Time-Dependent Analysis of Paenarthrobacter nicotinovorans pAO1 Nicotine-Related Proteome.

Paenarthrobacter nicotinovorans is a soil Gram-positive nicotine-degrading microorganism (NDM) that harbors a 165 kb pAO1 catabolic megaplasmid. The nicotine catabolic genes on pAO1 have been sequenced, but not all the details on the regulation and interplay of this pathway with the general metabolism of the cell are available. To address this issue at the protein level, a time-based shotgun proteomics study was performed. P. nicotinovorans was grown in the presence or absence of nicotine, and the cells were harvested at three different time intervals: 7, 10, and 24 h after inoculation. The cells were lysed, separated on SDS-PAGE, and digested by in-gel digestion using trypsin, and the resulting peptide mixture was analyzed using nanoliquid chromatography tandem mass spectrometry. We found an extensive number of proteins that are both plasmidal- and chromosomal-encoded and that work together in the energetic metabolism via the Krebs cycle and nicotine pathway. These data provide insight into the adaptation of the bacterial cells to the nicotine metabolic intermediates and could serve as a basis for future attempts to genetically engineer the pAO1-encoded catabolic pathway for increased bioremediation efficiency or for the production of valuable chemicals. The mass-spectrometry-based proteomics data have been deposited to the PRIDE partner repository with the data set identifier PXD012577.

Introduction

The tobacco industry is producing about 300 000 tons of nicotine-containing wastes every year,[1] which are classified as toxic and hazardous.[2,3] These nicotine-containing wastes are slowly transformed from an environmental problem into a renewable resource by the identification and characterization of an ever increasing number of nicotine-degrading microorganisms (NDMs) that can break down nicotine and use it for their growth.[4] NDMs have already been successfully used to decontaminate the tobacco wastes[4] and also to convert nicotine and nicotine-containing wastes into compounds with industrial and pharmaceutical relevance, such as 6-hydroxy-nicotine,[5] 6-hydroxy-3-succinoyl-pyridine,[6,7] and 3-succinoyl-pyridine,[8] indicating the feasibility of using nicotine from tobacco wastes as a biomass resource.

Still, the microbial biotransformation of nicotine is hindered by the lack of knowledge on the biology and biochemistry of NDMs, especially on the regulation of the nicotine-degradation pathways and their integration into the general metabolism of the cell. Advancements in this direction would allow the engineering of the nicotine-degrading pathways and enzymes for an improved yield and a wider array of useful intermediates.

Here, we focus on Paenarthrobacter nicotinovorans pAO1+, a Gram-positive NDM that combines the nicotine catabolic pathway with the remarkable survival abilities of Arthrobacter species.[9] In P. nicotinovorans, the nicotine is degraded using the pyridine pathway. Briefly, nicotine is first hydroxylated at the pyridine ring and the activated molecule is converted to γ-N-methylaminobutyrate (CH3-4-GABA) and 2,6-dihydroxy-pyridine (2,6-DHP). CH3-4-GABA is degraded to succinate and methylamine, with the latter compound accumulating in the growth medium, while 2,6-DHP is hydroxylated to trihydroxypyridine (THP). The THP spontaneously dimerizes to 4,4′,5,5′-tetrahydroxy-3,3′-diazadiphenoquinone-(2,2′) (nicotine blue, NB), giving a characteristic blue color of the growth medium.[10]

The nicotine-degrading abilities of P. nicotinovorans have been linked to the presence of the pAO1 catabolic megaplasmid.[11] The genes related to nicotine metabolism form the so-called pAO1 nic-gene cluster—a 64 kb piece of DNA consisting of about 40 genes organized in four operons. The general organization of the nic-gene cluster is depicted in Figure . The molecular biology,[12] biochemistry,[13] and evolution[14] of nicotine degradation in P. nicotinovorans have been extensively studied, but only 19 nic-gene products have been fully characterized. The remaining ones have been only partially linked to nicotine metabolism and have putative or no known functions, and some might not be expressed at all. These genes could include the yet missing key gene regulators of nicotine metabolism or the nicotine transporters responsible for nicotine import. Also, little is known on how the cells cope with the accumulation of the resulting nicotine metabolism byproducts.

