- Email: [email protected]
Protective effects of bacterial glyceroglycolipid M874B against cell death caused by exposure to heat and hydrogen peroxide
BIOENGINEERING Vol. 89, No. 4, 345-349. 2000
JOURNAL OF BIOSCIENCE AND
Protective Effects of Bacterial G lyceroglycolipid M874B against Cell Death Caused by Exposure to Heat and Hydrogen Peroxide MOTOKO MATSUFUJI,‘* YASUNORI NAGAh4ATSU,2 AND AKIHIRO YOSHIMOT02 Central Research Laboratories, Mercian Corporation, 9-l Johnan I-chome, Fujisawa 251-0057 and Faculty of Applied Biological Science, Hiroshima University, 4-4 Kagamiyama I-chome, Higashi-Hiroshima 739-8526,= Japan Received 1 November 1999IAccepted 11 January 2000
It was revealed by bioassay using sodA and katA mutants of Bacillus subtilis that the bacterial monogalactosyldiacylglycerol M874B, previously characterized as an alkyl peroxyl radical scavenger, was also capable of protecting cells from death caused by heating and exogenous HzOz. Chemical assaysusing the Fenton reaction and xanthine-xanthine oxidase revealed that M874B could quench hydroxyl radicals but not superoxlde anions. Wheat monogalactosyldiacylglycerol, but neither digalactosyldiacylglycerol nor synthetic diacylglycerol, also had the same activities as those of M874B, although it was less efficient than M874B. These results suggest that monogalactosyldiacylglycerols such as M874B are a new type of oxygen radical scavengerscapable of quenching some reactive oxygen species. [Key words: glyceroglycolipid, monogalactosyldiacylglycerol, gen peroxide, sodA mutant, katA mutant, Bacillus subtilis] Aerobic living cells possess a defense mechanism against oxidative damage caused by reactive oxygen species that are generated during normal aerobic metabolism. Short-lived reactive oxygen species, including superoxide anions, Hz02 and hydroxyl radicals, are noted for their high reactivity and ability to cause damage to proteins, lipid membranes and DNA (1). Enzymatic digestion of these reactive oxygen species is catalyzed by superoxide dismutase (SOD), catalase and cytochrome c peroxidase (2-4). Loss of SOD or catalase in aerobically grown cells inhibits cell growth and causes stationaryphase death (5), and hypersensitivity to superoxide (6), hydroxyl radicals or H202 (7, 8). Most microbial cells are also sensitive to sudden heat exposure. The major factors causing cell death after heat exposure are still unknown although a lethal temperature shift induces many heat-shocked proteins (HSPs) (9, 10). Recently, Benov and Fridovich reported that superoxide anion was a major cause of cell death in both the aerobic heat shock and stationary-phase of growth (5, 11). Cell death has also been reported to be caused by H20z exposure (7, S), and at least one of the major causes is hydroxylradical-induced cell death. The activity of scavenging superoxide or hydroxyl radicals is measured as that of the loss of cell viability following exposure to heat or H202. In this case, the use of SOD-deficient and catalasedeficient mutants of bacteria is beneficial for the bioassays since both the mutants are highly sensitive to heat and H202, respectively. We reported previously that a glyceroglycolipid, M874B [1,2-di-O-(12-methyltetradecanoyl)-3-O-@-galactopyranosyl-sn-glycerol], isolated from Microbacterium sp. strain M874, had a quenching effect on alkyl peroxyl radicals (12) and thus were prompted to study its antioxidant activity against other reactive oxygen species. To elucidate this, we used bioassays employing the sodA mutant PS2495 and katA mutant PS2488 of B. subtilis, which were isolated and characterized by Setlow for their hypersensitivity to heat exposure which induces
oxygen radical scavenger, heat shock, hydro-
generation of superoxide anions and H202, respectively (6). In this paper, we report that M874B has a significant protective activity against cell death caused by exposure to heat or HzOz. We also examined the scavenging activity of M874B on superoxide and hydroxyl radicals by chemical assays using the xanthine-xanthine oxidase and Fenton reaction systems, respectively. It was ascertained that M874B had at least a pronounced quenching activity on hydroxyl radicals released from Hz02 although its inhibitory effect on the xanthine-xanthine oxidase system was not defined. MATERIALS
AND METHODS
Chemicals Wheat flour monogalactosyldiacylglycerol (MGDAG), wheat flour digalactosyldiacylglycerol (DGDAG) and bovine kidney superoxide dismutase were obtained from Funakoshi Chemicals (Tokyo). Glutathione, a-tocopherol, ascorbic acid, chloramphenicol (CM), 1,2-dimyristoyl-rat-glycerol (Cl4 : 0) (DAG), xanthine, xanthine oxidase, 1,10-O phenanthroline, cytochrome c from horse heart, and bovine liver catalase were purchased from Sigma Chemicals (St. Louis, MO, USA). Glyceroglycolipid M874B was obtained as described previously (14). The glycolipid samples were dissolved or suspended in sterile phosphate-buffered saline (PBS) by sonication and added to the reaction mixture. Microorganisms and cultivation Mutant strains PS2495 (sodA, Cmr) and PS2488 (katA, Cmr) and the corresponding parent strain PS832 of Bacillus subtilis (6) were used for this experiment. The bacteria were maintained on Luria-Bertani (LB) agar (log of bacto-tryptone, 5 g of yeast extract, 10 g of NaCl, 15 g of agar in 1 I of deionized water; pH 7.4) supplemented or not supplemented with CM (3 pg/ml). The bacteria were grown at 28°C for 18 h in a 250-ml Erlenmeyer flask containing 30ml of LB medium with or without CM on a rotary shaker (220 rpm). All the 18-h cultures of the parent strain and mutant sodA or katA strains showed an optical density of about 1 .O-at 660 nm and were used for the bioassays without dilution, unless otherwise stated.
* Corresponding author. 345
346
MATSUFUJI ET AL.
