- Email: [email protected]
Reductive repression in Escherichia coli K-12 is mediated by oxygen radicals
ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS Vol. 250, No. 1, October, pp. 54-62,1986
Reductive Repression in Escherichia co/i K-l 2 Is Mediated by Oxygen Radicals’ R. HERTZ Department
AND
J. BAR-TANA
of Biochemistrg Hebrew University-Hadassah P.O. Box 1172, Jerusalem 91010, Israel
Received September
Medical School,
3,1985, and in revised form June 17,1986
Cyclic AMP (CAMP) content and the expression of CAMP-dependent phenotypes were positively correlated with respiration capacity in respiration-deficient mutants of Escherichia coli K-12 (“reductive repression,” R. Hertz, and J. Bar-Tana, (1982) Arch. Biochem. Biophgs. 213,193-199). Reductive repression in respiration-deficient mutants could not be accounted for by respective changes in either the energy charge of adenine nucleotides or the redox state of pyridine nucleotides but could be ascribed to an increased formation of oxygen radicals under conditions of limited respiration. Scavengers of superoxide radicals eliminated reductive repression in respiration-deficient mutants with a concomitant increase in CAMP content. Such scavengers also effected a partial escape from permanent glucose catabolite repression, thus indicating a possible role played by oxygen radicals in both repression modes. o 19% Academic PEW, IN.
Cyclic AMP (CAMP)’ content and the expression of CAMP-dependent phenotypes in Escherichia coli K-12 were reported recently to be positively correlated with the respiration capacity of the culture. Thus, growing ubiquinone (ubi A, ubi D)- or protoporphyrin (hem A)-deficient mutants in the presence of variable concentrations of p-hydroxybenzoate (p-HBA) or F-aminolevulinate (&ALA), respectively, resulted in a respective gradual increase in the ubiquinone or cytochrome content with a concomitant recovery of respiration capacity, CAMP content, P-galactosidase activity, and flagellation. Repression of CAMP-dependent phenotypes under conditions of limited electron transport-mediated respiration (“reductive repression”) and modulation of the CAMP content as a function
of respiration rate were the results of modulation of the adenylate cyclase activity (1). The present report evaluates the possible role played by the phosphate potential, the redox state of pyridine nucleotides, and the generation of oxygen radicals in mediating reductive repression in E. coli K-12. MATERIALS
Copyright All rights
0 1986 by Academic Press, Inc. of reproduction in any form reserved.
METHODS
Respiration-deficient, @‘-galactosidase constitutive strains HB 101 (ubi A, lac”) and HB 111 (hem A, lac’) were derived as previously described (1). The ubi A allele affects the prenylation step of the ubiquinone synthetic pathway by increasing the K, for p-hydroxybenzoic acid (2). The hem A allele affects the protoporphyrin synthetic pathway by eliminating the biosynthesis of &aminolevulinic acid (3). Bacteria were grown aerobically in 0.8% nutrient broth in 0.05 M potassium phosphate buffer (pH 7.0) or in H minimal medium (4), with additions as described. Cultures were grown to midlog phase with continuous gyratory shaking at 30°C. Bacterial density was determined by measuring the optical density at 650 nm. Respiration rate of washed cultures was measured in the presence of 20 mM succinate in H minimal me-
i Supported by the Herta Schnap Foundation for Cancer Research. 2 Abbreviations used: pHBA, p-hydroxybenzoic acid, &ALA, &aminolevulinic acid, CAMP, cyclic AMP, PMS, phenazine methosulfate. 0003-9861/86 $3.00
AND
54
REDUCTION
REPRESSION
MEDIATED
BY OXYGEN
dium using an oxygen electrode. Respiration capacity of electron transport-deficient strains was expressed as percentage of the respiration rate of the respective deficient strain grown in the presence of saturating concentrations of its limiting precursor, p-HBA or 6ALA, respectively. The 100% respiration rate for the HB 101 (ubi A, la&) strain grown in the presence of saturating pHBA amounted to 150 + 10 nmol Oz/mg protein/min. The 100% respiration rate for the HB 111 (hem A, lace) strain grown in the presence of saturating &ALA amounted to 300 2 20 nmol Oz/mg protein/min. Cyclic AMP content was determined as described previously (1). Adenine nucleotides were extracted (5, 6) and determined by the luciferase bioluminescence assay as described by Swedes et al. (7). Pyridine nucleotides were extracted and determined as described by Wimpenny and Firth (8). All nucleotides were extracted with 90-95% efficiency. P-Galactosidase activity was assayed at 28°C in toluene-treated cells as described by Pardee et al. (9). Cultures assayed for catalase were washed and suspended in 0.05 M phosphate buffer (pH 7.0) and were sonicated three times consecutively for 15 s. The sonicate was centrifuged at 37,OOOgfor 30 min and the activity was determined in the supernatant. The reaction mixture for catalase contained 0.05 M of phosphate buffer (pH 7.0) and 9 mM of hydrogen peroxide in a final volume of 1.0 ml. The activity was assayed by following the initial decrease in absorbance at 240 nm at 25°C (10). a-Glycerokinase was assayed as described by Newsholme et al. (46). Cultures assayed for a-glycerokinase were washed and suspended in 0.1 M Tris-HCl buffer (pH 7.4) and were sonicated twice for 30 sec. The sonicate (5-10 mg) was incubated in the presence of 0.1 M Tris-HCl (pH 7.4), 16 mM EDTA, 22 mM NaF, 5 mM MgS04, 33 mM mercaptoethanol, 5 mM ATP, and 4 mM (U-‘“C) glycerol (2.5 &i/rmol) in a final volume of 1.0 ml. Catalase complex I was measured by following the difference spectrum in the 60@ 650 nm range, using double beam spectrophotometry (11-13). Cultures assayed for superoxide dismutase activity were washed and suspended in 0.05 M phosphate buffer (pH 7.8) containing 0.1 mM EDTA and were sonicated three times consecutively for 15 s. The sonicate was centrifuged at 27,OOOgfor 60 min, dialyzed in 0.05 M phosphate buffer (pH 7.0) containing 0.1 mM EDTA, and was again centrifuged at 30,OOOg for 60 min. Superoxide dismutase activity was determined in the supernatant by assaying the extent of inhibition of nitrobluetetrazolium reduction in the presence of added bacterial protein (44). Cu-(salicylate)z was prepared according to (14). Cu(1ysine)z was prepared by refluxing 1.8 g of L-lysine with 16.0 g of CuO in 200 ml of Ha0 for 120 min. The cooled mixture was diluted into water, filtered, allowed to crystallize for 48 h, and the crystals were washed in acetone and dried in air.
RADICALS
Significance test.
IN Eschetichia
coli
was analyzed by the Mann-Whitney
55 U
RESULTS
Adenine and Pyridine Nucleotides as Mediators of Reductive Repression The ATP, ADP, AMP, and CAMP contents and the /3-galactosidase activity as a function of respiration capacity in E. coli K-12 were evaluated by growing the ubi A strain in the presence of increasing concentrations of p-HBA added to the growth medium. As shown in Table I, increasing the respiration capacity within the range of 20-100% was linearly correlated with the ATP content as well as with the CAMP content and the P-galactosidase activity as previously reported (1). However, the ATP/ ADP ratio did not vary significantly throughout the range of 20-100% respiration capacity and the AMP content was found to be lower than 0.5 nmol/mg protein, whether in the presence or absence of added p-HBA. Similarly, no significant change in the redox charge (15) of pyridine nucleotides or phosphopyridine nucleotides was observed as a function of respiration capacity of the ubi A strain grown either in glycerol or succinate (Table II). Also, the adenine and pyridine nucleotide charges of the HB 111, hem A strain, remained unchanged as a function of &ALA added to the growth medium, under conditions of an increase in respiration capacity from 20 to 100% with a concomitant threefold increase in CAMP content (not shown). These results seem to confirm previous reports where no significant changes in the “reductive catabolic charge” [NADH/(NAD + NADH)] (15) or energy charge (16) have been observed upon limiting the rate of growth of E. coli by nutrient availability or oxygen tension. The two- to threefold increase in CAMP content and CAMP-dependent activities as a function of respiration capacity in the absence of variation in adenine and pyridine nucleotide charges seem to indicate that reductive repression in E. coli K-12 is presumably not mediated by either the energy or the reductive charges.