Figure 1

The nic-gene cluster on the pAO1 plasmid of P. nicotinovorans. Bold letters indicate genes with known function, italic text indicates genes with putative functions related to nicotine or with no known functions, and blue text indicates genes that have been shown to have a nicotine-dependent expression by the proteomics experiment. The black text indicates genes that were not detected, as expressed in our proteomics experiment.

In our previous study, we used a shotgun proteomics approach based on nanoliquid chromatography tandem mass spectrometry (nano-LC–MS/MS) to investigate the physiology and nicotine catabolism of this bacterium on three different growth media. We showed how the nicotine degradation pathway can be switched to produce different end products based on the availability of C sources.[15] The data presented here further our work using the same experimental methodology but performing a time-based analysis of the nicotine-related proteome, thus expanding our current understanding of nicotine-induced biochemistry and physiology of P. nicotinovorans as well as on the role of plasmidal- vs chromosomal-encoded proteins in the nicotine catabolic pathway.

Results and Discussion

For proteomics analysis, three key points were sampled that correspond to major growth phases of the P. nicotinovorans pAO1 culture. The position of the sampling points related to bacteria growth phases, nicotine levels, and NB accumulation can be observed in Figure A.

Figure 2

Key point location of the samples on a typical growth curve of P. nicotinovorans (A) and the distribution of the identified proteins in the samples (B), where red indicates samples taken 7 h after inoculation, blue indicates samples taken 10 h after inoculation, and green indicates samples taken 24 h after inoculation; HAI, hours after inoculation. Nicotine-induced proteins are proteins detected when bacteria were cultivated on citrate media supplemented with nicotine (+) but are not detected when cultivated on citrate alone (−).

As in the lag phase of the culture, the number of cells is fewer to permit extensive analysis, and the first samples were harvested 7 h after inoculation (HAI). This corresponds to the middle log phase; when the bacterial culture is growing fast, NB starts to accumulate but no notable nicotine consumption can be detected. This indicates that the nicotine metabolism is being activated; hence, the enzymes involved accumulate. The second timepoint was in the early stationary phase corresponding to 10 HAI. The culture reached the maximum cell density when nicotine is being consumed, resulting in the maximum accumulation of NB. Hence, nicotine catabolism is fully activated, and all relevant enzymes are abundant. The last timepoint was in the late stationary phase at 24 HAI. At this stage, the entire nicotine is depleted from the medium and the cells should switch to using another C source: citrate or nicotine metabolism byproducts. At this point, NB is replaced by a dark brown pigment of unknown nature.

The approach allowed us to identify a total of 915 proteins grouped in 528 nonredundant protein clusters. As the search was performed against the genomes of two closely related strains, one might expect some degree of overlap between the two theoretical proteomes. Thus, each cluster was manually evaluated, and identical proteins were filtered out, the Paenarthrobacter aureus reference proteins being always preferred. The full list of 528 nonredundant proteins identified in this study is presented in Table S1, and their summary distribution based on growth conditions and HAI is depicted in Venn diagrams in Figure B.

pAO1 Nicotine-Related Proteins

The generated proteomics data are in good agreement with a variety of previously reported experimental studies on the pAO1 proteins involved in nicotine metabolism. Out of the 40 genes making up the nic-gene cluster, 27 genes have been experimentally shown to be involved in nicotine metabolism using various methods, including reverse transcription PCR, Western blotting, or functional assays (for an overview, see Table S2 and refs,[10,12] and.[13] The remaining genes are putative genes with no known function, and their association with nicotine metabolism is based solely on the genetic context. The current shotgun proteomics data confirm the nicotine-dependent expression for 19 of the experimentally characterized proteins (70% coverage). Most of these proteins are already highly expressed at 7 HAI, and their abundance decreases as nicotine levels drop (Figure ).

