BBRC Metalloantibiotic Mn(II)–bacitracin complex mimicking manganese superoxide dismutase Theeraphon Piacham a,b, Chartchalerm Isarankura-Na-Ayudhya a, Chanin Nantasenamat a, Sakda Yainoy a, Lei Ye b, Leif Bu¨low b, Virapong Prachayasittikul a,* a
Department of Clinical Microbiology, Faculty of Medical Technology, Mahidol University, Bangkok 10700, Thailand b Pure and Applied Biochemistry, Chemical Center, Lund University, P.O. Box 124, 22100 Lund, Sweden
Abstract Superoxide dismutase (SOD) activities of various metallobacitracin complexes were evaluated using the riboflavin-methionine-nitro blue tetrazolium assay. The radical scavenging activity of various metallobacitracin complexes was shown to be higher than those of the negative controls, e.g., free transition metal ions and metal-free bacitracin. The SOD activity of the complex was found to be in the order of Mn(II) > Cu(II) > Co(II) > Ni(II). Furthermore, the effect of bacitracin and their complexation to metals on various microorganisms was assessed by antibiotic susceptibility testing. Moreover, molecular modeling and quantum chemical calculation of the metallobacitracin complex was performed to evaluate the correlation of electrostatic charge of transition metal ions on the SOD activity. Ó 2006 Elsevier Inc. All rights reserved. Keywords: Antibiotic; Bacitracin; Superoxide dismutase; SOD; SOD mimic; Molecular modeling
Protective and beneficial roles of SOD have been demonstrated both pre-clinically and clinically in combating a broad range of diseases, including ischemic-reperfusion injury, inflammation, cancer, and AIDS [1–4]. However, some major drawbacks associated with the use of native enzymes, such as immunological problems and difficulties in penetrating membrane barriers to reach targeted sites, have halted the application of SOD as a therapeutic agent. Synthetic SOD mimics have been potentially studied in order to be developed into pharmaceutical candidates since it is conceptualized that removing superoxide anion can moderate the course of inflammation. Moreover, some of the synthetic metal complexes have shown to possess favorable SOD activity and promising clinical effects [5,6]. Bacitracin is a metal-dependent dodecapeptide antibiotic produced by Bacillus species (e.g., Bacillus licheniformis and Bacillus subtilis) [7]. The metalloantibiotic is effective against Gram-positive bacteria by inhibiting the synthetic *
Corresponding author. Fax: +662 849 6330. E-mail address:
[email protected] (V. Prachayasittikul).
pathway for cell wall formation. Because of its metal binding capability, bacitracin is also a potent inhibitor for metalloprotease and has been applied in immobilized metal affinity chromatography (IMAC) for metalloprotease purification [8]. In addition, antagonistic activity to diphtheria toxin [9] and inhibition of HIV [10] invasion at the cellular level have been demonstrated. Although bacitracin is one of the most widely used antibiotics, its bacterial resistance has rarely been reported. In this context, bacitracin infers great potential for designing new drugs against bacterial infection and other diseases. Bacitracin provides strong affinity to divalent metal ions such as Zn(II), Cu(II), Co(II), and Mn(II) in the formation of 1:1 complex. Its metal binding affinity in order of increasing affinity has been established as follows: Cu(II) > Ni(II) > Co(II) Zn(II) > Mn(II) [11]. Structural characterization of metallobacitracin showed that it is composed of a cyclic heptapeptide and a short N-terminal sequence containing a triazole ring (Fig. 1) [12]. The divalent metals interact with the cyclic and the linear peptides to form a strong bending structure that encapsulates the
O
HO NH2 O
O
O H N
N
HN
HN NH
O
Ile 1 N-terminal amino
His 10
O
HN NH
NH2
Thiazoline ring O
O
S
NH
N O
NH
N H
NH
O
O HN
NH
H2 N
O O
D-Glu 4 carboxylate
HO O
Fig. 1. The chemical structure of bacitracin A (PubChem, Accession code CID 439542).
