Applied Microbiology and Biotechnology

, Volume 98, Issue 19, pp 8235–8252 | Cite as

Genetic and metabolic analysis of the carbofuran catabolic pathway in Novosphingobium sp. KN65.2

  • Thi Phi Oanh Nguyen
  • Damian E. Helbling
  • Karolien Bers
  • Tekle Tafese Fida
  • Ruddy Wattiez
  • Hans-Peter E. Kohler
  • Dirk Springael
  • René De Mot
Applied genetics and molecular biotechnology

Abstract

The widespread agricultural application of carbofuran and concomitant contamination of surface and ground waters has raised health concerns due to the reported toxic effects of this insecticide and its degradation products. Most bacteria that degrade carbofuran only perform partial degradation involving carbamate hydrolysis without breakdown of the resulting phenolic metabolite. The capacity to mineralize carbofuran beyond the benzofuran ring has been reported for some bacterial strains, especially sphingomonads, and some common metabolites, including carbofuran phenol, were identified. In the current study, the catabolism of carbofuran by Novosphingobium sp. KN65.2 (LMG 28221), a strain isolated from a carbofuran-exposed Vietnamese soil and utilizing the compound as a sole carbon and nitrogen source, was studied. Several KN65.2 plasposon mutants with diminished or abolished capacity to degrade and mineralize carbofuran were generated and characterized. Metabolic profiling of representative mutants revealed new metabolic intermediates, in addition to the initial hydrolysis product carbofuran phenol. The promiscuous carbofuran-hydrolyzing enzyme Mcd, which is present in several bacteria lacking carbofuran ring mineralization capacity, is not encoded by the Novosphingobium sp. KN65.2 genome. An alternative hydrolase gene required for this step was not identified, but the constitutively expressed genes of the unique cfd operon, including the oxygenase genes cfdC and cfdE, could be linked to further degradation of the phenolic metabolite. A third involved oxygenase gene, cfdI, and the transporter gene cftA, encoding a TonB-dependent outer membrane receptor with potential regulatory function, are located outside the cfd cluster. This study has revealed the first dedicated carbofuran catabolic genes and provides insight in the early steps of benzofuran ring degradation.

Keywords

Methylcarbamate Sphingomonad Proteomics Metabolites cfd genes 

Introduction

Carbofuran (2,3-dihydro-2,2-dimethyl-7-benzofuranyl methylcarbamate) is a broad spectrum carbamate pesticide that has been used worldwide to control insects, nematodes, and acarids in vegetable, fruit, and forest crops (Chapalamadugu and Chaudhry 1992). In soil, carbofuran is moderately persistent with a half-life of up to 110 days. Good water solubility (700 mg/L at 25 °C) and low adsorption (mean Koc of 29.4) (http://www.epa.gov/safewater/pdfs/factsheets/soc/tech/carbofur.pdf) result in contamination of surface and groundwater (Caldas et al. 2010; Chiron et al. 1995; Chowdhury et al. 2012; De Llasera and Bernal-Gonzalez 2001; Tariq et al. 2006; Vryzas et al. 2009). Carbofuran is a potent cholinesterase inhibitor in mammals (Fahmy et al. 1970; Gupta 1994). The compound and its metabolites can penetrate the placental barrier and impose serious effects on the maternal-placental-fetal unit (Gupta 1994). The compound also acts as an endocrine disruptor (Goad et al. 2004), oxidative stress inducer (Milatovic et al. 2005), as well as a nephrotoxin (Kaur et al. 2012). Moreover, several carbofuran metabolites from biodegradation cause effects on the reproductive system in female rats (Gupta 1994). As a consequence, carbofuran was banned in many countries but is still in use in developing countries including Vietnam (Dasgupta et al. 2007).

Various carbofuran-degrading microorganisms, mostly Gram-negative bacteria (Castellanos et al. 2013; Chaudhry and Ali 1988; Chaudhry et al. 2002; Desaint et al. 2000; Feng et al. 1997; Kim et al. 2004; Ogram et al. 2000; Plangklang and Reungsang 2012; Shin et al. 2012; Yan et al. 2007) and some fungi (Salama 1998; Seo et al. 2007), have been isolated from carbofuran-treated sites, including several sphingomonad strains (Castellanos et al. 2013; Feng et al. 1997; Kim et al. 2004; Ogram et al. 2000; Shin et al. 2012; Yan et al. 2007). An enzyme hydrolyzing the carbamate functional group of carbofuran was isolated from Achromobacter sp. WM111 (Karns and Tomasek 1991), and the corresponding gene (mcd) located on plasmid pPDL11 was cloned (Tomasek and Karns 1989). Sequences homologous to the mcd gene have been detected in phylogenetically diverse carbofuran-degrading bacterial isolates (Desaint et al. 2000; Parekh et al. 1995; Topp et al. 1993). This widespread occurrence apparently results from gene transfer among bacterial populations in soils (Desaint et al. 2003). The Mcd enzyme is a metallohydrolase with carboxylesterase and phosphotriesterase activity, suggesting that its carbamate hydrolase activity may represent promiscuous activity on a nonnatural substrate (Naqvi et al. 2009). In line with this, Mcd-producing strains such as WM111 (Karns et al. 1986) and the methylotroph ER2 (Topp et al. 1993) only utilize the methylamine originating from the carbofuran methylcarbamyl side chain, while carbofuran phenol accumulates as a persistent metabolite. In addition to organisms that only degrade carbofuran to carbofuran phenol, several isolates have been reported that use the insecticide as sole source of carbon, nitrogen, and energy with concomitant mineralization of the compound beyond the benzofuran ring. Often, such carbofuran-utilizing isolates are sphingomonads that apparently play a crucial role in carbofuran biodegradation in the environment (Feng et al. 1997). For some sphingomonad strains, metabolites in carbofuran degradation were reported. Carbofuran phenol was identified as a metabolite for Sphingomonas sp. SB5 (Kim et al. 2004) and Novosphingobium FND-3 (Yan et al. 2007) indicating hydrolytic cleavage of the carbamate group as a common step. In both strains, further cleavage of the furanyl ring of carbofuran phenol was apparent (Kim et al. 2004; Yan et al. 2007), but the catabolic pathway has not yet been further elucidated. In this paper, we report on the isolation from a carbofuran-exposed agricultural soil in the Mekong Delta (Vietnam) of a carbofuran-mineralizing strain, identified as Novosphingobium sp. KN65.2. Characterization of KN65.2 mutants impaired in carbofuran degradation and mineralization revealed the unique cfd operon with dedicated carbofuran catabolic genes. By metabolic profiling of wild-type KN65.2 and its mutants, several novel carbofuran degradation products were identified, enabling the proposition of a biochemical pathway for mineralization.

Materials and methods

Bacterial strains, plasmids, media, and growth conditions

Strain KN65.2 was isolated from soil sampled within a vegetable field with a long history of carbofuran treatment in the Soc Trang province, Vietnam, by enrichment in minimal medium (MM) pH 7.0 (1.42 g Na2SO4, 1.36 g KH2PO4, 0.3 g (NH4)2SO4, 98.5 mg MgSO4 · 7H2O, 5.75 mg CaCl2 · 2H2O, 3.2 mg Na2-EDTA, 2.75 mg FeSO4 · 7H2O, 1.7 mg MnSO4 · H2O, 1.16 mg H3BO3, 1.15 mg ZnSO4 · 7H2O, 0.24 mg CuSO4, 0.24 mg CoCl2 · 6H2O, 0.1 mg MoO3, 1 L water) supplemented with 200 or 500 mg/L carbofuran (Sigma-Aldrich) as sole carbon source. The strain was routinely grown at 25 °C in Tryptone Soya Broth (TSB, Oxoid) or MM supplemented with carbofuran or glucose (2 g/L, Merck) when needed. For MM with carbofuran as sole carbon and nitrogen source, ammonium sulfate was omitted. In case of solid media, 15 mg/L agar (Sigma-Aldrich) was added. Strain KN65.2 was deposited in the BCCM/LMG culture collection (Ghent, Belgium) as LMG 28221. Escherichia coli BW20767 carrying vector pRL27 (Larsen et al. 2002) and E. coli S17-1 λpir (Simon et al. 1983) were grown in Luria Broth (LB) at 37 °C. Media were supplemented with 25 mg/L kanamycin (Km) when needed.

De novo genome sequencing of strain KN65.2

Sequencing of the genome of strain KN65.2 was performed by BaseClear (The Netherlands) on an Illumina GAIIx platform. The Genomics Workbench 3.7 (CLC bio) was used to assemble the 50-bp paired end reads. The draft genome sequence was analyzed by MaGe (https://www.genoscope.cns.fr/agc/microscope/home/index.php). Annotations of contigs of interest were inspected manually.

Taxonomic assignment of strain KN65.2

A maximum likelihood tree using PhyML (Guindon and Gascuel 2003; JTT matrix) was derived from a multiple amino acid sequence alignment for 30 other sphingomonad strains. A representative set of 25 enzymes/proteins with housekeeping functions was used for strain KN65.2 (draft genome sequence), the strains described by Aylward et al. (2013), and four additional sphingomonads (Spingobium japonicum BiD32, Sphingopyxis sp. MC1, Sphingomonas sp. MM-1, Novosphingobium sp. LH128). The concatenated amino acid sequences (~10,200 residues) were aligned with MUSCLE and realigned with MAFFT (Geneious Pro 6.1.4).

Resting cell carbofuran mineralization assay

Fifteen milliliters of an overnight preculture of strain KN65.2, grown in MM supplemented with glucose (2 g/L) until an optical density at 600 nm (OD600) of 1.0, was used to inoculate 7.5 mL MM supplemented with both 4 g/L glucose and 400 mg/L carbofuran, as well as 7.5 mL MM supplemented with 4 g/L glucose. The carbofuran-induced and -noninduced cultures were incubated at 25 °C, aerated by shaking at 125 rpm until an OD600 of 1.0, and subsequently centrifuged at 6,000 × g for 5 min. The cell pellets were washed in 5 mL phosphate-buffered saline (PBS, 1.24 g K2HPO4, 0.32 g KH2PO4, 8.8 g NaCl, 1 L water) and resuspended in 900 μL PBS buffer pH 7.0. A 300-μL aliquot of each culture was inoculated in 4.5 mL PBS buffer containing 10 μg/L [benzene ring-U-14C] carbofuran (specific activity 4.90 MBq/mg; Izotop). Triplicate cultures were incubated at 25 °C while shaking at 125 rpm. 14CO2 evolved from carbofuran mineralization was determined every 10 min during 2.5 h as described (Uyttebroek et al. 2006).

Plasposon mutagenesis

Plasposon delivery vector pRL27 carrying a hyperactive Tn5 transposase gene and a mini-Tn5 element encoding Km resistance (Larsen et al. 2002) was maintained in E. coli BW20767 and used to construct a mutant library of Novosphingobium sp. KN65.2 by conjugation. Strains KN65.2 and BW20767 were individually grown overnight until an OD600 of 1.2 was attained in 250 mL TSB and 25 mL LB supplemented with Km, respectively. The cells were washed twice in 0.01 M MgSO4 by centrifuging at 6,000 × g for 10 min at 25 °C (Beckman Coulter). Pellets of the recipient and donor were suspended in 5 mL and 500 μL 0.01 M MgSO4, respectively. The suspensions were mixed by gentle pipetting and dropped onto sterile Whatman paper filters (diameter 3 cm), after which the filters were placed on TSB agar plates. After incubation at 25 °C overnight, the cell biomass on the filters was recovered by gentle vortexing of the filters in 25 mL 0.01 M MgSO4. Exconjugants were subsequently selected by plating the cell suspension on MM agar plates containing glucose (2 g/L) and Km (25 mg/L).

