Genome mining and screening for plastic-degrading potential in marine bacteria

TL;DR

This study identifies marine bacteria from plastic fishing nets that can degrade plastics and explores their genetic potential for biodegradation.

Contribution

A new workflow for screening and genome mining of plastic-degrading marine bacteria from submerged plastic nets is introduced.

Findings

Eleven bacterial strains showed enhanced growth on plastic polymers and eight exhibited esterase/lipase activity.

Genes linked to plastic degradation were predominantly found in Actinomycetes strains.

Genome mining revealed enzymes potentially capable of degrading polyethylene terephthalate, low-density PE, and nylon.

Abstract

Marine plastic litter, including microplastics, has a profound impact on the ocean and its wildlife, and strategies to remove/eliminate it are needed. Microbial biodegradation, particularly by bacteria, offers a potential solution, where a link between hydrocarbon and plastic-degradation has been hypothesized. This study screened the plastic-degrading potential of 18 bacterial strains isolated from 1-month-old biofilms developed in three submerged plastic fishing nets (braided polyethylene (PE), braided nylon, thin nylon). In addition, three highly efficient hydrocarbon-degrading strains were also tested. Strains were cultivated on solid minimal media with fishing net small pieces (new/unused nets) added as a carbon source for 1 month, followed by tributyrin-agar assays to assess esterase/lipase activity. Eleven bacteria exhibited enhanced growth with net polymers, mainly from the genera Sulfitobacter, Rhodococcus, Bacillus, and Pseudomonas, and eight of which bacteria also demonstrated esterase/lipase activity. Then, genes encoding hydrocarbon or plastic-degrading enzymes (alkB and almA homologs, PETase-like enzymes) were screened by PCR in the 21 mentioned bacteria and in ca.100 other strains found in submerged nets biofilm. Amplification of the investigated genes was predominantly observed in Actinomycetes strains. Genome mining of six promising strains revealed hits with enzymes linked to degradation of synthetic polymers like polyethylene terephthalate, low-density PE and nylon. The workflow developed here enabled the selection of marine bacteria with plastic-degrading potential, sourced from biofilms of submerged plastic fishing nets and hydrocarbon-enriched environments.

• A comprehensive lab workflow was developed to assess plastic-degrading potential.

• Genome mining in Rhodococcus and Pseudomonas strains revealed plastic-degrading enzymes.

• Hydrocarbon-degrading bacteria could hold plastic-degrading capabilities.

The online version contains supplementary material available at 10.1007/s00253-026-13767-4.

Figures

Growth on MM agar with plastic nets (net A, braided PE; net B, braided nylon; and net C, thin nylon) as carbon source added: a) Example of the circular biomass growth using the discs approach, in MM agar + net B after 22 days growth of the strain NLBR_22; (b) Bar graph with the difference (in mm of diameter) between the growth in the media with nets and the control medium (MM Agar), after 7, 15, or 30 days of incubation (T7, T15, and T30), for the strains that presented better growth in the media with nets

Growth on MM agar with plastic nets (net A, braided PE; net B, braided nylon; and net C, thin nylon) as sole carbon source: (a) Example of the biomass growth using the wells approach, for the strain NLAR_29 in the plates MM Agar, and the media supplemented with net A, net B, and net C, after 30 days of growth; (b and c) 15-day biomass growth using the wells approach, in the media supplemented with net A and net B, for the strains NLCR_1 and NLCR_8, respectively

Maximum likelihood phylogenetic tree with sequences of potential PETase-like enzymes from the *Streptomyces* strains, their Swiss-Prot nearest neighbors and additional ones: *Streptomyces* sp. SM14 alpha/beta hydrolase ([DAC80635.1](DAC80635.1)) and the Leaf-branch compost cutinase (LCC) ([G9BY57](G9BY57).1) (Sulaiman et al. [2012](#CR72)). The sequence of the strain NLAR_46 was excluded because of its smaller size (152 aa), while NLBR_30 was excluded due to its distance from other PETases/Cutinases. The maximum likelihood phylogenetic tree was constructed in MEGA 11 using amino acid sequences

Overall genome mining results for the 1.7 L (*Pseudomonas khazarica*), CPN2 and CPN3 (*Rhodococcus erythropolis*), NLAR_29 (*Bacillus pumilus*), NLCR_1 (*Rhodococcus qingshengii*) and NLCR_8 (*Rhodococcus fascians*): Barplot indicates the number of protein hits above 40% identity, after using Blast protein–protein analysis, with enzymes linked to the degradation of either synthetic or bio-based/“biodegradable” plastics, displaying on the right side the presence (positive hit) or absence (no hit) of hits linked with the degradation of the synthetic plastics PET (polyethylene terephthalate); LDP

Full text

Introduction

The consumption-driven society that we live in has led to the expansion of the marine litter problematic in our ocean. Primarily dominated by plastic materials—approximately 80%—it is accepted that most of the marine litter comes from land activities (González-Fernández et al. 2021; UNEP 2016). Nevertheless, sea-based sources contribute as well to this issue, including abandoned lost or otherwise discarded fishing gear (ALDFG) (Lebreton et al. 2018), a familiar reality to Portugal as well (Prata et al. 2020).

Given the significant environmental impacts of marine plastic litter, such as endangering marine wildlife, adsorbing contaminants and breaking down into microplastics (Almeida et al. 2024; Amelia et al. 2021; León et al. 2019; Yu et al. 2019), nations are increasingly committed to tackling marine litter. For that, it is required, including the improvement of waste management practices, fostering environmental education, reinforcing international collaboration and legislation to reduce single-use plastics, from both public and private sectors, as well as from local to global action (Löhr et al. 2017). An example of such relevant effort is the ongoing negotiation of the Global Plastics Treaty of the United Nations Environment Programme (UNEP), for which they intend to establish a legally binding instrument to tackle plastic pollution, including in marine environments (UNEP 2023). While tackling the marine litter issue “upstream” is necessary, looking into “downstream” solutions, such as improving degradation and recycling processes of the collected plastic litter, is paramount. As a result, microbial degradation of plastic polymers, namely of synthetic plastics, is gaining attention in the research community (Kour et al. 2023; Ru et al. 2020), and has been pointed out as an eco-friendly and promising alternative to tackle environmental plastic pollution (Silva et al. 2018; Yuan et al. 2020).

Microorganisms from the genera Pseudomonas, Bacillus, Sphingomonas, Rhodococcus, and Streptomyces have been continuously reported for plastic degradation (Ali et al. 2021; Atanasova et al. 2021; Auta et al. 2018; Mor and Sivan 2008), either in soil or marine environments. In marine ecosystems, plastic-degrading groups are known to inhabit the plastisphere of collected marine litter items (Roager and Sonnenschein 2019), along with the presence of recognized hydrocarbon-degrading microorganisms in those items (Delacuvellerie et al. 2019; Vaksmaa et al. 2021; Yakimov et al. 2022). But assessment of the potential of other microorganisms for degradation is still needed.