metal inside the coordination sphere. On the complex of Zn(II)–bacitracin, the first coordination sphere comprises of three nitrogen-containing ligands and one oxygen-containing ligand from His-10 imidazole nitrogen, thiazoline nitrogen, Ile-1 amino nitrogen, and Glu-4 carboxylate oxygen, respectively [13,14]. It is possible that the metal binding sites in metallobacitracin exhibit similar arrangement to that of the metal coordination sphere in native MnSOD [15]. In our previous work, we have synthesized a SOD mimic by template polymerization of functional monomers. The obtained polymer contains three nitrogen-containing ligands from imidazole groups and an oxygen-containing ligand from carboxyl group, and displays interesting MnSOD activity [16]. Based on the facts that bacitracin possesses strong metal affinity and is resistant towards protease digestion, we are interested in exploring possible SOD activity of metalloantibiotics, as it may serve as an impetus towards the development of novel therapeutic SOD mimics. Quantum chemistry has immense applications in the area of chemistry and biochemistry [17]. The arsenal of quantum chemical techniques have been applied to the study of various chemical and biochemical problems including, for example, the correlation of structure–property relationships [18], elucidation of rapid enzymatic reac-
tions [19], and mechanism of enzyme inhibitor interaction [20], etc. Therefore, quantum chemical calculation has been employed to study the influence of different transition metal ions on the SOD activity as a function of the electrostatic charge of the transition metal ions. Construction of the metallobacitracin complexes has been made according to known coordination chemistry from the literature. We, therefore, hypothesize that there is a correlation between the electrostatic charge of transition metal ions and SOD activity. Materials and methods Preparation of metallobacitracin complex. The metallobacitracin were prepared by incubating each divalent metal ion (Mn(II), Co(II), Ni(II), and Cu(II)) with bacitracin in 1:1 molar ratio in methanol for 1 h. Then, solvent was removed under vacuum to give metallobacitracin complex powders (Mn(II)–bacitracin, Co(II)–bacitracin, Ni(II)–bacitracin, and Cu(II)–bacitracin). Catalysis of superoxide dismutation. Each metallobacitracin complex was tested for SOD activity using a previously described method [16,21]. The SOD activity of metallobacitracin was assayed by measuring inhibition of the photoreduction of nitro blue tetrazolium (NBT). This indirect assay is comprised of several reactions: the photochemically excited riboflavin was first reduced by methionine into a semiquinone, which donated an electron to oxygen to form the superoxide source. The superoxide readily converted NBT into a purple formazan product. In this
Results and discussion When the metallobacitracins were assayed, we found that Mn(II)–bacitracin exhibited the highest SOD activity, while the metal-free bacitracin did not (Table 1). The SOD activity of metallobacitracin was found to be in the order of Mn(II) > Cu(II) > Co(II) > Ni(II). In a subsequent experiment, Mn(II)–bacitracin complex was further investigated by varying its concentration in the assay. Our result indicated that approximately 1.2 mg of Mn(II)–bacitracin was needed in order to inhibit 50% of the NBT photoreduction (Fig. 2). The amount of Mn(II)–bacitracin is equivalent to 3.7 U of native SOD enzyme. We noticed that in the absence of bacitracin, MnCl2 in solution displayed very low intrinsic SOD activity. Although our presently used bacitracin (Sigma) is a mixture of derivatives, among which bacitracin A and B are the major components, the metal binding ligands from
Table 1 SOD activity and calculated electrostatic charge of the transition metal ions of metallobacitracin No.
Test compounda
Inhibition of NBT reduction (%)
Calculated electrostatic charge
1 2 3 4 5 6 7
Mn(II)–bacitracin Co(II)–bacitracin Ni(II)–bacitracin Cu(II)–bacitracin Metal-free bacitracinb Metal Native SODc
48 31 3.5 47 0 0 70
1.110 0.609 0.365 0.270 — — —
a Each metallobacitracin (1 mg/mL) was dissolved in methanol and assayed to evaluate SOD activity by NBT photoreduction inhibition method. b Negative control. c Native SOD (7.46 U) from bovine erythrocytes was used as positive control.