To assess whether the plasposon inserted randomly in the KN65.2 genome, ten randomly selected mutants were grown overnight in TSB, and the genomic DNA was isolated, digested by BamHI, and separated on 0.8 % agarose gel in Tris-acetate EDTA buffer. The restriction fragments were blotted onto an Amersham Hybond-N membrane (GE healthcare) (Sambrook and Russell 2001). The Km resistance gene located on the mini-Tn5 was amplified by PCR using primers Km722R (5′-GGAGAAAACTCACCGAGGCA-3′) and Km377F (5′-TGTTCCTGCGCCGGTTGCAT-3′) (Fida et al. 2014) and digoxygenin (DIG) labeled nucleotides (Roche Diagnostics). The (DIG)-labeled fragment was used as a probe in hybridization with the blotted DNA fragments. Detection of hybridization signals was performed with an anti-digoxigenin antibody coupled with an alkaline phosphatase by means of the CDP-star chemiluminescent substrate (Roche Diagnostics) according to the recommendations of the manufacturer.

A mutant library of 2,628 exconjugant colonies was stored in 15 % glycerol in 96-well microtiter plates (Greiner Bio-One) after growth in MM glucose (2 g/L) supplemented with Km (25 mg/L). The library was screened for mutants affected in carbofuran catabolism by incubating the mutants overnight at 25 °C in 100 μL MM containing glucose (2 g/L) and Km (25 mg/L) in 96-well microtiter plates after which 5 μL of each mutant culture was transferred onto two square Petri dishes (12 × 12 cm; Greiner Bio-One) containing MM agar supplemented with either glucose (2 g/L) or carbofuran (500 mg/L). The plates were incubated at 25 °C and inspected daily. Mutants showing no growth on carbofuran but growth on glucose or decreased growth on carbofuran as compared to the wild type were retained and their phenotype checked on agar and liquid MM supplemented with either glucose (2 g/L) or carbofuran (200 or 500 mg/L for liquid or agar medium, respectively). The screening media always contained 25 mg/L Km.

Phenotypic characterization of mutants affected in carbofuran catabolism

Growth of mutants affected in growth on carbofuran was assessed as follows. Overnight cultures of strain KN65.2 and the mutants grown in TSB to an OD600 of 0.8 were diluted 100-fold in MM supplemented with carbofuran (200 mg/L), after which 300 μL of each triplicate culture was transferred to a Honeycomb plate. Their growth at 25 °C was monitored in a Bioscreen C (Thermo Scientific) apparatus. A wideband filter between 420 and 580 nm was used to measure the OD of the cultures every 30 min during 5 days. The cultures were shaken for 2 min prior to each turbidity measurement.

Degradation of carbofuran was assayed as follows. Precultures of the mutants were grown in TSB at 25 °C until OD600 of 1.0 was attained and 10 μL of them used to inoculate10 mL MM supplemented with 200 mg/L carbofuran in 30-mL test tubes. The inoculated vials were incubated at 25 °C while shaking at 125 rpm. Vials inoculated with the wild-type KN65.2 strain or without inoculum were included as positive and negative controls, respectively. Samples of 700 μL of the suspensions were taken daily, centrifuged at 12,000 × g for 5 min and the residual carbofuran concentration in the supernatant analyzed by reverse phase high-pressure liquid chromatography (HPLC) (LaChrom, Merck Hitachi) using an Alltima HP C-18 column (4.6 mm × 250 mm, particle size 5 μ) and a UV-VIS detector set at 223 nm, or by means of ultrafast liquid chromatography (UFLC) (Nexera, Shimadzu) using a Platinum EPS C-18 column (4.6 mm × 150 mm, 100 Å, particle size 3 μm) that was heated to 40 °C and a UV-VIS detector set at 210 nm. In the case of HPLC, the mobile phase consisted of water/acetonitrile (40:60) with a flow rate of 1 mL/min. The sample of 20 μL was injected for measurement and carbofuran eluted after 4.4 min. In the case of UFLC, the mobile phase was water/acetonitrile (70:30) with a flow rate of 1 mL/min. One microliter of the sample was injected and carbofuran eluted after 5.7 min. In an identical way, UFLC was used to detect appearance and/or degradation of carbofuran phenol (elution time of 4.9 min). To this end, carbofuran phenol (99 % purity; Dr. Ehrenstorfer GmbH) was used as a reference compound.

Mineralization of [benzene ring-U-14C] (specific activity 4.90 MBq/mg) carbofuran in MM was performed as described previously (Uyttebroek et al. 2006). Briefly, precultures of the mutants were grown in TSB at 25 °C until an OD600 of 1.0 was attained and 5 μL of the precultures used to inoculate in triplicate 5 mL MM supplemented with 10 μg/L labeled (4.9 × 104 Bq/L) and 990 μg/L nonlabeled carbofuran. The inoculated vials were incubated at 25 °C while shaking at 125 rpm. Vials inoculated with the wild-type KN65.2 strain or without inoculum were included as positive and negative controls, respectively. Cumulative carbofuran mineralization curves were calculated by dividing the cumulative amount of evolved 14CO2 by the initial amount of 14C-carbofuran added to the medium.

Plasposon rescue and identification of the mutated genes

Genomic DNA of mutants impaired in growth on carbofuran was isolated using the Puregene Core Kit A (Qiagen) and digested overnight at 37 °C by BamHI (New England Biolabs) that does not cut inside the plasposon. The digested DNAs were purified using the QIAquick PCR purification kit (Qiagen) and self-ligated overnight at 16 °C by T4 DNA ligase (Westburg). The ligation mixtures were introduced into E. coli S17-1 λpir by heat-shock transformation (Sambrook and Russell 2001) and the transformants selected on LB agar supplemented with 25 mg/L Km. Km-resistant E.coli colonies were recovered and plasmids extracted using the Invisorb Spin Plasmid Mini Two kit (Westburg). The transposon flanking genes were analyzed by Sanger sequencing using pRL27-specific primer pair tpnRL17-1 (5′-AACAAGCCAGGGATGTAACG-3′) and tpnRL13-2 (5′-CAGCAACACCTTCTTCACGA-3′), complementary to the transposon and oriented toward the flanking genes (Larsen et al. 2002). The Geneious R6.1.4 software (http://www.geneious.com) was used to localize the rescued plasposon insertion sites in the draft genome sequence of strain KN65.2.

Transcriptional analysis

Ten milliliters of an overnight preculture of strain KN65.2, grown in MM supplemented with glucose (2 g/L) until an OD600 of 1.0, was used to inoculate 20 mL MM supplemented with glucose (2 g/L) and carbofuran (100 mg/L), as well as 20 mL MM supplemented with glucose (2 g/L). The cultures were incubated at 25 °C and aerated at 125 rpm. At different time points of incubation (10, 20, 30, 40, 50, 60, 80, 100, 120, 180, and 240 min), 700 μL of each culture was sampled for monitoring degradation. In parallel, 1-mL samples were taken for later RNA extraction at a time point (40 min) where carbofuran was no longer detected and carbofuran phenol was still present. Immediately after sampling, 200 μL stop solution containing 95 % ethanol and 5 % water-saturated phenol (Roth) was added to the samples for RNA analysis to avoid RNA degradation. The samples were snap-frozen and stored at −80 °C until extraction. RNA extraction was performed using the modified SV total RNA isolation protocol (Promega) and residual DNA removed by using the Turbo DNA-free kit (Ambion). RNA was precipitated by ethanol with 1:50 volume of 5 mg/L glycogen (Thermo Scientific) and the absence of residual DNA checked by PCR using primers targeting the cfdC gene (Online resource 1). The concentration of total RNA was quantified by a Nanodrop ND-1000 Spectrophotometer (Thermo Scientific). RNA extracts were diluted to an equal concentration prior to conversion to complementary DNA (cDNA) by RevertAid H Minus First-Strand cDNA synthesis kit (Fermentas). Primer3 (Untergasser et al. 2012) was used to design primers for RT-PCR detection of transcripts of selected genes as well as their transcriptional linkage. The target genes, sequences of the corresponding primers, and PCR conditions are listed in Online resource 1.

Sample preparation for protein and metabolite analyses of mutants affected in carbofuran catabolism

Precultures of 25 mL of Novosphingobium sp. KN65.2 and selected mutants were grown overnight in TSB until an OD600 of 1.0 was attained. The cells were collected by centrifugation at 6,000 × g for 15 min at 20 °C and washed twice in MM. The cell pellets were suspended in 1 mL MM and 500 μL inoculated into 9.5 mL MM supplemented with carbofuran (200 mg/L) and Km (25 mg/L). For each of the strains, the experiment was done in duplicate, and control incubations without inoculum were included. After 30 min of incubation, the cultures were centrifuged at 6,000 × g for 15 min at 4 °C and the supernatants filtered over 0.22-μm filters. The cell pellets and the filtrates were used for proteomic and metabolic analyses, respectively.

Differential proteomic analysis

Proteins extracted from the biomass of strain KN65.2 and selected mutants were analyzed by liquid chromatography tandem mass spectrometry (LC-MS/MS). Protein extraction, tryptic digestion, LC-MS/MS analysis, protein identification, and quantification are detailed in Online resource 2. Protein identification and quantification were performed using a label-free strategy on a UHPLC-HRMS platform composed of an Eksigent 2D liquid chromatograph and an AB SCIEX TripleTOF™ 5600. Protein searches were performed against a predicted proteome dataset derived from the genomic annotation for Novosphingobium sp. KN65.2 using ProteinPilot software v4.1. In addition, all peptides were manually inspected (Held et al. 2013).

Identification of carbofuran metabolites

Carbofuran metabolites were identified by HPLC coupled with a quadrupole-orbitrap mass spectrometer (QExactive, Thermo Scientific) using a method previously described (Helbling et al. 2010). XCalibur v2.0.7 software (Thermo Scientific) was used for chromatogram analysis and interpretation (Online resource 3). Structures of metabolites formed within the mutant incubation experiments were proposed following either target and nontarget analyses of mass acquisitions as previously described (Helbling et al. 2010). Target analysis was based on University of Minnesota Pathway Prediction System (UM-PPS; Gao et al. 2011) predictions of plausible metabolite structures over two generations of transformations. Nontarget analysis was with a previously described workflow incorporating a mass and retention time domain constraint, a background subtraction algorithm, a constrained molecular formula fit, and a plausibility check based on the presence of 13C monoisotopic masses (Helbling et al. 2010). Candidate metabolite masses identified with either the target or nontarget analysis were further confirmed or rejected as actual metabolites following manual inspection of MS spectra for the relative abundance of 13C monoisotopic masses and data-dependent MS/MS spectra acquisitions which were triggered when peaks were detected in the full scan at the exact masses of any candidate metabolite. The Mass Frontier software (Thermo Scientific) was used to compare acquired MS/MS fragments with predicted MS/MS fragment spectra of the proposed metabolite structures.

Nucleotide sequence accession number

The Whole Genome Shotgun project reporting on the genome sequence of Novosphingobium sp. KN65.2, including ORF predictions, was deposited at The European Nucleotide Archive under accession number CCBH000000000. The version described in this paper is the first version, CCBH010000000 (http://www.ebi.ac.uk/ena/data/view/CCBH010000001-CCBH010000243).

