JOURNAL OF BACTERIOLOGY, Nov. 1977, p. 505-510 Copyright C 1977 American Society for Microbiology
Vol. 132, No. 2
Printed in U.S.A.
Regulation of Superoxide Dismutase Synthesis in Escherichia coli: Glucose Effect H. MOUSTAFA HASSAN AND IRWIN FRIDOVICH* Department of Biochemistr,y, Duke University Medical Center, Durham, North Carolina 27710
Received for publication 13 May 1977
Growth of Escherichia coli based upon the fermentation of glucose, is associated with a low intracellular level of superoxide dismutase. Exhaustion of glucose, or depression of the pH due to accumulation of organic acids, causes these organisms to then obtain energy from the oxidative degradation of other substances present in a rich medium. This shift in metabolism is associated with a marked increase in the rate of synthesis of superoxide dismutase. Depression of the synthesis of superoxide dismutase by glucose is not due to catabolite repression since it is not eliminated by cyclic adenosine 3',5'-monophosphate and since a-methyl glucoside does not mimic the effect of glucose. Moreover, glucose itself no longer depresses superoxide dismutase synthesis when the pH has fallen low enough to cause a shift to a non-fermentative metabolism. It appears likely that superoxide dismutase is controlled directly or indirectly by the intracellular level of O., and that glucose depressed the level of this enzyme because glucose metabolism is not associated with as rapid a production of 0., as is the metabolism of many other substances. In accord with this view is the observation that paraquat, which can increase the rate of production of O,- by redox cycling, caused a rapid and marked increase in superoxide dismutase. The production of 0.. by the univalent reduction of molecular oxygen is a commonplace event, and the superoxide dismutases (SOD), which catalytically scavenge this free radical, appear to be essential defenses against its potential cytotoxicity. The work supporting these statements has been reviewed (5-7). Recent studies of Escherichia coli K-12, in glucoselimited chemostat culture. indicated that both oxygen uptake and content of SOD increased with increases in specific growth rate. The level of SOD, manipulated in this way, correlated with tolerance towards oxygen toxicity (10). These results and others, in which the level of SOD was modified by oxygen induction (9), support the view that SOD is essential for the aerobic growth of E. coli. During studies of the regulation of SOD synthesis in bacteria. we noted that the presence of glucose exerted profound effects, which gave the appearance of catabolite repression (H. M. Hassan and I. Fridovich, Fed. Proc. 36: 715, 1977). We now describe the effects of glucose and other nutrients on the synthesis of SOD in growing cultures of E. coli. MATERIALS ANt) MIETHODS Organisms. E. coli K-12 his thi (ATCC 23794) and E. coli B Bl2 (ATCC 29682) (provided by D. H. Hall) were used throughout.
Culture. Trypticase soy broth, trypticase peptone and yeast extract from the Baltimore Biological Laboratories and peptone and nutrient broth from Difco were used in preparing media. The Trypticase soy-yeast extract (TSY) medium contained 3% Trypticase soy broth and 0.5% yeast extract. This medium contains 0.25% glucose derived from the commercial Trypticase soy powder. The Trypticase peptone (TP) medium contained 2%c Trypticase peptone, 0.5% bactopeptone, 0.5% NaCl, and 0.25% K2HPO4. The TP medium was devoid of sugars. The TPY medium was identical to the TP medium except that 0.51% yeast extract was also present. The TPY medium was very similar to the TSY medium but it contained no glucose. Glucose-minimal medium contained (per liter): MgSO4 7H2O, 0.2 g; citric acid H2O, 2.0 g; K2HPO, 10.0 g; NaNH4HPO4 4H2O, 3.5 g; glucose, 5.0 g, and the essential growth factors B12 (1 mg) for E. colh B and histidine (300 mg) plus thiamine (30 mg) for E. coll K-12. Cultures were grown at 37°C on a rotatory shaker at 200 rpm with a ratio of flask volume to medium volume of 5:1. The size of the inoculum was varied to give an initial absorbancy at 600 nm in the range of 0.1 to 0.2. Growth was estimated in terms of turbidity measured as absorbancy at 600 nm,. and specific growth rates were calculated as described previously (10). Assays. Samples ot the cultures, collected at intervals, were chilled in ice and centrifuged at 10,000 x g for 15 min. The cells were washed in cold 50 mM potassium phosphate (pH 7.0) and, after suspension in 50 mM potassium phosphate, 0.1 mM
505
506
J. BACTERIOL.
HASSAN AND FRIDOVICH
ethylenediaminetetraacetic acid at pH 7.0, were disrupted by sonic oscillation for 3 min with a Branson W185 sonifier, operated at a power output of 50 W. The temperature during sonic treatment was maintained at 4 to 6°C by chilling in an ice-salt bath and by intermittent application of sonifier power. The homogenates were clarified at 30,000 x g for 1 h and then used directly for assays. Protein was estimated by the Lowry (12) method with pure bovine serum albumin as a calibrating standard. SOD was assayed as described previously (14). Samples of the culture medium were obtained by filtration through 0.45-gim membrane filters (Millipore Corp.) and assayed for residual glucose by the glucostat method (Sigma Chemical Co).
place of E. coli K-12. These changes in SOD were seen only with rich media that contained glucose. In contrast, growth in glucose-minimal medium was not accompanied by marked changes in SOD. Thus, growth of a log-phase inoculum, taken from a glucose-minimal medium, in glucose-minimal medium, exhibited a steady decline in pH during growth and no significant increase in SOD (Fig. 2). These results suggested that glucose might repress the biosynthesis of SOD.
