Abstract
Recombinant wild-pyrazinamidase from H37Rv M. tuberculosis was analyzed by gel electrophoresis under differential reducing conditions to evaluate its quaternary structure. PZAse was fractionated by size exclusion chromatography under non-reducing conditions. PZAse activity was measured and mass spectrometry analysis was performed to determine the identity of proteins by de novo sequencing and to determine the presence of disulfide bonds.
This study confirmed that M. tuberculosis wild type PZAse was able to form homo-dimers in vitro. Homo-dimers showed a slightly lower specific PZAse activity compared to monomeric PZAse. PZAse dimers were dissociated into monomers in response to reducing conditions. Mass spectrometry analysis confirmed the existence of disulfide bonds (C72-C138 and C138-C138) stabilizing the quaternary structure of the PZAse homo-dimer.
Keywords: pyrazinamidase, nicotinamidase, pyrazinamide, nicotinamide, dimer, multimer, reducing stress, tuberculosis, mycobacterium, drug resistance
INTRODUCTION
Pyrazinamidase (PZAse) is an essential enzyme in the mechanism of action of pyrazinamide (PZA) in M. tuberculosis. PZA is a pro-drug that enters bacteria by passive diffusion. It is converted to pyrazinoic acid (POA) by PZAse [1–3]. POA is expelled from the mycobacterium by an efflux mechanism [4]. In the acidic extracellular space is protonated and reabsorbed into the bacilli where the proton is released. This cycle results in POA accumulation and reduction of intracellular pH leading to a lethal disruption of membrane permeability [3]
Important characteristics of M. tuberculosis PZAse have been elucidated from the recently crystallized structure of M. tuberculosis PZAse [5] and from crystallized homologous hydrolases (N-Carbamoylsarcosine amidohydrolase from Arthrobacter, nicotinamidases from Pyrococcus horikoshii and Acinetobacter baumanii [6]), as well as from theoretical models of M. tuberculosis mutated PZAses [7]. M. tuberculosis PZAse has an active site that comprises residues D8, A134, and C138, and a metal binding site that comprises residues D49, H51, and H71[5]
In this study we analyze the capacity of M. tuberculosis PZAse to form homo-dimers in vitro stabilized by disulfide bonds.
MATERIALS AND METHODS
Cloning, expression and purification
The DNA from M. tuberculosis H37Rv PZA susceptible pncA wild type reference strain was extracted and the entire pncA gene was amplified and purified (575bp) as described previously [8]. Briefly, for purification, a six-histidine tag was added to the carboxy-terminal end. The pncA cloned plasmid was transformed into E. coli BL21(DE3)pLysS cells (Novagen) as described above for protein expression. Protein expression was induced with 1mM Isopropyl β-D-thiogalactoside (IPTG) and also purified using Histrap columns as described before [9]. Aliquots of the fractions obtained were analyzed by 12% SDS-PAGE, and those containing pure PZAse (molecular weight: 20.7 kDa) were combined. The purified protein was concentrated 10 times and then washed 3 times with 20 mM Tris-HCl, pH 7.9 by ultra-filtration using cellulose membranes with 10 kDa pore size in an Ultracel Amicon Ultrafiltration system (Millipore, Billerica, MA) at 4ºC.
Rabbit anti-PZAse sera
Two 2-month old New Zealand white rabbits, were immunized with recombinant PZAse in four doses given at intervals of four days. The first dose was 5 μg of antigen with 50 μg of saponin. The second, third, and fourth doses included 10, 20, and 40 μg of antigen respectively with 50 μg of saponin. Pre-immune and post-immune sera were tested in a Western immunoblot against recombinant PZAse to confirm the presence of anti-PZAse antibodies. The animal study component was reviewed and approved by the Institutional Review Board of Universidad Peruana Cayetano Heredia.
Polyacrylamide gel electrophoresis under reducing conditions
The purified PZAse (5 mg/ml) was resolved on a SDS-PAGE 12% polyacrylamide gels as described before [9] under reducing and non-reducing conditions. 20 μl of each of these solutions were loaded on each lane. One gel was stained with Coomassie blue method (staining solution: 0.25 g on 100 ml of 45% water, 45% methanol, 10% acetic acid). A second gel (under non-reducing conditions) was stained with Western blot. Proteins were transferred to a nitrocellulose membrane and revealed against the pre and post immune rabbit serum.