HERTZ
AND TABLE
BAR-TANA I
ATP AND ADP CONTENT AS A FUNCTION OF RESPIRATION CAPACITY IN A UBIQUINONE-DEFICIENT MUTANT OF E. coli K-12 Respiration capacity 23 34 51 65 86 100
8-Galactosidase 1.8 + 2.2 f 2.6 f 2.8 f 3.2 f 4.0 +
0.1 0.1* O.l* O.l* 0.2* 0.2*
CAMP 20.0 21.5 24.0 26.0 29.0 40.0
+ + + f f +
3.5 2.5 3.0 2.0 2.2* 3.0*
ATP
ADP
ATP/ADP
6.3 f 0.4 6.9 f 0.5 7.7 * 0.5* 7.2 + 0.25* 8.3 f 0.8* 7.9 zk0.7*
1.1 f 0.15 0.9 zk0.1 1.4 + 0.15 1.1 f 0.2 1.2 f 0.3 0.9 * 0.15
5.7 zk 1.1 7.7 f 1.4 5.5 Lk0.9 6.5 f 1.4 6.9 IL 2.4 8.6 A 2.2
Note. HB 101 was grown in H minimal medium supplemented with 0.2 mM L-lysine, 0.2 mM L-methionine, 0.2 fiM thiamine, 0.4% glycerol, and p-HBA as required to yield the percentage respiration capacity as stated. @Galactosidase activity (U/mg protein), CAMP (pmol/mg protein), ATP (nmol/mg protein), and ADP (nmol/ mg protein) content (mean k SD, TZ= 3) were determined as described under Materials and Methods. * Significantly different from the respective values observed at 23% respiration capacity, P < 0.05.
Oxygen Radicals as Mediators of Reductive Repression The possible role of oxygen radicals in mediating reductive repression was studied by (a) measuring the content of either hydrogen peroxide or hydrogen peroxide-induced catalase as a function of respiration capacity in electron transport-deficient mutants; (b) evaluating the CAMP content and CAMP-dependent phenotypes of the wild type in the presence of added artificial redox couples known to generate superoxides or hydrogen peroxide; (c) evaluating the extent of reductive repression in respiration-deficient strains grown in the presence of scavengers of superoxide radicals. Since hydrogen peroxide serves as an inducer of catalase in E. coli (18, 19, 44) as well as in the ubi A mutant (not shown), the steady state level of hydrogen peroxide as a function of respiration capacity could be inferred from the catalase activity observed in respiration-deficient strains in the presence or absence of added p-HBA. Thus, the 250% increase in catalase activity of the ubi A strain grown under conditions of limiting p-HBA (Table III) may indicate a higher intracellular content of hydrogen peroxide generated under conditions of reductive repression. The direct determination of the intracellular hydrogen peroxide content of the HB 101 strain was also at-
tempted by following spectrophotometritally the steady state level of catalase complex I in the presence of added nitrite or methanol (11, 12) and as a function of respiration capacity of the culture. However, no catalase complex I difference spectrum in the 600 to 680-nm range was observed in this strain even at highly densed cultures (20-50 ODsW), and the relatively low contribution made by catalase complex I to the difference spectrum of the soret band did not allow the spectrophotometric measurement of complex I in the 400-nm range. Also, the generation rate of extracellular hydrogen peroxide as determined by scopoletin fluorescence was lower than the sensitivity limit of the assay (17), nor did preincubation of the culture in the presence of added aminotriazole result in any detectable extracellular hydrogen peroxide. Two artificial redox couples capable of generating oxygen radical species were studied for their capacities as reductive repressors in the wild type. Paraquat was used to mediate one-electron reduction of molecular oxygen to yield the superoxide radical (18), and PMS was used to mediate either one- or two-electron reduction of molecular oxygen to yield the superoxide radical or hydrogen peroxide, respectively (20, 21). Alternatively, an increase in intracellular hydrogen peroxide was effected by inhibition of catalase by aminotriazole
REDUCTION
REPRESSION
MEDIATED
BY OXYGEN
RADICALS
IN Escherichia
coli
5’7
added to the growth medium (22). Since the respiration capacity of the ulri A mutant grown in the presence of 500 pM added pHBA is similar to that of the isogenic wild type (1, 2), the repressive potential of the various artificial redox couples was evaluated in the mutant grown in the presence of p-HBA. As shown in Table IV ,8-galactosidase activity was repressed by PMS, paraquat, or aminotriazole added to the growth medium at concentrations which did not affect the growth rate of the culture. The repression of /3-galactosidase activity was accompanied by a two- to threefold decrease in CAMP content and was partially restored in the presence of exogenous CAMP added to the growth medium. Similarly, PMS and paraquat repressed the P-galactosidase activity of the hem A strain grown in the presence of saturating &ALA and the addition of 5 mM of CAMP to the growth medium effected a complete escape from the observed repression. Organic complexes of Cu2+ (e.g., Cu-lysine, Cu-salicylate) were reported to scavenge superoxide radicals by catalyzing the dismutation of superoxides into hydrogen peroxide (23-25). Indeed, reductive repression of either P-galactosidase (Table V, Fig. 1) or a-glycerokinase (Table V) was eliminated in the ubi A strain in the presence of Cu complexes added to the culture medium, and the increase in P-galactosidase activity induced by these scavengers was correlated with an increase in CAMP content (Table V). The escape induced by Cu complexes was observed either in nutrient broth or in minimal medium with the efficacy of the scavenger being higher in minimal medium. It is worth noting that Cu complexes did not improve the respiration capacity of the ubi A strain (Table V), and did not activate the P-galactosidase activity of HB 101 grown in the presence of saturating p-HBA (Table V). Also, Cu complexes could not substitute for CAMP in a cya strain and the P-galactosidase activity was nil in the presence or absence of the added Cu complex. Hence, escape from reductive repression exerted by these scavengers could not be ascribed to reversion of the respiratory deficiency of the ubi
58
HERTZ
AND
BAR-TANA
TABLE CATALASE
AND SUPEROXIDE DISMUTASE ACTIVITIES AS A FUNCTION OF RESPIRATION IN A UBIQUINONE-DEFICIENT MUTANT OF E. coli K-12
Addition to growth medium p-HBA p-HBA
III
Respiration capacity
5 /.LM 500 HIM
Catalase
activity
25
50.8 f 2.1
100
19.1 f 2.6
CAPACITY
Superoxide dismutase activity 7.2 f 0.5 6.5 f 0.6
Note. HB 101 was grown in H minimal medium supplemented with 0.2 mM L-lysine, 0.2 mM L-methionine, 0.2 pM thiamine, 0.4% glycerol, and pHBA as stated to yield the percentage respiration capacity as stated. Catalase activity (nmol H202 decomposed/mg protein/min) was determined as described under Materials and Methods (data are means f SD; n = 7). Superoxide dismutase activity is expressed as the amount of protein (pg) which inhibited by 50% the reduction of nitrobiue tetrazolium (44) (mean f SD, n = 3).