RESULTS Effect of growth medium on SOD content of cells. As shown in Table 1. the specific 0 activity of SOD, present in cell-free extracts of 0 E. coli B, was strongly influenced by the culture medium. The level of SOD was not a 0 simple function of growth rate, as is the case (0 202 in glucose-limited chemostat culture (10), since growth was slower in TP medium than in the 0 nutrient broth plus 0.5% glucose medium, yet 0 the SOD content was two times higher with 0 cn) the former medium. Furthermore, cells grown in minimal medium containing succinate or lactate, as the sole carbon source, contained higher levels of SOD than did cells grown on glucose. Changes in SOI) during growth cycle in HOURS TSY medium. Figure 1 presents the changes FIG. 1. Kinetics of growth and SOD synthesis by in pH and SOD content during the growth of E. coli K-12 on TSY medium. Growth was E. coli K-12 in a TSY medium. Growth was initiated in the TSY medium with a 1% inoculum taken from started with an inoculum taken from a 17-h an early-stationary-phase culture in the same meculture in the same medium. It is apparent dium. that both pH and SOD concentrations declined for 5 h and then increased again. At zero time this medium was found to contain 13.9 mM glucose, which then diminished during 5 h of 7A-O* incubation to undetectably low levels. Similar results were seen when E. coli B was grown in 0 ,2 o
15a
0
TABLE 1. Effect of different growth media on the levels of superoxide dismutase in E. coli Ba Medium
Glucose-minimal
Generation time (min)
/
X 0
10
--
/
-
0
SOD
I0
'+o
SOD
_
0
D
(U/
mg)
46 6-8 Nutrient broth + 0.5% glucose 39 7-9 Minimal salts + 0.5% succinate 160 11-13 Minimal salts + 0.5% lactate 105 15-16 TP 42 16-18 TSY 32 21-24 (a The cells were grown in the different media at 37°C on a rotary shaker (200 rpm), and samples were removed at different intervals for growth and SOD assays. The values reported for SOD were taken at peak values (i.e., late-log phase of growth).
0
I
o
a
~-._pH
0
0
105
-i
cn
60
2
3
4
5
6
7
HOURS FIG. 2. Kinetics of growth and of SOD synthesis by E. coli B in a glucose-minimal medium. Growth was initiated in the glucose-minimal medium with a 5% inoculum taken from a logarithmic-phase culture in the same medium.
VOL. 132, 1977
507
REGULATION OF SUPEROXIDE DISMUTASE IN E. COLI
Effect of glucose on SOD synthesis. The appeared possible that glucose might repress influence of glucose was investigated by shift- the synthesis of SOD. ing cells from a glucose-minimal to a TSY Effect of cAMP on SOD synthesis. In E. medium. As shown in Fig. 3, growth under coli and other bacteria, the biosynthesis of these conditions was diauxic. There was an many inducible enzymes is regulated by the initial phase of rapid growth, which coincided intracellular level of cyclic adenosine 3',5'with a decrease in pH and with consumption of monophosphate (cAMP) (3). Glucose reduces glucose. This was followed by a lag in growth the level of cAMP in these bacteria and thus of approximately 3 h, after which growth was represses the synthesis of these enzymes (13). resumed. The growth lag was presumably as- Exogenous cAMP overcomes this repressive sociated with the synthesis of enzymes needed effect of glucose (17, 20). If the effects of glucose for utilization of other components of the me- on the synthesis of SOD were due to this sort dium, such as organic acids, amino acids, pu- of catabolite repression, then we might expect rines, and pyrimidines. The increase in pH that cAMP would eliminate this glucose effect. following the exhaustion of glucose is consist- E. coli B organisms were transferred from a ent with the use of organic acids and nitroge- glucose-minimal to a TSY medium in the abnous compounds as an energy source. SOD sence and the presence of 8 mM cAMP. The content declined slightly during the growth results demonstrate that cAMP did eliminate dependent upon glucose and then increased catabolite repression (Fig. 5). Thus, the lag in sharply during the lag that preceded the second growth rate, which occurred as glucose was phase of growth. The influence of glucose was exhausted, was not seen in the presence of further probed by shifting cells from glucoseminimal to a TSY medium, which was devoid of glucose. Growth under these conditions was associated with an almost immediate increase in SOD, and, in the complete absence of the repressing effect of low levels of glucose, the growth lag was imperceptibly short (Fig. 4). It 32 30
00-0 2
28 -
X 2
0
26 x
0
24 .
0
20
c.x
0 0
x 22 . 0
0I
0 ID 20141 0 I B0 -J
0-
(I) 2
15 a
0 -
I0
6
0 1.2 I0 0o8
O
2
3
4
5
6
7
8
9
10
24
HOURS FIG. 3. Effects of nutritional shift-up (glucose minimal TSY) on the kinetics of growth and SOD synthesis in E. coli B. Growth was initiated in the TSY medium with a 5% inoculum taken from a late-logarithmic-phase culture in glucose-minimal medium. -*
0
1
2
3
4
5
6
06
HOURS FIG. 4. Kinetics of growth and SOD synthesis in E. coli B on a glucose-free TSY medium. Conditions were identical to those described in the legend of Fig. 3 except that the TSY medium was devoid of glucose.