Size exclusion chromatography
One ml of purified PZAse (5 mg/ml) was resolved in a G75 Sephadex column (120 ml, 75 cm length), and equilibrated with sodium phosphate buffer (20 mM, 0.25 M NaCl and pH 7.4). Buffer (150 ml) was passed at a flow rate of 0.2 ml/minute and fractions of 2 ml were collected. The absorbance profile at 280 nm was monitored. Aliquots (1.5 ml) were freeze-dried and resuspended in 100 μl of water for further analysis..
PZAse activity
Each eluted fraction from the size exclusion chromatography was tested for PZAse activity with the Wayne test as described before [9] . Each eluted fraction was read in duplicate. The overall experiment was repeated two times. The specific activity was estimated as the amount of μmolar POA produced in a 1 min reaction divided by the total amount of enzyme as described before [9].
Mass spectrometry analysis
Three different bands obtained by polyacrylamide gel electrophoresis stained with Coomassie blue and with molecular masses estimated as 20.5 kDa, 39.7 kDa and 64 kDa (Figure 1) were individually analyzed by mass spectrometry.
Figure 1.
Complexes formation of the H37Rv PZAse. (A) SDS PAGE of soluble PZAse under reducing (1% 2-mercaptoethanol) and non-reducing (-) conditions. (B) Identification of the H37Rv PZAse multimeric complexes by Western immunoblot with pre and post immune sera of rabbit immunized with recombinant PZAse.
In-gel tryptic digestion
In-gel digestion was performed using Trypsin Gold (Promega, Madison, WI). Briefly, excised spots were destained with destaining solution (100 μl of 100 mM ammonium bicarbonate/acetonitrile, 1:1, vol/vol), followed by reduction with 10 mM DTT/100 mM ammonium bicarbonate for 30 min at 60 °C and alkylation with 55 mM iodoacetamide/100 mM ammonium bicarbonate for 20 min in the dark. The gel plugs were washed twice with 50% acetonitrile (ACN)/100 mM ammonium bicarbonate, 20 min each wash, and dehydrated with 100% ACN for 20 min. The gel pieces were subsequently dried and digested overnight in 25 μl of 10 ng/μl trypsin at 37°C. Tryptic peptides were then extracted twice, first with 50% ACN for 15 min, followed by 70% ACN for another 15 min. The extracted solutions were pooled together into a clean tube and dried using a vacuum centrifuge.
Nanoflow HPLC-MS/MS
The tryptic peptide mixture was analyzed by online nanoflow liquid chromatography tandem mass spectrometry (LC-MS/MS) on Accela system (1/20 split-flow) connected to the LTQ Orbitrap Velos instrument (Thermo Fisher Scientific, Bremen, Germany) through a Thermo nanoelectrospray ion source. 5 μl of the peptide mixture was sampled directly onto a capillary column in-house packed with 3-μm C18 beads with a 90-min gradient from 5% to 50% acetonitrile in 0.5% formic acid. To prevent repetition in the detection of peptides due to capillary column overloading, the system was rinsed thoroughly and different samples were intercalated among analyzes. The effluent from the capillary column was directly electrosprayed into the mass spectrometer.
For high-resolution CID and HCD MS/MS top10 method, full scans MS spectra (from m/z 300–1800) were acquired in the Orbitrap analyzer after accumulation to a target value of 1e6 in the linear ion trap. Resolution in the Orbitrap was set to r = 60,000 for full scan MS and r = 30,000 for MS/MS detection. The ten most intense signals with only charge state 2+ and 3+ were sequentially isolated and fragmented in the high-pressure linear ion trap by low energy Collision-Induced Dissociation (CID) with normalized collision energy of 35% and wideband-activation enabled. The same ten most intense peaks doubly and triply charged were also fragmented in parallel in the High-Energy Collision Dissociation (HCD) collision cell with normalized collision energy 40%. The resulting ion fragments from both CID and HCD were detected in the Orbitrap system. The ion selection thresholds were 500 counts for CID and 5,000 counts for HCD, accumulation time was 500 ms for full scans and 250 ms for both dissociation methods. A dynamic exclusion window of 0.5 Da was applied that prevented the same m/z from being selected for 180 s after its acquisition. Standard mass spectrometric conditions for all experiments were: spray voltage, 2.2 kV; no sheath and auxiliary gas flow; heated capillary temperature, 200 °C; predictive automatic gain control (AGC) enabled, and an S-lens RF level of 50–60% [10].
