ARCHIVES

OF BIOCHEMISTRY

AND BIOPHYSICS

Vol. 294, No. 1, April, pp. 138-143, 1992

Effects of Overproduction of Superoxide Dismutases in EscherkIM coll’on Inhibition of Growth and on Induction of Glucose-6-phosphate Dehydrogenase by Paraquat’ Stefan

I. Liochev and Irwin Fridovich2

Department

of Biochemistry,

Duke University

Medical Center, Durham,

Received September 27, 1991, and in revised form November

25,199l

Stationary phase inocula were more susceptible to the growth inhibitory effect of paraquat than were log phase inocula and this difference was exacerbated in strains overproducing superoxide dismutases (SOD). Glucose-6 phosphate dehydrogenase (G-6-PD), a member of the soxR regulon, was induced by paraquat promptly in the case of log phase cells; but only after a lag in stationary phase cells and this difference was also exaggerated in strains overproducing SOD. The negative consequences of overproduction of SOD on the adaptation of stationary phase cells to paraquat may be attributed to competition for cellular resources with an attendant delay in biosynthesis of other components of soxR. Since overproduction of SOD did not prevent log phase cells from inducing G6-PD in response to paraquat, it appears likely that soxR can respond to aspects of redox status other than 0;. This conclusion is in accord with data which is already in the literature. 0 1992 Academic Press, ln0.

Gross overproduction of either MnSOD3 (1) or FeSOD (2), achieved with multicopy plasmids bearing the relevant genes, has been associated with an apparently paradoxical increase in sensitivity toward the growth-inhibiting effects of paraquat. We have examined the basis of the increased sensitivity toward PQ2+ of the SOD-overproducers and have reported (3) that they were less able to induce glucose-6-phosphate dehydrogenase than were control 1 This work was supported by research grants to I.F. from the National Science Foundation, the National Institutes of Health, and the Council for Tobacco Research-U.S.A., Inc., and to S.I.L. from the Bulgarian Academy of Sciences. ’ To whom correspondence should be addressed. Fax: (919) 684-8885. 3 Abbreviations used: MnSOD, the manganese-containing superoxide dismutase; FeSOD, the iron-containing superoxide dismutase; PQ*+ paraquat, G-6-PD, glucose-6-phosphate dehydrogenase; sodA, the gene coding for MnSOD; sodR, the gene coding for FeSOD. 138

North Carolina 27710

strains. Since G-6-PD is a member of the soxR regulon (4, 5), we surmised that overproduction of SOD was diminishing induction of the soxR regulon and was thus preventing the cells from achieving a balanced defense against the several consequences of PQ2+ action. Two possible explanations for this effect, of overproduction of SOD on SOXR, were proposed (3). One of these was lowering of the steady-state level of 0; to a point which precluded its induction of soxR. The other was a simple competition for amino acids and ATP, such that overproduction of SOD necessarily diminished biosynthesis of other proteins. The latter proposal seemed tenable because, in the case of the MnSOD overproducer, fully f of the extractable protein of cells grown in the presence of 50 PM PQ2+ was SOD. These possibilities, and the effects of PQ2+, could be further explored by comparing the responses to PQ2+ of SOD-defective, wild type, and SOD-overproducing strains, in both the logarithmic and the stationary phases of the growth cycle. Log phase cells are replete with ribosomes and the other components of the protein biosynthetic apparatus, whereas in stationary phase cells this machinery has been sharply down regulated (6). The inductive responses to PQ2+ should thus occur more rapidly with the log phase than with the stationary phase cells. Moreover, when multiple copies of the SOD genes lead to disproportionate induction of SOD, induction of other members of the soxR regulon in stationary phase cells should be strongly limited as long as this limitation is due to competition for energy and for biosynthetic building blocks. On the other hand if the suppression of soxR by overproduced SOD is due to lowering of the steadystate level of O,, then we need expect little difference between log phase and stationary phase cells beyond the lag in growth and protein synthesis usually exhibited by such cells. Results supporting the view that overproduction of SOD diminishes induction of soxR due to competition for 0003-9861/92 $3.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.

