- Email: [email protected]
E84G mutation in dihydrofolate reductase from drug resistant strains of Mycobacterium tuberculosis (Mumbai, India) leads to increased interaction with Trimethoprim
International Journal of Mycobacteriology
H O S T E D BY
4 ( 2 0 1 5 ) 9 7 –1 0 3
Available at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/IJMYCO
E84G mutation in dihydrofolate reductase from drug resistant strains of Mycobacterium tuberculosis (Mumbai, India) leads to increased interaction with Trimethoprim 5 Archana Raju a, Savita Kulkarni b, M.K. Ray b, M.G.R. Rajan b, Mariam S. Degani
a,*
a
Institute of Chemical Technology, Department of Pharmaceutical Sciences and Technology, N.P. Marg, Matunga, Mumbai 400019, India Radiation Medicine Centre – BARC, Tuberculosis Immunology & Immunoassay Development Section, Tata Memorial Hospital, Annexe Bldg, Parel, Mumbai 400012, India b
A R T I C L E I N F O
A B S T R A C T
Article history:
Background: Dihydrofolate reductase (DHFR) (dfrA gene) is an essential enzyme for cell sur-
Received 4 February 2015
vival and an unexplored target in Mycobacterium tuberculosis (Mtb). This study was carried
Accepted 16 February 2015
out to analyze mutations in the dfrA gene amongst 20 clinical DNA samples from Mtb iso-
Available online 13 March 2015
lates obtained from Mumbai, India.
Keywords:
point mutation in one strain, leading to a glutamic acid to glycine change. In silico sim-
Mycobacterium tuberculosis
ulation studies revealed a surface alteration in the enzyme due to this E84G mutation.
Dihydrofolate reductase
The amplified mutant gene was cloned and expressed. The mutant protein was assessed
Point mutation
against known DHFR inhibitors: Methotrexate and Trimethoprim.
Methods: Sequencing of the PCR amplified dfrA gene from these DNA isolates revealed a
Molecular docking
Results: An increased affinity for inhibitor Trimethoprim and native substrate dihydrofo-
Enzyme assay
late was observed with the mutant. Methotrexate did not vary in its activity with both
Trimethoprim
the enzyme forms. Conclusions: The Glu84Gly point mutation may lead to a variation in the strain which may cause resistance in the future. Ó 2015 Asian African Society for Mycobacteriology. Production and hosting by Elsevier Ltd. All rights reserved.
Introduction Biosynthesis of a variety of cellular components involves the presence of reduced folates in both prokaryotes and eukaryotes. The normal mechanism of synthesis involves de novo biosynthesis along with salvage pathways in some bacteria and mammalian cells. Dihydrofolate reductase (DHFR)
(encoded by dfrA gene) has an essential role in the maintenance of cellular pools of reduced folate, tetrahydrofolate (THF) and its derivatives, including methylation of dUMP to dTMP, and in cell growth and proliferation. The enzyme is hence a target for several anti-bacterial and anti-cancer agents [1,2]. Thymidylate synthase (TS) is the enzyme which functions downstream of DHFR in the folate pathway. Two
* Corresponding author. E-mail addresses: [email protected], [email protected] (M.S. Degani). Peer review under responsibility of Asian African Society for Mycobacteriology. http://dx.doi.org/10.1016/j.ijmyco.2015.02.001 2212-5531/Ó 2015 Asian African Society for Mycobacteriology. Production and hosting by Elsevier Ltd. All rights reserved.
98
International Journal of Mycobacteriology
forms of TS enzyme coexist in Mycobacterium tuberculosis (Mtb), namely ThyA and ThyX [3]. ThyA is involved in de novo synthesis, while ThyX functions as a Pyrimidine Salvage Pathway enzyme. The conversion of dUMP to dTMP, carried out by both forms of thymidylate synthase, requires the presence of tetrahydrofolate, which is formed due to the activity of DHFR enzymes. Thus, an inhibited DHFR enzyme would result in the accumulation of dUMP, which would get converted to dUTP, and the incorporation of dUTP into DNA results in cell death [4]. Mtb, a human pathogen, claims more lives than any other known infectious organism around the world [5,6]. The current therapy involves a four-drug regime [7–9]. However, due to the emergence of multi-drug resistant (MDR) forms of the organism, the current treatment proves inefficient [10]. A sequence alignment of human and Mtb DHFRs indicates only 26% sequence identity [11]. However, mutations in this target may disrupt the inherent binding ability of this enzyme to known inhibitors. Hence, the present study was undertaken to investigate 20 clinical isolates of first-line drug-resistant tuberculosis (TB) for inherent mutations in the dfrA gene. The studies led to the identification of a point mutation replacing Adenine by Guanine at the 251st position in the dfrA gene (480 bp) of one of the 20 screened isolates. This leads to an amino acid change of glutamic acid to glycine at the 84th position of the enzyme. Docking software was employed to locate the position of this alteration to the surface of the protein. Thus, it was assumed that the modification would not alter the interaction directly at the active site, but might allosterically or structurally affect it. This was in accordance with the fact that DHFR is an essential enzyme, and the organism might not mutate in a manner to affect its inherent function in order to maintain its own viability. The mutated gene was cloned and expressed using pET21a vector. The expressed protein was assessed using SDS–PAGE and Km determination for native substrate, dihydrofolate (DHF). The mutant was also screened against known inhibitors of DHFR, Trimethoprim (TMP) and Methotrexate (MTX), by both in silico docking studies and in vitro enzyme assay. The results obtained indicate better binding of the mutant enzyme with the substrate, DHF and inhibitor, TMP when compared with the wild type.
Materials and methods Chemicals and standard drugs DHF was purchased from Sigma Aldrich. MTX was obtained as a gift sample from Khandelwal Laboratories Pvt. Ltd., Mumbai, India. TMP, buffer salts, reduced nicotinamide adenine dinucleotide phosphate (NADPH) and Dithiothreitol (DTT) were purchased from HiMedia.
Amplification of dfrA gene from 20 clinical isolates The chromosomal DNA of 20 clinical isolates was procured from RMC-BARC, Mumbai, India. The primary source of these strains was Sewri Tuberculosis Hospital, Mumbai, India. The genomic DNA was extracted from all the strains by the standard cetyl trimethyl ammonium bromide (CTAB) method [12].