508
HASSAN AND FRIDOVICH
J. BACTERIOL.
22 A 600
/_o0
20 0 O 0
8
20
0
24
x
X
I
O
2.2
.0-
I 6
0 0
0 2.0 (D
0
41
4
-
1-8
>
SOD
CLI
0
/P
0
2 -j
0
0
X
I 0
0
/ */
cn)
08
i 0
1
2
3
4
5
6
7
I0
IIII 1)5 8 9 10 11 12 13
HOURS
FIG. 5. Effect of exogenous cAMP on the kinetics of growth and SOD synthesis in E. coli B. Growth was initiated in TSY media, with or without 8 mM cAMP, with a 10% inoculum taken from a latelogarithmic-phase culture in glucose-minimal medium. Closed symbols represent absorbance at 600 nm, whereas open symbols represent SOD activity.
cAMP. cAMP did not, however, have any effect the kinetics of SOD synthesis under these conditions. Thus. SOD content diminished slightly during growth on glucose and then increased rapidly when growth shifted to noncarbohydrate components of the medium. This cAMP at 3 mM did allow an adenylate cyclasenegative mutant (cya C57; from W. Dobrogosz) to induce the enzymes required for the utilization of the lactose present in eosin methylene blue plates. We conclude from these results that the effect of glucose on SOD synthesis was probably not due to catabolite repression. Effect of a-methyl glucoside on SOD synthesis. It remained possible that adding 8 mM cAMP to the medium did not provide a high enough intracellular concentration of this cyclic nucleotide to overcome catabolite repression of SOD synthesis, another experimental approach was desirable. It has recently been shown that glucose lowers the intracellular level of cAMP by inhibiting adenylate cyclase and that the nonmetabolizable analogue a-methyl glucoside acts similarly (18). We therefore studied the effect of 0.5%7 a-methyl glucoside on SOD synthesis in the glucose-free TP medium. Figure 6 demonstrates that a-methyl glucoside did not on
-
O
2
3
4
5
6
7
8
S
HOURS FIG. 6. Effect of a-methyl glucoside on the kinetics of growth and SOD synthesis by E. coli B in the TP medium. Growth was initiated in the TP medium containing 0.5% a-methyl glucoside uwith a 5% inoculum taken from a late-logarithmic-phase culture in glucose-minimal medium. act like glucose
in preventing the induction of SOD, but it did decrease the specific growth rate and the yield of cells. We conclude that the effect of glucose in preventing the induction of SOD was not due to catabolite repression. Effect of elevated levels of glucose. Transfer of an inoculum from a glucose-minimal medium to a glucose-free TP medium resulted in an increase in SOD which commenced almost immediately (Fig. 7B). In contrast, when transfer was into a TP medium supplemented with 0.5% glucose, there was a 3-h lag before increased synthesis of SOD was apparent (Fig. 7A). The TSY medium previously used for such a nutritional step-up experiment (Fig. 3) contained only 0.25% glucose, which was virtually completely exhausted before the increase in SOD became apparent. In the TP medium containing 0.5% glucose there was still ample glucose remaining in the medium when SOD began to increase. This indicates that it is not the simple presence of glucose that keeps the level of SPD low. Increased SOD synthesis coincided with the inflection in the growth curve
REGULATION OF SUPEROXIDE DISMUTASE IN E. COLI
VOL. 132, 1977 268
2-8 - B
26
26
24
2-4
509
6X 25
22 0 0
2-2-
0 0 20
2.0
20 _
x
0 .1-8 0 (a 4c Is
x
1I8
(D
4% i6
I-
0
0
0
15 a 0
14
0
O
-i
SOD
p
en
-j
12
12/
I*0
10
0-8
0-8
0-6 O
2
3 4
5
05 6
7
HOU RS
HOURS
FIG. 7. Effect of 0.5% glucose on the kinetics of growth and SOD synthesis by E. coli B in TP medium. Growth was initiated in (A) TP medium plus 0.5% glucose or (B) TP medium. In both cases a 6% inoculum was taken from a late-logarithmic-phase culture in glucose-minimal medium.
which signaled a shift from primary dependglucose catabolism to other components of the medium, probably amino acids and other organic acids. We propose that the level of SOD in E. coli is controlled by the rate of production of its substrate, 0-. In that 34 case it must follow that fermentation of glucose is associated with a low rate of generation of -^oI 0.> compared with the oxidative catabolism of 26 amino acids and other organic acids. This leads to the prediction that a compound that is capa22j ble of causing the intracellular production of O°- would induce SOD, even when glucose fermentation was the primary energy source. 14 Relief of the glucose effect by paraquat. Paraquat (methyl viologen) is readily reduced I100 to a semiquinone, which rapidly reacts with molecular oxygen, giving rise to 02 (19) Paraquat has been shown to increase the rate 2 3 4 5 of production of 02- in biological systems (1, 2, HOURS 15, 16). We therefore tested the effect of adding FIG. 8. Effect of methyl viologen on the kinetics of paraquat to E. coli B growing in the TSY SOD synthesis by E. coli B in TSY medium. A 10% medium. Paraquat, at a final concentration of inoculum of washed cells taken from a late-logarith1.0 mM, caused an immediate and dramatic mic-phase culture in glucose-minimal medium was increase in the synthesis of SOD, even in the transferred into (a) TSY medium or (E) TSY containing 1 mM methyl viologen. presence of glucose (Fig. 8). ate dehydrogenase and isocitric dehydrogenase DISCUSSION (4). Ample glucose should thus minimize oxiWhen adequate glucose is available, E. coli dative metabolism in E. coli. In accord with derive most of their energy from the Embden- this expectation, E. coli B in the logarithmic Meyerhof pathway and then utilize the tricar- phase of growth in Trypticase soy broth exboxylic cycle mainly for the purpose of biosyn- hibited a smaller capacity for oxidative phosthesis (8). Glucose at a concentration equal to phorylation than did cells taken from the staor greater than 0.15% represses a-ketoglutartionary phase, after the glucose in the medium 50
ence upon
46
42
36
6
2
0
510
HASSAN AND FRIDOVICH
had been exhausted (11). We have seen that growth on glucose is associated with a low level of SOD and conclude that glucose does not directl depress the synthesis of SOD for the following reasons: e) cAMP did not reverse the glucose effect; (ii) ctmethyl glucoside did not mimic glucose; (iii) excess glucose did not prevent induction of SOD in the TSY and TP media once the pH had fallen to levels, which necessitated a shift from catabolism of glucose to use of nitrogenous compounds; and finally (iv) 1.0 mM methyl viologen caused a rapid and dramatic induction of SOD even in the presence of glucose. It appears likely that the level of SOD in E. (oli is controlled by the steady-state level of O. in the cells and that the fermentative catabolism of glucose leads to a lower rate of production of this radical than does the predominantly oxidative catabolism of nitrogenous compounds and organic acids. Cells placed in a complex medium such as the TSY medium at first derive most of their energy from fermentation and then shift to a predominantly oxidative metabolism when the glucose is exhausted or when the pH becomes dangerously low. The rate of synthesis of SOD is then a reflection of the rate of production of O,
J. BACTERIOL. monophosphate. J. Biol. Chem. 244:5828-5835. 4. Doelle, H. W., N. Hollywood, and A. W. Westwood. 1974. Effect of glucose concentration on a number of enzymes involved in the aerobic and anaerobic utili-
5.
