Hydrophobins from Aspergillus mediate fungal interactions with microplastics

Abstract

Microplastics cause negative environmental consequences such as the release of toxic additive leachates, increased greenhouse gas emissions during degradation, and threaten food chains. Microplastic particles are known to serve as a vector for transport of microbes (fungi and bacteria) to new environments, threatening biodiversity. Robust biofilm formation makes fungi a candidate to collect and remediate environmental microplastics. However, fungal-microplastic colonization mechanisms have yet to be explored. In this work, we aim to understand which fungal molecules mediate microplastics binding. We examine common fungal genus Aspergillus, which we found binds microplastics tightly, removing particles from suspension. Upon inoculation of Aspergilli with microplastics particles, up to 3.85 +/− 1.48 g of microplastics were flocculated per gram of dry fungal biomass; this phenomenon was observed across various plastics ranging in size from 0.05 – 5 mm. Gene knockouts revealed that hydrophobins drive microplastic-fungi binding, evidenced by a decrease in flocculation relative to wild-type Aspergillus fumigatus. Moreover, purified hydrophobins flocculated microplastics independently of the fungus, validating their ability to bind to microplastics. Our work elucidates a role for hydrophobins in fungal colonization of microplastics and highlights a target for mitigating the harm of microplastics through engineered fungal-microplastic interactions.

Graphical Abstract

Introduction:

Microplastics accumulation in the environment has led to myriad environmental, biodiversity, and human health issues^1–7. An estimated 275 million tons of plastic are disposed at end of life each year⁸,with the majority of these plastics degraded into microplastic particles dispersed in our air, water, and soils⁹. Their presence in the environment leads to the release of toxic xenobiotics through additive leaching¹⁰, and increases in greenhouse gas levels from environmental microplastics degradation and toxicity to phytoplankton that fix atmospheric CO₂¹¹. Microplastic particles present negative environmental consequences worldwide, as they are found ubiquitously in soil, groundwater, marine environments, estuaries, and landfills due to their low density and small size that allows them to easily be carried to new locations¹². As a result of this rapid transport, microplastics with biofilm-bound microbes are often transported to new environments, leading to disruption of the local microenvironment through the introduction of new, possibly invasive or pathogenic, species¹³. Moreover, microplastic particles have been observed in the intestinal tracts, tissues, and organs of marine organisms throughout the ecosystem¹⁴ and have been shown to lead to negative health consequence in mice, zebrafish, and other animals such as inflammation, metabolic disorders, and decreases in reproduction¹³. Microplastic particles subsequently work their way through animal and human food chains, potentially exacerbating these negative health effects through the entire food web, threatening biodiversity¹⁵. Consequently, microplastics have been found in human placentas¹⁶, testes¹⁷, and blood¹⁸ leading to negative health effects from leaching of toxic monomers, additives, and adsorbed environmental pollutants^19–23. More importantly, polyethylene microplastics in human arteries increase the likelihood of cardiovascular events, stroke, or death²⁴. These environmental and human health issues continue to worsen with exponential increases in plastic production and subsequent increases in environmental contamination²⁵.

Microorganisms frequently interact with microplastics, forming robust biofilms on their surface. These biofilms lead to the transport of new microbes to foreign environments, disrupting the local ecosystem and food chains^12,26. Moreover, microplastic biofilms are often enriched in pathogens such as those from the genera Pseudomonas^27,28 and Vibrio²⁸. Pathogens tend to be enriched in biofilms due to their ability to promote cell fitness via horizontal gene transfer of antibiotic resistance genes that improve microbial viability of other members of the microbial community²⁹. Additionally, pathogens such as Vibrios have been noted to evolve into hyperbiofilm-formers in stressed microenvironments²⁹. The presence and enrichment of pathogenic microbes in these biofilms can exacerbate animal health, biodiversity, and human health impacts by introducing new pathogens into microbial communities and food chains and by harboring increased horizontal transfer of antibiotic resistance genes between pathogens^27,28,30. Such pathogenic consequences are evidenced by disease outbreaks in marine environments tied to the migration of pathogens on plastics waste^31,32. The taxonomic profiles of microplastic-associated biofilms are well documented^33–35, as taxonomic changes to biofilm members vary dependent on sampling location, plastic type, and particle size^33,36,37. While there is a strong understanding of the types of microorganisms that bind to microplastic particles under various conditions, the specific biomolecules responsible for microbial binding to microplastics are poorly understood.