4 ( 2 0 1 5 ) 9 7 –1 0 3
These strains were selected based on their resistance against first-line drugs identified by employing the recommended WHO protocol [13]. Out of these isolated strains, 3 were resistant to Isoniazid, 12 to Rifampicin, 3 others to Isoniazid and Rifampicin, and 2 were resistant to Isoniazid, Rifampicin and Streptomycin. The drug susceptibility was performed by the standard proportion method as described in Standard Operating Protocol (SOP) prepared in the Revised National TB Control Programme Training Manual for ‘‘Mycobacterium tuberculosis Culture & Drug susceptibility testing’’. The document was prepared by the Central TB Division, Directorate General of Health Services, Ministry of Health and Family Welfare Government of India. Along with these DNA samples, the wild type Mtb H37Rv DNA sample was also procured as a standard. The complete dfrA gene (480 bp) was amplified from these isolates and standard using the forward primer 5P-GAC GCG TGTGCA TAT GGT GGG GCT GAT CTG-3P and reverse primer 5P-GCG ATG AGG ATC CGC GGC GCT CATGAG CGG-3P, containing NdeI and BamHI restriction sites, respectively [14]. The amplified fragments were checked on agarose gel and sequenced by employing dideoxy DNA sequencing method of Sanger et al. using ABI Prism 377-18 automated DNA sequencer and the Big Dye terminator kit (ABI, Foster City, CA). A point mutation was found in one of the 20 isolates.
In silico mutation analysis The ligands (Methotrexate and Trimethoprim) were sketched using the 2D-sketcher of Maestro (GLIDE, Schro¨dinger 9.0, LLC, NY). Low-energy 3D structures of ligands were generated with the help of Ligprep, by adding appropriate hydrogens, generating possible ionization states and tautomers. The ligands were subsequently subjected to energy minimization using the OPLS-2005 force field. The protein structure PDB code: 1DG8 which is the complex of Mtb DHFR enzyme and NADPH with resolution of 2.00 was retrieved from Brookhaven Protein Data Bank (www.rcsb.com) and used for validation of the docking algorithm. All water molecules were deleted, bond orders and charges of cofactor NADPH were assigned properly in the protein preparation step. The amino acid at the 84th position was altered to generate the mutant enzyme. The docking study was carried out using GLIDE software in both standard precision (SP) and extra precision (XP) mode with ‘must match’ of at least one hydrogen bond criteria; with Ile5, Asp27 and Ile94 (10) using both the wild type and the mutant protein structure.
Cloning, expression and purification of the mutant identified The mutant amplified fragment was cleaned using the QIAquick PCR purification kit (Qiagen). Both the amplified dfrA gene and pET21a expression vector (Novagen) were then digested with NdeI and BamHI, purified on a 1.0% agarose gel, eluted, and ligated together using T4 DNA ligase overnight at room temperature (20 °C). The resulting construct was sequenced to confirm that the dfrA gene was in the proper orientation for expression and that no other mutations were introduced during the PCR amplification. After propagation in Escherichia coli strain DH5a, the construct was purified and transformed into E. coli strain BL21 STAR (Invitrogen) for expression.
International Journal of Mycobacteriology
The DHFR protein was over-expressed under the pET21a T7 promoter in recombinant E. coli BL21 cells. Briefly, the recombinant cells were grown in Luria-Bertani (LB) medium (HiMedia) at 37 °C until optical density (OD600) reached 0.4– 0.5. Induction was performed by the addition of 1 mM isopropyl-L-thiogalactopyranoside (IPTG), and cells were grown for two and half hours more at 37 °C to reach an OD of 1.5. Cells were centrifuged, washed with cold phosphate-buffered saline (PBS) and the pellet frozen at 80 °C. The pellet was thawed and used for enzyme purification using Ni-IDA (Nickel-Iminodiacetic acid) column comprising of Sephabeads IP-IDA resin and urea and imidazole as eluents. The purified enzyme, obtained from the pellet after column chromatography, was analyzed using SDS–PAGE and characterized by Km determination for substrate dihydrofolate (DHF).
IC50 determination using enzyme assay The mutant DHFR activity was determined spectrophotometrically against known ligands MTX and TMP by monitoring the decreasing absorbance for 2 min at 340 nm and 37 °C. Standard assay mixture contained 50 mM KPO4 (pH 7.4), 5 mM DTT, and 60 lM NADPH, and the reaction was started after a 3 min incubation by the addition of 45 lM DHF [14,15]. The wild type enzyme obtained after optimized purification from recombinant yeast [16–18] was used as the control DHFR for the studies. The wild type and mutant enzymes were characterized by determining Km values for the substrate DHF using MS Excel graph plots. The inhibitors MTX and TMP were used in the reaction mixtures in 5 different concentrations. The assay was carried out in microtiter plates in duplicates in the assay and analyzed using the kinetic mode of the micro plate reader from Synergy H1 BioTek with Gen5 software. The Inhibitory concentration 50 (IC50) values were determined from graph plots generated by MS Excel and standard deviation was calculated.
4 (2 0 15 ) 9 7–10 3
99
Results The amplified dfrA gene of the 20 clinical isolates and the standard Mtb H37Rv were analyzed on agarose gel where 480 bp sequences were visualized. These amplified genes were sequenced and the sequencing revealed a point mutation in isolate number 19 which was absent in the wild type H37Rv strain when analyzed using Serial cloner 2.5 (alignment tool, sequence aligned with Mtb DHFR with GenBank accession number NC_000962). A transition from Adenine to Guanine at position 251 in the dfrA gene of an isolate, which was resistant to Isoniazid, Rifampicin and Streptomycin (MIC at concentrations of 0.2 lg/ml for Isoniazid, 40 lg/ml for Rifampicin and 4 lg/ml for Streptomycin), led to the identification of an alteration of amino acid at the 84th position (glutamic acid to glycine) which was found to result in a surface modification (slightly unaligned b sheet from amino acid 79 to 85) through in silico simulation studies (Fig. 1). This might be the result of a change in surface charge from negative to neutral. Hence the interaction with the nearby side chain is minimized thereby resulting in distancing of the entire b sheet. As a result of the in silico studies, it was revealed that the known inhibitors MTX and TMP interacted with the wild type and the mutant differently (Table 1-Docking scores). MTX showed complete alignment in the docked images of wild type as well as mutant since the slight alignment variation did not cause any contact changes within the active site (Fig. 2). However, TMP showed altered and, curiously, a better docking score with the mutant than the wild type which is supported by the change in alignment of TMP in the active site of mutant as depicted in Fig. 3. Also, any alteration in the Mtb DHFR surface geometry has been reported to affect the final binding of TMP in the active site as reported by Nerukh et al. [19]. Hence, this surface modification could have led to the differential binding of TMP in the active site pocket of the mutant enzyme.
