Polychlorinated Biphenyl Transformation, Peroxidase and Oxidase Activities of Fungi and Bacteria Isolated from a Historically Contaminated Site

Causing major health and ecological disturbances, polychlorinated biphenyls (PCBs) are persistent organic pollutants still recovered all over the world. Microbial PCB biotransformation is a promising technique for depollution, but the involved molecular mechanisms remain misunderstood. Ligninolytic enzymes are suspected to be involved in many PCB transformations, but their assessments remain scarce. To further inventory the capabilities of microbes to transform PCBs through their ligninolytic enzymes, we investigated the role of oxidase and peroxidase among a set of microorganisms isolated from a historically PCB-contaminated site. Among 29 isolated fungi and 17 bacteria, this work reports for the first time the PCB-transforming capabilities from fungi affiliated to Didymella, Dothiora, Ilyonectria, Naganishia, Rhodoturula, Solicoccozyma, Thelebolus and Truncatella genera and bacteria affiliated to Peribacillus frigotolerans, Peribacillus muralis, Bacillus mycoides, Bacillus cereus, Bacillus toyonensis, Pseudarthrobacter sp., Pseudomonas chlororaphis, Erwinia aphidicola and Chryseobacterium defluvii. In the same way, this is the first report of fungal isolates affiliated to the Dothiora maculans specie and Cladosporium genus that displayed oxidase (putatively laccase) and peroxidase activity, respectively, enhanced in the presence of PCBs (more than 4-fold and 20-fold, respectively, compared to controls). Based on these results, the observed activities are suspected to be involved in PCB transformation.


Introduction
Polychlorinated biphenyls (PCBs) are synthetic compounds manufactured by humans and used from 1929 to their prohibition at the end of 1980s. Mainly found in electrical transformer devices, paints, plasticizers or flame retardants [1], they were classified as persistent organic pollutants according to the Stockholm Convention from 2001. Breivik et al. (2002) have estimated that approximately 13 million tons of PCBs were synthetized worldwide, with 20% to 35% released in ecosystems [1,2]. Recognized as endocrine disruptors and carcinogenic substances [3,4], they represent ecological and health concerns. Due to their chemical characteristics, they are stable and persistent molecules that accumulate in ecosystems and the food web through their bioaccumulation and biomagnification, making their degradation a challenge [5,6].

PCR Amplification, Gene Sequencing and Taxonomic Affiliation
Fungal DNA extractions were conducted for each isolate as described by Liu et al. [33]. The PCRs were carried out according to the conventional protocols using a Biorad C1000 thermal cycler (Biorad, Hercules, CA, USA), using 4 µL of HOT FIREPol ® Blend Master Mix Ready to Load (5X) (Solis Biodyne, Tartu, Estonia), 0.8 µM of each primer and 20 ng of fungal genomic DNA or small amount of a bacterial colony picked up using a sterile toothpick, in a final volume of 25 µL. ITS and 16S rDNA amplicons, obtained with the ITS1F (5 -CTTGGTVATTTAGAGGAAGTAA-3 ) and ITS4 (5 -TCCTCCGCTTATTGATATGC-3 ) primer pair and the pA (5 -AGAGTTTGATCCTCGCTCAG-3 ) and pH (5 -AAGGAGG TGATCCAGCCGCA-3 ) primer pair, respectively, were sequenced (Microsynth France, Vaulx-en-Velin, France https://www.microsynth.com/standard-services.html, accessed on 19 April 2023). For taxonomic affiliation, fungal isolates were first identified according to general morphotype principles of fungal classification. Then, ITS sequences were compared against tree databases: the rRNA/ITS databases with fungi (taxid:4751) as organism option using Blastn+ [34]; the UNITE and the INSD databases using massBLASTer (BLAST+ 2.13.0) in PlutoF [35]. For bacteria, the 16S rDNA sequences were compared to the Refseq genome database using Blastn+ [34]. The 16S rDNA sequences were further aligned with reference sequences and phylogenetic trees using the Maximum Likelihood method, and a Kimura 2-parameter model with 1000 replicates was constructed using Seaview [36]. Amplicons of laccase-encoding genes were obtained using the degenerate primer pairs Cu1BF/Cu2R targeting laccase-encoding genes from Basidiomycota and Cu1AF/Cu2R targeting Ascomycota ones [37]. Fungal dye-peroxidases and Basidiomycota class II peroxidase genes were amplified using DYP360F/DYP485R [38] and AA2-F and AA2-R [39] primer pairs, respectively. The F_DYPR and R_DYPR primers from Tian et al. 2016 were used to target bacterial Dye-Peroxidase genes [40]. The bacterial bphA gene was amplified using the 4 different pairs of primers from Correa et al. 2010 [41]. All positive amplicons were cloned using TOPO ® TA Cloning ® Kit for Sequencing (Invitrogen, Waltham, MA, USA), and ten clones per amplicon were analyzed after sequencing by Microsynth France.

