Bacterial isolates used in this study were previously obtained from a biopurification system used for pesticide removal from contaminated waters located in Kortrijk, Belgium, operational since 2008 (16). All strains were routinely grown on Luria-Bertani (LB) agar plates or in liquid LB medium (Sigma–Aldrich, Merck, St Louis, MO, USA) at 37 °C and 200 rpm. The biopurification system composition, analysis of pesticide types and concentrations (19), high molecular mass (HMM) plasmid DNA purification, high-throughput sequencing and computational analysis were previously reported (20). Escherichia coli DH5α (Promega, Madison, WI, USA) and E. coli BL21 Star^TM (DE3) (Invitrogen Life Technologies, Carlsbad, CA, USA) strains were employed for lac80 cloning and heterologous expression, respectively. They were routinely cultured in LB medium supplemented when appropriate with 100 μg/mL of ampicillin (Sigma–Aldrich, Merck). The pUC59 (GenScript Biotech, Piscataway, NJ, USA) and pET22b(+) (Novagen, Darmstadt, Germany) plasmids were used for lac80 cloning and expression, respectively. The latter plasmid enabled us to introduce a hexahistidine tag (His₆-Tag) at the C-terminus of the protein. Molecular biology techniques were performed using the standard protocols (21). All the reagents were purchased from Sigma–Aldrich, Merck unless otherwise stated.

Laccase-like multicopper oxidase (LMCO)-encoding lac80 gene was identified in silico in the present study from the plasmid dataset reported by Martini et al. (20). To retrieve novel LMCOs, profile hidden Markov models (pHMMs) were used as previously described (9). The lac80 gene and aminoacidic sequences were compared to those reported in GenBank (22) and Protein Data Bank (23), respectively. The protein module structure was analyzed using the simple modular architecture research tool (SMART) (24). The presence and location of the signal peptide in lac80 was checked using the neural networks and hidden Markov models trained on Gram-negative and Gram-positive bacteria with SignalP 5.0 (25). Multiple sequence alignment of lac80 with related LMCO sequences was performed using ClustalΩ (26). Laccase gene was also blasted to the Laccase Engineering Database (27) by the basic local alignment search tool (BLASTP) algorithm (28). Phylogenetic relationship was inferred by using the maximum likelihood phylogenetic method. The tree was constructed using MEGA X (29) and a multiple sequence alignment by MUSCLE of Lac80 sequence with a selection of previously characterized and some uncharacterized bacterial LMCOs, and 100 bootstrapping replications were used as a test of phylogeny.

A molecular polymerase chain reaction (PCR)-based method was used to screen the bacterial collection for retrieving this novel LMCO-encoding gene. The strain harbouring lac80 was identified by using two sets of primers (set 1: 5’-CCACCGTCTGGGGTCTTG-3’ and 5’-GTCGATGCGCCGTATTTC-3’, amplicon size 549 bp and set 2: 5’-TCACCGGGCGATGCTGGC-3’ and 5’-GAGGAGGTGATGGCCGAGATC-3’, amplicon size 733 bp). PCR was performed as follows: an initial denaturation step at 94 °C for 4 min, followed by 35 cycles of denaturation at 94 °C for 20 s, annealing at 54 °C for 30 s and extension at 72 °C for 50 s. Then, a final extension at 72 °C for 2 min was used. The sequence of lac80 was submitted to GenBank (22) under the accession number MT130716.

In situ lysis gel electrophoresis was performed to obtain further information on the genomic location of lac80 by analyzing the DNA of the bands corresponding to plasmid(s) or chromosome, as previously described (16). The visualized bands were purified using the AccuPrep PCR/gel purification kit (Bioneer, Daejeon, Korea). Then, a semi-nested PCR for lac80 was run to reduce nonspecific amplification of DNA template. Purified samples and primers Lac80F1 (5’-CCACCGTCTGGGGTCTTG-3’) and Lac80R2 (5’-TCACCGGGCGATGCTGGC-3’) were used in the first PCR and Lac80F1 and Lac80R1 (5’-GTCGATGCGCCGTATTTC-3’) in the second PCR run. To exclude the amplification of contaminating DNA, several controls were included in the PCR with DNA recovered from randomly selected positions of the agarose gel.