Heat exposure test Aliquots (100~1) of the 18-h sodA mutant PS2495cultures were poured into a 96-well cell culture plate (NUNC) and 20 ~1 each of glycolipid solution at various concentrations or sterile PBS (as control) was added. The plate was immediately shaken at the indicated temperature on a plate shaker, M icrotherm (CAMLAB, USA), for an appropriate time and the cultures were then diluted serially IO-fold with sterile PBS, followed by plating to observe for colony formation on a YMPG agar plate (10 g of glucose, 3 g of yeast extract, 3 g of malt extract, 5 g of Polypepton and 15 g of agar; pH 7.0). The colony assay was performed in triplicate and the colonies were counted after 20 h incubation at 28°C. The viable counts are representedby the means of triplicate assays. A comparative study of heat sensitivity between the parent strain PS863 and the sodA mutant PS2495 was performed using 18-h cultures in the same manner as described above. Aliquots (100 ~1) of the 18-h HzOz exposure test katA mutant PS2488 culture and 20 ~1 each of the glycolipid solutions at various concentrations or sterile PBS (as control) were poured into a 96-well cell culture plate. To each of the m ixtures was added 20 ~1 of 1 m M HzOz and the plate was incubated on a plate shaker, M icrotherm, at 28°C for an appropriate time. After serial IO-fold dilution with sterile PBS, 100/-11each of the diluent was plated on a YMPG agar plate. Colonies were counted after incubation for 20 h at 28°C. The colony count are represented by the means of triplicate assays. In the HzOz dose-responseexperiment with the parent strain PS832 and the katA mutant PS2488, 100~1 aliquots of both the 18-h cultures were added to a 96-well cell culture plate and 20 ~1 of Hz02 solution at various concentrations was added to each. The plate was shaken at 28°C for 1 h and the colonies were counted in the same manner as described above. Chemical assays for superoxide and hydroxyl radicals
Xanthine-xanthine oxidase was used as the superoxideanion generating system and the reaction was monitored by measuring the extent of reduction of cytochrome c (13). The assay m ixture contained 50 P M xanthine, various amounts of glycolipid, 20pM of cytochrome c, 100P M of EDTA and 1.5 units of xantine oxidase in 3 m l of 50mM phosphate buffer (pH 7.8). These constituents were directly added to a 3-ml cuvette and the reaction was initiated by adding xanthine oxidase. The reaction m ixture was allowed to stand at 25°C for 5 m in and the extent of cytochrome c reduction by the superoxide anions generated was recorded by the absorbance at 550nm. Superoxide anion production was expressed as nmoles of cytochrome c reduced/min in the reaction m ixture. The Fenton reaction was used to generate hydroxyl radicals. The assay m ixture contained 2 m M of Hz02 1 m M of ferrous sulfate, various amounts of glycolipid and 5 m M of NaCl in a total volume of 1 m l (pH 7.0). The reaction was started by adding Hz02 and allowed to proceed at 25°C for IOmin. The oxidation of ferrous iron by HzOz to generate hydroxyl radicals was monitored by a calorimetric assay of the amount of unconsumed ferrous iron (14). Thus, at the end of the reaction, 1 m l of 1, IO-0phenanthroline aqueous solution was added to the assay m ixture. After allowing the m ixture to stand for 10min at room temperature, the absorbance of the red complex was measuredat 510 nm.
J. BIOSCI. BIOENG.,
RESULTS Effect of the glyceroglycolipid M874B on H202-induced cell death B. subtilis produces some major catalases
which are products of katA and katE. The katA appears to be the major catalasein growing cells and plays a significant role in the resistance of growing cells to HzOz (15-17). The katX product was produced even in the spores. Loss of catalaseresults in increasedsensitivity to Hz02 (18). Figure 1 illustrates the hypersensitivity of the katA mutant strain PS2488 to HzOz exposure as compared with the Hz02 sensitivity of the katA mutant PS2488and the parent strain PS832. Cells that survived in the presenceof exogenousHzOz were enumeratedby the colony formation assay. The katA mutant cells were hypersensitive to Hz02 and the cell survival was decreasedto approximately one-fourth by exposure to even as low as 1 m M of HzOz; the cells of the parent strain were, however, resistant to even over 1 O O m M of H202. We thus examined the protective effect of the glyceroglycolipid M874B against Hz02 stressby assayingits effect of inhibiting HzOz-induced death of katA mutant PS2488 cells. Glyceroglycolipid M874B and wheat flour MGDAG, both known to be protective against damage induced by alkyl peroxyl radicals, namely, tert-butylhydroperoxideinduced cell death (12), were also demonstrated to have a protective effect against HzOz-induced cell death (Fig. 2). The activity of M874B was dose-dependentand greater than that of wheat flour MGDAG. M874B exhibited distinct protective activity against HzOz-induced damage at the concentration of 0.05 mg/ml, whereas wheat flour MGDAG exhibited approximately the samelevel of activity at the concentration of 1.Omg/ml or more. The doseresponsecurve of this activity of M874B was almost the same as that of its quenching activity against alkyl peroxyl radicals (12). Table 1 shows the effects of some related glycerolipids and antioxidants, catalase and SOD on HzOz-induced death. Wheat flour DGDAG and DAG had no protective effect at either the concentrations of 1 mg/ml or 5 mg/ml, suggestingthat at least the structure of MGDAG was required for the protective activity. Antioxidants such as ascorbate at 1 m M were not proI
0
1 10 H,O, concentration (mM)
100
FIG. 1. Assay mixtures containing /carA mutant cells or parent cells and various concentrations of Hz02 in a 96-well cell culture plate were incubated at 28’C for 1 h and the viable cells were counted as described in the text. The value was expressedas a % of the viable cell count in each control assay without Hz02. The viable cell count of strains PS2488 and PS832 in the control assay without Hz02 were 1.26 x 107/ml and 1.57 x lO’/ml, respectively: PS2488 (0), PS832 (0).
VOL.
GLYCEROGLYCOLIPID
89, 2000
0
0.05 0.1 0.25 0.5 1 2.5 Concentration (mg/ml)
5
FIG. 2. Assay m ixtures containing k&t mutantcells,1 mM Hz02 of M874B or MGDAG in a 96-we11cell and various concentrations cultureplatewereincubated at 28°C for 1 h andthe viablecellswere countedas described in the text. The values are the m e a n+SD of triplicateassays: M874B @I), MGDAG (B).
tective. O n the other hand, a-tocopherol exhibited a pronounced activity of inhibiting H20z-induceddeath at 1 m M . However, this activity of a-tocopherol was somewhat lower than that of M874B as it was comparableto that of 0.7 m M less of M874B (0.5 mg/ml) when compared on a m o lar basis. Exogenouscatalaseexerted complete protection against damageinduced by 1 m M Hz02 when it was added at a high concentration of 5OOU/ml in this assay, while SOD was not protective against HzOz-induceddeath even when added in excess. Effect of glyceroglycolipid M874B on heat-induced death Superoxideanions are one of the m a jor causes of heat lability in aerobically grown bacteria (5, 6, 11, 15). Therefore, loss of viability by heating is more significant in sod-deficientmutant cells than in sod-containing parent cells of bacteria. Heat sensitivity of the so&l mutant strain PS2495and parent strain PS832of B. subtilis was examinedby the colony formation assayof cells that heat treatment at temperaturesranging from 40°C to 53°C. The results are shown in F ig. 3. Viable counts of neither strain decreasedsignificantly following heating at 45°C or less for 30 m in, but they rapidly decreasedby heating at 47°C or more. It was evident that the sodA mutant was much more sensitive to heat-induceddeath than the parent PS832strain. The most significant differTABLE
Addition
M874B PROTECTIVE AGAINST HEAT AND H202 STRESS
40 42
44
347
46 48 50 52 54
Temperature
(“C )
FIG. 3. Assay mixtures containing sodA mutant cells or parent cells in a %-weII cell culture plate were shaken at 28,40,45,47,50 or 53°C for 30 min and the viable cells were counted as described in the text. The values were expressedas a % of the viable cell count *SD in each control assay shaken at 28OC. The viable cell counts of PS2495 and PS832 in the control assay shaken at 28°C were 8.16 x 107/ml and 10.8 x 107/ml, respectively: so&l mutant PS2495(0), parental PS832 (0).
