Mutation of katG in a clinical isolate of Mycobacterium tuberculosis : effects on catalase-peroxidase for isoniazid activation

1Biochemistry research Division, Department of Chemistry, Faculty of Sciences and technology, airlangga University; Surabaya, Indonesia; e-mail: purkan@fst.unair.ac.id; 2Biochemistry research Division, Faculty of Mathematics and Natural Sciences, Bandung Institute of technology, Bandung, Indonesia; 3School of Pharmacy, Bandung Institute of technology, Bandung, Indonesia; 4Department of Chemical engineering, Institut teknologi Sepuluh Nopember, Surabaya, Indonesia; e-mail: heriseptyakusuma@gmail.com

Mutations in katg gene are often associated with isoniazid (INh) resistance in Mycobacterium tuberculosis strain.this research was perfomed to identify the katg mutation in clinical isolate (l8) that is resistant to INH at 1 μg/ml.In addition to characterize the catalase-peroxidase of KatG L8 and perform the ab initio structural study of the protein to get a more complete understanding in drug activation and the resistance mechanism.The katG gene was cloned and expressed in Escherichia coli, then followed by characterization of catalase-peroxidase of KatG.The structure modelling was performed to know a basis of alterations in enzyme activity.a substitution of a713g that correspond to asn238Ser replacement was found in the l8 katG.The Asn238Ser modification leads to a decline in the activity of catalase-peroxidase and INH oxidation of the L8 KatG protein.The catalytic efficiency (K cat /K M ) of mutant katg asn238Ser respectively decreases to 41 and 52% for catalase and peroxidase.The mutant KatG asn238Ser also shows a decrease of 62% in INH oxidation if compared to a wild type katg (katgwt).the mutant asn238Ser might cause instability in the substrate binding site of katg, because of removal of a salt bridge connecting the amine group of asn238 to the carbo xyl group of Glu233, which presents in KatGwt.The lost of the salt bridge in the substrate binding site in mutant katg asn238Ser created changes unfavorable for enzyme activities, which in turn emerge as INh resistance in the l8 isolate of M. tuberculosis.

K e y w o r d s: Mycobacterium tuberculosis, INH resistance, katG, catalase-peroxidase.
T uberculosis (TB) is well known as an infec tious disease that is caused by Mycobacterium tuberculosis.The disease is currently ranked as the seventh most common causes of death in the world, and still estimated to remain in the top 10 causes of death untill 2030.Recently, TB is re ported as the second leading cause of death in adults and recorded as the deadliest of all infectious disea ses.Indonesia is classified as a country with the large number of TB cases and occupies the fifth rank of the 22 countries with high potential of TB.In the country, there are about 500,000 new cases of TB annually and 175,000 of them are deaths [1,2].Two percent of new cases and twelve percent of the recur ring cases of TB found in Indonesia are the multid rugresistant (MDR) cases [1].A better understan ding in the antituberculous drug resistance is needed to make easy in the TB therapy.
Isoniazid (isonicotinic acid hydrazide; INH) has commonly been recommended by the World Health Organization (WHO) to treat tuberculosis since 1952 because the drug has a high bactericidal effect and a low price [3].The bactericidal effect of INH as TB drug depends on catalaseperoxidase of M. tuberculosis which is encoded by katg gene.The enzyme plays a role to convert isoniazid absorbed by the M. tuberculosis to be in active form, an isoni cotinoyl acyl radical, to trigger the death of myco bacteria.This might occur because an isonicotinoyl acyl radical is a potential inhibitor for enoylacyl doi: https://doi.org/10.15407/ubj88.05.071 carrier protein reductase (InhA) and β-ketoacyl-acyl carrier protein synthase (KasA), the two key en zymes for the biosynthesis of mycolic acids, a cell wall component of mycobacteria [4,5].
KatG mediates the sensitivity of M. tuberculosis to INH.The katg deficiency strain of M. tuberculosis, which is resistant to INH, can restore the sensitivity to INH when it is complemented with a functional katg [6].Meanwhile, total deletion of katg gene in M. tuberculosis raised resistance to INH [68].Nevertheless, the loss of catalaseperoxi dase in M. tuberculosis has not yet explained com pletely the mechanism of INH resistance, because the total deletion of katg is rarely found in clinical isolates [4,9].
It has been reported that 5070% of INHre sistant M. tuberculosis strains are associated with the mutation in the katg gene [3,4].The mutation types are very diverse, with missense mutations being the most common alteration.The mutations in katg also reveal unique types in INH resistant strain from different geographical areas.The frequency of mutation types in katg is often only associated with drug resistance and rarely linked directly to the change of INH sensitivity or to the effect of the en zymatic activity of KatG.The katg mutations that affect cata lase-peroxidase activity have been found in either all part of the gene and encode the Nter minal or the Cterminal part of the protein [10,5].Many variants of KatG associated with INH resist ance exhibits a decrease in catalaseperoxidase ac tivities.The decreasing scale of the activity among the mutants of katg does not directly correlate with INH resistance [11,12].This is the basis of the argu ment that the INH resistance in clinical isolates is not linked directly to the ability of variants of KatG in INH activation.
The mutant KatG Ser315Thr, that is commonly found in clinical isolates and associated with INH resistance, decreases the activities of catalaseper oxidase and INH oxidation [13,14].The amino acid Ser315 in KatG is closely put in the active environ ment.So that, the genetic modifications in this part are easily understood as the producer of importantly affected enzymatic function and thus isoniazid re sistance.As many as 50% variants of katg as sociated with INH resistance are not modified in Ser315Thr.Biochemical analysis of variants of katg other than Ser315Thr is necessary to examine the re lation between INH resistance with the changes in the function and structure of the mutants.Some clinical isolates of INHresistant M. tuberculosis from Indonesian TB patients are not mu tated at codon 315 of katg.Among these isolates, an isolate, namely L8 isolate, shows resistance to INH at 1 μg/ml.This paper shows the correlation of katg mutation in the isolate with the biochemical proper ties of catalaseperoxidase, especially for isoniazid oxidation.This paper also examines the ab initio structural study of the mutant KatG L8 with the aim of gaining a more complete understanding of drug activation and the resistance mechanism.