Certain microorganisms have the ability to produce enzymes capable of breaking down plastic polymers. A notorious example is the work of Yoshida et al. (2016), which unveiled Ideonella sakaiensis 201-F6 as the first bacteria known to degrade polyethylene terephthalate (PET) polymer through the production of potent hydrolytic enzymes: PET hydrolase or PETase, that breaks the initial polymer and MHET (mono-(2-hydroxyethyl) terephthalate) hydrolase or MHETase, that breaks the resulting product into terephthalate (TPA) and ethylene glycol (EG). More enzymes associated with plastic degradation, namely esterases, cutinases, hydroxylases, or laccases are commonly enumerated, also including enzymes involved in the degradation of hydrocarbons such as alkane monooxygenases (Herrero Acero et al. 2011; Santo et al. 2013; Yoon et al. 2012). As new information on plastic-degrading enzymes and/or microorganisms is uncovered, curated databases have been recently established such as Plastic DB (Gambarini et al. 2022) and PAZY (Buchholz et al. 2022). They can be used to further mine new plastic-degrading enzymes in bacterial genomes or even in metagenomes (Kim et al. 2022), leading to the discovery of new strains with potential plastic-degradation capacity.

It is widely accepted that floating marine litter items undergo weathering and deterioration by abiotic factors (such as UV light and wave currents), rendering them more accessible for microbial action (Andrady 2011; Fotopoulou and Karapanagioti 2017; Gewert et al. 2015). Still, a lot remains to be answered regarding the degradation of plastics in the marine environment (Van Sebille et al. 2015). The recognized stages of plastic biodegradation typically involve: (i) the attachment of bacteria to the polymer’s surface and the subsequent formation of biofilms; (ii) the biodeterioration of the plastic material; (iii) the enzymatic degradation facilitated by bacteria; and (iv) a final stage involving mineralization or assimilation of the degradation by-products (Ali et al. 2021; Patrício Silva 2022). This general process is commonly accepted, but in fact, little is known about the precise steps involved in the microbial enzymatic catalysis of various plastic polymers. The KEGG database (Kanehisa and Goto 2000) nowadays has only pathway models associated with the degradation of three different plastics and/or its monomers: (i) PET, which is the only one that holds the complete catalysis of the primary polymer until its monomers; (ii) nylon 6-oligomers; and (iii) the monomer styrene. Considering the variety of plastic polymers commonly used, gathering knowledge on the biodegradation of other plastic polymers is needed.

Microbial biofilms attached to marine plastic litter could serve as a reservoir of novel bacteria or enzymes, with enhanced capabilities for plastic biodegradation (Roager and Sonnenschein 2019). Hence, in this work we aimed to develop a workflow capable of identifying bacterial strains with plastic-degrading potential, i.e., with appropriate genomic (i.e., via genome mining or PCR amplification of candidate genes) and phenotypic (i.e., bacterial growth in the presence of plastic polymers) features compatible with plastic degradation. The identified strains could then be suitable candidates for future in vitro testing and validation.

The marine bacteria screened in this study were isolated from biofilms developed on submerged plastic fishing nets. In addition, we also included three highly efficient hydrocarbon-degrading strains, considering that most plastic polymers are classified as petroleum-based products.

Materials and methods

Library of bacterial strains used and overall workflow of laboratory screening assays

In a previous work (Perdigão et al. 2025), a total of 122 bacterial strains from 46 different genera were isolated from biofilm communities growing on 3 different plastic fishing nets (braided polyethylene (PE), braided nylon, and thin nylon) that have been submerged for 1 month in marine waters. In addition, three marine bacterial strains from the genera Rhodococcus and Pseudomonas, known for their ability to degrade petroleum hydrocarbons, were added to this library. When applied as a consortium in laboratory experiments with natural seawater, these strains demonstrated effectiveness in the degradation of crude oil and fuel (Perdigão et al. 2021, 2024). The combined phylogenetic information and NCBI accession numbers of the bacterial strains addressed in this work can be found in Table S1.

Growth in solid minimal medium with plastic nets as carbon source

To assess the ability to grow on plastic as the intended carbon source added, 18 bacterial strains selected from net biofilms and the 3 hydrocarbon-degrading strains were grown in solid minimal medium supplemented with each net (net A, braided PE; net B, braided nylon; net C, thin nylon) for 1 month. Table 1 comprises the codes, phylogenetic classification, and culture media in which the strains tested were isolated from. Table 1. Phylogenetic classification and culture media from which bacteria were originally isolated (marine agar, MA; plate count agar, PCA; and Bushnell Haas, BH), for the 21 bacterial strains grown in solid minimal medium with plastic added as carbon sourceCodeClosest identificationCulture medium > 1% abundance in net biofilms communitiesNLCR_33Erythrobacteraceae familyMANLBR_19Sulfitobacter sp.MANLCR_36Sulfitobacter sp.MANLCR_45Sulfitobacter geojensisMANLBR_20Pseudophaeobacter arcticusMAPotential plastic-degrading bacteria (based on their taxonomy)NLAR_30Exiguobacterium oxidotoleransPCANLCR_9Pseudomonas sp.PCANLAR_37Rhodococcus sp.PCANLCR_8Rhodococcus sp.PCANLCR_1Rhodococcus sp.PCANLAR_9Bacillus sp.MANLAR_29Bacillus sp.PCANLBR_22Bacillus sp.PCANLBR_26Bacillus sp.PCANLBR_39Bacillus sp.PCANLCR_22Priestia sp.PCANLCR_23Peribacillus sp.PCANLCR_29Bacillus sp.PCAHydrocarbon-degrading bacteriaCPN2Rhodococcus erythropolisPCACPN3Rhodococcus erythropolisPCA1.7 LPseudomonas sp.PCA

For this screening, the 18 bacterial strains were selected based on their phylogenetic classification, taking into account their genera’s abundance in the biofilm communities (relative abundance > 1%: Erythrobacter (Erythrobacteraceae), Sulfitobacter, Pseudophaeobacter) (Table S2) and the genera previously described in the literature as having potential for plastic-degradation (Exiguobacterium, Pseudomonas, Rhodococcus, and Bacillus (Auta et al. 2018; Chauhan et al. 2018; Devi et al. 2019)).

First, each net (new, unused) was untangled and cut into small pieces or fibers (2–5 mm) with the help of sterile tweezers and scissors, weighed, and supplemented individually to the culture flasks at 1% (w/v). The solid minimal salt medium (MM), an oligotrophic medium, was prepared as follows:(per liter of deionized water): 2.7 g of Na_2_HPO_4_·2H_2_O; 1.4 g of KH_2_PO_4_; 0.5 g of (NH_4_)2SO_4_; 0.2 g of MgSO_4_·7H_2_O; 15 g of agar; and 10 mL of a trace elements solution (per liter: 12 g of Na_2_EDTA·2H_2_O + 2 g of NaOH + 0.4 g of MnSO_4_·4H_2_O + 0.4 g of ZnSO_4_·7H_2_O, 0.5 mL of H_2_SO_4_, 10 g Na_2_SO_4_, 0.1 g of Na_2_MoO_4_·2H_2_O + 2 g of FeSO_4_·7H_2_O + 0.1 g of CuSO_4_·5H_2_O and 1 g of CaCl_2_).