PCB Transformation Analysis
Each fungal isolate was grown on Petri dishes with CYM media. In triplicate, 10 calibrated fresh mycelial inocula (diameter 5 mm) were added to a liquid minimal media [42,43]. After incubation for 2 days at 24 • C on a rotary shaker (120 rpm), Aroclor 1254 (25 mg·L −1 dissolved in methanol (Sigma-Aldrich ® , Saint Quentin Fallavier, France) was added, and each culture was incubated for 3 additional days under the same conditions. After incubation, cultures were frozen at −80 • C. Each bacterial isolate was cultivated in quadruplicate in liquid LB medium until mid-exponential growth time and washed 3-times with NaCl 0.9% (1.2 M, 70% w/v) solution. Then, 50% of washed cells were transferred in glass tube containing 5 mL of a media (Yeast extract 10 g·L −1 , NH 4 Cl 2 g·L −1 , NaCl 1 g·L −1 , KH 2 PO 4 6 g·L −1 , Na 2 HPO 4 25.6 g·L −1 and Aroclor 1254 25 mg·L −1 (Sigma-Aldrich, dissolved in methanol)), incubated during 48 h at 30 • C on a rotary shaker (120 rpm). After incubation, cultures were frozen at −80 • C. Indicator PCB (iPCB) assays of the whole culture (including extracellular medium and microbial cells) were performed using gas chromatography coupled with mass spectrometry (GC-MS) (AGROLAB FRANCE SARL, Dijon, France). Positive controls without fungal or bacterial cells and with Aroclor 1254 were also assayed in triplicate. iPCBs correspond to PCB congeners 28, 52, 101, 118, 138, 153 and 180 found to be the most abundant PCB congeners in environmental and biological matrices and the most common and easily determined congeners [44]. The percent of iPCB transformation efficiency was calculated for each isolate by subtracting the average iPCB quantity of each isolate to the positive control. The results were then transformed in percentages by considering control as 100%. Comparison of iPCB amount between controls and biotransformation assays was conducted using a Pairwise Wilcoxon test using the R software (https://www.r-project.org/, accessed on 18 November 2023).