100
80
Inhibition (%)
regard, the SOD activity was inversely related to the amount of formazan formed. In a control experiment, we used metal-free bacitracin to give a background visible absorbance value. Antimicrobial susceptibility testing. Clinical strains of b-Streptococcus group A, Streptococcus pneumoniae ATCC 49619, Staphylococcus aureus ATCC 25923, Enterococcus spp. ATCC 29212, Escherichia coli ATCC 25922, and Pseudomonas aeruginosa ATCC 15442 were used for antibiotic susceptibility testing. Each metallobacitracin complex (Mn(II), Co(II), Ni(II), and Cu(II)) was dissolved in 70% ethanol, followed by dilution to five concentrations (0.005, 0.05, 0.5, 5, and 50 U/15 ll). Fifteen microliters of each concentration was saturated and dried on a filter paper disk with diameter of 0.45 mm. Standard bacitracin disk at concentration of 0.04 U was obtained from Becton–Dickinson, MD, USA. Antibiotic susceptibility testing was modified from guidelines of the National Committee on Clinical Laboratory Standards (NCCLS) [22]. Mueller-Hinton sheep blood agar (Difco, Becton–Dickinson, MD, USA) was inoculated with a direct colony suspension of each organism equivalent to a 0.5 McFarland turbidity standard. Disks with metallobacitracin at different concentrations were dispensed on the agar plate, followed by measurement of the zone diameters after 24 h of incubation in 5% CO2 at 35 °C. Molecular model of metallobacitracin. The calculations were carried out on a Pentium 4 3.0 GHz personal computer using the quantum chemical software Spartan’04 [23] on a Windows XP operating system. Molecular mechanics, Hartree-Fock, and density functional theory were used to obtain the proposed structure of metallobacitracin complexes. The molecular structure of bacitracin A was taken from PubChem [24] under the accession code CID439542. Coordination of bacitracin A to different transition metal ions including Mn(II), Co(II), Ni(II), and Cu(II) was constructed. The molecular models of metallobacitracin complex were based on proton NMR studies of the structure of Co(II)–bacitracin A1 [14]. For all metallobacitracin structures, formal charge of +2 was assigned to the transition metal ions. The molecular models were optimized with Merck Molecular Force Field (MMFF). The size of the metallobacitracin complexes ruled out the use of ab initio quantum chemical calculations (e.g., density functional theory) due to its high computational demands. Therefore, a smaller model that is representative of the metal–ligand interaction was constructed by truncating the bacitracin backbone while leaving the metal center and coordinated ligands intact. The geometry of the simplified metallobacitracin complexes was pre-optimized using MMFF. The structure was then subjected to full geometry optimization using the semi-empirical PM3 method. The electrostatic charge of the transition metal ions was derived from the PM3 optimized structures.
60
40
IC50 = 1.2 mg/ml 20
0 0
1
2
3
4
Mn-Bacitracin concentration (mg/mL)
Fig. 2. Inhibition of NBT photoreduction by increasing the concentration of Mn(II)–bacitracin.