Results

Isolation and identification of carbofuran-degrading strain KN65.2

Strain KN65.2 uses carbofuran as a sole source of carbon and nitrogen (Online resource 4). A pink to light red color was apparent in the cultures, which was more intense in the medium without additional nitrogen source (ammonium sulfate). Mineralization experiments with 14C-ring-labeled carbofuran demonstrated its ability to degrade the aromatic moiety of the pesticide. The mineralization of carbofuran by a cell suspension of resting KN65.2 cells is shown in Fig. 1. No difference in carbofuran mineralization kinetics was observed between carbofuran-grown and glucose-grown cells suggesting the rate-determining step(s) in carbofuran degradation to be constitutive (Fig. 1).
Fig. 1

Mineralization of ring-labeled 14C-carbofuran (10 μg/L) in PBS buffer by resting cells of strain KN65.2 precultured in MM with glucose (2 g/L) (white circle) or in MM with glucose (2 g/L) and carbofuran (200 mg/L) (black circle)

The 16S ribosomal RNA (rRNA) gene sequence of strain KN65.2 indicated its affiliation with sphingomonads. This was confirmed by further analysis of the draft genome sequence of strain KN65.2, consisting of 243 contigs with an average length of 20.6 kb and a GC content of 63.1 %. A phylogenetic tree based on an alignment of concatenated amino acid sequences from 25 proteins with housekeeping function was constructed to scrutinize the taxonomic position of strain KN65.2 among sphingomonads (Online resource 5). The inferred tree topology, resolving four distinct clusters for Novosphingobium, Sphingobium, Sphingomonas, and Sphingopyxis, corresponds well with the genome-based phylogeny described by Aylward et al. (2013). Strain KN65.2 is very closely related to sediment isolate Novosphingobium pentaromativorans US6-1 (Luo et al. 2012) and marine isolate Novosphingobium sp. PP1Y (D’Argenio et al. 2011), sharing with these strains ~98 % amino acid identity for about 10,200 amino acids analyzed.

Isolation and phenotypic characterization of mutants with altered carbofuran utilization

The mini-Tn5-carrying plasposon from pRL27 was successfully introduced into the genome of Novosphingobium sp. KN65.2 by conjugation with E. coli BW20767. Southern blot hybridization of ten randomly selected mutants showed that the plasposon had inserted as one copy and randomly within the genome (data not shown). The KN65.2 mutant library was screened for mutants with an altered phenotype regarding growth on solid medium with carbofuran as sole carbon source, with their growth on glucose remaining unaffected. Twenty-eight such mutants were obtained. Further characterization of these 28 mutants was based on their phenotypes in carbofuran-supplemented liquid MM regarding growth on, degradation of, and mineralization of carbofuran. Bioscreen analysis confirmed that these mutants were affected in growth on MM supplemented with carbofuran, showing either lack of detectable growth, growth at reduced rate to a cell density lower than the wild type, or delayed growth after a prolonged lag phase to a final cell density similar to the wild type (Fig. 2). Carbofuran degradation of the mutants was assessed by HPLC-monitoring of residual carbofuran in the culture supernatant. While four of the mutants retained a degradation rate comparable to the wild-type strain, degradation of carbofuran was decreased in rate for most mutants. Carbofuran degradation kinetics of representative mutants is shown in Fig. 3 and Online resource 6. With the exception of two mutants, the ability to degrade the aromatic ring, as measured by labeled CO2 release from aromatic ring 14C-labeled carbofuran, was not abolished, although the mineralization was either occurring more slowly or reduced to a lower final total extent of CO2 production as compared to the wild-type strain. Carbofuran mineralization activity of representative mutants is shown in Fig. 3 and Online resource 6.
Fig. 2

Representative growth patterns of Novosphingobium sp. KN65.2 (black circle) and selected mutants in MM supplemented with carbofuran (200 mg/L) inoculated with 105 cells/mL. Delayed (white circle), reduced (inverted black triangle), and lack of detectable growth (white triangle) were recorded for mutants 3A6 (group I), 18E9 (group I), and 21D7 (group II), respectively

Fig. 3

Degradation of carbofuran (200 mg/L; 0.9 mM) (black circle), formation of carbofuran phenol (white circle) in the medium, and mineralization of ring-labeled 14C-carbofuran (10 μg/L 14C-labeled carbofuran supplemented with 990 μg/L nonlabeled carbofuran) (black diamond) by Novosphingobium sp. KN65.2 (a) and selected mutants with representative phenotypes affected in carbofuran degradation (bf). Mutant name and mutant group (I through V) are indicated. The profiles of additional mutants are shown in Online resource 6

Based on the combinations of the three carbofuran-related phenotypes (growth-degradation-mineralization), the selected mutants were tentatively assigned to five different groups (I to V) of phenotypically similar mutants (Table 1). Ten mutants, at least one from each group, were selected to verify whether carbofuran phenol, the presumed first hydrolytic metabolite, was detected in the culture medium during carbofuran degradation. In Fig. 3 and Online resource 6, the time course of carbofuran phenol appearance, if any, in the culture medium is shown, along with the degradation of carbofuran and its apparent mineralization (measured as release of labeled CO2 originating from the aromatic ring) for these mutants. The carbofuran hydrolysis product could not be detected in the wild-type KN65.2 medium, although its presence before the first sampling (after 1 day) cannot be ruled out (Fig. 3a). Carbofuran phenol was detected after 1 day of incubation for the three group I mutants (3A6, 21D12, 23F7), but it was further converted and no longer detected the next day (Fig. 3b; Online resource 6b, c). Although showing wild-type mineralization, the group I mutants showed a delayed or reduced growth phenotype. The group II mutant 21D7 exhibited accumulation of carbofuran phenol until day 2, after which its concentration slowly decreased to become undetectable at day 6 (Fig. 3c). The representative of group III, mutant 11B3, also accumulated carbofuran phenol in the medium but without apparent further conversion (Fig. 3d). This pattern is also displayed by the mutants from group IV (11A4, 16D4, 24H6) and from group V (8A8, 9G4), although the groups III, IV, and V differ in aromatic ring mineralization capacity (Fig. 3d–f; Online resource 6d–f). However, in these mutants, no stoichiometric conversion of carbofuran to carbofuran phenol was noted, since the carbofuran phenol concentration only amounted maximally until around 50 % of the added carbofuran concentration (Fig. 3f and Online resource 6f).
Table 1

Characteristics of Novosphingobium sp. KN65.2 mutants. Phenotypes for growth (G, 200 mg/L carbofuran), degradation (De; 200 mg/L carbofuran), and mineralization (M; 10 μg/L14 C-labeled carbofuran and 990 μg/L nonlabeled carbofuran) in carbofuran-containing minimal medium

Group

Mutant

Phenotypes in carbofuran medium

Gene

Metabolite analysis of mutants

G

De

M

1

2*

3

4*

5*

6*

7*

X*

I

3A6

D

WT

WT

 

+

ND

+

+

ND

ND

ND

ND

3D7,12C8, 18D2, 29E11, 29 F11

D

NA

WT

 

NA

NA

NA

NA

NA

NA

NA

NA

21D12

D

WT

WT

 

+

ND

+

+

ND

ND

ND

ND

23 F7

D

WT

WT

 

ND

+

ND

ND

ND

ND

ND

NA

10D9, 17G2, 22A9, 23C7

D

NA

WT

 

NA

NA

NA

NA

NA

NA

NA

NA

18E9

D, R

NA

WT

 

NA

NA

NA

NA

NA

NA

NA

NA

II

21D7

WT

WT

 

+

ND

+

+

ND

ND

ND

ND

6C1

NA

WT

 

NA

NA

NA

NA

NA

NA

NA

NA

III

11B3

D

R

cfdH

+

ND

ND

ND

ND

ND

+

ND

12E8, 20D2, 15E9

NA

R

 

NA

NA

NA

NA

NA

NA

NA

NA

IV

11A4

D

R

cfdE

+

ND

ND

ND

ND

+

ND

+

24H6

D

R

cfdI

+

ND

ND

ND

+

ND

ND

ND

16D4

D

R

cfdD

+

ND

ND

ND

ND

ND

ND

+

9D6, 9 F5, 10G6, 20H6

NA

R

 

NA

NA

NA

NA

NA

NA

NA

NA

V

9G4

D

cfdC

+

ND

ND

ND

ND

ND

ND

ND

8A8

D

cftA

+

ND

ND

ND

ND

ND

ND

ND

WT wild-type phenotype, D decreased in rate, R reduced in final OD (growth assay) or accumulated 14CO2 production (mineralization assay), X unidentified metabolite, ND not detected or only in very low amount, NA not analyzed, − no detectable growth or no detectable mineralization

Mutants used for metabolite analysis are underlined. The corresponding inactivated ORFs and genes with a proposed function in carbofuran degradation (cftA, cfdABCDEFGH operon, cfdI) are described in Online resource 7 and Table 2. Metabolites that accumulated in the mutants are marked with + and those not previously described are labeled with *. The (proposed) structures are shown in Fig. 6 and Online resource 3

Identification of genes affected in carbofuran utilization mutants

The plasposon with flanking genomic DNA of the KN65.2 mutants impaired in carbofuran degradation was cloned and sequenced. This enabled mapping of the insertion sites on the contigs of the draft genome sequence of strain KN65.2. The characteristics of the wild-type proteins encoded by the interrupted coding regions are compiled in Online resource 7. Some genes were hit by different insertions (up to four) in independent mutants, while mutants 9D6 and 9F5 (group IV) carried the insertion at an identical position.

Overall, in 27 independent mutants, 19 different genes were affected. Thirteen mutants displayed the group I phenotype, characterized by a delayed growth or reduced growth rate with carbofuran as sole carbon source. Several genes disrupted in group I mutants likely function in one-carbon (C1) metabolism as their deduced proteins act as components of the activated methyl cycle (in mutants 17G2 and 21D12), folate biosynthesis (in mutant 3A6), or folate-dependent reactions (in mutant 29E11). The latter mutant appears to be affected in a homolog (with 68 % amino acid identity) of a tetrahydrofolate-dependent O-demethylase from Sphingomonas paucimobilis SYK-6 (Masai et al. 2004). The genomic clustering of the 3A6- and 29E11-mutated genes with other folate metabolism-related genes supports this hypothesis (Bermingham and Derrick 2002) (Online resource 8 and 9, cluster 1). The group I phenotype was also linked with inactivation of two putative regulatory genes (in mutants 3D7 and 12C8), of a presumed dehydrogenase (in mutant 18D2), of four genes predicted to be involved in metabolism of coenzyme A-activated substrates (in mutants 10D9/23F7, 18E9, 22A9, and 29F11), and of a putative dioxygenase gene (in mutant 23C7) in which also the insertions of the group IV mutants 11A4 and 10G6 were mapped (further discussed below). Several of the mutated genomic regions of group I and group II mutants show an operon-like organization (Online resource 8 and 9). The genes downstream of the ketothiolase inactivated in mutant 18E9 may be required for reactions similar to the propionyl-CoA to succinyl-CoA conversion. The putative dehydrogenase affected in mutant 18D2 appears to be part of a putative oxidative catabolic operon. The cobW gene is the first gene of the cobalamin biosynthetic operon (in mutant 17G2). All gene products affected in group I mutants, except those associated with mutants 23C7 and 18D2, share a high percentage of amino sequence identity and a highly similar genetic context with counterparts in other sphingomonads, in particular Novosphingobium spp., in line with the phylogenetic analysis of housekeeping genes (Online resource 5, 7, and 9).

The group II phenotype, characterized by lack of growth on carbofuran concomitant with and degradation and mineralization of carbofuran similar to those of the wild type, was associated with deficiency in another dehydrogenase (in mutant 21D7) and in another gene function of CoA-activated substrate conversion (in mutant 6C1). The dehydrogenase gene affected in mutant 21D7 is located upstream of a convergent set of genes predicted to be involved in R-CoA conversions (Online resource 8 and 9). The insertions in mutants 6C1 and 21D7 mapped in the same contig. As observed for most of the group I mutants, gene products affected in group II mutants are well conserved among Novosphingobium spp. (Online resource 7 and 9).