Materials and Methods
Bacterial strains and plasmids.A clinical iso late of M. tuberculosis was obtained from sputum of a TB patient in Bandung, Indonesia.escherichia coli TOP10 (Invitrogen, Carlsbad, CA) was used as hosts for cloning of katg of INH resistant and sensitive M. tuberculosis.The e. coli BL21 (De3) (Promega, Madison, USA) was employed as an expression host of KatG.The plasmid pGEM ® T (Invitrogen) and pCold IIDNA (Kinki University) were used as a cloning and expression vector, respectively.
preparation of chromosomal dna.Chro mosomal DNAs of M. tuberculosis of wild type and clinical isolates were prepared by an alkaline lysis method in 5 mM Tris-Cl buffer (pH 8.5) contai ning 0.5% (b/v) Tween20 and 0.2 mg/ml proteinase K at 50 °C for 60 min.The mixture then was heated at 95 °C for 5 min.Cellular debris was collected at 12,000 g for 10 min and the supernatant containing chromosomal DNA of M. tuberculosis was used for PCR [15].
Amplification of katG gene.The full length of the katg gene was amplified by the polymerase chain reaction (PCR) technique using FG and RG primers (Table 1).PCR was performed in a total reac tion volu me of 50 μl containing 50 ng cromosomal DNA; 1xPCR buffer (20 mM Tris-HCl pH 8.4, 50 mM KCl); 0.1 mM of each primer, 200 μM dNTPs ; 1.5 mM MgCl 2 ; and 0.25 unit of taq DNA polymerase (Amersham).The amplifica tion process was done by a GeneAmp® PCR System 2700 (PerkinElmer), and set at the following steps: predenaturation at 94 °C for 4 min, 25 cycles of de naturation at 94 °C for 1 min, annealing at 57 °C for 1 min, and an extension at 72 °C for 3 min.The process was terminated by postelongation at 72 °C for 7 min.All PCR products were analyzed using agarose gel electrophoresis and purified by GFX pu rification kit (Amersham).
dna sequencing.The nucleotide of the katg gene in the recombinant plasmid was se quenced by an automatic nucleotide sequencer (ABI PRISM, Macrogen, Seoul, Korea).The recombinant pGEM ® T, namely pGEMTkatg was used for DNA template.All oligonucleotides primers used for the sequencing were presented in Table 1.
alignment analysis.The katg genes of wild type and clinical isolate of M. tuberculosis were ana lyzed in silico by aligning the nucleotide sequenc es of the genes and their deduced amino acid se quences, using the SeqManTMII and MegAlignTM DNASTAR program (Lasergene).The nucleotides of the genes were also compared with katg nucleotides in Genbank (accession number X68081).
subcloning of the katG gene.The katg gene fragment in pGEMTkatg was taken by digestion of plasmid recombinant with NdeII and XbaI.The digestion product was purified and inserted into plasmid pCold II DNA, which previously had been digested with the same restriction enzymes.The ligation product was transformed to e. coli BL21 (De3) and the transformed bacteria were then grown on a selective LB agar plate.