6. 7. 8.
9.
10.
11.
12. 13.
14.
ACKNOWLEDG)GMENTS This work was supported by Public Health Service r1Xsearch grants GM-10287 and HL-17603 from the National Institute of General Medical Sciences and the National Heart and Lung Institute, respectively, and research grant DAHC-0470-G-0194 from the United States Army Research Office, Research Triangle Park, N.C.
16.
LITERATU'IRE CITEI)
17.
1. Asada, K., M. Takahashi, and K. Tanaka. 1977. Formation of active oxygen and its fate in chloroplasts. In 0. Hayaishi and K. Asada (ed. Proceedings of the Kyoto Symposium on Active Oxygen: Biochemical and Medical Aspects, 29 November 1976, Kyoto, Japan, in press. Academic Press Inc., New York. 2. Bus, J., S. Z. (agen. M. 0lgaard, and J. E. Gibson. 1976. A mechanism of paraquat toxicity in nmice and rats. Toxicol. Appl. Pharmacol. 35:501-513. 3. de (rombrugge, B., R. L. Perlman, H. E. Varonue,
and I. Pastan. 1969. Regulation of inducible enzyme synthesis in Escherichia co/i by cyclic adenosine-3 .5 -
15.
18.
19. 20.
zation of glucose in turbidostat cultures of Escherichia co/i. Microbios 9:221-232. Fridovich, I. 1972. Superoxide radical and superoxide dismutase. Acc. Chem. Res. 5:321-326. Fridovich, I. 1974. Superoxide dismutase. Advan. Enzymol. 41:35-97. Fridovich. I. 1975. Superoxide dismutase. Ann. Rev. Biochem. 44:147-159. Gray, C. T., J. W. T. Wimpenny, and M. R. Mossman. 1966. Regulation of metabolism in facultative bacteria. II. Effects of aerobiosis. anaerobiosis and nutrition on the formation of Krebs cycle enzymes in Escherichia coli. Biochim. Biophys. Acta 117:33-41. Hassan, H. M1., and I. Fridovich. 1977. Enzymatic defenses against the toxicity of oxygen and of streptonigrin in Escherichia coli. J. Bacteriol. 129:15741583. Hassan, H. MI., and I. Fridovich. 1977. Physiological function of superoxide dismutase in glucose-limited chemostate-cultures of Escherich ia coli. J. Bacteriol. 130:805-811. Hempfling, WA. P. 1970. Repression of oxidative phosphorylation in Escherichia coli B by growth in glucose and other carbohydrates. Biochem. Biophys. Res. Commun. 41:9-15. Lowry, 0. H., N. .1. Rosebrough, A. L. Farr, and R. J. Randall. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:265-275. Makman, R. S.. and E. W. Sutherland. 1965. Adenosine 3'.5-phosphate in Escherichia coli. J. Biol. ('hem. 240: 1309-1314. Mc(ord. J. M., and I. Fridovich. 1969. Superoxide dismutase: an enzymic function of erythrocuprein. J. Biol. Chem. 244:6049-6055. NMiller, R. W., and F. I). H. Macdowell. 1975. The tiron free radical as a sensitive indicator of chloroplastic photoautoxidation. Biochim. Biophys. Acta 387:176-187. Mlontgomery. M. R. 1976. Interaction of paraquat with pulmonary microsomal enzyme systems. Toxicol. Appl. Pharmacol. 37:106-107. Perlman. R. L., and I. Pastan. 1968. Regulation of /Bgalactosidase synthesis in Escherichia coli by cvclic adenosine 3.5'-monophosphate. J. Biol. Chem. 243:5420-5427. Peterkofsky. A., and C. Gazdar. 1974. Glucose inhibition of adenylate cyclase in intact cells of Escherichia coli B. Proc. Natl. Acad. Sci. U.S.A. 71:2324-2328. Stancliffe, T. (., and A. Pirie. 1971. The production of superoxide radicals in reactions of the herbicide diquat. FEBS Lett. 17:297-299. U'lIman. A., and J. M1onod. 1968. Cyclic AMP as an antagonist of catabolite repression in Escherichia coli. FEBS Lett. 2:57-60.
