Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Characterization of the caprolactam degradation pathway in Pseudomonas jessenii using mass spectrometry-based proteomics

Characterization of the caprolactam degradation pathway in Pseudomonas jessenii using mass... Some bacterial cultures are capable of growth on caprolactam as sole carbon and nitrogen source, but the enzymes of the catabolic pathway have not been described. We isolated a caprolactam-degrading strain of Pseudomonas jessenii from soil and identified proteins and genes putatively involved in caprolactam metabolism using quantitative mass spectrometry-based proteomics. This led to the discovery of a caprolactamase and an aminotransferase that are involved in the initial steps of caprolactam conversion. Additionally, various proteins were identified that likely are involved in later steps of the pathway. The caprolactamase consists of two subunits and demonstrated high sequence identity to the 5-oxoprolinases. Escherichia coli cells expressing this caprolactamase did not convert 5-oxoproline but were able to hydrolyze caprolactam to form 6-aminocaproic acid in an ATP-dependent manner. Characterization of the aminotransferase revealed that the enzyme deaminates 6-aminocaproic acid to produce 6-oxohexanoate with pyruvate as amino acceptor. The amino acid sequence of the aminotransferase showed high similarity to subgroup II ω- aminotransferases of the PLP-fold type I proteins. Finally, analyses of the genome sequence revealed the presence of a caprolactam catabolism gene cluster comprising a set of genes involved in the conversion of caprolactam to adipate. . . . . . Keywords Caprolactam Caprolactamase Omega aminotransferase Mass spectrometry Proteomics Pseudomonas jessenii Introduction plants (Fortmann and Rosenberg 1984;Kalinová etal. 2014, 2016). The toxic nature of these chemicals, including muta- Caprolactam is a large-volume industrial compound mainly genic effects and plant growth inhibition, has been shown in used in the production of Nylon 6, a versatile synthetic poly- different studies (Sheldon 1989; Vogel 1989). mer applied in fabrics, utensils, mechanical parts, etc. Several microorganisms are able to degrade caprolactam Synthesis of Nylon 6 is achieved by ring-opening polymeri- and/or these by-products, including strains of Pseudomonas, zation of caprolactam at high temperatures. During this pro- Alcaligenes,and Acinetobacter (Kulkarni and Kanekar 1998; cess, several side products are generated, including 6- Rajoo et al. 2013; Baxi and Shah 2002). Previously, a capro- aminocaproic acid (6-aminohexanoic acid (6-ACA)) lactam degradation pathway was proposed (Caspi et al. 2014). unreacted monomers, as well as dimers, cyclic dimers, and It starts with a caprolactam-ring cleavage to form 6-ACA, oligomers of 6-ACA. Caprolactam and these by-products are followed by the deamination to 6-oxohexanoate and subse- present as contaminants in waste water of nylon production quent oxidation to yield adipate. Adipate can be further con- verted via β-oxidation reactions of the fatty acid metabolism pathway. Whereas this pathway seems straightforward, infor- Electronic supplementary material The online version of this article mation about the proteins involved in bacterial caprolactam (https://doi.org/10.1007/s00253-018-9073-7) contains supplementary degradation is rare. At first sight, one would expect a hydro- material, which is available to authorized users. lytic enzyme involved in lactam ring opening. The conversion of the related D,L-α-amino-ε-caprolactam by a combination of * Dick B. Janssen aracemase anda L-amino acid lactam hydrolase to yield L- d.b.janssen@rug.nl lysine has been described (Payoungkiattikun et al. 2017; Biochemical Laboratory, Groningen Biomolecular Sciences and Fukumura et al. 1978; Ahmed et al. 1982, 1986). Biotechnology Institute, University of Groningen, Nijenborgh 4, Additionally, a putative hydrolase and aminotransferase have 9747 AG Groningen, The Netherlands been reported for caprolactam metabolism. Recent work 6700 Appl Microbiol Biotechnol (2018) 102:6699–6711 suggests that a dehydrogenase may oxidize 6-oxohexanoate to at 30 °C. Subsequently, cells were transferred two times to adipate in Arthrobacter sp. KI72 and in Acinetobacter sp. fresh liquid medium after which pure cultures were isolated NCIMB 9871. Subsequently, an aminotransferase might cat- on MM agarose plates supplemented with 0.2% glucose and alyze the conversion of 6-ACA to 6-oxohexanoate in the cell 0.5 mM caprolactam. To confirm growth of the resulting pure (Takehara et al. 2018; Iwaki et al. 1999). cultures on caprolactam, cells were reinoculated in selective The environmental relevance of caprolactam, and the im- liquid MM. Identification of isolated organisms was based on portance to understand the biodegradation pathway of this the analyses of the 16S rRNA gene sequence, received from synthetic compound, prompted us to investigate the biochem- the partial genome sequence obtained in this study, using the istry of the early degradation steps. In this paper, we describe a online EzTaxon database (Chun et al. 2007). P. jessenii strain newly isolated caprolactam degrading strain of Pseudomonas GO3 is deposited at DSMZ (Braunschweig, Germany) under jessenii. To investigate the pathway, and identify the enzymes accession number DSM 106008. involved, we used label-free quantitative mass spectrometry- based proteomics (van der van der Wal and Demmers 2015; Genome sequencing and annotation Fabre et al. 2014; Wasinger et al. 2013). From the various caprolactam-induced proteins, we further examined an unex- In order to obtain a partial genome sequence, total DNA from pected ATP-dependent caprolactamase that forms 6-ACA and P. jessenii GO3 cells was isolated essentially as described a class-II omega-aminotransferase that converts 6-ACA to 6- previously (Poelarends et al. 1998; Sambrook et al. 2001). oxohexanoate. The activities of the enzymes were examined The resulting genomic DNA was subjected to paired end se- with heterologously expressed proteins. Furthermore, a cap- quencing by BaseClear BV (Leiden, The Netherlands). The rolactam degradation gene cluster containing all genes for the genome sequencing was done using a HiSeq2500 system conversion of caprolactam to adipate was detected by genome (Illumina Inc., Eindhoven, The Netherlands). Sequencing of sequencing. the GO3 genome yielded 1.7 million reads of ~ 100 bp in length. These were assembled by BaseClear, using the CLC Genomics Workbench version 7.0.4 (Qiagen, Venlo, Materials and methods The Netherlands), resulting in 1274 contigs with a total length of 7 mega base pairs (Mb). This whole genome project has Growth conditions been deposited at DDBJ/ENA/GenBank under the accession PDLL00000000. The version described in this paper is Escherichia coli C41 (DE3) cells (Lucigen, Halle-Zoersel, PDLL01000000. Subsequently, all generated contigs were Belgium) were grownat37°CinanLBmedium used for automated annotation using the RAST server (Aziz (Sambrook et al. 2001). When required, ampicillin (50 μg/ et al. 2008). mL) was added. The caprolactam-degrader P. jessenii GO3 was cultured at 30 °C in nitrogen-free minimal medium Molecular techniques (MM) (Gabor et al. 2004), supplemented with caprolactam (4 mM) as carbon and nitrogen source. For the preparation Standard recombinant DNA techniques were performed es- of agar plates, the medium was supplemented with 2% agar sentially as described previously (Sambrook et al. 2001). or 1.6% H O-rinsed agarose. Restriction enzymes and polymerase were used according to To determine the caprolactam tolerance of P. jessenii GO3, the instructions of the supplier (New England Biolabs, cells were precultured in MM supplemented with 0.1% cap- Ipswich, MA, USA). Primers used in this study are listed in rolactam at 30 °C. Cells from these precultures were diluted Table 1. 50- to100-foldin 200 μL fresh medium supplemented with varying concentrations of caprolactam (0.05to1.2%). Subsequently, growth was monitored by measuring the absor- Plasmid constructions bance at OD using a microplate spectrophotometer (PowerWave, BioTek, Winooski, VT, USA). For the purification of a 6-ACA aminotransferase (PjAT), plasmid pET-PjAT was constructed containing an in-frame Enrichment of caprolactam-degrading bacteria fusion of the PjAT-encoding gene to a hexahistidine tag, be- hind the T7 promoter region. For this purpose, the PjAT gene For the isolation of bacterial strains, 1 g of residential grass- was amplified using P. jessenii GO3 chromosomal DNA and land soil from Groningen (The Netherlands) was used to in- primers ATfw and ATrev. The resulting 1419 base pair prod- oculate 50 mL nitrogen-free MM supplemented with 0.2% uct was then digested with NdeI/EcoRI and ligated into NdeI/ glucose and 0.5 mM caprolactam as sole carbon and nitrogen EcoRI-digested pET20b (Novagen-Merck, Amsterdam, source. Cultures were incubated for 7 days in an orbital shaker The Netherlands). Appl Microbiol Biotechnol (2018) 102:6699–6711 6701 Table 1 Primers used ATfw TTCCTTCTCTAGAATGAACCAGTCAGTATCCTCGC ATrev TTCCTGAATTCTTAATGATGATGATGATGATGGCCGCCCGGACCAACCCACTGAGTGGTGTC OPfw ATCAAGCTTAATGAACACAGTAGACCCGATC OPrev AAGGAAAAAAGCGGCCGCTCAATGACCGGGAGTCAGTTC For the expression of caprolactamase subunits α and β, acetonitrile, were used as the mobile phases. A gradient from plasmid pET-OP was constructed containing both genes, be- 5 to 40% acetonitrile was performed at a flow rate of 300 nL/ hind the T7 promoter region. Since both genes are likely lo- min. Eluted peptides were analyzed using a linear ion trap cated in one operon, they were amplified together using Orbitrap hybrid mass spectrometer (LTQ-Orbitrap XL, P. jessenii GO3 chromosomal DNA and primers OPfw and Thermo Scientific). MS scans were acquired in the range from OPrev. The resulting 3871 base pair fragment was then 300 to 2000 m/z. The five most intense ions per scan were digested using HindIII and NotI and ligated into HindIII/ selected for MS/MS fragmentation (35% normalized collision NotI-digested pET20b . energy) and detected in the linear ion trap. Peak lists were obtained from raw data files using Proteomics by mass spectrometry Proteome Discoverer (version 1.1, Thermo Fisher Scientific). Mascot (version 2.1, Matrix Science, London, Pseudomonas jessenii GO3 cells were cultured in 50 mL MM UK) was used for searching against the annotated P. jessenii supplemented with 0.2% glucose and 5 mM (NH ) SO or GO3 genomic DNA sequence. Peptide tolerance was set to 4 2 4 with 4 mM caprolactam. When the culture reached the late 10 ppm and the fragment ion tolerance to 2.0 Da, using exponential growth phase, cells were harvested by centrifuga- semitrypsin as protease specificity and allowing for up to tion at 3000×g for 15 min. Cell pellets were washed once with two missed cleavages. Oxidation of methionine residues, 50 mM potassium phosphate buffer, pH 7.8, and stored at − deamidation of asparagine and glutamine, and S- 20 °C prior to use. carboamidomethylation of cysteines were specified as vari- For mass spectrometry, cell pellets were resuspended in able modifications. The MS/MS-based peptide and protein 50 mM potassium phosphate buffer, pH 7.8, and lysed using identifications were further validated with the program a Vibra Cell sonicator (Sonics, Newtown, CT, USA) at 0 °C. Scaffold (version 4.6.1, Proteome Software Inc., Portland, To remove unbroken cells and cell debris, the samples were OR, USA). Peptide identifications were accepted when the centrifuged at 17,000×g for 60 min at 4 °C. Soluble protein probability was greater than 95%. Protein identifications were was precipitated by the addition of 20% trichloroacetic acid based on at least two unique peptides identified by MS/MS, (TCA). After incubating the samples for 1 h on ice, samples each with a confidence of identification probability higher were centrifuged at 17,000×g for 30 min at 4 °C. Protein than 99%. pellets were washed with ice-cold acetone to remove residual For each growth condition, at least two replicates of two TCA. Dry protein extracts were then resuspended in 50 μL independent cultures were analyzed. Normalized intensity- 50 mM NaOH. Reduction of the samples was performed with based absolute quantification (iBAQ) values from Scaffold 5 μL of 500 mM dithiothreitol (DTT) in 350 μL100 mM were used as a measure for the abundance of the identified NH HCO for 30 min at 25 °C, followed by derivatization proteins. Average iBAQ values were calculated for the differ- 4 3 of sulfhydryls by 30 min incubation at room temperature with ent samples and subsequently log2 transformed. In case the 10 μL of 550 mM iodoacetamide. Trypsin digestion was per- protein was not detected, the log2-transformed iBAQ value formed overnight at 37 °C by addition of 4 μg trypsin gold was manually set to 8.5, 2.5-fold below the approximate limit (mass spectrometry grade, Promega (Leiden, of detection. The effect of growth conditions on specific pro- The Netherlands)), followed by a second trypsin digestion tein amounts was calculated by dividing the average log2 for 3hat 37°Cusing 1 μg trypsin gold. Samples were pre- iBAQ value for each protein in extracts from caprolactam- pared for injection by addition of 2.5% formic acid. grown cells by the corresponding iBAQ value in protein ex- For LC-MS, peptides were first trapped on a precolumn tracts from control cells. A protein was considered upregulat- (EASY-Column C18, 100 μm × 20 mm, 5 μmparticle size, ed when the log2-fold ratio was more than two and downreg- Thermo Scientific, Ermelo, The Netherlands) and separated ulated when the log2-fold ratio was less than 0.5. on a capillary column (C18 PepMap 300, 75 μm×100 mm, The mass spectrometry proteomics data have been depos- 3-μm particle size, Thermo Scientific) mounted on a Proxeon ited to the ProteomeXchange Consortium via the PRIDE Easy-nLCII system (Thermo Scientific). Solutions of 0.1% (Vizcaíno et al. 2016) partner repository with the dataset iden- tifier PXD008544 and https://doi.org/10.6019/PXD008544. formic acid in water, and 0.1% formic acid in 100% 6702 Appl Microbiol Biotechnol (2018) 102:6699–6711 Expression and purification Standard reaction mixtures contained 100 mM potassium phosphate buffer (pH 8), 5 U/mL alanine dehydrogenase, The aminotransferase PjAT and caprolactamase were both 2mM NAD , 0.05 mM PLP, 2 mM substrate, 0.2 mM pyru- produced in E. coli C41 cells under control of the T7 promot- vate, and varying concentrations of enzyme in a total volume er. PjAT was expressed and purified as previously described of 300 μL in flat-bottom 96-well microtiter plates. Reactions for related aminotransferases (Palacio et al. 2016). were carried out at 30 °C and analyzed using a microtiter plate For the expression of the α and β subunits of reader (Synergy Mx Microplate Reader, BioTek Instruments, caprolactamase, 0.5 mL of an overnight grown LB culture of Bad Friedrichshall, Germany). Reaction mixtures lacking py- pET-OP transformed cells was transferred to 50 mL ruvate (150 μL) were prewarmed before the reaction was ini- autoinduction medium (ForMedium) containing ampicillin tiated by the addition of 150-μL pyruvate solution. Each re- and incubated for 48 h in a rotary shaker at 17 °C. To prepare action was analyzed in triplicate. Initial rates were used to cell-free extracts, cells were washed in buffer A (50 mM am- determine specific activities in units per mg protein monium bicarbonate, pH 8.5, 10 mM MgCl ), resuspended in (μmol/min/mg). Protein content was determined using the buffer A, and lysed using a Sonics Vibra Cell sonicator at Bradford method, with bovine serum albumin as the standard. 