The codon optimization strategy employed in the present study is referred to as ‘one amino acid-one codon’. In this method, the most preferred codon of the E. coli expression system for a given amino acid is utilized in the target sequence (30). The sequence of lac80 was obtained from the data set. The OPTIMIZER web server (31) was used for rare codon detection. The E. coli rare codon analyzer2 (32) was utilized for gene sequence optimization. GenScript web server (33) was used to analyze the designed sequence codon adaptation index (CAI).

The codon-optimized sequence of lac80 with the native signal peptide removed was synthesized by GenScript Biotech (Piscataway, NJ, USA) and cloned into pUC59 plasmid (Leiden, The Netherlands) for DNA amplification. For protein expression, lac80 was cloned from pUC59 into the expression vector pET22b(+) (Novagen Inc, Madison, WI, USA), and introduced into E. coli BL21 Star^TM (DE3) cells (Thermo Fisher Scientific, Waltham, MA, USA). Multiple conditions were then evaluated for optimizing Lac80 expression. Briefly, 2 L unbaffled Erlenmeyer flasks containing 750 mL of LB medium (Merck KGaA, Darmstadt, Germany) or richer Terrific broth (TB, in g/L: tryptone 12, yeast extract 24, K₂HPO₄/L 9.4, KH₂PO₄ 2.2, glycerol 8; components from Merck KGaA) supplemented with ampicillin 100 µg/mL (Merck KGaA) were inoculated with the starter culture (A_{600 nm}~0.1), then incubated at 37 °C and 200 rpm. When the early exponential phase of growth was reached (corresponding to an A_{600 nm} of approx. 0.6 for LB medium and approx. 1 for TB medium), different concentrations of isopropyl-β-d-thiogalactopyranoside (IPTG, from 0.4 to 1 mM) (Merck KGaA) were tested to induce protein expression. At the same time, none or 0.25 mM CuSO₄ (Merck KGaA) was added to facilitate metal incorporation into the enzyme active site. Cells were then incubated at 37 or 20 °C under different agitation regimes (200 or 100 rpm or without shaking), and harvested at 0, 2, 4 or 24 h after induction, by centrifuging at 7500×g for 10 min (Thermo Fisher Scientific, Langenselbold, Germany). Multiple combinations of incubation conditions were also evaluated, such as incubation for 4 h after induction at 37 or 20 °C and 200 rpm, followed by overnight incubation at the same temperature but without shaking. Total proteins in the cell-free fermentation broths were concentrated by precipitation with φ(trichloroacetic acid)=10%. Cell pellets were instead resuspended in 20 mM phosphate buffer, pH=6.7, containing 10 µg/mL deoxyribonuclease I (DNaseI) (Merck KGaA), 0.19 mg/mL phenylmethylsulfonylfluoride (PMSF) and 0.7 µg/mL pepstatin (Merck KGaA). Then, cells were disrupted by sonication for 6 cycles of 30 s each on ice, followed by centrifugation at 34 000×g for 60 min at 4 °C (Beckman Coulter Inc., Brea, CA, USA). This enabled us to recover soluble (cytoplasmic) and insoluble (inclusion bodies) fractions, which were then analyzed through: (i) sodium dodecyl sulfate-polyacrylamide (12% m/V) gel electrophoresis (SDS-PAGE) and Coomassie Brilliant Blue staining, (ii) Western blot with anti-His-Tag antibody horseradish peroxidase (HRP) conjugate (Novagen Inc., Madison) and detection by chemiluminescence (ECL Western blotting detection system, GE Healthcare Sciences, Little Chalfont, UK), and (iii) routine laccase activity assay with 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid) (ABTS) (Merck KGaA) as substrate (see below). Total protein concentration in the fractions was estimated by the biuret assay (34).