Microbes often form biofilms on solid surfaces through secretion of biosurfactants and/or surface proteins. For example, bacteria often rely on flagella or pili to attach to surfaces and form biofilms^38,39. Additionally, many bacteria secrete extracellular polymeric substances (EPS) containing proteins and lipopolysaccharides (LPS) that promote biofilm hydrophobicity and allow for surface binding⁴⁰. Similarly, fungal adhesion to extracellular surfaces is canonically driven by surface proteins called adhesins⁴¹. Adhesins are responsible for cell-cell adhesion, biofilm formation, and adhesion to hydrophobic surfaces in model yeasts like S. cerevisiae⁴². Common fungi such as Aspergilli secrete adhesins belonging to the class hydrophobins that allow them to form strong hyphal networks and adhere to extracellular surfaces⁴³. Hydrophobins are a class of small, secreted fungal proteins (~10–15kDa) that form amphipathic layers at hydrophobic/hydrophilic interfaces⁴³, allowing them to bridge fungi to extremely hydrophobic substrates. Though they are known to form strong biofilms on solid surfaces, the interactions of fungi with (micro)plastics are understudied. However, there is growing interest in the fungal members of microplastic-associated biofilms and their interactions due to the inherent pathogenicity of many fungi and their propensity for horizontal gene transfer^36,44. Aspergillus niger has been documented to interact with and bind to polystyrene (PS) and Poly(methyl methacrylate) (PMMA), removing PS and PPMA from solution⁴⁵, but binding mechanisms were not studied. These fungi-microplastics relationships are essential to understand how microplastics are colonized, mitigation of health risks from microplastic-bound pathogenic fungi, potential toxicity effects, and how to better remove microbes from microplastics for recovery.

In this study, we leverage Aspergillus fumigatus to better understand the manner in which fungi bind to microplastics because it is a reported opportunistic pathogen and is found ubiquitously across soil and marine environments^46,47. We isolated a novel strain of Aspergillus fumigatus that forms extremely hydrophobic biofilms, recovering nearly 100% of microplastics from suspensions. These observations, concurrent with those found in A. niger, imply that there is a conserved mechanism across all Aspergilli for microplastics binding. Our observations of microplastics flocculation across several strains covering the genus phylogeny validate this hypothesis. Moreover, we confirmed that microplastics recovery occurs ubiquitously across various single and mixed plastic types, demonstrating that fungal-microplastics interactions are conserved on model post-consumer plastic waste streams. We hypothesized that microplastics bind to hydrophobins on Aspergilli hyphae due to their abundance and ubiquity across the genus and due to their inherent hydrophobicity^43,48,49. Additionally, hydrophobins have been reported to bind to and enhance enzymatic plastics deconstruction ⁵⁰. For example, RolA from A. oryzae has been reported to bind to polybutylene succinate-coadipate and recruit esterases for its deconstruction^51,52 and hydrophobins have been used to enhance enzymatic PET hydrolysis⁵⁰. Here, we show that hydrophobin proteins from A. fumigatus are the primary driver for microplastic binding by Aspergillus through gene knockouts and confirming pure hydrophobin binding to microplastic particles. The understanding that hydrophobins are responsible for microplastic binding can be used to reverse biofilm formation by pathogenic strains such as Aspergillus fumigatus, subsequently mitigating potential pathogenicity of microplastics, reducing negative health effects on animals. Additionally, hydrophobins can be used in the absence of (pathogenic) hosts to provide sustainable microplastics recovery from aqueous environments, alleviating negative consequences to biodiversity.

Materials and methods:

Organism Isolation:

Aspergillus fumigatus AF-UD1 was isolated from the gut of a yellow mealworm (Tenebrio molitor larvae) fed HDPE for 20 days. 10 mealworm guts were extracted, suspended in 1 mL of PBS, and vortexed to homogenize. 50 μL of gut contents were plated on fungal Medium B (defined previously⁵³). Individual colonies were re-plated on Medium B to isolate the organism. The organism was originally isolated in a co-culture with an un-identified bacterial strain. AF-UD1 was isolated from the co-culture by plating on potato dextrose agar with 50 μg/mL penicillin and 50 μg/mL streptomycin.

Organism Identification:

Whole genome sequencing was carried out on genomic DNA from AF-UD1, with details of each method outlined below.

DNA Extraction:

High molecular weight DNA was extracted from mycelium using the protocol of Puppo et al (2017)⁵⁴ with minor modifications. Flash-frozen biomass was ground to a fine powder in a frozen mortar with liquid nitrogen followed by very gentle extraction in 3X CTAB extraction buffer (3% CTAB (hexadecyltrimethylamonium bromide), 1.4 M NaCl, 100 mM Tris pH 8.0, 20 mM EDTA, 1% 2-mercaptoethanol) for 1h at 65 °C. The mixture was cooled down and gently extracted with 24:1 Chloroform:Isoamyl alcohol. The upper phase was collected and gently extracted again with 24:1 Chloroform:Isoamyl alcohol. The aqueous phase was transferred to a new tube and 1/10th volume 3 M Sodium acetate was added, gently mixed, before precipitating the DNA with iso-propanol. The sample was kept in −20 °C overnight to facilitate precipitation. DNA precipitate was collected by centrifugation, washed with 70% ethanol, air dried for 5 minutes and dissolved thoroughly in elution buffer at room temperature followed by RNAse treatment. DNA purity was measured with Nanodrop, DNA concentration measured with Qubit HS kit (Invitrogen) and DNA size was validated by Femto Pulse System (Agilent).