ence in viability between the two strains was observed when the bacterial cells were heat-treatedat 50°C. Benov and Fridovich (11) reported based on their experimental results using E. coli that the differencein heat sensitivity between the sodA sodB mutant strain and the parent strain was most pronouncedwhen the cells were exposed to 48°C which corresponded to the temperature at which cellular SOD in the bacteria is inactivated. In order to determine the ability of glyceroglycolipids to protect cells from heat-induceddeath, we assayedthe cell viability when the sodA mutant PS2495 cells were exposed to 50°C for 3 0 m in in the presence of glyceroglycolipids. F igure 4 shows the effect of M874B and wheat flour MGDAG and DGDAG on the heatinduced cell death. It was obvious that M874B and MGDAG, but not DGDAG, had a pronounced protective effect against heat-inducedcell death. The activity of M874B was dose-dependentand more marked than that of MGDAG. Further experiments with heat exposure demonstrated that the protective effect of M874B was more marked when the cells were exposedto the higher temperature of 50°C or 53°C than to 47°C (Fig. 5),
1. Effects of some related glyceroglycolipids and antioxidants on HzOz-induced cell death and heat-induced cell death Effect on HrOr-induced Heat-induced Concentration cell death cell death
(viable cell counts: x lW/ml) 28.1 22.0 1 mg/ml 50.6 44.4 5 mg/ml 57.6 69.2 MGDAG 1 mg/ml 22.2 30.4 DGDAG 1 mg/ml 22.0 17.2 DAG(C14) 1 mg/ml 26.2 25.4 Ascorbate 1mM 5.0 25.8 Glutathione 1mM 12.4 a-Tocopherol 1mM 45.2 23.0 Catalase 100 units/ml 17.3 500 units/ml 61.6 28.1 Superoxide dismutase loo0 units/ml
None M874B
0
2.5 0.5 1 Concentration (mg/ml)
5
FIG. 4. Assay mixtures containing sodA mutant cells and various concentrations of glycohpid or PBS (as control) in a %-we11ceII culture plate were shaken at 50°C for 30min and the viable cells were counted as described in the text. The viable cell count in the control assay shaken at 28°C for 30 min was 8.8 x 107/ml. The values are expressedas the mean -+ SD of triplicate assays:M874B @), MGDAG
348
MATSUFUJI ET AL.
J. BIOSCI. BIOENG.,
60 h E % 5 E z z a, 3.>
TABLE 2. Effect of glyceroglycolipid M874B and wheat MGDAG and DAG on superoxide-induced cytochrome c reduction Addition
40
None M874B 20 MGDAG DAG
0 47 “C
50 “C
53 “C
Heat temperature FIG. 5. Assay mixtures containing sodA mutant cells and various concentrations of M874B in a 96-well cell culture plate were shaken at 47, 50 or 53°C for 30min and the viable cells were counted as described in the text. The viable cell counts in the control assay without glycolipids shaken at 28°C for 30 m m was 8.8 x 107/ml. The values are the mean of viable cells 2 SD of triplicate assays: none @), 1 mg/ml (m), 2.5 mg/ml (0). being also more significant against heat-induced death at 53°C than that at 50°C. In fact, cells that survived heat treatment at 53°C for 30min were constituted about 1 x 106/ml in the absenceof M874B, but about 23 X 106/ m l in the presence of 1 mg/ml of M874B and 35 x 106/ m l in the presence of 2.5 mg/ml. It must be noted that M874B was still fully protective against cell death induced by such a high temperature as 53’C. On the other hand, DGDAG and DAG, analogous to M874B, were not protective against heat-induced cell death at the concentration of 5 mg/ml, nor did ascorbate, glutathione and cu-tocopherol at 1 m M have the ability to inhibit heat-inducedcell death (data not shown). Effect of glyceroglycolipid M874B on the Fenton reacThe Fenton tion and the xanthine oxidase system reaction and xanthine-xanthine oxidase are well defined generation systems of superoxide anions and hydroxyl radicals, respectively. To clarify whether the protective activity of M874B against heat- or HzOz-induced bacterial cell death is mediated by its oxygen-radical scavenging activity, we examinedthe inhibitory effect of M874B on both of these oxygen-radical generation systems. Figure 6 shows the effects of M874B, MGDAG and DAG on
0
I
I
I
I
I
100
200
300
400
500
Concentration
(w/ml)
FIG. 6. The Fenton reaction in the presence of glycolipids at various concentrations were carried out at 25°C for 10 min. The ferrous iron remaining unconsumed were calorimetrically determined by measuring the Ass0 of the red complex with 1, IO-O-phenanthroiine. The values are expressed as a % of the AjSo in each control assay without Hz02: M874B (0), MGDAG (0) DAG (0).