Oxidase and Peroxidase Activity Assay
Each fungal isolate was cultivated as described above except that minimum media was supplemented with Aroclor 1254 1 mg·L −1 (dissolved in 5 µL methanol), methanol (5 µL) or no supplementation for control conditions. Each bacterial isolate was cultivated in a media containing Yeast Carbon Base-NutriSelect ® 11.7 g·L −1 , (NH 4 ) 2 SO 4 3.5 g·L −1 , CH 3 COCOOH 10 g·L −1 supplemented with Aroclor 1254 1 mg·L −1 (dissolved in 5 µL methanol), methanol (5 µL) or no supplementation for both control conditions. After a total of 7 days of incubation, cells were centrifuged 10 min at 4000 g. For oxidase activity assay, 50 µL of supernatant was transferred in 96-well plate preloaded with 200 µL of phosphate citrate buffer pH 4 containing ABTS (2,2'-azino-bis(3-ethylbenzothiazoline-6sulfonic acid)) (5.8 mM) or Guaiacol (53 mM) as substrates (Sigma-Aldrich ® ). The same protocol was used for peroxidase activity except that H 2 O 2 (100 µM) was added to each well. Oxidase and peroxidase activities were quantitatively determined by measuring absorption change of each substrate for 4 h using TECAN infinite 200 ® PRO microtiter plate reader (Tecan, Salzburg, Austria) at 740 nm and 470 nm for ABTS and Guaiacol, respectively. Molar extinction coefficients of ABTS and Guaiacol ( 1 g·L −1 , KH2PO4 6 g·L −1 , Na2HPO4 25.6 g·L −1 and Aroclor 1254 25 mg·L −1 (Sigma-A dissolved in methanol)), incubated during 48 h at 30 °C on a rotary shaker (120 rpm incubation, cultures were frozen at −80 °C. Indicator PCB (iPCB) assays of the who ture (including extracellular medium and microbial cells) were performed using ga matography coupled with mass spectrometry (GC-MS) (AGROLAB FRANCE SAR jon, France). Positive controls without fungal or bacterial cells and with Aroclor 125 also assayed in triplicate. iPCBs correspond to PCB congeners 28, 52, 101, 118, 138, 1 180 found to be the most abundant PCB congeners in environmental and biologic trices and the most common and easily determined congeners [44]. The percent o transformation efficiency was calculated for each isolate by subtracting the averag quantity of each isolate to the positive control. The results were then transformed centages by considering control as 100%. Comparison of iPCB amount between c and biotransformation assays was conducted using a Pairwise Wilcoxon test using software (https://www.r-project.org/, accessed on the 18 November 2023).

Oxidase and Peroxidase Activity Assay
Each fungal isolate was cultivated as described above except that minimum was supplemented with Aroclor 1254 1 mg·L −1 (dissolved in 5 µL methanol), meth µL) or no supplementation for control conditions. Each bacterial isolate was cultiv a media containing Yeast Carbon Base-NutriSelect ® 11.7 g·L −1 , (NH4)2SO4 3.5 CH3COCOOH 10 g·L −1 supplemented with Aroclor 1254 1 mg·L −1 (dissolved in 5 µL anol), methanol (5 µL) or no supplementation for both control conditions. After a 7 days of incubation, cells were centrifuged 10 min at 4000 g. For oxidase activity 50 µL of supernatant was transferred in 96-well plate preloaded with 200 µL of pho citrate buffer pH 4 containing ABTS (2,2'-azino-bis(3-ethylbenzothiazoline-6-s acid)) (5.8 mM) or Guaiacol (53 mM) as substrates (Sigma-Aldrich ® ). The same p was used for peroxidase activity except that H2O2 (100 µM) was added to each we dase and peroxidase activities were quantitatively determined by measuring abso change of each substrate for 4 h using TECAN infinite 200 ® PRO microtiter plate (Tecan, Salzburg, Austria) at 740 nm and 470 nm for ABTS and Guaiacol, respec Molar extinction coefficients of ABTS and Guaiacol (ɛM) were obtained by measurin full oxidation in the concentration range from 10 µM to 1000 µM at the previousl tioned wavelength. The estimation of total excreted proteins was measured wit massie (Bradford) Protein Assay Kit (Thermo Scientific, Illkirch, France). Specific ac are expressed in µmol of substrate oxidized per minute and per mg of total excrete teins (µmol·min −1 ·µg −1 ). A pairwise Wilcoxon test in R software was conducted to co oxidase activities measured in the presence of PCBs compared to controls (with or w methanol).