Figure 3

Heat map representation (weighted spectrum counts) of the pAO1 megaplasmid encoded proteins detected when the bacteria was grown on citrate media supplemented (+) or not (−) with nicotine. All proteins except the single-stranded DNA-binding proteins are related to nicotine metabolism. The numbers are fold change relative to the expression levels in the 7 HAI sample (referenced R) and the bold numbers are statistically significant fold changes (Fisher’s exact test, Benjamini–Hochberg multiple test correction, and significance level p < 0.001). INF indicates that the protein was not detected in the reference sample. The heat map was generated with Scaffold perSPECtives, linkage method—single, distance metric—rank-based Euclidean.

The key enzymes known to be involved in the first steps of the nicotine degradation pathway, namely, nicotine dehydrogenase (NDH), 6-hydroxy-d and l-nicotine oxidases, and 6-hydroxypseudooxynicotine dehydrogenase (KDHM), have a rather low abundance in the log and early stationary phase and are missing from the late stationary phase. The key enzymes processing the resulting intermediates and performing the last steps of the pathway are significantly more abundant in the late stationary phase. Two of these enzymes, the 2-oxoglutaramate amidase (NIT) and the hypothetical polyketide cyclase (PKC), are believed to be involved in the cleavage of the THP ring,[16] while the NAD(P)H-nicotine blue oxidoreductase (NBOR) is believed to be involved in maintaining the NB pigment in an oxidation stage favorable for cleavage.[15,17] The abundance of these enzymes in the late stationary phase when nicotine could not be detected in the medium, and, thus should not further activate the transcription of the nic-genes, could be related to the decrease in the NB pigment levels observed in Figure A. Still, we are hesitant to conclude that this is a clear indication that an NB degradation takes place and that these are the enzymes involved. It is known that the closely related Corynebacterium glutamicum strains[18] are characterized by a low protein degradation rate when a nutrient supply is available and this might be the case for P. nicotinovorans cells too.

The major missing nic-related proteins are all transcriptional regulators and membrane transporters, most probably due to their low abundance and solubility, respectively. Moreover, the proteomics data provide additional experimental evidence for the nicotine-dependent expression of six genes belonging to the nic-cluster that have only putative functions: folD—a putative methylene-tetrahydrofolate dehydrogenase cyclohydrolase homolog, purU—a putative formyltetrahydrofolate deformylase, the above-mentioned pkc—putative polyketide cyclase, coxD and coxG—subunits of a putative carbon monoxide dehydrogenase subunit, and modC—molybdenum transport ATPase.

pAO1 Core Functions Proteins

The nic-genes represent only 25% of the 165 open reading frames on pAO1. The rest of the genes encode an oxidative xylose degradation pathway[19] as well as core plasmid functions such as replication, partition, and conjugation.[11,20] Only three proteins outside of the nic-genes cluster were detected in our proteomics approach. Detected in all growth phases and disregarding the nutrients available is SSB_PAENI—a putative single-stranded DNA-binding protein. The P. nicotinovorans SSB SSB_PAENI protein is 65.3% identical at the amino acid level with the SSB protein from Mycolicibacterium smegmatis, which is known to interact with RecAn[21] and to play an important role in DNA replication, repair, and recombination.

The two other proteins are Q8GAD0_PAENI, a putative parA protein related to plasmid partition, and Q8GAD1_PAENI, yet uncharacterized protein. Both Q8GAD0_PAENI and Q8GAD1_PAENI are widely spread within the members of the Arthrobacter genus, but their role remains elusive.

Chromosome-Encoded Proteins

The heat map representation of weighted spectrum counts for the chromosomal proteins detected in the proteomics experiment is depicted in Figure . Despite that a large number (21 proteins, 60%) of the proteins with a significant fold change have unknown GOs, the few proteins with known functions offer a glimpse on the impact of nicotine metabolism on gene expression, translation, and general metabolic processes of the cells.