Lac80 was purified starting from recombinant E. coli cells grown under the conditions allowing the highest soluble protein production. Hence, glycerol stocks of E. coli BL21 Star^TM (DE3)/pET22(b)::lac80 were inoculated into 80 mL LB medium with 100 µg/mL ampicillin, grown overnight at 37 °C and 200 rpm. This pre-culture was then used to inoculate (at A_{600 nm}~0.1) 750 mL of selective LB medium. Flasks were incubated as before until an A_{600 nm} of 0.6 was reached. After induction of protein expression with 0.4 mM isopropyl β-d-1-thiogalactopyranoside (IPTG) and addition of 0.25 mM CuSO₄, incubation was prolonged overnight at 20 °C and 100 rpm. Cells were disrupted as reported above and the soluble, cytoplasmic fraction recovered by centrifugation at 34 000×g and 4 °C for 60 min (Beckman Coulter Inc.). C-terminus His₆-tagged Lac80 was purified by affinity chromatography on a 5-mL Ni²⁺-HiTrap chelating affinity column (GE Healthcare Sciences) equilibrated with 20 mM phosphate buffer, pH=6.7, 20 mM imidazole and 300 mM Na₂SO₄ (Merck KGaA). The recombinant protein was eluted with 20 mM phosphate buffer, pH=6.7, 250 mM imidazole and 300 mM Na₂SO₄, then loaded onto a size-exclusion PD10 Sephadex G25 column (GE Healthcare Sciences) equilibrated with 20 mM phosphate buffer, pH=6.7. Lac80 was further concentrated with 30 K Amicon Ultra-2 centrifugal filter devices (Merck KGaA).

Protein purity was checked with 10% m/V SDS-PAGE. For molecular mass determination, molecular mass markers (GE Healthcare Sciences) were used. Protein concentration was estimated by densitometric analysis of SDS-PAGE bands using Quantity One analysis software (35) (Bio Rad, Hercules, USA) and spectrophotometrically using Lac80 theoretical molar absorption coefficient at 280 nm calculated based on the amino acid sequence (ε_{280 nm}=58.44 mM^-1cm^-1).

Laccase activity was routinely measured at 25 °C in 1 mL reaction mixture containing 50 mM sodium acetate, pH=5.0, 5 mM ABTS (ε_{420 nm}=36 mM^-1cm^-1) and 100 μL protein sample. Oxidation of the substrate was monitored with a spectrophotometer (V460; Jasco, Easton, MD, USA) as the change in the absorbance at 420 nm for 5 min. All assays were conducted in triplicate and negative controls without enzyme were run in parallel. One activity unit (U) was defined as the amount of Lac80 required to oxidize 1 μmol of ABTS per minute. The ability of Lac80 to oxidize various phenolic and non-phenolic compounds was checked by the following additional substrates at designed concentrations and wavelengths: 100 mM 2,6-dimethylphenol (DMP) (ε_{468 nm}=49.6 mM^-1cm^-1), 100 mM pyrocatechol (ε_{450 nm}=2.21 mM^-1cm^-1), 100 mM pyrogallol (ε_{450 nm}=4.4 mM^-1cm^-1), 8 mM vanillic acid (ε_{316 nm}=2.34 mM^-1cm^-1), 10 mM K₄Fe(CN)₆ (ε_{405 nm}=0.9 mM^-1cm^-1), 2 mM tyrosine (ε_{475 nm}=3.6 mM^-1cm^-1), 5 mM l-3,4-dihydroxyphenylalanine (l-DOPA) (ε_{475 nm}=3.7 mM^-1cm^-1), 100 mM guaiacol (ε_{468 nm}=12 mM^-1cm^-1), 10 mM syringic acid (ε_{300 nm}=8.5 mM^-1cm^-1), 4 mM ferulic acid (ε_{287 nm}=12.4 mM^-1cm^-1), 1 mM syringaldazine (ε_{525 nm}=65 mM^-1cm^-1) and 1 mM syringaldehyde (ε_{320 nm}=8.5 mM^-1cm^-1). The optimal pH for the activity on different substrates was evaluated at 25 °C in a multi-component buffer (10 mM Trizma base, 15 mM sodium carbonate, 15 mM phosphoric acid, 250 mM potassium chloride (36), in the pH range 2.0–8.0. All substrates and buffer components were purchased from Merck KGaA.