0 °C. To remove unbroken cells and cell debris, the samples The amination of 6-oxohexanoate with alanine as donor were centrifuged at 17,000×g for 30 min. was also followed by coupling to alanine dehydrogenase. The reaction mixtures contained 100 mM potassium phos- Enzyme kinetics phate buffer (pH 8), 2 mM substrate, 0.1 mM NADH, 0.05 mM PLP, 8 U/mL alanine dehydrogenase, 5 mM ammo- To analyze aminotransferase activity in cell-free extracts, re- nium bicarbonate, 5 mM L-alanine, and varying concentra- actions were followed using HPLC analyses. Cell-free ex- tions of enzyme in a total volume of 300 μL. Reactions were tracts were prepared in 50 mM potassium phosphate buffer initiated by addition of 150 μLof L-Ala and carried out as (pH 8) containing 0.3 mM pyridoxal 5'-phosphate (PLP). described previously. Conversion was followed by measuring Standard reaction mixtures contained 50 mM potassium phos- NADH depletion at 340 nm. phate buffer (pH 8), 2 mM amine donor (pyruvate or α- Caprolactamase activity was determined by analyzing re- ketoglutarate), 5 mM 6-ACA, 0.3 mM PLP, and cell-free ex- action mixtures using an Acquity TQD mass spectrometer tract, in a total volume of 300 μL. Reactions were carried out (Waters, Etten-Leur, The Netherlands). Standard reaction mix- at 28 °C. With different time intervals, 50 μL samples were tures contained 2 mM substrate (caprolactam or 5- taken and quenched by the addition of 50 μL2MHCl. After oxoproline), 5 mM ATP, 10 mM MgCl , and cell-free extract, incubating the samples for 10 min on ice, samples were cen- in 50 mM ammonium bicarbonate, pH 8.5. Samples were trifuged for 10 min at 17,000×g and neutralized using 100 μL taken and quenched by the addition of 2% formic acid. 1 M NaOH. Amino acids (6-ACA, alanine, glutamate) in the Separation of the reaction content was performed by UPLC reaction mixtures were quantified by precolumn o- using a Waters Acquity UPLC HSS T3 Column (1.8 μm, phthalaldehyde (OPA) derivatization and subsequent HPLC 2.1 × 150 mm) and a linear gradient (eluent A: water, 0.1% analyses. To this purpose, 10-μL sample was mixed with formic acid; eluent B: 100 acetonitrile, 0.1% formic acid). The 40 μL 0.4 M boric acid (pH 9.7) and 10 μL OPA solution samples were analyzed in positive ion mode. To determine (Fisher Scientific) and incubated for 20 min at 30 °C. Then, substrate reduction and product formation, multiple reaction 3 μL of the reaction sample were injected by an autosampler monitoring (MRM) was performed, measuring the following and analyzed by HPLC using a C18 OPA Adsorbosphere fragments: caprolactam m/z =96; 6-ACA m/z =114; 5- column connected to a Jasco FP-920 detector (excitation oxoproline m/z = 84; glutamate m/z =102. 350 nm; emission 450 nm). Compounds were eluted using a linear gradient (eluent A, 20 mM sodium acetate, pH 7.2, 0.5% (vol/vol) tetrahydrofuran, 0.018% (vol/vol) TEA and eluent B, 90% acetonitrile) at a flow rate of 0.5 mL/min. Results The activity of purified PjAT with pyruvate as the amine acceptor was estimated by coupling the reaction to alanine Isolation of bacterial strains using caprolactam dehydrogenase and measuring the increase in NADH absor- as a sole nitrogen source bance that occurs as a result of oxidative deamination of the produced alanine. Since pyruvate is a competitive inhibitor of In order to isolate a bacterial strain possessing a caprolactam alanine dehydrogenase, low pyruvate concentrations were metabolism pathway, soil microorganisms were enriched for used to minimize the lag time of the reaction. The concentra- the ability to grow on caprolactam as sole nitrogen source. A tion of alanine dehydrogenase in the assays was in excess pure culture was obtained by repeated transfer to fresh medi- (5 units/mL), to give aminotransferase-dependent velocities. um plates. The bacterial strain that was growing best on Appl Microbiol Biotechnol (2018) 102:6699–6711 6703 0.4 selective medium was used to study caprolactam metabolism in detail and was designated strain GO3. 0.3 Growth analyses revealed that strain GO3 was able to use caprolactam both as a sole nitrogen and carbon source. 0.2 Interestingly, 6-aminohexanoic acid (6-ACA), a described in- termediate in caprolactam degradation (Caspi et al. 2014), was 0.1 not a possible growth substrate for this strain. In mineral me- dium (MM) supplemented with 0.05% caprolactam as sole 00.2 0.4 0.6 carbon and nitrogen source, the calculated μ was max −1 Caprolactam (%) 0.37 h . Growth analyses using different concentrations of Fig. 1 Specific growth rates of P. jessenii GO3 in media supplemented caprolactam in MM revealed that the μ is reduced by max with different concentrations of caprolactam higher concentration of caprolactam, with a calculated critical caprolactam concentration of 0.46% (Fig. 1). A total of 137 different proteins were identified in the com- bined independent replicate experiments, corresponding to Draft genome sequence 2.2% of the predicted P. jessenii proteome. Among these, 109 proteins were identified in both experiments and were After paired end sequencing of the genomic DNA from strain subjected to further bioinformatic analysis (Fig. 2a). GO3, genome assembly resulted in 1274 contigs (N50: Seventeen of these proteins showed at least a 2-fold increase 10,682 bp), covering 6,993,317 bp. The GC content is 60% in log2 protein abundance in the caprolactam-grown cells as with 6231 predicted coding sequences. Among these predict- compared to the glucose cultures (Fig. 2b, Table 2;Supporting ed genes, 4754 were assigned a predicted function (76%). information Table S1 and Table S2). Interestingly, some pro- Furthermore, 3 rRNA and 61 tRNA genes were identified in teins are highly induced on caprolactam, while others are just the draft genome. above the level of detection (LOD ~ 21, Fig. 2c). Conversely, Analyses of the 16S rRNA gene revealed that the organism in glucose-grown cells, these proteins were below the level of is a Pseudomonas species closely related to P. jessenii CIP detection (iBAQ < 21, data not shown). 105274 (99.5% identity). The draft genome was compared to related Pseudomonas species of which the complete se- Hypothetical caprolactam degradation pathway quence is published, including the caprolactam-degrading or- ganism Pseudomonas mosselii SJ10 (Park et al. 2014) Based on the identified caprolactam-induced proteins (Table 2). Previous studies showed that in other (Table 3), in combination with previous data (Esikova caprolactam-utilizing Pseudomonas strains, the genes in- et al. 2012), a complete putative P. jessenii caprolactam volved in caprolactam metabolism are plasmid localized degradation pathway was built, including all enzymes that (Boronin et al. 1984). Using gel electrophoresis of DNA ex- play a role in the pathway (Fig. 3). To enable growth on tracts, we did not find a plasmid in P. jessenii GO3 (data not caprolactam as sole carbon and nitrogen source, an active shown), suggesting a chromosomal location of the catabolic uptake of caprolactam might be needed, which may be genes. dependent on ABC transporter proteins. Four ABC trans- porter substrate binding proteins are significantly induced Identification of caprolactam degradation enzymes during growth on caprolactam, including ORF1056, ORF3044, ORF1114, and ORF2532 (Table 3). Database A hypothetical caprolactam degradation pathway (Esikova searches demonstrated homology to various ABC trans- et al. 2012) involves two unidentified enzymes: the ring- porters, including the spermidine/putrescine binding pro- cleavage enzyme, presumably a hydrolase, and the enzyme teins (ORF1056 and ORF3044), the branched chain ami- involved in the deamination reaction, which could be an ami- no acid binding proteins (ORF1114), and the amino acid notransferase, an oxidase, or an amine dehydrogenase, all binding proteins (ORF2532). Interestingly, ORF1114 and three producing an ω-ketoacid. To identify the enzymes in- ORF2532 were predominantly induced when the cells volved in the pathway, the P. jessenii GO3 proteome was were grown on caprolactam as sole carbon and nitrogen examined for caprolactam-induced proteins. To this purpose, source (Fig. 2c). Possibly, one or several of these pro- P. jessenii cells were grown in minimal medium supplemented tein(s) are involved in the uptake of caprolactam. with caprolactam (4 mM), or glucose plus ammonium sulfate. Inside the cell caprolactam is most likely converted to 6- Cell-free extracts were prepared from both cultures and sub- ACA. This conversion might be dependent on a jected to quantitative proteome analysis using a label-free caprolactamase, catalyzing the opening of the lactam ring. approach. Inspection of the caprolactam-induced proteins for a -1 µ (h ) max 6704 Appl Microbiol Biotechnol (2018) 102:6699–6711 Table 2 General genomic features of various Pseudomonas species General features Pj GO3 Pb NFM421 Pk 1855–344 Pm SJ10 Size (Mb) 7.0 6.8 6.8 6.2 GC (%) 60.0 60.8 60.7 63.4 CDS 6231 6097 5856 5413 Protein with predicted function (%) 76.3 The genome data are adopted from the original papers. These numbers may differ from numbers obtained with updated annotations CDS coding sequences, Pb Pseudomonas brassicacearum NFM421 (Ortet et al. 2011), Pk Pseudomonas kilonensis 1855-344 (Eng et al. 2015), Pm P. mosselii SJ10 (Park et al. 2014) lactamase-related protein resulted in the identification of two 5-oxoproline, hydantoin, and caprolactam share a lactam moi- distinct polypeptides, ORF4270 and ORF4271. Database ety, it is likely that proteins encoded by these two ORFs are searches revealed identity to subunits A and B of the involved in the caprolactamase reaction. hydantoinase from Pseudomonas sp. (Table 3,respectively, Further conversion of 6-ACA can proceed through deam- 32 and 30% identity) and to the putative 5-oxoprolinase sub- ination catalyzed by an aminotransferase, producing 6- units A and B from Pseudomonas putida (OplA, 78% identity oxohexanoate. Inspection of the caprolactam-induced proteins and OplB, 86% identity). The ORF4270 and ORF4271 resulted in the identification of a protein with homology to the encoded sequences also displayed weak similarity to eukary- Vibrio fluvialis ω-amino acid aminotransferase (ORF4266, otic 5-oxoprolinases which are known to catalyze the ATP- 43% identity). The well-characterized V. fluvialis enzyme cat- dependent hydrolytic decyclization of 5-oxoproline, produc- alyzes the pyruvate-dependent transamination of ω-amino ing L-glutamate (Saccharomyces cerevisiae OXP1/YKL215c, acids and other amines to aldehydes or ketones (Shin et al. 22 and 23% identity, respectively) (Seddon et al. 1984). Since 2003). Fig. 2 Proteomic analysis of A B P. jessenii GO3 cells, obtained from two independent replicate 100 Total cultures. A Number of P. jessenii proteins identified by mass Downregulated <0.5 spectrometry in replicate culture 1 (red) and replicate culture 2 Upregulated >2 (blue). The overlapping region represents the number of proteins identified in both experiments (purple). B Bar graph representing the total number of identified P. jessenii proteins (green), including several caprolactam-induced proteins (red) and several caprolactam- repressed proteins (blue). A protein was considered upregulated when the protein level in caprolactam-induced cells divided by the protein level in non-induced cells was larger than two; a protein was considered downregulated when this ratio was smaller than 0.5. C Bar graph representing the average iBAQ log2 protein amounts of the identified upregulated proteins in caprolactam-grown cells. For all these hits, the protein amounts in non-induced cells were below the level of detection (not depicted in this plot) (Color figure online) No. of idenfied proteins Appl Microbiol Biotechnol (2018) 102:6699–6711 6705 Table 3 Caprolactam-induced proteins Name Contig Size (bp) Enzyme Seq. identity to known EC number Accession proteins (%, organism) number 4265 5 1451 Succinate-semialdehyde dehydrogenase 81, E. coli 1.2.1.24 3JZ4 4266 5 1367 Omega aminotransferase 43, V. fluvialis 2.6.1.18 3NUI 4270 5 2114 Hydantoin utilization protein A 32, Pseudomonas sp. 3.5.2.- Q01262 4271 5 1745 Hydantoin utilization protein B 30, Pseudomonas sp. 3.5.2.