each bacitracin congener are nearly identical (histidine, triazoline ring, N-terminal NH2 group, and glutamic acid). Either each Mn(II)–bacitracin congener differs in SOD activity is subjected to our further investigation. It is worthy to state that 4 out of 12 amino acid residues in bacitracin are in D-form, which results in protease resistance potentiation. We also tested the stability of Mn(II)–bacitracin in the presence of EDTA (a strong chelating agent) and BSA (one of the strongest biological chelators). Varying concentration of EDTA or BSA was introduced into the assay solution followed by assessment of the SOD activity. Our results indicated that Mn(II)–bacitracin at 1 mg/mL maintained its SOD activity even in the presence of EDTA up to 0.37 mg/mL while the BSA (1.2 mg/mL) could not affect the SOD activity. Antibiotic susceptibility testing of the Mn(II)–bacitracin on a clinical strain b-Streptococcus group A gave a concentration dependent on growth inhibition activity (Fig. 3A). Meanwhile, various metal complexes gave non-significant
A
0.5 units
5 units
0.005 units
50 units
B
oxygen [14]. On the other hand, it has been shown that Cu(II) takes on a different coordination chemistry in which Cu(II) is a tetragonally distorted geometry with two coordinated nitrogens and two coordinated oxygens, particularly, His-10 imidazole nitrogen, thiazoline nitrogen, Glu-4 carboxylate oxygen, and Asp-11 carboxylate oxygen [25]. In the construction of metallobacitracin, the initial structure of bacitracin A was obtained from PubChem and the coordination to the metal ions was drawn in Spartan’04 followed by geometry optimization with MMFF. The calculated molecular model of Mn(II)–bacitracin is shown in Fig. 4. It is interesting to note that the established order of binding affinity of the transition metal ions was found to be inversely correlated with the observed SOD activity reported in this work. The order of metal binding affinity and SOD activity is Cu(II) > Ni(II) > Co(II) Zn(II) > Mn(II) [11] and Mn(II) > Cu(II) > Co(II) > Ni(II), respectively. However, it should be noted that the negative correlation is valid for Mn(II), Co(II), and Ni(II) but not Cu(II). It is possible that the observed trend, in which Cu(II)–bacitracin did not follow the order of increasing SOD activity with decreasing metal binding affinity, is because Cu(II) takes on a different coordination chemistry from the other divalent metal ions. Likewise, the negative correlation for Mn(II), Co(II), and Ni(II) can be attributed to the similar coordination chemistry adopted by these divalent metal ions as proposed earlier. Furthermore, the calculated electrostatic charge of the transition metal ions (see Table 1) derived from the PM3 optimized geometry (Mn(II)–bacitracin serves as a representative model of the metallobacitracins calculated as shown in Fig. 5 yielding a positive correlation between the SOD activity and the electrostatic charge. This positive correlation trend is valid for Mn(II), Co(II), and Ni(II) but not for Cu(II). Again, this may be attributed to the differences observed among the three divalent metal ions and Cu(II), particularly, its coordination chemistry. Thus, the evidences suggest that the SOD activity is greater when the metal binding affinity and electrostatic charge are weak and strong, respectively. In summary, we have found that the metallocyclic peptide, bacitracin, has an interesting SOD activity. The Mn(II)–bacitracin complex is potentially useful as an
0.05 units
0.04 units (standard disc). Mn(II)-bacitracin
Ni(II)bacitracin
Cu(II)bacitracin
Co(II)-bacitracin Fig. 3. Antibiotic susceptibility testing with varying concentrations of bacitracin (A) and various metal ions (B).
difference of anti-streptococcus (Fig. 3B). However, the Mn(II)–bacitracin was found to exert greater inhibitory activity than the other metallobacitracins on S. aureus and Enterococcus spp. (Table 2). For the construction of the molecular models of metallobacitracin, we assumed that Mn(II) and Ni(II) adopt a similar coordination chemistry to that of Co(II) because they all have a vacant d shell. Proton NMR studies established that Co(II) is coordinated to three nitrogens and one oxygen, namely, His-10 imidazole nitrogen, thiazoline nitrogen, Ile-1 amino nitrogen, and Glu-4 carboxylate Table 2 Results of susceptibility testing by metallobacitracins Microorganisms
Mn(II)–bacitracina
Co(II)–bacitracina
Ni(II)–bacitracina
0.05
0.5
5
50
0.05
0.5
5
50
0.05
0.5
5
50
0.05
0.5
5
50
b-Streptococcus group A Streptococcus pneumoniae Staphylococcus aureus Enterococcus spp. Escherichia coli Pseudomonas aeruginosa
1.30 0.70 n.z. n.z. n.z. n.z.