A mutant, to nonspecific activation of pgalactosidase, or to bypassing CAMP-dependent promotion. The presence of either Cu-salicylate or Cu-lysine effected also an escape of the HB 101 strain from permanent glucose catab-
olite repression exerted by growth in glucase. Thus, the P-galactosidase activity of HB 101 grown in minimal medium in the presence of saturating p-HBA amounted to 4.6 and 1.5 U/mg protein in the presence of either 0.4% glycerol or 0.4% glucose as
TABLE REDUCTIVE
REPRESSION
IN
E.
coli K-12
EXERTED
IV BY PMS,
PARAQUAT,
AND AMINOTRIAZOLE
@-Galactosidase Addition
to growth
medium
HB 101 None PMS 20 ELM Paraquat 80 pM Paraquat 160 PM Aminotriazole 0.1 M Paraquat 160 pM + aminotriazole 0.1 M HB 111 None PMS 26 PM PMS 52 /.hM Paraquat 320
@M
-CAMP
+cAMP
CAMP content
0.3 0.2* 0.3* 0.3* 0.1*
4.3 * 0.3 3.3 f 0.3** 3.0 z!Y0.2** 2.0 + 0.2 3.5 f 0.2**
12.5 f 3.0 4.7 f 0.5* 6.5 + 0.5*
1.3 f 0.1*
2.4 2 0.2**
4.7 f 0.6*
2.2 + 0.1 1.5 + 0.1*
2.5 + 0.2 2.7 + 0.2**
1.6 + O.l*
2.9 + 0.1**
1.6 + 0.2*
2.5 + 0.2**
3.9 f 2.1 k 2.0 + 1.6 f 1.9 f
5.0 + 0.6*
Note HB 101 was grown in nutrient broth supplemented with 0.2 PM thiamine and 500 PM p-HBA, in the presence or absence of 5 mM added CAMP as stated. HB 111 was grown in nutrient broth supplemented with capacity of all grown cultures amounted to 0.4% glycerol, 0.2 pM thiamine, and 10 pM &ALA. Respiration 100%. Three-four hours before they were harvested the cultures were diluted into fresh medium and either PMS, paraquat, or aminotriazole was added to the growth medium as stated. P-Galactosidase activity (U/mg protein) was determined as described under Materials and Methods (mean f SD, n = 3). The intracellular CAMP content (pmoi/mg protein) was determined in cultures grown in the absence of CAMP added to the growth medium as described under Materials and Methods (mean f SD, n = 3). * Significantly different from the respective no addition value, P i 0.05. ** Significantly different from the respective (-CAMP) value, P < 0.05.
REDUCTION
REPRESSION
MEDIATED
BY OXYGEN TABLE
RADICALS
IN Escherichia
coli
59
V
ESCAPEFROMREDUCTIVEREPRESSIONEXERTEDBYORGANIC CU COMPLEXES Addition to growth medium
Respiration capacity
None Cu-salicylate 10 PM Cu-salicylate 125 FM Cu-salicylate 250 PM Cu-salicylate 1000 PM p-HBA 500 PM p-HBA 500 /LM + Cusalicylate 500 PM
CAMP P-Galactosidase
10
10 10 10
1.0 t 0.1
ol-Glycerokinase 8.5 f 0.8
content 9.5 * 2.0
12.8 f1.3* 0.15 * 0.15 * 0.2* 0.2*
10 100
2.3 f 3.2 + 3.8 f 4.0 +
100
4.6 I+ 0.3*
12.7 f 1.5
17.5 f 2.0* 14.5 f 0.8*
26.0 Z+I4.0* 26.0 f 5.0*
15.3 IL 0.8*
23.0 + 4.0*
Note. HB 101 was grown either in nutrient broth (for P-galactosidase and CAMP determinations) supplemented with 0.2 FM thiamine, p-HBA, and Cu-salicylate as stated, or in H minimal medium (for o-glycerokinase determination) supplemented with 0.2 mM L-lysine, 0.2 mM L-methionine, 0.2 PM thiamine, 0.4% glycerol, pHBA, and Cu-salicylate as stated. Respiration capacity (%),@galactosidase (U/mg protein), a-glycerokinase (mU/mg protein), and CAMP content (pmol/mg protein) were determined as described under Materials and Methods (mean + SD, 12 = 3). * Significantly different from the respective no addition value. P < 0.05.
carbon sources, respectively. However, in the presence of 50 PM of Cu-salicylate added to the growth medium, the activity observed in either glycerol or glucose media amounted to 4.2 U/mg protein. The escape from glucose catabolite repression exerted
by Cu-complexes was also observed in a CAMP phosphodiesterase-deficient mutant (47) and could not be ascribed, therefore, to inactivation of the phosphodiesterase by organic Cu complexes with a concomitant increase in CAMP content. DISCUSSION
3-EB l-----l 2.25
0.063
FIG. 1. Escape from reductive repression exerted by Cu-lysine: HBlOl was grown in nutrient broth supplemented with 0.2 FM thiamine and Cu-lysine as indicated. fl-Galactosidase activity was determined as described under Methods (mean + SD, rz = 3). The pgalactosidase activity in the presence of 500 pM pHBA added to the growth medium amounted to 4.0 U/mg protein.
The role of oxygen radicals in mediating reductive repression in E. coli K-12 was indicated here by three lines of evidence: (a) Intracellular CAMP content was decreased and the /3-galactosidase activity was repressed in the HB 101 strain under conditions of growth in the presence of nonlimiting p-HBA and added oxidizing agents capable of generating oxygen radicals. The repressed activity could be restored in the presence of exogenous CAMP added to the growth medium. (b) The catalase activity was significantly higher in electron transport-deficient strains, thus indicating a higher generation rate of hydrogen peroxide under conditions of inhibition of the electron transport chain. (c) Reductive repression of p-HBA-limited, electron transport-deficient strains could be eliminated in the presence of scavengers of
60
HERTZ
AND
superoxide radicals. Hence, reductive repression of ubiquinoneor cytochromedeficient strains of E. coli K-12 requires a direct reduction of molecular oxygen into superoxide radicals mediated by reduced intermediates of the electron transport chain under conditions of inhibition of oxygen reduction into water. Overproduction of oxygen radicals under these conditions may resemble the generation of superoxide radicals in antimycin-inhibited mitochondria (43). Alternatively, reductive repression of the wild type may be effected by generating a cyanide-resistant respiratory pathway mediated by artificial redox couples (e.g., PMS, paraquat) which may compete for electrons with intermediates of the electron transport chain (44). Whatever the mechanism of superoxide production, the oxygen radicals produced under conditions of reductive repression are assumed to inhibit the adenylate cyclase with a concomitant repression of CAMP-dependent phenotypes (1). The exceptional sensitivity of the cyclase protein to inactivation by oxygen radicals and the possible interrelationship between reductive repression and oxidation-dependent “heat-shock stress” (26, 27, 45) still remain to be evaluated. The present report seems to indicate that superoxide radicals or hydroxyl radicals rather than hydrogen peroxide are implicated as effecters of reductive repression. Thus, scavenging the superoxide radicals by organic Cu complexes, while dismutating these radicals into hydrogen peroxide (24), resulted in an escape from reductive repression. Reductive repression by aminotriazole-dependent inactivation of catalase could still be ascribed to the accumulation of superoxides or hydroxyl radicals as a result of inactivation of superoxide dismutase by hydrogen peroxide (28). Since reductive repression was also observed under anaerobic growth conditions in a mutant deficient of both ubiquinone and menaquinone (29), inactivation of the cyclase may be exerted also by nonoxygen radicals which may accumulate anaerobically in the absence of a functional anaerobic electron transport. Alternatively, since cultures may not be entirely
BAR-TANA