Aiming to break furthermore the pieces of nets, the media were subjected to ultrasonic pulses, at 50% amplitude in a regime of 30″/10″ (pulse/break) for a total of 1′ 30″, in the SONOPULS Ultrasonic homogenizer (Bandelin Electronic GmbH & Co; GM 2070.2, frequency 20 kHz), followed by 30 min in an ultrasonic cleaning bath (VWR, model USC100T, frequency: 45 kHz) (inspired from the methodology of Charnock (2021) and Divyalakshmi and Subhashini (2016)). The media were sterilized by autoclave (20 min at 121 °C) and plated in 90 mm petri dishes (also made of plastic) under aseptic conditions inside a laminar flow chamber. Solid minimal salt medium with no addition of nets or other carbon sources was also prepared to be used as a control (MM Agar).

Prior to the assay, the selected 21 strains were unfrozen from a cryopreserved stock and cultivated for 3–4 days at 28 °C in their respective growth media (marine agar—MA, plate count agar—PCA, and Bushnell Haas, BH) by the streaking method.

For the assay, biomass was collected from the pure isolated strains, homogenized in sterile saline solution (0.85% v/v), and the optical density (OD) of the inoculum adjusted to 0.1 (OD measured at 600 nm).

Two approaches were employed in this screening: discs (6 mm diameter) added to the agar surface of each media, and wells (~6 mm diameter) cut within the media with a sterile well-cutter. No more than 6 discs or wells were made on each plate. The bacterial inoculum was added to the discs (15 µL) and wells (20 µL), and the plates incubated at 28 °C for 1 month. Results were observed after 7, 15, and 30 days, measuring the diameter of growth above the discs and registering the growth around the wells. It is important to note that in these culture-dependent assays, the intended carbon source added was the plastic nets; however, agar (15 g/L) and EDTA (0.12 g/L) were also carbon sources present in the mineral salt medium, both in the control MM Agar (media without plastic) and MM supplemented with nets. In addition, following a positive PCR amplification of PETase homologs in four Streptomyces strains using the culture-independent approach, their biomass growth was also evaluated in discs and wells.

Screening for lipolytic activity—tributyrin agar assays

The previously mentioned 21 strains (Table 1) were screened for their lipolytic activity, associated with the action of esterases/lipases, by a Tributyrin Agar (TBA) assay adapted from the works of Charnock (2021), Molitor et al. (2020), and Carrazco-Palafox et al. (2018). Since the base methodologies differed, an initial optimization of the method was made using Bacillus subtilis ATCC 6633 and Staphylococcus aureus ATCC 29213, two strains with lipolytic activity that were used as positive controls in these assays. The composition of TBA was as follows:(for 1 L of deionized water): 3 g of yeast extract, 5 g of peptone, 15 g of bacteriological agar, and supplemented with tributyrin 1% (v/v) (PanReac AppliChem GmbH, ITW Reagents). One of the parameters optimized was the supplementation of TBA with Ca^2+^ (as CaCl_2_ at 2.5 mM) and Mg^2+^ (as MgSO_4_ at 5 mM), which had been reported to improve halo visualization (indicative of lipolytic activity). In case the bacteria would not grow in TBA medium, the media in which the bacteria were first isolated from were also tested (MA, PCA, and BH). Different temperature incubations (28 and 37 ºC) were also tested since 28 °C is the one preferred by the selected strains, but 37 °C is the optimum growth temperature for the strains used as positive controls. Results were seen after 24, 48, 72, and 96 h.

Overall, the method optimization showed that TBA supplemented with Ca^2^⁺ and Mg^2^⁺ enhances halo visualization within 24 h. B. subtilis ATCC 6633 and S. aureus ATCC 29213 serve as effective positive controls, also forming halos at 28 °C, though smaller than at 37 °C. Since not all strains grow at 37 °C, 28 °C was selected as more suitable. Most strains grow on TBA medium, but alternative media with tributyrin can be used when needed. Tributyrin (1% v/v) can be added before autoclaving, with media prepared in half-filled flasks and agitated at high speed to ensure proper emulsification.

Based on the method optimization, the assay was conducted with the TBA medium (TBA, TBA + Ca^2+^ and TBA + Mg^2+^), at both 28 °C and 37 °C for 48 h. Since the controls were also valid at 28 °C (although better seen at 37 °C), the selected strains grew in TBA media. The biomass of each strain was cultivated in the media by a single line streaking technique and incubated at 28 °C and 37 °C for 48 h. The strains Bacillus subtilis ATCC 6633 and Staphylococcus aureus ATCC 29213 were used as positive controls, while Escherichia coli ATCC 25922 was used as a negative control. Positive results were indicated by the formation of a clear halo around the biomass streaking line.

PCR detection of genes encoding potential plastic-degrading enzymes

Given that certain enzymes, such as hydrolases and cutinases, have been reported in the degradation of plastic polymers, together with hydrocarbon-degrading enzymes, in this study direct amplification of a selection of genes, by PCR, was employed to all 125 bacterial strains in the library (122 strains from nets biofilms plus 3 hydrocarbon-degrading strains). The DNA of all strains was extracted using the E.Z.N.A.® Bacterial DNA Kit (Omega Bio-Tek, GA, United States).

The high phylogenetic diversity in the library challenged the selection of the genes of interest and adaptation of PCR protocols. However, based on the works of Charnock (2021), Kawai et al. (2014), Perez-de-Mora et al. (2010), Wang and Shao (2012) and Kohno et al. (2002), the following enzymes (and encoding genes) were chosen to detect directly by PCR: Alkane monooxygenases (alkB1, alkB homologs), Flavin-binding long chain–monooxygenase (almA), cutinase-type polyesterase (cut190) and genes encoding PETase-like enzymes. Table 2 contains information of the primers used, expected amplicon sizes and references of the corresponding papers. Table 2. Primers used in the present study for the amplification of the desired genes, encoding the potential plastic-degrading enzymesEnzymeGeneExpected amplicon sizePrimersReferencesAlkane 1-monooxygenasealkB1330 bpalkB1_F 5′-TCG AGCA CAT CCG CGG CCA CCA −3′alkB1_R 5′-CCG TAG TGC TCG ACG TAG TT −3′(Kohno et al. 2002)Alkane monooxygenasealkB homologs550 bpRT-alkB_F 5′ AAY ACI GCI CAY GAR CTI GGI CAY AA 3′RT-alkB_R 5′ GCR TGR TGR TCI GAR TGI CGY TG 3(Perez-de-Mora et al. 2010)Cutinase-type polyesteraseCut190915 bpCut-F 5′ GTG CGA ATT CGC AGG CAG GCG G 3′Cut-R 5′ CCG CAC GTG GTA CGT GCG GTG C 3′(Kawai et al. 2014)PET-hydrolyzing enzymesISF6_4831480 bpPET_F 5′—GTC ATC ACC ATC GAC ACC A 3′PET_R 5′—GTA GCG SGT GTC GTT GTC—3′(Charnock 2021)Flavin-binding LC—monooxygenasealmA1100 bpAlmAdf 5′-GGN GGN ACN TGG GAY CTN TT-3′AlmAdr 5′-ATR TCN GCY TTN AGN GTC C-3′(Wang and Shao 2012)