Concentration (@ml) 100 250 500 100 250 500 100 250 500
Reduction of cytochrome c (nmoles/min) 1.71 1.66 1.66 1.56 1.61 1.52 1.56 1.52 1.56 1.52
hydroxyl radical generation in the Fenton reaction. The reaction was monitored by a calorimetric assay (A& of Fe2+ with 1, 10-O phenanthroline (13). Inhibition of hydroxyl radical generation was expressedas the extent of decreasein Fe2+ oxidation (or consumption) by Hz02 in the reaction m ixture. Bovine liver catalase (500 units/ml) as a positive control completely inhibited the Fez+ consumption in this reaction. It was revealed that M874B inhibited the Fez+ consumption by Hz02 in a dose-dependentmanner. Wheat MGDAG also exhibited the same activity although it was less efficient than M874B, but DAG had no inhibitory activity. On the other hand, no inhibitory effect of M874B was found on the xanthine-xanthine oxidase reaction (Table 2). The reaction was monitored by measuringthe extent of reduction of cytochrome c (with A&. No inhibition of cytochrome c reduction was observed with the addition of M874B, MGDAG or DAG, even at the maximum dose of 5OO,~g/ml, while bovine kidney SOD (500 units/ml) as a positive control completely inhibited cytocyrome c reduction. DISCUSSION We describe here that the glyceroglycolipid M874B exhibits significantly protective activity against oxidative stress resulting from exposure of bacterial cells to Hz02 or heat. Evidence has been obtained that indicates that a major cause of HzOz-induced or heat-induced cell death is the release of hydroxyl radicals or superoxide anions (5-8, 11). Therefore, we examined the scavengingactivity of M874B on hydroxyl radicals and superoxide anions using the Fenton reaction and the xanthine-xanthine oxidase systems, and demonstratedthat M874B is a hydroxyl radical scavengerbut not a superoxide anion scavenger. These results explain the protective effect of M874B on HzOz-induced cell death, but not its effect on heatinduced cell death. Thus, the protective effect of M874B against heat-induced cell death may be mediated by its scavenging effect of other reactive oxygen species than superoxideanions. In this respect, we previously reported that M874B scavengedalkyl peroxyl radicals (12). Thus, M874B is capable of quenching at least alkyl peroxyl radicals and hydroxyl radicals. Some positive control experiments demonstrated that M874B was similar to LYtocopherol in its protective effects against tert-BHPinduced cell death or lipid peroxidation (12) and HzOzinduced cell death (Table 1). It must be noted that the activity of M874B was comparableto that of cu-tocopherol and cY-tocopherolis one of the strongest known hydroxyl radical scavengersamong natural antioxidants. However,
VOL.
GLYCEROGLYCOLIPID
89, 2000
M874B is different from a-tocopherol in its reductivity of 1,1-diphenyl-2-picrylhydrazyl which is rendered colourless by this activity exerted by most antioxidants (18). M874B did not exhibit reductivity on 1,1-diphenyl-Z picrylhydrazyl while a-tocopherol did. It is noteworthy finding that wheat MGDAG and other bacterial MGDAG (S365B) tested previously (12) and in this study, but not wheat DGDAG and synthetic DAG mimic M874B in their antioxidant properties. This means that such unique antioxidative properties are characteristic of the structure of a monogalactosyldiacylglycerol. Thus, the addition or removal of one galactosyl residue to or from M874B causes complete loss of its antioxidative properties even though the compounds may differ only in one diacyl group. In this respect, wheat MGDAG contains mainly Iinoleic acid (18 : 3); while M874B contains a major anteiso Cl5 acid (12-methyltetradecanoic acid) (12). Further detailed study on the structure-activity relationships remains to be performed, especially concerning both the constituents of sugar and fatty acid. Our finding that monogalactosyldiacylglycerols including M874B are unique antioxidants may provide an important clue in the resolution of the functional role of the membrane lipid component MGDAG in cell biology. ACKNOWLEDGMENT We would like to thank Dr. Peter Setlow, Department of Biochemistry, School of Medicine, University of Connecticut Health Center, for his kind gift of /catA mutant PS2488, so&l mutant PS2495 and parental strain PS832. REFERENCES 1. Halliwell, B. and Gutteridge, J. M.: Oxygen toxicity, oxygen radicals, transition metals and desease. Biochem. J., 219, l-14 (1984). 2. MacCord, J. M. and Firdovich, I.: Superoxide dismutase: an enzymatic function for erythrocuprein (hemocuprein). J. Biol. Chem., 244, 6049-6055 (1969). 3. Asada, K., Kanematsu, S., Okada, S., and Hayakawa, T.: Chemical and biochemical aspects of superoxide and superoxide dismutase, p. 136-153. Elsevier, New York (1980). 4. Kaput, J., Goltz, S., and Blobel, G.: Nucleotide sequence of the yeast nuclear gene for cytochrome c peroxidase precursor. Functional implications of the pre sequence for protein transport into mitochondria. J. Biol. Chem., 257, 15054-15058
M874B PROTECTIVE AGAINST HEAT AND HzOz STRESS
349
(1982). 5. Benov, L. and Fridovich, I.: A superoxide dismutase mimic protects sod4 sodB Escherichia co/i against aerobic heating and stationary phase death. Arch. Biochem. Biophys., 322, 291-294 (1995). 6. Casillas-Martinez, L. and Setlow, P.: Alkyl hydroperoxide reductase, catalase, MrgA and superoxide dismutase are not involved in resistance of Bacillus subtilb spores heat or oxidizing agents. J. Bacterial., 179, 7420-7425 (1997). 7. Brandi. G.. Cattabeni. F.. Albano. A.. and Cantoi. 0.: Role of hydroxyl radicals ‘in kscherichia coli killing induced by hydrogen peroxide. Free Radic. Res. Commun., 6, 47-55 (1989). 8. Imlay, J. and Linn, S.: Mutagenesis and stress responses induced in Escherichia coli by hydrogen peroxide. J. Bacterial., 169, 2967-2976 (1988). 9. Lee, P. C., Bother, B. R., and Ames, B. N.: AppppA, heat shock, stress, and cell oxidation. Proc. Natl. Acad. Sci. USA, 80, 7496-7500 (1983). 10. Miller, M. J., Xuong, N. H., and Geiduscbek, E. P.: Quantitative analysis of the heat shock response of Saccharomyces cerevisiae. J. Bacterial., 151, 311-327 (1982). 11. Benov, L. and Fridovich, I.: Superoxide dismutase protects against aerobic heat shock in Escherichia coli. J. Bacterial., 177, 3344-3346 (1995). 12. Matsufuji, M., Taguchi, K., Inagaki, M., Higuchi, R., Ohta, S., and Yoshimoto, A.: Glyceroglycolipids preventing tertbutylhydroperoxide-induced cell death from Microbacterium sp. and Corynebacterium aquaticum strains. J. Biosci. Bioeng., 89, 170-175 (2000). 13. Reddy, A. Ch. P. and Lokesb, B. R.: Studies on the inhibitory effects of curcumin and eugenol on the formation of reactive oxygen species and the oxidation of ferrous iron. Mol. Cell. Biochem., 137, l-8 (1994). 14. Flohe, L. and Otting, H.: Superoxide dismutase assays. Methods Enzymol., 105, 93-104 (1981). 15. Bol, D. K. and Yasbin, R. E.: Characterization of an inducible oxidative stress system in BaciNus subtilis. J. Bacterial., 172, 3503-3506 (1990). 16. Davidson, J. F., Whyte, B., Bissinger, P. H., and Schiestl, R. H.: Oxidative stress is involved in heat-induced cell death in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA, 93, 5116-5121 (1996). 17. Storz, G., Jacobsou, F. S., Tartaglia, L. A., Morgan, R. W., Silveira, L. A., and Ames, B. N.: An alkyl hydroperoxide reductase induced by oxidative stress in Salmonella typhimurium and Escherichia coli: genetic characteristics and cloning of aph. J. Bacteriol., 171, 2049-2055 (1989). 18. Blois, M. S.: Antioxidant determinations by the use of a stable free radical. Science, 181, 1199-1200 (1958).
