Upper respiratory microbial communities of healthy populations are shaped by niche and age

Abstract

Background: Alterations in upper respiratory microbiomes have been implicated in shaping host health trajectories, including by limiting mucosal pathogen colonization. However, limited comparative studies of respiratory microbiome development and functioning across age groups have been performed. Herein, we perform shotgun metagenomic sequencing paired with pathogen inhibition assays to elucidate differences in nasal and oral microbiome composition and intermicrobial interactions across healthy 24-month-old infant (n = 229) and adult (n = 100) populations.

Results: We find that beta diversity of nasal and oral microbiomes varies with age, with nasal microbiomes showing greater population-level variation compared to oral microbiomes. Infant microbiome alpha diversity was significantly lower across nasal samples and higher in oral samples, relative to adults. Accordingly, we demonstrate significant differences in genus- and species-level composition of microbiomes between sites and age groups. Antimicrobial resistome patterns likewise varied across body sites, with oral microbiomes showing higher resistance gene abundance compared to nasal microbiomes. Biosynthetic gene clusters encoding specialized metabolite production were found in higher abundance across infant oral microbiomes, relative to adults. Investigation of pathogen inhibition revealed greater inhibition of gram-negative and gram-positive bacteria by oral commensals, while nasal isolates had higher antifungal activity.

Conclusions: In summary, we identify significant differences in the microbial communities inhabiting nasal and oral cavities of healthy infants relative to adults. These findings inform our understanding of the interactions impacting respiratory microbiome composition and functions related to colonization resistance, with important implications for host health across the lifespan. Video Abstract.

Fig. 1

Nasal and oral microbial communities of infants compared to adults. Nonmetric multidimensional scaling (NMDS) ordination plot of Bray–Curtis dissimilarity indices of A infant nasal and oral microbiomes, B adult nasal and oral microbiomes, C adult and infant nasal microbiomes, and D adult and infant oral microbiomes. Shotgun metagenomic sequencing reads were examined from microbiomes normalized to an equal sequencing depth. Points represent individual samples, with circles denoting infant microbiomes and triangles adult microbiomes. Nasal samples are colored blue, and oral samples are colored red. Plot stress is listed for each comparison. E Boxplots displaying richness, Shannon, inverse Simpson, and Fisher diversity metrics of nasal (blue) and oral (red) microbiomes from infants compared to adults. Plot is faceted by body site. Boxes show the 25th and 75th percentiles with the median represented by a horizontal line and whiskers showing 1.5 × the interquartile range. Overlayed data points are jittered to avoid overplotting. Significance values *p < 0.05 and ****p < 0.001 Mann–Whitney U-test

Fig. 2

Nasal taxa across infants and adults. Boxplots comparing the relative abundance of the 10 most abundant A nasal genera and B top 20 nasal species from these genera between infants and adults. Stars above species names in B highlight known respiratory pathogens. Y-axes display log + 1 of abundance values. Boxes show the 25th and 75th percentiles with the median represented by a horizontal line and whiskers showing 1.5 × the interquartile range. Overlayed data points (gray) are jittered to avoid overplotting. Significance values *p < 0.05, ** < 0.01, *** < 0.001, and ****p < 0.0001 Mann–Whitney U-test with FDR correction

Fig. 3

Oral taxa across infants and adults. Boxplots comparing the relative abundance of the 10 most abundant A oral genera and B top 20 oral species from these genera between infants and adults. Stars above species names in B highlight known respiratory pathogens. Y-axes display log + 1 of abundance values. Boxes show the 25th and 75th percentiles with the median represented by a horizontal line and whiskers showing 1.5 × the interquartile range. Overlayed data points (gray) are jittered to avoid overplotting. Significance values *p < 0.05, ** < 0.01, *** < 0.001, and ****p < 0.0001 Mann–Whitney U-test with FDR correction

Fig. 4

Antimicrobial resistance gene content. Volcano plots comparing antimicrobial resistance genes identified from A oral and B nasal metagenomes that are differentially abundant between infants (light orange points) and adults (dark orange points), or not significantly different (gray points). Y-axis represents inverse of log of FDR significance values, and x-axis represents the log fold change between groups. FDR cutoff of < 0.05 was used. Select genes are labeled, with bold text indicating a gene originated from a respiratory pathogen (H. influenzae, M. catarrhalis, or S. pneumoniae). Metagenomes (n = 89 infant nasal, n = 212 infant oral, n = 48 adult nasal, n = 97 adult oral) were subsampled to 500,000 paired reads and aligned to the CARD database. Counts of antimicrobial resistance genes with MAPQ greater or equal to 10 were compared using edgeR. C Boxplots of fragments per kilobase per million mapped reads (FPKM) aligning to antimicrobial resistance genes with MAPQ greater or equal to 10 from H. influenzae, M. catarrhalis, or S. pneumoniae between infants and adults. Boxes show the 25th and 75th percentiles with the median represented by a horizontal line and whiskers showing 1.5 × the interquartile range. Significance values *p < 0.05, ** < 0.01, *** < 0.001, and ****p < 0.0001. Mann–Whitney U-test

Fig. 5

Biosynthetic gene cluster abundance differs with age. A The plot indicates reads from nasal (blue) and oral (red) infant and adult microbiomes that mapped to BGCs (aryl polyene, NRPS, PKS, RiPP, siderophore, terpene, and hybrid clusters) identified from upper respiratory bacteria (eHOMD habitat listed as “nasal,” “nasal, oral,” or “oral”). B Reads from nasal (blue) and oral (red) infant and adult microbiomes mapping to BGCs from upper respiratory bacteria, grouped by cluster type. All metagenomes were subsampled to 160 K reads. Fifty-one infant nasal, 22 adult nasal, and 1 adult oral metagenomes were not included in analysis due to low sequencing depth. Normalized reads were pseudoaligned onto an in-house BGC library. NRPS, non-ribosomal peptide synthases; PKS, polyketide synthases; RiPP, ribosomally synthesized and post-translationally modified peptides. Significance value *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 Yuen Welch’s T-test with 0.001 trim

Fig. 6

Colonization resistance varies across upper respiratory tract sites. Heatmaps of pathogen inhibition by A nasal and B oral isolates. Pathogens listed on the y-axis are grouped by gram positive (top), gram negative (middle), and fungal (bottom). Inhibition scores (“none,” “partial,” “complete,” or “not determined”) are colored according to legends inhibition. C Fractional inhibition scores were calculated for nasal and oral isolates across pathogen types. Fractional inhibition scores for isolates with genus-level taxonomic assignments are displayed for D nasal and E oral isolates. Error bars represent standard error of the mean