Vol. 132, No. 2
Printed in U.S.A.
Regulation of Superoxide Dismutase Synthesis in Escherichia coli: Glucose Effect H. MOUSTAFA HASSAN AND IRWIN FRIDOVICH* Department of Biochemistr,y, Duke University Medical Center, Durham, North Carolina 27710
Received for publication 13 May 1977
Growth of Escherichia coli based upon the fermentation of glucose, is associated with a low intracellular level of superoxide dismutase. Exhaustion of glucose, or depression of the pH due to accumulation of organic acids, causes these organisms to then obtain energy from the oxidative degradation of other substances present in a rich medium. This shift in metabolism is associated with a marked increase in the rate of synthesis of superoxide dismutase. Depression of the synthesis of superoxide dismutase by glucose is not due to catabolite repression since it is not eliminated by cyclic adenosine 3',5'-monophosphate and since a-methyl glucoside does not mimic the effect of glucose. Moreover, glucose itself no longer depresses superoxide dismutase synthesis when the pH has fallen low enough to cause a shift to a non-fermentative metabolism. It appears likely that superoxide dismutase is controlled directly or indirectly by the intracellular level of O., and that glucose depressed the level of this enzyme because glucose metabolism is not associated with as rapid a production of 0., as is the metabolism of many other substances. In accord with this view is the observation that paraquat, which can increase the rate of production of O,- by redox cycling, caused a rapid and marked increase in superoxide dismutase. The production of 0.. by the univalent reduction of molecular oxygen is a commonplace event, and the superoxide dismutases (SOD), which catalytically scavenge this free radical, appear to be essential defenses against its potential cytotoxicity. The work supporting these statements has been reviewed (5-7). Recent studies of Escherichia coli K-12, in glucoselimited chemostat culture. indicated that both oxygen uptake and content of SOD increased with increases in specific growth rate. The level of SOD, manipulated in this way, correlated with tolerance towards oxygen toxicity (10). These results and others, in which the level of SOD was modified by oxygen induction (9), support the view that SOD is essential for the aerobic growth of E. coli. During studies of the regulation of SOD synthesis in bacteria. we noted that the presence of glucose exerted profound effects, which gave the appearance of catabolite repression (H. M. Hassan and I. Fridovich, Fed. Proc. 36: 715, 1977). We now describe the effects of glucose and other nutrients on the synthesis of SOD in growing cultures of E. coli. MATERIALS ANt) MIETHODS Organisms. E. coli K-12 his thi (ATCC 23794) and E. coli B Bl2 (ATCC 29682) (provided by D. H. Hall) were used throughout.
Culture. Trypticase soy broth, trypticase peptone and yeast extract from the Baltimore Biological Laboratories and peptone and nutrient broth from Difco were used in preparing media. The Trypticase soy-yeast extract (TSY) medium contained 3% Trypticase soy broth and 0.5% yeast extract. This medium contains 0.25% glucose derived from the commercial Trypticase soy powder. The Trypticase peptone (TP) medium contained 2%c Trypticase peptone, 0.5% bactopeptone, 0.5% NaCl, and 0.25% K2HPO4. The TP medium was devoid of sugars. The TPY medium was identical to the TP medium except that 0.51% yeast extract was also present. The TPY medium was very similar to the TSY medium but it contained no glucose. Glucose-minimal medium contained (per liter): MgSO4 7H2O, 0.2 g; citric acid H2O, 2.0 g; K2HPO, 10.0 g; NaNH4HPO4 4H2O, 3.5 g; glucose, 5.0 g, and the essential growth factors B12 (1 mg) for E. colh B and histidine (300 mg) plus thiamine (30 mg) for E. coll K-12. Cultures were grown at 37°C on a rotatory shaker at 200 rpm with a ratio of flask volume to medium volume of 5:1. The size of the inoculum was varied to give an initial absorbancy at 600 nm in the range of 0.1 to 0.2. Growth was estimated in terms of turbidity measured as absorbancy at 600 nm,. and specific growth rates were calculated as described previously (10). Assays. Samples ot the cultures, collected at intervals, were chilled in ice and centrifuged at 10,000 x g for 15 min. The cells were washed in cold 50 mM potassium phosphate (pH 7.0) and, after suspension in 50 mM potassium phosphate, 0.1 mM
505
506
J. BACTERIOL.
HASSAN AND FRIDOVICH
ethylenediaminetetraacetic acid at pH 7.0, were disrupted by sonic oscillation for 3 min with a Branson W185 sonifier, operated at a power output of 50 W. The temperature during sonic treatment was maintained at 4 to 6°C by chilling in an ice-salt bath and by intermittent application of sonifier power. The homogenates were clarified at 30,000 x g for 1 h and then used directly for assays. Protein was estimated by the Lowry (12) method with pure bovine serum albumin as a calibrating standard. SOD was assayed as described previously (14). Samples of the culture medium were obtained by filtration through 0.45-gim membrane filters (Millipore Corp.) and assayed for residual glucose by the glucostat method (Sigma Chemical Co).
place of E. coli K-12. These changes in SOD were seen only with rich media that contained glucose. In contrast, growth in glucose-minimal medium was not accompanied by marked changes in SOD. Thus, growth of a log-phase inoculum, taken from a glucose-minimal medium, in glucose-minimal medium, exhibited a steady decline in pH during growth and no significant increase in SOD (Fig. 2). These results suggested that glucose might repress the biosynthesis of SOD.