KatG gene expression.
A single colony of e. coli BL21 (DE3) containing recombinant plasmid (pCold IIkatg) was cultured in LB liquid medium containing ampicillin 100 μg/ml, then followed by shaking at 37 °C to obtain an optical density (OD) at λ 600 nm approximately of 0.4-0.5.The culture was then immediately cooled at 15 °C for 30 minu tes without shaking.To induce the expression of re combinant protein, the culture was added by 0.1 mM IPTG, and followed by shaking at 15 °C for 24 h.The culture was then centrifugatedat 5,000 g for 10 min to harvest the cells.The resulted cells were washed and resuspended in 50 mM potassium phos phate buffer (pH 7.0), and then disrupted by sonica tion.The cellular debris was removed by centrifuga tion at 12,000 g for 15 min.The KatG protein that remained in the supernatant was then purified.
Protein purification.The recombinant protein of KatG was purified by affinity chromatography using HisTrapTMHP column (GE Healthcare, Frei burg, Germany) containing Nisepharose matrix.The purification steps were run according to the manufacturer's protocol.The recombinant protein was eluted gradually by 50 mM K-phosphate buffer, pH 7.0, containing imidazole of 50200 mM.The purified protein was analyzed by a sodium dodecyl sulphatepolyacrylamide gel electrophoresis (SDS PAGE).
catalase-peroxidase activity and inH oxidation assays.The recombinant KatG was assayed re garding its activity on catalaseperoxidase and INH oxidation.Catalase activity was determined by the Patti and BonetMaury method, based on the formed color from the reaction of titanium with H 2 O 2 [17].The reaction was carried out in 10 mM Kphosphate buffer pH 7.0 with a total volu me of 1 ml containing 12.5 mM H 2 O 2 substrate and KatG protein.The enzymatic reaction was stopped by the addition of 2.5 ml of titanium reagent, in turn the formed yel low color was observed at λ 410 nm [17].One unit of cata lase activity was defined as the amount of en zyme decomposing 1 mmol of H 2 O 2 per min.
Peroxidase activity was determined by reacting of 100 μM o-dianisidine in 100 ml of 50 mM potas sium buffer (pH 4.5) containing 25 mM tertbutyl hydroperoxide (tBHP) with 12.5 mM H 2 O 2 .The absorbance of the reaction product, i.e. odiani sidin quinone diimine was observed at λ 460 nm (ε 460 = 11.86 mM 1 •cm 1 ) [13].One unit of peroxidase activi ty was defined as the amount of enzyme that catalyzed the formation of one μmol product per min at 30 °C.
The activity of katg on INH oxidation was carried out by the Shoeb method [18] of 0.65 g KatG protein in 10 mM phosphate buffer pH 7.0 was reacted with 13.1 mM phenol and 6.25 mM H 2 O 2 in a total volume of 3 ml at 37 °C for 7 min, then followed by addition of 2.4 mM INH.The reaction was continued at 37 °C for 1 h [18].The absorbance of the reaction product , i.e. benzo-quinone was monitored at λ 444 nm (ε 444 = 0.393 M 1 •cm 1 ).One unit of the activity was defined as the amount of enzyme that catalyzes the formation of benzoquinone product per min at 37 °C.structure alignment.Threedimensional structure of the mutant KatG protein was generated by SWISSMODEL, an automated protein homology modeling server, using the known crystal structure of a wild type KatG protein structure (PDB code 1SJ2) as a template.The minimization of the struc tural model was conducted with Amber 10 [19].The modeled structure was visualized using PyMOL 1.3 (www.pymol.org).Root mean square deviation of the model was compared to the 1SJ2 structure by Super Pose version 10.

