COMMUNICATION:
The Role of Mn(II)-Peroxidase Activity of Mycobacterial Catalase-Peroxidase in Activation of the Antibiotic Isoniazid*

(Received for publication, January 3, 1997, and in revised form, February 14, 1997)

Richard S. Magliozzo Dagger and Jovita A. Marcinkeviciene §

From the Departments of Physiology and Biophysics and § Biochemistry, Albert Einstein College of Medicine of Yeshiva University, Bronx, New York 10461

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS AND DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES


ABSTRACT

The catalase-peroxidase of Mycobacteria smegmatis exhibits Mn(II)-peroxidase activity characterized by a low Km for Mn(II) (5 µM) and a high Km for t-butyl hydroperoxide (100 mM). This activity, monitored by the formation of Mn(III)-malate or -malonate, is inhibited by Co(II) but not by superoxide dismutase. Optical evidence for binding of Mn(II) to the resting (ferric) enzyme is found in a change in intensity of the Soret peak upon titration with Mn(II). A potential role for Mn(III) in the antimycobacterial action of the antibiotic isoniazid is suggested by the rapid reduction of Mn(III)-malonate by this drug. The stoichiometry of the reaction is consistent with two single electron transfer steps per mole of isoniazid.


INTRODUCTION

Isoniazid (isonicotinic acid hydrazide) has been used as a first line antibiotic for the treatment of tuberculosis for decades; yet its mechanism of action is only partially understood. The accumulation of evidence that certain isoniazid-resistant strains of mycobacteria have reduced catalase-peroxidase activity (1-3) suggested that this enzyme converts isoniazid into a bacteriocidal agent (4, 5). Susceptibility to isoniazid could be produced in isoniazid insensitive Escherichia coli or Mycobacteria smegmatis (2) upon introduction and expression of the Mycobacteria tuberculosis katG gene encoding catalase-peroxidase. A target of the isoniazid-derived bacteriocidal agent has been identified as an enzyme required for mycolic acid synthesis (inhA-encoded enoyl-ACP reductase), the activity of which is effectively inhibited by activated isoniazid (6-8).

The structure and catalytic function of the purified catalase-peroxidase from M. smegmatis (9) (and other microorganisms (10-13)) has been partially characterized; in its resting state it contains ferric heme according to optical and EPR spectroscopy, and it catalyzes classical peroxidatic reactions utilizing H2O2 or alkyl peroxides. A recent report on the oxidation of isoniazid by catalase-peroxidase showed that the ferric enzyme could be activated by reduction with hydrazine under aerobic conditions (9). That report also contained evidence for an apparent enhancement of the enzymatic rate of isoniazid oxidation when Mn(II) was included in the reaction and showed that manganese alone could induce the aerobic oxidation of the drug to give the same products as those produced by the purified enzyme. Mn(II) has been included in reaction protocols designed to demonstrate the isoniazid-dependent inhibition of the enoyl-ACP reductase by M. tuberculosis and M. smegmatis catalase-peroxidases (6-8), although no definition of the metal ion's role has emerged in the literature.

The data presented here demonstrate that catalase-peroxidase purified from M. smegmatis catalyzes the peroxidation of Mn(II) to Mn(III). This activity, characteristic of the fungal class of manganese-dependent peroxidases like that from the white rot fungus, Phanerochaete chrysosporium (14-16), has also been reported for the lignin peroxidase from the same organism (17).

It has been clearly demonstrated that the activity of catalase-peroxidase is responsible for the efficacy of isoniazid, whereas the importance of manganese in the mechanism of action of this drug in mycobacterial pathogens has not been investigated. The finding that isoniazid in the presence of manganese and air can lead to drug oxidation (9) and inhibition of inhA in the absence of catalase-peroxidase1 raises the interesting possibility that enzymatic Mn(III) production stimulates the activation of isoniazid and potentiates intracellular damage by the drug even in the absence of oxygen required for the nonenzymatic reaction.

