Geography and location are the primary drivers of office microbiome composition

Abstract

North Americans spend the majority of their time indoors where they are exposed to the microbiome of the built environment (BE) they inhabit. Despite the ubiquity of microbes in BEs, and their potential impacts on health and building materials, basic questions about the microbiology of these environments remain unanswered. We present a study on the impacts of geography, material type, human interaction, location in a room, seasonal variation, and indoor and microenvironmental parameters on bacterial communities in offices. Our data elucidates several important features of microbial communities in BEs. First, under normal office environmental conditions, bacterial communities do not differ based on surface material (e.g., ceiling tile or carpet), but do differ based on the location in a room (e.g., ceiling or floor), two features which are often conflated, but which we are able to separate here. We suspect that previous work showing differences in bacterial composition with surface material were likely detecting differences based on different usage patterns. Next, we find that offices have cityspecific bacterial communities, such that we can accurately predict which city an office microbiome sample is derived from, but officespecific bacterial communities are less apparent. This differs from previous work which has suggested officespecific compositions of bacterial communities. We again suspect that the difference from prior work arises from different usage patterns. As has been previously shown, we observe that human skin contributes heavily to the composition of BE surfaces. ImportanceOur study highlights several points that should impact the design of future studies of the microbiology of the BEs. focus sampling effort on surveying different locations in offices and in different cities, but not necessarily different materials or different offices in the same city. Next, disturbance due to repeat sampling, though detectable, is small compared to other variables, opening up a range of longitudinal study designs in the built environment. Next, studies requiring more samples than can be sequenced on a single sequencing run (which is increasingly common) must control for run effects by including some of the same samples on all sequencing runs as technical replicates. Finally, detailed tracking of indoor and material environment covariates is likely not essential for BE microbiome studies, as the normal range of indoor environmental conditions is likely not large enough to impact bacterial communities.

Airborne bacterial assemblage in a zero carbon building: A case study

“…Correlation between OTU richness and RH was not observed in the autumn in any location, and temperature was not significantly correlated with OTU richness, regardless of season, and location (Spearman's correlation between each location and RH and temperature P < .05). The effects of temperature and RH on the indoor microbiome may be dependent on the BEs in question, as there have been mixed findings regarding the roles of these environmental conditions on indoor air or surface microbial assemblages …”

Environmental DNA metabarcoding: Transforming how we survey animal and plant communities

“…For example, recently in a study focused on bacterial biodiversity using the 16S locus, it was shown that a run effect can be confounded with a sample effect if it is not accounted for (e.g., by splitting sample groups across multiple Illumina runs, Chase et al, 2016); however, it remains to be seen whether such technical artefacts are also prevalent for loci used for metabarcoding plant and animals from eDNA (COI, 18S, ITS, etc. ), and more research is needed.…”

Analyzing differences between microbiome communities using mixture distributions

“…To illustrate the method, we apply it to two microbiome datasets. The first is a study of environmental microbiomes collected in office environments available from Qiita (http://qiita.microbio.me, Study ID: 10423). The samples are restricted to dust collected at university campuses in Toronto, Ontario, Canada and Flagstaff, Arizona, USA with at least 5000 classified reads, resulting in 981 samples.…”

“…The results for the dust data, collected by Chase et al Inference is performed on order, family, genus, and species levels. Univariate methods use the aggregates at each taxonomic level, whereas the distance methods use each component separately in the distance calculation.…”

“…Studies have linked changes in the microbiome to a variety of health outcomes, including inflammatory bowel disease, skin conditions, and immunity . Variance in microbiome composition can be a result of many factors, including genetic and environmental, but if targeted changes to bacterial composition can be used to alleviate symptoms or to cure, then the population burden of disease can potentially be reduced. Recent advances in technology have also made measuring the composition of a microbiome more accessible, resulting in steadily increasing cohort sizes.…”