JOURNAL OF BIOSCIENCE AND
Protective Effects of Bacterial G lyceroglycolipid M874B against Cell Death Caused by Exposure to Heat and Hydrogen Peroxide MOTOKO MATSUFUJI,‘* YASUNORI NAGAh4ATSU,2 AND AKIHIRO YOSHIMOT02 Central Research Laboratories, Mercian Corporation, 9-l Johnan I-chome, Fujisawa 251-0057 and Faculty of Applied Biological Science, Hiroshima University, 4-4 Kagamiyama I-chome, Higashi-Hiroshima 739-8526,= Japan Received 1 November 1999IAccepted 11 January 2000
It was revealed by bioassay using sodA and katA mutants of Bacillus subtilis that the bacterial monogalactosyldiacylglycerol M874B, previously characterized as an alkyl peroxyl radical scavenger, was also capable of protecting cells from death caused by heating and exogenous HzOz. Chemical assaysusing the Fenton reaction and xanthine-xanthine oxidase revealed that M874B could quench hydroxyl radicals but not superoxlde anions. Wheat monogalactosyldiacylglycerol, but neither digalactosyldiacylglycerol nor synthetic diacylglycerol, also had the same activities as those of M874B, although it was less efficient than M874B. These results suggest that monogalactosyldiacylglycerols such as M874B are a new type of oxygen radical scavengerscapable of quenching some reactive oxygen species. [Key words: glyceroglycolipid, monogalactosyldiacylglycerol, gen peroxide, sodA mutant, katA mutant, Bacillus subtilis] Aerobic living cells possess a defense mechanism against oxidative damage caused by reactive oxygen species that are generated during normal aerobic metabolism. Short-lived reactive oxygen species, including superoxide anions, Hz02 and hydroxyl radicals, are noted for their high reactivity and ability to cause damage to proteins, lipid membranes and DNA (1). Enzymatic digestion of these reactive oxygen species is catalyzed by superoxide dismutase (SOD), catalase and cytochrome c peroxidase (2-4). Loss of SOD or catalase in aerobically grown cells inhibits cell growth and causes stationaryphase death (5), and hypersensitivity to superoxide (6), hydroxyl radicals or H202 (7, 8). Most microbial cells are also sensitive to sudden heat exposure. The major factors causing cell death after heat exposure are still unknown although a lethal temperature shift induces many heat-shocked proteins (HSPs) (9, 10). Recently, Benov and Fridovich reported that superoxide anion was a major cause of cell death in both the aerobic heat shock and stationary-phase of growth (5, 11). Cell death has also been reported to be caused by H20z exposure (7, S), and at least one of the major causes is hydroxylradical-induced cell death. The activity of scavenging superoxide or hydroxyl radicals is measured as that of the loss of cell viability following exposure to heat or H202. In this case, the use of SOD-deficient and catalasedeficient mutants of bacteria is beneficial for the bioassays since both the mutants are highly sensitive to heat and H202, respectively. We reported previously that a glyceroglycolipid, M874B [1,2-di-O-(12-methyltetradecanoyl)-3-O-@-galactopyranosyl-sn-glycerol], isolated from Microbacterium sp. strain M874, had a quenching effect on alkyl peroxyl radicals (12) and thus were prompted to study its antioxidant activity against other reactive oxygen species. To elucidate this, we used bioassays employing the sodA mutant PS2495 and katA mutant PS2488 of B. subtilis, which were isolated and characterized by Setlow for their hypersensitivity to heat exposure which induces
oxygen radical scavenger, heat shock, hydro-
generation of superoxide anions and H202, respectively (6). In this paper, we report that M874B has a significant protective activity against cell death caused by exposure to heat or HzOz. We also examined the scavenging activity of M874B on superoxide and hydroxyl radicals by chemical assays using the xanthine-xanthine oxidase and Fenton reaction systems, respectively. It was ascertained that M874B had at least a pronounced quenching activity on hydroxyl radicals released from Hz02 although its inhibitory effect on the xanthine-xanthine oxidase system was not defined. MATERIALS
AND METHODS
Chemicals Wheat flour monogalactosyldiacylglycerol (MGDAG), wheat flour digalactosyldiacylglycerol (DGDAG) and bovine kidney superoxide dismutase were obtained from Funakoshi Chemicals (Tokyo). Glutathione, a-tocopherol, ascorbic acid, chloramphenicol (CM), 1,2-dimyristoyl-rat-glycerol (Cl4 : 0) (DAG), xanthine, xanthine oxidase, 1,10-O phenanthroline, cytochrome c from horse heart, and bovine liver catalase were purchased from Sigma Chemicals (St. Louis, MO, USA). Glyceroglycolipid M874B was obtained as described previously (14). The glycolipid samples were dissolved or suspended in sterile phosphate-buffered saline (PBS) by sonication and added to the reaction mixture. Microorganisms and cultivation Mutant strains PS2495 (sodA, Cmr) and PS2488 (katA, Cmr) and the corresponding parent strain PS832 of Bacillus subtilis (6) were used for this experiment. The bacteria were maintained on Luria-Bertani (LB) agar (log of bacto-tryptone, 5 g of yeast extract, 10 g of NaCl, 15 g of agar in 1 I of deionized water; pH 7.4) supplemented or not supplemented with CM (3 pg/ml). The bacteria were grown at 28°C for 18 h in a 250-ml Erlenmeyer flask containing 30ml of LB medium with or without CM on a rotary shaker (220 rpm). All the 18-h cultures of the parent strain and mutant sodA or katA strains showed an optical density of about 1 .O-at 660 nm and were used for the bioassays without dilution, unless otherwise stated.
* Corresponding author. 345
346
MATSUFUJI ET AL.