Determination of disulfide bonds
PZAse was digested with Trypsin Gold without any previous treatment with reducing or alkylating agents. The tryptic digestion was performed in Tris-HCl under slightly acidic conditions (pH 6.8) for 12 h at 37°C. The presence of four cysteine residues in the PZAse produces four peptides containing only one cysteine residue after the trypsin digestion: (i) DFHIDPGDHFSGTPDYSSSWPPHC72VSGTPGADFHPSLDTSAIEAVFYK, (ii) ALIIVDVQNDFC14EGGSLAVTGGAALAR, (iii) GVDEVDVVGIATDHC138VR, and (iv) TASVELVC184SS. After trypsin digestion, the exact masses of the possible heterodimers composed by the combination of any pair of the four peptides described above were theoretically calculated. Theoretical molecular masses were convoluted in a charge envelope and the m/z values found were searched against experimental m/z values generated by LC-MS/MS to verify possible combinations and determine the disulfide bond arrangements. All peptides, monomers and possible dimers, were submitted to CID/HCD dissociation and partial amino acid sequence was determined for each dimer.
MS Data Analysis
Tandem mass spectra in Thermo RAW files were used to search a combined forward and reverse pyrazinamidase/nicotinamidase pncA (PZAse encoding gene) database from Mycobacterium tuberculosis H37Rv by Protein Discoverer 1.3 software containing the Sequest algorithm. The spectral search parameters considered doubly and triply charged tryptic peptides ions with a 50 ppm precursor mass window and 10 ppm tolerance was set to fragment ions. Carbamidomethylation (+57.02146) of cysteine was set as a static peptide modification and oxidation of methionine (+15.9949) was set as a variable peptide modification. Candidate peptides were filtered to a 1% protein FDR.
Structural model of the PZAse homo-dimer
Quaternary structure docking was performed by HADDOCK web server [31]. The PZAse structure was obtained from the Protein Data Bank (ID: 3PL1) [25]. The complex structure model was refined using molecular dynamics (MD) with NAMD 2.9. The structure file and a parameter and parameters were generated in VMD 1.9 with CHARMM27 protein topology and parameters. More specific details are shown in the supplementary material.
RESULTS
The H37Rv PZAse was expressed and purified as a soluble protein at a final concentration of 5 mg/ml. The SDS-PAGE of the purified H37Rv PZAse stained with coomassie blue and Western blot revealed under non-reducing conditions. The migration pattern of the purified PZAse revealed bands corresponding to molecular weights of 20.5 kDa, 39.7 kDa and 64 kDa. However, under reducing conditions, a single band of 20.5 kDa was observed predominant (Figure 1).
The chromatogram from the size exclusion chromatography and the profile of PZAse activity, showed peaks (p16, p27 and p37) associated to a 39.7 kDa band and a more intense band of 20.5 kDa (Figure 2). This evidence suggests that the peak p37 is likely to be the PZAse monomer, while peaks p27 and p16 are likely to be predominantly PZAse homo-dimers. The specific PZAse activity of fraction p37 (a mixture of monomers and to a lesser extend dimers) was estimated to be 30.1 μmoles POA mg−1 PZAse min−1, which is within the range of previously reported activity (median 38.4 μmoles POA mg−1 PZAse min−1) [9]. The specific PZAse activities of fraction p27 (likely to be pure dimer) and p16 were 26.7 and 5.7 μmoles POA mg−1 PZAse min−1respectively.
Figure 2.

Separation of protein multimeric complexes by size exclusion chromatography (SEC) (A) Protein distribution detected by absorbance at 280 nm in ( - ); and PZAse activity (△) detected by the Wayne assay of the purified recombinant PZAse. (B) SDS-PAGE of protein fractions associated to the 3 PZAse activity peaks collected during the size exclusion chromatography (p16, p27, p37).
LC-MS/MS analysis identified PZAse with at least 2 unique peptides per sample with probability greater than 99%. Sample of 20 kDa were identified with 33.87% sequence coverage whereas samples of 40 kDa and 60 kDa were both identified with 19.35% sequence coverage. These results unequivocally identified PZAse in their multimeric forms. Confirming this finding, the de novo sequencing of the generated peptides that were not linked by a disulfide bond confirmed the identity of the PZAse. The combination of peptides linked by a disulfide bond were (i)+(iii) (6971.1 Da), and (iii)+(iii) (3563.72 Da). The MS/MS fragmentation of these dimers was sequenced de novo, confirming once again that a disulfide bond links these peptides (Figure 4 in supplementary material). Therefore it appears that at least disulfide bonds C72-C138 and C138-C138 stabilize the dimers.