Table S3 summarizes the preparation of PCR reaction mixtures, final primer concentrations, and cycles applied for the amplification of all desired genes. The PCR products were afterwards checked for an amplicon of expected size by 1.5% agarose gel electrophoresis with SYBR Safe (Thermo Fisher Scientific, MA, United States). Positive PCR products were sent for Sanger sequencing at the Genomics i3S Scientific Platform (Porto, Portugal). The translated sequences of the PET amplified products were blasted (BLASTP) against the SwissProt database in NCBI, and the closest 5 sequences were used to construct a phylogenetic tree.

Draft genome sequencing and mining for potential plastic-degrading enzymes

Based on the results of the culture-dependent (higher growth in solid minimal medium supplemented with nets compared to the control (MM Agar), using the discs and wells approach and tributyrin agar assays (lipolytic activity)) and culture-independent screening assays (PCR detection of genes of alkB homologs and almA holomologs), 6 strains were chosen for whole genome sequencing: CPN2 (Rhodococcus erythropolis); CPN3 (Rhodococcus erythropolis); 1.7 L (Pseudomonas sp.); NLCR_1 (Rhodococcus sp.); NLCR_8 (Rhodococcus sp.); and NLAR_29 (Bacillus sp.). The strains were grown in their respective culture media (Table 1), and afterwards, biomass was collected and preserved in 100 µL TE buffer at −20 °C until further DNA extraction.

Total genomic DNA was extracted with the commercial kit E.Z.N.A.® Bacterial DNA Kit (Omega, bio-tek) following the supplier’s protocol, except on the final elution step, where the DNA was eluted in DNAse/RNAse free water to prevent the presence of EDTA, which inhibits the company’s NGS library preparation. Following up the DNA quantification with the kit Quant-it dsDNA Assay in the Qubit fluorometer (Invitrogen), extracted DNA samples were sent for Whole Genome Sequencing (in the Illumina sequencing platform using 2 × 250 bp paired end reads) at MicrobesNG (http://www.microbesng.com) (Birmingham, UK). The company service provided the following data: assemblies, trimmed reads, and raw sequence data.

The reads were trimmed using Trimmomatic (Bolger et al. 2014) and the quality was assessed with the software: Samtools (Li et al. 2009) and BWA mem (Li and Durbin 2009). A summary of the Illumina data after trimming is displayed in Table S4. De novo assembly of the reads was made using SPAdes, and assembly metrics were calculated using QUAST (Table S5). The company also identified the closest available reference genome using Kraken (Wood and Salzberg 2014), and automated annotations were performed with Prokka. Nevertheless, in this work we only used the assembly files for further downstream processing.

Initially, assembly files of each bacterial strain were uploaded to the Kbase platform (Arkin et al. 2018) and their quality assessed by CheckM (v1.0.18) to confirm their suitability for further downstream analyses. The quality results are compiled in Table S6. In addition, each assembled genome was annotated using the RASTtk (v1.073) database and mined against a curated database (stated below) of potential plastic-degrading enzymes by the BLASTp analysis (v2.13.0. +), to overcome any bias of using a general annotated file.

This database was created by compiling two datasets of plastic-degrading enzymes (based on their EC number): known metabolic pathways for plastics or their monomers from KEGG (www.kegg.jp, PET, Nylon, and styrene pathways) and PAZY (Buchholz); www.pazy.eu) which lists exclusively biochemically characterized plastic-active enzymes. The enzymes from the upper KEGG metabolic pathways of PET, Nylon 6 oligomers, and styrene were used (Table S7). Data were retrieved from the databases in September 2023. A list of all features was mapped on UniProt (www.uniprot.org/) to compile a unified database for the genome mining efforts. In contrast to the hydrocarbon styrene, short-chain alkane degradation pathways were excluded from the genome mining since they would introduce a larger set of genes involved in general hydrocarbon metabolism, which could likely overwhelm the dataset and mask the more specific enzyme groups relevant for plastic degradation.

In parallel, genome mining in the Plastic DB (https://plasticdb.org/) database was also performed. For both approaches (Kbase with curated database and PlasticDB database), a threshold of protein similarity > 40% was applied, a value above which is sometimes assumed of similar protein functionality (Pearson 2013).

The curated database (in FASTA format) is provided in the supplementary file “Curated database_plastic degrading enzymes.fasta”. The genomes of the 6 selected strains were deposited at GenBank (NCBI) under the accession numbers SAMN48103133 (Rhodococcus erythropolis CPN2), SAMN48103134 (Rhodococcus erythropolis CPN3), SAMN48103135 (Pseudomonas khazarica 1.7L), SAMN48103136 (Rhodococcus qingshengii NLCR_1), SAMN48103137 (Rhodococcus fascians NLCR_8), SAMN48103138 (Bacillus pumilus NLAR_29). These six selected strains are cryopreserved in the CIIMAR Microbial Culture Collection (CM2C) and will be made available upon request (cm2c@ciimar.up.pt).

Results

Growth in solid minimal medium with plastic nets as carbon source and lipolytic activity

After growing the selected 21 strains in minimal media supplemented with pieces of plastic fishing nets, 11 strains grew better in the presence of plastic when compared to the control medium composed of only mineral salt medium and agar (MM Agar), combining the results of both approaches: discs and wells (Table 3). Table 3. Summary of the results of the culture dependent and culture independent screening assays performed for the initial group of 21 strains. Agar-based assaysPCR-based assaysStrain codeClosest identificationBest growth in medium with nets (discs)Best growth in medium with nets (wells)Lypolitic activityalkB homologsalmA homologs1.7 LPseudomonas* sp. + (net A, C)- + **** + **** + CPN2Rhodococcus erythropolis*** + (net A, B)- + **** + **** + CPN3Rhodococcus erythropolis*--** + **** + **** + NLAR_9Bacillus sp.-----NLAR_29Bacillus* sp.-** + ^a^ (net A, C) + --NLAR_30Exiguobacterium oxidotolerans-----NLAR_37Rhodococcus sp.- + (net B) + **** + -NLBR_19Sulfitobacter sp.-- + --NLBR_20Pseudophaeobacter arcticus-- + **** + -NLBR_22Bacillus sp.** + (net A, B, C)- + --NLBR_26Bacillus sp. + (net A)----NLBR_39Bacillus sp.-----NLCR_1Rhodococcus* sp.-** + (net A, B) + **** + **** + NLCR_8Rhodococcus* sp.-** + (net A, B) + **** + **** + NLCR_9Pseudomonas sp.-----NLCR_22Priestia sp.- + (net B) + --NLCR_23Peribacillus sp.-- + --NLCR_29Bacillus sp.-----NLCR_33Erythrobacteraceae family + ^a^ (net C)----NLCR_36Sulfitobacter sp.-----NLCR_45Sulfitobacter geojensis- + **^a^ (net A)---No detection of genes encoding PETases-like enzymes for this group of bacteria. * highlight the six strains chosen for genome sequencing. Net A, braided PE; net B, braided nylon; and net C, thin nylon+ Positive result- Negative result^a^Bacteria grew only in media supplemented with nets