Figure 4

Heat map representation (weighted spectrum counts) of all chromosomal proteins detected when the bacteria were grown on citrate media supplemented (+) or not (−) with nicotine and GO distribution of the proteins with a significant fold change (Fisher’s exact test, Benjamini–Hochberg multiple test correction, and significance level p < 0.001). The heat map was generated with Scaffold perSPECtives, linkage method—single, distance metric—rank-based Euclidean. GO annotations were performed with the Scaffold using NCBI as an annotation source.

Protein and mRNA Synthesis

The 50S ribosomal protein L29 (A0A175J0V0_PAENI) registers one of the highest fold changes among all the proteins encoded by the chromosomes. Disregarding the growth phase, the L29 ribosomal protein is always detected in higher amounts in the presence of nicotine, accumulating in levels up to 3.2 times higher (Fisher’s exact test p value of 0.00028 and p < 0.00115) compared to when the cells are grown in the absence of nicotine. L29 ribosomal protein is one of the proteins actively involved in translation, surrounding the nascent polypeptide as it exits the tunnel outside the 50S ribosomal subunit.[22] Moreover, the 30S ribosomal protein S7 (RS7_PAEAT), despite registering only slightly higher levels in the middle log and early stationary phases (fold change nicotine vs no nicotine 1.2 at both 7 and 10 HAI), also accumulates in significantly higher amounts in the late stationary phase (fold change of 2.5, Fisher’s exact test p value of 0.00014, and p < 0.0012). The S7 ribosomal protein is known to be involved in contacting the tRNA during the peptide chain synthesis[23] as well as functioning as a translational repressor for the str operon, regulating the expression of various operon members to different degrees by binding to mRNA.[24] The elongation factor Tu (EF-Tu, A0A175J0V4_PAENI) is another dysregulated protein related to protein synthesis. EF-Tu is responsible for the binding of an aminoacyl-tRNA to the ribosome and is downregulated in the middle log phase (fold change of 0.6 at 7 HAI, Fisher’s exact test p value of 0.0001, and p < 0.00137) and is strongly upregulated in the late stationary phase (fold change of 1.7 at 24 HAI, Fisher’s exact test p value of 0.00013, and p < 0.0012) (Figure A—ribosomal proteins and protein synthesis).

Figure 5

Chromosome-encoded proteins with statistically significant fold change induced by the presence of nicotine in the growth medium of P. nicotinovorans pAO1. The fold change was calculated using weighted spectra and the corresponding no nicotine sample as the reference; SPB, substrate-binding protein.

Two more proteins that were found to be strongly dysregulated when the nicotine catabolic pathway is active are related to RNA metabolism: the DNA-directed RNA polymerase subunit beta (RPOB_PAEAT) responsible for mRNA synthesis and a polyribonucleotide nucleotidyltransferase (A0A175J953_PAENI) responsible for the phosphorolysis of single-stranded polyribonucleotides and hence mRNA degradation. Both enzymes have similar levels at 7 HAI in all samples and are strongly upregulated in the early stationary phase (fold change of 2.4 at 10 HAI, Fisher’s exact test p value of 0.00015, and p < 0.00115 and fold change of 2.6 at 10 HAI, Fisher’s exact test p value of 0.00043, and p < 0.00115). This would indicate a very active mRNA metabolism, which correlates well with an active protein synthesis.

Integration of Nicotine Degradation Pathway into the General Metabolism

The time-based analysis of the most significant dysregulated proteins when the cells are grown on nicotine allows us to postulate how the interplay between nicotine catabolism and general metabolism of the cell takes place.