Experimental results are presented in this communication for the formation of Mn(III) from Mn(II) as a function of enzyme concentration, Mn(II) concentration, and the concentration of t-butyl hydroperoxide. Other results demonstrate the inhibition of Mn(II) peroxidation by Co(II) and the stoichiometry of the reaction between Mn(III) and isoniazid. Homologies between the amino acid sequences of mycobacterial catalase-peroxidases and fungal manganese peroxidase are presented in the context of identifying potential Mn(II) ligands in the bacterial enzyme.


EXPERIMENTAL PROCEDURES

Catalase-peroxidase was purified from M. smegmatis according to a published procedure (9). The purified protein had an optical absorbance ratio greater than 0.51 = A408/A280. Initial rates of Mn(II) oxidation were calculated from the increase in the ultraviolet absorbance due to the formation of Mn(III)-malate or Mn(III)-malonate according to a published method (16). Hydrogen peroxide and ethyl hydroperoxide, as well as t-butyl hydroperoxide, supported Mn(III) production by the enzyme. t-Butyl hydroperoxide was used for all experiments presented here. Hydrogen peroxide was not used for assays due to difficulties encountered both in the presence of low levels or higher (millimolar) concentrations of this oxidant. The high catalase activity of the catalase-peroxidase leads to rapid consumption of peroxide and bubbling in reaction mixtures under conditions expected to support measurable rates of Mn(III) formation. Hydrogen peroxide has also been shown to rapidly reduce Mn(III)-chelates back to the Mn(II) species.

Excess malate or malonate (45 mM) stabilizes Mn(III) and forms the chromophores with absorbance maxima at 290 nm (Mn(III)-malate, epsilon 290 = 4.5 mM-1 cm-1) or 270 nm (Mn(III)-malonate, epsilon 270 = 11.6 mM-1 cm-1). The formation of Mn(III) in the reaction catalyzed by the enzyme was confirmed spectrophotometrically using difference spectroscopy, by observation of the visible absorbance maximum at 458 nm due to Mn(III)-malonate (16). The formation of Mn(III) was not inhibited by bovine copper-zinc superoxide dismutase (generously provided by Dr. Howard Steinman, Albert Einstein College of Medicine) or commercial bacterial iron superoxide dismutase (90 units/ml).

The Km value for Mn(II) was calculated using a least squares fit to the equation v = V[A]/(Km + [A]) (19), where v is velocity, V is maximal velocity, Km is the Michaelis constant, and [A] is the substrate concentration. The Km for t-butyl hydroperoxide was estimated from a double reciprocal plot of rates versus concentration because the very high concentrations of this poor oxidizing substrate required to achieve saturation interfered with the assay of Mn(III) formation. Initial velocities of appearance of Mn(III)-malonate were determined for a range of concentrations of substrate A at a fixed concentration of the second substrate.

The inhibition of Mn(III)-malonate formation was studied as a function of Co(II) concentration in the mM range, using 100 µM Mn(II), 45 mM malonate, and 26 mM t-butyl hydroperoxide, pH 4.5. The Mn(III) chelates were slowly reduced by excess t-butyl hydroperoxide, but no correction for this slow reaction is made in the calculation of initial rates.

Mn(III)-malonate was prepared using Mn(III) acetate dissolved in 200 mM malonate, pH 4.5, and was standardized optically. This complex readily decays to the colorless Mn(II) complex, and it was used immediately after preparation. Mn(III) malonate could be instantaneously reduced by isoniazid. The stoichiometry of this reaction was determined by "titration" of a known amount of Mn(III)-malonate with aliquots of a freshly prepared solution of isoniazid in water. This reduction was followed spectrophotometrically at 458 nm starting either with authentic Mn(III)-malonate or the Mn(III)-malonate formed in situ by the enzyme. For the latter, the absorbance band was observed by difference spectroscopy: the spectrum of a reaction mixture recorded immediately after initiation of the reaction by the addition of peroxide was subtracted from a spectrum of the same reaction mixture after incubation for 10 min. The difference spectrum revealed the broad absorbance band at 458 nm due to Mn(III)-malonate. The same starting spectrum was subtracted from the spectrum of the incubated sample (t = 10 min) immediately after the addition of the drug to demonstrate the loss of the 458 nm absorbance.