2.10 2.00 0.85 0.60 n.z. n.z.
2.50 2.90 1.50 1.20 n.z. n.z.
2.90 3.60 2.10 1.90 0.80 0.55
1.40 0.90 n.z. n.z. n.z. n.z.
2.10 1.80 n.z. n.z. n.z. n.z.
2.50 2.80 1.10 0.85 n.z. n.z.
2.80 3.20 1.80 1.55 0.85 0.70
1.20 0.80 n.z. n.z. n.z. n.z.
1.90 2.10 n.z. n.z. n.z. n.z.
2.50 2.80 1.20 0.90 n.z. n.z.
2.90 3.40 1.60 1.50 0.80 0.55
1.20 0.60 n.z. n.z. n.z. n.z.
1.85 2.00 n.z. n.z. n.z. n.z.
2.40 2.90 1.20 0.90 n.z. n.z.
2.80 3.60 1.55 1.40 0.80 0.60
n.z. represents no inhibition zone. a Metallobacitracins of various concentrations (units) and their radius of the inhibition zone (cm).
Cu(II)–bacitracina
Fig. 4. Molecular model of Mn(II)–bacitracin complexes.
Fig. 5. Structure of the Mn(II)–ligand models derived from metallobacitracin complexes.
effective agent against oxidative stress (for O2 scavenging). On the other hand, probably this Mn–bacitracin may be involved in the respiratory burst mechanism of white blood cells that could enhance bacterial killing by synergistic process to convert superoxide radical into hydrogen peroxide which is used by enzyme myeloperoxidase to convert normally unreactive halide ions into reactive hypohalous acids that are toxic to bacteria. Also of its antibiotic mechanism could be useful for bacterial and oxidative stress treatments. With more in-depth study the potential of developing metallobacitracin-based therapeutics may be realized in the future.
Acknowledgments We acknowledge financial support for the integrated program ‘‘Biomimetic Materials Science (Biomics)’’ by the Swedish Strategic Research Foundation (SSF). T.P. and S.Y. are grateful for the graduate fellowships from ‘‘The Ministry Staff Development Project’’ by the Ministry of University Affairs, Thailand. C.N. is a recipient of the Royal Golden Jubilee PhD scholarship under supervision of V.P. V.P. and C.I. are also greatly indebted to the Thailand Toray Science Foundation (TTSF) and Mahidol University for financial support. Finally, we would like to thank Assoc.
Prof. Dr. Li-June Ming from the Department of Chemistry and Institute for Biomolecular Science, University of South Florida for helpful suggestions regarding molecular modeling of the structures of metallobacitracin complexes. References [1] S.R. Maxwell, Prospects for the use of antioxidant therapies, Drugs 49 (1995) 345–361. [2] W. Land, J.L. Zweler, Prevention of reperfusion-induced, free radical-mediated acute endothelial injury by superoxide dismutase as an effective tool to delay/prevent chronic renal allograft failure: a review, Transplant Proc. 29 (1997) 2567–2568. [3] S.L. Church, J.W. Grant, L.A. Ridnour, L.W. Oberley, P.E. Swanson, P.S. Meltzer, J.M. Trent, Increased manganese superoxide dismutase expression suppresses the malignant phenotype of human melanoma cells, Proc. Natl. Acad. Sci. USA 90 (1993) 3113–3117. [4] M.A. Edeas, I. Emerit, Y. Khalfoun, Y. Lazizi, L. Cernjavski, A. Levy, A. Lindenbaum, Clastogenic factors in plasma of HIV-1 infected patients activate HIV-1 replication in vitro: inhibition by superoxide dismutase, Free Radic. Biol. Med. 23 (1997) 571–578. [5] D. Salvemini, D.P. Riley, S. Cuzzocrea, SOD mimetics are coming of age, Nat. Rev. Drug Discover. 1 (2002) 367–374. [6] D.P. Riley, Functional mimics of superoxide dismutase enzymes as therapeutic agents, Chem. Rev. 99 (1999) 2573–2588. [7] L.J. Ming, J.D. Epperson, Metal binding and structure-activity relationship of the metalloantibiotic peptide bacitracin, J. Inorg. Biochem. 91 (2002) 46–58. [8] M. Faisal, D.Y. Schafhauser, K.A. Garreis, E. Elsayed, J.F. La Peyre, Isolation and characterization of Perkinsus marinus proteases using bacitracin–sepharose affinity chromatography, Comp. Biochem. Physiol. B: Biochem. Mol. Biol. 123 (1999) 417–426. [9] R. Mandel, H.J. Ryser, F. Ghani, M. Wu, D. Peak, Inhibition of a reductive function of the plasma membrane by bacitracin and antibodies against protein disulfide-isomerase, Proc. Natl. Acad. Sci. USA 90 (1993) 4112–4116. [10] H.J. Ryser, E.M. Levy, R. Mandel, G.J. DiSciullo, Inhibition of human immunodeficiency virus infection by agents that interfere with thiol-disulfide interchange upon virus-receptor interaction, Proc. Natl. Acad. Sci. USA 10 (1994) 4559–4563.