- Q01262 4277 5 1238 Acetyl-CoA:oxalate CoA-transferase 39, E. coli 2.8.3.19 4HL6 4278 5 1202 Acetyl-CoA acetyltransferase 41, Ralstonia eutropha 2.3.1.9 4O99 4279 5 1541 3-Hydroxyacyl-CoA dehydrogenase 39, R. eutropha 1.1.1.35 4PZC 4282 5 1154 Acyl-CoA dehydrogenase 42, Thermus thermophilus 1.3.99.2 2DVL 2532 28 1031 ABC transporter, amino acid binding protein 59, Brucella ovis 4Z9N 3504 39 1922 Serine protein kinase 77, E. coli 2.7.11.1 P0ACY5 4150 48 2147 Fatty acid oxidation complex α subunit 94, P. fragi 4.2.1.17 5.3.3.8 1WDK 1.1.1.35 5.1.2.3 1056 157 1112 ABC transporter, putrescine binding protein 75, P. aeruginosa 3TTM 1114 161 1133 ABC transporter, branched chain amino 51, Agrobacterium fabrum 3IP5 acid binding protein 3044 330 2549 ABC transporter, Spermidine/putrescine 31, Streptococcus pneumoniae 4EQB binding protein 3301 363 344 Putative enzyme of the cupin superfamily 98, Pseudomonas sp. WP_ 4187 485 1325 Isocitrate lyase 82, E. coli 4.1.3.1 1IGW 5740 814 905 Histone H1-like protein HC2 43, Chlamydia pneumoniae Q9Z8F9 The product 6-oxohexanoate would then be converted to hydratase function might be performed by a protein with ho- adipate. In Acinetobacter, this reaction is catalyzed by a 6- mology to the Pseudomonas fragi fatty acid oxidation com- oxohexanoate dehydrogenase (Iwaki et al. 1999). Among plex α subunit (ORF4150). In Pseudomonas, this complex the caprolactam-induced proteins, one protein (ORF 4265) catalyzes multiple reactions of the beta fatty acid oxidation, was identified with homology to the E. coli succinate- including the enoyl-CoA hydratase and the 3-hydroxyacyl- semialdehyde dehydrogenase. In E. coli, this enzyme converts CoA dehydrogenase (Ishikawa et al. 2004). succinate semialdehyde to succinate, which is part of the bio- degradation of 4-aminobutyric acid (Donnelly and Cooper Gene organization 1981). Since succinate-semialdehyde and 6-oxohexanoate are structurally similar, it is plausible that this enzyme is in- The sequence information for most caprolactam-induced pro- volved in the conversion of 6-oxohexanoate. Additionally, teins was found on contig 5, a large segment of 40,376 bp this protein has 38% sequence identity to the Acinetobacter containing 36 putative open reading frames (ORF4261 to 6-oxohexanoate dehydrogenase. ORF4296). The genetic context of this contig was analyzed Finally, adipate most likely enters the β-oxidation pathway using the RAST server. This revealed that contig 5 comprises for degradation of fatty acids, which consists of multiple re- two gene clusters involved in the caprolactam degradation. actions (Janßen and Steinbüchel 2014). First adipate needs to The first gene cluster contains the genes putatively in- be activated by CoA, resulting in adipyl-CoA. This reaction volved in the conversion of caprolactam to adipate, including might be achieved by the caprolactam-induced protein with both subunits of the proposed caprolactam-induced homology to the E. coli acetyl-CoA:oxalate CoA-transferase caprolactamase (ORF4270, ORF4271), the omega amino- (ORF4277). In E. coli, this enzyme catalyzes the reversible transferase (ORF4266), and a 6-oxohexanoate dehydrogenase conversion of oxalate and acetyl-CoA to oxalyl-CoA and ac- (ORF4265) (Fig. 4a). In between these genes, three other etate. Then, adipyl-CoA can be converted in a multistep pro- ORFs are located (Fig. 4a, Pj, in blue). BLAST searches re- cess by means of an adipyl-CoA dehydrogenase (ORF4282), vealed that these open reading frames encode proteins with an enoyl-CoA hydratase, an 3-hydroxyadipyl-CoA dehydro- high homology to the L-2-hydroxyglutarate oxidase LhgO of genase (ORF4279), and a 3-ketoadipyl-CoA thiolase E. coli (ORF4267, 72% identity), the starvation induced pro- (ORF4278). Homologs of all of these proteins were clearly tein CsiD of P. putida (ORF4268, 75%), and the transcription- induced in caprolactam-grown cells, where the enoyl-CoA al regulator CsiR from E. coli (ORF4269, 59%). In E. coli, 6706 Appl Microbiol Biotechnol (2018) 102:6699–6711 Fig. 3 Hypothetical pathway for the biodegradation of caprolactam in P. jessenii cells. A The conversion of caprolactam to adipate. B The first steps of the β- fatty acid degradation these genes cluster together with gabD, gabT,and gabP in- supplemented with 6-ACA as sole nitrogen source; (2) the volved in the conversion of γ-aminobutyrate to succinate. The absence of a homolog of the caprolactamase genes gabD gene encodes a succinate-semialdehyde dehydrogenase ORF4270 and 4271 in E. coli, which appeared to be involved (DH), gabT a γ-aminobutyrate aminotransferase (AT) and in the first step in caprolactam degradation in P. jessenii gabP a γ-aminobutyrate permease. GabT and GabD represent (Fig. 4a, genes labeled with *). similar catalytic activities as the P. jessenii omega aminotrans- The second gene cluster includes genes involved in the ferase (ORF4266) and the 6-oxohexanoate dehydrogenase fatty acid β-oxidation, including the caprolactam-induced ace- (ORF4265), respectively. Clear differences between both gene tyl-CoA:oxalate CoA-transferase (ORF4277), acetyl-CoA clusters include (1) the absence of a homolog of the γ- acetyltransferase (ORF4278), 3-hydroxyacyl-CoA dehydro- aminobutyrate permease gabP in P. jessenii, which might ex- genase (ORF4279), and butyryl-CoA dehydrogenase plain why P. jessenii is not able to grow on minimal medium (ORF4282). In between these genes, two more open reading Appl Microbiol Biotechnol (2018) 102:6699–6711 6707 frames are located which according to BLAST searches en- revealed that 5-oxoproline is not a substrate for the identified code for proteins with high homology to an enoyl-CoA P. jessenii caprolactamase. hydratase from Mus musculus (ORF4280, 47% identity) and an IclR family transcriptional regulator from Pseudomonas Characterization of the omega aminotransferase testosteroni (ORF4281, 36%). A similar gene cluster contain- ing homologs of all six genes is present in the genome of other Most omega aminotransferases are PLP-fold type I enzymes bacteria (e.g., Pseudomonas aeruginosa PAO1), suggesting a that catalyze the transfer of an amino group from a β-, γ-or wider occurrence of adipate metabolism by the same pathway. other ω-amino acid or an amine to pyruvate or α- ketoglutarate (Schiroli and Peracchi 2015). In order to confirm ATP-dependent caprolactamase thepresenceof6-ACA ω-aminotransferase activity in P. jessenii GO3, a protein extract of caprolactam-induced cells To confirm the presence of an ATP-dependent caprolactamase was prepared and incubated with 6-ACA and pyruvate or α- activity in P. jessenii, a cell-free extract of caprolactam- ketoglutarate, and levels of 6-ACA and produced alanine or induced cells was prepared. Subsequently, the extract was glutamate were determined using OPA derivatization and incubated in the presence of caprolactam, ATP, and MgCl , HPLC. This revealed that caprolactam was indeed enzymati- and the production of 6-ACA was examined by UPLC-MS. cally converted with pyruvate as the amino acceptor. In the This revealed that caprolactam is indeed enzymatically con- absence of pyruvate, no conversion of 6-ACA was detected. verted to 6-ACA. When a similar assay was performed using When a similar assay was performed using cell-free extract of the cell-free extract of glucose-grown P. jessenii cells, no con- glucose-grown P. jessenii cells, no conversion of 6-ACA or version of caprolactam to 6-ACA was detected. This con- pyruvate was found. This confirmed the presence of a firmed that the caprolactamase activity present in P. jessenii caprolactam-inducible ω-transaminase activity in P. jessenii cells is induced during growth on caprolactam. Additionally, GO3 cells. to study the ATP dependence of the putative caprolactamase To establish if this ORF4266-encoded putative ω-amino in P. jessenii cells, cell-free extracts of caprolactam-grown acid aminotransferase (PjAT) is responsible for the conversion cells containing MgCl were prepared and tested for the for- of 6-ACA, activity assays were performed using E. coli- mation of 6-ACA in the absence of ATP. No 6-ACA formation expressed PjAT. Since E. coli contains multiple aminotrans- was detected, demonstrating that ATP indeed is required for ferases, PjAT was equipped with a C-terminal His6-tag, and the enzymatic hydrolyses of caprolactam to 6-ACA. the enzyme was purified using a Ni-NTA resin. This yielded To confirm the role of the putative caprolactamase ca. 30–35 mg purified enzyme/L of culture. The activity of the (ORF4270, ORF4271) in the conversion of caprolactam, the enzyme was examined in both directions, so with 6- genes were expressed in E. coli C41. To express both subunits oxohexanoate or with 6-ACA as substrate, together with L- (CapA, CapB) simultaneously, the entire operon including the alanine or pyruvate as amino acceptor. Activities were obtain- 36-bp intergenic region (Fig. 4) was cloned under control of a ed by coupling the reaction to that of alanine dehydrogenase single T7 promoter. Expression in E. coli resulted in the high- and following production or consumption of NADH spectro- level production of two proteins of the expected size, of which photometry at 340 nm. This revealed a specific activity of approximately 60% was present in the soluble fraction. 0.2 U/mg using 6-ACA and pyruvate as the substrates and UPLC-MS analysis was performed to detect the formation an activity of 4.5 U/mg using 5-oxohexanoate and alanine as of 6-ACA in a mixture containing caprolactam, ATP, MgCl , the substrates. These activities would suffice to enable strain and E. coli cell-free extract. Time course analyses demonstrat- GO3 to use caprolactam as a nitrogen source for growth. ed that 6-ACA levels increased and caprolactam levels de- creased in time. A specific activity of 0.14 U/mg was calcu- lated for the E. coli cell-free extract (Fig. 5). In reaction mix- tures containing caprolactam, ATP, MgCl ,and E. coli extract Discussion from cells not producing the caprolactamase, no detectable 6- ACA was observed even after 3 h of incubation. Thus, the In this work, we explored the caprolactam degradation path- caprolactamase activity detected in the induced E. coli (pET- way in the bacterium P. jessenii strain GO3. Previous studies OP) extract originates from the expressed α and β subunits of suggested an overall catabolic pathway for caprolactam me- the caprolactamase. Since the sequence of the α and β sub- tabolism, but the enzymes catalyzing the first two steps, i.e., units showed homology to enzymes annotated as 5- conversion of caprolactam to 6-oxohexanoate, remained un- oxoprolinase, reaction mixtures containing 5-oxoproline, known. Using proteomic studies, we identified an ATP- ATP, MgCl ,and E. coli cell-free extract were tested for the dependent lactamase involved in the conversion of caprolac- production of glutamate. UPLC-MS analyses demonstrated tam to 6-ACA and an ω-aminotransferase responsible for the no detectable glutamate even after 3 h of incubation. This subsequent conversion of 6-ACA to 6-oxohexanoate. 6708 Appl Microbiol Biotechnol (2018) 102:6699–6711 4271 4270 4269 4268 4267 4266 4265 capA* csiR csiD ygaF AT DH capB* Pj csiD ygaF gabD gabT gabP* csiR Ec 4282 4280 4278 adipyl-CoA enoyl-CoA acetyl-CoA dehydrogenase hydratase acetyltransferase Pj 4281 4277 transcriptional acetyl-CoA:adipate 3-hydroxyacyl-CoA regulator CoA transferase dehydrogenase Fig. 4 Schematic representation of the gene organization of contig 5, between the P. jessenii and E. coli gene clusters. b Gene cluster containing most of the caprolactam-induced genes, analyzed using the comprising the genes putatively involved in the fatty acid β- RAST server. a Gene cluster comprising the genes putatively involved degradation of adipate. Genes represented in blue: P. jessenii genes in the conversion of caprolactam to adipate in P. jessenii (Pj) and a similar encoding proteins that were not significantly upregulated (Color figure gene cluster with the genes involved in the conversion of γ-aminobutyrate online) in E. coli (Ec). Genes marked with an asterisk indicate clear differences Additionally, we identified various other enzymes and genes (Seddon and Meister 1986;Liet al. 1988). Subunit A, which putatively involved in the caprolactam catabolic pathway. is homologous to the N-terminal part of the 137 kDa subunit The ATP-dependent lactamase catalyzing caprolactam ring of the eukaryotic enzymes, catalyzes the phosphorylation of opening was identified by proteomic studies, sequence com- enzyme-bound 5-oxoproline, whereas subunit B, which is ho- parison to known proteins with lactamase activity, and func- mologous to the C-terminal part of eukaryotic oxoprolinase, is tional expression in E. coli. Sequence comparison indicated required for hydrolysis of the phosphorylated hydroxypyrrole. similarity to proteins annotated as 5-oxoprolinase or Other enzymes with significant sequence similarity to the hydantoinase. Oxoprolinases catalyze the ATP-dependent hy- caprolactamase are the ATP-dependent hydantoinases. Only drolysis of 5-oxoproline to glutamate (EC 3.5.2.9). The se- the N-terminal sequences of the two subunits of the enzyme quence of only four oxoprolinases with confirmed activity isolated from P. putida 77 have been reported (Ogawa et al. has been reported, including the enzymes from rat and cow 1995), but BLAST searches allow retrieval of complete se- (Ye et al. 1996; Watanabe et al. 2004), S. cerevisiae (Kumar quences from Marinobacterium profundum and Bachhawat 2010), and the human enzyme. The latter can (WP_067296627.1 and WP_067296623.1). Sequence align- harbor mutations of medical relevance (Calpena et al. 2015). ments show that in addition to the A and B subunits of the The eukaryotic 5-oxoprolinases are homodimers with sub- bacterial oxoprolinases mentioned previously, also the A and units of approximately 137 kDa, whereas the P. putida 5- B subunits of these putative hydantoinases and the α and β oxoprolinase, of which the sequence is not reported, consists subunits of caprolactamase described here correspond to the of two different subunits of approximately 75 and 63 kDa N-terminal part and C-terminal part, respectively, of the eu- karyotic oxoprolinases. All these ATP-dependent hydrolases belong to InterPro families IPR002821 and IPR003692. We 3.0 demonstrated the ATP dependence of the caprolactamase ac- 2.5 tivity using enzyme expressed in E. coli. Within the families, 2.0 there are clear differences. For example, the substrate range of the P. jessenii caprolactamase lacks 5-oxoprolinase activity. It 1.5 seems possible that many bacterial genes annotated in 1.0 GenBank as 5-oxoprolinase actually are (capro)lactamase 0.5 genes or cyclic dipeptide hydrolase genes, since there is a higher sequence similarity to the CapA and CapB ORFs than 0.0 0 20406080 100 120 to confirmed oxoprolinases. Time (min) Based on sequence and structural analysis, these Fig. 5 UPLC-MS analyses monitoring the formation of 6-ACA (black lactamases can be further grouped with the ATP-dependent line, triangles) and the degradation of caprolactam (grey line, closed carboxylases/lactamase superfamily, which includes carbox- circles). The reaction mixtures contained 2 mM caprolactam, 5 mM ylases acting on acetone and acetophenone (Weidenweber ATP, 10 mM MgCl ,and 75 μg/mL cell-free extract, in 50 mM ammonium bicarbonate, pH 8.5 et al. 2017). The structure of the acetophenone carboxylase Concentration (mM) Appl Microbiol Biotechnol (2018) 102:6699–6711 6709 from Aromatoleum aromaticum EbN1 was recently solved and CapB 99.3% sequence identity). This suggests the same (pdb 519w), revealing its quaternary structure as (αα′βγ) . caprolactam degradation pathways in both Pseudomonas spe- Sequence motifs indicative of ATP binding are conserved be- cies. Additionally, BLASTsearches against the non-redundant tween the α subunit of the carboxylase and the α subunit of protein database revealed highly similar open reading frames caprolactamase (ORF 4270), and the β subunit of the carbox- in annotated sequences of various other Pseudomonas strains, ylase is homologous to the β subunit of caprolactamase (ORF indicating that the ability to hydrolyze lactams or