EFFECTS

OF OVERPRODUCTION

OF SUPEROXIDE

DISMUTASE

139

IN E. COLZ

2 3 4

2.0 1.0 0.6 0.6 0.4

E 8 a”

E 8

0.2

a" 0. I 0.06 0.06

0.02

I

I

t

I

0.011 0 I

2345676

0

Hours

energy and raw materials, and that induction not dependent upon O;, are reported below.

I2

3

4

5

6

Hours

FIG. 1. Effects of PQe+ on growth of stationary phase inocula. Overnight cultures of control and FeSOD-overproducing E. coli were diluted 1:lOO into fresh medium containing PQ*+ at 0.00 (lines 1 and 2), 0.05 (lines 3 and 4), or 0.9 mM (lines 5 and 6). Growth of the control (lines 1, 3, and 5) and of the FeSOD-overproducer (lines 2, 4, and 6) were followed turbidimetrically at 600 nm.

MATERIALS

:I----

FIG. 2. Effects of PQ” on growth of log phase inocula. Overnight cultures of control and of FeSOD-overproducing E. coli were diluted 1: 100 into fresh medium and these dilutions were incubated for 2 h before being diluted to Am nm = 0.03-0.035 in medium lacking (lines 1 and 2), or containing (lines 3 and 4) 0.1 mM P&“‘. Growth of the control (lines 1 and 3) and of the FeSOD-overproducer (lines 2 and 4) were followed in terms of ABoonm.

of soxR is

AND METHODS

Bacterial strains. DKl, a recA deletion strain of E. coli K12 MC1061, was used as the host for the following plasmids: pDTl-5 (bearing sodA); pHS l-4 (bearing sodl3); pHC (control for pDTl-5); and pBR.322 (control for pHSl-4). The irains bearing pDTl-5 and pHC were obtained from Bloch (1) while the strains bearing pBR.322 and pHSl-4 were prepared as described in the Maniatis Handbook (7). Strain J1132, defective in both sodA and sodB, and strain AB1157 (wild type) were obtained from J. A. Imlay (8). Growth of cultures. LB medium was used throughout. It contained 10 g NaCl, 5 g yeast extract, and 10 g bactotryptone per liter and was made up in deionized water and adjusted to pH 7.5. Cultures were grown aerobically at 37°C with shaking at 200 rpm. Stationary phase cells were taken from overnight cultures and were used to inoculate experimental cultures. Logarithmic phase cells were obtained by diluting overnight cultures -lOO-fold with fresh medium and allowing -2 h prior to use in experimental cultures. Tetracycline (12.5 pg/ml) or ampicillin (50 pg/ml) was used to maintain selection pressure for pBR322 and pHSl4, or pHC and pDTl-5, respectively. Reagents. NADP+, PQ*+, glucose B-phosphate, cytochrome c type III, ampicillin, tetracycline, and xanthine were obtained from Sigma and MgCl, was from Mallinckrodt. Xanthine oxidase, prepared from fresh bovine cream (S), was kindly provided by K. V. Rajagopalan. SOD activity was assayed by the xanthine oxidase/cytochrome c method (lo), while G-6-PD was assayed in terms of NADP+ reduction followed at 340 nm (11). Extracts were prepared and assayed for protein and enzymatic activities as previously described (3).

RESULTS Effectsof paraquat on growth. When stationary phase cells were inoculated into fresh medium (Fig. 1) a growth

lag was evident and this was more pronounced with the FeSOD overproducer (line 2) than with the control strain (line 1). PQ2+ at 0.05 M elicited a complex growth pattern in which the initial lag was followed by two periods of growth acceleration, the first presumably reflecting biosynthesis of ribosomes and of other components necessary for protein biosynthesis and the second reflecting induction of enzymes which provide adaptation to the stress imposed by PQ2+. The FeSOD overproducer (line 4) exhibited these complex kinetics, in the presence of 0.05 mM PQ2’, to a greater degree than did the control strain (line 3). At 0.9 mM PQ2+ the initial growth lag was lengthened and the second spurt of growth was not seen within the 8 h of observation. Once again the FeSOD overproducer (line 6) was more strongly affected than the control strain (line 5). This biphasic growth pattern of stationary phase inocula in PQ2+-containing media was seen repeatedly during the course of these investigations. Moreover they have been reported previously although left unremarked (12). By way of explanation, a strain overproducing any single protein, such as FeSOD, would, because of this diversion of cellular resources, take longer to achieve the necessary up-regulation of the protein biosynthetic apparatus; this up-regulation is a necessary prerequisite to the adaptive biosynthesis of defensive and of reparative enzymes. When log phase inocula were used (Fig. 2) the results were strikingly different. Thus, no lags in growth were seen and the FeSOD overproducer (line 2) grew only