RESULTS Effect of growth medium on SOD content of cells. As shown in Table 1. the specific 0 activity of SOD, present in cell-free extracts of 0 E. coli B, was strongly influenced by the culture medium. The level of SOD was not a 0 simple function of growth rate, as is the case (0 202 in glucose-limited chemostat culture (10), since growth was slower in TP medium than in the 0 nutrient broth plus 0.5% glucose medium, yet 0 the SOD content was two times higher with 0 cn) the former medium. Furthermore, cells grown in minimal medium containing succinate or lactate, as the sole carbon source, contained higher levels of SOD than did cells grown on glucose. Changes in SOI) during growth cycle in HOURS TSY medium. Figure 1 presents the changes FIG. 1. Kinetics of growth and SOD synthesis by in pH and SOD content during the growth of E. coli K-12 on TSY medium. Growth was E. coli K-12 in a TSY medium. Growth was initiated in the TSY medium with a 1% inoculum taken from started with an inoculum taken from a 17-h an early-stationary-phase culture in the same meculture in the same medium. It is apparent dium. that both pH and SOD concentrations declined for 5 h and then increased again. At zero time this medium was found to contain 13.9 mM glucose, which then diminished during 5 h of 7A-O* incubation to undetectably low levels. Similar results were seen when E. coli B was grown in 0 ,2 o
15a
0
TABLE 1. Effect of different growth media on the levels of superoxide dismutase in E. coli Ba Medium
Glucose-minimal
Generation time (min)
/
X 0
10
--
/
-
0
SOD
I0
'+o
SOD
_
0
D
(U/
mg)
46 6-8 Nutrient broth + 0.5% glucose 39 7-9 Minimal salts + 0.5% succinate 160 11-13 Minimal salts + 0.5% lactate 105 15-16 TP 42 16-18 TSY 32 21-24 (a The cells were grown in the different media at 37°C on a rotary shaker (200 rpm), and samples were removed at different intervals for growth and SOD assays. The values reported for SOD were taken at peak values (i.e., late-log phase of growth).
0
I
o
a
~-._pH
0
0
105
-i
cn
60
2
3
4
5
6
7
HOURS FIG. 2. Kinetics of growth and of SOD synthesis by E. coli B in a glucose-minimal medium. Growth was initiated in the glucose-minimal medium with a 5% inoculum taken from a logarithmic-phase culture in the same medium.
VOL. 132, 1977
507
REGULATION OF SUPEROXIDE DISMUTASE IN E. COLI
Effect of glucose on SOD synthesis. The appeared possible that glucose might repress influence of glucose was investigated by shift- the synthesis of SOD. ing cells from a glucose-minimal to a TSY Effect of cAMP on SOD synthesis. In E. medium. As shown in Fig. 3, growth under coli and other bacteria, the biosynthesis of these conditions was diauxic. There was an many inducible enzymes is regulated by the initial phase of rapid growth, which coincided intracellular level of cyclic adenosine 3',5'with a decrease in pH and with consumption of monophosphate (cAMP) (3). Glucose reduces glucose. This was followed by a lag in growth the level of cAMP in these bacteria and thus of approximately 3 h, after which growth was represses the synthesis of these enzymes (13). resumed. The growth lag was presumably as- Exogenous cAMP overcomes this repressive sociated with the synthesis of enzymes needed effect of glucose (17, 20). If the effects of glucose for utilization of other components of the me- on the synthesis of SOD were due to this sort dium, such as organic acids, amino acids, pu- of catabolite repression, then we might expect rines, and pyrimidines. The increase in pH that cAMP would eliminate this glucose effect. following the exhaustion of glucose is consist- E. coli B organisms were transferred from a ent with the use of organic acids and nitroge- glucose-minimal to a TSY medium in the abnous compounds as an energy source. SOD sence and the presence of 8 mM cAMP. The content declined slightly during the growth results demonstrate that cAMP did eliminate dependent upon glucose and then increased catabolite repression (Fig. 5). Thus, the lag in sharply during the lag that preceded the second growth rate, which occurred as glucose was phase of growth. The influence of glucose was exhausted, was not seen in the presence of further probed by shifting cells from glucoseminimal to a TSY medium, which was devoid of glucose. Growth under these conditions was associated with an almost immediate increase in SOD, and, in the complete absence of the repressing effect of low levels of glucose, the growth lag was imperceptibly short (Fig. 4). It 32 30
00-0 2
28 -
X 2
0
26 x
0
24 .
0
20
c.x
0 0
x 22 . 0
0I
0 ID 20141 0 I B0 -J
0-
(I) 2
15 a
0 -
I0
6
0 1.2 I0 0o8
O
2
3
4
5
6
7
8
9
10
24
HOURS FIG. 3. Effects of nutritional shift-up (glucose minimal TSY) on the kinetics of growth and SOD synthesis in E. coli B. Growth was initiated in the TSY medium with a 5% inoculum taken from a late-logarithmic-phase culture in glucose-minimal medium. -*
0
1
2
3
4
5
6
06
HOURS FIG. 4. Kinetics of growth and SOD synthesis in E. coli B on a glucose-free TSY medium. Conditions were identical to those described in the legend of Fig. 3 except that the TSY medium was devoid of glucose.