cyclic pep- 4271). Thus, the elucidation of the caprolactam catabolism tides may not be unusual in Pseudomonas.Other nylonby- described here adds a new member to this diverse group of products that can be degraded by microorganisms include 6- ATP-dependent hydrolytic enzymes. By similarity to ACA dimers, cyclic dimers, and oligomers. Metabolism is oxoprolinase and hydantoinase, caprolactam hydrolysis by dependent on hydrolases, such as NylA, NylB, and NylC, the caprolactamase is expected to proceed by phosphorylation which have different specificities (Negoro 2000). NylA cata- of the enol (lactim) tautomer mediated by the α subunit, lyzes the hydrolysis of 6-ACA cyclic dimer, resulting in the followed by its hydrolysis with a role for the β subunit. formation of the 6-ACA dimer, which can be converted by Structures of lactamases that would provide detailed insight NylB, generating 6-ACA. NylC catalyzes the hydrolysis of 6- are lacking, however. ACA oligomers. BLAST searches using Flavobacterium sp. The 6-aminohexanoate aminotransferase (PjAT) catalyzing NylA, B, and C revealed that P. jessenii contains homologs of the second step in caprolactam catabolism was also identified NylA and NylB, but not of NylC, suggesting the absence of a by proteomic studies, sequence similarities, and functional complete pathway for metabolism of 6-ACA polymers. overexpression in E. coli. BLAST searches using the protein sequence of the PjAT demonstrated relatedness to the fold- Acknowledgements The authors acknowledge M. Schürman and S. Turk type I PLP enzymes, more in particular to the subgroup II (DSM, The Netherlands) for their support and helpful discussions and T. aminotransferases, which are now often grouped as class III Tiemersma-Wegman for assistance with the MS analysis. aminotransferases (Schiroli and Peracchi 2015;Steffen- Munsberg et al. 2015). These enzymes catalyze conversion Author’s contributions MO performed the proteomic and genomic stud- ies. MO and CP performed bioinformatic analysis and expression exper- of ω-amino acids to aldehydes (EC 2.6.1.18). In vitro charac- iments. MO, CP, and DJ wrote the paper. terization of PjAT confirmed the expected deamination of 6- ACA to 6-oxohexanoate. The enzyme was also active in the Funding information This research was supported financially by the reverse reaction: L-alanine-dependent amination of 6- Dutch Ministry of Economic Affairs and BE-Basic (www.be-basic.org), aminohexanoate. This conversion is of potential interest for a public/private research organization. a biosynthetic 6-ACA production pathway that was recently engineered into E. coli (Turk et al. 2015). Comparison with Compliance with ethical standards the V. fluvialis aminotransferase, which is the closest well- studied homolog, revealed conservation of several residues Ethical statement This article does not contain any studies with human participants or animals performed by any of the authors. important for proper activity (PDB code 4E3Q; (Midelfort et al. 2013)). The lysine required for the formation of the Conflict of interest The authors declare that they have no conflict of internal aldimine (Schiff base) with the PLP cofactor is present interest. at a conserved position. Further work on the biocatalytic and structural properties of the enzyme is ongoing. The PjAT ami- Open Access This article is distributed under the terms of the Creative notransferase has 27.3% pairwise sequence identity with the Commons Attribution 4.0 International License (http:// NylD1 aminotransferase recently described in the nylon olig- creativecommons.org/licenses/by/4.0/), which permits unrestricted use, omer degrader Arthrobacter sp. KI72 (Takehara et al. 2018). distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link The closest homolog of the latter enzyme in the P. jessenii to the Creative Commons license, and indicate if changes were made. genome is ORF2380 (48.6% pairwise identify), but an upreg- ulation of this protein was not detected by proteomic analysis of caprolactam-grown P. jessenii cultures. Previously, the full genome of P. mosselii, another capro- References lactam degrading organism isolated from wastewater of a ny- lon producing industrial complex in Korea, was sequenced by Ahmed SA, Esaki N, Soda K (1982) Purification and properties of α- amino-ϵ-caprolactam racemase from Achromobacter obae.FEBS Park and coworkers (Park et al. 2014). Interestingly, BLAST Lett 150:370–374 searches using the sequences of the genes that were found here Ahmed SA, Esaki N, Tanaka H, Soda K (1986) Mechanism of α-amino- to be involved in caprolactam degradation against the full ε-caprolactam racemase reaction. Biochemistry 25:385–388 P. mosselii genome sequence revealed genes with high simi- Aziz RK, Bartels D, Best AA, DeJongh M, Disz T, Edwards RA, larity to these proteins (aminotransferase 100%, CapA 99.6%, Formsma K, Gerdes S, Glass EM, Kubal M, Meyer F, Olsen GJ, 6710 Appl Microbiol Biotechnol (2018) 102:6699–6711 Olson R, Osterman AL, Overbeek RA, MacNeil LK, Paarmann D, Kalinová JP, Tříska J, Vrchotová N, Novák J (2016) Uptake of caprolac- tam and its influence on growth and oxygen production of Paczian T, Parrello B, Pusch GD, Reich C, Stevens R, Vassieva O, Vonstein V, Wilke A, Zagnitko O (2008) The RAST server: rapid Desmodesmusquadricauda algae. Environ Pollut 213:518–523 annotations using subsystems technology. BMC Genomics 9:75 Kulkarni RS, Kanekar PP (1998) Bioremediation of ε-caprolactam from Baxi N, Shah A (2002) ε-Caprolactam-degradation by Alcaligenes Nylon-6 waste water by use of Pseudomonas aeruginosa MCM faecalis for bioremediation of wastewater of a nylon-6 production B-407. Curr Microbiol 37:191–194 plant. Biotechnol Lett 24:1177–1180 Kumar A, Bachhawat AK (2010) OXP1/YKL215c encodes an ATP- Boronin A, Naumova R, Grishchenkov V, Ilijinskaya O (1984) Plasmids dependent 5-oxoprolinase in Saccharomyces cerevisiae: functional specifying ε-caprolactam degradation in Pseudomonas strains. characterization, domain structure and identification of actin-like FEMS Microbiol Lett 22:167–170 ATP-binding motifs in eukaryotic 5-oxoprolinases. FEMS Yeast Res 10:394–401 Calpena E, Deshpande AA, Yap S, Kumar A, Manning NJ, Bachhawat Li LY, Seddon AP, Meister A (1988) Interaction of the protein compo- AK, Espinós C (2015) New insights into the genetics of 5- nents of 5-oxoprolinase. Substrate-dependent enzyme complex for- oxoprolinase deficiency and further evidence that it is a benign bio- mation. J Biol Chem 263:6495–6501 chemical condition. Eur J Pediatr 174:407–411 Midelfort KS, Kumar R, Han S, Karmilowicz MJ, McConnell K, Caspi R, Altman T, Billington R, Dreher K, Foerster H, Fulcher CA, Gehlhaar DK, Mistry A, Chang JS, Anderson M, Villalobos A, Holland TA, Keseler IM, Kothari A, Kubo A, Krummenacker M, Minshull J, Govindarajan S, Wong JW (2013) Redesigning and Latendresse M, Mueller LA, Ong Q, Paley S, Subhraveti P, Weaver characterizing the substrate specificity and activity of Vibrio fluvialis DS, Weerasinghe D, Zhang P, Karp PD (2014) The MetaCyc data- aminotransferase for the synthesis of imagabalin. Protein Eng Des base of metabolic pathways and enzymes and the BioCyc collection Sel 26:25–33 of pathway/genome databases. Nucleic Acids Res 42:D459–D471 Negoro S (2000) Biodegradation of nylon oligomers. Appl Microbiol Chun J, Lee JH, Jung Y, Kim M, Kim S, Kim BK, Lim YW (2007) Biotechnol 54:461–466 EzTaxon: a web-based tool for the identification of prokaryotes Ogawa J, Kim JM, Nirdnoy W, Amano Y, Yamada H, Shimizu S (1995) based on 16S ribosomal RNA gene sequences. Int J Syst Evol Purification and characterization of an ATP-dependent Microbiol 57:2259–2261 amidohydrolase, N-methylhydantoin amidohydrolase, from Donnelly MI, Cooper RA (1981) Succinic semialdehyde dehydrogenases Pseudomonas putida 77. Eur J Biochem 229:284–290 of Escherichia coli. Eur J Biochem 113:555–561 Ortet P, Barakat M, Lalaouna D, Fochesato S, Barbe V, Vacherie B, Eng WW, Gan HM, Gan HY, Hudson AO, Savka MA (2015) Whole- Santaella C, Heulin T, Achouak W (2011) Complete genome se- genome sequence and annotation of octopine-utilizing quence of a beneficial plant root-associated bacterium, Pseudomonas kilonensis (previously P. fluorescens) strain 1855- Pseudomonas brassicacearum. J Bacteriol 193:3146–3111 344. Genome Announc 3:e00463–e00415 Palacio CM, Crismaru CG, Bartsch S, Navickas V, Ditrich K, Breuer M, Esikova T, Ponamoreva O, Baskunov B, Taran S, Boronin A (2012) Abu R, Woodley J, Baldenius K, Wu B (2016) Enzymatic network Transformation of low-molecular linear caprolactam oligomers by for production of ether amines from alcohols. Biotechnol Bioeng caprolactam-degrading bacteria. J Chem Technol Biotechnol 87: 113:1853–1861 1284–1290 Park G, Chu J, Hong S, Kwak Y, Khan AR, Jung BK, Ullah I, Shin J Fabre B, Lambour T, Bouyssié D, Menneteau T, Monsarrat B, Burlet- (2014) Complete genome sequence of the caprolactam-degrading Schiltz O, Bousquet-Dubouch M (2014) Comparison of label-free bacterium Pseudomonas mosselii SJ10 isolated from wastewater quantification methods for the determination of protein complexes of a nylon 6 production plant. J Biotechnol 192:263–264 subunits stoichiometry. EuPA Open Proteom 4:82–86 Payoungkiattikun W, Okazaki S, Ina A, Aran H, Asano Y (2017) Fortmann L, Rosenberg A (1984) Fate of ϵ-caprolactam in the aquatic Characterization of an α-amino-ɛ-caprolactam racemase with broad environment. Chemosphere 13:53–65 substrate specificity from Citreicella sp. SE45. J Ind Microbiol Fukumura T, Talbot G, Misono H, Teramura Y, Kato K, Soda K (1978) Biotechnol 44:677–685 Purification and properties of a novel enzyme, L-α-amino-ϵ- Poelarends GJ, Wilkens M, Larkin MJ, van Elsas JD, Janssen DB (1998) caprolactamase from Cryptococcus laurentii. FEBS Lett 89:298– Degradation of 1, 3-dichloropropene by Pseudomonas cichorii 170. Appl Environ Microbiol 64:2931–2936 Gabor EM, De Vries EJ, Janssen DB (2004) Construction, characteriza- Rajoo S, Ahn JO, Lee HW, Jung JK (2013) Isolation and characterization tion, and use of small-insert gene banks of DNA isolated from soil of a nov el ε - caprolac tam- d egrading microbe, and enrichment cultures for the recovery of novel amidases. Environ Acinetobactercalcoaceticus, from industrial wastewater by Microbiol 6:948–958 chemostat-enrichment. Biotechnol Lett 35:2069–2072 Ishikawa M, Tsuchiya D, Oyama T, Tsunaka Y, Morikawa K (2004) Sambrook J, Russell DW, Russell DW (2001) Molecular cloning: a lab- Structural basis for channelling mechanism of a fatty acid β- oratory manual, 3rd edn. Cold Spring Harbor Laboratory Press, New oxidation multienzyme complex. EMBO J 23:2745–2754 York Iwaki H, Hasegawa Y, Teraoka M, Tokuyama T, Bergeron H, Lau PC Schiroli D, Peracchi A (2015) A subfamily of PLP-dependent enzymes (1999) Identification of a transcriptional activator (ChnR) and a 6- specialized in handling terminal amines. Biochim Biophys Acta oxohexanoate dehydrogenase (ChnE) in the cyclohexanol catabolic 1854:1200–1211 pathway in Acinetobacter sp. strain NCIMB 9871 and localization Seddon AP, Meister A (1986) Trapping of an intermediate in the reaction of the genes that encode them. Appl Environ Microbiol 65:5158– catalyzed by 5-oxoprolinase. J Biol Chem 261:11538–11543 Seddon AP, Li LY, Meister A (1984) Resolution of 5-oxo-L-prolinase into Janßen HJ, Steinbüchel A (2014) Fatty acid synthesis in Escherichia coli a 5-oxo-L-proline-dependent ATPase and a coupling protein. J Biol and its applications towards the production of fatty acid based Chem 259:8091–8094 biofuels. Biotechnol Biofuels 7:1 Sheldon T (1989) Chromosomal damage induced by caprolactam in hu- Kalinová JP, Tříska J, Vrchotová N, Moos M (2014) Verification of pres- man lymphocytes. Mutat Res 224:325–327 ence of caprolactam in sprouted achenes of Fagopyrumesculentum Shin J, Yun H, Jang J, Park I, Kim B (2003) Purification, characterization, Moench and its influence on plant phenolic compound content. and molecular cloning of a novel amine: pyruvate transaminase from Food Chem 157:380–384 Vibrio fluvialis JS17. Appl Microbiol Biotechnol 61:463–471 Appl Microbiol Biotechnol (2018) 102:6699–6711 6711 Steffen-Munsberg F, Vickers C, Kohls H, Land H, Mallin H, Nobili Vizcaíno JA, Csordas A, del-Toro N, Dianes JA, Griss J, Lavidas I, Mayer G, Perez-Riverol Y, Reisinger F, Ternent T, Xu QW, Wang R, A, Skalden L, van den Bergh T, Joosten H, Berglund P (2015) Bioinformatic analysis of a PLP-dependent enzyme superfami- Hermjakob H (2016) 2016 update of the PRIDE database and its ly suitable for biocatalytic applications. Biotechnol Adv 33: related tools. Nucleic Acids Res 44:D447–456 566–604 Vogel E (1989) Caprolactam induces genetic alterations in early germ cell Takehara I, Fujii T, Tanimoto Y, Kato D, Takeo M, Negoro S (2018) stages and in somatic tissue of D. melanogaster. Mutat Res 224: Metabolic pathway of 6-aminohexanoate in the nylon 339–342 oligomer-degrading bacterium Arthrobacter sp. KI72: identifi- Wasinger VC, Zeng M, Yau Y (2013) Current status and advances in cation of the enzymes responsible for the conversion of 6- quantitative proteomic mass spectrometry. Int J Proteomics 2013: aminohexanoate to adipate. Appl Microbiol Biotechnol 102: 801–814 Watanabe T, Abe K, Ishikawa H, Iijima Y (2004) Bovine 5-oxo-L- Turk SC, Kloosterman WP, Ninaber DK, Kolen KP, Knutova J, Suir E, prolinase: simple assay method, purification, cDNA cloning, and Schürmann M, Raemakers-Franken PC, Müller M, de Wildeman detection of mRNA in the coronary artery. Biol Pharm Bull 27: SM, Raamsdonk LM, van der Pol R, Wu L, Temudo MF, van der 288–294 Hoeven RA, Akeroyd M, van der Stoel RE, Noorman HJ, Weidenweber S, Schuhle K, Demmer U, Warkentin E, Ermler U, Heider J Bovenberg RA, Trefzer AC 2016. Metabolic engineering toward (2017) Structure of the acetophenone carboxylase core complex: sustainable production of nylon-6. ACS Synth Biol 5:65–73 prototype of a new class of ATP-dependent carboxylases/hydro- van der Wal L, Demmers JAA (2015) Quantitative mass spectrometry- lases. Sci Rep 7:39674 based proteomics. In: Magdeldin S (ed) Recent advances in proteo- Ye GJ, Breslow E, Meister A (1996) The amino acid sequence of rat mics research. InTech, Rijeka. https://doi.org/10.5772/61756 kidney 5-oxo-L-prolinase determined by cDNA cloning. J Biol Available from: https://www.intechopen.com/books/recent- Chem 271:32293–32300 advances-in-proteomics-research/quantitative-mass-spectrometry- based-proteomics http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Applied Microbiology and Biotechnology Springer Journals

Characterization of the caprolactam degradation pathway in Pseudomonas jessenii using mass spectrometry-based proteomics

Download PDF
13 pages

Loading next page...