Genome Sequencing:

The draft genome of AF-UD1 was sequenced using PacBio Multiplexed 6–10kb Ultra-Low Input library sequenced using the REVIO. An input of 50 ng of genomic DNA was sheared to 6 kb - 10 kb using the Megaruptor^® 3 (Diagenode) or g-TUBE (Covaris). The sheared DNA was treated with DNA damage repair enzyme mix, end-repair/A-tailing mix and ligated with amplification adapters using SMRTbell Express Template Prep Kit 3.0 (PacBio) and purified with SMRTbell cleanup beads. The purified ligation product was split into two reactions and enriched using 10–18 cycles of PCR using barcoded amplification oligos (IDT) and SMRTbell^® gDNA Sample Amplification Kit (PacBio). Up to sixteen libraries were pooled in equimolar concentrations and the pooled libraries were size-selected using the 0.75% agarose gel cassettes with Marker S1 and High Pass protocol on the BluePippin (Sage Science). The size-selected pools were treated with DNA damage repair enzyme mix, end-repair/A-tailing mix and ligated with SMRTbell sequencing adapters, a nuclease enzyme mix and purified with SMRTbell cleanup beads. CCS data was filtered with the JGI QC pipeline to remove artifacts. CCS reads were assembled with Flye version 2.9-b1768 (https://github.com/fenderglass/Flye) and subsequently polished with two rounds of RACON version 1.4.13 ⁵⁵. The mitochondrial sequence was identified based on coverage, GC content, and BLAST hits to the NCBI nt database, used to filter the CCS reads to produce non-organelle CCS, and polished with two rounds of RACON version 1.4.13⁵⁵.

Gene calling was facilitated through transcriptome acquisition and a single RNA-seq library was produced as input for the Joint Genome institute (JGI) annotation pipeline. The JGI gene finding methods include the use of filtered reads, or trinity-assembled contigs as transcriptional evidence that the gene is translated. mRNA was isolated from an input of 200 ng of total RNA with oligo dT magnetic beads and fragmented to 300 bp - 400 bp with divalent cations at a high temperature. Using the TruSeq stranded mRNA kit (Illumina), the fragmented mRNA was reverse transcribed to create the first strand of cDNA with random hexamers and SuperScript^™ II Reverse Transcriptase (Thermo Fisher Scientific) followed by second strand synthesis. The double stranded cDNA fragments were treated with A-tailing, ligation with NEXTFLEX UDI Barcodes (PerkinElmer) and enriched using 10 cycles of PCR. The prepared libraries were quantified using KAPA Biosystems’ next-generation sequencing library qPCR kit and run on a Roche LightCycler 480 real-time PCR instrument. Sequencing of the flowcell was performed on the Illumina NovaSeq sequencer using NovaSeq XP V1.5 reagent kits, S4 flowcell, following a 2×151 indexed run recipe. RNA-Seq reads were trimmed for artifact sequence by kmer matching (kmer=25), allowing 1 mismatch, from the 3’ end of the reads, and filtered for spike-in reads, PhiX reads and reads containing any Ns.

Quality trimming of the genome was performed using the phred trimming method set at Q6. Finally, following trimming, reads under the length threshold were removed (minimum length 25 bases or 1/3 of the original read length - whichever is longer). Filtered reads were assembled into consensus sequences using Trinity v2.12.0⁵⁶. The genome was annotated using the JGI Annotation pipeline and made publicly available via JGI fungal genome portal MycoCosm⁵⁷. The genome is available on JGI MycoCosm at https://mycocosm.jgi.doe.gov/AspfumUD1_1/AspfumUD1_1.info.html

Protein Clustering and Phylogenetic Analysis:

Aspergillus species below were downloaded from MycoCosm and included in orthofinder v2.55 clustering with A. fumigatus UD1⁵⁸. Briefly, all GeneCatalog proteins were clustered into orthologous groups (orthogroups) by sequence similarity. A total of 5345 orthogroups contained every species, and were aligned to produce a species tree using default methods^59,60. The tree file was plotted along with MycoCosm assembly and gene count metrics by phytools⁶¹. Genome references are original publications unless sourced from AspGD⁶²: A. flavus NRRL 3357⁶³; A. fischeri NRRL 181⁶²; A. fumigatus Af293⁶⁴; A. fumigatus A1123⁶⁵; A. nidulans FGSC A4⁶²; A. niger NRRL3⁶⁶; A. novofumigatus IBT 16806⁶⁷; A. terreus NIH 2624⁶²; A. udagawae IFM 46973⁶⁸

Microplastics recovery assays:

Aspergillus pre-cultures were grown in YPD at 37 °C, 220 rpm in 5 mL cultures for 2 days directly from a freezer stock. The pre-culture was then inoculated into 100 mL YPD to and grown at 37 °C, 100 rpm in a 500 mL flask to grow Aspergillus ‘flocs’ of LDPE (Goodfellow Cambridge Limited, Huntingdon, England; catalog number LS563303), PP (post-consumer yogurt Chobani yogurt cups; PP disks/beads cut out using a 2 mm diameter hole punch), PET (post-consumer Dasani water bottles; PET disks/beads cut out using a 5 mm diameter hole punch), or UHMWPE (Sigma-Aldrich Chemical Company; catalog number 43272 – 100g). Or, cellulose acetate (0.45 μm, Sterlitech corporation, Auburn, WA, USA; catalog number CA0459025) used in place of microplastic particles. In addition, environmentally relevant cryomilled and UV weathered PVCmicronanoplastics with an average size of 5 um, were provided by the Rutgers Nanoscience and Advanced Materials Center and used ⁶⁹. Details on their synthesis and physicochemical properties are summarized in the Suppl. information section and detailed by the authors in the Das et al publication⁶¹). Approximately 25 mg microplastics (or cellulose acetate) particles were suspended in 5 mL of sterile mineral media (1g NaH₂PO₄, 0.5 g MgSO₄*7H₂O, 0.2 g KH₂PO₄, and 1 g yeast extract per 1 liter). 2–3 flocs from the large Aspergillus culture were dropped into the 5 mL culture containing plastic powder or disks. The cultures were allowed to shake at 30 °C, 220 rpm to allow the Aspergillus strains to slowly grow in a nutrient deprived environment. Every 2 hours, the culture was shaken to allow fungi to come into direct contact with the plastic. Once plastic was flocculated at one of these 2 hour intervals, the culture was removed for analysis. If no plastic was grabbed by the fungi (in the case of Aspergillus knockout experiments) after 36 hours, the culture was removed and discarded.