Heat exposure test Aliquots (100~1) of the 18-h sodA mutant PS2495cultures were poured into a 96-well cell culture plate (NUNC) and 20 ~1 each of glycolipid solution at various concentrations or sterile PBS (as control) was added. The plate was immediately shaken at the indicated temperature on a plate shaker, M icrotherm (CAMLAB, USA), for an appropriate time and the cultures were then diluted serially IO-fold with sterile PBS, followed by plating to observe for colony formation on a YMPG agar plate (10 g of glucose, 3 g of yeast extract, 3 g of malt extract, 5 g of Polypepton and 15 g of agar; pH 7.0). The colony assay was performed in triplicate and the colonies were counted after 20 h incubation at 28°C. The viable counts are representedby the means of triplicate assays. A comparative study of heat sensitivity between the parent strain PS863 and the sodA mutant PS2495 was performed using 18-h cultures in the same manner as described above. Aliquots (100 ~1) of the 18-h HzOz exposure test katA mutant PS2488 culture and 20 ~1 each of the glycolipid solutions at various concentrations or sterile PBS (as control) were poured into a 96-well cell culture plate. To each of the m ixtures was added 20 ~1 of 1 m M HzOz and the plate was incubated on a plate shaker, M icrotherm, at 28°C for an appropriate time. After serial IO-fold dilution with sterile PBS, 100/-11each of the diluent was plated on a YMPG agar plate. Colonies were counted after incubation for 20 h at 28°C. The colony count are represented by the means of triplicate assays. In the HzOz dose-responseexperiment with the parent strain PS832 and the katA mutant PS2488, 100~1 aliquots of both the 18-h cultures were added to a 96-well cell culture plate and 20 ~1 of Hz02 solution at various concentrations was added to each. The plate was shaken at 28°C for 1 h and the colonies were counted in the same manner as described above. Chemical assays for superoxide and hydroxyl radicals
Xanthine-xanthine oxidase was used as the superoxideanion generating system and the reaction was monitored by measuring the extent of reduction of cytochrome c (13). The assay m ixture contained 50 P M xanthine, various amounts of glycolipid, 20pM of cytochrome c, 100P M of EDTA and 1.5 units of xantine oxidase in 3 m l of 50mM phosphate buffer (pH 7.8). These constituents were directly added to a 3-ml cuvette and the reaction was initiated by adding xanthine oxidase. The reaction m ixture was allowed to stand at 25°C for 5 m in and the extent of cytochrome c reduction by the superoxide anions generated was recorded by the absorbance at 550nm. Superoxide anion production was expressed as nmoles of cytochrome c reduced/min in the reaction m ixture. The Fenton reaction was used to generate hydroxyl radicals. The assay m ixture contained 2 m M of Hz02 1 m M of ferrous sulfate, various amounts of glycolipid and 5 m M of NaCl in a total volume of 1 m l (pH 7.0). The reaction was started by adding Hz02 and allowed to proceed at 25°C for IOmin. The oxidation of ferrous iron by HzOz to generate hydroxyl radicals was monitored by a calorimetric assay of the amount of unconsumed ferrous iron (14). Thus, at the end of the reaction, 1 m l of 1, IO-0phenanthroline aqueous solution was added to the assay m ixture. After allowing the m ixture to stand for 10min at room temperature, the absorbance of the red complex was measuredat 510 nm.
J. BIOSCI. BIOENG.,
RESULTS Effect of the glyceroglycolipid M874B on H202-induced cell death B. subtilis produces some major catalases
which are products of katA and katE. The katA appears to be the major catalasein growing cells and plays a significant role in the resistance of growing cells to HzOz (15-17). The katX product was produced even in the spores. Loss of catalaseresults in increasedsensitivity to Hz02 (18). Figure 1 illustrates the hypersensitivity of the katA mutant strain PS2488 to HzOz exposure as compared with the Hz02 sensitivity of the katA mutant PS2488and the parent strain PS832. Cells that survived in the presenceof exogenousHzOz were enumeratedby the colony formation assay. The katA mutant cells were hypersensitive to Hz02 and the cell survival was decreasedto approximately one-fourth by exposure to even as low as 1 m M of HzOz; the cells of the parent strain were, however, resistant to even over 1 O O m M of H202. We thus examined the protective effect of the glyceroglycolipid M874B against Hz02 stressby assayingits effect of inhibiting HzOz-induced death of katA mutant PS2488 cells. Glyceroglycolipid M874B and wheat flour MGDAG, both known to be protective against damage induced by alkyl peroxyl radicals, namely, tert-butylhydroperoxideinduced cell death (12), were also demonstrated to have a protective effect against HzOz-induced cell death (Fig. 2). The activity of M874B was dose-dependentand greater than that of wheat flour MGDAG. M874B exhibited distinct protective activity against HzOz-induced damage at the concentration of 0.05 mg/ml, whereas wheat flour MGDAG exhibited approximately the samelevel of activity at the concentration of 1.Omg/ml or more. The doseresponsecurve of this activity of M874B was almost the same as that of its quenching activity against alkyl peroxyl radicals (12). Table 1 shows the effects of some related glycerolipids and antioxidants, catalase and SOD on HzOz-induced death. Wheat flour DGDAG and DAG had no protective effect at either the concentrations of 1 mg/ml or 5 mg/ml, suggestingthat at least the structure of MGDAG was required for the protective activity. Antioxidants such as ascorbate at 1 m M were not proI
0
1 10 H,O, concentration (mM)
100
FIG. 1. Assay mixtures containing /carA mutant cells or parent cells and various concentrations of Hz02 in a 96-well cell culture plate were incubated at 28’C for 1 h and the viable cells were counted as described in the text. The value was expressedas a % of the viable cell count in each control assay without Hz02. The viable cell count of strains PS2488 and PS832 in the control assay without Hz02 were 1.26 x 107/ml and 1.57 x lO’/ml, respectively: PS2488 (0), PS832 (0).
VOL.
GLYCEROGLYCOLIPID
89, 2000
0
0.05 0.1 0.25 0.5 1 2.5 Concentration (mg/ml)
5
FIG. 2. Assay m ixtures containing k&t mutantcells,1 mM Hz02 of M874B or MGDAG in a 96-we11cell and various concentrations cultureplatewereincubated at 28°C for 1 h andthe viablecellswere countedas described in the text. The values are the m e a n+SD of triplicateassays: M874B @I), MGDAG (B).
tective. O n the other hand, a-tocopherol exhibited a pronounced activity of inhibiting H20z-induceddeath at 1 m M . However, this activity of a-tocopherol was somewhat lower than that of M874B as it was comparableto that of 0.7 m M less of M874B (0.5 mg/ml) when compared on a m o lar basis. Exogenouscatalaseexerted complete protection against damageinduced by 1 m M Hz02 when it was added at a high concentration of 5OOU/ml in this assay, while SOD was not protective against HzOz-induceddeath even when added in excess. Effect of glyceroglycolipid M874B on heat-induced death Superoxideanions are one of the m a jor causes of heat lability in aerobically grown bacteria (5, 6, 11, 15). Therefore, loss of viability by heating is more significant in sod-deficientmutant cells than in sod-containing parent cells of bacteria. Heat sensitivity of the so&l mutant strain PS2495and parent strain PS832of B. subtilis was examinedby the colony formation assayof cells that heat treatment at temperaturesranging from 40°C to 53°C. The results are shown in F ig. 3. Viable counts of neither strain decreasedsignificantly following heating at 45°C or less for 30 m in, but they rapidly decreasedby heating at 47°C or more. It was evident that the sodA mutant was much more sensitive to heat-induceddeath than the parent PS832strain. The most significant differTABLE
Addition
M874B PROTECTIVE AGAINST HEAT AND H202 STRESS
40 42
44
347
46 48 50 52 54
Temperature
(“C )
FIG. 3. Assay mixtures containing sodA mutant cells or parent cells in a %-weII cell culture plate were shaken at 28,40,45,47,50 or 53°C for 30 min and the viable cells were counted as described in the text. The values were expressedas a % of the viable cell count *SD in each control assay shaken at 28OC. The viable cell counts of PS2495 and PS832 in the control assay shaken at 28°C were 8.16 x 107/ml and 10.8 x 107/ml, respectively: so&l mutant PS2495(0), parental PS832 (0).
