(S)–(-)-Nicotine (>99%) was obtained from Chemsky international Co., Ltd (Shanghai, China). 6 HN was a gift from Jiguo Qiu (Nanjing Agricultural University, China). TransStart® FastPfu DNA Polymerase for fragment amplification was purchased from TransGen Biotech (Beijing, China). Restriction enzymes used for plasmid construction and a premixed protein marker for protein electrophoresis were purchased from Takara Biotechnology Co., Ltd. (Dalian, China). Antibiotics, isopropyl β-D-1-thiogalactopyranoside (IPTG) and other reagents were purchased from Shanghai Sangon Biotech Co., Ltd. (Shanghai, China). A plasmid extraction kit, gel extraction kit and DNA purification kit were obtained from Omega Bio-tek, Inc (Norcross, GA, USA).

PCR program used for high-fidelity DNA polymerase was according to the manual: pre-denaturing at 95°C for 2 min; denaturing at 95°C for 20 s, annealing at certain temperature (according to the primer pair) for 20 s, extending at 72°C at a speed of 2 kb/min, 32 cycles; extending at 72°C for 5 min; keeping at 4°C. General PCR program used for fragment amplification was pre-denaturing at 94°C for 5 min; denaturing at 94°C for 30 s, annealing at certain temperature (according to the primer pair) for 30 s, extending at 72°C at a speed of 1 kb/min, 32 cycles; extending at 72°C for 5 min, keeping at 4°C. PCR program used for reverse transcription quantitative PCR (RT-qPCR) was: pre-denaturing at 94°C for 30 s; denaturing at 94°C for 5 s, annealing at 60°C for 15 s, extending at 72°C for 12 s, 40 cycles.

The bacterial strains and plasmids used in this study are listed in Table 1; the primers used are provided in Table S1. Wild-type S. melonis TY (collection number CGMCC 1.15791) can utilize nicotine as a sole source of carbon, nitrogen and energy for growth (Wang et al., 2011). Wild-type TY and derivatives were cultured aerobically at 30°C in LB medium or inorganic salt medium (ISM) supplemented with nicotine, as described previously (Yang, 2010). Escherichia coli strains were grown in LB broth at 37°C. When necessary, kanamycin and tetracycline were used at final concentrations of 50 and 10 μg/mL, respectively. IPTG was used to induce expression at gradient concentration, and 2, 6-diaminopimelic acid (2, 6-DAP) was used at a final concentration of 0.3 mM for E. coli WM3064.

The draft genome sequence of S. melonis TY was obtained using Illumina Hiseq2000 paired-end (PE) sequencing (101 bp for each read; 299.8-fold coverage) and then assembled into 68 scaffolds using SOAPdenovo version 2.04 (N50 length, 205,244 bp) (Li et al., 2010). The draft genome sequence of strain TY contains 4,100,783 bp with a GC content of 67.083% and 3748 predicted coding sequences (CDSs). The annotation was performed using Best-placed reference protein set; GeneMarkS+. Previously described nicotine-degrading genes, including ndhLSM, which catalyzes the first step of nicotine transformation in A. nicotinovorans (Dang Dai et al., 1968; Grether-Beck et al., 1994), and vppA and vppB of the upper VPP pathway of Ochrobactrum sp. SJY1 (Yu et al., 2014, 2015), were used for blast searches of the strain TY genome in NCBI.

RNAprep Pure Bacteria Kit (Tiangen Biotech, Beijing, China) was used to prepare total RNA from S. melonis TY grown in triplicate in control and nicotine-induction cultures. The RNA was reverse transcribed into cDNA using random hexamer primers and PrimeScript RT Reagent Kit with gDNA Eraser (Perfect Real Time) (Takara, Dalian, China). The respective cDNA fragments were used as templates in PCR with gene-specific primers (Table S1). Real-time quantitative PCR was performed using a Rotor-Gene Q real-time PCR detection system (Qiagen) with TransStart Top Green qPCR SuperMix (TransGen Biotech, Beijing, China). The strain TY genome was used as the positive control, and untranscribed RNA was used as the negative control. Melting curves and agarose gel analyses were applied to confirm the specificity of the PCR products. The threshold cycle (C_T) values for each target gene were normalized to the V3 region of the 16S rRNA gene. The 2^−ΔΔCT method was used to calculate the relative expression level, where ΔΔC_T = (C_{T, target}-C_{T, 16S})_induction—(C_{T, target}-C_{T, 16S})_control; to obtain more reliable results, the theoretical efficiency value 2 was replaced with the estimated PCR efficiency value (Livak and Schmittgen, 2001; Ramakers et al., 2003; Ruijter et al., 2009; Tuomi et al., 2010).

