Hospitalised children with COVID-19 display an aberrant intestinal microbiota and a shift in faecal compounds related with the metabolism of vitamins and lipids

Abstract

The SARS-CoV-2 virus and its rapid spread have made it a global health concern. The aim of this was to investigate the microbial and metabolic faecal profiles of paediatric patients hospitalised for COVID-19 to try to identify biomarkers of predisposition to severity. The study included 16 patients (aged 4-14 years old) from six different Spanish hospitals and 20 age-matched healthy controls. The gut microbiota was characterised by sequencing of 16S rDNA amplicons and internal transcribed space amplicons, while the metabolic profile was analysed by liquid chromatography high resolution mass spectrometry. A different microbial profile was observed between patients and controls, with a significantly higher abundance of sequences belonging to the phyla Bacteroidota and Pseudomonadota in patients. A different metabolic profile was observed between the two groups. Non-infected children had higher faecal levels of vitamins such as niacin, thiamine, and vitamin D3 derivatives, which were negatively correlated with the abundance of pathogenic bacteria, such as members of Enterobacteriaceae. Hospitalisation due to SARS-CoV-2 infection in children was associated with changes in the gut microbiota and an altered metabolomic profile. For the first time, several relevant biological compounds were found to be reduced in the faeces of children hospitalised with COVID-19 compared to healthy controls.

Fig 1. Principal Coordinate Analysis (PCoA).

The plot compares the bacterial communities between samples from controls (red circles) and COVID-19 patients (blue circles). The percentages shown on the axes represent the proportion of dissimilarities.

Fig 2. Microbial relative abundances.

Differences at the phylum level in faecal samples from COVID patients and controls (* p-value ≤ 0.05; ** p-value ≤ 0.01).

Fig 3. Profiles of bifidobacterial sequences.

Data are represented as relative abundance in patients and controls (* p-value ≤ 0.05; ** p-value ≤ 0.01). Data show percentages of Bifidobacterium species greater than 1% in at least one group.

Fig 4. Volcano plot.

Differential molecular features comparing control and patient groups in positive (panel A) and negative (panel B) ion modes. Blue and red squares indicate down-regulated and up-regulated compounds, respectively, in controls compared to patients.

Fig 5. Classification of metabolites.

Differential metabolites found in the faeces of controls (A) and patients (B) expressed as percentages.

Fig 6. Correlation analysis.

Spearman correlation analysis between metabolites and microbiota at phylum and family level. Metabolites codes are shown in Tables S2 and S3 in S1 File.