results and discussion
cloning of katG gene from inH-resistant M. tuberculosis strain.The open reading frame (ORF) of katg gene from L8 Isolate of INHresistant M. tuberculosis strain was amplified by PCR using FG and RG primers.The 2.2 kb fragment of PCR product then was inserted in pGEMT vector (3.0 kb) to construct pGEMTkatg recombinant.The recom binant plasmid exhibited two fragments (3.0 and 2.2 kb) then it was digested by both ecoRI and NotI restrction enzymes (Fig. 1).The 3.0 kb fragment cor responded to pGemT vector, and the 2.2 kb fragment had the size of the katg gene.The 2.2 kb fragment of L8 isolate carried a mutation, the guanine instead of adenine at position 713 (Fig. 3).
The katg gene was then subcloned to pColdII DNA as expression vector.Insertion of katg frag ment (2.2 kb) in vector pCold IIDNA (4.3 kb) yielded a pCold IIkatg recombinant (6.5 kb).Sin gle digestion of the recombinant with NdeI and XbaI restriction enzymes yielded a DNA fragment (6.5 kb), respectively, while the digestion with both the enzymes yielded two fragments consisting of 2.2 and 4.3 kb (Fig. 2).The nucleotides sequences of the katg gene from the L8 M. tuberculosis isolate showed a substitution of adenine to guanine at posi tion 713 compared to the katg from INH suscepti ble M. tuberculosis, the H37Rv strain (Fig. 3).The mutation altered the amino acid of KatG protein, the serine instead of asparagine at posistion 238 through in silico translation (Fig. 3).
the KatG expression.To express the katg, the e. coli BL21 (De3) that bring pColdkatg re combinant was grown in LB medium containing am phicilin, then followed the induction of the culture with IPTG.After the cell of e. coli was collected by centrifugation, the cell pellets were lysed to release its extract protein.Analysis of the extract protein in SDS PAGE showed a high intensity of protein band with a molecular mass of 80 kDa that belonged to KatG protein (Fig. 4).
catalase-peroxidase activities of mutant KatG asn238ser .The kinetic properties of catalase peroxidase of mutant KatG Asn238Ser and wild type had been determined, i.e. k M , k cat and k cat /k M .The mutant KatG Asn238Ser exhibited the k cat /k M value for catalase and peroxidase lower than KatGwt.The k cat /k M value mean as the catalytic efficiency of en zyme.The high k cat /k M suggested that an enzyme could bind its substrate effectively and decompose its substrate into a product well, and vice versa.KatGwt has k cat /k M value as 8.62×10 4 M 1 •S 1 for catalase and 1.99×10 5 M 1 •S 1 for peroxidase.Mean while KatG Asn238Ser mutants have a low k cat /k M value, i.e. 5.12×10 4 M 1 •S 1 for catalase and 0.96×10 5 M 1 •S 1 for peroxidase (Table 2).In addition to the ki netic proper ties, the mutant KatG Asn238Ser took also a declining activity to oxidize INH.The mutant KatG Asn238Ser activity in oxidation of isoniazid re mained 42.5% from that of KatGwt (Fig. 5).
structural Model of mutant KatG asn238ser .Structure modeling of KatG Asn238Ser was done to know the effect of amino acid substitution in the mu tant protein.Superposition of Cα framework of the mutant model with the KatGwt structure exhibited root mean square deviations (RMSD) of 0.073.This result showed that The mutant KatG Asn238Ser shared similiar structure with KatGwt of M. tuberculosis H37Rv.Moreover, the amino acid substitution of Asn238Ser found in the mutant KatG L8 elimi nate the salt bridge interaction around the substrate binding site, which connects between the amine group in side chain of Asn238 residue with the car boxyl group of Glu233 in KatGwt (Fig. 6).
KatG is the only enzyme in M. tuberculosis that could generate isoniazid susceptibility.There fore it plays a central role in the development of at least one type of isoniazid resistance [20].Mutations in katg are often associated with INH resistance.A substitution of adenine to guanine at position 713 in   Mutation is marked with a red oval to the substrate is lower than the KatGwt, because it has a value of k M for each catalase and peroxidase, 1.4 times higher than KatGwt (Table 2).Substitution Asn238Ser reduced 40% of substrate binding affinity for catalaseperoxidase of the mutant.Compa ring with KatGwt, the mutant KatG Asn238Ser also displayed a decrease of k cat value by 17% for catalase and 32% for peroxidase.This means that the mutant lost 17 and 32% ability of catalase and peroxidase respec titively in the converting of substrate into product.The catalytic efficiency that symbolized by k cat /k M for the mutant KatG Asn238Ser decreased by 41% for cata lase and 52% for peroxidase activity (Table 2).The mutant KatG Asn238Ser decreased in both the binding affinity and the converting activity of sub trate into product.Several papers reported that the catalaseperoxidase activity among variants of katG were not correlated with the resistance level of INH [11,12].This can be triggered due to the measure ment of enzyme activity in vitro, but it is connected directly with INH resistance phenotype which is ac tually derived from the in vivo process.
The INH oxidation test of KatG Asn238Ser showed that the mutant lost 62% of the activity compared with that of KatGwt (Table 2).The Asn238Ser modification in KatG of INH resistant isolate (L8) is connected with the decline in the activity of cata laseperoxidase and INH oxidation of the protein.
Decreasing of catalytic efficiency (k cat /k M ) of mutan KatG Asn238Ser as 41% for catalase and 52% for per oxidase accompanied the reduction of 62% in INH oxidation activity of the mutant.The decline of INH activation in many variants of KatG, i.e.Ala110Val, Asp735Ala, Ala139Pro, Ser315Asn, Leu619Pro, and