In-frame disruption of ndrA1 and ndrB2 in strain S. melonis TY was performed using the suicide plasmid pEX18Tc and a two-step homologous recombination method as described previously (Chen et al., 2014). Gene complementation were conducted the same as described in Chen et al. (2014).

Biotransformation tests were performed using wild-type TY and derivative strains TYΔndrA1, TYΔndrA1(pRK415- ndrA1), TYΔndrB2, and TYΔndrB2(pRK415-ndrB2) as well as an inactivated wild-type TY strain [heated at 80°C for 5 min to prevent nicotine absorption (Harwood et al., 1994)] and a control consisting only of nicotine. The biotransformation test was performed in a 150 mL flask containing 60 mg dry cell weight of resting cells (approximately 5 OD_600nm, with one OD_600nm unit equal to 0.40 g/L dry cell weight) and 0.5 g/L nicotine in 30 mL of sterilized ISM medium; the culture was incubated at 30°C and 200 rpm. Three sets of parallel experiments were performed for each strain. Biotransformation tests using 6 HN were conducted in a manner similar to nicotine that described above.

To test whether the two mutant strains retained the ability to metabolize 6 HN, the first product formed in the VPP pathway, strains TYΔndrA1, and TYΔndrB2 were washed three times with sterilized water and streaked onto ISM plates supplemented with 6 HN as the sole source carbon and nitrogen; the plates were placed in an incubator (30°C) for growth; wild-type TY was used as a positive control.

Nicotine catalysis activity of ndrA1A2A3 and ndrB1B2B3B4 was tested by heterologous expression. To test for nicotine dehydrogenase activity using whole cell or crude cell extracts, the two gene clusters were cloned (using primers shown in Table S1) into the broad host-range cloning vector pRK415 (Hind III- and EcoR I digested) and mated to P. putida KT2440 and Sphingomonas aquatilis (two non-nicotine-degrading strains) via E. coli WM3064. The culture conditions and enzyme activity detection are those previously reported (Ganas and Brandsch, 2009), while the detection of production 6 HN was replaced by UV-scan, and the host transformed with the control plasmid was used as the negative control, wild-type TY set as positive control.

Genes orf2 (coding for glycosyl transferase) and orf1 (coding for a XshC-Cox1-family protein) just downstream of ndrB4 exhibit similarity with mobA and coxF, respectively, which are essential for nicotine dehydrogenase activity in A. nicotinovorans pAO1 (Ganas and Brandsch, 2009) and may be required for nicotine dehydrogenase cofactor synthesis in strain TY. We cloned orf1 and orf2 together with ndrB1B2B3B4 into vector pRK415 (Hind III- and EcoR I digested) according to a previous report (Ganas and Brandsch, 2009), followed by transformation of the plasmid into E. coli DH5α and mating to S. aquatilis and P. putida KT2440 via E. coli WM3064. Moreover, we also cloned these six genes into vector pET-28a(+) (Hind III- and Nco I-digested) just downstream the T7 promoter and transformed the plasmid into E. coli BL21(DE3) to induce expression with various concentrations of IPTG (at 0, 0.1, 0.2, 0.5 and 1 mM). Activity tests were performed using whole cell and crude cell extracts of the above recombinant strains and corresponding negative controls with the host containing the original plasmid, pRK415 or pET-28a(+). SDS-PAGE was carried out for the E. coli BL21(DE3) derivative strains to detect induction under the test conditions.

This whole-genome shotgun project has been deposited at DDBJ/ENA/GenBank under accession LQCK00000000. The version described in this paper is version LQCK02000000.

Genome sequencing can provide much information for research into the functional mechanisms of microorganisms (Tang et al., 2013). The draft genome sequence of S. melonis TY was obtained by Illumina sequencing, and blast searching of the sequence using representative sequences for nicotine-degrading enzymes from A. nicotinovorans and Ochrobactrum sp. SJY1 was performed. Based on gene similarity and organization, we identified two sets of putative nicotine dehydrogenase genes in TY: ndrA1A2A3 and ndrB1B2B3B4. According to the annotation of genome sequence, ndrA1 encodes an aldehyde dehydrogenase iron-sulfur subunit; ndrA2 encodes a subunit of molybdopterin dehydrogenase; ndrA3 encodes a subunit of xanthine dehydrogenase; ndrB1 encodes an aldehyde dehydrogenase iron-sulfur subunit; ndrB2, ndrB3, and ndrB4 all encode subunit of xanthine dehydrogenase. These two sets of gene clusters, ndrA1A2A3 and ndrB1B2B3B4, show similarity to the three subunits of ndh: nicotine dehydrogenase subunit S (ndhS), nicotine dehydrogenase subunit M (ndhM) and nicotine dehydrogenase subunit L (ndhL). ndrA3, ndrB3, and ndrB4 show similarity to nicotine hydroxylase subunit L (vppA_L), and ndrA1 and ndrB1 show similarity to nicotine hydroxylase subunit S (vppA_S). All these genes exhibit 15–44% amino acid sequence identity with homologous nicotine-degrading proteins, with similar gene organization. The gene organization and amino acid identity shared with homologous genes ndh and vppA are shown in Figure 1.