Results
First, 146 fungi and 83 bacteria were isolated from a PCB-contaminated brow taxonomically characterized and assayed for their PCB-transformation capacity. A them, the 29 fungal isolates listed in Table 1 and the 17 bacterial isolates listed in were selected with respect to their taxonomic diversity and for their highest PCB formation efficiency, i.e. presenting more than 30% or 10% of iPCB depletion in 3 or for fungi or bacteria, respectively. Table 1. Identification name (ID), phylum and species blast taxonomic affiliation of the ITS (with percent of identity), the database used for taxonomic affiliation and average degradat ciency ± the standard deviation in percent for each assayed isolate. M ) were obtained by measuring their full oxidation in the concentration range from 10 µM to 1000 µM at the previously mentioned wavelength. The estimation of total excreted proteins was measured with Coomassie (Bradford) Protein Assay Kit (Thermo Scientific, Illkirch, France). Specific activities are expressed in µmol of substrate oxidized per minute and per mg of total excreted proteins (µmol·min −1 ·µg −1 ). A pairwise Wilcoxon test in R software was conducted to compare oxidase activities measured in the presence of PCBs compared to controls (with or without methanol).

Results
First, 146 fungi and 83 bacteria were isolated from a PCB-contaminated brownfield, taxonomically characterized and assayed for their PCB-transformation capacity. Among them, the 29 fungal isolates listed in Table 1 and the 17 bacterial isolates listed in Table 2 were selected with respect to their taxonomic diversity and for their highest PCB-transformation efficiency, i.e., presenting more than 30% or 10% of iPCB depletion in 3 or 2 days for fungi or bacteria, respectively.

Discussion
While PCBs remain a major health and ecological issue, microbial PCB bioremediation processes are an attractive way to depollute contaminated matrices. However, molecular mechanisms performed by microbes to transform PCBs are still misunderstood. Among the new PCB-transforming pathway, ligninolytic enzymes that are produced by fungi and bacteria are good candidates to be responsible for PCB transformation. Indeed, several fungal purified laccases have been observed to transform some PCBs [19,27,28,45,46]. In this way, to further investigate the distribution of potential PCB-transforming laccases and peroxidases among fungal and bacterial taxa, the present work aimed to inventory microbial isolates involved in PCB transformation, taxonomically characterize them, amplify ligninolytic genes on their genomic DNA and measure their secreted oxidase and peroxidase activities in the presence and the absence of PCBs. In addition, the bph gene PCR amplification was performed on bacterial DNA in order to determine which bacterial isolate is able to transform PCB through the well-known bph pathway and those potentially using another biochemical pathway.
Among the 29 fungal isolates analyzed, this is the first report of PCB transformation ability for fungal isolates belonging to Didymella, Dothiora, Ilyonectria, Naganishia, Rhodoturula, Solicoccozyma, Thelebolus and Truncatella genera (Table 1). In the same way, among the 17 bacterial isolates, this is the first report of PCB-transforming capabilities from bacteria phylogenetically close to Peribacillus frigotolerans, Peribacillus muralis, Bacillus mycoides, Bacillus cereus, Bacillus toyonensis, Pseudarthrobacter sp., Pseudomonas chlororaphis, Erwinia aphidicola and Chryseobacterium defluvii (Table 2 and Figure 1). This is also the first report of PCB-enhanced peroxidase activities from a fungal isolate affiliated to Cladosporium sp. and PCB-enhanced oxidase activities from an isolate closely related to Dothiora maculans (more than 20 and 4-fold, respectively). Concerning fungal peroxidases and according to the literature, only one study reports enhanced peroxidase activity by PCBs from Acremonium sclerotigenum [31]. This supports the need of more investigation regarding peroxidase activity from fungi with PCB-transforming capabilities that might reveal a new PCB transforming pathway through peroxidases. On the contrary, laccases (that are oxidases) activities from Pleurotus ostreatus, Fusarium solani, Thermothelomyces thermophila, Pleurotus sajor-caju LBM 105, Cladosporium sp. and Trametes versicolor have been already observed to be enhanced by PCBs [28,31,43,47,48]. In this way, the enhanced oxidase activity in the presence of PCBs from fungal isolates CRIS_F13 phylogenetically close to Dothiora maculans and from CRIS_F29 affiliated to Cladosporium sp. suggests that this measured oxidase activity is likely due to the presence of laccases. This hypothesis is supported by the laccase gene amplification from the fungal isolate CRIS_F13. These PCB-enhanced peroxidase and oxidase activities also suggest that the PCB transformation observed for these two isolates might be due to laccases and peroxidases. The substrate proximity of laccase and peroxidase and purified laccases that displayed PCB-transforming abilities supports this hypothesis [26][27][28][29][30]. In addition, even if laccase activities of fungal isolates CRIS_F2, F8, F18, F20 F22, F24, F25, F26 and CRIS_F27 are not specifically enhanced by PCBs, they might be involved in observed PCB transformations. In any case, these hypotheses must be confirmed by testing purified laccases and peroxidases from those fungal isolates in PCB-transformation assays.
Concerning the bacterial isolates, it is surprising that no bph gene amplification was observed among those affiliated to Pseudomonas sp. (CRIS_B17), Pseudomonas chlororaphis (CRIS_B9), Peribacillus frigoritolerans (CRIS_B3, B7, B8 and B11), Bacillus cereus (CRIS_B1), B. mycoides (CRIS_B2) and B. toyonensis (CRIS_B14 and B16). Indeed, there is a wide distribution of bph gene in Pseudomonas and Bacillus PCB-transforming bacteria [49][50][51], and the recent new Peribacillus genus was formerly Bacillus [52]. This suggests that primers used in this study might not be able to amplify all bph genes, or that still unknown PCBtransforming pathways are involved in the PCB-transformation realized by those isolates.
This last hypothesis can also be applied to the other fungal and bacterial isolates that transform PCB with the lack of oxidase and peroxidase activities or bph gene (for bacteria only). It is particularly true for the first report through this work of the PCBtransforming fungal isolates affiliated to Didymella, Ilyonectria, Naganishia, Rhodoturula, Solicoccozyma, Thelebolus and Truncatella genera and to the PCB-transforming bacterial isolates affiliated to Chryseobacterium defluvii and Pseudarthrobacter sp. with no bph genes reported so far for phylogenetically close strains. This highlights that further investigations about the PCB-transforming capabilities of these species are required to reveal new PCBtransforming pathways and to participate to better understand the PCB biotransformation process operated by the microorganisms.