anaerobic, reductive repression under anaerobic growth conditions could still result from the generation of oxygen radicals. The interplay between the induction of either superoxide dismutase or catalase and their possible repression under conditions of reductive repression is worth noting. Thus, the induction of the dismutase by superoxide radicals (38, 39) could be expected to increase its activity under conditions of reductive repression. On the other hand, since this enzyme appears to be controlled by permanent glucose catabolite repression (40), its expression could be expected to be repressed under conditions of reductive repression as a result of the concomitant decrease in CAMP content. Indeed, the superoxide dismutase activity remained unchanged in reductively repressed electron transport-deficient strains (Table III), presumably reflecting the antagonistic inductive-repressive effects exerted on this enzyme by oxygen radicals. In contrast to repression of the superoxide dismutase by glucose growth, catalase expression in E. coli was reported not to be modulated by permanent glucose catabolite repression nor by the CAMP content (41). Thus, the presumed increased generation of hydrogen peroxide under conditions of reductive repression in electron transportdeficient strains may effect an induction of catalase (l&19,45) unmasked by the concomitant decrease in CAMP content (Table III). Hence, the catalase activity in reductively repressed strains presumably reflects the intracellular level of its hydrogen peroxide inducer whereas the superoxide dismutase responds both to the level of its superoxide inducer and to the prevailing decreased CAMP content. It is worth noting that the relative increase in catalase activity of the ubi A mutant under limiting p-HBA as reported here (Table III) is in discrepancy with results previously reported by Hassan and Fridovich (42). The discrepancy observed may have resulted from strain, culture medium, or other experimental differences. Reductive repression is essentially similar to permanent glucose catabolite repression (30-32) with respect to the extent
REDUCTION
REPRESSION
MEDIATED
BY OXYGEN
of inhibition of the cyclase activity (5, 35), the scope and extent of repression of CAMP-promoted phenotypes (32), and the escape observed in the presence of added CAMP (33, 34). Furthermore, escape from either reductive or catabolite repression as exerted by scavengers of superoxide radicals seems to indicate an involvement of oxygen radicals in both repressive modes. The generation of superoxide radicals by the electron transport chain (36) may thus be assumed to result from glucose catabolism (37) in excess of the electron flux allowed through the limiting step of the chain, or a mutation of the electron transport chain which limits respiration capacity (1). The oxygen radicals produced by either one of these possibilities may inactivate the cyclase with a concomitant decrease in CAMP content and CAMP-promoted phenotypes, thus leading to permanent glucose catabolite or reductive repression, respectively. Hence, permanent glucose catabolite repression may be considered as a specific mode of reductive repression. ACKNOWLEDGMENT We thank Professor B. Hess, The Max Planck Institute fiir Ernahrungsphysiologie, Dortmund, Germany, for his help and advice. REFERENCES 1. HERTZ, R., AND BAR-TANA, J. (1982) Arch. Biochem Biophys. 213,193-199. 2. YOUNG, I. G., LEPPIK, R. A., HAMILTON, J. A., AND GIBSON, F. (1972) J. Bacterial. 110,18-25. 3. SASARMAN, A., SURDEANU, M., AND HORODNICEANU, T. (1968) J. BacterioL 96,1882-1884. 4. KAISER, A. D., AND HOGNESS, D. S. (1960) J. Mol. BioL 2,392-415. 5. PETERKOFSKY, A., AND GAZDAR, C. (19’73) Proc. Natl. Acad. Sci. USA 70,2149-2152. 6. CHAPMAN, A. G., FALL, L., AND ATKINSON, D. E. (1971) J. Bacterial 108,1072-1086. 7. SWEDES,J. S., SEDO, J., AND ATKINSON, D. E. (1975). J. BioL Chem. 250,6930-6938. 8. WIMPENNY, J. W. T., AND FIRTH, A. (1972) J. Bacteriol. 111, 24-32. 9. PARDEE, A. B., JACOB, F., AND MONOD, J. (1959). J. Mol. Biol. 1, 165-178.
RADICALS
IN Escherichia
coli
61
10. BEERS, R. F., AND SIZER, I. W. (1952) J. BioL Chem. 195,133-140. 11. CHANCE, B., AND OSHINO, N. (1971) Biochem. J. 122.225-233. 12. OSHINO, N., AND CHANCE, B. (1973) Arch, Biochem Biophys. 154,117-131. 13. HESS, B., KLEINHANS, H., AND SCHLUTER, H. (1970) Hoppe Seyler’s Z. PhysioL Chem. 351, 515-531. 14. DE ALVARE, L. R., GODA, K., AND KIMURA, T. (1976) B&hem. Biophys. Res. Commun. 69,687-694. 15. ANDERSEN, K. B., AND MEYENBURG, K. (1977) J. BioL Chem. 252,4151-4156. 16. WALKER-SIMMONS, M., AND ATKINSON, D. E. (1977) J. Bacterial. 130, 676-683. 17. ROOT, R. K., METCALF, J., OSHINO, N., AND CHANCE, B. (1975) J. Clin. Invest. 55, 945-955. 18. HASSAN, H. M., AND FRIDOVICH, I. (1980) J. BacterioL 141,156-163. 19. YOSHPE-PURER, Y., HENIS, Y., AND YOSHPE, J. (1977) Canad J. MicrobioL 23,84-91. 20. MCCORD, J. M., AND FRIDOVICH, I. (1970) J. Biol. Chem. 245,1374-1377. 21. NISHIKIMI, M., RAO, N. A., AND YAGI, K. (1972) Biochem. Biophys. Res. Commun. 46,849-854. 22. MARGOLIASH, E., NOVOGRODSKY, A., AND SCHEJTER, A. (1959). Biochem. J. 74,339-350. 23. WESER, U., SELLINGER, K. H., LONGFELDER, E., WERNER, W., AND STRAHLE, J. (1980) Biochim. Biophys. Acta 631,232-245. 24. WALKER, W. R., AND BEVERIDGE, S. J. (1979) Inorg Perspect. BioL Med 2, 93-125. 25. RICHTER, C., PRIPFL, T., AND WINTERHALTER, K. H. (1980) FEBS I&t. 111,95-98. 26. LEE, P. C., BOCHNER, B. R., AND AMES, B. N. (1983) Proc. NatL Acad. Sci. USA 80,7496-7500. 27. LEE, P. C., BOCHNER, B. R., AND AMES, B. N. (1983) J. BioL Chem. 258,6827-6834. 28. HASSAN, H. M. (1980) in Biological Roles of Copper (Ciba Foundation Symposium No. 79), pp. 125142, Elsevier, Amsterdam. 29. HERTZ, R., AND BAR-TANA, J. (1977) J. Bucteriol. 132,1034-1035. 30. MAGASANIK, B. (1961) Cold Spring Harbor Symp. Quant. Biol. 26,249-256. 31. MAKMAN, R. S., AND SUTHERLAND, E. W. (1965) J. BioL Chem. 240,1309-1314. 32. PASTAN, I., AND ADHYA, S. (1976) BacterioL Rev. 40,527-551. 33. ULLMANN, A., AND MONOD, J. (1968) FEBS Lett. 2, 57-60. 34. PERLMAN, R. L., GROMBRUGGHE, B., AND PASTAN, I. (1969) Nature (London) 223,810-812. 35. EPSTEIN, W. L., ROTHMAN-DENES, L. B., AND HESSE, J. (1975) Proe. NatL Acad. Sci USA 72, 2300-2304. 36. BOVERIS, A., OSHINO, N., AND CHANCE, B. (1972) Biochem. J. 128,617-630.
62
HERTZ
AND
37. WANNER, B. L., KODAIRA, R., AND NEIDHARDT, F. C. (19’78) J. Bacterial. 136,947-954. 38. HASSAN, H. M., AND FRIDOVICH, I. (1977) J. Biol. Chem. 252,7667-7672. 39. HASSAN, H. M., AND FRIDOVICH, I. (1978) J. Biol. Chem. 253,8143-8148. 40. HASSAN, H. M., AND FRIDOVICH, I. (1977) J. Batteriol. 132,505-510. 41. RICHTER, H. E., AND LOEWEN, P. C. (1982) Arch. Biochem. Biophys. 215,72-77. 42. HASSAN, H. M., AND FRIDOVICH, I. (1978) J. Biol. Chem 253,6445+X50.
BAR-TANA 43. CADENAS, E., AND BOVERIS, A. (1980) Biochem. J 188,31-37. 44. HASSAN, H. M., AND FRIDOVICH, I. (1979) Arch Biochem. Biophys. 196,385-395. 45. CHRISTMAN, M. F., MORGAN, R. W., JACOBSON, F. S., AND AMES, B. N. (1985) Cell 41,753-762. 46. NEWSHOLME, E. A., ROBINSON, J., AND TAYLOR, K. (1967) Biochim. Biophys. Acta 132,338-346. 47. BUETTNER, M. J., SPITZ, E., AND RICKENBERG, H. V. (1973) J. Bacterial. 114,1068-1078.