For mass-normalized microplastics recovery assays, flocs of A. fumigatus UD01, A. fumigatus (ATCC 1022), A. niger (ATCC 16888), A. nidulans (ATCC10074), A. flavus (ATCC 16833), or A. terreus (ATCC 1012) containing microplastics were removed from the culture tubes using sterile weighing spatulas onto pre-weighed dishes and allowed to dry. Remaining plastic was subsequently removed from the tube by flushing the tube with water and ensuring all remaining contents of the tube were deposited onto a pre-weighed dish. Both pre-weighed dishes were then weighed after drying. Mass of flocculated plastic was calculated by subtracting the mass of remaining plastic from the initial plastic mass. Additionally, the microplastic-flocculated cultures were weighed. Calculated flocculated plastic mass was subtracted from the mass of the dried flocculated plastic fungal culture to obtain dry biomass weight.

For microplastics recovery assays that included beta-mercaptoethanol (βME), the βME was added into the culture tube with plastic and prior to adding the fungal flocs. Flocs were added and the protocol outlined above was followed.

For recovery assay with pure RodA, 10 mg of green fluorescent LDPE particles (Cospheric LLC, Somis, CA; catalog number UVPMS-BG-1.00 35–45 um) were placed into 1 mL of solution containing purified RodA. The solution was lightly shaken to mix and then left stationary at room temperature overnight to allow separation.

Preparation of weathered PVC (w-PVC) particles:

Methods for weathering and materials testing of w-PVC particles can be found in Supplementary information file 2. Particles were prepared via methods in ref. ⁶⁹

Confocal microscopy:

Microplastics recovery assays were carried out as above, but with red fluorescent LDPE beads (Cospheric LLC, Somis, CA; catalog number UVPMS-BR-0.995 45–53 um). After flocculation, floc culture was washed with PBS three times. The culture was placed into a microscopy sample dish (ibidi ibiTreat: #1.5 polymer coverslip, tissue culture treated, sterilized) in 2 mL PBS. One microliter of calcofluor white was added to the culture and it was stored in darkness for 15 minutes to stain the fungi. Microscopy images were taken using a Stellaris 8 tauSTED/FLIM Confocal Microscope at the University of Delaware Bioimaging Center.

Scanning electron microscopy:

Microplastics recovery assays were carried out as above. After flocculation, floc culture was washed with PBS three times. Culture flocs were coated in platinum and imaged on the Apreo VolumeScope^™ Scanning Electron Microscope at the University of Delaware Bioimaging Center.

Hydrophobin knockout strains:

Mutant hydrophobin knockout strains were generously donated by Jean-Paul Latgé and Isabelle Mouyna from Aspergillus Unit, Institut Pasteur, 75015 Paris, France. The hydrophobin knockouts were generated from methods listed in the original publication⁴³.

RodA cloning in Yarrowia lipolytica:

RodA sequence was codon optimized for Yarrowia lipolytica and the resulting gene fragment was ordered from Twist Biosciences with an AscI restriction site on the N terminus and an NheI restriction site on the C terminus. The gene fragment was digested with AscI and NheI enzymes at 37 °C for 1 hour and then therestriction enzymes were heat inactivated at 80 °C for 20min. The vector for cloning was a homology donor for integration into the AXP site in Yarrowia lipolytica⁷⁰. The vector was also digested with AscI and NheI at 37 °C for 1 hour. The vector and insert were ligated using NEB DNA ligase. 5uL of ligation mix was added to 50 μL of NEB 10β competent cells which were heat shocked at 42 °C for 30 seconds and then recovered in 1mL of LB media for 1 hour at 37°C with shaking. 100 μL of transformed cells were plated onto an ampicillin containing LB plate and grown overnight at 37 °C. The sequence of the resulting plasmid was confirmed by Sanger sequencing. The RodA gene was integrated into the AXP site using the homology donor and a CRISPR containing plasmid. Integration was confirmed via colony PCR and then the strain was cured of all plasmids.