The optical absorbance spectrum of resting catalase-peroxidase (8 µM heme) was monitored in the Soret region as a function of Mn(II) concentration using difference spectroscopy (UVICON Spectrophotometer). A prediction of secondary structure was performed using programs in the EMBL Heidelberg Predict Protein server (20-23).


RESULTS AND DISCUSSION

Initial rates of Mn(III) production were linearly dependent on the concentration of M. smegmatis catalase-peroxidase, and rates were similar whether malate or malonate was used as the chelating molecule (Table I). The rate of Mn(III) production was similar to the rate of oxidation of the organic substrate 2,6-dimethoxyphenol under conditions similar to those of the assay but in the absence of added manganese.

Table I.

Initial velocity of Mn(III) formation as a function of enzyme concentration


Enzyme concentration Velocities

µM heme µM/min
1.5a 4.4
1.0 2.7
0.5 1.5
0.5b 1.1

a Initial concentrations are 0.1 mM Mn(II), 45 mM malate, 26 mM t-butyl hydroperoxide, pH 4.5.
b Initial concentrations: 1 mM Mn(II), 45 mM malonate, 26 mM t-butyl hydroperoxide, pH 4.5.

Fig. 1 shows a double reciprocal plot for Mn(III)-malonate production as a function of t-butyl hydroperoxide concentration. The Km for t-butyl peroxide estimated from this plot was approximately 100 mM. This high value compares well with the value calculated from the oxidation of o-dianisidine using t-butyl peroxide (data not shown). Under these conditions, the rate-limiting step is either the production of a hypervalent form of the enzyme, the rate of dissociation of Mn(III) from the enzyme, or possibly the rate of Compound II reduction by Mn(II) as found for the fungal MnP2 (16). As noted earlier (9), only the ferric enzyme is detected optically after manual mixing of peroxide and resting enzyme, preventing analysis of the rates of individual reactions between Mn(II) and hypervalent forms of the catalase-peroxidase. The sluggish rate of peroxidative reactions is in part a result of the poor activity of the bulky alkyl peroxide substrate.


Fig. 1. Lineweaver-Burke plot for initial velocities of Mn(III)-malonate production by catalase-peroxidase. The oxidizing substrate (S) was t-butyl hydroperoxide. Enzyme concentration, 1 µM heme; initial Mn(II) concentration, 0.1 mM in 45 mM malonate, pH 4.5. vel, velocity.
[View Larger Version of this Image (12K GIF file)]


The formation of Mn(III) under typical reaction conditions was not inhibited by 90 units/ml of superoxide dismutase. This experiment was performed to rule out an indirect oxidation of Mn(II) by superoxide released from the decay of any Compound III (oxyferrous catalase-peroxidase) formed on the enzyme.

The rate of Mn(II) oxidation is saturable at a low concentration of metal as shown in Fig. 2. The Km for Mn(II) calculated from these rates is 5 ± 0.6 µM. This value is nearly an order of magnitude lower than that reported for the fungal manganese peroxidase under similar conditions (in the presence of malonate) but compares well with the value reported for the fungal enzyme in the absence of chelator (16). (The values for the fungal enzyme are KD values determined from titrimetric data and are not kinetic constants.)


Fig. 2. Initial velocity of Mn(III)-malonate formation as a function of initial concentration of Mn(II). Inset, Lineweaver-Burke plot of the same data. vel, velocity.
[View Larger Version of this Image (16K GIF file)]


Direct evidence for Mn(II) binding near the heme of resting (ferric) catalase-peroxidase was found by difference spectroscopy when the enzyme was titrated with Mn(II) (Fig. 3). The small change in intensity of the Soret absorbance at 408 nm (0.002 absorbance units), which was complete upon the addition of 100 µM Mn(II), was similar to that reported for the fungal enzyme (16).