A putative CoA-transferase gene, yet another gene function of CoA-activated substrate conversion, was inactivated in the four independent mutants of group III (11B3, 12E8, 15E9, 20D2), displaying abolished growth on and reduced degradation and mineralization of carbofuran. Interestingly, the 8-kb contig containing this gene does not have a homologous counterpart in other Novosphingobium spp. strains for whom the genome sequencing is (partially) known nor in other bacteria. Moreover, within this contig, in the region upstream of this putative CoA-transferase gene, plasposon mutations were identified in three other genes, either assigned to the group V mutant 9G4 (putative hydroxylase) or to the group IV mutants 16D4/20H6 and 11A4/10G6 (putative dioxygenase). The group I mutant 23C7 has an insertion in the same dioxygenase gene as the latter mutants but displays a different phenotype. This may be due to location of the insertion in 23C7, just in front of the stop codon, which may still allow production of (partially) functional gene product. Because of its uniqueness to Novosphingobium sp. KN65.2 and the multiple transposon insertions leading to abolished growth on carbofuran, this 8-kb region is a good candidate for a specific determinant of carbofuran degradation (Fig. 4) and, hence, crucial for productive metabolism of carbofuran by strain KN65.2. This genomic region with eight convergent ORFs displays an operon-like organization and was designated as carbofuran degradation region cfdABCDEFGH (Table 2). A putative regulator gene of the TetR family (Cuthbertson and Nodwell 2013), designated cfdA, precedes seven metabolic genes, of which cfdG and cfdH are likely involved in conversion of CoA-activated intermediates (Hamed et al. 2008; Heider 2001). CfdH (containing insertions in group III mutants 20D2, 12E8, 15E9, and 11B3) belongs to the same protein family as the gene product inactivated in the group I mutant 29F11, but they share only marginal sequence homology (32 % amino acid identity), suggesting a different involvement in carbofuran catabolism. CfdC (containing the insertion in group V mutant 9G4), CfdE (containing insertions in group IV mutants 11A4 and 10G6 as well in group I mutant 23C7), and CfdF show only moderate homology to hypothetical proteins, but a triplet of distantly related equivalent enzymes, HsaA (hydroxylase)-HsaC (dioxygenase)-HsaD (hydrolase), respectively, is involved in cholesterol catabolism by actinomycetes such as Mycobacterium, Rhodococcus, and Gordonia (Chen et al. 2012; Dresen et al. 2010; van der Geize et al. 2007). The gene product of cfdD (containing insertions in group IV mutants 16D4 and 20H6) is a member of the diverse fumarylacetoacetate hydrolase family, typically involved in bacterial meta-cleavage pathways for degradation of aromatic compounds, but also including various (de)hydratase, decarboxylase, and isomerization enzymes (Ran et al. 2013). Notably, the cholesterol catabolic enzyme HsaE (2-hydroxypentadienoate hydratase) belongs to the same family.
Fig. 4

Gene organization (a) and transcriptional analysis (b) of the Novosphingobium sp. KN65.2 cfdABCDEFGH gene cluster. aArrows indicate the localization and direction of ORFs. Differential shading refers to putative functions in carbofuran catabolism as indicated. A few codons are missing from the cfdA gene on the assembled contig. Solid lines indicate the position of plasposon insertion sites. Mutants used for metabolite analysis are underlined, and the corresponding mutant group is indicated between brackets. b Transcript (upper panels) and transcriptional linkage analysis by RT-PCR (lower panels) of cfdA (A) through cfdH (H). Left panels from cells grown on MM glucose supplemented with carbofuran. Right panels from cells grown on MM glucose. Size markers (M) are indicated. The expected size of the amplicons generated from adjacent genes (pairs AB through GH) is described in Online resource 1

Table 2

Homology and Pfam analysis of the cfd genes and context of the cftA gene involved in carbofuran degradation in Novosphingobium sp. KN65.2. Genes carrying a plasposon insertion are labeled with * and the corresponding mutant names are indicated

Cluster

Gene

Mutant

ORFa

Size (AA)

Homology

Pfam

Identity

Protein (accession nr)

Organism

Accession

Description

cfd

cfdA

SPHv1_1690003

194b

42 %

putative TetR family transcriptional regulator (GAD47807)

N. tardaugens NBRC 16725

PF00440

TetR family N-terminal HTH motif

 

cfdB

SPHv1_1690002

33

NA

NA

NA

NA

NA

 

cfdC*

9G4

SPHv1_1690001

385

48 %

putative hydroxylase (GAD47808)

N. tardaugens NBRC 16725

PF02771/PF08028

N-terminal domain/C-terminal domain acyl-CoA dehydrogenase

   

32 %

3-HSAc hydroxylase HsaA (ABG96325)

R. jostii RHA1

   

31 %

3-HSA hydroxylase HsaA (NP_218087)

Myc. tuberculosis H37Rv

 

cfdD*

16D4, 20H6

SPHv1_860006

325

45 %

putative hydrolase (ADD29301)

M. ruber DSM1279

PF01557

fumarylacetoacetate hydrolase superfamily

   

25 %

putative fumarylacetoacetase (GAB04531)

G. amarae NBRC 15530

 

cfdE*

10G6, 11A4, 23C7

SPHv1_860004

425

45 %

putative dioxygenase (GAB04530)

G. amarae NBRC 15530

PF00903/ PF00903

glyoxalase/bleomycin resistance protein/dioxygenase superfamily

   

32 %

3,4-DHSAd dioxygenase HsaC (ABG96327)

R. jostii RHA1

   

31 %

3,4-DHSA dioxygenase HsaC (NP_218085)

Myc. tuberculosis H37Rv

 

cfdF

SPHv1_860003

298

72 %

hypothetical protein (EQB11812)

S. lactosutens DS20

PF12697

alpha/beta-hydrolase family 6

   

22 %

4,9-DSHAe hydrolase HsaD (ABG96326)

R. jostii RHA1

   

21 %

4,9-DSHA hydrolase HsaD (NP_218086)

Myc. tuberculosis H37Rv

 

cfdG

SPHv1_860002

272

51 %

hypothetical protein (EJE50735)

Acidovorax sp. CF316

PF00378

crotonase/enoyl-CoA hydratase/isomerasesuperfamily

 

cfdH*

11B3, 12E8, 15E9, 20D2

SPHv1_860001

391

54 %

putative CoA-transferase (EDP65782)

Alpha-proteobacterium BAL199

PF02515

CoA-transferase family III

cft

SPHv1_2290059

180

59 %

putative ECF σ factor (EQB13445)

N. lindaniclasticum LE124

PF08281

sigma70 region4

 

SPHv1_2290058

341

50 %

putative σ factor regulator (EQB13446)

N. lindaniclasticum LE124

PF04773

FecR protein

 

cftA*

8A8

SPHv1_2290057

1067

76 %

putative TonB-dependent receptor (EQB13447)

N. lindaniclasticum LE124

PF07660/PF07715/PF14905

secretin-TonB N-terminus domain/TonB- dependent receptor plug domain/outer membrane protein beta-barrel family

 

SPHv1_2290056

422

75 %

putative protein phosphatase (EQB13448)

N. lindaniclasticum LE124

PF00149

Calcineurin-like phosphoesterase

cfdI*

24H6

SPHv1_1160001/ SPHv1_1560004f

584

51 %

putative monooxygenase (EKX01755)

Pseudomonas sp. M1

PF01494

FAD binding domain 3

AA amino acids, G. Gordonia, M. Meiothermus, Myc. Mycobacterium, N. Novosphingobium, R. Rhodococcus, S. Sphingobium, NA not applicable (no significant homology)

aORF according to MAGE annotation

bN-terminal amino acid sequence missing

c3,4-HSA 3-hydroxy-9,10-seconandrost-1,3,5(10)-triene-9,17-dione

d3,4-DHSA 3,4-dihydroxy-9,10-seconandrost-1,3,5(10)-triene-9,17-dione

e4,9-DSHA 4,5–9,10-diseco-3-hydroxy-5,9,17-trioxoandrosta-1(10),2-diene-4-oic acid

fthe coding region of cfdI is located on two separate contigs in the initial assembly; the gene context of cfdI suggests that it is not part of an operon

In addition to the putative mono-oxygenating enzyme CfdC and the dioxygenating enzyme CfdE, a third predicted putative oxygenase was affected in two other independent group IV mutants (9D6/9F5, 24H6), displaying abolished growth along with diminished degradation and mineralization of carbofuran. Also, the gene containing this oxygenase appears unique to strain KN65.2, and the gene was named cfdI. cfdI apparently codes for a mono-oxygenase that belongs to a family different from this of CfdC and is not linked to the cfdABCDEFGH region nor part of another operon.

The insertion in the remaining group V mutant 8A8, displaying abolished growth on carbofuran, no carbofuran mineralization, and carbofuran degradation with accumulation of carbofuran phenol (Fig. 3f), was located in a gene encoding a putative TonB-dependent receptor. This gene is located in a separate contig and hence a genomic position different from the insertion in cfdC in mutant 9G4 that showed the same phenotype as mutant 8A8 (Online resource 6f). This outer membrane protein, designated CftA (encoded by cftA, carbofuran transport-related gene A), bears significant homology with putative transporters in hexachlorocyclohexane-degrading Novosphingobium lindaniclasticum LE124 (Saxena et al. 2013) and marine isolate Brevundimonas sp. BAL3, but it is apparently not conserved in sphingomonads or other alpha-proteobacteria (Fig. 5). Interestingly, the gene context in strains KN65.2, LE124, and BAL3 is very similar, composing a four-gene module together with an RNA polymerase σ factor, a σ factor regulator, and a phosphatase (Fig. 5).
Fig. 5

Genomic context of the Novosphingobium sp. KN65.2 cftA gene that is insertionally inactivated in group V mutant strain 8A8. Differential shading refers to predicted functions as reported at the right. The plasposon insertion site is indicated by a solid line. The corresponding homologous genomic regions of Novosphingobium lindaniclasticum LE124 and Brevundimonas sp. BAL3 are shown for comparison. The percentages of amino acid sequence identities of equivalent genes between two organisms are indicated in the gray arrows. No homology was detected for the genes flanking the four-gene syntenic region

Transcriptional analysis of the cfd gene cluster

Since the cfd gene cluster appears to play a crucial role in carbofuran metabolism in strain KN65.2 and shows an operon-like structure, transcriptional analysis of the gene cluster was performed on KN65.2 cells grown in MM supplemented with glucose and carbofuran and in MM glucose medium lacking the pesticide. The transcripts of cfdA-H genes were detected in cells grown on MM supplemented with glucose and carbofuran, but also on glucose medium lacking the pesticide (Fig. 4), suggesting constitutive expression of the cfd operon. In addition, pairs of adjacent genes in the cluster, from cfdA through cfdH, yielded RT-PCR amplicons of expected sizes for RNA extracted from both carbofuran-supplemented and carbofuran-lacking media, indicating that the genes indeed constitute an operon.

Identification of carbofuran metabolites

To obtain more insight in carbofuran catabolism by Novosphingobium sp. KN65.2 and assign a possible role to the gene functions identified by mutational analysis, the wild type and ten mutants representing the different carbofuran degradation and mineralization phenotypes (Table 1) were subjected to target and nontarget screening for carbofuran catabolic products.

