlation between microbial species and metabolites in the NEC-D group and the NON-NEC group. (A) The microbial species-metabolite correlation heat map in the context of NEG for both groups. (B) The microbial species-metabolite correlation heat map concerning POS among the groups. The metabolic names associated with the serial numbers of the metabolites depicted in the figure are fully detailed in Supplementary Table S1.