Fig. 3. Titration of catalase-peroxidase with Mn(II) observed by difference spectroscopy. Mn(II) solution was added into the sample cuvette to give 30, 50, and 100 µM Mn(II), whereas equivalent volumes of water were added into the reference cuvette.
[View Larger Version of this Image (16K GIF file)]


The oxidation of 100 µM Mn(II) was inhibited by Co(II). The rate was reduced 34% by 4 mM Co(II) and 55% by 8 mM Co(II). The oxidation of the organic substrate 2,6-dimethoxyphenol was not inhibited by Co(II) in the same concentration range that gave inhibition of Mn(II) peroxidation. These results suggest that metal binds at a site different from the organic substrate binding site.

The reduction of Mn(III) by isoniazid was demonstrated spectrophotometrically by following the loss of the Mn(III)-malonate absorbance band at 458 nm. This experiment was performed starting with Mn(III)-malonate prepared de novo (Fig. 4) or that produced by the enzymatic peroxidation of Mn(II). The reaction was instantaneous and the stoichiometry of the complete reduction of a known amount of the authentic complex is consistent with the reduction of 2 mol of Mn(III) by 1 mol of drug. This suggests a mechanism involving two single electron transfer steps that produce radical intermediates as shown in Scheme I. 
<UP>Isoniazid</UP>+<UP>Mn</UP>(<UP>III</UP>)→<UP>Isoniazid<SUP>&z.ccirf;</SUP></UP>+<UP>Mn</UP>(<UP>II</UP>)
<UP>Isoniazid<SUP>&z.ccirf;</SUP></UP>+<UP>Mn</UP>(<UP>III</UP>)→<UP>products</UP>+<UP>Mn</UP>(<UP>II</UP>)
<UP>Scheme I</UP>
A radical intermediate formed by catalase-peroxidase oxidation of isoniazid has been suggested to be responsible for the inhibition of the inhA protein mentioned above (8). However, the presence of manganese together with catalase-peroxidase in those experiments raises the issue of the true origin and identity of the species involved in inhibition of the inhA enzyme, especially because no peroxide was added to reaction mixtures. The oxidation products include isonicotinic acid and isonicotinamide (8, 9), both of which are produced by the enzyme and by manganese-catalyzed reactions in the absence of enzyme (9) from a presumed radical intermediate(s).


Fig. 4. Titration of Mn(III)-malonate with isoniazid. The x intercept yields a stoichiometry of 2:1 for reduction of Mn(III) by isoniazid. Inset, optical spectra showing the absorbance due to Mn(III)-malonate and its disappearance upon addition of isoniazid. Mn(II)-malonate is colorless.
[View Larger Version of this Image (26K GIF file)]


According to the x-ray crystal structure of the MnP (24), there are three acidic residues, Glu35, Glu39, and Asp179, that constitute part of the manganese binding site. The propionate side chain of the protoporphyrin IX group is also identified as a metal ligand. The side chain of arginine 177 is considered important for stabilization of the high negative charge of the acidic site and for hydrogen bonding to Glu35. Alignment of the amino acid sequence derived from the M. tuberculosis (3) (or M. bovis (25) or M. intracellulare (26)) katG gene with the sequence of fungal manganese peroxidase in the region of the proximal imidazole (residue His173 in MnP) reveals homologies upon insertion of a six-residue gap in the manganese peroxidase sequence.
<AR><R><C>172</C><C><UP>    S</UP><A><AC><UP>H</UP></AC><AC>&cjs1170;</AC></A><UP>TVAR......<UNL>ADKV</UNL>     </UP></C><C><UP>181 MnP</UP></C></R><R><C></C><C><UP>    G</UP><A><AC><UP>H</UP></AC><AC>&cjs1170;</AC></A><UP>TFG<UNL>K</UNL>THGAGP</UP><UNL><UP>ADLV</UP></UNL></C><C><UP>katG</UP></C></R></AR>
<UP><SC>Sequences</SC></UP><UP> 1 <SC>and</SC> 2</UP>
The key aspartate residue (Asp179) in the fungal enzyme is found at the beginning of a loop. The analogous region of the catalase-peroxidase is predicted with very high probability (80-90%) to form a loop. The location of one presumed Mn(II) ligand in a loop suggests that catalase-peroxidase could maintain manganese coordination near the heme despite the insertion. The positively charged residue (Arg177) considered important in the MnP structure for stabilization of the manganese binding site is conserved in catalase-peroxidase as a lysine residue.

