Postmortem Nasopharyngeal Microbiome Analysis of Zambian Infants With and Without Respiratory Syncytial Virus Disease: A Nested Case Control Study

Abstract

Background: Respiratory syncytial virus (RSV) is the most common cause of bronchiolitis and lower respiratory tract infections in children in their first year of life, disproportionately affecting infants in developing countries. Previous studies have found that the nasopharyngeal (NP) microbiome of infants with RSV infection has specific characteristics that correlate with disease severity, including lower biodiversity, perturbations of the microbiota and differences in relative abundance. These studies have focused on infants seen in clinical or hospital settings, predominantly in developed countries.

Methods: We conducted a nested case control study within a random sample of 50 deceased RSV+ infants with age at death ranging from 4 days to 6 months and 50 matched deceased RSV- infants who were all previously enrolled in the Zambia Pertussis and RSV Infant Mortality Estimation (ZPRIME) study. All infants died within the community or within 48 hours of facility admittance. As part of the ZPRIME study procedures, all decedents underwent one-time, postmortem NP sampling. The current analysis explored the differences between the NP microbiome profiles of RSV+ and RSV- decedents using the 16S ribosomal DNA sequencing.

Results: We found that Moraxella was more abundant in the NP microbiome of RSV+ decedents than in the RSV- decedents. Additionally, Gemella and Staphylococcus were less abundant in RSV+ decedents than in the RSV- decedents.

Conclusions: These results support previously reported findings of the association between the NP microbiome and RSV and suggest that changes in the abundance of these microbes are likely specific to RSV and may correlate with mortality associated with the disease.

FIGURE 1.

Microbial composition of RSV− and RSV+ samples showing microbe diversity within each sample. On average, the 5 most abundant microbes in RSV+ samples were Streptococcus (22.2%), Haemophilus (19.7%), Moraxella (11.0%), Klebsiella (10.2%) and Corynebacterium (6.8%). The 5 most abundant microbes in RSV− samples were Streptococcus (29.0%), Staphylococcus (13.5%), Escherichia (8.5%), Klebsiella (6.4%) and Haemophilus (5.7%).

FIGURE 2.

Alpha diversity for RSV+ and RSV− Zambian decedents. No significant differences were found when conducting Wilcoxon rank-sum test for (A) genera (P = 0.621 for observed, 0.287 for Shannon and 0.206 for Simpson) or (B) species (P = 0.790, 0.173, 0.120, respectively).

FIGURE 3.

A: NMDS displaying the Bray-Curtis dissimilarity index for genera comparison between RSV+ and RSV− deceased infants. The best solution stress value was 0.248. Some clustering of RSV+ decedents is visible to the bottom with some clustering of RSV− decedents to the top left. B: PCoA showing general clustering of RSV+ samples in the bottom right-hand corner without RSV− samples. However, otherwise there is substantial overlap between the 2 groups.

FIGURE 4.

Box plots of relative abundance indicating significant (A) intragenera or (B) intraspecies differences in RSV+ and RSV− samples. A: RSV+ samples show lower abundances in the genera of Gemella (P = 0.020) and Staphylococcus (P = 0.018) and a higher abundance of Moraxella (P = 0.006). B: RSV+ samples show higher abundances of the species Moraxella sp. (P = 0.016) and lower abundances of G. haemolysans (P = 0.006). Differences in Corynebacterium accolens (P = 0.049) and Prevotella scopos (P = 0.004) appear to be driven by a few outliers. C: Heatmap displaying the Bray-Curtis dissimilarity index for genera comparison between RSV+ and RSV− decedents. Higher abundances of Haemophilus and Moraxella are seen in RSV+ decedents. Lower abundances of Gemella, Staphylococcus and Streptococcus are present in RSV+ decedents.