The screening of wild-type culture supernatant of carbofuran-supplemented MM revealed the formation of only a single metabolite with an exact mass of 343.1178 Da (compound X). Based on the high-mass accuracy data and the acquired isotopic pattern, its proposed molecular formula is C19H18O6; however, its structure is not yet identified. Incubations with the mutants resulted in the suggestion of seven other carbofuran metabolites, of which two (1, 3) were described previously (Table 1; Fig. 6). The MS spectra and proposed structures of the metabolites are documented in Online resource 3. Metabolite 1 (carbofuran phenol) resulting from carbamate hydrolysis was previously reported to also be formed during carbofuran degradation by Sphingomonas sp. SB5 (Kim et al. 2004), Pseudomonas sp. 50432 (Chaudhry et al. 2002), and Novosphingobium sp. FND-3 (Yan et al. 2007) and its occurence in the present work was confirmed with an authentic standard. Metabolite 3 (3-(2-hydroxy-2-methylpropyl)benzene-1,2-diol reflects oxidative furan ring opening of metabolite 1 and was likewise identified during carbofuran degradation by Sphingomonas sp. SB5 and Novosphingobium sp. FND-3 (Kim et al. 2004; Yan et al. 2007). For the latter strain, an open-ring metabolite retaining the intact carbamate bond was also reported, but such a compound was not identified for strain KN65.2, indicating that furan ring opening required prior hydrolysis of the carbamate ester bond. The structures of metabolites 2 (3-(2-hydroxy-2-methylpropyl)cyclohexa-3,5-diene-1,2-dione), 4 ((2E,4Z)-2,8-dihydroxy-8-methyl-6-oxonona-2,4-dienoic acid), and 5 (3-hydroxy-3-methylbutanoic acid) were proposed primarily based on the assigned molecular formulae, monoisotopic patterns, MS/MS spectra, and plausible biochemical mechanisms (Online resource 3). Metabolite 6 is the product of an apparent hydroxylation of carbofuran, proposed to modify a methyl group attached to the furan ring (Fig. 6). An apparent hydroxylation product of metabolite 3 accumulated in mutant 11B3, but the exact location of the hydroxyl group in this compound 7 could not be resolved (Online resource 3).
Fig. 6

Tentative pathway for the degradation of carbofuran in Novosphingobium sp. KN65.2 with proposed functions of cfd gene products. 1 carbofuran phenol; 2 3-(2-hydroxy-2-methylpropyl)cyclohexa-3,5-diene-1,2-dione; 3 3-(2-hydroxy-2-methylpropyl)benzene-1,2-diol; 4 (2E,4Z)-2,8-dihydroxy-8-methyl-6-oxonona-2,4-dienoic acid; 5 3-hydroxy-3-methyl butanoic acid; 6 2-(hydroxymethyl)-2-methyl-2,3-dihydrobenzofuran-7-yl methylcarbamate. A possible structure for the putative hydroxylation product of 3, accumulating in the cfdH mutant, is shown in Online resource 3 (metabolite 7). Compound 8 represents the putative intermediate formed during conversion of metabolite 1 to metabolite 2. The presumed hydrolysis product of metabolite 4, 2-oxopent-4-enoate (9), was not demonstrated, but the enzymes required for its conversion to the central metabolites pyruvate and acetyl-CoA are encoded by the Novosphingobium sp. KN65.2 genome. Further hypothetical steps for conversion of metabolite 6 to a side-chain-hydroxylated analog of compound 3 are illustrated in Online resource 10. Asterisk, newly proposed metabolite

Proteomic analysis of mutants affected in carbofuran degradation

The proteome of selected mutants from each group were compared with the proteome of the wild-type strain to (i) confirm the abolishment of the plasposon-affected gene products and (ii) examine whether mutations affected protein expression of other genes, for instance by polar effects of the insertion on downstream genes (Online resource 2). Emphasis was on the mutants affected in the cfdABCDEFGH operon (mutants 9G4, 16D4, 11A4, and 11B3), the cfdI gene (24H6), and the cftA gene (8A8). All of the genes that had an insertion showed also decreased expression of the corresponding proteins compared to the wild-type strain. Mutations in genes of the cfdABCDEFGH gene cluster clearly also affected the expression of downstream genes and of the unlinked cfdI. Even a moderate effect on the expression of proteins encoded by genes located upstream of the affected genes is observed. Interestingly, lack of CftA (in group V mutant 8A8) abolished the expression of the proteins encoded by the cfdABCDEFGH operon and by the cfdI gene. Notably, various mutants (24H6, 8A8) showed increased expression (until 100-fold) of a protein that showed significant similarity with an ABC transporter and an increased expression of a protein with similarity to glyoxalase.

Discussion

Carbofuran-mineralizing isolate KN65.2 belongs to the genus Novosphingobium

Novosphingobium sp. strain KN65.2 is one of the few strains with demonstrated ability to utilize carbofuran as sole source of carbon, nitrogen, and energy with use/mineralization of the aromatic ring moiety, adding to the remarkable specialization of the phylogenetic clade of sphingomonads as degraders and users of carbofuran. Mineralization of the aromatic ring was, prior to Novosphingobium sp. KN65.2, also demonstrated with Sphingomonas sp. CF06 (Feng et al. 1997), Sphingomonas sp. TA, and Sphingomonas sp. CD (Ogram et al. 2000), but likely also other reported carbofuran-degrading sphingomonads use the aromatic moiety since they grow efficiently on carbofuran as sole source of carbon and energy. Many non-sphingomonad carbofuran-degrading strains have been reported, but most of them only catalyze the hydrolysis of the carbamate bond and do not degrade the aromatic ring. Reported carbofuran-degrading non-sphingomonad strains that also mineralize the aromatic ring are Pseudomonas, Flavobacterium (Chaudhry and Ali 1988; Head et al. 1992), and Arthrobacter (Ramanand et al. 1991), but unequivocal identification/taxonomic assignment of those strains is lacking without 16S rRNA gene sequence data. Carbofuran-utilizing sphingomonads were isolated from geographically distant locations including US (Feng et al. 1997; Ogram et al. 2000), China (Yan et al. 2007), Korea (Kim et al. 2004; Shin et al. 2012), and Columbia (Castellanos et al. 2013). Likely, sphingomonads have some physiological and/or genetic features that facilitate adaptation to degrade and utilize this particular compound.

Using plasposon mutagenesis, a collection of Novosphingobium sp. KN65.2 mutants affected in growth on and/or degradation of carbofuran was generated. This approach allowed to readily identify the disrupted genes and their putative affected functions by means of the plasposon rescue approach, in order to explain the mutant phenotypes. However, since upstream plasposon insertions can cause (negative) polar effects on the expression of downstream genes within transcriptional units, one should be cautious to ascribe phenotypic effects on carbofuran catabolism to individual disrupted genes. Indeed, several genes could be allocated to operon-like structures. In several cases, those operons could be linked to specific functions like C1 metabolism and metabolism of CoA-activated substrates. Among these operons, those carrying the cfd and cft genes appear to be dedicated to carbofuran metabolism. No indications were found that the contigs carrying the identified carbofuran degradation genes would be plasmid-borne. The loss of plasmid-encoded carbofuran-degrading ability in medium lacking the insecticide has been reported for unclassified strain MS2d (Head et al. 1992), but the carbofuran catabolic phenotype of Novosphingobium sp. KN65.2 appears to be quite stable under nonselective cultivation conditions. The mutants were classified in five groups according to their phenotype related to growth, degradation, and mineralization of carbofuran.

Novosphingobium sp. KN65.2 mutants affected in growth on carbofuran but retaining ring mineralization capacity

Group I mutants differ particularly from the other mutants by their ability to still grow on carbofuran though with retarded onset of growth and/or decreased growth rate compared to the wild type. Moreover, these mutants are not affected in carbofuran mineralization. A majority of the group I mutants are affected in genes that are associated with C1 metabolism. We suspect that C1 metabolism is related to the conversion of methylamine. In carbofuran-degrading sphingomonad bacteria, the first step in carbofuran degradation is the hydrolytic release of the methylamine moiety (Kim et al. 2004), which is similar to the degradation of carbofuran by other genera that use the Mcd enzyme system for hydrolysis (Tomasek and Karns 1989), and methylamine utilization in bacteria is linked with C1 metabolism. The fate of methylamine in carbofuran-utilizing sphingomonads is not clear, but enzymes of methylamine-dependent respiration were induced in the mcd-expressing methylotroph strain ER2. Attempts were made to grow Novosphingobium sp. KN65.2 on methylamine as sole source of carbon and energy at concentrations ranging from 10 to 40 mM, but no growth was observed. In fact, an operon (mauBEDA) predicted to encode the subunits of a putative amine dehydrogenase that catalyzes the oxidative deamination of a primary amine to an aldehyde and ammonia is present in the KN65.2 genome. Apparently, this system for methylamine utilization is found in only a limited number of sphingomonads, including the closely related Novosphingobium sp. PP1Y. In case methylamine is used, the reduced growth of the mutants with abolished genes in C1 metabolism compared to the wild-type strain may be explained by effects on carbofuran-associated growth by toxic intermediates in methylamine catabolism such as formaldehyde. These effects might be either directed to overall cell metabolism or specifically to carbofuran metabolism. Effects of the mutations on at least carbofuran phenol conversion are clear as carbonfuran phenol transiently accumulated, and also, the HPLC-MS analysis showed the accumulation of several metabolites.

Another major group of genes identified in group I mutants are associated with the metabolism of CoA-activated substrates and hence fatty acid-like substrates, indicating that carbofuran degradation involves enzymatic activities related to metabolism of fatty acid-like compounds. This was supported by the identification of other inactivated genes with a similar functionality in group II and group III mutants that show a more severely affected carbofuran degradation phenotype. Mineralization of the aromatic ring is not affected in these mutants, and possibly, retarded growth is due to the inability to utilize nonlabeled moieties of carbofuran. Furthermore, the group I mutants show transient accumulation of various upstream metabolites including carbofuran phenol indicating also effects on upstream enzymes and possibly overall metabolic activity which also can explain the retarded growth. Alternatively, some downstream metabolites may again exert toxicity on cellular metabolism. Effects of the mutations on at least carbofuran phenol conversion are apparent and other metabolites accumulated. We speculate that, since the affected genes are conserved in other Novosphingobium sp. strains, it concerns genes of (a) common downstream pathway(s) related to fatty acid metabolism that also can be used for conversion of (some) downstream metabolites of carbofuran.

Mutants of group II (21D7 and 6C1) show a peculiar phenotype in the sense that mineralization is similar to the wild type while degradation is slower and no growth occurs. At least mutant 21D7 also shows a prolonged transient production of carbofuran phenol and several metabolites accumulated. The discrepancy between mineralization kinetics and formation of carbofuran phenol can be explained by different substrate concentrations (1 and 200 mg/L) used in the different assays, but apparently, the proteins for mineralization of the carbofuran ring are still expressed and as active as in the wild-type strain. The genes affected in group II mutants are located close to each other and hence seem to make part of another gene cluster associated with fatty acid-related metabolism (Online resource 8, cluster 6). The gene affected in mutant 21D7 is conserved among sphingomonads indicating also a common role in fatty acid-related metabolic conversions of downstream carbofuran metabolites channeled into this pathway. The observation that mineralization is not affected while growth is abolished, points again to the accumulation of a metabolite that affects general cell metabolism rather than mineralization.

Novosphingobium sp. KN65.2 mutants affected in carbofuran ring mineralization capacity

Group III mutants were characterized by reduced degradation rate, no growth on carbofuran, and reduced carbofuran mineralization and accumulation of carbofuran phenol. All these mutants carried the insertion in the same gene, i.e., cfdH, which makes part of the apparent crucial cfdABCDEFGH operon. The gene is located in the downstream part of the cfdABCDEFGH cluster and encodes another R-CoA conversion enzyme, like cfdG that is located just upstream of cfdH, adding to the suggestion that conversions similar to those of fatty acid metabolism are crucial in carbofuran utilization. Mineralization extent was only 50 % of that displayed by the wild type indicating that only part of the aromatic ring moiety became mineralized.