1
 
/lp/springer_journal/characterization-of-the-caprolactam-degradation-pathway-in-pseudomonas-gFx40UkY7c

References (56)

Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s)
Subject
Life Sciences; Microbiology; Microbial Genetics and Genomics; Biotechnology
ISSN
0175-7598
eISSN
1432-0614
DOI
10.1007/s00253-018-9073-7
pmid
29850960
Publisher site
See Article on Publisher Site

Abstract

Some bacterial cultures are capable of growth on caprolactam as sole carbon and nitrogen source, but the enzymes of the catabolic pathway have not been described. We isolated a caprolactam-degrading strain of Pseudomonas jessenii from soil and identified proteins and genes putatively involved in caprolactam metabolism using quantitative mass spectrometry-based proteomics. This led to the discovery of a caprolactamase and an aminotransferase that are involved in the initial steps of caprolactam conversion. Additionally, various proteins were identified that likely are involved in later steps of the pathway. The caprolactamase consists of two subunits and demonstrated high sequence identity to the 5-oxoprolinases. Escherichia coli cells expressing this caprolactamase did not convert 5-oxoproline but were able to hydrolyze caprolactam to form 6-aminocaproic acid in an ATP-dependent manner. Characterization of the aminotransferase revealed that the enzyme deaminates 6-aminocaproic acid to produce 6-oxohexanoate with pyruvate as amino acceptor. The amino acid sequence of the aminotransferase showed high similarity to subgroup II ω- aminotransferases of the PLP-fold type I proteins. Finally, analyses of the genome sequence revealed the presence of a caprolactam catabolism gene cluster comprising a set of genes involved in the conversion of caprolactam to adipate. . . . . . Keywords Caprolactam Caprolactamase Omega aminotransferase Mass spectrometry Proteomics Pseudomonas jessenii Introduction plants (Fortmann and Rosenberg 1984;Kalinová etal. 2014, 2016). The toxic nature of these chemicals, including muta- Caprolactam is a large-volume industrial compound mainly genic effects and plant growth inhibition, has been shown in used in the production of Nylon 6, a versatile synthetic poly- different studies (Sheldon 1989; Vogel 1989). mer applied in fabrics, utensils, mechanical parts, etc. Several microorganisms are able to degrade caprolactam Synthesis of Nylon 6 is achieved by ring-opening polymeri- and/or these by-products, including strains of Pseudomonas, zation of caprolactam at high temperatures. During this pro- Alcaligenes,and Acinetobacter (Kulkarni and Kanekar 1998; cess, several side products are generated, including 6- Rajoo et al. 2013; Baxi and Shah 2002). Previously, a capro- aminocaproic acid (6-aminohexanoic acid (6-ACA)) lactam degradation pathway was proposed (Caspi et al. 2014). unreacted monomers, as well as dimers, cyclic dimers, and It starts with a caprolactam-ring cleavage to form 6-ACA, oligomers of 6-ACA. Caprolactam and these by-products are followed by the deamination to 6-oxohexanoate and subse- present as contaminants in waste water of nylon production quent oxidation to yield adipate. Adipate can be further con- verted via β-oxidation reactions of the fatty acid metabolism pathway. Whereas this pathway seems straightforward, infor- Electronic supplementary material The online version of this article mation about the proteins involved in bacterial caprolactam (https://doi.org/10.1007/s00253-018-9073-7) contains supplementary degradation is rare. At first sight, one would expect a hydro- material, which is available to authorized users. lytic enzyme involved in lactam ring opening. The conversion of the related D,L-α-amino-ε-caprolactam by a combination of * Dick B. Janssen aracemase anda L-amino acid lactam hydrolase to yield L- d.b.janssen@rug.nl lysine has been described (Payoungkiattikun et al. 2017; Biochemical Laboratory, Groningen Biomolecular Sciences and Fukumura et al. 1978; Ahmed et al. 1982, 1986). Biotechnology Institute, University of Groningen, Nijenborgh 4, Additionally, a putative hydrolase and aminotransferase have 9747 AG Groningen, The Netherlands been reported for caprolactam metabolism. Recent work 6700 Appl Microbiol Biotechnol (2018) 102:6699–6711 suggests that a dehydrogenase may oxidize 6-oxohexanoate to at 30 °C. Subsequently, cells were transferred two times to adipate in Arthrobacter sp. KI72 and in Acinetobacter sp. fresh liquid medium after which pure cultures were isolated NCIMB 9871. Subsequently, an aminotransferase might cat- on MM agarose plates supplemented with 0.2% glucose and alyze the conversion of 6-ACA to 6-oxohexanoate in the cell 0.5 mM caprolactam. To confirm growth of the resulting pure (Takehara et al. 2018; Iwaki et al. 1999). cultures on caprolactam, cells were reinoculated in selective The environmental relevance of caprolactam, and the im- liquid MM. Identification of isolated organisms was based on portance to understand the biodegradation pathway of this the analyses of the 16S rRNA gene sequence, received from synthetic compound, prompted us to investigate the biochem- the partial genome sequence obtained in this study, using the istry of the early degradation steps. In this paper, we describe a online EzTaxon database (Chun et al. 2007). P. jessenii strain newly isolated caprolactam degrading strain of Pseudomonas GO3 is deposited at DSMZ (Braunschweig, Germany) under jessenii. To investigate the pathway, and identify the enzymes accession number DSM 106008. involved, we used label-free quantitative mass spectrometry- based proteomics (van der van der Wal and Demmers 2015; Genome sequencing and annotation Fabre et al. 2014; Wasinger et al. 2013). From the various caprolactam-induced proteins, we further examined an unex- In order to obtain a partial genome sequence, total DNA from pected ATP-dependent caprolactamase that forms 6-ACA and P. jessenii GO3 cells was isolated essentially as described a class-II omega-aminotransferase that converts 6-ACA to 6- previously (Poelarends et al. 1998; Sambrook et al. 2001). oxohexanoate. The activities of the enzymes were examined The resulting genomic DNA was subjected to paired end se- with heterologously expressed proteins. Furthermore, a cap- quencing by BaseClear BV (Leiden, The Netherlands). The rolactam degradation gene cluster containing all genes for the genome sequencing was done using a HiSeq2500 system conversion of caprolactam to adipate was detected by genome (Illumina Inc., Eindhoven, The Netherlands). Sequencing of sequencing. the GO3 genome yielded 1.7 million reads of ~ 100 bp in length. These were assembled by BaseClear, using the CLC Genomics Workbench version 7.0.4 (Qiagen, Venlo, Materials and methods The Netherlands), resulting in 1274 contigs with a total length of 7 mega base pairs (Mb). This whole genome project has Growth conditions been deposited at DDBJ/ENA/GenBank under the accession PDLL00000000. The version described in this paper is Escherichia coli C41 (DE3) cells (Lucigen, Halle-Zoersel, PDLL01000000. Subsequently, all generated contigs were Belgium) were grownat37°CinanLBmedium used for automated annotation using the RAST server (Aziz (Sambrook et al. 2001). When required, ampicillin (50 μg/ et al. 2008). mL) was added. The caprolactam-degrader P. jessenii GO3 was cultured at 30 °C in nitrogen-free minimal medium Molecular techniques (MM) (Gabor et al. 2004), supplemented with caprolactam (4 mM) as carbon and nitrogen source. For the preparation Standard recombinant DNA techniques were performed es- of agar plates, the medium was supplemented with 2% agar sentially as described previously (Sambrook et al. 2001). or 1.6% H O-rinsed agarose. Restriction enzymes and polymerase were used according to To determine the caprolactam tolerance of P. jessenii GO3, the instructions of the supplier (New England Biolabs, cells were precultured in MM supplemented with 0.1% cap- Ipswich, MA, USA). Primers used in this study are listed in rolactam at 30 °C. Cells from these precultures were diluted Table 1. 50- to100-foldin 200 μL fresh medium supplemented with varying concentrations of caprolactam (0.05to1.2%). Subsequently, growth was monitored by measuring the absor- Plasmid constructions bance at OD using a microplate spectrophotometer (PowerWave, BioTek, Winooski, VT, USA). For the purification of a 6-ACA aminotransferase (PjAT), plasmid pET-PjAT was constructed containing an in-frame Enrichment of caprolactam-degrading bacteria fusion of the PjAT-encoding gene to a hexahistidine tag, be- hind the T7 promoter region. For this purpose, the PjAT gene For the isolation of bacterial strains, 1 g of residential grass- was amplified using P. jessenii GO3 chromosomal DNA and land soil from Groningen (The Netherlands) was used to in- primers ATfw and ATrev. The resulting 1419 base pair prod- oculate 50 mL nitrogen-free MM supplemented with 0.2% uct was then digested with NdeI/EcoRI and ligated into NdeI/ glucose and 0.5 mM caprolactam as sole carbon and nitrogen EcoRI-digested pET20b (Novagen-Merck, Amsterdam, source. Cultures were incubated for 7 days in an orbital shaker The Netherlands). Appl Microbiol Biotechnol (2018) 102:6699–6711 6701 Table 1 Primers used ATfw TTCCTTCTCTAGAATGAACCAGTCAGTATCCTCGC ATrev TTCCTGAATTCTTAATGATGATGATGATGATGGCCGCCCGGACCAACCCACTGAGTGGTGTC OPfw ATCAAGCTTAATGAACACAGTAGACCCGATC OPrev AAGGAAAAAAGCGGCCGCTCAATGACCGGGAGTCAGTTC For the expression of caprolactamase subunits α and β, acetonitrile, were used as the mobile phases. A gradient from plasmid pET-OP was constructed containing both genes, be- 5 to 40% acetonitrile was performed at a flow rate of 300 nL/ hind the T7 promoter region. Since both genes are likely lo- min. Eluted peptides were analyzed using a linear ion trap cated in one operon, they were amplified together using Orbitrap hybrid mass spectrometer (LTQ-Orbitrap XL, P. jessenii GO3 chromosomal DNA and primers OPfw and Thermo Scientific). MS scans were acquired in the range from OPrev. The resulting 3871 base pair fragment was then 300 to 2000 m/z. The five most intense ions per scan were digested using HindIII and NotI and ligated into HindIII/ selected for MS/MS fragmentation (35% normalized collision NotI-digested pET20b . energy) and detected in the linear ion trap. Peak lists were obtained from raw data files using Proteomics by mass spectrometry Proteome Discoverer (version 1.1, Thermo Fisher Scientific). Mascot (version 2.1, Matrix Science, London, Pseudomonas jessenii GO3 cells were cultured in 50 mL MM UK) was used for searching against the annotated P. jessenii supplemented with 0.2% glucose and 5 mM (NH ) SO or GO3 genomic DNA sequence. Peptide tolerance was set to 4 2 4 with 4 mM caprolactam. When the culture reached the late 10 ppm and the fragment ion tolerance to 2.0 Da, using exponential growth phase, cells were harvested by centrifuga- semitrypsin as protease specificity and allowing for up to tion at 3000×g for 15 min. Cell pellets were washed once with two missed cleavages. Oxidation of methionine residues, 50 mM potassium phosphate buffer, pH 7.8, and stored at − deamidation of asparagine and glutamine, and S- 20 °C prior to use. carboamidomethylation of cysteines were specified as vari- For mass spectrometry, cell pellets were resuspended in able modifications. The MS/MS-based peptide and protein 50 mM potassium phosphate buffer, pH 7.8, and lysed using identifications were further validated with the program a Vibra Cell sonicator (Sonics, Newtown, CT, USA) at 0 °C. Scaffold (version 4.6.1, Proteome Software Inc., Portland, To remove unbroken cells and cell debris, the samples were OR, USA). Peptide identifications were accepted when the centrifuged at 17,000×g for 60 min at 4 °C. Soluble protein probability was greater than 95%. Protein identifications were was precipitated by the addition of 20% trichloroacetic acid based on at least two unique peptides identified by MS/MS, (TCA). After incubating the samples for 1 h on ice, samples each with a confidence of identification probability higher were centrifuged at 17,000×g for 30 min at 4 °C. Protein than 99%. pellets were washed with ice-cold acetone to remove residual For each growth condition, at least two replicates of two TCA. Dry protein extracts were then resuspended in 50 μL independent cultures were analyzed. Normalized intensity- 50 mM NaOH. Reduction of the samples was performed with based absolute quantification (iBAQ) values from Scaffold 5 μL of 500 mM dithiothreitol (DTT) in 350 μL100 mM were used as a measure for the abundance of the identified NH HCO for 30 min at 25 °C, followed by derivatization proteins. Average iBAQ values were calculated for the differ- 4 3 of sulfhydryls by 30 min incubation at room temperature with ent samples and subsequently log2 transformed. In case the 10 μL of 550 mM iodoacetamide. Trypsin digestion was per- protein was not detected, the log2-transformed iBAQ value formed overnight at 37 °C by addition of 4 μg trypsin gold was manually set to 8.5, 2.5-fold below the approximate limit (mass spectrometry grade, Promega (Leiden, of detection. The effect of growth conditions on specific pro- The Netherlands)), followed by a second trypsin digestion tein amounts was calculated by dividing the average log2 for 3hat 37°Cusing 1 μg trypsin gold. Samples were pre- iBAQ value for each protein in extracts from caprolactam- pared for injection by addition of 2.5% formic acid. grown cells by the corresponding iBAQ value in protein ex- For LC-MS, peptides were first trapped on a precolumn tracts from control cells. A protein was considered upregulat- (EASY-Column C18, 100 μm × 20 mm, 5 μmparticle size, ed when the log2-fold ratio was more than two and downreg- Thermo Scientific, Ermelo, The Netherlands) and separated ulated when the log2-fold ratio was less than 0.5. on a capillary column (C18 PepMap 300, 75 μm×100 mm, The mass spectrometry proteomics data have been depos- 3-μm particle size, Thermo Scientific) mounted on a Proxeon ited to the ProteomeXchange Consortium via the PRIDE Easy-nLCII system (Thermo Scientific). Solutions of 0.1% (Vizcaíno et al. 2016) partner repository with the dataset iden- tifier PXD008544 and https://doi.org/10.6019/PXD008544. formic acid in water, and 0.1% formic acid in 100% 6702 Appl Microbiol Biotechnol (2018) 102:6699–6711 Expression and purification Standard reaction mixtures contained 100 mM potassium phosphate buffer (pH 8), 5 U/mL alanine dehydrogenase, The aminotransferase PjAT and caprolactamase were both 2mM NAD , 0.05 mM PLP, 2 mM substrate, 0.2 mM pyru- produced in E. coli C41 cells under control of the T7 promot- vate, and varying concentrations of enzyme in a total volume er. PjAT was expressed and purified as previously described of 300 μL in flat-bottom 96-well microtiter plates. Reactions for related aminotransferases (Palacio et al. 2016). were carried out at 30 °C and analyzed using a microtiter plate For the expression of the α and β subunits of reader (Synergy Mx Microplate Reader, BioTek Instruments, caprolactamase, 0.5 mL of an overnight grown LB culture of Bad Friedrichshall, Germany). Reaction mixtures lacking py- pET-OP transformed cells was transferred to 50 mL ruvate (150 μL) were prewarmed before the reaction was ini- autoinduction medium (ForMedium) containing ampicillin tiated by the addition of 150-μL pyruvate solution. Each re- and incubated for 48 h in a rotary shaker at 17 °C. To prepare action was analyzed in triplicate. Initial rates were used to cell-free extracts, cells were washed in buffer A (50 mM am- determine specific activities in units per mg protein monium bicarbonate, pH 8.5, 10 mM MgCl ), resuspended in (μmol/min/mg). Protein content was determined using the buffer A, and lysed using a Sonics Vibra Cell sonicator at Bradford method, with bovine serum albumin as the standard. 