Figure 1 – Overlap of wild type (pink) and mutant (green) DHFR enzyme depicting area of surface modification around 84th amino acid residue.
100
International Journal of Mycobacteriology
4 ( 2 0 1 5 ) 9 7 –1 0 3
Table 1 – In silico Docking scores and in vitro IC50 values of wild type and mutant Mtb DHFR with MTX and TMP. Ligand
Wild type DHFR Docking Score
MTX TMP
9.38 4.43
Mutant DHFR IC50 (lM)
Docking Score
0.008 ± 0.0007 16.0 ± 3.0
9.38 6.60
IC50 (lM) 0.008 ± 0.0004 10.5 ± 2.0
˚ surface generated for wild type (pink) and (b) Figure 2 – Interaction diagram of DHFR enzyme with MTX (orange) (a) with 8 A ˚ ) (d) Ligplot of MTX with mutant (6 A ˚ ). with mutant (green) (c) Ligplot of MTX with wild type (6 A
In order to establish the validity of the docking, a construct with the mutant gene insert was prepared and purified. The new construct (pET21a-MtbDHFR#19) was transformed in E. coli BL21 and later used for protein expression. Purification of the Mtb DHFR using the Ni-IDA column was confirmed using the SDS–PAGE which revealed a 19 kDa protein band. Formation of inclusion bodies in the E. coli host leads to enzyme loss. Since this issue was not encountered, the purified enzyme was considered for further studies. The wild type enzyme was purified from recombinant yeast. A comparative study of the two enzymes was undertaken since both were obtained in pure form.
The enzymes were characterized by Km determination and assessed for activity against the ligands MTX and TMP. The Km value of the mutant enzyme (4.4 ± 0.02 lM) is slightly less than the Km value for the wild type enzyme (4.6 ± 0.05 lM) indicating that the mutant has a slightly higher affinity towards the substrate DHF. The comparative wild type and mutant enzyme IC50 values are as shown in Table 1. When the IC50 values were compared with the reported IC50 of the wild type [15], the mutant shows a visibly better interaction with TMP as indicated by the docking studies thus validating the in silico obtained results.
International Journal of Mycobacteriology
4 (2 0 15 ) 9 7–10 3
101
˚ surface generated for wild type (pink) and (b) with Figure 3 – Interaction diagram of DHFR enzyme with TMP (blue) (a) with 8 A ˚ ) (d) Ligplot of TMP with mutant (6 A ˚ ). mutant (green) depicting change in alignment (c) Ligplot of TMP with wild type (6 A
Discussion The current work aimed at checking for mutation in the dfrA gene from Mtb and analyzing the effect of the mutant enzyme on inhibitor binding. Thus, the DNA from 20 clinical isolates of Mtb showing variable resistance to first-line drugs were used for amplification of the dfrA gene, and one DNA isolate revealed a point mutation when compared with the wild type Mtb H37Rv at the 84th position from glutamic acid to glycine, which led to a surface modification in the 3-D structure of the protein when analyzed using in silico methods. Similar studies carried out by Nosova et al., with other enzymes of Mtb, have indicated the association of these enzyme mutations, with resistance to second-line drugs [20]. Resistance to known drugs was also found to be associated with the DHFR enzyme mutations in other organisms like Plasmodium vivax [21]. Since the strain with the mutation was found to be resistant to the first-line drugs Isoniazid, Rifampicin and Streptomycin, the studies were carried forward to in vitro stages. The mutant protein was cloned, expressed and studied by Km determination and for inhibition due to known drugs by
employing an enzyme assay. On comparison with the wild type, it revealed a peculiarly higher affinity towards the native substrate DHF and inhibitor TMP. The data with TMP showing variation co-related with the in silico docking studies. This data can be justified since a change in the surface geometry of DHFR has been reported to alter the final binding of TMP with DHFR [19]. In this experimentation, the inhibitory concentration of TMP has shown an improvement, indicating that the mutation has no detrimental effect on inhibitor binding as was expected, but makes the enzyme more susceptible to inhibition. However, it has been reported that in a number of pathogens, resistance towards DHFR inhibitors develops by the simple accumulation of point mutations in the dfrA gene [22,23]. Further, the mutant enzyme also showed an increased affinity towards the native substrate, DHF. These increased interactions may be due to the two conformational states showing varying affinities towards anti-folate drugs that have been proposed for Mtb DHFR [24].
102
International Journal of Mycobacteriology
Conclusions Point mutations in the target genes have been reported as the principal mechanism of development of drug resistance in Mtb [25,26]. Also, mutations in dfrA have been reported to play a role in Mtb resistance development [27]. Additionally, Isoniazid (INH) is known to inhibit Mtb DHFR in vitro [28], and dfrA mutations have also been reported to contribute to the detection of Isoniazid-resistant Mtb isolates along with other mutations [29,30]. The Glu84Gly (E84G) mutation in DHFR conferring increased affinity towards Trimethoprim in Mtb or other organisms has been reported for the first time in this paper. Considering the above-mentioned literature and the fact that the said mutant strain is resistant to Isoniazid, this point mutation may be proposed to lead to a variation in the strain which might contribute towards resistance in the future. The in vitro analysis of this particular mutant in the near future would affirm this assumption.
Conflict of interest There are no conflicts of interest.
Acknowledgements Authors are thankful to the Indian Council of Medical Research, Delhi, India (PHA-BMS-45/94/2012) for financial assistance. The authors would like to thank Dr. Carol H. Sibley for the recombinant yeast containing wild type Mtb DHFR. R E F E R E N C E S
[1] G.B. Elion, Nobel lecture, the purine path to chemotherapy, Biosci. Rep. 9 (1989) 509–529. [2] G.H. Hitchings, Nobel lecture in physiology or medicine, selective inhibitors of dihydrofolate reductase, In Vitro Cell Dev. Biol. 25 (1989) 303–310. [3] J. Rengarajan, C.M. Sassetti, V. Naroditskaya, A. Sloutsky, B.R. Bloom, E.J. Rubin, The folate pathway is a target for resistance to the drug para-aminosalicylic acid (PAS) in Mycobacteria, Mol. Microbiol. 53 (2004) 275–282. [4] A.D. Villela, Z.A. Sanchez-Quitian, R.G. Ducati, D.S. Santos, L.A. Basso, Pyrimidine salvage pathway in Mycobacterium tuberculosis, Curr. Med. Chem. 18 (2011) 1286–1298. [5] P.J. Dolin, M.C. Raviglione, A. Kochi, Global tuberculosis incidence and mortality during 1990–2000, Bull. World Health Organ. 72 (1994) 213–220. [6] A.M. Rouhi, Tuberculosis: a tough adversary, Chem. Eng. News 77 (1999) 52–70. [7] B.R. Bloom, C.J. Murray, Tuberculosis: commentary on a reemergent killer, Science 257 (1992) 1055–1064. [8] D.E. Snider, W.L. Roper, The new tuberculosis, N. Engl. J. Med. 326 (1992) 703–705. [9] J.B. Bass, L.S. Farer, P.C. Hopewell, R. O’Brien, R.F. Jacobs, F. Ruben, et al, Treatment of tuberculosis and tuberculosis infection in adults and children, Am. J. Respir. Crit. Care Med. 149 (1994) 1359–1374. [10] H. Nakajima, Tuberculosis: A Global Emergency, World Health Organization, 1993.
