Exploring nasopharyngeal microbiota profile in children affected by SARS-CoV-2 infection

Abstract

The relationship between COVID-19 and nasopharyngeal (NP) microbiota has been investigated mainly in the adult population. We explored the NP profile of children affected by COVID-19, compared to healthy controls (CTRLs). NP swabs of children with COVID-19, collected between March and September 2020, were investigated at the admission (T₀), 72 h to 7 days (T₁), and at the discharge (T₂) of the patients. NP microbiota was analyzed by 16S rRNA targeted-metagenomics. Data from sequencing were investigated by QIIME 2.0 and PICRUSt 2. Multiple machine learning (ML) models were exploited to classify patients compared to CTRLs. The NP microbiota of COVID-19 patients (N = 71) was characterized by reduction of α-diversity compared to CTRLs (N = 59). The NP microbiota of COVID-19 cohort appeared significantly enriched in Streptococcus, Haemophilus, Staphylococcus, Veillonella, Enterococcus, Neisseria, Moraxella, Enterobacteriaceae, Gemella, Bacillus, and reduced in Faecalibacterium, Akkermansia, Blautia, Bifidobacterium, Ruminococcus, and Bacteroides, compared to CTRLs (FDR < 0.001). Exploiting ML models, Enterococcus, Pseudomonas, Streptococcus, Capnocytopagha, Tepidiphilus, Porphyromonas, Staphylococcus, and Veillonella resulted as NP microbiota biomarkers, in COVID-19 patients. No statistically significant differences were found comparing the NP microbiota profile of COVID-19 patients during the time-points or grouping patients on the basis of high, medium, and low viral load (VL). This evidence provides specific pathobiont signatures of the NP microbiota in pediatric COVID-19 patients, and the reduction of anaerobic protective commensals. Our data suggest that the NP microbiota may have a specific disease-related signature since infection onset without changes during disease progression, regardless of the SARS-CoV-2 VL.

Importance: Since the beginning of pandemic, we know that children are less susceptible to severe COVID-19 disease. A potential role of the nasopharyngeal (NP) microbiota has been hypothesized but to date, most of the studies have been focused on adults. We studied the NP microbiota modifications in children affected by SARS-CoV-2 infection showing a specific NP microbiome profile, mainly composed by pathobionts and almost missing protective anaerobic commensals. Moreover, in our study, specific microbial signatures appear since the first days of infection independently from SARS-CoV-2 viral load.

Keywords: COVID-19; SARS-CoV-2; children; nasopharyngeal microbiota; respiratory tract.

Fig 1

NP microbiota ecology. Alpha-diversity representation of COVID-19 and CTRLs groups based on Goods coverage (A), Simpson (B), Shannon (C), Chao-1 (D), observed features (E), Faith PD (F) indices. P values were obtained by Kruskal–Wallis test.

Fig 2

Beta-diversity representation of COVID-19 and CTRLs groups, based on Bray–Curtis, Euclidian distance, and unweighted and weighted UniFrac algorithms. P values were obtained by PERMANOVA (P value < 0.05 for all analyses).

Fig 3

LEfSe cladogram and plot representation of taxa differences obtained for the NP microbiota of COVID-19 and CTRLs groups. Bar and nodes highlighted in red and green were significantly more abundant more abundant in COVID-19 and CTRLs, respectively (P value FDR < 0.001).

Fig 4

Graphical representation of hierarchical analysis of bacterial distribution at genus level of COVID-19 and CTRLs groups. In the heatmap, the hierarchical complete linkage dendrogram is based on the ASVs Pearson’s correlation coefficient. The color scale characterizes the Z-score for each variable: red, high level; blue, low level. Red circle, cluster mainly composed by COVID-19; green circle, cluster mainly composed by CTRL; blue star, cluster composed by bacterial taxa negatively correlated with the disease; yellow star, cluster composed by bacterial taxa negatively correlated with SARS-CoV-2 infection.

Fig 5

Important bacterial taxa selected by model classification analysis. The bars represent the importance scores of each ASV in the prediction of models.

Fig 6

Functional analysis of the NP microbiota of COVID-19 patients and CTRLs. PICRUSt was used to infer the functional content of the NP microbiota based on 16S rRNA metataxonomic data. Log2 fold change representation of the abundances of functional pathways shows significant difference between COVID-19 and CTRLs groups. Red and green colors indicate the increase and the decrease of abundances of pathways in COVID-19 affected patients NP.