Conclusions
For the first time, this study reports direct PCB transformation capabilities of fungal isolates belonging to Didymella, Dothiora, Ilyonectria, Naganishia, Rhodoturula, Solicoccozyma, Thelebolus and Truncatella genera and to bacterial isolates affiliated to Peribacillus frigotolerans, Bacillus mycoides, B. cereus, B. toyonensis, Pseudarthrobacter sp., Pseudomonas chlororaphis, Chryseobacterium defluvii Erwinia aphidicola and Peribacillus muralis. Concerning oxidase and peroxidase activities, only two fungal isolates showed PCB-enhanced activities with a first report of PCB-enhanced peroxidase activity from a fungal isolate affiliated to Cladosporium sp. and PCB-enhanced oxidase activity from another closely related to Dothiora maculans (more than 20 and 4-fold respectively). These activities suggest the involvement of laccases and peroxidases in the PCB-transforming action of those fungal isolates. The unobserved oxidase and peroxidase activities for 18 fungal isolates out of 29 and all the bacteria (added to no bph gene amplification for these last), suggest that still-unknown PCB-transformation pathways operate. In this way, further investigations focusing on PCB-transforming capabilities are required to reveal new PCB-transformation pathways, especially for fungal isolates affiliated to Didymella, Ilyonectria, Naganishia, Rhodoturula, Solicoccozyma, Thelebolus and Truncatella genera and to the PCB-transforming bacterial isolates affiliated to Chryseobacterium defluvii and Pseudarthrobacter sp. The next step of this work is to characterize at the species level the most efficient iPCB transformers and those showing an enhanced enzymatic activity in the presence of PCBs. For this purpose, PCB up-regulated peroxidases and laccases encoding genes from these isolates will be identified by RNA seq to further predict in silico how they could transform iPCBs by modeling Protein Ligand Docking on PCBs and up-regulated enzymes. Other enzyme encoding genes that could be identified as possible PCB transformers will undergo the same process. Then, the enzyme-PCB binding would have to be confirmed in vitro using purified enzymes from these isolates and by chemically characterizing the releasing PCB byproducts.