Polluted sites are hot spots of plasmids potentially carrying catabolic genes (37). To identify novel laccase-like multicopper oxidase (LMCO)-encoding genes in the plasmid dataset from the 35 strains previously isolated from a biopurification system [REMOVED HYPERLINK FIELD] (20), a profile hidden Markov model-based search was performed. One candidate gene, named lac80, was retrieved using this approach. This gene showed 89% identity at nucleotide level with a yet uncharacterized multicopper oxidase from Ochrobactrum anthropi strain OAB (GenBank accession number: CP008819). Lac80 is composed of 502 amino acids, with a predicted signal peptide of 20 amino acids at the N-terminal (Fig. 1a). Predicted isoelectric point and molecular mass of this protein, excluding the signal peptide, were 5.43 and 51.2 kDa, respectively. As reported in Table S1, full-length Lac80 protein showed the highest amino acid sequence identity with two putative multicopper oxidases, one from Alphaproteobacteria (WP_024899901.1, 100% sequence identity) and the other from Ochrobactrum rhizosphaerae (WP_094574672.1, 99.8%). In addition, Lac80 showed 30% sequence identity with the Lac15 from a marine microbial metagenome (Protein Data Bank (PDB): 4F7K_A) and 29.5% identity with a laccase from Pseudomonas thermotolerans (PDB: 6VOW). These two latter proteins were crystallized and characterized as functional LMCOs. The phylogenetic tree shown in Fig. 1b confirms the novelty of our enzyme since Lac80 clusters most closely with uncharacterized LMCOs from Paracoccus sp., Ochrobactrum sp. and Rhizobiales. These findings were also confirmed by the best protein matches identified by the Laccase Engineering Database (LccED, https://lcced.biocatnet.de/) (27) by the basic local alignment search tool (BLASTP) algorithm for laccase-like protein.

In silico sequence analysis of Lac80 identified the three conserved copper-oxidase domains characteristic of laccases, with protein families (Pfam) database accession numbers of PF07732, PF00394 and PF07731 (Fig. 1a). Multiple sequence alignment with bacterial laccases or multicopper oxidase sequences reported in Fig. S1 revealed that Lac80 contains metal-binding amino acid residues that are conserved in bacterial laccases.

To identify the strain containing lac80, a PCR-based screening (see Materials and Methods) was performed using gDNA from the 35 strains (38). Of all the tested strains, only one yielded positive PCR results. Identification of the PCR product was corroborated by sequencing. The strain carrying the plasmid-encoded lac80 gene was identified as Ochrobactrum sp. BF15. To determine the genomic localization of lac80, an additional PCR screening was performed using the DNA bands cut from the in situ lysis gel electrophoresis (16) (Fig. 2a and Fig. 2b). Positive amplification from plasmid bands (Fig. 2c) and further sequencing confirmed that lac80 is encoded in a plasmid.

Although laccases have already been found in Ochrobactrum spp. (39-41), this is the first report of a plasmid-encoded LMCO in this genus. Ochrobactrum species are aerobic, non-fermenting, Gram-negative bacilli found in a variety of environmental sources such as water, soil, plants and animals. Genome sequencing of Ochrobactrum strains has shown the presence of multiple circular chromosomes and large plasmids (42-46). In addition, plasmid conjugation machinery has been identified in Ochrobactrum plasmids (20, 42), and horizontal gene transfer by conjugation was demonstrated in this genus (47, 48). Ochrobactrum species was previously found to degrade a wide spectrum of recalcitrant and xenobiotic compounds, such as organophosphates (49), tetrachloroethene (50), dichloro-diphenyl-trichloroethane (DDT) (51), 2,4-dichlorophenoxyacetic acid, endosulfan (52) and aniline (53), among many others. Interestingly, ligninolytic activity was also reported from Ochrobactrum species (39, 41). However, this is the first report of a plasmid-encoded functional laccase in the Ochrobactrum genus. Granja-Travez et al. (39) reported the purification and characterization of a laccase, named OcCueO, from Ochrobactrum sp. Although this laccase has the typical three-domain multicopper oxidase structure and displays activity against different substrates, alignment of OcCueO with Lac80 showed a low percentage of identity at the amino acid level (25.75%) (Fig. S2). This is not surprising, considering that OcCueO appears to be encoded in the chromosome, whereas lac80 is located in a plasmid, indicating a possible acquisition via horizontal gene transfer and consequently different evolution pathways.