Growth of the bacterial strains was evident above the agar and disc, in a circular shape (Fig. 1a), and subsequently its diameter was annotated. Following this approach, 5 strains grew better in media with nets (1.7 L, CPN 2, NLBR_22, NLBR_26, NLCR_33). Figure 1b compiles the difference (in mm of diameter) between the growth upon media with nets and MM Agar (control), after 30 days of incubation. Table S8 displays the total biomass growth (in mm of biomass diameter) of all tested strains in the discs assay. The Erythrobacteraceae NLCR_33 was the only one that grew solely on a medium supplemented with nets, namely with net C (Thin nylon) after 30 days of incubation in discs. The Pseudomonas sp. 1.7 L preferred the medium with net C as well, growing up to 6 mm more compared to the control, but not for the other nylon net, net B. The Bacillus sp. NLBR_22 grew up to 3 mm more in diameter in all media with nets, compared to the control. The strains CPN2 and NLBR_26 (R. erythropolis and Bacillus sp., respectively) grew better on the media with net A (braided PE), reaching up to 6 mm more than MM Agar. Four of the tested strains (the P. arcticus NLBR_20 and the Sulfitobacter spp. NLBR_19, NLCR_36 and NLCR_45) could not grow onto the solid minimal media. On the contrary, the Bacillus sp. NLBR_39 grew all over the culture media, after 30 days, independently of the presence of nets.Fig. 1. Growth on MM agar with plastic nets (net A, braided PE; net B, braided nylon; and net C, thin nylon) as carbon source added: a) Example of the circular biomass growth using the discs approach, in MM agar + net B after 22 days growth of the strain NLBR_22; (b) Bar graph with the difference (in mm of diameter) between the growth in the media with nets and the control medium (MM Agar), after 7, 15, or 30 days of incubation (T7, T15, and T30), for the strains that presented better growth in the media with nets

As for the approach of adding bacterial inoculum into wells cut from the agar plates, biomass growth was seen around the wells or even attached to the adjacent pieces of nets (Fig. 2a). The strains that grew more on media with nets, compared to the control medium (MM Agar), following the wells approach, were: NLAR_29, NLCR_22, NLCR_1, NLCR_8, NLCR_45, NLAR_37. The slow-growing S. geojensis NLCR_45 strain only grew in net A after 30 days of incubation. The Rhodococcus sp. NLAR_37 and the Priestia sp. NLCR_22 grew better in net B after 15 days of incubation. The Bacillus sp. NLAR_29 grew in media supplemented with nets, particularly on net C (thin nylon) and net A (braided PE) (Fig. 2a). The Rhodococcus spp. NLCR_1 and NLCR_8 grew both around pieces of net A (HDPE) and around the well of net B (Fig. 2b, c), but this biofilm attachment was observed first in strain NLCR_8, after just 7 days of cultivation. For the additional Streptomyces strains tested, biomass growth was observed in both discs and wells, and the results are presented in Table S9 in the supplementary material. Overall, growth in discs was similar between the control medium (MM Agar) and media supplemented with nets. However, under the wells approach, some strains showed enhanced growth in the presence of nets. Strain NLAR_46 grew better in media supplemented with nets, particularly in net A, consistently across all incubation times (Fig. S1). Strain NLCR_57 also showed improved growth in the presence of nets, especially in net A and net C. In addition, strains NLAR_46 and NLCR_59 formed biomass around pieces of net C within the well (Fig. S2 a, c), while NLCR_46 also grew around pieces of net A (Fig. S2 b). In contrast, strain NLBR_30 did not show improved growth in media containing nets.Fig. 2. Growth on MM agar with plastic nets (net A, braided PE; net B, braided nylon; and net C, thin nylon) as sole carbon source: (a) Example of the biomass growth using the wells approach, for the strain NLAR_29 in the plates MM Agar, and the media supplemented with net A, net B, and net C, after 30 days of growth; (b and c) 15-day biomass growth using the wells approach, in the media supplemented with net A and net B, for the strains NLCR_1 and NLCR_8, respectively

Regarding the lipolytic activity within the group of 21 bacterial strains, the overall screening results are displayed in Table 3. Positive lipolytic activity was considered with the formation of a halo surrounding the biomass line streaking in TBA medium (normal and/or with the addition of Ca^2+^ and Mg^2+^). Briefly, a total of 12 strains, namely from the genera Rhodococcus (CPN2, CPN3, NLAR_37, NLCR_1, NLCR_8), Pseudomonas (1.7L), Bacillus (NLAR_29, NLBR_22), Sulfitobacter (NLBR_19), Pseudophaeobacter (NLBR_20), Priestia (NLCR_22), and Peribacillus (NLCR_23) demonstrated lipolytic activity.

PCR detection of genes encoding potential plastic-degrading enzymes

This section presents the results of PCR-based direct amplification of the selected genes across the bacterial library selected for this work.

The amplification of the genes alkB1 and cut190 led to the co-amplification of other sized (unspecific) products, following the program and primers used in this work, and so the results obtained for these 2 primer sets were not considered within the screening assays of this work. In contrast, the detection of the alkB and almA homolog genes and genes encoding PETase-like enzymes, using degenerated primers, amplified in most cases a single clear band in the expected fragment size.

For the initial group of 21 bacterial strains (listed in Table 3), both alkB and almA homologs were simultaneously detected in the three known hydrocarbon-degrading strains (1.7 L, CPN2, CPN3), as well as in two strains isolated from net C (NLCR_1 and NLCR_8). In addition, alkB homologs were detected in strains NLAR_37 and NLBR_20. However, since PETase-like genes were not detected in this group, we continued screening the rest of the bacterial library.

As a result, the bacterial strains in which at least one gene of interest was amplified are summarized in Table S10. An example of the resulting amplified products obtained for each gene of interest, by visualization in an electrophoresis gel, is shown in Fig. S3.

Overall, from all the strains tested in this work, alkB homolog genes were detected in a total of 15 strains, belonging to the genera of Acinetobacter, Pseudomonas,* Rhodococcus*, Pseudophaeobacter, and Dietzia. Regarding almA homologs, amplifications were also observed in Acinetobacter, Pseudomonas, and the Actinomycetes Rhodococcus and Oerskovia. Concerning genes encoding PETase-like enzymes, their detection was seen only in the four strains belonging to the Streptomyces genus.