0 °C. To remove unbroken cells and cell debris, the samples The amination of 6-oxohexanoate with alanine as donor were centrifuged at 17,000×g for 30 min. was also followed by coupling to alanine dehydrogenase. The reaction mixtures contained 100 mM potassium phos- Enzyme kinetics phate buffer (pH 8), 2 mM substrate, 0.1 mM NADH, 0.05 mM PLP, 8 U/mL alanine dehydrogenase, 5 mM ammo- To analyze aminotransferase activity in cell-free extracts, re- nium bicarbonate, 5 mM L-alanine, and varying concentra- actions were followed using HPLC analyses. Cell-free ex- tions of enzyme in a total volume of 300 μL. Reactions were tracts were prepared in 50 mM potassium phosphate buffer initiated by addition of 150 μLof L-Ala and carried out as (pH 8) containing 0.3 mM pyridoxal 5'-phosphate (PLP). described previously. Conversion was followed by measuring Standard reaction mixtures contained 50 mM potassium phos- NADH depletion at 340 nm. phate buffer (pH 8), 2 mM amine donor (pyruvate or α- Caprolactamase activity was determined by analyzing re- ketoglutarate), 5 mM 6-ACA, 0.3 mM PLP, and cell-free ex- action mixtures using an Acquity TQD mass spectrometer tract, in a total volume of 300 μL. Reactions were carried out (Waters, Etten-Leur, The Netherlands). Standard reaction mix- at 28 °C. With different time intervals, 50 μL samples were tures contained 2 mM substrate (caprolactam or 5- taken and quenched by the addition of 50 μL2MHCl. After oxoproline), 5 mM ATP, 10 mM MgCl , and cell-free extract, incubating the samples for 10 min on ice, samples were cen- in 50 mM ammonium bicarbonate, pH 8.5. Samples were trifuged for 10 min at 17,000×g and neutralized using 100 μL taken and quenched by the addition of 2% formic acid. 1 M NaOH. Amino acids (6-ACA, alanine, glutamate) in the Separation of the reaction content was performed by UPLC reaction mixtures were quantified by precolumn o- using a Waters Acquity UPLC HSS T3 Column (1.8 μm, phthalaldehyde (OPA) derivatization and subsequent HPLC 2.1 × 150 mm) and a linear gradient (eluent A: water, 0.1% analyses. To this purpose, 10-μL sample was mixed with formic acid; eluent B: 100 acetonitrile, 0.1% formic acid). The 40 μL 0.4 M boric acid (pH 9.7) and 10 μL OPA solution samples were analyzed in positive ion mode. To determine (Fisher Scientific) and incubated for 20 min at 30 °C. Then, substrate reduction and product formation, multiple reaction 3 μL of the reaction sample were injected by an autosampler monitoring (MRM) was performed, measuring the following and analyzed by HPLC using a C18 OPA Adsorbosphere fragments: caprolactam m/z =96; 6-ACA m/z =114; 5- column connected to a Jasco FP-920 detector (excitation oxoproline m/z = 84; glutamate m/z =102. 350 nm; emission 450 nm). Compounds were eluted using a linear gradient (eluent A, 20 mM sodium acetate, pH 7.2, 0.5% (vol/vol) tetrahydrofuran, 0.018% (vol/vol) TEA and eluent B, 90% acetonitrile) at a flow rate of 0.5 mL/min. Results The activity of purified PjAT with pyruvate as the amine acceptor was estimated by coupling the reaction to alanine Isolation of bacterial strains using caprolactam dehydrogenase and measuring the increase in NADH absor- as a sole nitrogen source bance that occurs as a result of oxidative deamination of the produced alanine. Since pyruvate is a competitive inhibitor of In order to isolate a bacterial strain possessing a caprolactam alanine dehydrogenase, low pyruvate concentrations were metabolism pathway, soil microorganisms were enriched for used to minimize the lag time of the reaction. The concentra- the ability to grow on caprolactam as sole nitrogen source. A tion of alanine dehydrogenase in the assays was in excess pure culture was obtained by repeated transfer to fresh medi- (5 units/mL), to give aminotransferase-dependent velocities. um plates. The bacterial strain that was growing best on Appl Microbiol Biotechnol (2018) 102:6699–6711 6703 0.4 selective medium was used to study caprolactam metabolism in detail and was designated strain GO3. 0.3 Growth analyses revealed that strain GO3 was able to use caprolactam both as a sole nitrogen and carbon source. 0.2 Interestingly, 6-aminohexanoic acid (6-ACA), a described in- termediate in caprolactam degradation (Caspi et al. 2014), was 0.1 not a possible growth substrate for this strain. In mineral me- dium (MM) supplemented with 0.05% caprolactam as sole 00.2 0.4 0.6 carbon and nitrogen source, the calculated μ was max −1 Caprolactam (%) 0.37 h . Growth analyses using different concentrations of Fig. 1 Specific growth rates of P. jessenii GO3 in media supplemented caprolactam in MM revealed that the μ is reduced by max with different concentrations of caprolactam higher concentration of caprolactam, with a calculated critical caprolactam concentration of 0.46% (Fig. 1). A total of 137 different proteins were identified in the com- bined independent replicate experiments, corresponding to Draft genome sequence 2.2% of the predicted P. jessenii proteome. Among these, 109 proteins were identified in both experiments and were After paired end sequencing of the genomic DNA from strain subjected to further bioinformatic analysis (Fig. 2a). GO3, genome assembly resulted in 1274 contigs (N50: Seventeen of these proteins showed at least a 2-fold increase 10,682 bp), covering 6,993,317 bp. The GC content is 60% in log2 protein abundance in the caprolactam-grown cells as with 6231 predicted coding sequences. Among these predict- compared to the glucose cultures (Fig. 2b, Table 2;Supporting ed genes, 4754 were assigned a predicted function (76%). information Table S1 and Table S2). Interestingly, some pro- Furthermore, 3 rRNA and 61 tRNA genes were identified in teins are highly induced on caprolactam, while others are just the draft genome. above the level of detection (LOD ~ 21, Fig. 2c). Conversely, Analyses of the 16S rRNA gene revealed that the organism in glucose-grown cells, these proteins were below the level of is a Pseudomonas species closely related to P. jessenii CIP detection (iBAQ < 21, data not shown). 105274 (99.5% identity). The draft genome was compared to related Pseudomonas species of which the complete se- Hypothetical caprolactam degradation pathway quence is published, including the caprolactam-degrading or- ganism Pseudomonas mosselii SJ10 (Park et al. 2014) Based on the identified caprolactam-induced proteins (Table 2). Previous studies showed that in other (Table 3), in combination with previous data (Esikova caprolactam-utilizing Pseudomonas strains, the genes in- et al. 2012), a complete putative P. jessenii caprolactam volved in caprolactam metabolism are plasmid localized degradation pathway was built, including all enzymes that (Boronin et al. 1984). Using gel electrophoresis of DNA ex- play a role in the pathway (Fig. 3). To enable growth on tracts, we did not find a plasmid in P. jessenii GO3 (data not caprolactam as sole carbon and nitrogen source, an active shown), suggesting a chromosomal location of the catabolic uptake of caprolactam might be needed, which may be genes. dependent on ABC transporter proteins. Four ABC trans- porter substrate binding proteins are significantly induced Identification of caprolactam degradation enzymes during growth on caprolactam, including ORF1056, ORF3044, ORF1114, and ORF2532 (Table 3). Database A hypothetical caprolactam degradation pathway (Esikova searches demonstrated homology to various ABC trans- et al. 2012) involves two unidentified enzymes: the ring- porters, including the spermidine/putrescine binding pro- cleavage enzyme, presumably a hydrolase, and the enzyme teins (ORF1056 and ORF3044), the branched chain ami- involved in the deamination reaction, which could be an ami- no acid binding proteins (ORF1114), and the amino acid notransferase, an oxidase, or an amine dehydrogenase, all binding proteins (ORF2532). Interestingly, ORF1114 and three producing an ω-ketoacid. To identify the enzymes in- ORF2532 were predominantly induced when the cells volved in the pathway, the P. jessenii GO3 proteome was were grown on caprolactam as sole carbon and nitrogen examined for caprolactam-induced proteins. To this purpose, source (Fig. 2c). Possibly, one or several of these pro- P. jessenii cells were grown in minimal medium supplemented tein(s) are involved in the uptake of caprolactam. with caprolactam (4 mM), or glucose plus ammonium sulfate. Inside the cell caprolactam is most likely converted to 6- Cell-free extracts were prepared from both cultures and sub- ACA. This conversion might be dependent on a jected to quantitative proteome analysis using a label-free caprolactamase, catalyzing the opening of the lactam ring. approach. Inspection of the caprolactam-induced proteins for a -1 µ (h ) max 6704 Appl Microbiol Biotechnol (2018) 102:6699–6711 Table 2 General genomic features of various Pseudomonas species General features Pj GO3 Pb NFM421 Pk 1855–344 Pm SJ10 Size (Mb) 7.0 6.8 6.8 6.2 GC (%) 60.0 60.8 60.7 63.4 CDS 6231 6097 5856 5413 Protein with predicted function (%) 76.3 The genome data are adopted from the original papers. These numbers may differ from numbers obtained with updated annotations CDS coding sequences, Pb Pseudomonas brassicacearum NFM421 (Ortet et al. 2011), Pk Pseudomonas kilonensis 1855-344 (Eng et al. 2015), Pm P. mosselii SJ10 (Park et al. 2014) lactamase-related protein resulted in the identification of two 5-oxoproline, hydantoin, and caprolactam share a lactam moi- distinct polypeptides, ORF4270 and ORF4271. Database ety, it is likely that proteins encoded by these two ORFs are searches revealed identity to subunits A and B of the involved in the caprolactamase reaction. hydantoinase from Pseudomonas sp. (Table 3,respectively, Further conversion of 6-ACA can proceed through deam- 32 and 30% identity) and to the putative 5-oxoprolinase sub- ination catalyzed by an aminotransferase, producing 6- units A and B from Pseudomonas putida (OplA, 78% identity oxohexanoate. Inspection of the caprolactam-induced proteins and OplB, 86% identity). The ORF4270 and ORF4271 resulted in the identification of a protein with homology to the encoded sequences also displayed weak similarity to eukary- Vibrio fluvialis ω-amino acid aminotransferase (ORF4266, otic 5-oxoprolinases which are known to catalyze the ATP- 43% identity). The well-characterized V. fluvialis enzyme cat- dependent hydrolytic decyclization of 5-oxoproline, produc- alyzes the pyruvate-dependent transamination of ω-amino ing L-glutamate (Saccharomyces cerevisiae OXP1/YKL215c, acids and other amines to aldehydes or ketones (Shin et al. 22 and 23% identity, respectively) (Seddon et al. 1984). Since 2003). Fig. 2 Proteomic analysis of A B P. jessenii GO3 cells, obtained from two independent replicate 100 Total cultures. A Number of P. jessenii proteins identified by mass Downregulated <0.5 spectrometry in replicate culture 1 (red) and replicate culture 2 Upregulated >2 (blue). The overlapping region represents the number of proteins identified in both experiments (purple). B Bar graph representing the total number of identified P. jessenii proteins (green), including several caprolactam-induced proteins (red) and several caprolactam- repressed proteins (blue). A protein was considered upregulated when the protein level in caprolactam-induced cells divided by the protein level in non-induced cells was larger than two; a protein was considered downregulated when this ratio was smaller than 0.5. C Bar graph representing the average iBAQ log2 protein amounts of the identified upregulated proteins in caprolactam-grown cells. For all these hits, the protein amounts in non-induced cells were below the level of detection (not depicted in this plot) (Color figure online) No. of idenfied proteins Appl Microbiol Biotechnol (2018) 102:6699–6711 6705 Table 3 Caprolactam-induced proteins Name Contig Size (bp) Enzyme Seq. identity to known EC number Accession proteins (%, organism) number 4265 5 1451 Succinate-semialdehyde dehydrogenase 81, E. coli 1.2.1.24 3JZ4 4266 5 1367 Omega aminotransferase 43, V. fluvialis 2.6.1.18 3NUI 4270 5 2114 Hydantoin utilization protein A 32, Pseudomonas sp. 3.5.2.- Q01262 4271 5 1745 Hydantoin utilization protein B 30, Pseudomonas sp. 3.5.2.