4 ( 2 0 1 5 ) 9 7 –1 0 3
[11] R. Li, R. Sirawaraporn, P. Chitnumsub, W. Sirawaraporn, J. Wooden, F. Athappilly, et al, Three-dimensional structure of M. tuberculosis dihydrofolate reductase reveals opportunities for the design of novel tuberculosis drugs, J. Mol. Biol. 295 (2000) 307–323. [12] S. Honore-Bouakline, J.P. Vincensini, V. Giacuzzo, P.H. Lagrange, J.L. Herrmann, Rapid diagnosis of extrapulmonary tuberculosis by PCR: impact of sample preparation and DNA extraction, J. Clin. Microbiol. 41 (2003) 2323–2329. [13] P. Chakraborty, S. Kulkarni, R. Rajan, K. Sainis, Drug resistant clinical isolates of Mycobacterium tuberculosis from different genotypes exhibit differential host responses in THP-1 cells, PLoS One 8 (2013) e62966. [14] E.L. White, L.J. Ross, A. Cunningham, V. Escuyer, Cloning, expression, and characterization of Mycobacterium tuberculosis dihydrofolate reductase, FEMS Microbiol. Lett. 232 (2004) 101–105. [15] B.L. Hillcoat, P.F. Nixon, R.L. Blakley, Effect of substrate decomposition on the spectrophotometric assay of dihydrofolate reductase, Anal. Biochem. 21 (1967) 178–189. [16] A.B. Gerum, J.E. Ulmer, D.P. Jacobus, N.P. Jensen, D.R. Sherman, C.H. Sibley, Novel Saccharomyces cerevisiae screen identifies WR99210 analogues that inhibit Mycobacterium tuberculosis dihydrofolate reductase, Antimicrob. Agents Chemother. 46 (2002) 3362–3369. [17] A. Raju, M.S. Degani, M.A. Khedkar, S.N. Niphadkar, Novel affinity chromatographic technique for purification of dihydrofolate reductase from recombinant yeast, Int. J. Drug Des. 4 (2013) 994–997. [18] A. Raju, M.A. Khedkar, M.S. Degani, Optimization of Mycobacterium tuberculosis DHFR production from recombinant Saccharomyces cerevisiae, Int. J. Curr. Microbiol. Appl. Sci. 2 (2013) 70–79. [19] D. Nerukh, N. Okimoto, A. Suenaga, M. Taiji, Ligand diffusion on protein surface observed in molecular dynamics simulation, J. Phys. Chem. Lett. 3 (2012) 3476–3479. [20] E.Y. Nosova, A.A. Bukatina, Y.D. Isaeva, M.V. Makarova, Y. Galkina, A.M. Moroz, Analysis of mutations in the gyrA and gyrB genes and their association with the resistance of Mycobacterium tuberculosis to levofloxacin, moxifloxacin and gatifloxacin, J. Med. Microbiol. 62 (2013) 108–113. [21] W.J. Lee, H.H. Kim, Y.K. Choi, K.M. Choi, M.A. Kim, J.Y. Kim, et al, Analysis of the dihydrofolate reductase-thymidylate synthase gene sequences in Plasmodium vivax field isolates that failed chloroquine treatment, Malar. J. 9 (2010) 331–337. [22] J.E. Hyde, The dihydrofolate reductase-thymidylate synthase gene in the drug resistance of malaria parasites, Pharmacol. Ther. 48 (1990) 45–59. [23] B.I. Schweitzer, A.P. Dicker, J.R. Bertino, Dihydrofolate reductase as a therapeutic target, FASEB J. 4 (1990) 2441–2452. [24] M.V. Dias, P. Tyrakis, R.R. Domingues, A.F. Paes Leme, T.L. Blundell, Mycobacterium tuberculosis dihydrofolate reductase reveals two conformational states and a possible low affinity mechanism to antifolate drugs, Structure 22 (2014) 94–103. [25] S.T. Cole, A. Telenti, Drug resistance in Mycobacterium tuberculosis, Eur. Respir. J. Suppl. 20 (1995) 701–713. [26] J.M. Musser, Antimicrobial agent resistance in mycobacteria: molecular genetic insights, Clin. Microbiol. Rev. 8 (1995) 496– 514. [27] C.U. Koser, R.N. Veerapen-Pierce, D.K. Summers, J.A. Archer, Role of mutations in dihydrofolate reductase dfrA (Rv2763c) and thymidylate synthase thyA (Rv2764c) in Mycobacterium tuberculosis drug resistance, Antimicrob. Agents Chemother. 54 (2010) 4522–4523. [28] A. Argyrou, M.W. Vetting, B. Aladegbami, J.S. Blanchard, Mycobacterium tuberculosis dihydrofolate reductase is a target for isoniazid, Nat. Struct. Mol. Biol. 13 (2006) 408–413.
International Journal of Mycobacteriology
[29] Y.M. Ho, Y.J. Sun, S.Y. Wong, A.S.G. Lee, Contribution of dfrA and inhA mutations to the detection of isoniazid-resistant Mycobacterium tuberculosis isolates, Antimicrob. Agents Chemother. 53 (2009) 4010–4012. [30] T. Jagielski, Z. Bakula, K. Roeske, M. Kaminski, A. Napiorkowska, E. Augustynowicz-Kopec, et al, Detection of
4 (2 0 15 ) 9 7–10 3
103
mutations associated with isoniazid resistance in multidrugresistant Mycobacterium tuberculosis clinical isolates, J. Antimicrob. Chemother. (2014), http://dx.doi.org/10.1093/jac/ dku161.