The key functional importance of Asp179 in the fungal MnP is illustrated by a mutation study in which a 250-fold decrease in kcat for Mn(II) peroxidation and a 50-fold increase in the dissociation constant for Mn(II) is found for the mutant enzyme, D179N (18). The fact that the bacterial catalase-peroxidases are twice the length of other peroxidases makes further sequence alignments more speculative. Nevertheless, if the acidic region around residues 35-39 in the MnP structure is considered to contain a consensus sequence for other manganese-binding ligands, the catalase-peroxidase sequence provides an excellent match in residues 519-522 with the two acidic residues in conserved positions and high homology for the other amino acids. This sequence is near the C terminus of the catalase-peroxidases from M. tuberculosis and M. bovis but is not conserved in M. intracellulare.

Another matching sequence conserved in the three mycobacterial species is found in residues 399-403 of catalase-peroxidase (EELADE). Both acidic sequences (591-522 and 399-403) are predicted with very high probability to be in helical regions, as is the sequence in the fungal enzyme containing residues 35-39 (in helix B).
<AR><R><C>33</C><C>     <UP>CG<B>E</B>DAH<B>E</B>VI</UP></C><C><UP>     MnP</UP></C></R><R><C><UP>517</UP></C><C><UP>     TL<B>E</B>EIQ<B>E</B>SF</UP></C><C>katG</C></R><R><C>398</C><C>     <UP> P<B>E</B>ELA<B>D</B>E</UP></C><C>katG</C></R></AR>
<UP><SC>Sequences</SC></UP><UP> 3–5</UP>
The mutation of residue 35 or 39 of fungal MnP leads to 50-fold decrease in the affinity for Mn(II) and 250-fold decreases in kcat (28). Unfortunately, these activities cannot be directly compared with those reported here because of the complicating catalase activity that precludes use of hydrogen peroxide as the oxidizing substrate with catalase-peroxidase.

The results presented in this communication suggest that mycobacterial catalase-peroxidase catalyzes the direct oxidation of Mn(II) to Mn(III) and show that isoniazid rapidly reduces Mn(III) in a 1:2 stoichiometry. The possibility that the oxidative activation of isoniazid and its conversion into a bacteriocidal agent is initiated or mediated by Mn(III) generated by catalase-peroxidase in vivo may provide new insights into the mechanism of action of this antibiotic.


FOOTNOTES

*   This work was supported by National Institutes of Health Grants GM40168, RR02583, and AI33696.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Dagger    To whom correspondence should be addressed: Dept. of Physiology and Biophysics, Albert Einstein College of Medicine of Yeshiva University, 1300 Morris Park Ave., Bronx, NY 10461. Tel.: 718-430-2175; Fax: 718-430-8935.
1   Zabinski, R., and Blanchard, J. S. (1997) J. Am. Chem. Soc., in press.
2   The abbreviation used is: MnP, manganese-dependent peroxidase.

ACKNOWLEDGEMENTS

We thank Drs. John Blanchard and Jack Peisach for helpful discussions and Dr. Giovanna Scapin for expertise in running Predict Protein.