To further explore the relationship between the detected PETase-like sequences and known PET-degrading enzymes, as well as their similarity to each other, a maximum likelihood phylogenetic tree was constructed using the translated sequences from two of the four Streptomyces strains (Fig. 3, Fig. S4). The sequence from strain NLAR_46 was excluded from the analysis due to its shorter length, which affected the quality of the alignment, while the sequence from strain NLBR_30 was excluded due to its distance from other PETases/Cutinases. The PETase-like sequences of the two Streptomyces strains NLCR_57 and NCR_59 clustered together and are closely related to the extracellular cutinase of Amycolatopsis mediterranei (Verma et al. 2011). Even so, the following hits on Swiss-Prot are related to other types of PET-degrading cutinases, from the species of Thermobifida alba (Hu et al. 2010), Thermobifida fusca (Hegde and Veeranki 2013; Lykidis et al. 2007), and Thermobifida cellulosilytica (Herrero Acero et al. 2011). Contrarily to these 2 strains, the sequence from NLBR_30 presented no hit against the Swiss-Prot database (above 40% identity percentage), resulting in a separated branch compared to the input sequences cutinase/PET hydrolase used in the tree.Fig. 3. Maximum likelihood phylogenetic tree with sequences of potential PETase-like enzymes from the Streptomyces strains, their Swiss-Prot nearest neighbors and additional ones: Streptomyces sp. SM14 alpha/beta hydrolase (DAC80635.1) and the Leaf-branch compost cutinase (LCC) (G9BY57.1) (Sulaiman et al. 2012). The sequence of the strain NLAR_46 was excluded because of its smaller size (152 aa), while NLBR_30 was excluded due to its distance from other PETases/Cutinases. The maximum likelihood phylogenetic tree was constructed in MEGA 11 using amino acid sequences with 240 aa length (Fig. S4) and the bootstrap method with 500 replications and WAG + G model. Branch length indicates the number of substitutions per site and numbers indicate the percentage of bootstrap replicates in which the associated taxa clustered together

Genome mining for potential plastic-degrading enzymes

As a result of the genome mining for potentially encoding plastic-degrading enzymes, using both the developed curated database in this work and the PlasticDB database, a table comprising all hits above 40% identity (threshold used in BLAST protein–protein) with plastic-degrading enzymes is presented in supplementary Table S11.

Six strains were selected for genome sequencing based on the screening assays. Four of them (1.7L, CPN2, NLCR_1, and NLCR_8) tested positive in all assays (Table 1): they grew better in net-supplemented media than in the control (MM agar), showed lipolytic activity, and carried homologs of alkB and almA. The hydrocarbon-degrading strain CPN3 shared these traits, except it grew similarly in media supplemented with nets and the control. In contrast, strain NLBR_29 lacked positive amplification of alkB and almA genes but showed lipolytic activity and grew only in net-supplemented media.

Overall, from all 6 genomes combined, there were 114 hits with potential plastic-degrading enzymes, above 40% identity, of which 72 were related to synthetic polymers degradation. A summary of the overall protein hits is displayed in Fig. 4.Fig. 4. Overall genome mining results for the 1.7 L (Pseudomonas khazarica), CPN2 and CPN3 (Rhodococcus erythropolis), NLAR_29 (Bacillus pumilus), NLCR_1 (Rhodococcus qingshengii) and NLCR_8 (Rhodococcus fascians): Barplot indicates the number of protein hits above 40% identity, after using Blast protein–protein analysis, with enzymes linked to the degradation of either synthetic or bio-based/“biodegradable” plastics, displaying on the right side the presence (positive hit) or absence (no hit) of hits linked with the degradation of the synthetic plastics PET (polyethylene terephthalate); LDPE (low-density polyethylene); nylon (or polyamides); the monomer styrene; PBAT (poly(butylene adipate-co-terephthalate)) and PVA (polyvinyl alcohol). Corresponding to each plastic type, the names of enzyme hits and their respective class (hydrolases, isomerases, oxidoreductases, or transferases) are represented

The Pseudomonas strain 1.7 L displayed the highest number of hits followed by the Rhodococcus strains NLCR_1 > CPN2 > CPN3 > NLCR_8 and the Bacillus strain NLAR_29, but when it comes to enzymes linked with synthetic plastics degradation, results were even among all strains except for NLAR_29, which presented only one hit (PVA dehydrogenase) related to the water-soluble synthetic polymer PVA (Polyvinyl alcohol).

Nevertheless, in the genomes of the 5 strains, other than NLCR_29, a total of 15 different enzymes linked with synthetic plastics degradation were found, mostly belonging to the enzyme classes of hydrolases such as PETase (41%—50% identity), polyamidase (42% – 48%), carboxylesterase (40% – 43%), and the class oxidoreductases like alkane monooxygenase (52% – 66%), alkane hydroxylase (42% and 98%), and protocatechuate 3,4-dioxygenase (40% – 60%) enzymes. Pseudomonas strain 1.7 L offered the highest protein identity percentages in some protein hits such as alkane hydrolase (98.1%), polyesterase (88.3%), and an esterase (71.2%), regarding synthetic plastic degradation.

Discussion

In this study, we developed a comprehensive laboratory workflow to explore the plastic-degrading potential of bacterial strains found in biofilms attached to plastic marine litter. Our goal was to screen and identify bacterial strains with appropriate genomic and phenotypic features compatible with plastic degradation that could turn them into suitable candidates for future in vitro testing and validation. We achieved this by combining both culture-dependent and independent methodologies tailored to a group of marine bacteria retrieved from biofilms growing on 1-month submerged plastic nets as well as from strains with hydrocarbon-degrading capabilities.

The agar-based screening assays were conducted using the same plastic polymers as the ones present in the nets from where the 122 stains were isolated, in the previous in situ experiment (Perdigão et al. 2025)—polyethylene, nylon 6, and nylon 6,6. However, this workflow is not limited to these specific polymers and can be adapted for other polymers as well. It is important to note that in the employed culture-dependent assays, the bacterial growth could also be influenced by other compounds present in the media (e.g., agar and EDTA in the mineral medium, cellulose from disks), other than the addition of the plastic polymers themselves, although we consider that the influence is small. Differences in growth were still observed between bacteria grown on plastic-net media and on MM agar alone (media without plastic), with the latter providing a baseline to account for any growth attributable to agar or EDTA. The plastic-degrading potential screening was complemented with culture-independent methods. Genome mining or PCR-based approaches only inform on the genomic occurrence of said gene(s) but do not guarantee its expression. Therefore, the combination of both culture-dependent and independent approaches offers a more comprehensive and reliable screening workflow, overcoming some limitations of each approach.

For the culture-independent methods applied in this study, we broaden our search for genes and enzymes associated with other types of plastic polymers, especially synthetic, given their petroleum nature and possible link between the degradation of those components. In this same line of petroleum hydrocarbons being the base of plastics, in this study we also screened three additional highly capable hydrocarbon-degrading strains for plastic degradation potential.