H O S T E D BY
4 ( 2 0 1 5 ) 9 7 –1 0 3
Available at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/IJMYCO
E84G mutation in dihydrofolate reductase from drug resistant strains of Mycobacterium tuberculosis (Mumbai, India) leads to increased interaction with Trimethoprim 5 Archana Raju a, Savita Kulkarni b, M.K. Ray b, M.G.R. Rajan b, Mariam S. Degani
a,*
a
Institute of Chemical Technology, Department of Pharmaceutical Sciences and Technology, N.P. Marg, Matunga, Mumbai 400019, India Radiation Medicine Centre – BARC, Tuberculosis Immunology & Immunoassay Development Section, Tata Memorial Hospital, Annexe Bldg, Parel, Mumbai 400012, India b
A R T I C L E I N F O
A B S T R A C T
Article history:
Background: Dihydrofolate reductase (DHFR) (dfrA gene) is an essential enzyme for cell sur-
Received 4 February 2015
vival and an unexplored target in Mycobacterium tuberculosis (Mtb). This study was carried
Accepted 16 February 2015
out to analyze mutations in the dfrA gene amongst 20 clinical DNA samples from Mtb iso-
Available online 13 March 2015
lates obtained from Mumbai, India.
Keywords:
point mutation in one strain, leading to a glutamic acid to glycine change. In silico sim-
Mycobacterium tuberculosis
ulation studies revealed a surface alteration in the enzyme due to this E84G mutation.
Dihydrofolate reductase
The amplified mutant gene was cloned and expressed. The mutant protein was assessed
Point mutation
against known DHFR inhibitors: Methotrexate and Trimethoprim.
Methods: Sequencing of the PCR amplified dfrA gene from these DNA isolates revealed a
Molecular docking
Results: An increased affinity for inhibitor Trimethoprim and native substrate dihydrofo-
Enzyme assay
late was observed with the mutant. Methotrexate did not vary in its activity with both
Trimethoprim
the enzyme forms. Conclusions: The Glu84Gly point mutation may lead to a variation in the strain which may cause resistance in the future. Ó 2015 Asian African Society for Mycobacteriology. Production and hosting by Elsevier Ltd. All rights reserved.
Introduction Biosynthesis of a variety of cellular components involves the presence of reduced folates in both prokaryotes and eukaryotes. The normal mechanism of synthesis involves de novo biosynthesis along with salvage pathways in some bacteria and mammalian cells. Dihydrofolate reductase (DHFR)
(encoded by dfrA gene) has an essential role in the maintenance of cellular pools of reduced folate, tetrahydrofolate (THF) and its derivatives, including methylation of dUMP to dTMP, and in cell growth and proliferation. The enzyme is hence a target for several anti-bacterial and anti-cancer agents [1,2]. Thymidylate synthase (TS) is the enzyme which functions downstream of DHFR in the folate pathway. Two
* Corresponding author. E-mail addresses: [email protected], [email protected] (M.S. Degani). Peer review under responsibility of Asian African Society for Mycobacteriology. http://dx.doi.org/10.1016/j.ijmyco.2015.02.001 2212-5531/Ó 2015 Asian African Society for Mycobacteriology. Production and hosting by Elsevier Ltd. All rights reserved.
98
International Journal of Mycobacteriology
forms of TS enzyme coexist in Mycobacterium tuberculosis (Mtb), namely ThyA and ThyX [3]. ThyA is involved in de novo synthesis, while ThyX functions as a Pyrimidine Salvage Pathway enzyme. The conversion of dUMP to dTMP, carried out by both forms of thymidylate synthase, requires the presence of tetrahydrofolate, which is formed due to the activity of DHFR enzymes. Thus, an inhibited DHFR enzyme would result in the accumulation of dUMP, which would get converted to dUTP, and the incorporation of dUTP into DNA results in cell death [4]. Mtb, a human pathogen, claims more lives than any other known infectious organism around the world [5,6]. The current therapy involves a four-drug regime [7–9]. However, due to the emergence of multi-drug resistant (MDR) forms of the organism, the current treatment proves inefficient [10]. A sequence alignment of human and Mtb DHFRs indicates only 26% sequence identity [11]. However, mutations in this target may disrupt the inherent binding ability of this enzyme to known inhibitors. Hence, the present study was undertaken to investigate 20 clinical isolates of first-line drug-resistant tuberculosis (TB) for inherent mutations in the dfrA gene. The studies led to the identification of a point mutation replacing Adenine by Guanine at the 251st position in the dfrA gene (480 bp) of one of the 20 screened isolates. This leads to an amino acid change of glutamic acid to glycine at the 84th position of the enzyme. Docking software was employed to locate the position of this alteration to the surface of the protein. Thus, it was assumed that the modification would not alter the interaction directly at the active site, but might allosterically or structurally affect it. This was in accordance with the fact that DHFR is an essential enzyme, and the organism might not mutate in a manner to affect its inherent function in order to maintain its own viability. The mutated gene was cloned and expressed using pET21a vector. The expressed protein was assessed using SDS–PAGE and Km determination for native substrate, dihydrofolate (DHF). The mutant was also screened against known inhibitors of DHFR, Trimethoprim (TMP) and Methotrexate (MTX), by both in silico docking studies and in vitro enzyme assay. The results obtained indicate better binding of the mutant enzyme with the substrate, DHF and inhibitor, TMP when compared with the wild type.
Materials and methods Chemicals and standard drugs DHF was purchased from Sigma Aldrich. MTX was obtained as a gift sample from Khandelwal Laboratories Pvt. Ltd., Mumbai, India. TMP, buffer salts, reduced nicotinamide adenine dinucleotide phosphate (NADPH) and Dithiothreitol (DTT) were purchased from HiMedia.
Amplification of dfrA gene from 20 clinical isolates The chromosomal DNA of 20 clinical isolates was procured from RMC-BARC, Mumbai, India. The primary source of these strains was Sewri Tuberculosis Hospital, Mumbai, India. The genomic DNA was extracted from all the strains by the standard cetyl trimethyl ammonium bromide (CTAB) method [12].