RodA expression and purification:

Yarrowia lipolytica RodA was grown for 4 days at 28°C with agitation at 220 rpm in 50 mL YPD. Culture was transferred to 50 mL falcon tube and centrifuged at 4000 rpm to separate pellet and supernatant. Supernatant was ultra-centrifuged for 1hr at 100,000g in SW32Ti rotor using OptiSeal 32mL tubes and adapters from Beckman Coulter. Supernatant was removed and pellet was resuspended in 1 mL of 2% SDS. The sample was transferred to microcentrifuge tube and boiled for 10 min at 98 °C. In 5mL thinwall open top tubes (Bechman Coulter), sample was ultra-centrifuged at 100,000 g in 5 mL of SDS for 1 hr at 20 °C. This process was repeated 2 times. All SDS was removed and the pellet was resuspended in 5mL DI water. The sample was then ultra-centrifuged at 100,000 g for 1 hr at 20 °C. This process was repeated 2 times. Resulting hydrophobin pellet was transferred to a 1.5 mL microcentrifuge tube with 1mL DI water. RodA expression and purification protocol was adapted from⁷¹. Purified RodA was then ran on a tris-trycine SDS-PAGE gel stained with silver stain using concentrations found in Supplementary Table 1 calculated via BCA assay. Purity was confirmed via SDS-PAGE gel (Supplementary Figure 3) that shows a dominant band in RodA culture supernatant after going through the purification process. This band is not present in the wild type strain (Supplementary Figure 3).

Results:

A novel microbial isolate from the yellow mealworm gut microbiome flocculates microplastics from suspension

We discovered a microbial isolate from the gut of Tenebrio molitor that flocculates microplastics, pulling them out of suspension. The isolate rapidly flocculated suspended ultra-high molecular weight polyethylene (UHMWPE) particles and floating red fluorescent LDPE particles within seconds (Fig. 1A). Moreover, fungal-microplastics flocs are highly stable in solution; we observed that microplastics remain bound to fungi at 4°C and at room temperature for up to 6 months after conclusion of flocculation experimentation (data not shown). We evaluated the extent of microplastics flocculation capabilities of this isolate by using 25 mg (0.4% wt./vol) polypropylene (PP), poly(ethylene terephthalate) (PET), surface oxidized UHMWPE, and low-density polyethylene (LDPE) plastics to ensure that the strain can bind to microplastics independent of polymer chemistry and hydrophobicity. 25 mg/mL was selected as it is well above the range of reported microplastics concentrations of 1 – 2000 mg/L reported from marine and wastewater environments^72,73 and can be reliably measured in the laboratory. The isolate captured 96.0 +/− 4.0% of 200 μm LDPE particles, 97.1 +/− 0.6% of 50 μm UHMWPE particles, 100 +/− 0% of 5 mm PET beads, and 90.9 +/− 8.1% of 2 mm PP beads meaning flocculation is independent of both polymer chemistry and particle size (Fig. 1B). Mixed plastic types did not interfere with flocculation. Pairwise combinations of plastics were still recovered with 85–100% efficiency and 92% recovery when a 40 mg (0.8% wt. vol) mixture of all 4 microplastic types and sizes was tested (Fig. 1B). These combinatorial samples were performed as they more closely emulate the mixed microplastics streams that occur in marine environments. Samples containing PP and PET beads have higher variance due to their larger particle size. If one PP or PET bead was not recovered, it significantly decreases the plastic recovery on the per mass basis. It is likely that the increased surface area of the larger particles requires more binding interactions to retain the plastic within the biofilm, leading to some particles not being captured due to lack of available fungal surface area. Nonetheless, microplastics flocculation is nearly 100% in both ‘pure’ and mixed plastic cases, with pristine and post-consumer plastics of varying chemistries and sizes. Additionally, the flocculation phenotype is maintained upon plastics UV photooxidation and weathering, as ~5 μm weathered PVC particles⁶⁹ were recovered with a mass yield similar to PP beads (Supplementary Fig. 1). Lastly, the flocculation phenotype was also conserved when repeating assays using cellulose acetate, suggesting that the binding phenotype is not plastic-specific (Supplementary Fig. 2).

Figure 1: Novel fungal isolate ubiquitously capture microplastics from solution.

(A) Polyethylene particles captured from solution via flocculation by fungal isolate; 50 μm UHMWPE particles (top) and ~40 μm red fluorescent LDPE beads (bottom) were used to demonstrate microplastics binding. The images in the left hand panels represent ultra-high molecular weight polyethylene (UHMWPE) particles in suspension (top) and red fluorescent low-density polyethylene (LDPE) particles (bottom) sitting at the air liquid interface or stuck to culture tube walls. After addition of the fungal inoculum, the microplastic particles attach to fungal hyphae, becoming embedded into the mycelium. The images on the right hand side are fungal mycelia with embedded microplastic particles, UHMWPE (top) and RLDPE (bottom). (B) Microplastics recovery of a variety of ‘pristine’ and post-consumer plastics shows ubiquitous recovery near 100%. Microplastics recovery was calculated by subtracting the remaining, un-flocculated plastic mass from the initial mass, dividing that by the total initial mass, and multiplying by 100%. Error bars represent standard error across three replicates.

Microplastics flocculation is common amongst Aspergillus species.

We acquired the whole genome for our microplastic-flocculating isolate and taxonomically placed it as an Aspergillus though phylogenetic analysis of its internal transcribed spacer (ITS) (Fig. 2a). The strain was identified as Aspergillus fumigatus due to grouping with published Aspergillus fumigatus genomes on a species tree constructed using OrthoFinder FastTree^{58–60,64,65} and was thus named Aspergillus fumigatus UD1, hereafter referred to as AF-UD1 (Fig. 2b). Having identified our isolate, we next asked if this ability for microplastics colonization was conserved across the genus by assessing five common Aspergillus species spanning a range of phylogenetic distances from AF-UD1 (Fig 2). Each strain successfully flocculated microplastics between 1–5 g of plastic per g of dry biomass (Fig. 3). The recovery of microplastics by all strains implies that there are conserved molecular phenomena occurring in Aspergilli cultures that permit microplastics capture.