508
HASSAN AND FRIDOVICH
J. BACTERIOL.
22 A 600
/_o0
20 0 O 0
8
20
0
24
x
X
I
O
2.2
.0-
I 6
0 0
0 2.0 (D
0
41
4
-
1-8
>
SOD
CLI
0
/P
0
2 -j
0
0
X
I 0
0
/ */
cn)
08
i 0
1
2
3
4
5
6
7
I0
IIII 1)5 8 9 10 11 12 13
HOURS
FIG. 5. Effect of exogenous cAMP on the kinetics of growth and SOD synthesis in E. coli B. Growth was initiated in TSY media, with or without 8 mM cAMP, with a 10% inoculum taken from a latelogarithmic-phase culture in glucose-minimal medium. Closed symbols represent absorbance at 600 nm, whereas open symbols represent SOD activity.
cAMP. cAMP did not, however, have any effect the kinetics of SOD synthesis under these conditions. Thus. SOD content diminished slightly during growth on glucose and then increased rapidly when growth shifted to noncarbohydrate components of the medium. This cAMP at 3 mM did allow an adenylate cyclasenegative mutant (cya C57; from W. Dobrogosz) to induce the enzymes required for the utilization of the lactose present in eosin methylene blue plates. We conclude from these results that the effect of glucose on SOD synthesis was probably not due to catabolite repression. Effect of a-methyl glucoside on SOD synthesis. It remained possible that adding 8 mM cAMP to the medium did not provide a high enough intracellular concentration of this cyclic nucleotide to overcome catabolite repression of SOD synthesis, another experimental approach was desirable. It has recently been shown that glucose lowers the intracellular level of cAMP by inhibiting adenylate cyclase and that the nonmetabolizable analogue a-methyl glucoside acts similarly (18). We therefore studied the effect of 0.5%7 a-methyl glucoside on SOD synthesis in the glucose-free TP medium. Figure 6 demonstrates that a-methyl glucoside did not on
-
O
2
3
4
5
6
7
8
S
HOURS FIG. 6. Effect of a-methyl glucoside on the kinetics of growth and SOD synthesis by E. coli B in the TP medium. Growth was initiated in the TP medium containing 0.5% a-methyl glucoside uwith a 5% inoculum taken from a late-logarithmic-phase culture in glucose-minimal medium. act like glucose
in preventing the induction of SOD, but it did decrease the specific growth rate and the yield of cells. We conclude that the effect of glucose in preventing the induction of SOD was not due to catabolite repression. Effect of elevated levels of glucose. Transfer of an inoculum from a glucose-minimal medium to a glucose-free TP medium resulted in an increase in SOD which commenced almost immediately (Fig. 7B). In contrast, when transfer was into a TP medium supplemented with 0.5% glucose, there was a 3-h lag before increased synthesis of SOD was apparent (Fig. 7A). The TSY medium previously used for such a nutritional step-up experiment (Fig. 3) contained only 0.25% glucose, which was virtually completely exhausted before the increase in SOD became apparent. In the TP medium containing 0.5% glucose there was still ample glucose remaining in the medium when SOD began to increase. This indicates that it is not the simple presence of glucose that keeps the level of SPD low. Increased SOD synthesis coincided with the inflection in the growth curve
REGULATION OF SUPEROXIDE DISMUTASE IN E. COLI
VOL. 132, 1977 268
2-8 - B
26
26
24
2-4
509
6X 25
22 0 0
2-2-
0 0 20
2.0
20 _
x
0 .1-8 0 (a 4c Is
x
1I8
(D
4% i6
I-
0
0
0
15 a 0
14
0
O
-i
SOD
p
en
-j
12
12/
I*0
10
0-8
0-8
0-6 O
2
3 4
5
05 6
7
HOU RS
HOURS
FIG. 7. Effect of 0.5% glucose on the kinetics of growth and SOD synthesis by E. coli B in TP medium. Growth was initiated in (A) TP medium plus 0.5% glucose or (B) TP medium. In both cases a 6% inoculum was taken from a late-logarithmic-phase culture in glucose-minimal medium.
which signaled a shift from primary dependglucose catabolism to other components of the medium, probably amino acids and other organic acids. We propose that the level of SOD in E. coli is controlled by the rate of production of its substrate, 0-. In that 34 case it must follow that fermentation of glucose is associated with a low rate of generation of -^oI 0.> compared with the oxidative catabolism of 26 amino acids and other organic acids. This leads to the prediction that a compound that is capa22j ble of causing the intracellular production of O°- would induce SOD, even when glucose fermentation was the primary energy source. 14 Relief of the glucose effect by paraquat. Paraquat (methyl viologen) is readily reduced I100 to a semiquinone, which rapidly reacts with molecular oxygen, giving rise to 02 (19) Paraquat has been shown to increase the rate 2 3 4 5 of production of 02- in biological systems (1, 2, HOURS 15, 16). We therefore tested the effect of adding FIG. 8. Effect of methyl viologen on the kinetics of paraquat to E. coli B growing in the TSY SOD synthesis by E. coli B in TSY medium. A 10% medium. Paraquat, at a final concentration of inoculum of washed cells taken from a late-logarith1.0 mM, caused an immediate and dramatic mic-phase culture in glucose-minimal medium was increase in the synthesis of SOD, even in the transferred into (a) TSY medium or (E) TSY containing 1 mM methyl viologen. presence of glucose (Fig. 8). ate dehydrogenase and isocitric dehydrogenase DISCUSSION (4). Ample glucose should thus minimize oxiWhen adequate glucose is available, E. coli dative metabolism in E. coli. In accord with derive most of their energy from the Embden- this expectation, E. coli B in the logarithmic Meyerhof pathway and then utilize the tricar- phase of growth in Trypticase soy broth exboxylic cycle mainly for the purpose of biosyn- hibited a smaller capacity for oxidative phosthesis (8). Glucose at a concentration equal to phorylation than did cells taken from the staor greater than 0.15% represses a-ketoglutartionary phase, after the glucose in the medium 50
ence upon
46
42
36
6
2
0
510
HASSAN AND FRIDOVICH
had been exhausted (11). We have seen that growth on glucose is associated with a low level of SOD and conclude that glucose does not directl depress the synthesis of SOD for the following reasons: e) cAMP did not reverse the glucose effect; (ii) ctmethyl glucoside did not mimic glucose; (iii) excess glucose did not prevent induction of SOD in the TSY and TP media once the pH had fallen to levels, which necessitated a shift from catabolism of glucose to use of nitrogenous compounds; and finally (iv) 1.0 mM methyl viologen caused a rapid and dramatic induction of SOD even in the presence of glucose. It appears likely that the level of SOD in E. (oli is controlled by the steady-state level of O. in the cells and that the fermentative catabolism of glucose leads to a lower rate of production of this radical than does the predominantly oxidative catabolism of nitrogenous compounds and organic acids. Cells placed in a complex medium such as the TSY medium at first derive most of their energy from fermentation and then shift to a predominantly oxidative metabolism when the glucose is exhausted or when the pH becomes dangerously low. The rate of synthesis of SOD is then a reflection of the rate of production of O,
J. BACTERIOL. monophosphate. J. Biol. Chem. 244:5828-5835. 4. Doelle, H. W., N. Hollywood, and A. W. Westwood. 1974. Effect of glucose concentration on a number of enzymes involved in the aerobic and anaerobic utili-
5.