Reductive Repression in Escherichia co/i K-l 2 Is Mediated by Oxygen Radicals’ R. HERTZ Department
AND
J. BAR-TANA
of Biochemistrg Hebrew University-Hadassah P.O. Box 1172, Jerusalem 91010, Israel
Received September
Medical School,
3,1985, and in revised form June 17,1986
Cyclic AMP (CAMP) content and the expression of CAMP-dependent phenotypes were positively correlated with respiration capacity in respiration-deficient mutants of Escherichia coli K-12 (“reductive repression,” R. Hertz, and J. Bar-Tana, (1982) Arch. Biochem. Biophgs. 213,193-199). Reductive repression in respiration-deficient mutants could not be accounted for by respective changes in either the energy charge of adenine nucleotides or the redox state of pyridine nucleotides but could be ascribed to an increased formation of oxygen radicals under conditions of limited respiration. Scavengers of superoxide radicals eliminated reductive repression in respiration-deficient mutants with a concomitant increase in CAMP content. Such scavengers also effected a partial escape from permanent glucose catabolite repression, thus indicating a possible role played by oxygen radicals in both repression modes. o 19% Academic PEW, IN.
Cyclic AMP (CAMP)’ content and the expression of CAMP-dependent phenotypes in Escherichia coli K-12 were reported recently to be positively correlated with the respiration capacity of the culture. Thus, growing ubiquinone (ubi A, ubi D)- or protoporphyrin (hem A)-deficient mutants in the presence of variable concentrations of p-hydroxybenzoate (p-HBA) or F-aminolevulinate (&ALA), respectively, resulted in a respective gradual increase in the ubiquinone or cytochrome content with a concomitant recovery of respiration capacity, CAMP content, P-galactosidase activity, and flagellation. Repression of CAMP-dependent phenotypes under conditions of limited electron transport-mediated respiration (“reductive repression”) and modulation of the CAMP content as a function
of respiration rate were the results of modulation of the adenylate cyclase activity (1). The present report evaluates the possible role played by the phosphate potential, the redox state of pyridine nucleotides, and the generation of oxygen radicals in mediating reductive repression in E. coli K-12. MATERIALS
Copyright All rights
0 1986 by Academic Press, Inc. of reproduction in any form reserved.
METHODS
Respiration-deficient, @‘-galactosidase constitutive strains HB 101 (ubi A, lac”) and HB 111 (hem A, lac’) were derived as previously described (1). The ubi A allele affects the prenylation step of the ubiquinone synthetic pathway by increasing the K, for p-hydroxybenzoic acid (2). The hem A allele affects the protoporphyrin synthetic pathway by eliminating the biosynthesis of &aminolevulinic acid (3). Bacteria were grown aerobically in 0.8% nutrient broth in 0.05 M potassium phosphate buffer (pH 7.0) or in H minimal medium (4), with additions as described. Cultures were grown to midlog phase with continuous gyratory shaking at 30°C. Bacterial density was determined by measuring the optical density at 650 nm. Respiration rate of washed cultures was measured in the presence of 20 mM succinate in H minimal me-
i Supported by the Herta Schnap Foundation for Cancer Research. 2 Abbreviations used: pHBA, p-hydroxybenzoic acid, &ALA, &aminolevulinic acid, CAMP, cyclic AMP, PMS, phenazine methosulfate. 0003-9861/86 $3.00
AND
54
REDUCTION
REPRESSION
MEDIATED
BY OXYGEN
dium using an oxygen electrode. Respiration capacity of electron transport-deficient strains was expressed as percentage of the respiration rate of the respective deficient strain grown in the presence of saturating concentrations of its limiting precursor, p-HBA or 6ALA, respectively. The 100% respiration rate for the HB 101 (ubi A, la&) strain grown in the presence of saturating pHBA amounted to 150 + 10 nmol Oz/mg protein/min. The 100% respiration rate for the HB 111 (hem A, lace) strain grown in the presence of saturating &ALA amounted to 300 2 20 nmol Oz/mg protein/min. Cyclic AMP content was determined as described previously (1). Adenine nucleotides were extracted (5, 6) and determined by the luciferase bioluminescence assay as described by Swedes et al. (7). Pyridine nucleotides were extracted and determined as described by Wimpenny and Firth (8). All nucleotides were extracted with 90-95% efficiency. P-Galactosidase activity was assayed at 28°C in toluene-treated cells as described by Pardee et al. (9). Cultures assayed for catalase were washed and suspended in 0.05 M phosphate buffer (pH 7.0) and were sonicated three times consecutively for 15 s. The sonicate was centrifuged at 37,OOOgfor 30 min and the activity was determined in the supernatant. The reaction mixture for catalase contained 0.05 M of phosphate buffer (pH 7.0) and 9 mM of hydrogen peroxide in a final volume of 1.0 ml. The activity was assayed by following the initial decrease in absorbance at 240 nm at 25°C (10). a-Glycerokinase was assayed as described by Newsholme et al. (46). Cultures assayed for a-glycerokinase were washed and suspended in 0.1 M Tris-HCl buffer (pH 7.4) and were sonicated twice for 30 sec. The sonicate (5-10 mg) was incubated in the presence of 0.1 M Tris-HCl (pH 7.4), 16 mM EDTA, 22 mM NaF, 5 mM MgS04, 33 mM mercaptoethanol, 5 mM ATP, and 4 mM (U-‘“C) glycerol (2.5 &i/rmol) in a final volume of 1.0 ml. Catalase complex I was measured by following the difference spectrum in the 60@ 650 nm range, using double beam spectrophotometry (11-13). Cultures assayed for superoxide dismutase activity were washed and suspended in 0.05 M phosphate buffer (pH 7.8) containing 0.1 mM EDTA and were sonicated three times consecutively for 15 s. The sonicate was centrifuged at 27,OOOgfor 60 min, dialyzed in 0.05 M phosphate buffer (pH 7.0) containing 0.1 mM EDTA, and was again centrifuged at 30,OOOg for 60 min. Superoxide dismutase activity was determined in the supernatant by assaying the extent of inhibition of nitrobluetetrazolium reduction in the presence of added bacterial protein (44). Cu-(salicylate)z was prepared according to (14). Cu(1ysine)z was prepared by refluxing 1.8 g of L-lysine with 16.0 g of CuO in 200 ml of Ha0 for 120 min. The cooled mixture was diluted into water, filtered, allowed to crystallize for 48 h, and the crystals were washed in acetone and dried in air.
RADICALS
Significance test.
IN Eschetichia
coli
was analyzed by the Mann-Whitney
55 U
RESULTS
Adenine and Pyridine Nucleotides as Mediators of Reductive Repression The ATP, ADP, AMP, and CAMP contents and the /3-galactosidase activity as a function of respiration capacity in E. coli K-12 were evaluated by growing the ubi A strain in the presence of increasing concentrations of p-HBA added to the growth medium. As shown in Table I, increasing the respiration capacity within the range of 20-100% was linearly correlated with the ATP content as well as with the CAMP content and the P-galactosidase activity as previously reported (1). However, the ATP/ ADP ratio did not vary significantly throughout the range of 20-100% respiration capacity and the AMP content was found to be lower than 0.5 nmol/mg protein, whether in the presence or absence of added p-HBA. Similarly, no significant change in the redox charge (15) of pyridine nucleotides or phosphopyridine nucleotides was observed as a function of respiration capacity of the ubi A strain grown either in glycerol or succinate (Table II). Also, the adenine and pyridine nucleotide charges of the HB 111, hem A strain, remained unchanged as a function of &ALA added to the growth medium, under conditions of an increase in respiration capacity from 20 to 100% with a concomitant threefold increase in CAMP content (not shown). These results seem to confirm previous reports where no significant changes in the “reductive catabolic charge” [NADH/(NAD + NADH)] (15) or energy charge (16) have been observed upon limiting the rate of growth of E. coli by nutrient availability or oxygen tension. The two- to threefold increase in CAMP content and CAMP-dependent activities as a function of respiration capacity in the absence of variation in adenine and pyridine nucleotide charges seem to indicate that reductive repression in E. coli K-12 is presumably not mediated by either the energy or the reductive charges.