[11] J.T. Garbutt, A.L. Morehouse, A.M. Hanson, Food additives, metal binding properties of bacitracin, J. Agric. Food Chem. 9 (1961) 285– 289. [12] L.J. Ming, Structure and function of ‘‘metalloantibiotics’’, Med. Res. Rev. 23 (2003) 697–762. [13] F. Drabløs, D.G. Nicholson, M. Rønning, EXAFS study of zinc coordination in bacitracin A, Biochim. Biophys. Acta 1431 (1999) 433–442. [14] J.D. Epperson, L.J. Ming, Proton NMR studies of Co(II) complexes of the peptide antibiotic bacitracin and analogues: insight into structure-activity relationship, Biochemistry 39 (2000) 4037–4045. [15] G.E. Borgstahl, M. Pokross, R. Chehab, A. Sekher, E.H. Snell, Cryotrapping the six-coordinate, distorted-octahedral active site of manganese superoxide dismutase, J. Mol. Biol. 296 (2000) 951–959. [16] T. Piacham, C. Isarankura Na Ayudhya, V. Prachayasittikul, L. Bu¨low, L. Ye, A polymer supported manganese catalyst useful as a superoxide dismutase mimic, Chem. Commun. (2003) 1254–1255. [17] C.J. Cramer, Essentials of Computational Chemistry: Theories and Models, second ed., John Wiley & Sons, New Jersey, 2004. [18] C. Nantasenamat, T. Naenna, C. Isarankura Na Ayudhya, V. Prachayasittikul, Quantitative prediction of imprinting factor of molecularly imprinted polymers by artificial neural network, J. Comput. Aided Mol. Des. (2005), doi:10.1007/s10822-005-9004-4. [19] A. Warshel, Computer simulations of enzyme catalysis: methods, progress, and insights, Annu. Rev. Biophys. Biomol. Struct. 32 (2003) 425–443. [20] E. Jenwitheesuk, R. Samudrala, Improved prediction of HIV-I protease-inhibitor binding energies by molecular dynamics simulations, BMC Struct. Biol. 3 (2003) 2. [21] W.F. Beyer Jr., I. Fridovich, Assaying for superoxide dismutase activity: some large consequences of minor changes in conditions, Anal. Biochem. 161 (1987) 559–566. [22] NCCLS. Performance Standards for Antimicrobial Susceptibility Testing. Eleventh Informational Supplement, NCCLS document M 100-911 (ISBN 1-56238-426-0). NCCLS, Pensylvania, USA 2001. [23] Spartan’04, Wavefunction, Inc., Irvine, CA. [24] PubChem, . [25] J. Peisach, W.E. Blumberg, Structural implications derived from the analysis of electron paramagnetic resonance spectra of natural and artificial copper proteins, Arch. Biochem. Biophys. 165 (1974) 691–708.