4 ( 2 0 1 5 ) 9 7 –1 0 3
These strains were selected based on their resistance against first-line drugs identified by employing the recommended WHO protocol [13]. Out of these isolated strains, 3 were resistant to Isoniazid, 12 to Rifampicin, 3 others to Isoniazid and Rifampicin, and 2 were resistant to Isoniazid, Rifampicin and Streptomycin. The drug susceptibility was performed by the standard proportion method as described in Standard Operating Protocol (SOP) prepared in the Revised National TB Control Programme Training Manual for ‘‘Mycobacterium tuberculosis Culture & Drug susceptibility testing’’. The document was prepared by the Central TB Division, Directorate General of Health Services, Ministry of Health and Family Welfare Government of India. Along with these DNA samples, the wild type Mtb H37Rv DNA sample was also procured as a standard. The complete dfrA gene (480 bp) was amplified from these isolates and standard using the forward primer 5P-GAC GCG TGTGCA TAT GGT GGG GCT GAT CTG-3P and reverse primer 5P-GCG ATG AGG ATC CGC GGC GCT CATGAG CGG-3P, containing NdeI and BamHI restriction sites, respectively [14]. The amplified fragments were checked on agarose gel and sequenced by employing dideoxy DNA sequencing method of Sanger et al. using ABI Prism 377-18 automated DNA sequencer and the Big Dye terminator kit (ABI, Foster City, CA). A point mutation was found in one of the 20 isolates.
In silico mutation analysis The ligands (Methotrexate and Trimethoprim) were sketched using the 2D-sketcher of Maestro (GLIDE, Schro¨dinger 9.0, LLC, NY). Low-energy 3D structures of ligands were generated with the help of Ligprep, by adding appropriate hydrogens, generating possible ionization states and tautomers. The ligands were subsequently subjected to energy minimization using the OPLS-2005 force field. The protein structure PDB code: 1DG8 which is the complex of Mtb DHFR enzyme and NADPH with resolution of 2.00 was retrieved from Brookhaven Protein Data Bank (www.rcsb.com) and used for validation of the docking algorithm. All water molecules were deleted, bond orders and charges of cofactor NADPH were assigned properly in the protein preparation step. The amino acid at the 84th position was altered to generate the mutant enzyme. The docking study was carried out using GLIDE software in both standard precision (SP) and extra precision (XP) mode with ‘must match’ of at least one hydrogen bond criteria; with Ile5, Asp27 and Ile94 (10) using both the wild type and the mutant protein structure.
Cloning, expression and purification of the mutant identified The mutant amplified fragment was cleaned using the QIAquick PCR purification kit (Qiagen). Both the amplified dfrA gene and pET21a expression vector (Novagen) were then digested with NdeI and BamHI, purified on a 1.0% agarose gel, eluted, and ligated together using T4 DNA ligase overnight at room temperature (20 °C). The resulting construct was sequenced to confirm that the dfrA gene was in the proper orientation for expression and that no other mutations were introduced during the PCR amplification. After propagation in Escherichia coli strain DH5a, the construct was purified and transformed into E. coli strain BL21 STAR (Invitrogen) for expression.
International Journal of Mycobacteriology
The DHFR protein was over-expressed under the pET21a T7 promoter in recombinant E. coli BL21 cells. Briefly, the recombinant cells were grown in Luria-Bertani (LB) medium (HiMedia) at 37 °C until optical density (OD600) reached 0.4– 0.5. Induction was performed by the addition of 1 mM isopropyl-L-thiogalactopyranoside (IPTG), and cells were grown for two and half hours more at 37 °C to reach an OD of 1.5. Cells were centrifuged, washed with cold phosphate-buffered saline (PBS) and the pellet frozen at 80 °C. The pellet was thawed and used for enzyme purification using Ni-IDA (Nickel-Iminodiacetic acid) column comprising of Sephabeads IP-IDA resin and urea and imidazole as eluents. The purified enzyme, obtained from the pellet after column chromatography, was analyzed using SDS–PAGE and characterized by Km determination for substrate dihydrofolate (DHF).
IC50 determination using enzyme assay The mutant DHFR activity was determined spectrophotometrically against known ligands MTX and TMP by monitoring the decreasing absorbance for 2 min at 340 nm and 37 °C. Standard assay mixture contained 50 mM KPO4 (pH 7.4), 5 mM DTT, and 60 lM NADPH, and the reaction was started after a 3 min incubation by the addition of 45 lM DHF [14,15]. The wild type enzyme obtained after optimized purification from recombinant yeast [16–18] was used as the control DHFR for the studies. The wild type and mutant enzymes were characterized by determining Km values for the substrate DHF using MS Excel graph plots. The inhibitors MTX and TMP were used in the reaction mixtures in 5 different concentrations. The assay was carried out in microtiter plates in duplicates in the assay and analyzed using the kinetic mode of the micro plate reader from Synergy H1 BioTek with Gen5 software. The Inhibitory concentration 50 (IC50) values were determined from graph plots generated by MS Excel and standard deviation was calculated.
4 (2 0 15 ) 9 7–10 3
99
Results The amplified dfrA gene of the 20 clinical isolates and the standard Mtb H37Rv were analyzed on agarose gel where 480 bp sequences were visualized. These amplified genes were sequenced and the sequencing revealed a point mutation in isolate number 19 which was absent in the wild type H37Rv strain when analyzed using Serial cloner 2.5 (alignment tool, sequence aligned with Mtb DHFR with GenBank accession number NC_000962). A transition from Adenine to Guanine at position 251 in the dfrA gene of an isolate, which was resistant to Isoniazid, Rifampicin and Streptomycin (MIC at concentrations of 0.2 lg/ml for Isoniazid, 40 lg/ml for Rifampicin and 4 lg/ml for Streptomycin), led to the identification of an alteration of amino acid at the 84th position (glutamic acid to glycine) which was found to result in a surface modification (slightly unaligned b sheet from amino acid 79 to 85) through in silico simulation studies (Fig. 1). This might be the result of a change in surface charge from negative to neutral. Hence the interaction with the nearby side chain is minimized thereby resulting in distancing of the entire b sheet. As a result of the in silico studies, it was revealed that the known inhibitors MTX and TMP interacted with the wild type and the mutant differently (Table 1-Docking scores). MTX showed complete alignment in the docked images of wild type as well as mutant since the slight alignment variation did not cause any contact changes within the active site (Fig. 2). However, TMP showed altered and, curiously, a better docking score with the mutant than the wild type which is supported by the change in alignment of TMP in the active site of mutant as depicted in Fig. 3. Also, any alteration in the Mtb DHFR surface geometry has been reported to affect the final binding of TMP in the active site as reported by Nerukh et al. [19]. Hence, this surface modification could have led to the differential binding of TMP in the active site pocket of the mutant enzyme.