Group IV mutants are characterized by slower degradation, only slight mineralization, no growth, and accumulation of carbofuran phenol. The genes affected belong to the cfd operon (cfdD and cfdE), in addition to cfdI. These genes could all be linked to aromatic degradation and likely are linked with the processing of the aromatic ring moiety of carbofuran, explaining the reduced mineralization. The metabolite profile of mutant 24H6 (cfdI) indicates that degradation can still proceed down to metabolite 5, suggesting that this putative mono-oxygenase, with moderate similarity (~47 % identity) to members of the subclass A flavoprotein mono-oxygenases (van Berkel et al. 2006), is either not involved in opening the ring structures of carbofuran phenol or is not the sole enzyme with such activity. It was noticed by proteomic analysis that in the cfdI mutant, as well as in the four cfd mutants analyzed, the amount of protein SPHv1_320038 was strongly upregulated (Online resource 2). This protein contains a similar domain (PF00903) as CfdE. However, CfdE is composed of two such domains in tandem, while SHPv1_320038 has only one. Apparent orthologs (>90 % amino acid identity) of the latter putative oxygenase from Novosphingobium sp. KN65.2 are encoded in the genomes of some phylogenetically close relatives (strains US6-1, PP1Y, LE124, LH128; Online resource 5), but the enzyme seems to be absent in most sphingomonads.

Both group V mutants, showing no mineralization, no growth, and decreased degradation, combined with carbofuran phenol accumulation, were affected in cfdC (9G4) or in cftA (8A8). cfdC encodes a putative mono-oxygenase, and we suspect that it is involved in direct attack of carbofuran phenol. Indeed, although carbofuran phenol was found in all mutants analyzed, the highest level of carbofuran phenol accumulated upon incubation with mutants 9G4 and 8A8, and none of the seven other metabolites were detected, suggesting a true bottleneck for its further degradation in these mutants. Furthermore, several further metabolites were detected in the cfdD and cfdE mutants still producing CfdC, supporting a key role for this enzyme in carbofuran phenol degradation. cftA encodes a putative TonB-dependent receptor and makes part a four-gene module together with an RNA polymerase σ factor, a σ factor regulator, and a phosphatase (Fig. 5). The organization is reminiscent of the tripartite regulatory systems with an ECF-type σ factor, a σ factor regulator (membrane-bound anti-σ factor), and a TonB-dependent transporter mediating transenvelope signal transduction of iron-siderophore complexes (Koebnik 2005). The additional putative protein phosphatase gene of KN65.2 located downstream may have a regulatory role (Kaczmarczyk et al. 2011), possibly exerted in the periplasm since its gene product contains a predicted signal peptide for Sec-dependent transport. The strong effect of the cftA mutation on expression of all the genes of the cfdABCDEFGH operon and of cfdI points toward its importance. Indeed, this mutation strongly reduced the amount of the Cfd enzymes (CfdC through CfdF), as well as the CfdI mono-oxygenase, suggesting a pronounced adverse effect on expression of the respective operon and gene (Online resource 2). Concurrently, the amount of an ABC-type transporter protein (SPHv1_600010) increased about ~100-fold. A comparable strong increase of this cytoplasmic component of a putative efflux system (~140-fold) was noted in the mutant lacking CfdI. This might be rationalized by a regulatory function for this outer membrane protein as part of a module mediating transenvelope signal transduction responsive to some metabolite (Noinaj et al. 2010). The strong induction of the efflux transporter may be due to toxic effects of accumulating carbofuran phenol when its further degradation is blocked. The gene context of cftA in strains KN65.2, LE124, and BAL3 is very similar, composing the same four-gene module (Fig. 5). An increased level of certain TonB-dependent receptors, detected by differential proteomic analysis in dioxin-degrading Sphingomonas wittichii RW1, suggests that this type of transporters plays a significant but as yet unresolved role in xenobiotic degradation (Hartmann and Armengaud 2014).

Constitutive expression of the carbofuran degradation genes

Different observations indicate that at least part of the genes are constitutively expressed. The cfd genes that appear crucial for metabolism are transcribed constitutively and not subject to induction by carbofuran. This is in line with the observed very similar carbofuran mineralization kinetics of carbofuran-induced and -noninduced resting cells. On the other hand, several genes were identified whose products can be linked to some kind of regulation such as the cfdA in the cfdABCDEFGH operon and the gene cluster containing cftA. Especially, the role of the latter in regulation of carbofuran metabolism is of interest since an insertion in cftA resulted into a complete abolishment of carbofuran phenol degradation and growth on carbofuran in mutant 8A8 and affected expression of all genes of the cfdABCDEFGH operon and cfdI. However, currently, it is unclear how this observation can be linked to the apparent constitutive expression of carbofuran mineralization and of the cfdABCDEFGH operon.

Proposed pathway for carbofuran biodegradation in Novosphingobium sp. KN65.2

The tentative pathway for carbofuran degradation shown in Fig. 6 is based on the identified metabolites and integrates current information from the phenotypic characterization of the carbofuran degradation mutants. Overall, both the metabolite analysis and the genetic analysis point toward three major steps in carbofuran degradation by strain KN65.2, i.e., hydrolysis of the carbamate bond, processing of the aromatic moiety, and further degradation of the cleaved aromatic ring through CoA-activated metabolites. In line with several reports for other carbofuran-degrading bacteria, the initial step is hydrolysis of the carbamate to the corresponding carbofuran phenol, methylamine, and carbon dioxide. However, the constitutively expressed hydrolase gene involved was not identified in this study. The Mcd enzyme might be a candidate for this initial step, but no homolog of the Achromobacter mcd gene is present in the Novosphingobium sp. KN65.2 draft genome. This enzyme also hydrolyzes carbaryl (1-naphthyl methylcarbamate) (Naqvi et al. 2009), even more efficiently than the carbaryl hydrolase CehA from Rhizobium sp. AC100 does (Hashimoto et al. 2002). The mobile element carrying cehA is also present in Novosphingobium sp. KN65.2, encoding a nearly identical copy of CehA (99.5 % amino acid identity), but the enzyme identified in strain AC100 lacks detectable activity on carbofuran (Hashimoto et al. 2002). The possible involvement of this CehA homolog in carbofuran degradation in strain KN65.2 is currently investigated. A homolog of another carbaryl hydrolase, CahA from Arthrobacter sp. RC100 (Hashimoto et al. 2006), is absent from the KN65.2 genome. Shin et al. (2012) isolated several sphingomonads degrading both carbofuran and carbaryl, but the enzymatic or genetic basis of these combined activities was not studied. As discussed above, the methylamine released from carbofuran can be likely metabolized by strain KN65.2 although we currently do not know whether this is indeed the case.

Mutant analysis assigns a crucial role for the enzymes encoded by the cfdABCDEFGH operon. Based on the current data, the most likely candidate for opening of the furan ring of carbofuran phenol is the enzyme encoded by cfdC. CfdC is distantly related to the flavin-dependent mono-oxygenase HsaA catalyzing hydroxylation during cholesterol metabolism by M. tuberculosis (Dresen et al. 2010). Ipso-hydroxylation would produce quinone 2 via the putative intermediate 8. This reaction is analogous to ipso-substitution reactions with para-alkoxyphenols in Sphingobium xenophagum (Gabriel et al. 2007) and with sulfonamides in Microbacterium sp. strain BR1 (Ricken et al. 2013). A reductase for CfdC is not encoded within the cfd operon, but such function may be provided by an unlinked gene. A homolog of the HsaA-flavin reductase HsaB (Dresen et al. 2010) is actually encoded by the gene located in the genomic region with two plasposon insertions (group II mutants 6C1, 21D7), immediately upstream of the aldehyde dehydrogenase gene inactivated in mutant 21D7 (Online resource 8 and 9, gene 6.6). Alternatively, CfdC may represent a single-component flavin-dependent mono-oxygenase (van Berkel et al. 2006).

Meta-cleavage of the previously described 3-substituted catechol 3, resulting from the reduction of quinone 2, would then yield intermediate 4, which can be hydrolyzed to products 5 and 9. Formation of the latter compound was not demonstrated, but 2-oxopent-4-enoate is a common intermediate in the metabolism of aromatics. A putative operon encoding the three enzymes for its conversion to pyruvate and acetyl-CoA (2-oxopent-4-enoatehydratase, 4-hydroxy-2-oxovalerate aldolase, and acylating acetaldehyde dehydrogenase) is located ~17 kb upstream of cluster 2 (data not shown). CfdE, a putative dioxygenase with significant homology to HsaC that catalyzes meta-cleavage of a 3-substituted catechol during cholesterol degradation (Dresen et al. 2010), can be proposed for this step. Metabolite 4 was indeed not detected in the corresponding mutant (11A4), but a novel side-chain hydroxylation product of carbofuran (6) accumulated (Fig. 6).

Two candidates for the hydrolysis of compound 4 to produce metabolite 5 are potentially encoded by the cfdABCDEFGH operon. CfdD belongs to the fumarylacetoacetase family whereas CfdF is a putative α/β-hydrolase. A cfdD mutant was isolated, but the polar effect on the downstream genes prevents reliable conclusions to be drawn. CfdF belongs to the same α/β-hydrolase subfamily as HsaD that splits a metabolite, similar in structure to compound 4, during mycobacterial cholesterol catabolism (van der Geize et al. 2007).

During degradation of carbofuran by strain KN65.2, a red color formed, as well as for most of the mutants except for those of group V (8A8, 9G4; data not shown). Formation of a carbofuran metabolite with a nominal mass of 343.4 Da similar to compound X identified in this study and generating a red color in solution was previously reported during degradation of carbofuran by Sphingomonas sp. SB5 (Kim et al. 2004). Another red compound produced by this strain was identified as 5-(2-hydroxy-2-methyl-propyl)-2,2-dimethyl-2,3-dihydro-naphtho[2,3-6]furan-4,6,7,9-tetrone (C18H18O6) (Park et al. 2006). These authors suggested that this pigmented compound is a product of the condensation of several carbofuran degradation metabolites. The identity of this red compound in our study is currently further explored.

A number of observations suggest that, in addition to the route shown in Fig. 6, a second catabolic pathway may be operating in strain KN65.2. The amount of carbofuran phenol accumulating in nonmineralizing mutants showing complete conversion of carbofuran is much less than expected from a stoichiometric conversion (Fig. 3f and Online resource 6f). The low amount of colored metabolite(s) accumulating in the medium rules out that a major fraction of carbofuran would end up in these pigmented products (data not shown). A parallel pathway may proceed through side-chain-hydroxylated carbofuran (compound 6) that is not detected in the wild type and, hence, does not represent a dead-end metabolite. Moreover, mutant 9G4 does not degrade carbofuran phenol (data not shown). A good candidate for catalyzing this conversion is the mono-oxygenase CfdI, and the hydroxylation product may then be further metabolized by sequential activity of carbofuran hydrolase and CfdC as proposed in Online resource 10. This concurrent route could account for the formation of a side-chain-hydroxylated analog of metabolite 3, i.e., putative compound 7.

In conclusion, a combined genomic-metabolomic-proteomic approach revealed a number of steps and the corresponding catabolic genes in the biodegradation pathway of carbofuran by Novosphingobium sp. KN65.2, but the hydrolase gene required for initiating mineralization remains to be identified. Further analysis using nonpolar mutants affected in specific genes identified here and biochemical characterization of the respective encoded enzymes are required to further elucidate the tentative scheme currently envisaged. Of particular interest is how TonB-dependent uptake is involved in catabolism of this insecticide.