Given the usual composition of plastics with a mixture of polymers, plasticizers, or other additives (UNEP 2016; www.plasticisers.org), an initial physico-chemical characterization of the input plastics is advised to ensure that the bacteria receive only the targeted single polymer without unintended carbon sources. Here, we used single-polymers plastics. The application of commercial polymers in the form of powder or pellets to culture media is also applied in the screening for plastic-degrading potential by other authors (Charnock 2021; Joshi et al. 2022).

In this study, some of the bacterial strains that belong to the low-abundance members of the net biofilm communities (< 1% abundance) were the ones that showed overall better results, namely from the genera of Rhodococcus, Bacillus, and Pseudomonas. In contrast, some other low-abundance members, previously reported for low-density polyethylene (LDPE) degradation potential (Khandare et al. 2021; Maroof et al. 2022), such as E. oxidotolerans NLAR_30, did not exhibit potential for PE degradation (no growth on solid media with PE net nor detection of genes encoding hydrocarbon-degrading monooxygenases). Recent studies have, however, highlighted the polystyrene (PS) degradation potential of Exiguobacterium species isolated from soil samples, and their production of enzymes involved in the degradation pathway of styrene, namely dioxygenase and hydrolases (Chauhan et al. 2018; Parthasarathy et al. 2022). Only two strains representative of the net biofilm communities (> 1% abundance)—Erythrobacteraceae (NLCR_33) and Sulfitobacter (NLCR_45)—were also found to be abundant in PE or polyamide (nylon-6) plastic biofilm communities in previous studies (Oberbeckmann et al. 2018; Wallbank et al. 2022). These strains, while growing on media supplemented with nets, failed on lipolytic activity and on the detection of the genes of interest by PCR in this study.

Present results corroborate the growing suggestion that the predominant members within the mature biofilms of marine plastic debris might not play a significant role in plastic degradation, often assuming a more generalist function (Vaksmaa et al. 2021; Wallbank et al. 2022). Instead, early-stage colonizers (< 1 month), members of the rare biosphere, or biofilm layers close to the plastic surface, might hold key plastic-degrading microorganisms in marine environments (Kirstein et al. 2019).

An in situ incubation of PE plastics in the Mediterranean Sea revealed after just 2 days the predominance of OHCB (obligate hydrocarbonoclastic bacteria) members, such as Alcanivorax and Marinobacter, especially onto the surface of weathered PE (Erni-Cassola et al. 2020). Yet, this predominance of OHCB was not sustained a few days later, supporting the idea that more mature biofilms in the plastisphere may conceal polymer-specific microorganisms and make the detection of potential polymer degraders difficult.

The potential of hydrocarbon-degraders for plastic biodegradation was also emphasized in the work of Vaksmaa et al. (2021), where the authors detected the presence of members of Marinobacter and Alcanivorax in the plastisphere of at least one polymer type (above 0.5% abundance) and Erythrobacter (Erythrobacteraceae family) above 1% in either PE or Polypropylene (PP). In addition, Zadjelovic et al. (2020) uncovered alkane and potentially plastic-degrading enzymes in a genome of the Alcanivorax sp. strain 24, namely esterases and peroxidases but also 3 alkB genes and 2 FAD-dependent monooxygenases/almA genes (screened on the PCR-based assays in the present study). Furthermore, a hydrocarbon-acclimated consortia (comprising Alcanivorax members) was shown recently to be able to initiate PET degradation (Denaro et al. 2020).

Our screening assays revealed that the 3 hydrocarbon-degrading strains tested in our study (CPN2, CPN3—R. erythropolis; 1.7L—Pseudomonas sp.) were amongst the best performers. These strains exhibited improved growth on plastic-containing media (except R. erythropolis CPN3), lipolytic activity, and contained homologs of both alkB and almA genes (encoding alkane-monooxygenases and FAD-dependent monooxygenases, respectively). The alkB genes are commonly found in these genera (Belhaj et al. 2002; Táncsics et al. 2015).

Both genes were also detected in the Acinetobacter lwoffii strain NLBR_41. The almA gene, in fact, is associated with long-chain n-alkane degradation within this genus (Throne-Holst et al. 2007). In our study, we detected almA in the Actinomycete Oerskovia (NLCR_13 strain). Although this genus has been reported for its remarkable capacity to degrade alkanes and polycyclic aromatic hydrocarbons (PAHs) (Lješević et al. 2019), no reference to the presence of this gene has been made until now. Belonging to another Actinomycetes genus previously linked to the degradation of a wide range of alkanes, we detected in the strain Dietzia sp. NLCR_54 alkB genes but no almA, aligning with the findings of Chen et al. (2017). As for the presence of genes encoding PETase-like enzymes in our broad group of bacteria, positive amplification of PETase-like encoding genes was only observed in Streptomyces strains (NLAR_46, NLBR_30, NLCR_57, NLCR_59). This was seen also in the screenings of Charnock (2021). However, in our study, co-amplification of other sized bands together with our expected single band (results not shown) occurred in the Actinomycete Dietzia sp. strain NLCR_54. In summary, alkane-monooxygenases (alkB gene), FAD-dependent monooxygenases (almA gene), and PETase-like enzymes were detected in this work mainly in Actinomycetes species.

Actinomycetes found in marine environments, particularly those belonging to the Streptomyces genus, exhibit significant potential for producing natural products (Girão et al. 2022). In recent years, these microorganisms have gained attention for their role in degrading synthetic plastic polymers, including PET, originating mainly from soil environments (Amobonye et al. 2021; Sriyapai et al. 2018). An example of such is the PETase-like enzyme identified in the Streptomyces sp. SM14 strain originating from a marine sponge, which was confirmed to possess polyesterase activity (Almeida et al. 2019). In our current study, we did not investigate the PET-degrading abilities of Streptomyces strains. However, following PCR detection of PETase homologs, an additional 1-month growth screening was conducted in minimal medium supplemented with nets. Although the nets were composed of other polymers (PE and Nylon), strains NLAR_46, NLCR_57, and NLCR_59 showed enhanced growth in media supplemented with nets and formed biomass attached to plastic pieces, particularly net A and net C. These findings support the applicability of the workflow developed in this study and also their potential for future exploration, including genome mining and degradation assays, for their ability to break down plastic polymers, especially PET.

In a metagenomic analysis of both marine and sediment samples, Danso et al. (2018) associated the presence of newly identified PET hydrolases primarily with the Actinobacteria (or Actinomycetota), Proteobacteria (or Pseudomonadota), and Bacteroidetes (or Bacteroidota) phyla. Interestingly, the authors pinpoint the phylum Bacteroidota as the primary host for PET hydrolases in the marine environment and stress the existing research gap regarding the plastic-degradation capacity among marine members of this phylum.