Figure 1 – Overlap of wild type (pink) and mutant (green) DHFR enzyme depicting area of surface modification around 84th amino acid residue.
100
International Journal of Mycobacteriology
4 ( 2 0 1 5 ) 9 7 –1 0 3
Table 1 – In silico Docking scores and in vitro IC50 values of wild type and mutant Mtb DHFR with MTX and TMP. Ligand
Wild type DHFR Docking Score
MTX TMP
9.38 4.43
Mutant DHFR IC50 (lM)
Docking Score
0.008 ± 0.0007 16.0 ± 3.0
9.38 6.60
IC50 (lM) 0.008 ± 0.0004 10.5 ± 2.0
˚ surface generated for wild type (pink) and (b) Figure 2 – Interaction diagram of DHFR enzyme with MTX (orange) (a) with 8 A ˚ ) (d) Ligplot of MTX with mutant (6 A ˚ ). with mutant (green) (c) Ligplot of MTX with wild type (6 A
In order to establish the validity of the docking, a construct with the mutant gene insert was prepared and purified. The new construct (pET21a-MtbDHFR#19) was transformed in E. coli BL21 and later used for protein expression. Purification of the Mtb DHFR using the Ni-IDA column was confirmed using the SDS–PAGE which revealed a 19 kDa protein band. Formation of inclusion bodies in the E. coli host leads to enzyme loss. Since this issue was not encountered, the purified enzyme was considered for further studies. The wild type enzyme was purified from recombinant yeast. A comparative study of the two enzymes was undertaken since both were obtained in pure form.
The enzymes were characterized by Km determination and assessed for activity against the ligands MTX and TMP. The Km value of the mutant enzyme (4.4 ± 0.02 lM) is slightly less than the Km value for the wild type enzyme (4.6 ± 0.05 lM) indicating that the mutant has a slightly higher affinity towards the substrate DHF. The comparative wild type and mutant enzyme IC50 values are as shown in Table 1. When the IC50 values were compared with the reported IC50 of the wild type [15], the mutant shows a visibly better interaction with TMP as indicated by the docking studies thus validating the in silico obtained results.
International Journal of Mycobacteriology
4 (2 0 15 ) 9 7–10 3
101
˚ surface generated for wild type (pink) and (b) with Figure 3 – Interaction diagram of DHFR enzyme with TMP (blue) (a) with 8 A ˚ ) (d) Ligplot of TMP with mutant (6 A ˚ ). mutant (green) depicting change in alignment (c) Ligplot of TMP with wild type (6 A
Discussion The current work aimed at checking for mutation in the dfrA gene from Mtb and analyzing the effect of the mutant enzyme on inhibitor binding. Thus, the DNA from 20 clinical isolates of Mtb showing variable resistance to first-line drugs were used for amplification of the dfrA gene, and one DNA isolate revealed a point mutation when compared with the wild type Mtb H37Rv at the 84th position from glutamic acid to glycine, which led to a surface modification in the 3-D structure of the protein when analyzed using in silico methods. Similar studies carried out by Nosova et al., with other enzymes of Mtb, have indicated the association of these enzyme mutations, with resistance to second-line drugs [20]. Resistance to known drugs was also found to be associated with the DHFR enzyme mutations in other organisms like Plasmodium vivax [21]. Since the strain with the mutation was found to be resistant to the first-line drugs Isoniazid, Rifampicin and Streptomycin, the studies were carried forward to in vitro stages. The mutant protein was cloned, expressed and studied by Km determination and for inhibition due to known drugs by
employing an enzyme assay. On comparison with the wild type, it revealed a peculiarly higher affinity towards the native substrate DHF and inhibitor TMP. The data with TMP showing variation co-related with the in silico docking studies. This data can be justified since a change in the surface geometry of DHFR has been reported to alter the final binding of TMP with DHFR [19]. In this experimentation, the inhibitory concentration of TMP has shown an improvement, indicating that the mutation has no detrimental effect on inhibitor binding as was expected, but makes the enzyme more susceptible to inhibition. However, it has been reported that in a number of pathogens, resistance towards DHFR inhibitors develops by the simple accumulation of point mutations in the dfrA gene [22,23]. Further, the mutant enzyme also showed an increased affinity towards the native substrate, DHF. These increased interactions may be due to the two conformational states showing varying affinities towards anti-folate drugs that have been proposed for Mtb DHFR [24].
102
International Journal of Mycobacteriology
Conclusions Point mutations in the target genes have been reported as the principal mechanism of development of drug resistance in Mtb [25,26]. Also, mutations in dfrA have been reported to play a role in Mtb resistance development [27]. Additionally, Isoniazid (INH) is known to inhibit Mtb DHFR in vitro [28], and dfrA mutations have also been reported to contribute to the detection of Isoniazid-resistant Mtb isolates along with other mutations [29,30]. The Glu84Gly (E84G) mutation in DHFR conferring increased affinity towards Trimethoprim in Mtb or other organisms has been reported for the first time in this paper. Considering the above-mentioned literature and the fact that the said mutant strain is resistant to Isoniazid, this point mutation may be proposed to lead to a variation in the strain which might contribute towards resistance in the future. The in vitro analysis of this particular mutant in the near future would affirm this assumption.
Conflict of interest There are no conflicts of interest.
Acknowledgements Authors are thankful to the Indian Council of Medical Research, Delhi, India (PHA-BMS-45/94/2012) for financial assistance. The authors would like to thank Dr. Carol H. Sibley for the recombinant yeast containing wild type Mtb DHFR. R E F E R E N C E S
[1] G.B. Elion, Nobel lecture, the purine path to chemotherapy, Biosci. Rep. 9 (1989) 509–529. [2] G.H. Hitchings, Nobel lecture in physiology or medicine, selective inhibitors of dihydrofolate reductase, In Vitro Cell Dev. Biol. 25 (1989) 303–310. [3] J. Rengarajan, C.M. Sassetti, V. Naroditskaya, A. Sloutsky, B.R. Bloom, E.J. Rubin, The folate pathway is a target for resistance to the drug para-aminosalicylic acid (PAS) in Mycobacteria, Mol. Microbiol. 53 (2004) 275–282. [4] A.D. Villela, Z.A. Sanchez-Quitian, R.G. Ducati, D.S. Santos, L.A. Basso, Pyrimidine salvage pathway in Mycobacterium tuberculosis, Curr. Med. Chem. 18 (2011) 1286–1298. [5] P.J. Dolin, M.C. Raviglione, A. Kochi, Global tuberculosis incidence and mortality during 1990–2000, Bull. World Health Organ. 72 (1994) 213–220. [6] A.M. Rouhi, Tuberculosis: a tough adversary, Chem. Eng. News 77 (1999) 52–70. [7] B.R. Bloom, C.J. Murray, Tuberculosis: commentary on a reemergent killer, Science 257 (1992) 1055–1064. [8] D.E. Snider, W.L. Roper, The new tuberculosis, N. Engl. J. Med. 326 (1992) 703–705. [9] J.B. Bass, L.S. Farer, P.C. Hopewell, R. O’Brien, R.F. Jacobs, F. Ruben, et al, Treatment of tuberculosis and tuberculosis infection in adults and children, Am. J. Respir. Crit. Care Med. 149 (1994) 1359–1374. [10] H. Nakajima, Tuberculosis: A Global Emergency, World Health Organization, 1993.