Figure 2: Novel plastics-flocculating isolate from mealworm gut is an Aspergillus.

(A) Phylogenetic tree of 45 Aspergillus strains built using complete ITS sequences, with strains used in this study boxed in red. A neighbor joining tree was constructed with 100 bootstrap iterations with ClustalW alignment, using Metarhizium anisopliaei as the outgroup. Tree is rooted to the outgroup. (B) Species tree confirming taxonomic identification of AF UD1. The tree was built by FastTree based on orthofinder clustering.

Figure 3: Plastics flocculation is conserved across Aspergilli.

Biomass normalized flocculation of two plastic types by Aspergillus strains across the genus. Mass of flocculated plastic was calculated by subtracting the mass of remaining plastic from the initial plastic mass. Images above each bar are 5 mL liquid cultures of each strain with flocculated ~200 μm goodfellow LDPE particles. Error bars represent standard error across three independent measurements

Microplastics flocculation is driven by redox-sensitive protein interactions

Confocal and scanning electron microscope (SEM) were used to observe microplastics flocculation and better understand underlying molecular phenomena. Microplastic particles are embedded both on the AF-UD1 surface and within the hyphal network (Fig. 4A, Supplementary movie). SEM images show a dense network of hyphae and extracellular polymeric substances (EPS) that pull plastic particles into the AF-UD1 network (Fig. 4B). The dense EPS and embedded nature of the microplastics formed by AF-UD1 creates a stable floc that can be mechanically perturbed without a loss of plastic. The formation of robust biofilm suggests that microplastics are pulled into the fungal matrix through hydrophobic interactions with secreted or membrane bound chemicals or biomolecules produced by the fungus.

Figure 4: Hydrophobins drive microplastics flocculation in Aspergilli.

(A) Confocal microscopy image of Aspergillus fumigatus AF-UD1 (blue) stained with calcofluor white using hyphal interactions to grab ~40 μm red fluorescent LDPE beads (red). (B) SEM image showing dense hyphal network of Aspergillus fumigatus AF-UD1 holding ~200 μm goodfellow LDPE microplastic particles in a floc. (C) Images showing ~200 μm goodfellow LDPE microplastics flocculation by AF-UD1 in the absence (left) and no flocculation in the presence (right) of beta-mercaptoethanol.

Aspergilli adhesion to extracellular surfaces is canonically driven by surface proteins⁷⁴. Hydrophobins are a highly surface-abundant class of proteins in Aspergillus, that have surfactant-like properties, namely amphiphilicity, making them very likely candidates to bind to extremely hydrophobic plastics^43,74. Moreover, hydrophobins are predominant proteins in the outermost hydrophobic layer^43,74 of Aspergillus fumigatus that form at hydrophobic/hydrophilic interfaces⁷⁵. Hydrophobins are characterized by eight conserved cysteine residues that form disulfide bonds that are responsible for stabilizing a large, hydrophobic solvent exposed interface⁷⁵. We disturbed these disulfide bonds using beta-mercaptoethanol (βME)⁷⁶, removing hydrophobins from the AF-UD1 surface, to determine if hydrophobins play a role in microplastic binding. Microplastics flocculation ability was eliminated upon the addition of βME to the culture, consistent with surface proteins such as hydrophobins that rely on disulfide bonds for structure being integral to microplastics recovery processes (Fig. 4C).

Hydrophobins are necessary for microplastics flocculation

The role of hydrophobins in microplastics flocculation was directly assessed by repeating microplastics flocculation assays using Aspergillus fumigatus strains with each hydrophobin knocked out of the genome. The Aspergillus fumigatus genome encodes 7 different hydrophobin genes, each expressing a different hydrophobin⁴³. Genes for hydrophobin expression are RODA, RODB, RODC, RODD, RODE, RODG, and RODF, corresponding to proteins RodA through RodF⁴³. Knocking out each hydrophobin gene reduced microplastics flocculation by Aspergillus fumigatus (Fig. 5A). AFΔRodA, AFΔRodB, AFΔRodE, AFΔRodG, and total knockout strain AFΔRodA-G all showed statistically significant decreases in flocculation relative to wild type AF-UD1, with AFΔRodG and AFΔRodA-G failing to flocculate plastics entirely (Fig. 5A). Rod A and RodG likely play an integral role in microplastics flocculation due to the observed significant decreases. Importantly, the inhibition of microplastics flocculation by the total knockout strain (ΔRodA-G) indicates that hydrophobins are necessary for microplastics flocculation.

Figure 5: Hydrophobins are responsible for microplastics flocculation by AF-UD1.

(A) Microplastics flocculation by Aspergillus strains with hydrophobin genes knocked out from the genome. Aspergillus fumigatus strains with each hydrophobin gene knocked out have reduced flocculation ability, indicating the importance of hydrophobins in microplastics recovery. Statistical significance testing was performed using a 2-tailed, homoscedastic t-test. * indicates p-value < 0.1, ** indicates p-value <0.05. (B) Recovery of ~40 μm green fluorescent LDPE beads by pure RodA (right) relative to a water control (left). After the addition of RodA, microplastic particles stick to the hydrophobin and are pulled from the air-water interface (dashed) to the bottom of the tube where the RodA sits.