HERTZ
AND TABLE
BAR-TANA I
ATP AND ADP CONTENT AS A FUNCTION OF RESPIRATION CAPACITY IN A UBIQUINONE-DEFICIENT MUTANT OF E. coli K-12 Respiration capacity 23 34 51 65 86 100
8-Galactosidase 1.8 + 2.2 f 2.6 f 2.8 f 3.2 f 4.0 +
0.1 0.1* O.l* O.l* 0.2* 0.2*
CAMP 20.0 21.5 24.0 26.0 29.0 40.0
+ + + f f +
3.5 2.5 3.0 2.0 2.2* 3.0*
ATP
ADP
ATP/ADP
6.3 f 0.4 6.9 f 0.5 7.7 * 0.5* 7.2 + 0.25* 8.3 f 0.8* 7.9 zk0.7*
1.1 f 0.15 0.9 zk0.1 1.4 + 0.15 1.1 f 0.2 1.2 f 0.3 0.9 * 0.15
5.7 zk 1.1 7.7 f 1.4 5.5 Lk0.9 6.5 f 1.4 6.9 IL 2.4 8.6 A 2.2
Note. HB 101 was grown in H minimal medium supplemented with 0.2 mM L-lysine, 0.2 mM L-methionine, 0.2 fiM thiamine, 0.4% glycerol, and p-HBA as required to yield the percentage respiration capacity as stated. @Galactosidase activity (U/mg protein), CAMP (pmol/mg protein), ATP (nmol/mg protein), and ADP (nmol/ mg protein) content (mean k SD, TZ= 3) were determined as described under Materials and Methods. * Significantly different from the respective values observed at 23% respiration capacity, P < 0.05.
Oxygen Radicals as Mediators of Reductive Repression The possible role of oxygen radicals in mediating reductive repression was studied by (a) measuring the content of either hydrogen peroxide or hydrogen peroxide-induced catalase as a function of respiration capacity in electron transport-deficient mutants; (b) evaluating the CAMP content and CAMP-dependent phenotypes of the wild type in the presence of added artificial redox couples known to generate superoxides or hydrogen peroxide; (c) evaluating the extent of reductive repression in respiration-deficient strains grown in the presence of scavengers of superoxide radicals. Since hydrogen peroxide serves as an inducer of catalase in E. coli (18, 19, 44) as well as in the ubi A mutant (not shown), the steady state level of hydrogen peroxide as a function of respiration capacity could be inferred from the catalase activity observed in respiration-deficient strains in the presence or absence of added p-HBA. Thus, the 250% increase in catalase activity of the ubi A strain grown under conditions of limiting p-HBA (Table III) may indicate a higher intracellular content of hydrogen peroxide generated under conditions of reductive repression. The direct determination of the intracellular hydrogen peroxide content of the HB 101 strain was also at-
tempted by following spectrophotometritally the steady state level of catalase complex I in the presence of added nitrite or methanol (11, 12) and as a function of respiration capacity of the culture. However, no catalase complex I difference spectrum in the 600 to 680-nm range was observed in this strain even at highly densed cultures (20-50 ODsW), and the relatively low contribution made by catalase complex I to the difference spectrum of the soret band did not allow the spectrophotometric measurement of complex I in the 400-nm range. Also, the generation rate of extracellular hydrogen peroxide as determined by scopoletin fluorescence was lower than the sensitivity limit of the assay (17), nor did preincubation of the culture in the presence of added aminotriazole result in any detectable extracellular hydrogen peroxide. Two artificial redox couples capable of generating oxygen radical species were studied for their capacities as reductive repressors in the wild type. Paraquat was used to mediate one-electron reduction of molecular oxygen to yield the superoxide radical (18), and PMS was used to mediate either one- or two-electron reduction of molecular oxygen to yield the superoxide radical or hydrogen peroxide, respectively (20, 21). Alternatively, an increase in intracellular hydrogen peroxide was effected by inhibition of catalase by aminotriazole
REDUCTION
REPRESSION
MEDIATED
BY OXYGEN
RADICALS
IN Escherichia
coli
5’7
added to the growth medium (22). Since the respiration capacity of the ulri A mutant grown in the presence of 500 pM added pHBA is similar to that of the isogenic wild type (1, 2), the repressive potential of the various artificial redox couples was evaluated in the mutant grown in the presence of p-HBA. As shown in Table IV ,8-galactosidase activity was repressed by PMS, paraquat, or aminotriazole added to the growth medium at concentrations which did not affect the growth rate of the culture. The repression of /3-galactosidase activity was accompanied by a two- to threefold decrease in CAMP content and was partially restored in the presence of exogenous CAMP added to the growth medium. Similarly, PMS and paraquat repressed the P-galactosidase activity of the hem A strain grown in the presence of saturating &ALA and the addition of 5 mM of CAMP to the growth medium effected a complete escape from the observed repression. Organic complexes of Cu2+ (e.g., Cu-lysine, Cu-salicylate) were reported to scavenge superoxide radicals by catalyzing the dismutation of superoxides into hydrogen peroxide (23-25). Indeed, reductive repression of either P-galactosidase (Table V, Fig. 1) or a-glycerokinase (Table V) was eliminated in the ubi A strain in the presence of Cu complexes added to the culture medium, and the increase in P-galactosidase activity induced by these scavengers was correlated with an increase in CAMP content (Table V). The escape induced by Cu complexes was observed either in nutrient broth or in minimal medium with the efficacy of the scavenger being higher in minimal medium. It is worth noting that Cu complexes did not improve the respiration capacity of the ubi A strain (Table V), and did not activate the P-galactosidase activity of HB 101 grown in the presence of saturating p-HBA (Table V). Also, Cu complexes could not substitute for CAMP in a cya strain and the P-galactosidase activity was nil in the presence or absence of the added Cu complex. Hence, escape from reductive repression exerted by these scavengers could not be ascribed to reversion of the respiratory deficiency of the ubi
58
HERTZ
AND
BAR-TANA
TABLE CATALASE
AND SUPEROXIDE DISMUTASE ACTIVITIES AS A FUNCTION OF RESPIRATION IN A UBIQUINONE-DEFICIENT MUTANT OF E. coli K-12
Addition to growth medium p-HBA p-HBA
III
Respiration capacity
5 /.LM 500 HIM
Catalase
activity
25
50.8 f 2.1
100
19.1 f 2.6
CAPACITY
Superoxide dismutase activity 7.2 f 0.5 6.5 f 0.6
Note. HB 101 was grown in H minimal medium supplemented with 0.2 mM L-lysine, 0.2 mM L-methionine, 0.2 pM thiamine, 0.4% glycerol, and pHBA as stated to yield the percentage respiration capacity as stated. Catalase activity (nmol H202 decomposed/mg protein/min) was determined as described under Materials and Methods (data are means f SD; n = 7). Superoxide dismutase activity is expressed as the amount of protein (pg) which inhibited by 50% the reduction of nitrobiue tetrazolium (44) (mean f SD, n = 3).