Our experimental setup uses citrate as the main carbon source available for growth. Under aerobic conditions, citrate can be directly fed into the tricarboxylic acid cycle, only a citrate transporter being required in addition to the basic enzymes available in the cell.[25] When P. nicotinovorans pAO1 is grown on media containing nicotine and citrate, both the citrate metabolism and the nicotine degradation pathway are active. A strong nicotine-induced downregulation could be observed for a C4-dicarboxylate ABC transporter substrate-binding protein (A0A175J7D5_PAENI) presumably involved in citrate/succinate uptake, as well as for a succinate-CoA ligase (A0A175JA53_PAENI), a known enzyme from the citric acid cycle. This would indicate that when nicotine is catabolized, its end products are fed to the citric cycle; hence, supplementary citrate import is not of major importance.

The known end products of the nicotine degradation in P. nicotinovorans pAO1 are methylamine, succinic acid, and NB. Both NB and methylamine are excreted into the medium.[26,27] Nevertheless, several reports indicate that NB is actually reimported into the cell, reduced by a NAD(P)H-NBOR,[17] and slowly converted into alpha-ketoglutarate.[15,16] Hence, both the end products succinic acid and alpha-ketoglutarate should be integrated into the general pathways of the cells and used for growth.

Our data further support these findings. In the middle log phase, the main byproduct is succinate, which is easily integrated into the citrate cycle, while the NB is excreted and accumulates in the medium. As nicotine is depleted from the medium and the cells reach the early stationary phase, the succinate is no longer available, and the cells start reimporting the NB and converting it to alpha-ketoglutarate. Consequently, a 5.6-fold change is registered at 10 HAI for an alpha-ketoglutarate decarboxylase (A0A175J1U6_PAENI, Fisher’s exact test p value of 0.00038, p < 0.00115 and fold change of 1.1 at 7 HAI, Fisher’s exact test p value of 0.42, p < 0.00137). In the late stationary phase, when the NB is consumed, the expression of the alpha-ketoglutarate decarboxylase drops again, reaching levels that are not statistically different from those registered in the absence of nicotine (fold change of 1.6 at 24 HAI, Fisher’s exact test p value of 0.18, and p < 0.0012). The same expression pattern could be detected also for a pyruvate carboxylase (A0A175J9R5_PAENI), an enzyme involved in gluconeogenesis.

An increase in the levels of alpha-ketoglutarate decarboxylase after 10 h (Figure C—general metabolism, high on nicotine) could indicate that the Krebs cycle is depleted by alpha-ketoglutarate. However, the integrity of the Krebs cycle is restored by the increase in the levels of pyruvate carboxylase, an anaplerotic reaction that produces oxaloacetate, suggesting that the Krebs cycle is partially restored, and even the production of succinate, fumarate, and malate is impaired. Furthermore, full restoration of the Krebs cycle is completed by an increased production of succinate, which is replenished by succinate-semialdehyde, the product of conversion of alpha-ketoglutarate by alpha-ketoglutarate decarboxylase.

It is worth noting that once nicotine is depleted, the nicotine-induced alteration of the Krebs cycle mentioned earlier is reverted to its normal functioning, but increased expression of succinyl-CoA ligase, thus diverting alpha-ketoglutarate from alpha-ketoglutarate decarboxylase into the Krebs cycle through succinyl-CoA ligase. However, both the source and the fate of the succinate-semialdehyde produced from the decarboxylation of alpha-ketoglutarate decarboxylase are yet to be investigated. While succinate-semialdehyde can, indeed, be produced by alpha-ketoglutarate decarboxylase, it can then be depleted by succinate-semialdehyde dehydrogenase through conversion to succinate. If true, then the Krebs cycle is only partially altered (Figure ). Indeed, this is a logical interpretation. It has also been reported that in bacteria, succinate-semialdehyde dehydrogenase can also convert succinate-semialdehyde to succinate during the fission of the pyridine ring,[28] and, moreover, the succinate-semialdehyde dehydrogenase part of the nicotine catabolic pathway[12] is expressed in our data set. Therefore, it is reasonable to postulate that nicotine catabolism replenishes the Krebs cycle via anaplerotic reactions, not only through pyruvate carboxylase but also through alpha-ketoglutarate decarboxylase and succinate-semialdehyde dehydrogenase. Therefore, our proteomics-based time-course experiments of the nicotine catabolic pathway directly link the involvement of alpha-ketoglutarate decarboxylase and pyruvate carboxylase, which have an end product, succinate-semialdehyde, which is later used by succinate-semialdehyde dehydrogenase in the nicotine pathway to both deplete succinate-semialdehyde and replenish the Krebs cycle with succinate. Once the nicotine (and succinate-semialdehyde) supply is depleted (24 h, see the graph), both pyruvate carboxylase and alpha-ketoglutarate decarboxylase are downregulated and the succinyl-CoA ligase takes over to restore the Krebs cycle.