Hydrophobin knockout data suggest that hydrophobins are necessary for microplastics flocculation, but use of purified hydrophobin in isolation of the host is necessary to confirm their propensity for microplastic flocculation. We thus expressed RodA, reported as the hydrophobin in A. fumigatus responsible for cell wall surface hydrophobicity⁴³, in the heterologous host Yarrowia lipolytica and subsequently purified the protein (purity confirmed via SDS-PAGE, Supplementary Fig. 3) to directly assess the microplastics flocculation ability by RodA in the absence of the host organism. Pure RodA flocculated microplastics from solution (Fig. 5B). Initially, green LDPE (GLDPE) particles sit atop the water due to their lower density than water and purified RodA resides in the bottom of the tube due to a density greater than that of water. GLDPE particles aggregate in that area after shaking, becoming entrapped in the purified hydrophobin, demonstrated by green particles near the bottom of the tube, below the air-liquid interface. Microplastics recovery by pure RodA and supporting hydrophobin knockout data demonstrate that hydrophobins are essential for microplastics flocculation by Aspergilli.

Discussion:

Understanding microplastics colonization is essential to mitigate toxic xenobiotic leaching, greenhouse gas emissions, biodiversity effects caused by microplastics accumulation, and animal health defects caused by foreign and/or pathogenic microorganisms entering the food chain through microplastic-bound biofilms. In this study, we evaluate Aspergillus as a model genus of fungi to better understand fungal interactions with microplastics due to their ubiquity across soil and marine microbial communities^46,47 and potential for microplastics remediation through strong binding interactions. We showed that Aspergilli are efficient microplastics binding and recovery agents. We verified that the microplastics binding phenotype is conserved across the genus by demonstrating microplastics recovery with a subset of Aspergilli across 5 sub-genera: Fumigati, Nidulantes, Wentii, Terrei, and Nigri⁷⁷. These strains each contain 5–8 hydrophobins, each between 10–20 kDa in size and containing 8–10 disulfide bond-forming cysteines that lead to amphiphilicity⁴⁸. This conservation in hydrophobin abundance and size suggests that there are analogous hydrophobins in each species that contribute to the conserved surface hydrophobicity and microplastics binding phenotype. Our data also suggest that microplastics binding and capture occurs independent of plastic type and size, capturing all single and mixed plastics with nearly 100% recovery, consistent with previous studies detailing 100% recovery of 200 nm PS and 5μm PMMA by Aspergillus niger⁴⁵. We also demonstrate flocculation of weathered PVC particles, implying that microplastic flocculation is also effective on plastics representative of those in the environment. This plastic type-independent microplastics binding is consistent with reports that biofilm taxonomic composition does not vary with plastic type⁷⁸. Rather, environmental factors such as temperature, pH, and salinity drive microplastics binding interactions and dictate which taxa persist in microplastic microbial communities³⁴. Microplastics binding interactions require hydrophobic interactions between the microorganism and hydrophobic microplastic surface^79,80, meaning that any mechanism altering surface hydrophobicity would be agnostic to plastic type. However, we note that this reliance on hydrophobic interactions means that hydrophobins do not exclusively bind to plastics. A. fumigatus flocculates cellulose acetate similarly to microplastics, demonstrating that hydrophobin binding is not specific to microplastics (Supplementary Fig. 2). Thus, competitive binding of alternative should be considered when deploying hydrophobins for microplastic capture in practice.

Microplastic-bound biofilms have historically been studied by identifying the dominant bacterial members present^27,33,35,78, overlooking the role of the biomolecules that drive colonization and the contributions of fungal biofilm species^36,44. We demonstrate that hydrophobins are important to microplastics binding by showing a decrease in microplastics flocculation upon knockout of each hydrophobin gene out of the genome. More importantly, we showed that pure RodA, the most abundant A. fumigatus hydrophobin, flocculates microplastics in isolation from the host organism, demonstrating that pure hydrophobin proteins bind directly to microplastic particles and flocculate them. The discovery of this relationship between hydrophobins and microplastics in Aspergillus biofilms allows for advances in microplastics remediation, pathogenicity, and biodeconstruction efforts by providing the physiological context in which Aspergilli bind to biofilms. Understanding how these microplastic-fungi interactions form can allow for new hydrophobin-based technologies to capture microplastics, alleviating toxic leaching¹⁰ and greenhouse gas emissions¹¹ caused by their accumulation. Moreover, technologies can arise to reverse binding, thereby mitigating diseases throughout the food chain caused by microplastic mediated pathogen transport such as in coral reefs³², to fish³¹, and to humans^24,29,43. Moreover, the mitigation of fungal biofilm formation on microplastics can decrease threats to biodiversity from the travel of invasive species to new ecosystems via microplastics transport^26,81.

