Human microbiome transfer in the built environment differs based on occupants, objects, and buildings

Abstract

Compared to microbiomes on other skin sites, the bacterial microbiome of the human hand has been found to have greater variability across time. To increase understanding regarding the longitudinal transfer of the hand microbiome to objects in the built environment, and vice versa, 22 participants provided skin microbiome samples from their dominant hands, as well as from frequently and infrequently touched objects in their office environments. Additional longitudinal samples from home environments were obtained from a subset of 11 participants. We observed stability of the microbiomes of both the hand and built environments within the office and home settings; however, differences in the microbial communities were detected across the two built environments. Occupants' frequency of touching an object correlated to that object having a higher relative abundance of human microbes, yet the percent of shared microbes was variable by participants. Finally, objects that were horizontal surfaces in the built environment had higher microbial diversity as compared to objects and the occupants' hands. This study adds to the existing knowledge of microbiomes of the built environment, enables more detailed studies of indoor microbial transfer, and contributes to future models and building interventions to reduce negative outcomes and improve health and well-being.

Figure 1

Alpha diversity and top taxa: (a) violin plot showing the Shannon diversity metric for samples in the office environment; (b) taxonomic summary of observed relative abundances of the top 10 most abundant taxa within each sample type at the genus level for the office environment; (c) violin plot showing the Shannon diversity metric for samples in the home environment; and (d) taxonomic summary of observed relative abundances of the top 10 most abundant taxa within each sample type at the genus level for the home environment.

Figure 2

Longitudinal stability of microbial communities (weighted UniFrac) within each participant colored by sample type in the office and home environments: (a) boxplot representing median weighted UniFrac distances within each participant colored by sample type between the weeks of sampling for the office environment. Sample sizes can be found in Supplementary Table 1; (b) boxplot representing median weighted UniFrac distances within each participant colored by sample type between pairs of the weeks of sampling for the home environment. Sample sizes can be found in Supplementary Table 2.

Figure 3

Microbial sharing in the built environment: (a) boxplot of microbial sharing with the participant hand by sample type in office. Sample sizes for each sample type were mouse (n = 65); keyboard (n = 64); desk (n = 63); (b) circle plot representing the microbial sharing by participant stratified by sample type in office. Lager circles and darker purple represent higher proportions shared communities. Sample sizes for each participant in office can be found in Supplementary Table 3; (c) boxplot representing the microbial sharing with the participant hand by sample type in home. Sample sizes for each sample type were nightstand (n = 31); bathroom (n = 33); bedroom floor (n = 31); living room floor (n = 31); (d) circle plot representing microbial sharing by participant stratified by sample type. Larger circles and darker purple represent higher proportions of shared communities. Sample sizes for each participant in home can be found in Supplementary Table 4. Partner hand samples were removed from these figures because not all participants had a partner.

Figure 4

Shared OTUs by sample type: (a) heat map representation of the summed shared OTUs between each sample type for the office environment; (b) heat map representation of the summed shared OTUs between each sample type for the home environment.