Figure 6

Upregulated (blue) and downregulated (red) proteins involved in nicotine catabolism and integration of the end products into the Krebs cycle. NDH, nicotine dehydrogenase; 6HLNO, 6-hydroxy-l-nicotine oxidase; KDH, ketone dehydrogenase; PONH, 2,6-dihydroxypseudooxynicotine hydrolase; DHPH, 2,6-dihydroxypyridine-3-hydroxylase; NBOR, nicotine blue oxidoreductase; MABO, γ-N-methylaminobutyrate oxidase; FolD, methylene-tetrahydrofolate dehydrogenase/cyclohydrolase; PurU, formyl-tetrahydrofolate deformylase; MAO, monoamine-oxidase; SAD, succinic semialdehyde dehydrogenase; PKC, putative polyketide cyclase; NIT, ω-amidase. Enzymes with an asterisk (*) have putative functions. The full arrows indicate a single catalytic reaction; the dashed arrows indicate multiple enzymatic steps.

Another highly dysregulated enzyme is catalase (A0A175J0M2_PAENI). The enzyme is upregulated when nicotine is present, and it is kept this way through the early stationary phase (fold change of 2.5 at 7 HAI, Fisher’s exact test p value of 0.0001, and p < 0.00137 and fold change of 3.5 at 10 HAI, Fisher’s exact test p value of 0.00023, and p < 0.00137). In the late stationary phase, when nicotine is depleted and the NB levels are low, the catalase expression levels are reduced and not statistically different from those registered in the absence of nicotine (fold change of 2.3 at 24 HAI, Fisher’s exact test p value of 0.33, and p < 0.0012). This pattern correlates very well with the demonstrated oxidative stress generated by the production of NB during the nicotine catabolism, with the catalase being an efficient protection mechanism for the cells.

Conclusions

The differences in protein expression patterns reported here are a follow-up of our previous study and allow not only to expand the knowledge about nicotine metabolism but more importantly to relate the observed differences in protein abundance to the accumulation of known nicotine intermediates and metabolites. Although we have previously reported the link between the Krebs cycle and nicotine catabolism in this bacterium, the five new chromosome-encoded enzymes related to the general metabolism of the cell reported here indicate that this is done through anaplerotic pathways, allowing us to understand the bacterial management of energy through the use of the Krebs cycle, nicotine pathway, or both. Our time-course proteomics experiments also allow us to determine when the Krebs cycle is active, when the nicotine pathway becomes active, and when both of them work together for an efficient energetic metabolism via the expression of various proteins through chromosomal–plasmidal gene regulation. Overall, these experiments can also lead to a better understanding of the pAO1-encoded catabolic pathway of P. nicotinovorans and the energy supply-based regulated expression of the plasmidal and chromosomal genes.

Materials and Methods

As the data reported here are a follow-up of our previous proteomics study dealing with the regulation of nicotine metabolism base on the available C sources in the growth medium,[15] we used the same experimental methodology and data analysis methods to analyze the samples taken at different time intervals from the same citrate-based growth medium. Thus, although the methods described below are expanded versions of descriptions from our previous work, they were used to generate the current proteomics data dealing with new samples and distinct proteomics assays.