A mutant, to nonspecific activation of pgalactosidase, or to bypassing CAMP-dependent promotion. The presence of either Cu-salicylate or Cu-lysine effected also an escape of the HB 101 strain from permanent glucose catab-
olite repression exerted by growth in glucase. Thus, the P-galactosidase activity of HB 101 grown in minimal medium in the presence of saturating p-HBA amounted to 4.6 and 1.5 U/mg protein in the presence of either 0.4% glycerol or 0.4% glucose as
TABLE REDUCTIVE
REPRESSION
IN
E.
coli K-12
EXERTED
IV BY PMS,
PARAQUAT,
AND AMINOTRIAZOLE
@-Galactosidase Addition
to growth
medium
HB 101 None PMS 20 ELM Paraquat 80 pM Paraquat 160 PM Aminotriazole 0.1 M Paraquat 160 pM + aminotriazole 0.1 M HB 111 None PMS 26 PM PMS 52 /.hM Paraquat 320
@M
-CAMP
+cAMP
CAMP content
0.3 0.2* 0.3* 0.3* 0.1*
4.3 * 0.3 3.3 f 0.3** 3.0 z!Y0.2** 2.0 + 0.2 3.5 f 0.2**
12.5 f 3.0 4.7 f 0.5* 6.5 + 0.5*
1.3 f 0.1*
2.4 2 0.2**
4.7 f 0.6*
2.2 + 0.1 1.5 + 0.1*
2.5 + 0.2 2.7 + 0.2**
1.6 + O.l*
2.9 + 0.1**
1.6 + 0.2*
2.5 + 0.2**
3.9 f 2.1 k 2.0 + 1.6 f 1.9 f
5.0 + 0.6*
Note HB 101 was grown in nutrient broth supplemented with 0.2 PM thiamine and 500 PM p-HBA, in the presence or absence of 5 mM added CAMP as stated. HB 111 was grown in nutrient broth supplemented with capacity of all grown cultures amounted to 0.4% glycerol, 0.2 pM thiamine, and 10 pM &ALA. Respiration 100%. Three-four hours before they were harvested the cultures were diluted into fresh medium and either PMS, paraquat, or aminotriazole was added to the growth medium as stated. P-Galactosidase activity (U/mg protein) was determined as described under Materials and Methods (mean f SD, n = 3). The intracellular CAMP content (pmoi/mg protein) was determined in cultures grown in the absence of CAMP added to the growth medium as described under Materials and Methods (mean f SD, n = 3). * Significantly different from the respective no addition value, P i 0.05. ** Significantly different from the respective (-CAMP) value, P < 0.05.
REDUCTION
REPRESSION
MEDIATED
BY OXYGEN TABLE
RADICALS
IN Escherichia
coli
59
V
ESCAPEFROMREDUCTIVEREPRESSIONEXERTEDBYORGANIC CU COMPLEXES Addition to growth medium
Respiration capacity
None Cu-salicylate 10 PM Cu-salicylate 125 FM Cu-salicylate 250 PM Cu-salicylate 1000 PM p-HBA 500 PM p-HBA 500 /LM + Cusalicylate 500 PM
CAMP P-Galactosidase
10
10 10 10
1.0 t 0.1
ol-Glycerokinase 8.5 f 0.8
content 9.5 * 2.0
12.8 f1.3* 0.15 * 0.15 * 0.2* 0.2*
10 100
2.3 f 3.2 + 3.8 f 4.0 +
100
4.6 I+ 0.3*
12.7 f 1.5
17.5 f 2.0* 14.5 f 0.8*
26.0 Z+I4.0* 26.0 f 5.0*
15.3 IL 0.8*
23.0 + 4.0*
Note. HB 101 was grown either in nutrient broth (for P-galactosidase and CAMP determinations) supplemented with 0.2 FM thiamine, p-HBA, and Cu-salicylate as stated, or in H minimal medium (for o-glycerokinase determination) supplemented with 0.2 mM L-lysine, 0.2 mM L-methionine, 0.2 PM thiamine, 0.4% glycerol, pHBA, and Cu-salicylate as stated. Respiration capacity (%),@galactosidase (U/mg protein), a-glycerokinase (mU/mg protein), and CAMP content (pmol/mg protein) were determined as described under Materials and Methods (mean + SD, 12 = 3). * Significantly different from the respective no addition value. P < 0.05.
carbon sources, respectively. However, in the presence of 50 PM of Cu-salicylate added to the growth medium, the activity observed in either glycerol or glucose media amounted to 4.2 U/mg protein. The escape from glucose catabolite repression exerted
by Cu-complexes was also observed in a CAMP phosphodiesterase-deficient mutant (47) and could not be ascribed, therefore, to inactivation of the phosphodiesterase by organic Cu complexes with a concomitant increase in CAMP content. DISCUSSION
3-EB l-----l 2.25
0.063
FIG. 1. Escape from reductive repression exerted by Cu-lysine: HBlOl was grown in nutrient broth supplemented with 0.2 FM thiamine and Cu-lysine as indicated. fl-Galactosidase activity was determined as described under Methods (mean + SD, rz = 3). The pgalactosidase activity in the presence of 500 pM pHBA added to the growth medium amounted to 4.0 U/mg protein.
The role of oxygen radicals in mediating reductive repression in E. coli K-12 was indicated here by three lines of evidence: (a) Intracellular CAMP content was decreased and the /3-galactosidase activity was repressed in the HB 101 strain under conditions of growth in the presence of nonlimiting p-HBA and added oxidizing agents capable of generating oxygen radicals. The repressed activity could be restored in the presence of exogenous CAMP added to the growth medium. (b) The catalase activity was significantly higher in electron transport-deficient strains, thus indicating a higher generation rate of hydrogen peroxide under conditions of inhibition of the electron transport chain. (c) Reductive repression of p-HBA-limited, electron transport-deficient strains could be eliminated in the presence of scavengers of
60
HERTZ
AND
superoxide radicals. Hence, reductive repression of ubiquinoneor cytochromedeficient strains of E. coli K-12 requires a direct reduction of molecular oxygen into superoxide radicals mediated by reduced intermediates of the electron transport chain under conditions of inhibition of oxygen reduction into water. Overproduction of oxygen radicals under these conditions may resemble the generation of superoxide radicals in antimycin-inhibited mitochondria (43). Alternatively, reductive repression of the wild type may be effected by generating a cyanide-resistant respiratory pathway mediated by artificial redox couples (e.g., PMS, paraquat) which may compete for electrons with intermediates of the electron transport chain (44). Whatever the mechanism of superoxide production, the oxygen radicals produced under conditions of reductive repression are assumed to inhibit the adenylate cyclase with a concomitant repression of CAMP-dependent phenotypes (1). The exceptional sensitivity of the cyclase protein to inactivation by oxygen radicals and the possible interrelationship between reductive repression and oxidation-dependent “heat-shock stress” (26, 27, 45) still remain to be evaluated. The present report seems to indicate that superoxide radicals or hydroxyl radicals rather than hydrogen peroxide are implicated as effecters of reductive repression. Thus, scavenging the superoxide radicals by organic Cu complexes, while dismutating these radicals into hydrogen peroxide (24), resulted in an escape from reductive repression. Reductive repression by aminotriazole-dependent inactivation of catalase could still be ascribed to the accumulation of superoxides or hydroxyl radicals as a result of inactivation of superoxide dismutase by hydrogen peroxide (28). Since reductive repression was also observed under anaerobic growth conditions in a mutant deficient of both ubiquinone and menaquinone (29), inactivation of the cyclase may be exerted also by nonoxygen radicals which may accumulate anaerobically in the absence of a functional anaerobic electron transport. Alternatively, since cultures may not be entirely
BAR-TANA