6. 7. 8.
9.
10.
11.
12. 13.
14.
ACKNOWLEDG)GMENTS This work was supported by Public Health Service r1Xsearch grants GM-10287 and HL-17603 from the National Institute of General Medical Sciences and the National Heart and Lung Institute, respectively, and research grant DAHC-0470-G-0194 from the United States Army Research Office, Research Triangle Park, N.C.
16.
LITERATU'IRE CITEI)
17.
1. Asada, K., M. Takahashi, and K. Tanaka. 1977. Formation of active oxygen and its fate in chloroplasts. In 0. Hayaishi and K. Asada (ed. Proceedings of the Kyoto Symposium on Active Oxygen: Biochemical and Medical Aspects, 29 November 1976, Kyoto, Japan, in press. Academic Press Inc., New York. 2. Bus, J., S. Z. (agen. M. 0lgaard, and J. E. Gibson. 1976. A mechanism of paraquat toxicity in nmice and rats. Toxicol. Appl. Pharmacol. 35:501-513. 3. de (rombrugge, B., R. L. Perlman, H. E. Varonue,
and I. Pastan. 1969. Regulation of inducible enzyme synthesis in Escherichia co/i by cyclic adenosine-3 .5 -
15.
18.
19. 20.
zation of glucose in turbidostat cultures of Escherichia co/i. Microbios 9:221-232. Fridovich, I. 1972. Superoxide radical and superoxide dismutase. Acc. Chem. Res. 5:321-326. Fridovich, I. 1974. Superoxide dismutase. Advan. Enzymol. 41:35-97. Fridovich. I. 1975. Superoxide dismutase. Ann. Rev. Biochem. 44:147-159. Gray, C. T., J. W. T. Wimpenny, and M. R. Mossman. 1966. Regulation of metabolism in facultative bacteria. II. Effects of aerobiosis. anaerobiosis and nutrition on the formation of Krebs cycle enzymes in Escherichia coli. Biochim. Biophys. Acta 117:33-41. Hassan, H. M1., and I. Fridovich. 1977. Enzymatic defenses against the toxicity of oxygen and of streptonigrin in Escherichia coli. J. Bacteriol. 129:15741583. Hassan, H. MI., and I. Fridovich. 1977. Physiological function of superoxide dismutase in glucose-limited chemostate-cultures of Escherich ia coli. J. Bacteriol. 130:805-811. Hempfling, WA. P. 1970. Repression of oxidative phosphorylation in Escherichia coli B by growth in glucose and other carbohydrates. Biochem. Biophys. Res. Commun. 41:9-15. Lowry, 0. H., N. .1. Rosebrough, A. L. Farr, and R. J. Randall. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:265-275. Makman, R. S.. and E. W. Sutherland. 1965. Adenosine 3'.5-phosphate in Escherichia coli. J. Biol. ('hem. 240: 1309-1314. Mc(ord. J. M., and I. Fridovich. 1969. Superoxide dismutase: an enzymic function of erythrocuprein. J. Biol. Chem. 244:6049-6055. NMiller, R. W., and F. I). H. Macdowell. 1975. The tiron free radical as a sensitive indicator of chloroplastic photoautoxidation. Biochim. Biophys. Acta 387:176-187. Mlontgomery. M. R. 1976. Interaction of paraquat with pulmonary microsomal enzyme systems. Toxicol. Appl. Pharmacol. 37:106-107. Perlman. R. L., and I. Pastan. 1968. Regulation of /Bgalactosidase synthesis in Escherichia coli by cvclic adenosine 3.5'-monophosphate. J. Biol. Chem. 243:5420-5427. Peterkofsky. A., and C. Gazdar. 1974. Glucose inhibition of adenylate cyclase in intact cells of Escherichia coli B. Proc. Natl. Acad. Sci. U.S.A. 71:2324-2328. Stancliffe, T. (., and A. Pirie. 1971. The production of superoxide radicals in reactions of the herbicide diquat. FEBS Lett. 17:297-299. U'lIman. A., and J. M1onod. 1968. Cyclic AMP as an antagonist of catabolite repression in Escherichia coli. FEBS Lett. 2:57-60.