The putative LMCO gene was synthesized by optimizing its codon usage for E. coli expression and removing the signal peptide sequence. The synthetic gene was then cloned into the pET22b(+) plasmid, and expressed as a His₆-tagged protein in E. coli BL21 Star^TM (DE3). Different expression conditions were explored, as detailed in Materials and Methods. Briefly, recombinant cells were grown in two dissimilarly composed cultivation media and increasing concentrations of IPTG were tested to induce Lac80 expression. Following IPTG addition, incubation under diverse temperatures and shaking regimes was compared, together with the addition of CuSO₄. Under most of the above conditions, SDS-PAGE and Western blot analyses revealed that His₆-Lac80 (molecular mass of 52.5 kDa) accumulated in the inclusion bodies (IBs) in an insoluble form. The recombinant protein was never detected in the extracellular broth, indicating that it could not be secreted. Laccase activity assays that used ABTS as substrate proved that Lac80 accumulated in the IBs was completely inactive, as generally expected for recombinant proteins packed into IBs in E. coli (12, 54). This is not surprising, as other LMCOs expressed in E. coli were described to accumulate as inactive protein into IBs. As previously reported, the addition of copper under microaerobic conditions was used to facilitate proper protein folding, a condition that facilitates copper incorporation into the LMCO active sites (55). Thus we grew E. coli BL21 Star^TM (DE3)/pET22(b)::lac80 cells in LB medium, supplemented with 0.25 mM CuSO₄ at the moment of induction, and incubated overnight at 20 °C and 100 rpm or without shaking. This procedure allowed us to obtain traces of active recombinant Lac80 in the soluble cytoplasmic fraction detectable by both Western blot analysis and activity assay (Fig. 3). Although under these conditions most of the recombinant enzyme was still packed into IBs, approx. 50 µg of soluble Lac80 per L corresponding to approx. 10.6 µg per g cells was produced in the cytoplasmatic fraction. Activity assay (see below) confirmed that this soluble Lac80 was biologically active. Therefore, it was purified by HisTag affinity chromatography, following the procedure described in Materials and Methods. As a precaution, we prepared all the buffers with Na₂SO₄ rather than with NaCl, since chlorine ions were shown to inhibit laccase activity (8). Approx. 42 µg per L culture and 9 μg per g cells of highly pure Lac80 were recovered from the elution peak, with a purification yield of approx. 84%.

We tested the ability of Lac80 to oxidize different phenolic and non-phenolic compounds that are typical LMCO substrates. Initial substrate screening was conducted at pH=5.0. Lac80 was active on five of the thirteen screened substrates: the non-phenolic ABTS (activity=1.25 U/mg) and the phenolic 2,6-DMP (activity=10.4 U/mg), pyrocatechol (activity=0.85 U/mg), pyrogallol (activity=3.2 U/mg) and vanillic acid (activity=0.87 U/mg). The highest specific activity at pH=5.0 was reported using 2,6-DMP as substrate, whereas oxidation of the other substrates was overall less efficient. Since for oxidation with LMCOs the optimal pH can vary among substrates, the activity of Lac80 on the five positive substrates was further evaluated at different pH values (from 2.0 to 8.0), as shown in Fig. 4. In all cases, the pH-dependent activity showed a bell-shaped curve, with optimum activity in the acidic range. These results show that Lac80 was active in a wide pH range: on 2,6-DMP, for instance, the recombinant LMCO was active in the pH interval from 4.0 to 8.0, whereas on pyrocatechol Lac80 showed activity from pH=3.0 to 8.0. Notably, on this latter substrate, approx. 65 and 53% of activity was also maintained at pH=7.0 and 8.0, respectively. A review of the literature shows that most of the thirteen substrates we tested with Lac80 were indeed oxidized by bacterial laccases in more than half of the reported cases (56, 57). Unlike other LMCOs, Lac80 was inactive on tyrosine, thereby excluding the possibility of being a tyrosinase.

Interestingly, the recently characterized laccase OcCueO from Ochrobactrum sp. showed activity against ABTS and DMP, two substrates that Lac80 was able to degrade (39). However, OcCueO displayed activity against guaiacol, while Lac80 showed no activity against this substrate. Given the differences at the amino acid level, these results are expectable. It is important to highlight here the prominence of Ochrobactrum species as a source of phenol oxidases.