Microplastics accumulate fungal pathogens in terrestrial ecosystems

Abstract

Microplastic (MP) is a pervasive pollutant in nature that is colonised by diverse groups of microbes, including potentially pathogenic species. Fungi have been largely neglected in this context, despite their affinity for plastics and their impact as pathogens. To unravel the role of MP as a carrier of fungal pathogens in terrestrial ecosystems and the immediate human environment, epiplastic mycobiomes from municipal plastic waste from Kenya were deciphered using ITS metabarcoding as well as a comprehensive meta-analysis, and visualised via scanning electron as well as confocal laser scanning microscopy. Metagenomic and microscopic findings provided complementary evidence that the terrestrial plastisphere is a suitable ecological niche for a variety of fungal organisms, including important animal and plant pathogens, which formed the plastisphere core mycobiome. We show that MPs serve as selective artificial microhabitats that not only attract distinct fungal communities, but also accumulate certain opportunistic human pathogens, such as cryptococcal and Phoma-like species. Therefore, MP must be regarded a persistent reservoir and potential vector for fungal pathogens in soil environments. Given the increasing amount of plastic waste in terrestrial ecosystems worldwide, this interrelation may have severe consequences for the trans-kingdom and multi-organismal epidemiology of fungal infections on a global scale.

Figure 1

Fungal colonisation of MP fragments visualised by SEM. White arrows indicate specific structures. (a) Mat of fungal conidia, including potentially germinating spores (1), closely attached to the plastic surface presumably through a form of self-produced mucilage (2). (b) Numerous conidia lining a crack in the plastic surface. (c) Clumping of conidia in association with a hypha. (d) Mycelial meshwork. (e) Extensive intertwined hyphal filaments adhering to the plastic surface via small peripheral bulges (3). (f) Conidia-producing hypha (conidiophore) with exposed vesicle after discharge (asterisk). Scale bars are 30 µm.

Figure 2

Alpha diversity metrics for the entire rarefied dataset according to substrates and sites. Samples were rarefied at the smallest library size (2496 sequences). Significance testing (ANOVA) revealed significantly higher fungal richness (observed, Chao1; both p < 0.01) and Shannon diversity (p < 0.01) of soil in comparison to plastic samples (Supplementary Fig. 2).

Figure 3

Variation in fungal community composition between plastic and soil visualised by NMDS. All four plots are based on Bray–Curtis dissimilarity matrices of square root transformed relative abundances of OTUs. Taxa subsets of the most dominant ecological guilds were compiled following meta-analysis and included only OTUs classified to genus or species level. Ellipsoids represent a 95% confidence interval surrounding the data points of each factor group. Significant variation between substrates was tested by PERMANOVA (included p-values) (Supplementary Table 4), while similarities within and between groups were additionally assessed using ANOSIM (sample statistic ‘R’) (Supplementary Table 6). NMDS ordination stress values included.

Figure 4

Phylogenetic distributions within the PCM. Displayed are those 22 OTUs accounting for ≥ 0.5% of plastisphere reads and contributing ≥ 0.3% to the plastic/soil dissimilarity based on SIMPER analysis with occurrence on MP in every site. OTUs classified at the species level (black font) were annotated with trait data. OTUs classified to at genus or higher level were not annotated (grey font). (*) indicates ‘unclassified’. Branch colours code for the indicated fungal classes, where Spiz. = Spizellomycetes, Micr. = Microbotryomycetes, Trem. = Tremellomycetes, Leot. = Leotiomycetes. Ring 1 indicates the success of OTU assignment based on the presence of a representative type strain or isolate ITS sequence match in the UNITE database. Ring 2 shows the most likely trophic mode for each taxon. Ring 3 displays the host (kingdom) range of the fungal taxa. The presence of at least one triangle in Ring 4 indicates potential human pathogenicity of the respective fungus, while the number and colour of the triangles code for possible virulence based on the known infection sites of the human body. Ring 5 and 6 indicate the relative abundance of each taxon on soil and plastic, respectively.

Figure 5

Differentially abundant fungal species of the soil and plastisphere mycobiome identified by DESeq2 analysis. Negative values indicate a significantly (p < 0.05) higher abundance of species on MP. Species are plotted according to their log2 fold change differential abundance and color-coded according to classes. Size factor normalisation was implemented in DESeq2. Bubble size corresponds to the mean read counts across all samples.