ence in viability between the two strains was observed when the bacterial cells were heat-treatedat 50°C. Benov and Fridovich (11) reported based on their experimental results using E. coli that the differencein heat sensitivity between the sodA sodB mutant strain and the parent strain was most pronouncedwhen the cells were exposed to 48°C which corresponded to the temperature at which cellular SOD in the bacteria is inactivated. In order to determine the ability of glyceroglycolipids to protect cells from heat-induceddeath, we assayedthe cell viability when the sodA mutant PS2495 cells were exposed to 50°C for 3 0 m in in the presence of glyceroglycolipids. F igure 4 shows the effect of M874B and wheat flour MGDAG and DGDAG on the heatinduced cell death. It was obvious that M874B and MGDAG, but not DGDAG, had a pronounced protective effect against heat-inducedcell death. The activity of M874B was dose-dependentand more marked than that of MGDAG. Further experiments with heat exposure demonstrated that the protective effect of M874B was more marked when the cells were exposedto the higher temperature of 50°C or 53°C than to 47°C (Fig. 5),
1. Effects of some related glyceroglycolipids and antioxidants on HzOz-induced cell death and heat-induced cell death Effect on HrOr-induced Heat-induced Concentration cell death cell death
(viable cell counts: x lW/ml) 28.1 22.0 1 mg/ml 50.6 44.4 5 mg/ml 57.6 69.2 MGDAG 1 mg/ml 22.2 30.4 DGDAG 1 mg/ml 22.0 17.2 DAG(C14) 1 mg/ml 26.2 25.4 Ascorbate 1mM 5.0 25.8 Glutathione 1mM 12.4 a-Tocopherol 1mM 45.2 23.0 Catalase 100 units/ml 17.3 500 units/ml 61.6 28.1 Superoxide dismutase loo0 units/ml
None M874B
0
2.5 0.5 1 Concentration (mg/ml)
5
FIG. 4. Assay mixtures containing sodA mutant cells and various concentrations of glycohpid or PBS (as control) in a %-we11ceII culture plate were shaken at 50°C for 30min and the viable cells were counted as described in the text. The viable cell count in the control assay shaken at 28°C for 30 min was 8.8 x 107/ml. The values are expressedas the mean -+ SD of triplicate assays:M874B @), MGDAG
348
MATSUFUJI ET AL.
J. BIOSCI. BIOENG.,
60 h E % 5 E z z a, 3.>
TABLE 2. Effect of glyceroglycolipid M874B and wheat MGDAG and DAG on superoxide-induced cytochrome c reduction Addition
40
None M874B 20 MGDAG DAG
0 47 “C
50 “C
53 “C
Heat temperature FIG. 5. Assay mixtures containing sodA mutant cells and various concentrations of M874B in a 96-well cell culture plate were shaken at 47, 50 or 53°C for 30min and the viable cells were counted as described in the text. The viable cell counts in the control assay without glycolipids shaken at 28°C for 30 m m was 8.8 x 107/ml. The values are the mean of viable cells 2 SD of triplicate assays: none @), 1 mg/ml (m), 2.5 mg/ml (0). being also more significant against heat-induced death at 53°C than that at 50°C. In fact, cells that survived heat treatment at 53°C for 30min were constituted about 1 x 106/ml in the absenceof M874B, but about 23 X 106/ m l in the presence of 1 mg/ml of M874B and 35 x 106/ m l in the presence of 2.5 mg/ml. It must be noted that M874B was still fully protective against cell death induced by such a high temperature as 53’C. On the other hand, DGDAG and DAG, analogous to M874B, were not protective against heat-induced cell death at the concentration of 5 mg/ml, nor did ascorbate, glutathione and cu-tocopherol at 1 m M have the ability to inhibit heat-inducedcell death (data not shown). Effect of glyceroglycolipid M874B on the Fenton reacThe Fenton tion and the xanthine oxidase system reaction and xanthine-xanthine oxidase are well defined generation systems of superoxide anions and hydroxyl radicals, respectively. To clarify whether the protective activity of M874B against heat- or HzOz-induced bacterial cell death is mediated by its oxygen-radical scavenging activity, we examinedthe inhibitory effect of M874B on both of these oxygen-radical generation systems. Figure 6 shows the effects of M874B, MGDAG and DAG on
0
I
I
I
I
I
100
200
300
400
500
Concentration
(w/ml)
FIG. 6. The Fenton reaction in the presence of glycolipids at various concentrations were carried out at 25°C for 10 min. The ferrous iron remaining unconsumed were calorimetrically determined by measuring the Ass0 of the red complex with 1, IO-O-phenanthroiine. The values are expressed as a % of the AjSo in each control assay without Hz02: M874B (0), MGDAG (0) DAG (0).
Concentration (@ml) 100 250 500 100 250 500 100 250 500
Reduction of cytochrome c (nmoles/min) 1.71 1.66 1.66 1.56 1.61 1.52 1.56 1.52 1.56 1.52
hydroxyl radical generation in the Fenton reaction. The reaction was monitored by a calorimetric assay (A& of Fe2+ with 1, 10-O phenanthroline (13). Inhibition of hydroxyl radical generation was expressedas the extent of decreasein Fe2+ oxidation (or consumption) by Hz02 in the reaction m ixture. Bovine liver catalase (500 units/ml) as a positive control completely inhibited the Fez+ consumption in this reaction. It was revealed that M874B inhibited the Fez+ consumption by Hz02 in a dose-dependentmanner. Wheat MGDAG also exhibited the same activity although it was less efficient than M874B, but DAG had no inhibitory activity. On the other hand, no inhibitory effect of M874B was found on the xanthine-xanthine oxidase reaction (Table 2). The reaction was monitored by measuringthe extent of reduction of cytochrome c (with A&. No inhibition of cytochrome c reduction was observed with the addition of M874B, MGDAG or DAG, even at the maximum dose of 5OO,~g/ml, while bovine kidney SOD (500 units/ml) as a positive control completely inhibited cytocyrome c reduction. DISCUSSION We describe here that the glyceroglycolipid M874B exhibits significantly protective activity against oxidative stress resulting from exposure of bacterial cells to Hz02 or heat. Evidence has been obtained that indicates that a major cause of HzOz-induced or heat-induced cell death is the release of hydroxyl radicals or superoxide anions (5-8, 11). Therefore, we examined the scavengingactivity of M874B on hydroxyl radicals and superoxide anions using the Fenton reaction and the xanthine-xanthine oxidase systems, and demonstratedthat M874B is a hydroxyl radical scavengerbut not a superoxide anion scavenger. These results explain the protective effect of M874B on HzOz-induced cell death, but not its effect on heatinduced cell death. Thus, the protective effect of M874B against heat-induced cell death may be mediated by its scavenging effect of other reactive oxygen species than superoxideanions. In this respect, we previously reported that M874B scavengedalkyl peroxyl radicals (12). Thus, M874B is capable of quenching at least alkyl peroxyl radicals and hydroxyl radicals. Some positive control experiments demonstrated that M874B was similar to LYtocopherol in its protective effects against tert-BHPinduced cell death or lipid peroxidation (12) and HzOzinduced cell death (Table 1). It must be noted that the activity of M874B was comparableto that of cu-tocopherol and cY-tocopherolis one of the strongest known hydroxyl radical scavengersamong natural antioxidants. However,
VOL.