Within this phylum, only one strain, Bacillus sp. NLBR_22, demonstrated improved growth across all three nets in the agar-based approach. Notably, NLBR_22 possessed lipolytic activity but did not carry any of the genes screened in the PCR assays. However, Bacillus sp. NLAR_29—despite showing no amplification for the targeted genes—was selected for genome sequencing to explore its potential for plastic degradation. Aside from lipolytic activity, NLAR_29 thrived on net C (thin nylon; growth in wells) in the agar-based assays, which could suggest the presence of other plastic-degrading enzymes, interesting to explore in its genome.

Closely related to Bacillus pumilus species, which have previously been noted for their ability to degrade either PE (Harshvardhan and Jha 2013) or polyurethane (PU) (Nair and Kumar 2007), our study’s genome mining revealed the existence of only one potential homolog of synthetic plastic-degrading enzyme in NLAR_29, specifically for PVA plastic. It stood out from the rest of the mined genomes (the Rhodococcus spp. CPN2, CPN3, NLCR_1, NLCR_8, and the Pseudomonas 1.7L) as the sole strain with hits for PVA, and where most of its protein hits were linked to biobased or “biodegradable” plastics.

The other strains in our study exhibited enzymes associated with both PET and PE degradation, and nearly all displayed connections to the synthetic co-polyester PBAT (Poly(butylene adipate-co-terephthalate)). Interestingly, the strain NLCR_8, closely related to Rhodococcus fascians (ANI 97.5%)—a species documented in previous studies for its PBAT (Poly(butylene adipate-co-terephthalate)) biodegradation capabilities (Soulenthone et al. 2021)—did not feature in its genome any enzyme hits linked to PBAT degradation. However, it did possess three homologs of PETase, with one exhibiting the highest protein identity of 50%.

In this research, homologs of PETase were exclusively identified within the genomes of Rhodococcus strains. A thorough examination of nearly 670 Rhodococcus species genomes by Zampolli et al. (2022) unveiled a significant abundance of enzymes associated with polyester biodegradation, including members of the species explored in our study—R. erythropolis, R. qingshengii, and R. fascians. The authors also emphasized the PE degradation potential of R. qingshengii strains. In a recent study by Rong et al. (2024), a glutathione peroxidase was implicated in the depolymerization of LDPE, after transcriptomic and genomic analyses of the Rhodococcus strain C-2, genetically closely related to R. qingshengii (ANI > 95%). Our R. qingshengii closely related strain (ANI > 98.4%), NLCR_1, demonstrated hits for LDPE degradation, featuring 2 laccases and 4 alkane monooxygenases.

Concerning the two hydrocarbon-degrading strains of R. erythropolis, both exhibited similar plastic-degrading enzymes; however, only strain CPN2 displayed hits for styrene degradation. Although only a monomer of polystyrene, the metabolic pathway of styrene was included in our genome mining because the initial enzymatic depolymerization of PS is still unknown (Ho et al., 2018), Ho et al. 2018); 2018 still, styrene has been shown to be utilized and converted into intermediates of the tricarboxylic acid (TCA) cycle by bacteria of the genera Rhodococcus and Pseudomonas (Hou & Majumder, 2021) and references therein.

The genome of our Pseudomonas strain 1.7L closely relates to Pseudomonas khazarica, a marine bacterium known for its ability to utilize PAHs (Tarhriz et al. 2020). Prior to our study, no genome mining for plastic-degrading potential had been conducted for this Pseudomonas species. Nevertheless, other Pseudomonas species have been investigated, revealing the presence of enzymes capable of degrading PET and PE (Bollinger et al. 2020; Hou et al. 2022). For instance, Howard et al. (2023) linked polyester-degrading capacity to Pseudomonas stutzeri (PS13) after identifying two putative polyesterases and a putative MHETase in its genome. In our study, the genome of the 1.7L strain not only showed hits for PET, LDPE, and PE polymer degradation but also demonstrated enzymes linked to styrene and PBAT degradation.

In the search for new microorganisms and enzymes involved in plastic degradation, the use of multiomics approaches, such as metagenomics, metatranscriptomics or even metaproteomics, is gaining terrain (Viljakainen and Hug 2021). These techniques allow us to unravel complex microbial interactions and enzymatic processes involved in plastic degradation. Nonetheless, culture-dependent approaches have proven useful in the past (Yoshida et al. 2016) in discovering new plastic-degrading enzymes or microorganisms, and their application might be more cost-effective and easier to perform in the laboratory.

When tackling plastic pollution, it is essential to consider a holistic approach combining governance, politics, preventative measures, and technologies, where biodegradation could have a crucial role. Our findings underscore the plastic-degrading potential of marine bacteria, including those associated with the plastisphere of marine litter items such as fishing nets, as well as those present in hydrocarbon-contaminated environments.

Conclusions

This study aimed to screen and identify bacterial strains with genomic and phenotypic features that could potentially contribute to plastic degradation. Marine bacteria isolated from 1-month-old biofilms growing on submersed plastic fishing nets and three strains obtained from hydrocarbon-enriched environments were selected, and a comprehensive laboratory workflow that integrated both culture-dependent and independent methods was used. Aside from presenting a comprehensive laboratory workflow, this study stands out by offering a detailed, step-by-step genome mining protocol and a curated database that integrates two major plastic-degrading enzyme datasets, including relevant metabolic pathway information retrieved from the KEGG database. Hence, it provides a valuable, ready-to-use resource for researchers aiming to explore their microbial strain’s library potential for plastic degradation.

Among the 21 tested strains, selected based on their representation of the biofilm community and previous reports of members from the same species involvement in plastic degradation, 11 bacteria presented improved growth in the presence of net polymers (PE and Nylon) when compared to the control minimal media. These strains predominantly belonged to the genera Erythrobacteraceae, Sulfitobacter, Rhodococcus, Bacillus, and Pseudomonas, with 8 of them also demonstrating lipolytic activity. The detection of alkB, almA homologs, as well as genes encoding PETase enzymes using degenerated primers proved to be more suitable for applied screening assays, with the majority of positive amplifications observed in Actinomycetes strains. Using the curated databases, we could successfully mine for potential plastic-degrading enzymes, namely for synthetic polymers, in the six promising bacterial strains obtained through the agar and PCR tests. Furthermore, three of the four Streptomyces strains that showed PET hydrolytic potential (by a positive amplification of PETase homologs in direct PCR) also displayed enhanced growth in media supplemented with nets, supporting the applicability of the workflow developed in this study. These strains represent promising candidates to be further explored for their ability to break down plastic polymers in laboratory, especially PET.

Our findings suggest that biofilms on plastic fishing nets and hydrocarbon-impacted environments can serve as reservoirs of bacteria and enzymes with plastic-degrading potential, and thus a promising source for further bioprospecting. Future laboratory experiments should now validate the plastic degradation capacity of the selected strains obtained in this work.

Supplementary Information

Below is the link to the electronic supplementary material. ESM 1(FASTA 17.7 MB)ESM 2(PNG 423 KB)High resolution image (TIF 3.72 MB)ESM 3(PNG 417 KB)High resolution image (TIF 4.23 MB)ESM 4(PNG 425 KB)High resolution image (TIF 4.29 MB)ESM 5(PDF 1.31 MB)