140

LIOCHEV

AND

FRIDOVICH

Hours

Hours

FIG. 3. Comparison of effects of PQ2+ on stationary and log phase inocula. An aliquot from an overnight culture of the control strain (DKl/ pBr) was diluted 1:lOO with fresh medium and was then incubated for 1.5 h to put it into log phase. Aliquots of the overnight and of the log phase subculture were then diluted to approximately equal Asoollmin fresh medium + 0.1 mM PQ x+. Line 1, log phase inoculum, line 2, stationary phase inoculum; line 3, log phase inoculum + 0.1 mM PQ2+; line 4, stationary phase inoculum + 0.1 mM PQ’+.

FIG. 4. Effect of overproduction of FeSOD on induction of G-6-PD. Aliquots of overnight cultures were diluted 1:lOO into fresh medium and were considered to be in log phase when the initial ABOO nmhad increased fivefold. Other aliquots of the overnight cultures were then diluted 1:20 into fresh medium and Pp was added to 0.1 mM to all cultures. At intervals thereafter aliquots were removed for assay of G-6-PD. Lines 1 and 2, log phase inocula of control or of FeSOD-overproducing strains; line 3, stationary phase inoculum of the control strain; line 4, stationary phase inoculum of the FeSOD overproducer.

lag in the induction of G-6-PD by PQ2+ and this lag was greater in the FeSOD overproducer (line 4) than in the control strain (line 3). Approximately equal levels of G6-PD activity were ultimately achieved in all cases. The difference between log phase and stationary phase inocula was even more profound when control and MnSOD-overproducing strains were compared. As shown in Fig. 5, 0.1 mM PQ2’ caused prompt induction of G-6PD in log phase cells, with the control strain (line 1) in-

somewhat slower than the control (line 1). Moreover, PQ’+ at 0.1 mM was less inhibitory toward the log phase inocula (lines 3 and 4) than half that concentration had been toward stationary phase inocula (Fig. 1). The different growth patterns and responses to PQ2+of log phase and stationary phase inocula are made clearer by the data in Fig. 3. In this case an aliquot taken from an overnight culture of the control strain was put into log phase by incubation for 1.5 h after loo-fold dilution with fresh medium. Approximately equal numbers of cells, as judged by &oonm, were then taken from the original overnight culture and from the log phase subculture and were inoculated into medium fO.l mM PQ2+. In the absence of PQ’+ the stationary phase inoculum (line 2) exhibited the expected growth lag, whereas the log phase inoculum (line 1) did not. PQ2+, at 0.1 mM, imposed a much greater growth inhibition on the stationary phase inoculum (line 4) than it did on the log phase inoculum (line 3). Log phase cells, already possessing an up-regulated protein synthetic capacity, could rapidly adapt to the oxidative I I I I I I I ‘0 I 2 3 4 5 6 stress imposed by PQ’+ and were thus less affected by it. Hours Induction of glucose-6-phosphate dehydrogenase. When log phase inocula were exposed to 0.1 mM PQ2+ FIG. 5. Effect of overproduction of MnSOD on induction of G-6-PD. they showed a rapid induction of G-6-PD and this was Log and stationary phase cultures were exposed to 0.1 mM PQ2+ and induction of G-6-PD was followed as in the legend to Fig. 4. Line 1, log the same in the control or in the FeSOD overproducer, phase inoculum of the control strain; line 2, log phase inoculum of the as shown by lines 1 and 2 in Fig. 4. In contrast, when MnSOD-overproducer; line 3, stationary phase inoculum of the control stationary phase inocula were used there was a marked strain: line 4, stationary phase inoculum of the MnSOD-overproducer.