GLYCEROGLYCOLIPID
89, 2000
M874B is different from a-tocopherol in its reductivity of 1,1-diphenyl-2-picrylhydrazyl which is rendered colourless by this activity exerted by most antioxidants (18). M874B did not exhibit reductivity on 1,1-diphenyl-Z picrylhydrazyl while a-tocopherol did. It is noteworthy finding that wheat MGDAG and other bacterial MGDAG (S365B) tested previously (12) and in this study, but not wheat DGDAG and synthetic DAG mimic M874B in their antioxidant properties. This means that such unique antioxidative properties are characteristic of the structure of a monogalactosyldiacylglycerol. Thus, the addition or removal of one galactosyl residue to or from M874B causes complete loss of its antioxidative properties even though the compounds may differ only in one diacyl group. In this respect, wheat MGDAG contains mainly Iinoleic acid (18 : 3); while M874B contains a major anteiso Cl5 acid (12-methyltetradecanoic acid) (12). Further detailed study on the structure-activity relationships remains to be performed, especially concerning both the constituents of sugar and fatty acid. Our finding that monogalactosyldiacylglycerols including M874B are unique antioxidants may provide an important clue in the resolution of the functional role of the membrane lipid component MGDAG in cell biology. ACKNOWLEDGMENT We would like to thank Dr. Peter Setlow, Department of Biochemistry, School of Medicine, University of Connecticut Health Center, for his kind gift of /catA mutant PS2488, so&l mutant PS2495 and parental strain PS832. REFERENCES 1. Halliwell, B. and Gutteridge, J. M.: Oxygen toxicity, oxygen radicals, transition metals and desease. Biochem. J., 219, l-14 (1984). 2. MacCord, J. M. and Firdovich, I.: Superoxide dismutase: an enzymatic function for erythrocuprein (hemocuprein). J. Biol. Chem., 244, 6049-6055 (1969). 3. Asada, K., Kanematsu, S., Okada, S., and Hayakawa, T.: Chemical and biochemical aspects of superoxide and superoxide dismutase, p. 136-153. Elsevier, New York (1980). 4. Kaput, J., Goltz, S., and Blobel, G.: Nucleotide sequence of the yeast nuclear gene for cytochrome c peroxidase precursor. Functional implications of the pre sequence for protein transport into mitochondria. J. Biol. Chem., 257, 15054-15058
M874B PROTECTIVE AGAINST HEAT AND HzOz STRESS
349
(1982). 5. Benov, L. and Fridovich, I.: A superoxide dismutase mimic protects sod4 sodB Escherichia co/i against aerobic heating and stationary phase death. Arch. Biochem. Biophys., 322, 291-294 (1995). 6. Casillas-Martinez, L. and Setlow, P.: Alkyl hydroperoxide reductase, catalase, MrgA and superoxide dismutase are not involved in resistance of Bacillus subtilb spores heat or oxidizing agents. J. Bacterial., 179, 7420-7425 (1997). 7. Brandi. G.. Cattabeni. F.. Albano. A.. and Cantoi. 0.: Role of hydroxyl radicals ‘in kscherichia coli killing induced by hydrogen peroxide. Free Radic. Res. Commun., 6, 47-55 (1989). 8. Imlay, J. and Linn, S.: Mutagenesis and stress responses induced in Escherichia coli by hydrogen peroxide. J. Bacterial., 169, 2967-2976 (1988). 9. Lee, P. C., Bother, B. R., and Ames, B. N.: AppppA, heat shock, stress, and cell oxidation. Proc. Natl. Acad. Sci. USA, 80, 7496-7500 (1983). 10. Miller, M. J., Xuong, N. H., and Geiduscbek, E. P.: Quantitative analysis of the heat shock response of Saccharomyces cerevisiae. J. Bacterial., 151, 311-327 (1982). 11. Benov, L. and Fridovich, I.: Superoxide dismutase protects against aerobic heat shock in Escherichia coli. J. Bacterial., 177, 3344-3346 (1995). 12. Matsufuji, M., Taguchi, K., Inagaki, M., Higuchi, R., Ohta, S., and Yoshimoto, A.: Glyceroglycolipids preventing tertbutylhydroperoxide-induced cell death from Microbacterium sp. and Corynebacterium aquaticum strains. J. Biosci. Bioeng., 89, 170-175 (2000). 13. Reddy, A. Ch. P. and Lokesb, B. R.: Studies on the inhibitory effects of curcumin and eugenol on the formation of reactive oxygen species and the oxidation of ferrous iron. Mol. Cell. Biochem., 137, l-8 (1994). 14. Flohe, L. and Otting, H.: Superoxide dismutase assays. Methods Enzymol., 105, 93-104 (1981). 15. Bol, D. K. and Yasbin, R. E.: Characterization of an inducible oxidative stress system in BaciNus subtilis. J. Bacterial., 172, 3503-3506 (1990). 16. Davidson, J. F., Whyte, B., Bissinger, P. H., and Schiestl, R. H.: Oxidative stress is involved in heat-induced cell death in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA, 93, 5116-5121 (1996). 17. Storz, G., Jacobsou, F. S., Tartaglia, L. A., Morgan, R. W., Silveira, L. A., and Ames, B. N.: An alkyl hydroperoxide reductase induced by oxidative stress in Salmonella typhimurium and Escherichia coli: genetic characteristics and cloning of aph. J. Bacteriol., 171, 2049-2055 (1989). 18. Blois, M. S.: Antioxidant determinations by the use of a stable free radical. Science, 181, 1199-1200 (1958).