anaerobic, reductive repression under anaerobic growth conditions could still result from the generation of oxygen radicals. The interplay between the induction of either superoxide dismutase or catalase and their possible repression under conditions of reductive repression is worth noting. Thus, the induction of the dismutase by superoxide radicals (38, 39) could be expected to increase its activity under conditions of reductive repression. On the other hand, since this enzyme appears to be controlled by permanent glucose catabolite repression (40), its expression could be expected to be repressed under conditions of reductive repression as a result of the concomitant decrease in CAMP content. Indeed, the superoxide dismutase activity remained unchanged in reductively repressed electron transport-deficient strains (Table III), presumably reflecting the antagonistic inductive-repressive effects exerted on this enzyme by oxygen radicals. In contrast to repression of the superoxide dismutase by glucose growth, catalase expression in E. coli was reported not to be modulated by permanent glucose catabolite repression nor by the CAMP content (41). Thus, the presumed increased generation of hydrogen peroxide under conditions of reductive repression in electron transportdeficient strains may effect an induction of catalase (l&19,45) unmasked by the concomitant decrease in CAMP content (Table III). Hence, the catalase activity in reductively repressed strains presumably reflects the intracellular level of its hydrogen peroxide inducer whereas the superoxide dismutase responds both to the level of its superoxide inducer and to the prevailing decreased CAMP content. It is worth noting that the relative increase in catalase activity of the ubi A mutant under limiting p-HBA as reported here (Table III) is in discrepancy with results previously reported by Hassan and Fridovich (42). The discrepancy observed may have resulted from strain, culture medium, or other experimental differences. Reductive repression is essentially similar to permanent glucose catabolite repression (30-32) with respect to the extent
REDUCTION
REPRESSION
MEDIATED
BY OXYGEN
of inhibition of the cyclase activity (5, 35), the scope and extent of repression of CAMP-promoted phenotypes (32), and the escape observed in the presence of added CAMP (33, 34). Furthermore, escape from either reductive or catabolite repression as exerted by scavengers of superoxide radicals seems to indicate an involvement of oxygen radicals in both repressive modes. The generation of superoxide radicals by the electron transport chain (36) may thus be assumed to result from glucose catabolism (37) in excess of the electron flux allowed through the limiting step of the chain, or a mutation of the electron transport chain which limits respiration capacity (1). The oxygen radicals produced by either one of these possibilities may inactivate the cyclase with a concomitant decrease in CAMP content and CAMP-promoted phenotypes, thus leading to permanent glucose catabolite or reductive repression, respectively. Hence, permanent glucose catabolite repression may be considered as a specific mode of reductive repression. ACKNOWLEDGMENT We thank Professor B. Hess, The Max Planck Institute fiir Ernahrungsphysiologie, Dortmund, Germany, for his help and advice. REFERENCES 1. HERTZ, R., AND BAR-TANA, J. (1982) Arch. Biochem Biophys. 213,193-199. 2. YOUNG, I. G., LEPPIK, R. A., HAMILTON, J. A., AND GIBSON, F. (1972) J. Bacterial. 110,18-25. 3. SASARMAN, A., SURDEANU, M., AND HORODNICEANU, T. (1968) J. BacterioL 96,1882-1884. 4. KAISER, A. D., AND HOGNESS, D. S. (1960) J. Mol. BioL 2,392-415. 5. PETERKOFSKY, A., AND GAZDAR, C. (19’73) Proc. Natl. Acad. Sci. USA 70,2149-2152. 6. CHAPMAN, A. G., FALL, L., AND ATKINSON, D. E. (1971) J. Bacterial 108,1072-1086. 7. SWEDES,J. S., SEDO, J., AND ATKINSON, D. E. (1975). J. BioL Chem. 250,6930-6938. 8. WIMPENNY, J. W. T., AND FIRTH, A. (1972) J. Bacteriol. 111, 24-32. 9. PARDEE, A. B., JACOB, F., AND MONOD, J. (1959). J. Mol. Biol. 1, 165-178.
RADICALS
IN Escherichia
coli
61
10. BEERS, R. F., AND SIZER, I. W. (1952) J. BioL Chem. 195,133-140. 11. CHANCE, B., AND OSHINO, N. (1971) Biochem. J. 122.225-233. 12. OSHINO, N., AND CHANCE, B. (1973) Arch, Biochem Biophys. 154,117-131. 13. HESS, B., KLEINHANS, H., AND SCHLUTER, H. (1970) Hoppe Seyler’s Z. PhysioL Chem. 351, 515-531. 14. DE ALVARE, L. R., GODA, K., AND KIMURA, T. (1976) B&hem. Biophys. Res. Commun. 69,687-694. 15. ANDERSEN, K. B., AND MEYENBURG, K. (1977) J. BioL Chem. 252,4151-4156. 16. WALKER-SIMMONS, M., AND ATKINSON, D. E. (1977) J. Bacterial. 130, 676-683. 17. ROOT, R. K., METCALF, J., OSHINO, N., AND CHANCE, B. (1975) J. Clin. Invest. 55, 945-955. 18. HASSAN, H. M., AND FRIDOVICH, I. (1980) J. BacterioL 141,156-163. 19. YOSHPE-PURER, Y., HENIS, Y., AND YOSHPE, J. (1977) Canad J. MicrobioL 23,84-91. 20. MCCORD, J. M., AND FRIDOVICH, I. (1970) J. Biol. Chem. 245,1374-1377. 21. NISHIKIMI, M., RAO, N. A., AND YAGI, K. (1972) Biochem. Biophys. Res. Commun. 46,849-854. 22. MARGOLIASH, E., NOVOGRODSKY, A., AND SCHEJTER, A. (1959). Biochem. J. 74,339-350. 23. WESER, U., SELLINGER, K. H., LONGFELDER, E., WERNER, W., AND STRAHLE, J. (1980) Biochim. Biophys. Acta 631,232-245. 24. WALKER, W. R., AND BEVERIDGE, S. J. (1979) Inorg Perspect. BioL Med 2, 93-125. 25. RICHTER, C., PRIPFL, T., AND WINTERHALTER, K. H. (1980) FEBS I&t. 111,95-98. 26. LEE, P. C., BOCHNER, B. R., AND AMES, B. N. (1983) Proc. NatL Acad. Sci. USA 80,7496-7500. 27. LEE, P. C., BOCHNER, B. R., AND AMES, B. N. (1983) J. BioL Chem. 258,6827-6834. 28. HASSAN, H. M. (1980) in Biological Roles of Copper (Ciba Foundation Symposium No. 79), pp. 125142, Elsevier, Amsterdam. 29. HERTZ, R., AND BAR-TANA, J. (1977) J. Bucteriol. 132,1034-1035. 30. MAGASANIK, B. (1961) Cold Spring Harbor Symp. Quant. Biol. 26,249-256. 31. MAKMAN, R. S., AND SUTHERLAND, E. W. (1965) J. BioL Chem. 240,1309-1314. 32. PASTAN, I., AND ADHYA, S. (1976) BacterioL Rev. 40,527-551. 33. ULLMANN, A., AND MONOD, J. (1968) FEBS Lett. 2, 57-60. 34. PERLMAN, R. L., GROMBRUGGHE, B., AND PASTAN, I. (1969) Nature (London) 223,810-812. 35. EPSTEIN, W. L., ROTHMAN-DENES, L. B., AND HESSE, J. (1975) Proe. NatL Acad. Sci USA 72, 2300-2304. 36. BOVERIS, A., OSHINO, N., AND CHANCE, B. (1972) Biochem. J. 128,617-630.
62
HERTZ
AND
37. WANNER, B. L., KODAIRA, R., AND NEIDHARDT, F. C. (19’78) J. Bacterial. 136,947-954. 38. HASSAN, H. M., AND FRIDOVICH, I. (1977) J. Biol. Chem. 252,7667-7672. 39. HASSAN, H. M., AND FRIDOVICH, I. (1978) J. Biol. Chem. 253,8143-8148. 40. HASSAN, H. M., AND FRIDOVICH, I. (1977) J. Batteriol. 132,505-510. 41. RICHTER, H. E., AND LOEWEN, P. C. (1982) Arch. Biochem. Biophys. 215,72-77. 42. HASSAN, H. M., AND FRIDOVICH, I. (1978) J. Biol. Chem 253,6445+X50.
BAR-TANA 43. CADENAS, E., AND BOVERIS, A. (1980) Biochem. J 188,31-37. 44. HASSAN, H. M., AND FRIDOVICH, I. (1979) Arch Biochem. Biophys. 196,385-395. 45. CHRISTMAN, M. F., MORGAN, R. W., JACOBSON, F. S., AND AMES, B. N. (1985) Cell 41,753-762. 46. NEWSHOLME, E. A., ROBINSON, J., AND TAYLOR, K. (1967) Biochim. Biophys. Acta 132,338-346. 47. BUETTNER, M. J., SPITZ, E., AND RICKENBERG, H. V. (1973) J. Bacterial. 114,1068-1078.