REFERENCES

  1. Sriprakash, K. S., and Ramakrishnan, T. (1970) J. Gen. Microbiol. 60, 125-132 [Medline] [Order article via Infotrieve]
  2. Zhang, Y., Heym, B., Allen, B., Young, D., and Cole, S. (1992) Nature 358, 591-593 [CrossRef][Medline] [Order article via Infotrieve]
  3. Heym, B., Alzari, P. M., Honore, N., and Cole, S. T. (1995) Mol. Microbiol. 15, 235-245 [Medline] [Order article via Infotrieve]
  4. Shoeb, H. A., Bowman, B. U., Jr., Ottolenghi, A. C., and Merola, A. J. (1985) Antimicrob. Agents Chemother. 27, 399-403 [Medline] [Order article via Infotrieve]
  5. Shoeb, H. A., Bowman, B. U., Jr., Ottolenghi, A. C., and Merola, A. J. (1985) Antimicrob. Agents Chemother. 27, 404-407 [Medline] [Order article via Infotrieve]
  6. Quemard, A., Dessen, A., Sugantino, M., Jacobs, W. R., Jr., Sacchettini, J. C., and Blanchard, J. S. (1996) J. Am. Chem. Soc. 118, 1561-1562 [CrossRef]
  7. Quemard, A., Sacchettini, J. C., Dessen, A., Vilcheze, C., Bittman, R., Jacobs, W. R., and Blanchard, J. B. (1995) Biochemistry 34, 8235-8241 [Medline] [Order article via Infotrieve]
  8. Johnsson, K., King, D. S., and Schultz, P. G. (1995) J. Am. Chem. Soc. 117, 5009-5011
  9. Marcinkeviciene, J. A., Magliozzo, R. S., and Blanchard, J. S. (1995) J. Biol. Chem. 270, 22290-22295 [Abstract/Free Full Text]
  10. Hicks, D. B. (1995) Biochim. Biophys. Acta 1229, 347-355 [Medline] [Order article via Infotrieve]
  11. Yumoto, I., Fukumori, Y., and Yamanaka, T. (1990) J. Biochem. (Tokyo) 108, 583-587 [Abstract]
  12. Hochman, A., and Goldberg, I. (1991) Biochim. Biophys. Acta 1077, 299-307 [Medline] [Order article via Infotrieve]
  13. Youn, H.-D., Yim, Y.-I., Kim, K., Hah, Y. C., and Kang, S.-O. (1995) J. Biol. Chem. 270, 13740-13747 [Abstract/Free Full Text]
  14. Glenn, J. K., and Gold, M. H. (1985) Arch. Biochem. Biophys. 242, 329-341 [Medline] [Order article via Infotrieve]
  15. Glenn, J. K., Akileswaran, L., and Gold, M. H. (1986) Arch. Biochem. Biophys. 251, 688-696 [Medline] [Order article via Infotrieve]
  16. Wariishi, H., Valli, K, and Gold, M. H. (1992) J. Biol. Chem. 267, 23688-23695 [Abstract/Free Full Text]
  17. Khindaria, A., Barr, D. P., and Aust, S. D. (1995) Biochemistry 34, 7773-7779 [Medline] [Order article via Infotrieve]
  18. Kusters-van Someren, M., Kishi, K., Lundell, T., and Gold, M. H. (1995) Biochemistry 34, 10620-10627 [Medline] [Order article via Infotrieve]
  19. Cleland, W. W. (1979) Methods Enzymol. 63, 103-138 [Medline] [Order article via Infotrieve]
  20. Rost, B. (1996) Methods Enzymol. 266, 525-539 [CrossRef][Medline] [Order article via Infotrieve]
  21. Rost, B., and Sander, C. J. (1993) Mol. Biol. 232, 584-599 [CrossRef]
  22. Rost, B., and Sander, C. (1994) Proteins 19, 55-72 [Medline] [Order article via Infotrieve]
  23. Rost, B., and Sander, C. (1994) Proteins 20, 216-226 [Medline] [Order article via Infotrieve]
  24. Sundaramoorthy, M., Kishi, K., Gold, M. H., and Poulos, T. L. (1994) J. Biol. Chem. 269, 32759-32767 [Abstract/Free Full Text]
  25. Heym, B., Alzari, P. M., Honore, N., and Cole, S. T. (1995) Mol. Microbiol. 15, 235-245 [Medline] [Order article via Infotrieve]
  26. Morris, S. L., Nari, J., and Rouse, D. A. (1992) J. Gen. Microbiol. 138, 2363-2370 [Medline] [Order article via Infotrieve]
  27. Mayfield, M. B., Godfrey, B. J., and Gold, M. H. (1994) Gene (Amst.) 142, 231-235 [CrossRef][Medline] [Order article via Infotrieve]
  28. Kishi, K., Kusters-van Someren, M., Mayfield, M. B., Sun, J., Loehr, T. M., and Gold, M. H. (1996) Biochemistry 34, 8986-8994 [CrossRef]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.