Chemicals, Bacterial Strains, and Growing Conditions

All chemicals were from Sigma-Aldrich (St. Louis, MO, USA) unless stated otherwise. DNase I and RNase A were from Roche (Basel, Switzerland). HPLC grade water and acetonitrile were from Fisher Chemical (Pittsburgh, PA, USA). LC–MS grade formic acid (FA) was from Fluka (Buchs, Switzerland) and iodoacetamide was from Calbiochem (San Diego, CA, USA). The P. nicotinovorans pAO1 strain was a kind gift from Professor Dr. Roderich Brandsch and is deposited in DSMZ (Deutsche Sammlung von Mikroorganismen und Zellkulturen, Braunschweig, Germany) with ID DSM-420 as well as American Type Culture Collection (ATCC) with ID ATCC 49919. The strain was grown on citrate medium consisting of 0.2% Na-citrate, 34 mM Na2HPO4, 22 mM KH2PO4, 0.2% (NH4)2SO4, pH 7.0, 5% mineral solution, 0.1 mg mL–1 biotin, and 35 μg mL–1 kanamycin supplemented or not with 0.05% nicotine. The mineral solution[29] consisted of 6.5 mM CaCl2, 20 mM ZnSO4, 75 mM H3BO3, 0.8 mM FeSO4, 0.8 mM MnSO4, 0.4 mM CuSO4, 0.4 mM CoSO4, 15 mM KH2PO4, 8 mM g MgSO4, and 25 mM EDTA sterilized by filtration. Saturated precultures were prepared by growing the bacteria on citrate medium in the stationary phase for 24 h and then were used to inoculate the main 100 mL cultures at a 1:100 dilution. The cultures were incubated on a rotary shaker (Model 3013, GFL, Burgwedel, Germany) at 28 °C and 180 rpm, and the samples were collected at 7, 10, and 24 h post inoculation. The samples were taken every hour and the bacterial growth was followed spectrophotometrically at 660 nm and the accumulation of NB was determined at 585 nm, while nicotine levels were measured by HPLC, as described before.[30]

Sample Preparation

Whole P. nicotinovorans pAO1 cultures were centrifuged at 4500g for 20 min, and the bacterial pellets were washed twice in 10 mM Tris/HCl at pH 7.4 for the removal of the NB and other nicotine-related metabolites. The cells were lysed according to the protocol of Vandera et al.,[31] with modifications as indicated by Mihăşan et al.[15] Briefly, the cells were treated with 50 mg mL–1 lysozyme for 60 min at 37 °C and then lysed by boiling at 95 °C and vigorous shaking in the presence of 0.3% SDS. Unbroken cells and cellular debris were removed by centrifugation, and the cell-free lysates were stored at −20 °C until further processing. Protein concentrations were determined using the BCA assay and BSA as a standard.

There were two conditions (bacteria growing on citrate medium supplemented with 0.05% nicotine and citrate medium alone) with triplicate samples at three different timepoints (7, 10, and 24 h after inoculation). One hundred micrograms of total proteins from each of the nine biological samples were fractionated by SDS-PAGE on 9–16% gradient gels using a PROTEAN II xi Cell (Bio-Rad, Hercules, CA, USA). Proteins fractions were visualized using standard Coomassie Brilliant Blue R250 staining. All lanes corresponding to one biological condition were cut into 20 gel bands and then subjected to trypsin digestion.[32] For this, each gel piece was first thoroughly washed and destained (HPLC grade water for 60 min, and then treated with 50% (v/v) acetonitrile (ACN)/50 mM ammonium bicarbonate (ABC) for 60 min), dehydrated (100% ACN for 60 min), and dried using a Speed Vac. The Cys residues were reduced (10 mM dithiothreitol (DTT) in 25 mM ABC for 60 min) and alkylated (100 mM iodoacetamide in 25 mM ABC for 60 min) in the dark. After another step of dehydration and drying, 200 μL of trypsin solution (10 ng/μL) was added to each gel piece and incubated overnight at 37 °C. Peptide extraction was carried out in two steps, first using 5% FA in 50/50 (v/v) 50 mM ABC/ACN and then using 5% FA in ACN (60 min each). Extracted peptides were pooled, dried, and then cleaned by reversed-phase chromatography using C18 ZipTips (EMD Millipore, Billerica, MA, USA) by following the recommended protocol provided by the supplier.