The modeling of the PZAse dimer structure for the C72-C138 214 dimer, found that the best cluster had 50 models with an average Van der Walls stabilization energy (VdW_E) of −52.8 ± 2.0 kcal/mol, an average electrostatic energy (El_E) of −197.3 ± 16.2 kcal/mol, and an average desolvation energy (Des_E) of −6.9 ± 5.5 kcal/mol. The best scored cluster for the C138−C138 dimer had 68 models with VdW_E of −62.9 ± 4.1 kcal/mol, El_E of −212.1 ± 19.8 kcal/mol and Des_E of 8.3 ± 4.1 kcal/mol. The MD simulation found a distance between the disulfide cysteines sulphur atoms for the C72-C138 and the C138-C138 dimers as 2.05 Å and 2.11 Å respectively. The orientation of the disulfide bond of the type C172-C2138 allows the active site of the first subunit to be exposed to the solvent in contrast with the C1138-C2138 dimer, where the catalytic residue C138 is blocked in both subunits, lacking PZAse activity (Figure 3).
Figure 3.
The most favorable structures found by the HADDOCK server: (A) C138-C138 dimer after 1ns of molecular Dynamics. Both metal binding sites are shown in black, as well as the catalytic sites in white. Both of the active sites are inactivated as the catalytic cysteines shown in color form part of the disulfide bond and the (B) C138-C72 dimer after 1ns of molecular Dynamics. Both metal binding sites are shown in black, as well as the catalytic sites in white. The catalytic site from the PZAse, which binds through C72, is complete and permeable to the solvent. The disulfide bond between C72-C138 is also shown in color.
DISCUSSION
This is the first report of M. tuberculosis homo-dimers stabilized by disulfide bonds retaining enzymatic activity in vitro. PZAse homo-dimer was only observed under non-reducing conditions.
In agreement with previous studies, the most abundant form of the PZAse in vitro is the monomeric form [11]. With a lesser abundance we found evidence of PZAse homo-multimers. A similar result has been found on previous reports of homologous nicotinamidases from Leisnmania infantum, Acinetobacter baumanii and Oceanobacillus iheyensis that are able to organize as dimers [6, 12, 13].
This study demonstrates that disulfide bonds stabilize the PZAse dimer. There are several ways in which the subunits can accommodate to form multimers, however only C72-C138 and C138-C138 were detected experimentally. According to the X-Ray crystalline structure of the H37Rv M. tuberculosis PZAse (PDB code: 3PL1), C184 is a naturally exposed cysteine likely to form disulfide bonds with itself and C72 or C14, although C14 was not found to establish disulfide bonds.
The interaction between C72 and C138 does not show steric or electrostatic impairments associated with this disulfide bond. Participation of C138 impairs PZAse activity of the particular subunit because C138 is a catalytic aminoacid. The enzymatic activity of the multimer complexes interacting with C138 is associated with the specific activity of the PZAse subunits that do not bind through C138, which explains the observed PZAse activity in the multimeric complexes. The estimation of the specific activity on the eluted protein fraction of the monomer is slightly higher than the corresponding eluted protein fraction of the dimer.
Only one crystal structure of monomeric M. tuberculosis PZAse has been reported [5] Despite multiple efforts and the importance of further understanding the structure of mutated PZAses, no other group had succeeded in its crystallization. The fact that no dimers were observed in the only M. tuberculosis PZAse crystal structure is probably due to the specific conditions used in the crystalization process. Eventually the possibility of PZAse homo-oligomers formation could be the reason why PZAse crystals are not easily reproduced.
M. tuberculosis PZAse has nicotinamidase activity. It hydrolyzes nicotinamide to nicotinic acid and ammonia in the β-Nicotinamide adenine dinucleotide (NAD+) salvage pathway [14]. A function of NAD is the modulation of anti oxidation and oxidative stress, as the NADH/NAD+ ratio is a measure of cellular reducing potential Intracellular redox changes in M. tuberculosis has been previously reported [15]. Cultures of M. tuberculosis in presence of certain antibiotics were associated to a NADH/NAD+ ratio change of 2.65 fold compared to untreated mycobacterium.
Studies in mice liver indicate that excessive intracellular NADH can produce ‘reductive stress’, which may result from its capacity to induce release of ferrous iron from ferritin, an enzyme which is present in M. tuberculosis
Any evolutionary advantage or biological role of the PZAse homo-dimers in vivo might be associated to its specific nicotinamidase activity. To clarify this, further studies are necessary.
Conclusion
This study demonstrates that the M. tuberculosis PZAse is able to form homo-dimers with PZAse activity in vitro.