To demonstrate a correlation between nicotine degradation and the two sets of putative nicotine dehydrogenase genes, the levels of mRNA expression of seven putative target genes involved in nicotine degradation in S. melonis TY were estimated using RT-qPCR and the 2^ΔΔCT method in the presence or absence of nicotine. The V3 region of the 16S rRNA gene was used as the reference. The results showed that transcription of these genes was significantly upregulated in the presence of nicotine compared to in its absence (Figure 2), suggesting that transcription of ndrA1A2A3 and ndrB1B2B3B4 was induced by nicotine.

To examine whether ndrA1A2A3 or ndrB1B2B3B4 is involved in nicotine degradation in vivo, ndrA1 or ndrB2 was knocked out via kanamycin resistance gene insertion using a gene replacement technique based on homologous recombination. Double crossover was verified by PCR (Figure 3) and sequencing analysis. Although neither mutant strain, TYΔndrA1, or TYΔndrB2, could grow on ISM containing nicotine as the sole carbon and nitrogen source (Figure 4), both mutants were able to grow on ISM supplemented with 1 g/L (NH₄)₂SO₄ and 1 g/L glucose as the nitrogen and carbon sources, respectively. Furthermore, following complementation of the corresponding gene, each mutant displayed restored capacity to grow on nicotine, similar to wild-type TY (Figure 4).

Based on the results of gene knockout and complementation, it appears that ndrA1 and ndrB2 are involved in the capability of strain TY to utilize nicotine as a sole carbon and nitrogen source, suggesting that one of these two sets of gene clusters encodes a nicotine dehydrogenase. To further confirm the function of ndrA1A2A3 and ndrB1B2B3B4, we performed a nicotine and 6 HN biotransformation test using the wild-type TY strain, the mutant strains and strains complemented with the two genes. Although TYΔndrA1 and TYΔndrB2 lost the ability to degrade and transform nicotine, they retained 6 HN conversion capacity (Figure 4). Complementation of the corresponding gene restored the ability of both TYΔndrA1(pRK415-ndrA1) and TYΔndrB2(pRK415-ndrB2) to degrade and transform nicotine.

Because TYΔndrA1 and TYΔndrB2 can grow on ISM medium supplemented with 6 HN as the sole carbon and nitrogen source (Figure 5), combined with the biotransformation results, it appears that the two mutant strains not only retain the ability to transform 6 HN but can also degrade it completely. This result indicates that disruption of ndrA1 or ndrB2 only affects the first step in nicotine metabolism, supporting our speculation that a nicotine dehydrogenase is encoded by one of these two sets of genes.

To shed light on the function of these two gene clusters, we tested the nicotine-degrading activity of ndrA1A2A3 and ndrB1B2B3B4 through heterologous expression in P. putida KT2440 and S. aquatilis; the latter was selected because it is a member of the same genus as strain TY but does not exhibit nicotine degradation ability. However, neither whole cell nor crude cell extracts exhibited nicotine dehydrogenase activity.

The cofactor molybdopterin cytosine dinucleotide (MCD) is needed for Ndh nicotine dehydrogenase function in A. nicotinovorans pAO1 (Sachelaru et al., 2006; Ganas and Brandsch, 2009), we thought MCD or other cofactor is needed by the putative nicotine dehydrogenase in strain TY. Thus, we identified two genes, orf1 and orf2, contiguous on the genome with ndrB1B2B3B4 that show similarity with coxF and mobA (involved in MCD biosynthesis in A. nicotinovorans pAO1), and orf2 shows LAAG amino acid motif as the canonical MocA (Neumann et al., 2011). We cloned these six genes into pRK415 for expression in E. coli DH5α, S. aquatilis and P. putida KT2440 and into pET-28a(+) for expression in E. coli BL21(DE3). However, whole cell and crude cell extracts showed no nicotine dehydrogenase activity. In addition, SDS-PAGE was carried out on extracts from E. coli BL21(DE3)pET28a-ndrB1B2B3B4orf1orf2 and its control strain E. coli BL21(DE3)pET28a, and several bands were appeared in the former compared to the latter (data not shown).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.