Microbial communities colonising plastics during transition from the wastewater treatment plant to marine waters

Abstract

Background: Plastics pollution and antimicrobial resistance (AMR) are two major environmental threats, but potential connections between plastic associated biofilms, the 'plastisphere', and dissemination of AMR genes are not well explored.

Results: We conducted mesocosm experiments tracking microbial community changes on plastic surfaces transitioning from wastewater effluent to marine environments over 16 weeks. Commonly used plastics, polypropylene (PP), high density polyethylene (HDPE), low density polyethylene (LDPE) and polyethylene terephthalate (PET) incubated in wastewater effluent, river water, estuarine water, and in the seawater for 16 weeks, were analysed via 16S rRNA gene amplicon and shotgun metagenome sequencing. Within one week, plastic-colonizing communities shifted from wastewater effluent-associated microorganisms to marine taxa, some members of which (e.g. Oleibacter-Thalassolituus and Sphingomonas spp., on PET, Alcanivoracaceae on PET and PP, or Oleiphilaceae, on all polymers), were selectively enriched from levels undetectable in the starting communities. Remarkably, microbial biofilms were also susceptible to parasitism, with Saprospiraceae feeding on biofilms at late colonisation stages (from week 6 onwards), while Bdellovibrionaceae were prominently present on HDPE from week 2 and LDPE from day 1. Relative AMR gene abundance declined over time, and plastics did not become enriched for key AMR genes after wastewater exposure.

Conclusion: Although some resistance genes occurred during the mesocosm transition on plastic substrata, those originated from the seawater organisms. Overall, plastic surfaces incubated in wastewater did not act as hotspots for AMR proliferation in simulated marine environments.

Fig. 1

An overview of experimental setup. A One of the three mesocosm fibreglass tanks with suspended nylon mesh bags containing the plastic units. B One of the three storage containers used to carry out serial incubations, with river water incubation tank. C Nylon mesh bag layout. Each bag contained 12 experimental units of one plastic type (HDPE, LDPE, PET or PP). D Schematic of the experimental design where plastic was incubated in WWTP effluent, river water, brackish water and seawater to simulate the pathway plastic takes from WWTP to a marine environment

Fig. 2

Heatmap of relative abundances of prokaryotic 16S rRNA gene amplicon reads. “Seawater Day 0—Seawater 16 Week” refer to the planktonic mesocosm seawater communities at corresponding timepoints. HDPE, LDPE, PET and PP refer to the compositions of plastic-associated biofilms that emerged after pre-incubation of above plastics in WW effluent (24 h), river (24 h) and brackish (24 h) waters and their exposure to the seawater in the mesocosms for periods between 0 days and 16 weeks. “Effluent”, “River” and “Brackish” correspond to planktonic microbial communities in fresh samples of WW effluent, river and brackish waters used for sequential pre-incubation of plastics. “Wall” samples correspond to microbial communities developing on the wall of mesocosm tanks

Fig. 3

16S rRNA amplicon sequencing results Non-MultiDimensional Scale (NMDS) analysis for enrichments on LDPE, PET and PP for 140 days (20 weeks). Ellipses cluster samples of same timepoint. Major changes in the communities occur between timepoint 0 and 1 week, at the later colonisation stages, community compositions tend to co-cluster. This analysis has informed the selection of samples for shotgun sequencing

Fig. 4

Relative abundance of bacterial families in microbial communities from different water and plastic surface samples based on number of reads mapped per million base pairs (Mbp) of assembled metagenome. Samples include wastewater treatment plant (WWTP) effluent, river water, seawater, and biofilms on low-density polyethylene (LDPE), polyethylene terephthalate (PET), and polypropylene (PP) plastics. Plastics were incubated sequentially in WWTP effluent, river water, brackish water, and seawater for 1 week each prior to sampling. Protein sequence identity cut off at 90% + . ANOVA-generated p-values for interactions between sample type and bacterial family are shown on significant heatmap tiles. Statistical significance: ***p < 0.001; **p < 0.01; *p < 0.05; p < 0.1

Fig. 5

Bray Curtis based non-metric multidimensional scaling plot of plastisphere sample community structure following successive incubations of plastic in WWTP effluent, river water, brackish water and seawater from WGS data. Centroids of baseline communities (planktonic WWTP effluent, river water and seawater) are labelled. Coloured bubbles represent the areas covered by each timepoint. Timepoints of 1-day and 1-week are shown by different shapes and plastics (LDPE, PET and PP) by colour

Fig. 6

Relative abundance of ARG classes across different microbial communities within wastewater treatment plant effluent, river water, seawater, and microbial communities colonising PET, PP and PE after successive passage through these water types and 1 week in seawater mescocosm. Z-score has been calculated by row, based on the DeepARG deep-learning model

Fig. 7

AMR gene copy number per Mbp of assembled metagenomes normalised for microbial communities in wastewater treatment plant effluent, river water, seawater, and communities that colonise LDPE, PET and PP plastics after pre-exposure to wastewater, river water brackish water and seawater at 1-week. Abundance changes are shown by the colour gradient. Significant ANOVA generated p-values for interactions between sample type and ARG are shown on significant heatmap tiles. Statistical significance: ***p < 0.001; **p < 0.01; *p < 0.05, p < 0.1. Based on CARD database [49]