Supplementary Material
Acknowledgments
We thank the technical team from the Laboratorio Universitario de Proteómica - Instituto de Biotecnología, Universidad Nacional Autónoma de Mexico. We are grateful with Dr. Vincent Kiley for his revision and editing of the manuscript and comments. We thank Alen Zimic for his idea in the interpretation of some results of this study.
Financial support
This research was funded by the Wellcome Trust Intermediate Fellowship awarded to PS and by the National Institute of Allergy and Infectious Diseases, National Institutes of Health US, under the terms of Award 1R01TW008669-01 awarded to MZ. D.R. was supported by an scholarship of the Franco-Peruvian School of Life Sciences.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Konno K, Feldmann FM, McDermott W. Pyrazinamide susceptibility and amidase activity of tubercle bacilli. Am Rev Respir Dis. 1967;95(3):461–9. doi: 10.1164/arrd.1967.95.3.461. [DOI] [PubMed] [Google Scholar]
- 2.Scorpio A, et al. Characterization of pncA mutations in pyrazinamide-resistant Mycobacterium tuberculosis. Antimicrob Agents Chemother. 1997;41(3):540–3. doi: 10.1128/aac.41.3.540. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Zhang Y, et al. Mode of action of pyrazinamide: disruption of Mycobacterium tuberculosis membrane transport and energetics by pyrazinoic acid. J Antimicrob Chemother. 2003;52(5):790–5. doi: 10.1093/jac/dkg446. [DOI] [PubMed] [Google Scholar]
- 4.Zhang Y, et al. Role of acid pH and deficient efflux of pyrazinoic acid in unique susceptibility of Mycobacterium tuberculosis to pyrazinamide. J Bacteriol. 1999;181(7):2044–9. doi: 10.1128/jb.181.7.2044-2049.1999. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Petrella S, et al. Crystal structure of the pyrazinamidase of Mycobacterium tuberculosis: insights into natural and acquired resistance to pyrazinamide. PLoS One. 2011;6(1):e15785. doi: 10.1371/journal.pone.0015785. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Fyfe PK, et al. Specificity and mechanism of Acinetobacter baumanii nicotinamidase: implications for activation of the front-line tuberculosis drug pyrazinamide. Angew Chem Int Ed Engl. 2009;48(48):9176–9. doi: 10.1002/anie.200903407. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Quiliano M, et al. Structure-Activity relationship in mutated pyrazinamidases from Mycobacterium tuberculosis. Bioinformation. 2011;6(9):335–9. doi: 10.6026/97320630006335. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.van Soolingen D, et al. Occurrence and stability of insertion sequences in Mycobacterium tuberculosis complex strains: evaluation of an insertion sequence-dependent DNA polymorphism as a tool in the epidemiology of tuberculosis. J Clin Microbiol. 1991;29(11):2578–86. doi: 10.1128/jcm.29.11.2578-2586.1991. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Sheen P, et al. Effect of pyrazinamidase activity on pyrazinamide resistance in Mycobacterium tuberculosis. Tuberculosis (Edinb) 2009;89(2):109–13. doi: 10.1016/j.tube.2009.01.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Shevchenko A, et al. In-gel digestion for mass spectrometric characterization of proteins and proteomes. Nat Protoc. 2006;1(6):2856–60. doi: 10.1038/nprot.2006.468. [DOI] [PubMed] [Google Scholar]
- 11.Zhang H, et al. Characterization of Mycobacterium tuberculosis nicotinamidase/pyrazinamidase. FEBS J. 2008;275(4):753–62. doi: 10.1111/j.1742-4658.2007.06241.x. [DOI] [PubMed] [Google Scholar]
- 12.Gazanion E, et al. The Leishmania nicotinamidase is essential for NAD+ production and parasite proliferation. Mol Microbiol. 82(1):21–38. doi: 10.1111/j.1365-2958.2011.07799.x. [DOI] [PubMed] [Google Scholar]
- 13.Sanchez-Carron G, et al. Biochemical and mutational analysis of a novel nicotinamidase from Oceanobacillus iheyensis HTE831. PLoS One. 8(2):e56727. doi: 10.1371/journal.pone.0056727. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Boshoff HI, et al. Biosynthesis and recycling of nicotinamide cofactors in mycobacterium tuberculosis. An essential role for NAD in nonreplicating bacilli. J Biol Chem. 2008;283(28):19329–41. doi: 10.1074/jbc.M800694200. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Rao SP, et al. The protonmotive force is required for 355 maintaining ATP homeostasis and viability of hypoxic, nonreplicating Mycobacterium tuberculosis. Proc Natl Acad Sci U S A. 2008;105(33):11945–50. doi: 10.1073/pnas.0711697105. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.