Activity
Isolate Kinetic parameters  The failure role of KatG in the INH activa ting has been shown in detail by the mutant KatG Ser315Thr [12,13,27].The Ser315Thr substitution in KatG impacts on the shifting of substrate binding chan nel from 6.0 to 4.7 Å [27].Consequently the mutant failed to bind the INH, and subsequently decreased 160 times in the INH activation compared with KatGwt [14].
By using a structure model, the basis of a de crease in the catalytic efficiency for catalase-per oxidase and INH oxidation activity in the mutant KatG Asn238Ser was described.In the KatGwt structure, the amino acid 238 is put closely to the substrate binding site, i.e.Asn137, Val230, Asn231, Pro232 and Ser315 residues.It is found that Asn231 residue makes hydrogen bond with Glu233 and Van Der Walls interaction with Asn236.These interactions Fig. 6.Illustration of Asn238Ser substitution effect in KatG L8.Superposition of mutant KatG structure (L8) (yellow) to the katgwt (red) (a). the blue rod-shaped amino acids represented the residues for subtrate binding ; green rod-shaped amino acids represented catalytic residues, and magenta are the residue for cross links.In the katgwt, the asn238 stabilized the active site environment through linkage to the glu233 and asn236 (B).In the mutant katg l8, the Ser238 residue could not make interaction with glu233 are stabilized by the presence of salt bridge which connects between the amine group in side chain of Asn238 residue with the carboxyl group of Glu233 in KatGwt (Fig. 6).This salt bridge was lost due to modification of Asn238Ser in the mutant KatG L8.The lost of the salt bridge created unfavorable for enzyme activities, then in turn it emerged the INH resistance in the L8 isolate of M. tuberculosis.The model analysis of KatG Asn238Ser should be further con firmed by a real crystal structure of the mutant.
Conflicts of interest.All authors declare that there is no conflict of interest.

Fig. 3 .
Fig. 3. Nucleotides alignment of the katg of INh resistant M. tuberculosis strain (l8) againts the katg from INH sensitive strain (H37Rv) and GenBank.Comparing with H37Rv and genbank katG, The L8 katG exhibited a varian nucleotide at position 713, guanine instead of adenine, then substituted asn with Ser at position 238.Mutation is marked with a red oval

Fig. 5 .
Fig. 5.The INH oxidation activity of wild type and mutant katg acknowledgements A partial funding of this research was supported by DIPA Secretariat of Research and De velopment Agency, The Ministry of Health, Repub lic of Indonesia, Number: 0682/03411.