4 ( 2 0 1 5 ) 9 7 –1 0 3
[11] R. Li, R. Sirawaraporn, P. Chitnumsub, W. Sirawaraporn, J. Wooden, F. Athappilly, et al, Three-dimensional structure of M. tuberculosis dihydrofolate reductase reveals opportunities for the design of novel tuberculosis drugs, J. Mol. Biol. 295 (2000) 307–323. [12] S. Honore-Bouakline, J.P. Vincensini, V. Giacuzzo, P.H. Lagrange, J.L. Herrmann, Rapid diagnosis of extrapulmonary tuberculosis by PCR: impact of sample preparation and DNA extraction, J. Clin. Microbiol. 41 (2003) 2323–2329. [13] P. Chakraborty, S. Kulkarni, R. Rajan, K. Sainis, Drug resistant clinical isolates of Mycobacterium tuberculosis from different genotypes exhibit differential host responses in THP-1 cells, PLoS One 8 (2013) e62966. [14] E.L. White, L.J. Ross, A. Cunningham, V. Escuyer, Cloning, expression, and characterization of Mycobacterium tuberculosis dihydrofolate reductase, FEMS Microbiol. Lett. 232 (2004) 101–105. [15] B.L. Hillcoat, P.F. Nixon, R.L. Blakley, Effect of substrate decomposition on the spectrophotometric assay of dihydrofolate reductase, Anal. Biochem. 21 (1967) 178–189. [16] A.B. Gerum, J.E. Ulmer, D.P. Jacobus, N.P. Jensen, D.R. Sherman, C.H. Sibley, Novel Saccharomyces cerevisiae screen identifies WR99210 analogues that inhibit Mycobacterium tuberculosis dihydrofolate reductase, Antimicrob. Agents Chemother. 46 (2002) 3362–3369. [17] A. Raju, M.S. Degani, M.A. Khedkar, S.N. Niphadkar, Novel affinity chromatographic technique for purification of dihydrofolate reductase from recombinant yeast, Int. J. Drug Des. 4 (2013) 994–997. [18] A. Raju, M.A. Khedkar, M.S. Degani, Optimization of Mycobacterium tuberculosis DHFR production from recombinant Saccharomyces cerevisiae, Int. J. Curr. Microbiol. Appl. Sci. 2 (2013) 70–79. [19] D. Nerukh, N. Okimoto, A. Suenaga, M. Taiji, Ligand diffusion on protein surface observed in molecular dynamics simulation, J. Phys. Chem. Lett. 3 (2012) 3476–3479. [20] E.Y. Nosova, A.A. Bukatina, Y.D. Isaeva, M.V. Makarova, Y. Galkina, A.M. Moroz, Analysis of mutations in the gyrA and gyrB genes and their association with the resistance of Mycobacterium tuberculosis to levofloxacin, moxifloxacin and gatifloxacin, J. Med. Microbiol. 62 (2013) 108–113. [21] W.J. Lee, H.H. Kim, Y.K. Choi, K.M. Choi, M.A. Kim, J.Y. Kim, et al, Analysis of the dihydrofolate reductase-thymidylate synthase gene sequences in Plasmodium vivax field isolates that failed chloroquine treatment, Malar. J. 9 (2010) 331–337. [22] J.E. Hyde, The dihydrofolate reductase-thymidylate synthase gene in the drug resistance of malaria parasites, Pharmacol. Ther. 48 (1990) 45–59. [23] B.I. Schweitzer, A.P. Dicker, J.R. Bertino, Dihydrofolate reductase as a therapeutic target, FASEB J. 4 (1990) 2441–2452. [24] M.V. Dias, P. Tyrakis, R.R. Domingues, A.F. Paes Leme, T.L. Blundell, Mycobacterium tuberculosis dihydrofolate reductase reveals two conformational states and a possible low affinity mechanism to antifolate drugs, Structure 22 (2014) 94–103. [25] S.T. Cole, A. Telenti, Drug resistance in Mycobacterium tuberculosis, Eur. Respir. J. Suppl. 20 (1995) 701–713. [26] J.M. Musser, Antimicrobial agent resistance in mycobacteria: molecular genetic insights, Clin. Microbiol. Rev. 8 (1995) 496– 514. [27] C.U. Koser, R.N. Veerapen-Pierce, D.K. Summers, J.A. Archer, Role of mutations in dihydrofolate reductase dfrA (Rv2763c) and thymidylate synthase thyA (Rv2764c) in Mycobacterium tuberculosis drug resistance, Antimicrob. Agents Chemother. 54 (2010) 4522–4523. [28] A. Argyrou, M.W. Vetting, B. Aladegbami, J.S. Blanchard, Mycobacterium tuberculosis dihydrofolate reductase is a target for isoniazid, Nat. Struct. Mol. Biol. 13 (2006) 408–413.
International Journal of Mycobacteriology
[29] Y.M. Ho, Y.J. Sun, S.Y. Wong, A.S.G. Lee, Contribution of dfrA and inhA mutations to the detection of isoniazid-resistant Mycobacterium tuberculosis isolates, Antimicrob. Agents Chemother. 53 (2009) 4010–4012. [30] T. Jagielski, Z. Bakula, K. Roeske, M. Kaminski, A. Napiorkowska, E. Augustynowicz-Kopec, et al, Detection of
4 (2 0 15 ) 9 7–10 3
103
mutations associated with isoniazid resistance in multidrugresistant Mycobacterium tuberculosis clinical isolates, J. Antimicrob. Chemother. (2014), http://dx.doi.org/10.1093/jac/ dku161.