The effect of sample type and location on industrial workplace sink and hand dryer microbiomes

Abstract

One major issue in tackling antimicrobial resistance (AMR) is the ability to effectively track resistance spread in environments where surveillance is limited. Such environments include those experiencing high volumes of hand washing and drying from multiple users. This study characterised the microbial populations and antimicrobial resistomes of two different sample types from a pharmaceutical industrial site as part of an AMR environmental surveillance programme. Paired samples were collected from hand dryers and adjacent sinks in distinct sampling locations: from toilets adjacent to 'wet' labs, and locations associated with 'dry' activities. Microbial populations in hand dryers were significantly different to those of sinks, whereas there was no significant difference based on sample location. The opposite effect was observed for resistomes, where profiles differed significantly based on sample location, but not sample type. When both sample type and location were considered together, differences in microbiomes were driven primarily by hand dryer profiles from different locations. Analysis of metagenomically-assembled genomes revealed the presence of many poorly characterised organisms, and suggested no specific families predominated in terms of ARG carriage. This study emphasises the impact of human activities in determining the resistome of commonly used appliances, and the need for continued AMR surveillance programmes.

Keywords: Antimicrobial resistance; Environmental surveillance; Hand dryer; Microbiome; Sink; Workplace.

Fig. 1

Taxonomic classification of sink and hand dryer metagenomes, showing relative abundance at Family level

Fig. 2

Alpha diversity analyses of sink and hand dryer microbiomes based on taxonomic OTUs at the Family level using both (A) Shannon, and (B) Simpson indices

Fig. 3

PCoA of beta diversity of sink and hand dryer microbiomes based on taxonomic OTUs at the Family level, showing profile separation based on sample type. PERMANOVA revealed statistically significant differences between community profiles at Family level between hand dryers and sinks, using both distance measurements (Bray-Curtis p = 0.002; Jaccard p = 0.004)

Fig. 4 A

Mean ARG count in different sample types. Error bars show the standard deviation within each group. PCoA showing beta diversity of ARGs across sample types using B. Bray-Curtis and C. Jaccard distances

Fig. 5 A

Mean ARG counts by both sample type (HD vs. sink) and location (Lane vs. E&M), showing significantly fewer ARGs in hand dryer samples from lane locations compared to hand dryers from E&M locations (p = 0.0134). B. Heatmap showing relative abundance of ARGs per sample. ARG profile relationships between samples are highlighted, with grouping seen among hand dryer samples and sink samples, respectively. Further groupings are observed within sample types by location, with hand dryer samples from E&M (red), separate to hand dryers from lanes (blue). Similarly, sink samples from E&M (green) groups separately to sink samples from lane (purple). PCoA based on ARG profiles beta diversity showing separation by location within sample types based on C. Bray-Curtis dissimilarity and D. Jaccard distances. Hand dryer samples from E&M (red), separate to hand dryers from lanes (blue). Similarly, sink samples from E&M (green) groups separately to sink samples from lane (purple)

Fig. 6

Total ARG counts identified in MAGs from each sample, showing the breakdown of ARG source by bacterial family following MAG identification