- Q01262 4277 5 1238 Acetyl-CoA:oxalate CoA-transferase 39, E. coli 2.8.3.19 4HL6 4278 5 1202 Acetyl-CoA acetyltransferase 41, Ralstonia eutropha 2.3.1.9 4O99 4279 5 1541 3-Hydroxyacyl-CoA dehydrogenase 39, R. eutropha 1.1.1.35 4PZC 4282 5 1154 Acyl-CoA dehydrogenase 42, Thermus thermophilus 1.3.99.2 2DVL 2532 28 1031 ABC transporter, amino acid binding protein 59, Brucella ovis 4Z9N 3504 39 1922 Serine protein kinase 77, E. coli 2.7.11.1 P0ACY5 4150 48 2147 Fatty acid oxidation complex α subunit 94, P. fragi 4.2.1.17 5.3.3.8 1WDK 1.1.1.35 5.1.2.3 1056 157 1112 ABC transporter, putrescine binding protein 75, P. aeruginosa 3TTM 1114 161 1133 ABC transporter, branched chain amino 51, Agrobacterium fabrum 3IP5 acid binding protein 3044 330 2549 ABC transporter, Spermidine/putrescine 31, Streptococcus pneumoniae 4EQB binding protein 3301 363 344 Putative enzyme of the cupin superfamily 98, Pseudomonas sp. WP_ 4187 485 1325 Isocitrate lyase 82, E. coli 4.1.3.1 1IGW 5740 814 905 Histone H1-like protein HC2 43, Chlamydia pneumoniae Q9Z8F9 The product 6-oxohexanoate would then be converted to hydratase function might be performed by a protein with ho- adipate. In Acinetobacter, this reaction is catalyzed by a 6- mology to the Pseudomonas fragi fatty acid oxidation com- oxohexanoate dehydrogenase (Iwaki et al. 1999). Among plex α subunit (ORF4150). In Pseudomonas, this complex the caprolactam-induced proteins, one protein (ORF 4265) catalyzes multiple reactions of the beta fatty acid oxidation, was identified with homology to the E. coli succinate- including the enoyl-CoA hydratase and the 3-hydroxyacyl- semialdehyde dehydrogenase. In E. coli, this enzyme converts CoA dehydrogenase (Ishikawa et al. 2004). succinate semialdehyde to succinate, which is part of the bio- degradation of 4-aminobutyric acid (Donnelly and Cooper Gene organization 1981). Since succinate-semialdehyde and 6-oxohexanoate are structurally similar, it is plausible that this enzyme is in- The sequence information for most caprolactam-induced pro- volved in the conversion of 6-oxohexanoate. Additionally, teins was found on contig 5, a large segment of 40,376 bp this protein has 38% sequence identity to the Acinetobacter containing 36 putative open reading frames (ORF4261 to 6-oxohexanoate dehydrogenase. ORF4296). The genetic context of this contig was analyzed Finally, adipate most likely enters the β-oxidation pathway using the RAST server. This revealed that contig 5 comprises for degradation of fatty acids, which consists of multiple re- two gene clusters involved in the caprolactam degradation. actions (Janßen and Steinbüchel 2014). First adipate needs to The first gene cluster contains the genes putatively in- be activated by CoA, resulting in adipyl-CoA. This reaction volved in the conversion of caprolactam to adipate, including might be achieved by the caprolactam-induced protein with both subunits of the proposed caprolactam-induced homology to the E. coli acetyl-CoA:oxalate CoA-transferase caprolactamase (ORF4270, ORF4271), the omega amino- (ORF4277). In E. coli, this enzyme catalyzes the reversible transferase (ORF4266), and a 6-oxohexanoate dehydrogenase conversion of oxalate and acetyl-CoA to oxalyl-CoA and ac- (ORF4265) (Fig. 4a). In between these genes, three other etate. Then, adipyl-CoA can be converted in a multistep pro- ORFs are located (Fig. 4a, Pj, in blue). BLAST searches re- cess by means of an adipyl-CoA dehydrogenase (ORF4282), vealed that these open reading frames encode proteins with an enoyl-CoA hydratase, an 3-hydroxyadipyl-CoA dehydro- high homology to the L-2-hydroxyglutarate oxidase LhgO of genase (ORF4279), and a 3-ketoadipyl-CoA thiolase E. coli (ORF4267, 72% identity), the starvation induced pro- (ORF4278). Homologs of all of these proteins were clearly tein CsiD of P. putida (ORF4268, 75%), and the transcription- induced in caprolactam-grown cells, where the enoyl-CoA al regulator CsiR from E. coli (ORF4269, 59%). In E. coli, 6706 Appl Microbiol Biotechnol (2018) 102:6699–6711 Fig. 3 Hypothetical pathway for the biodegradation of caprolactam in P. jessenii cells. A The conversion of caprolactam to adipate. B The first steps of the β- fatty acid degradation these genes cluster together with gabD, gabT,and gabP in- supplemented with 6-ACA as sole nitrogen source; (2) the volved in the conversion of γ-aminobutyrate to succinate. The absence of a homolog of the caprolactamase genes gabD gene encodes a succinate-semialdehyde dehydrogenase ORF4270 and 4271 in E. coli, which appeared to be involved (DH), gabT a γ-aminobutyrate aminotransferase (AT) and in the first step in caprolactam degradation in P. jessenii gabP a γ-aminobutyrate permease. GabT and GabD represent (Fig. 4a, genes labeled with *). similar catalytic activities as the P. jessenii omega aminotrans- The second gene cluster includes genes involved in the ferase (ORF4266) and the 6-oxohexanoate dehydrogenase fatty acid β-oxidation, including the caprolactam-induced ace- (ORF4265), respectively. Clear differences between both gene tyl-CoA:oxalate CoA-transferase (ORF4277), acetyl-CoA clusters include (1) the absence of a homolog of the γ- acetyltransferase (ORF4278), 3-hydroxyacyl-CoA dehydro- aminobutyrate permease gabP in P. jessenii, which might ex- genase (ORF4279), and butyryl-CoA dehydrogenase plain why P. jessenii is not able to grow on minimal medium (ORF4282). In between these genes, two more open reading Appl Microbiol Biotechnol (2018) 102:6699–6711 6707 frames are located which according to BLAST searches en- revealed that 5-oxoproline is not a substrate for the identified code for proteins with high homology to an enoyl-CoA P. jessenii caprolactamase. hydratase from Mus musculus (ORF4280, 47% identity) and an IclR family transcriptional regulator from Pseudomonas Characterization of the omega aminotransferase testosteroni (ORF4281, 36%). A similar gene cluster contain- ing homologs of all six genes is present in the genome of other Most omega aminotransferases are PLP-fold type I enzymes bacteria (e.g., Pseudomonas aeruginosa PAO1), suggesting a that catalyze the transfer of an amino group from a β-, γ-or wider occurrence of adipate metabolism by the same pathway. other ω-amino acid or an amine to pyruvate or α- ketoglutarate (Schiroli and Peracchi 2015). In order to confirm ATP-dependent caprolactamase thepresenceof6-ACA ω-aminotransferase activity in P. jessenii GO3, a protein extract of caprolactam-induced cells To confirm the presence of an ATP-dependent caprolactamase was prepared and incubated with 6-ACA and pyruvate or α- activity in P. jessenii, a cell-free extract of caprolactam- ketoglutarate, and levels of 6-ACA and produced alanine or induced cells was prepared. Subsequently, the extract was glutamate were determined using OPA derivatization and incubated in the presence of caprolactam, ATP, and MgCl , HPLC. This revealed that caprolactam was indeed enzymati- and the production of 6-ACA was examined by UPLC-MS. cally converted with pyruvate as the amino acceptor. In the This revealed that caprolactam is indeed enzymatically con- absence of pyruvate, no conversion of 6-ACA was detected. verted to 6-ACA. When a similar assay was performed using When a similar assay was performed using cell-free extract of the cell-free extract of glucose-grown P. jessenii cells, no con- glucose-grown P. jessenii cells, no conversion of 6-ACA or version of caprolactam to 6-ACA was detected. This con- pyruvate was found. This confirmed the presence of a firmed that the caprolactamase activity present in P. jessenii caprolactam-inducible ω-transaminase activity in P. jessenii cells is induced during growth on caprolactam. Additionally, GO3 cells. to study the ATP dependence of the putative caprolactamase To establish if this ORF4266-encoded putative ω-amino in P. jessenii cells, cell-free extracts of caprolactam-grown acid aminotransferase (PjAT) is responsible for the conversion cells containing MgCl were prepared and tested for the for- of 6-ACA, activity assays were performed using E. coli- mation of 6-ACA in the absence of ATP. No 6-ACA formation expressed PjAT. Since E. coli contains multiple aminotrans- was detected, demonstrating that ATP indeed is required for ferases, PjAT was equipped with a C-terminal His6-tag, and the enzymatic hydrolyses of caprolactam to 6-ACA. the enzyme was purified using a Ni-NTA resin. This yielded To confirm the role of the putative caprolactamase ca. 30–35 mg purified enzyme/L of culture. The activity of the (ORF4270, ORF4271) in the conversion of caprolactam, the enzyme was examined in both directions, so with 6- genes were expressed in E. coli C41. To express both subunits oxohexanoate or with 6-ACA as substrate, together with L- (CapA, CapB) simultaneously, the entire operon including the alanine or pyruvate as amino acceptor. Activities were obtain- 36-bp intergenic region (Fig. 4) was cloned under control of a ed by coupling the reaction to that of alanine dehydrogenase single T7 promoter. Expression in E. coli resulted in the high- and following production or consumption of NADH spectro- level production of two proteins of the expected size, of which photometry at 340 nm. This revealed a specific activity of approximately 60% was present in the soluble fraction. 0.2 U/mg using 6-ACA and pyruvate as the substrates and UPLC-MS analysis was performed to detect the formation an activity of 4.5 U/mg using 5-oxohexanoate and alanine as of 6-ACA in a mixture containing caprolactam, ATP, MgCl , the substrates. These activities would suffice to enable strain and E. coli cell-free extract. Time course analyses demonstrat- GO3 to use caprolactam as a nitrogen source for growth. ed that 6-ACA levels increased and caprolactam levels de- creased in time. A specific activity of 0.14 U/mg was calcu- lated for the E. coli cell-free extract (Fig. 5). In reaction mix- tures containing caprolactam, ATP, MgCl ,and E. coli extract Discussion from cells not producing the caprolactamase, no detectable 6- ACA was observed even after 3 h of incubation. Thus, the In this work, we explored the caprolactam degradation path- caprolactamase activity detected in the induced E. coli (pET- way in the bacterium P. jessenii strain GO3. Previous studies OP) extract originates from the expressed α and β subunits of suggested an overall catabolic pathway for caprolactam me- the caprolactamase. Since the sequence of the α and β sub- tabolism, but the enzymes catalyzing the first two steps, i.e., units showed homology to enzymes annotated as 5- conversion of caprolactam to 6-oxohexanoate, remained un- oxoprolinase, reaction mixtures containing 5-oxoproline, known. Using proteomic studies, we identified an ATP- ATP, MgCl ,and E. coli cell-free extract were tested for the dependent lactamase involved in the conversion of caprolac- production of glutamate. UPLC-MS analyses demonstrated tam to 6-ACA and an ω-aminotransferase responsible for the no detectable glutamate even after 3 h of incubation. This subsequent conversion of 6-ACA to 6-oxohexanoate. 6708 Appl Microbiol Biotechnol (2018) 102:6699–6711 4271 4270 4269 4268 4267 4266 4265 capA* csiR csiD ygaF AT DH capB* Pj csiD ygaF gabD gabT gabP* csiR Ec 4282 4280 4278 adipyl-CoA enoyl-CoA acetyl-CoA dehydrogenase hydratase acetyltransferase Pj 4281 4277 transcriptional acetyl-CoA:adipate 3-hydroxyacyl-CoA regulator CoA transferase dehydrogenase Fig. 4 Schematic representation of the gene organization of contig 5, between the P. jessenii and E. coli gene clusters. b Gene cluster containing most of the caprolactam-induced genes, analyzed using the comprising the genes putatively involved in the fatty acid β- RAST server. a Gene cluster comprising the genes putatively involved degradation of adipate. Genes represented in blue: P. jessenii genes in the conversion of caprolactam to adipate in P. jessenii (Pj) and a similar encoding proteins that were not significantly upregulated (Color figure gene cluster with the genes involved in the conversion of γ-aminobutyrate online) in E. coli (Ec). Genes marked with an asterisk indicate clear differences Additionally, we identified various other enzymes and genes (Seddon and Meister 1986;Liet al. 1988). Subunit A, which putatively involved in the caprolactam catabolic pathway. is homologous to the N-terminal part of the 137 kDa subunit The ATP-dependent lactamase catalyzing caprolactam ring of the eukaryotic enzymes, catalyzes the phosphorylation of opening was identified by proteomic studies, sequence com- enzyme-bound 5-oxoproline, whereas subunit B, which is ho- parison to known proteins with lactamase activity, and func- mologous to the C-terminal part of eukaryotic oxoprolinase, is tional expression in E. coli. Sequence comparison indicated required for hydrolysis of the phosphorylated hydroxypyrrole. similarity to proteins annotated as 5-oxoprolinase or Other enzymes with significant sequence similarity to the hydantoinase. Oxoprolinases catalyze the ATP-dependent hy- caprolactamase are the ATP-dependent hydantoinases. Only drolysis of 5-oxoproline to glutamate (EC 3.5.2.9). The se- the N-terminal sequences of the two subunits of the enzyme quence of only four oxoprolinases with confirmed activity isolated from P. putida 77 have been reported (Ogawa et al. has been reported, including the enzymes from rat and cow 1995), but BLAST searches allow retrieval of complete se- (Ye et al. 1996; Watanabe et al. 2004), S. cerevisiae (Kumar quences from Marinobacterium profundum and Bachhawat 2010), and the human enzyme. The latter can (WP_067296627.1 and WP_067296623.1). Sequence align- harbor mutations of medical relevance (Calpena et al. 2015). ments show that in addition to the A and B subunits of the The eukaryotic 5-oxoprolinases are homodimers with sub- bacterial oxoprolinases mentioned previously, also the A and units of approximately 137 kDa, whereas the P. putida 5- B subunits of these putative hydantoinases and the α and β oxoprolinase, of which the sequence is not reported, consists subunits of caprolactamase described here correspond to the of two different subunits of approximately 75 and 63 kDa N-terminal part and C-terminal part, respectively, of the eu- karyotic oxoprolinases. All these ATP-dependent hydrolases belong to InterPro families IPR002821 and IPR003692. We 3.0 demonstrated the ATP dependence of the caprolactamase ac- 2.5 tivity using enzyme expressed in E. coli. Within the families, 2.0 there are clear differences. For example, the substrate range of the P. jessenii caprolactamase lacks 5-oxoprolinase activity. It 1.5 seems possible that many bacterial genes annotated in 1.0 GenBank as 5-oxoprolinase actually are (capro)lactamase 0.5 genes or cyclic dipeptide hydrolase genes, since there is a higher sequence similarity to the CapA and CapB ORFs than 0.0 0 20406080 100 120 to confirmed oxoprolinases. Time (min) Based on sequence and structural analysis, these Fig. 5 UPLC-MS analyses monitoring the formation of 6-ACA (black lactamases can be further grouped with the ATP-dependent line, triangles) and the degradation of caprolactam (grey line, closed carboxylases/lactamase superfamily, which includes carbox- circles). The reaction mixtures contained 2 mM caprolactam, 5 mM ylases acting on acetone and acetophenone (Weidenweber ATP, 10 mM MgCl ,and 75 μg/mL cell-free extract, in 50 mM ammonium bicarbonate, pH 8.5 et al. 2017). The structure of the acetophenone carboxylase Concentration (mM) Appl Microbiol Biotechnol (2018) 102:6699–6711 6709 from Aromatoleum aromaticum EbN1 was recently solved and CapB 99.3% sequence identity). This suggests the same (pdb 519w), revealing its quaternary structure as (αα′βγ) . caprolactam degradation pathways in both Pseudomonas spe- Sequence motifs indicative of ATP binding are conserved be- cies. Additionally, BLASTsearches against the non-redundant tween the α subunit of the carboxylase and the α subunit of protein database revealed highly similar open reading frames caprolactamase (ORF 4270), and the β subunit of the carbox- in annotated sequences of various