Notes

Acknowledgments

This research was funded by the Flemish Interuniversity Council (VLIR-UOS) of Belgium (BBTP2007-0012-1087), the joint support of the International Foundation for Science and Organisation for the Prohibition of Chemical Weapons (IFS/OPCW) (C/4563-1), and the EU FP7 projects BIOTREAT (EU grant 266039) and AQUAREHAB (EU grant ENV 2008.3.1.1.1.). We thank Kenneth Simoens for technical support.

Supplementary material

253_2014_5858_MOESM1_ESM.pdf (905 kb)
ESM 1(PDF 905 kb)

References

  1. Aylward FO, McDonald BR, Adams SM, Valenzuela A, Schmidt RA, Goodwin LA, Woyke T, Currie CR, Suen G, Poulsen M (2013) Comparison of 26 sphingomonad genomes reveals diverse environmental adaptations and biodegradative capabilities. Appl Environ Microbiol 79(12):3724–3733. doi:10.1128/Aem.00518-13 PubMedPubMedCentralCrossRefGoogle Scholar
  2. Bermingham A, Derrick JP (2002) The folic acid biosynthesis pathway in bacteria: evaluation of potential for antibacterial drug discovery. Bioessays 24:637–648. doi:10.1002/Bies.10114 PubMedCrossRefGoogle Scholar
  3. Caldas SS, Demoliner A, Costa FP, D'Oca MGM, Primel EG (2010) Pesticide residue determination in groundwater using solid-phase extraction and high-performance liquid chromatography with diode array detector and liquid chromatography-tandem mass spectrometry. J Braz Chem Soc 21(4):642–650. doi:10.1590/S0103-50532010000400009 CrossRefGoogle Scholar
  4. Castellanos J, Sánchez J, Uribe D, Moreno L, Melgarejo LM (2013) Characterization of carbofuran degrading bacteria obtained from potato cultivated soils with different pesticide application records. Rev Fac Nal Agr Medellín 66:6899–6908Google Scholar
  5. Chapalamadugu S, Chaudhry GR (1992) Microbiological and biotechnological aspects of metabolism of carbamates and organophosphates. Crit Rev Biotechnol 12(5–6):357–389. doi:10.3109/07388559209114232 PubMedCrossRefGoogle Scholar
  6. Chaudhry GR, Ali HD (1988) Bacterial metabolism of carbofuran. Appl Environ Microbiol 54(1414):1419Google Scholar
  7. Chaudhry GR, Mateen A, Kaskar B, Sardessai M, Bloda M, Bhatti AR, Walia SK (2002) Induction of carbofuran oxidation to 4-hydroxycarbofuran by Pseudomonas sp. 50432. FEMS Microbiol Lett 214(2):171–176. doi:10.1016/s0378-1097(02)00851-0 PubMedCrossRefGoogle Scholar
  8. Chen HP, Zhu SH, Casabon I, Hallam SJ, Crocker FH, Mohn WW, Indest KJ, Eltis LD (2012) Genomic and transcriptomic studies of an RDX (hexahydro-1,3,5-trinitro-1,3,5-triazine)-degrading actinobacterium. Appl Environ Microbiol 78(21):7798–7800. doi:10.1128/Aem.02120-12 PubMedPubMedCentralCrossRefGoogle Scholar
  9. Chiron S, Valverde A, FernandezAlba A, Barcelo D (1995) Automated sample preparation for monitoring groundwater pollution by carbamate insecticides and their transformation products. J AOAC Int 78(6):1346–1352PubMedGoogle Scholar
  10. Chowdhury MAZ, Banik S, Uddin B, Moniruzzaman M, Karim N, Gan SH (2012) Organophosphorus and carbamate pesticide residues detected in water samples collected from paddy and vegetable fields of the Savar and Dhamrai Upazilas in Bangladesh. Int J Environ Res Public Health 9(9):3318–3329. doi:10.3390/ijerph9093318 PubMedPubMedCentralCrossRefGoogle Scholar
  11. Cuthbertson L, Nodwell JR (2013) The TetR family of regulators. Microbiol Mol Biol Rev 77(3):440–475. doi:10.1128/Mmbr.00018-13 PubMedPubMedCentralCrossRefGoogle Scholar
  12. D'Argenio V, Petrillo M, Cantiello P, Naso B, Cozzuto L, Notomista E, Paolella G, Di Donato A, Salvatore F (2011) De novo sequencing and assembly of the whole genome of Novosphingobium sp. strain PP1Y. J Bacteriol 193(16):4296PubMedPubMedCentralCrossRefGoogle Scholar
  13. Dasgupta S, Meisner C, Wheeler D, Xuyen K, Lam NT (2007) Pesticide poisoning of farm workers - implications of blood test results from Vietnam. Int J Hyg Environ Health 210(2):121–132. doi:10.1016/j.ijheh.2006.08.006 PubMedCrossRefGoogle Scholar
  14. De Llasera MPG, Bernal-Gonzalez M (2001) Presence of carbamate pesticides in environmental waters from the northwest of Mexico: determination by liquid chromatography. Water Res 35(8):1933–1940CrossRefGoogle Scholar
  15. Desaint S, Hartmann A, Parekh NR, Fournier JC (2000) Genetic diversity of carbofuran-degrading soil bacteria. FEMS Microbiol Ecol 34(2):173–180. doi:10.1111/j.1574-6941.2000.tb00767.x PubMedCrossRefGoogle Scholar
  16. Desaint S, Arrault S, Siblot S, Fournier JC (2003) Genetic transfer of the mcd gene in soil. J Appl Microbiol 95(1):102–108. doi:10.1046/j.1365-2672.2003.01965.x PubMedCrossRefGoogle Scholar
  17. Dresen C, Lin LY, D'Angelo I, Tocheva EI, Strynadka N, Eltis LD (2010) A flavin-dependent monooxygenase from Mycobacterium tuberculosis involved in cholesterol catabolism. J Biol Chem 285(29):22264–22275. doi:10.1074/jbc.M109.099028 PubMedPubMedCentralCrossRefGoogle Scholar
  18. Fahmy MAH, Fukuto TR, Myers RO, March RB (1970) Selective toxicity of new N-phosphorothioylcarbamate esters. J Agric Food Chem 18(5):793–796. doi:10.1021/jf60171a014 PubMedCrossRefGoogle Scholar
  19. Feng XH, Ou LT, Ogram A (1997) Plasmid-mediated mineralization of carbofuran by Sphingomonas sp. strain CF06. Appl Environ Microbiol 63(4):1332–1337PubMedPubMedCentralGoogle Scholar
  20. Fida TT, Breugelmans P, Lavigne R, van der Meer JR, De Mot R, Vaysse PJ, Springael D (2014) Identification of opsA, a gene involved in solute stress mitigation and survival in soil, in the polycyclic aromatic hydrocarbon-degrading Novosphingobium sp. strain LH128. Appl Environ Microbiol 80(11):3350–3361. doi:10.1128/AEM.00306-14
  21. Gabriel FL, Cyris M, Jonkers N, Giger W, Guenther K, Kohler HP (2007) Elucidation of the ipso-substitution mechanism for side-chain cleavage of alpha-quaternary 4-nonylphenols and 4-t-butoxyphenol in Sphingobium xenophagum Bayram. Appl Environ Microbiol 73(10):3320–3326. doi:10.1128/AEM.02994-06 PubMedPubMedCentralCrossRefGoogle Scholar
  22. Gao JF, Ellis LBM, Wackett LP (2011) The university of Minnesota pathway prediction system: multi-level prediction and visualization. Nucleic Acids Res 39:W406–W411. doi:10.1093/Nar/Gkr200 PubMedPubMedCentralCrossRefGoogle Scholar
  23. Goad RT, Goad JT, Atieh BH, Gupta RC (2004) Carbofuran-induced endocrine disruption in adult male rats. Toxicol Mech Methods 14(4):233–239. doi:10.1080/15376520490434476 PubMedCrossRefGoogle Scholar
  24. Guindon S, Gascuel O (2003) A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst Biol 52(5):696–704. doi:10.1080/10635150390235520 PubMedCrossRefGoogle Scholar
  25. Gupta RC (1994) Carbofuran toxicity. J Toxicol Environ Health 43(4):383–418PubMedCrossRefGoogle Scholar
  26. Hamed RB, Batchelar ET, Clifton IJ, Schofield CJ (2008) Mechanisms and structures of crotonase superfamily enzymes—how nature controls enolate and oxyanion reactivity. Cell Mol Life Sci 65:2507–2527. doi:10.1007/s00018-008-8082-6 PubMedCrossRefGoogle Scholar
  27. Hartmann EM, Armengaud J (2014) Shotgun proteomics suggests involvement of additional enzymes in dioxin degradation by Sphingomonas wittichii RW1. Environ Microbiol 16:162–176. doi:10.1111/1462-2920.12264 PubMedCrossRefGoogle Scholar
  28. Hashimoto M, Fukui M, Hayano K, Hayatsu M (2002) Nucleotide sequence and genetic structure of a novel carbaryl hydrolase gene (cehA) from Rhizobium sp. strain AC100. Appl Environ Microbiol 68(3):1220–1227. doi:10.1128/Aem.68.3.1220-1227.2002 PubMedPubMedCentralCrossRefGoogle Scholar
  29. Hashimoto M, Mizutani A, Tago K, Ohnishi-Kameyama M, Shimojo T, Hayatsu M (2006) Cloning and nucleotide sequence of carbaryl hydrolase gene (cahA) from Arthrobacter sp. RC100. J Biosci Bioeng 101:410–414. doi:10.1263/Jbb.101.410 PubMedCrossRefGoogle Scholar
  30. Head IM, Cain RB, Suett DL (1992) Characterization of a carbofuran-degrading bacterium and investigation of the role of plasmids in catabolism of the insecticide carbofuran. Arch Microbiol 158(4):302–308. doi:10.1007/Bf00245249 PubMedCrossRefGoogle Scholar
  31. Heider J (2001) A new family of CoA-transferases. FEBS Lett 509(3):345–349. doi:10.1016/s0014-5793(01)03178-7 PubMedCrossRefGoogle Scholar
  32. Helbling DE, Hollender J, Kohler HPE, Singer H, Fenner K (2010) High-throughput identification of microbial transformation products of organic micropollutants. Environ Sci Technol 44:6621–6627. doi:10.1021/es100970m PubMedCrossRefGoogle Scholar
  33. Held JM, Schilling B, D'Souza AK, Srinivasan T, Behring JB, Sorensen DJ, Benz CC, Gibson BW (2013) Label-free quantitation and mapping of the ErbB2 tumor receptor by multiple protease digestion with data-dependent (MS1) and data-independent (MS2) acquisitions. Int J Proteomics 2013:791985. doi:10.1155/2013/791985 PubMedPubMedCentralCrossRefGoogle Scholar
  34. Kaczmarczyk A, Campagne S, Danza F, Metzger LC, Vorholt JA, Francez-Charlot A (2011) Role of Sphingomonas sp. strain Fr1 PhyR-NepR-sigmaEcfG cascade in general stress response and identification of a negative regulator of PhyR. J Bacteriol 193(23):6629–6638. doi:10.1128/JB.06006-11 PubMedPubMedCentralCrossRefGoogle Scholar
  35. Karns JS, Tomasek PH (1991) Carbofuran hydrolase—purification and properties. J Agric Food Chem 39:1004–1008. doi:10.1021/Jf00005a041 CrossRefGoogle Scholar