Due to their inherent hydrophobicity, hydrophobins have been shown to interact with plastic substrates, namely in the context of biological deconstruction of hydrolyzable plastics by fungal enzymes. For example, Aspergillus oryzae expresses hydrophobin RolA that recruits a cutinase that hydrolyses polybutylene-succinate-coadipate^51,82. Moreover, RolA incubation with PET substrate prior to treatment with a PETase improved PET deconstruction from 17% to 26% weight loss⁸³. While these studies have focused on fungal enzymes and natural complexing with hydrophobins, we detail efficient microbial microplastics binding via hydrophobins. Our work highlights one strategy by which microbes colonize suspended microplastic particles, which would be the first step of biological deconstruction. Engineering this process may lead to more efficient/rapid plastics bioconstruction. For example, PET deconstruction by a PETase was improved 328-fold relative to pure PETase and 9-fold relative to surface displayed PETase by co-displaying the PETase with HFBI, a hydrophobin from T. reesei, on heterologous host Pichia pastoris⁵⁰. The work presented in this manuscript can build on such studies by providing a library of hydrophobins for plastics binding from Aspergilli that can be used to similarly enhance biological (micro)plastic deconstruction efforts.

Biologically compatible (micro)plastics binding technologies further enhance bioremediation efforts by providing a microplastics capture mechanism that interfaces with (bio)deconstruction efforts. Existing microplastics capture technologies used in wastewater treatment plants (WWTPs), an extremely large source of microplastics⁸⁴, such as ultrafiltration, reverse osmosis, and chemical flocculation fail to provide a mechanism through which plastics can be deconstructed. Without conversion of (micro)plastics waste into non-plastic, non-toxic products, the (micro)plastics waste crisis remains unresolved. Importantly, plastics wastes need to be upcycled into consumer products or recycled into plastics of equal value to the recycled waste to meet economic demands required to compete with plastics production from petrochemical refining⁸⁵. Hydrophobins can thus be used to capture microplastics from aqueous environments with nearly 100% efficiency and can be utilized concurrently to engineer improved biological plastics deconstruction technologies that may be able to circumvent economic barriers with conventional mechanical or chemical plastics recycling⁸⁵. Continued research on the interactions between fungal systems and microplastics is essential to identify hydrophobins capable of increased plastic binding that can ultimately be used to develop biological plastics deconstruction technologies that can mitigate the (micro)plastics waste accumulation crisis.

Environmental Implications:

The exponentially increasing accumulation of plastics in the environment leads to the direct release of toxic compounds such as xenobiotics and increased emission of greenhouse gases as a result of their environmental degradation^10,11. Environmental plastics degradation generates microplastic particles that exacerbate these negative effects by leading to the distribution of microplastics particles to waterways and soil throughout the globe^12,15. This widespread transport leads to microplastics ingestion by animals that is linked to negative health consequences such as hepatic lipid disorder and bile acids metabolism disorder in mice, inflammation and metabolism disruption in zebrafish^13,86, among other animal health impacts, directly contributing to food chain disruption by increasing mortality rates and via trophic transfer can affect human health ^3,69.. Moreover, microplastics serve as a vector for transport of microorganisms^26,80, meaning that microplastic ingestion can also lead to the delivery of pathogenic microbes to animal species through ingestion, leading to the spread of disease throughout animal populations ¹³. Microplastic mediated microorganism transport also disrupts soil and marine microenvironments through the delivery of foreign microbes²⁶ that can be invasive²⁶ directly changing microbial communities. These changes in microbial community composition can lead to reduced growth and nutrient uptake by plants⁸⁷, thereby percolating changes through the food web again.

In this work, we demonstrate that hydrophobins can be used as a sustainable microplastics capture and removal agent. However, the direct use of pathogenic fungal strains for microplastics capture can lead to undesired environmental consequence such as the disruption of microbial community biodiversity from the introduction of a non-native strain. Moreover, pathogenic strains bound to unrecovered microplastic particles could proliferate in the environment, leading to further biodiversity and human health issues from their spread. Alternatively, non-pathogenic Aspergillus species can be used for microplastics capture, but fungal community sequencing should be carried out to ensure the species is already present in the microbial community in the area of remediation interest to limit biodiversity disruption. Additionally, fungal-microplastic floccs contain water, which could require processing to dry them, dependent on downstream needs. To limit environmental impacts and process economics concerns from drying, pure hydrophobins can be deployed in waterways or in water treatment facilities to capture microplastics due to their inert nature.. We have shown that hydrophobins can be used to remove microplastics from real-world environments such as water treatment plants or the ocean by confirming that microplastics capture is agnostic to both particle size and chemical identity and by demonstrating microplastics capture at concentrations ranging from 20–40 mg/mL, greater than those in marine environments^72,73. While these data show the potential for microplastics remediation using pure hydrophobins, this is likely more costly than using an Aspergillus strain. Heterologous expression of a candidate hydrophobin in hosts such as Y. lipolytica or P. pastoris can limit downstream purification costs through the use of efficient secretion systems (cite). While we leave economic evaluation of hydrophobin based microplastic remediation technologies to future studies, this work shows promise for hydrophobin based microplastics removal. This technology can lead to the alleviation of toxicity concerns from microplastics degradation and subsequently mitigating animal health and biodiversity consequences caused by their ingestion

Synopsis:

Aspergillus bind microplastics via hydrophobin proteins that may be used as a capture agent to mitigate microplastics pollution in oceans and waterways.