other Pseudomonas strains, ylase is homologous to the β subunit of caprolactamase (ORF indicating that the ability to hydrolyze lactams or cyclic pep- 4271). Thus, the elucidation of the caprolactam catabolism tides may not be unusual in Pseudomonas.Other nylonby- described here adds a new member to this diverse group of products that can be degraded by microorganisms include 6- ATP-dependent hydrolytic enzymes. By similarity to ACA dimers, cyclic dimers, and oligomers. Metabolism is oxoprolinase and hydantoinase, caprolactam hydrolysis by dependent on hydrolases, such as NylA, NylB, and NylC, the caprolactamase is expected to proceed by phosphorylation which have different specificities (Negoro 2000). NylA cata- of the enol (lactim) tautomer mediated by the α subunit, lyzes the hydrolysis of 6-ACA cyclic dimer, resulting in the followed by its hydrolysis with a role for the β subunit. formation of the 6-ACA dimer, which can be converted by Structures of lactamases that would provide detailed insight NylB, generating 6-ACA. NylC catalyzes the hydrolysis of 6- are lacking, however. ACA oligomers. BLAST searches using Flavobacterium sp. The 6-aminohexanoate aminotransferase (PjAT) catalyzing NylA, B, and C revealed that P. jessenii contains homologs of the second step in caprolactam catabolism was also identified NylA and NylB, but not of NylC, suggesting the absence of a by proteomic studies, sequence similarities, and functional complete pathway for metabolism of 6-ACA polymers. overexpression in E. coli. BLAST searches using the protein sequence of the PjAT demonstrated relatedness to the fold- Acknowledgements The authors acknowledge M. Schürman and S. Turk type I PLP enzymes, more in particular to the subgroup II (DSM, The Netherlands) for their support and helpful discussions and T. aminotransferases, which are now often grouped as class III Tiemersma-Wegman for assistance with the MS analysis. aminotransferases (Schiroli and Peracchi 2015;Steffen- Munsberg et al. 2015). These enzymes catalyze conversion Author’s contributions MO performed the proteomic and genomic stud- ies. MO and CP performed bioinformatic analysis and expression exper- of ω-amino acids to aldehydes (EC 2.6.1.18). In vitro charac- iments. MO, CP, and DJ wrote the paper. terization of PjAT confirmed the expected deamination of 6- ACA to 6-oxohexanoate. The enzyme was also active in the Funding information This research was supported financially by the reverse reaction: L-alanine-dependent amination of 6- Dutch Ministry of Economic Affairs and BE-Basic (www.be-basic.org), aminohexanoate. This conversion is of potential interest for a public/private research organization. a biosynthetic 6-ACA production pathway that was recently engineered into E. coli (Turk et al. 2015). Comparison with Compliance with ethical standards the V. fluvialis aminotransferase, which is the closest well- studied homolog, revealed conservation of several residues Ethical statement This article does not contain any studies with human participants or animals performed by any of the authors. important for proper activity (PDB code 4E3Q; (Midelfort et al. 2013)). The lysine required for the formation of the Conflict of interest The authors declare that they have no conflict of internal aldimine (Schiff base) with the PLP cofactor is present interest. at a conserved position. Further work on the biocatalytic and structural properties of the enzyme is ongoing. The PjAT ami- Open Access This article is distributed under the terms of the Creative notransferase has 27.3% pairwise sequence identity with the Commons Attribution 4.0 International License (http:// NylD1 aminotransferase recently described in the nylon olig- creativecommons.org/licenses/by/4.0/), which permits unrestricted use, omer degrader Arthrobacter sp. KI72 (Takehara et al. 2018). distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link The closest homolog of the latter enzyme in the P. jessenii to the Creative Commons license, and indicate if changes were made. genome is ORF2380 (48.6% pairwise identify), but an upreg- ulation of this protein was not detected by proteomic analysis of caprolactam-grown P. jessenii cultures. Previously, the full genome of P. mosselii, another capro- References lactam degrading organism isolated from wastewater of a ny- lon producing industrial complex in Korea, was sequenced by Ahmed SA, Esaki N, Soda K (1982) Purification and properties of α- amino-ϵ-caprolactam racemase from Achromobacter obae.FEBS Park and coworkers (Park et al. 2014). Interestingly, BLAST Lett 150:370–374 searches using the sequences of the genes that were found here Ahmed SA, Esaki N, Tanaka H, Soda K (1986) Mechanism of α-amino- to be involved in caprolactam degradation against the full ε-caprolactam racemase reaction. Biochemistry 25:385–388 P. mosselii genome sequence revealed genes with high simi- Aziz RK, Bartels D, Best AA, DeJongh M, Disz T, Edwards RA, larity to these proteins (aminotransferase 100%, CapA 99.6%, Formsma K, Gerdes S, Glass EM, Kubal M, Meyer F, Olsen GJ, 6710 Appl Microbiol Biotechnol (2018) 102:6699–6711 Olson R, Osterman AL, Overbeek RA, MacNeil LK, Paarmann D, Kalinová JP, Tříska J, Vrchotová N, Novák J (2016) Uptake of caprolac- tam and its influence on growth and oxygen production of Paczian T, Parrello B, Pusch GD, Reich C, Stevens R, Vassieva O, Vonstein V, Wilke A, Zagnitko O (2008) The RAST server: rapid Desmodesmusquadricauda algae. Environ Pollut 213:518–523 annotations using subsystems technology. BMC Genomics 9:75 Kulkarni RS, Kanekar PP (1998) Bioremediation of ε-caprolactam from Baxi N, Shah A (2002) ε-Caprolactam-degradation by Alcaligenes Nylon-6 waste water by use of Pseudomonas aeruginosa MCM faecalis for bioremediation of wastewater of a nylon-6 production B-407. Curr Microbiol 37:191–194 plant. Biotechnol Lett 24:1177–1180 Kumar A, Bachhawat AK (2010) OXP1/YKL215c encodes an ATP- Boronin A, Naumova R, Grishchenkov V, Ilijinskaya O (1984) Plasmids dependent 5-oxoprolinase in Saccharomyces cerevisiae: functional specifying ε-caprolactam degradation in Pseudomonas strains. characterization, domain structure and identification of actin-like FEMS Microbiol Lett 22:167–170 ATP-binding motifs in eukaryotic 5-oxoprolinases. FEMS Yeast Res 10:394–401 Calpena E, Deshpande AA, Yap S, Kumar A, Manning NJ, Bachhawat Li LY, Seddon AP, Meister A (1988) Interaction of the protein compo- AK, Espinós C (2015) New insights into the genetics of 5- nents of 5-oxoprolinase. Substrate-dependent enzyme complex for- oxoprolinase deficiency and further evidence that it is a benign bio- mation. J Biol Chem 263:6495–6501 chemical condition. Eur J Pediatr 174:407–411 Midelfort KS, Kumar R, Han S, Karmilowicz MJ, McConnell K, Caspi R, Altman T, Billington R, Dreher K, Foerster H, Fulcher CA, Gehlhaar DK, Mistry A, Chang JS, Anderson M, Villalobos A, Holland TA, Keseler IM, Kothari A, Kubo A, Krummenacker M, Minshull J, Govindarajan S, Wong JW (2013) Redesigning and Latendresse M, Mueller LA, Ong Q, Paley S, Subhraveti P, Weaver characterizing the substrate specificity and activity of Vibrio fluvialis DS, Weerasinghe D, Zhang P, Karp PD (2014) The MetaCyc data- aminotransferase for the synthesis of imagabalin. Protein Eng Des base of metabolic pathways and enzymes and the BioCyc collection Sel 26:25–33 of pathway/genome databases. Nucleic Acids Res 42:D459–D471 Negoro S (2000) Biodegradation of nylon oligomers. Appl Microbiol Chun J, Lee JH, Jung Y, Kim M, Kim S, Kim BK, Lim YW (2007) Biotechnol 54:461–466 EzTaxon: a web-based tool for the identification of prokaryotes Ogawa J, Kim JM, Nirdnoy W, Amano Y, Yamada H, Shimizu S (1995) based on 16S ribosomal RNA gene sequences. Int J Syst Evol Purification and characterization of an ATP-dependent Microbiol 57:2259–2261 amidohydrolase, N-methylhydantoin amidohydrolase, from Donnelly MI, Cooper RA (1981) Succinic semialdehyde dehydrogenases Pseudomonas putida 77. Eur J Biochem 229:284–290 of Escherichia coli. Eur J Biochem 113:555–561 Ortet P, Barakat M, Lalaouna D, Fochesato S, Barbe V, Vacherie B, Eng WW, Gan HM, Gan HY, Hudson AO, Savka MA (2015) Whole- Santaella C, Heulin T, Achouak W (2011) Complete genome se- genome sequence and annotation of octopine-utilizing quence of a beneficial plant root-associated bacterium, Pseudomonas kilonensis (previously P. fluorescens) strain 1855- Pseudomonas brassicacearum. J Bacteriol 193:3146–3111 344. Genome Announc 3:e00463–e00415 Palacio CM, Crismaru CG, Bartsch S, Navickas V, Ditrich K, Breuer M, Esikova T, Ponamoreva O, Baskunov B, Taran S, Boronin A (2012) Abu R, Woodley J, Baldenius K, Wu B (2016) Enzymatic network Transformation of low-molecular linear caprolactam oligomers by for production of ether amines from alcohols. Biotechnol Bioeng caprolactam-degrading bacteria. J Chem Technol Biotechnol 87: 113:1853–1861 1284–1290 Park G, Chu J, Hong S, Kwak Y, Khan AR, Jung BK, Ullah I, Shin J Fabre B, Lambour T, Bouyssié D, Menneteau T, Monsarrat B, Burlet- (2014) Complete genome sequence of the caprolactam-degrading Schiltz O, Bousquet-Dubouch M (2014) Comparison of label-free bacterium Pseudomonas mosselii SJ10 isolated from wastewater quantification methods for the determination of protein complexes of a nylon 6 production plant. J Biotechnol 192:263–264 subunits stoichiometry. EuPA Open Proteom 4:82–86 Payoungkiattikun W, Okazaki S, Ina A, Aran H, Asano Y (2017) Fortmann L, Rosenberg A (1984) Fate of ϵ-caprolactam in the aquatic Characterization of an α-amino-ɛ-caprolactam racemase with broad environment. Chemosphere 13:53–65 substrate specificity from Citreicella sp. SE45. J Ind Microbiol Fukumura T, Talbot G, Misono H, Teramura Y, Kato K, Soda K (1978) Biotechnol 44:677–685 Purification and properties of a novel enzyme, L-α-amino-ϵ- Poelarends GJ, Wilkens M, Larkin MJ, van Elsas JD, Janssen DB (1998) caprolactamase from Cryptococcus laurentii. FEBS Lett 89:298– Degradation of 1, 3-dichloropropene by Pseudomonas cichorii 170. Appl Environ Microbiol 64:2931–2936 Gabor EM, De Vries EJ, Janssen DB (2004) Construction, characteriza- Rajoo S, Ahn JO, Lee HW, Jung JK (2013) Isolation and characterization tion, and use of small-insert gene banks of DNA isolated from soil of a nov el ε - caprolac tam- d egrading microbe, and enrichment cultures for the recovery of novel amidases. Environ Acinetobactercalcoaceticus, from industrial wastewater by Microbiol 6:948–958 chemostat-enrichment. Biotechnol Lett 35:2069–2072 Ishikawa M, Tsuchiya D, Oyama T, Tsunaka Y, Morikawa K (2004) Sambrook J, Russell DW, Russell DW (2001) Molecular cloning: a lab- Structural basis for channelling mechanism of a fatty acid β- oratory manual, 3rd edn. Cold Spring Harbor Laboratory Press, New oxidation multienzyme complex. EMBO J 23:2745–2754 York Iwaki H, Hasegawa Y, Teraoka M, Tokuyama T, Bergeron H, Lau PC Schiroli D, Peracchi A (2015) A subfamily of PLP-dependent enzymes (1999) Identification of a transcriptional activator (ChnR) and a 6- specialized in handling terminal amines. Biochim Biophys Acta oxohexanoate dehydrogenase (ChnE) in the cyclohexanol catabolic 1854:1200–1211 pathway in Acinetobacter sp. strain NCIMB 9871 and localization Seddon AP, Meister A (1986) Trapping of an intermediate in the reaction of the genes that encode them. Appl Environ Microbiol 65:5158– catalyzed by 5-oxoprolinase. J Biol Chem 261:11538–11543 Seddon AP, Li LY, Meister A (1984) Resolution of 5-oxo-L-prolinase into Janßen HJ, Steinbüchel A (2014) Fatty acid synthesis in Escherichia coli a 5-oxo-L-proline-dependent ATPase and a coupling protein. J Biol and its applications towards the production of fatty acid based Chem 259:8091–8094 biofuels. Biotechnol Biofuels 7:1 Sheldon T (1989) Chromosomal damage induced by caprolactam in hu- Kalinová JP, Tříska J, Vrchotová N, Moos M (2014) Verification of pres- man lymphocytes. Mutat Res 224:325–327 ence of caprolactam in sprouted achenes of Fagopyrumesculentum Shin J, Yun H, Jang J, Park I, Kim B (2003) Purification, characterization, Moench and its influence on plant phenolic compound content. and molecular cloning of a novel amine: pyruvate transaminase from Food Chem 157:380–384 Vibrio fluvialis JS17. Appl Microbiol Biotechnol 61:463–471 Appl Microbiol Biotechnol (2018) 102:6699–6711 6711 Steffen-Munsberg F, Vickers C, Kohls H, Land H, Mallin H, Nobili Vizcaíno JA, Csordas A, del-Toro N, Dianes JA, Griss J, Lavidas I, Mayer G, Perez-Riverol Y, Reisinger F, Ternent T, Xu QW, Wang R, A, Skalden L, van den Bergh T, Joosten H, Berglund P (2015) Bioinformatic analysis of a PLP-dependent enzyme superfami- Hermjakob H (2016) 2016 update of the PRIDE database and its ly suitable for biocatalytic applications. Biotechnol Adv 33: related tools. Nucleic Acids Res 44:D447–456 566–604 Vogel E (1989) Caprolactam induces genetic alterations in early germ cell Takehara I, Fujii T, Tanimoto Y, Kato D, Takeo M, Negoro S (2018) stages and in somatic tissue of D. melanogaster. Mutat Res 224: Metabolic pathway of 6-aminohexanoate in the nylon 339–342 oligomer-degrading bacterium Arthrobacter sp. KI72: identifi- Wasinger VC, Zeng M, Yau Y (2013) Current status and advances in cation of the enzymes responsible for the conversion of 6- quantitative proteomic mass spectrometry. Int J Proteomics 2013: aminohexanoate to adipate. Appl Microbiol Biotechnol 102: 801–814 Watanabe T, Abe K, Ishikawa H, Iijima Y (2004) Bovine 5-oxo-L- Turk SC, Kloosterman WP, Ninaber DK, Kolen KP, Knutova J, Suir E, prolinase: simple assay method, purification, cDNA cloning, and Schürmann M, Raemakers-Franken PC, Müller M, de Wildeman detection of mRNA in the coronary artery. Biol Pharm Bull 27: SM, Raamsdonk LM, van der Pol R, Wu L, Temudo MF, van der 288–294 Hoeven RA, Akeroyd M, van der Stoel RE, Noorman HJ, Weidenweber S, Schuhle K, Demmer U, Warkentin E, Ermler U, Heider J Bovenberg RA, Trefzer AC 2016. Metabolic engineering toward (2017) Structure of the acetophenone carboxylase core complex: sustainable production of nylon-6. ACS Synth Biol 5:65–73 prototype of a new class of ATP-dependent carboxylases/hydro- van der Wal L, Demmers JAA (2015) Quantitative mass spectrometry- lases. Sci Rep 7:39674 based proteomics. In: Magdeldin S (ed) Recent advances in proteo- Ye GJ, Breslow E, Meister A (1996) The amino acid sequence of rat mics research. InTech, Rijeka. https://doi.org/10.5772/61756 kidney 5-oxo-L-prolinase determined by cDNA cloning. J Biol Available from: https://www.intechopen.com/books/recent- Chem 271:32293–32300 advances-in-proteomics-research/quantitative-mass-spectrometry- based-proteomics

Journal

Applied Microbiology and Biotechnology – Springer Journals

Published: May 31, 2018

There are no references for this article.