  36. Karns JS, Mulbry WW, Nelson JO, Kearney PC (1986) Metabolism of carbofuran by a pure bacterial culture. Pestic Biochem Physiol 25(2):211–217. doi:10.1016/0048-3575(86)90048-9 CrossRefGoogle Scholar
  37. Kaur B, Khera A, Sandhir R (2012) Attenuation of cellular antioxidant defense mechanisms in kidney of rats intoxicated with carbofuran. J Biochem Mol Toxicol 26(10):393–398. doi:10.1002/jbt.21433 PubMedCrossRefGoogle Scholar
  38. Kim IS, Ryu JY, Hur HG, Gu MB, Kim SD, Shim JH (2004) Sphingomonas sp. strain SB5 degrades carbofuran to a new metabolite by hydrolysis at the furanyl ring. J Agric Food Chem 52:2309–2314. doi:10.1021/jf035502l PubMedCrossRefGoogle Scholar
  39. Koebnik R (2005) TonB-dependent trans-envelope signalling: the exception or the rule? Trends Microbiol 13:343–347. doi:10.1016/j.tim.2005.06.005 PubMedCrossRefGoogle Scholar
  40. Larsen RA, Wilson MM, Guss AM, Metcalf WW (2002) Genetic analysis of pigment biosynthesis in Xanthobacter autotrophicus Py2 using a new, highly efficient transposon mutagenesis system that is functional in a wide variety of bacteria. Arch Microbiol 178(3):193–201. doi:10.1007/s00203-002-0442-2 PubMedCrossRefGoogle Scholar
  41. Luo YR, Kang SG, Kim SJ, Kim MR, Li N, Lee JH, Kwon KK (2012) Genome sequence of benzo(a)pyrene-degrading bacterium Novosphingobium pentaromativorans US6-1. J Bacteriol 194(4):907. doi:10.1128/JB.06476-11 PubMedPubMedCentralCrossRefGoogle Scholar
  42. Masai E, Sasaki M, Minakawa Y, Abe T, Sonoki T, Miyauchi K, Katayama Y, Fukuda M (2004) A novel tetrahydrofolate-dependent O-demethylase gene is essential for growth of Sphingomonas paucimobilis SYK-6 with syringate. J Bacteriol 186(9):2757–2765PubMedPubMedCentralCrossRefGoogle Scholar
  43. Milatovic D, Gupta RC, Dekundy A, Montine TJ, Dettbarn WD (2005) Carbofuran-induced oxidative stress in slow and fast skeletal muscles: prevention by memantine and atropine. Toxicol 208:13–24. doi:10.1016/j.tox.2004.11.004 CrossRefGoogle Scholar
  44. Naqvi T, Cheesman MJ, Williams MR, Campbell PM, Ahmed S, Russell RJ, Scott C, Oakeshott JG (2009) Heterologous expression of the methyl carbamate-degrading hydrolase Mcd. J Biotechnol 144(2):89–95. doi:10.1016/j.jbiotec.2009.09.009 PubMedCrossRefGoogle Scholar
  45. Noinaj N, Guillier M, Barnard TJ, Buchanan SK (2010) TonB-dependent transporters: regulation, structure, and function. Annu Rev Microbiol 64:43–60. doi:10.1146/annurev.micro.112408.134247 PubMedPubMedCentralCrossRefGoogle Scholar
  46. Ogram AV, Duan YP, Trabue SL, Feng X, Castro H, Ou LT (2000) Carbofuran degradation mediated by three related plasmid systems. FEMS Microbiol Ecol 32(3):197–203. doi:10.1111/j.1574-6941.2000.tb00712.x CrossRefGoogle Scholar
  47. Parekh NR, Hartmann A, Charnay MP, Fournier JC (1995) Diversity of carbofuran-degrading soil bacteria and detection of plasmid-encoded sequences homologous to the mcd gene. FEMS Microbiol Ecol 17(3):149–160. doi:10.1111/j.1574-6941.1995.tb00138.x CrossRefGoogle Scholar
  48. Park MR, Lee S, Han TH, Oh BT, Shim JH, Kim IS (2006) A new intermediate in the degradation of carbofuran by Sphingomonas sp. strain SB5. J Microbiol Biotechnol 16(8):1306–1310Google Scholar
  49. Plangklang P, Reungsang A (2012) Isolation and characterisation of a carbofuran degrading Burkholderia sp. PCL3 from carbofuran-phytoremediated rhizosphere soil. Chem Ecol 28:253–266. doi:10.1080/02757540.2011.645032 CrossRefGoogle Scholar
  50. Ramanand K, Sharmila M, Singh N, Sethunathan N (1991) Metabolism of carbamate insecticides by resting cells and cell-free preparations of a soil bacterium, Arthrobacter sp. Bull Environ Contam Toxicol 46(3):380–386PubMedCrossRefGoogle Scholar
  51. Ran T, Gao Y, Marsh M, Zhu W, Wang M, Mao X, Xu L, Xu D, Wang W (2013) Crystal structures of Cg1458 reveal a catalytic lid domain and a common catalytic mechanism for the FAH family. Biochem J 449(1):51–60. doi:10.1042/BJ20120913 PubMedCrossRefGoogle Scholar
  52. Ricken B, Corvini PFX, Cichocka D, Parisi M, Lenz M, Wyss D, Martinez-Lavanchy PM, Muller JA, Shahgaldian P, Tulli LG, Kohler HPE, Kolvenbach BA (2013) ipso-Hydroxylation and subsequent fragmentation: a novel microbial strategy to eliminate sulfonamide antibiotics. Appl Environ Microbiol 79(18):5550–5558. doi:10.1128/Aem.00911-13 PubMedPubMedCentralCrossRefGoogle Scholar
  53. Salama AK (1998) Metabolism of carbofuran by Aspergillus niger and Fusarium graminearum. J Environ Sci Health B 33:253–266. doi:10.1080/03601239809373142 PubMedCrossRefGoogle Scholar
  54. Sambrook J, Russell D (2001) Molecular cloning: a laboratory manual, 3rd edn. Cold Spring Harbor Laboratory, Cold Spring Harbor, New YorkGoogle Scholar
  55. Saxena A, Anand S, Dua A, Sangwan N, Khan F, Lal R (2013) Novosphingobium lindaniclasticum sp. nov., a hexachlorocyclohexane (HCH)-degrading bacterium isolated from an HCH dumpsite. Int J Syst Evol Microbiol 63:2160–2167. doi:10.1099/ijs.0.045443-0 PubMedCrossRefGoogle Scholar
  56. Seo J, Jeon J, Kim S-D, Kang S, Han J, Hur H-G (2007) Fungal biodegradation of carbofuran and carbofuran phenol by the fungus Mucor ramannianus: identification of metabolites. Water Sci Technol 55(1–2):163–167. doi:10.2166/wst.2007.051 PubMedCrossRefGoogle Scholar
  57. Shin DH, Kim DU, Seong CN, Song HG, Ka JO (2012) Genetic and phenotypic diversity of carbofuran-degrading bacteria isolated from agricultural soils. J Microbiol Biotechnol 22(4):448–456. doi:10.4014/jmb.1108.08087 PubMedCrossRefGoogle Scholar
  58. Simon R, Priefer U, Pühler A (1983) A broad host range mobilization system for in-vivo genetic engineering: transposon mutagenesis in Gram-negative bacteria. Nat Biotechnol 1:784–791. doi:10.1038/nbt1183-784 CrossRefGoogle Scholar
  59. Tariq MI, Afzal S, Hussain I (2006) Degradation and persistence of cotton pesticides in sandy loam soils from Punjab, Pakistan. Environ Res 100:184–196. doi:10.1016/j.envres.2005.05.002 PubMedCrossRefGoogle Scholar
  60. Tomasek PH, Karns JS (1989) Cloning of a carbofuran hydrolase gene from Achromobacter sp. strain WM111 and its expression in Gram-negative bacteria. J Bacteriol 171(7):4038–4044PubMedPubMedCentralGoogle Scholar
  61. Topp E, Hanson RS, Ringelberg DB, White DC, Wheatcroft R (1993) Isolation and characterization of an N-methylcarbamate insecticide-degrading methylotrophic bacterium. Appl Environ Microbiol 59(10):3339–3349PubMedPubMedCentralGoogle Scholar
  62. Untergasser A, Cutcutache I, Koressaar T, Ye J, Faircloth BC, Remm M, Rozen SG (2012) Primer3 - new capabilities and interfaces. Nucleic Acids Res 40(15):e115. doi:10.1093/nar/gks596 PubMedPubMedCentralCrossRefGoogle Scholar
  63. Uyttebroek M, Breugelmans P, Janssen M, Wattiau P, Joffe B, Karlson U, Ortega-Calvo JJ, Bastiaens L, Ryngaert A, Hausner M, Springael D (2006) Distribution of the Mycobacterium community and polycyclic aromatic hydrocarbons (PAHs) among different size fractions of a long-term PAH-contaminated soil. Environ Microbiol 8:836–847. doi:10.1111/j.1462-2920.2005.00970.x PubMedCrossRefGoogle Scholar
  64. van Berkel WJH, Kamerbeek NM, Fraaije MW (2006) Flavoprotein monooxygenases, a diverse class of oxidative biocatalysts. J Biotechnol 124(4):670–689. doi:10.1016/j.jbiotec.2006.03.044 PubMedCrossRefGoogle Scholar
  65. van der Geize R, Yam K, Heuser T, Wilbrink MH, Hara H, Anderton MC, Sim E, Dijkhuizen L, Davies JE, Mohn WW, Eltis LD (2007) A gene cluster encoding cholesterol catabolism in a soil actinomycete provides insight into Mycobacterium tuberculosis survival in macrophages. Proc Natl Acad Sci U S A 104:1947–1952. doi:10.1073/pnas.0605728104 PubMedPubMedCentralCrossRefGoogle Scholar
  66. Vryzas Z, Vassiliou G, Alexoudis C, Papadopoulou-Mourkidou E (2009) Spatial and temporal distribution of pesticide residues in surface waters in northeastern Greece. Water Res 43(1):1–10. doi:10.1016/j.watres.2008.09.021 PubMedCrossRefGoogle Scholar
  67. Yan QX, Hong Q, Han P, Dong XJ, Shen YJ, Li SP (2007) Isolation and characterization of a carbofuran-degrading strain Novosphingobium sp. FND-3. FEMS Microbiol Lett 271(2):207–213. doi:10.1111/j.1574-6968.2007.00718.x PubMedCrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2014

Authors and Affiliations

  • Thi Phi Oanh Nguyen
    • 1
    • 2
  • Damian E. Helbling
    • 3
  • Karolien Bers
    • 1
  • Tekle Tafese Fida
    • 1
  • Ruddy Wattiez
    • 4
  • Hans-Peter E. Kohler
    • 3
  • Dirk Springael
    • 1
  • René De Mot
    • 5
  1. 1.Division of Soil and Water Management, Department of Earth and Environmental SciencesKU LeuvenLeuvenBelgium
  2. 2.Department of Biology, College of Natural SciencesCan Tho UniversityCan ThoVietnam
  3. 3.Department of Environmental MicrobiologySwiss Federal Institute of Aquatic Science and Technology (EAWAG)DübendorfSwitzerland
  4. 4.Department of Proteomics and Microbiology, Research Institute for BiosciencesUniversity of MonsMonsBelgium
  5. 5.Centre of Microbial and Plant GeneticsKU LeuvenLeuvenBelgium
  6. 6.School of Civil and Environmental EngineeringCornell UniversityIthacaUSA
  7. 7.Department of Civil and Chemical EngineeringGeorgia Institute of TechnologyAtlantaUSA

Personalised recommendations