Nanoliquid Chromatography Tandem Mass Spectrometry

The resulting peptide mixture from each gel piece was loaded onto a 150 μm × 100 mm reversed-phase M-class peptide BEH130 C18 with 1.7 μm 130A UPLC column (Waters, Milford, MA, USA) coupled to a NanoAcquity UPLC (Waters, Milford, MA, USA) system and separated using the following solvent system: solvent A, 0.1% FA in HPLC water; and solvent B, ACN containing 0.1% FA. The separation was performed over a 180 min gradient at a flow rate of 400 nL/min as follows: 1%–45% organic solvent B over 1–120 min, 45%–85% B (120–140 min), constant 85% B (140–160 min), 85%–2% B (160–165 min), and then return to the initial conditions of 1% B (165–180 min). The separated peptides were analyzed using a Q-TOF Xevo G2 MS (Waters) interfaced with the UPLC system through a Picotip Emitter Silicatip nanoelectrospray needle (New Objective, Woburn, MA, USA). MS data acquisition involved 0.5 s survey, MS scans with the m/z range of 350–2000, and automatic data-dependent analysis (DDA) of the top six ions with the highest intensity and the charge of 2+, 3+, or 4+. The MS/MS (recorded over m/z of 50–2000) was triggered when the MS signal intensity exceeded 500 counts/s. In survey MS scans, the six most intense peaks were selected for collision-induced dissociation (CID) and fragmented until the total MS/MS ion counts reached 6000 or for up to 1.1 s each. The entire procedure was also described and used in previous studies.[15,33−35]

Data Analysis

The raw files were converted to peak list files using ProteinLynx Global Server v.2.4 (Waters, Milford, MA, USA) using the default parameters (background subtraction of polynomial order five adaptive with a threshold of 30%, two smoothings with a window of three channels in the Savitzky–Golay mode, and centroid calculation of top 80% of peaks based on a minimum peak width of four channels at half height). Database searches were performed with Mascot v.2.5.1 (Matrix Science, London, U.K.) using a customized database containing the complete protein set from the reference genome of P. aurescens strain TC1 (Uniprot UP000000637),[36] the protein set from the draft genome of P. nicotinovorans strain Hce-1 (Uniprot UP000078426),[37] and the complete protein set of pAO1 megaplasmid extracted from its DNA sequence (GenBank AJ507836.1).[11] Mascot was also set up to search for contaminants in the common Repository of Adventitious Proteins database (January 1, 2012; the Global Proteome Machine, www.thegpm.org/crap). For the estimation of false-positive levels, a decoy database with a reverse database appended at the end of the forward database[38] was also used. Search parameters were with strict trypsin specificity (up to three missed cleavage sites), fragment ion mass tolerance of 1.30 Da, a parent ion tolerance of 0.8 Da, variable modification—oxidation of methionine and fixed modification carbamidomethyl-cysteine.

The resulting data files were analyzed in Scaffold (v.4.8.2, Proteome Software Inc., Portland, OR, USA) using the MudPIT option and further validated. The protein and peptide false discovery rate was set to 0.1%, and the positive hits were accepted if the protein contained at least two identified peptides. Protein probabilities were assigned by the Protein Prophet algorithm.[39] Proteins that could not be differentiated based on MS/MS analysis alone were grouped into clusters to satisfy the principles of parsimony. All hits from the contaminants database were manually filtered out. Label-free relative quantification was performed using weighted spectra and outputted as fold change. Statistical significance of the fold change was assessed by Fisher’s exact[40] test performed within Scaffold. The Benjamini–Hochberg multiple test correction[41] and a significant threshold of p < 0.001 were applied.

The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium[42] via the PRIDE partner repository[43] with the data set identifier PXD012577.