EFFECTS OF OVERPRODUCTION

OF SUPEROXIDE

DISMUTASE

l-II1 -024

01 ’ ’ 0

2 4

L

IO PQ++

3

J

I rl 20 25

OLuM)

FIG. 7. Induction of SOD by PQ*‘. Log phase cultures were exposed to the indicated concentrations of PQ2+ and at 2.5 h thereafter samples were taken for assay of SOD activity. Line 1, FeSOD-overproducer; line 2, control (pBR322) strain; line 3, SOD-null.

I IO PC)++ (AM

141

IN E. COLI

25

FIG. 6. Induction of G-6-PD in the SOD-null. PQ’+ was added to the indicated concentrations to cultures in logarithmic growth and samples were removed at 2.5 h thereafter for assay of G-6-PD. Line 1, SOD-null strain; line 2, control (pBR322) strain.

parent, this induction was diminished, presumably because damage due directly or indirectly to 0, prevented up-regulation of the protein biosynthetic apparatus. DISCUSSION Overproduction of MnSOD has been reported to sensitize E. coli to the growth inhibiting effects of PQ2’ (1). Since this was done with stationary phase cells, we repeated these conditions and noted that induction of G6-PD by PQ” was suppressed by overproduction of MnSOD (3). We now find that log phase cells are less prone to these negative effects of PQ2’ than are stationary phase cells. Marked lags in both growth and induction of G-6-PD were seen with stationary phase inocula and presumably reflect the need to rebuild the down-regulated protein biosynthetic apparatus. Overproduction of SOD, or of any other protein, would divert cellular resources from the biosynthesis of ribosomes, and of the other components essential for protein synthesis. This would slow

ducing more rapidly than the MnSOD-overproducer (line 2). The control cells in the stationary phase (line 3) exhibited a 2-h lag in the induction of G-6-PD, while the stationary phase MnSOD-overproducer did not detectably induce G-6-PD during 6 h of incubation with PQ2’ (line 4). Since the MnSOD-overproducer responds to PQ2+with a larger biosynthesis of MnSOD than does the FeSOD overproducer, its greater deficit in G-6-PD biosynthesis is not surprising. When the induction of G-6-PD by log phase inocula was examined, as a function of the concentration of PQ”’ (Fig. 6), the SOD-null strain (line 1) was seen to be much more responsive than the control strain (line 2). This indicates that the high levels of O;, achievable in the absence of SOD, can elicit induction of the soxR regulon. TABLE I The effect of paraquat on SOD induction by log phase Effects of PQ’+ on Stationary PhaseInocula of the SOD-Null inocula was similarly examined (Fig. 7). The FeSODand Wild Type overproducer did not significantly increase its net content of SOD (line 1). The control strain (line 2) did show a SOD PQ” G-6-PD fourfold induction of SOD in response to PQ2’, whereas (units/mg) Strain (units/mg) (PM) Aeoonm the SOD-null had no activity in any case (line 3). 0 26.7 AB1157 3.65 0.08 Table I summarizes the effects of PQ2+ on stationary 4 31.1 AB1157 3.55 0.10 phase SOD-null and control inocula, in terms of growth AB1157 10 3.50 40.1 0.13 and of induction of G-6-PD and of SOD. The control AB1157 25 3.15 52.6 0.28 strain responded to PQ2’ with inductions of both SOD SOD-null 0 2.55 0 0.09 and of G-6-PD and grew to an Aem,,, of over 3.5. The SOD-null 4 1.64 0 0.23 10 0.76 0 0.19 growth of the control strain was only weakly inhibited by SOD-null 25 0.66 0 0.09 25 PM PQ”‘. In contrast the SOD-null achieved an Am nm SOD-null of only 2.5 in the absence of PQ2+;the SOD-null exhibited a Inocula from overnight cultures were diluted 1:20 into fresh medium some induction of G-6-PD at 4 PM PQ2+, but at higher + the indicated concentrations of PQ” and 4 h thereafter A mnm, SOD, concentrations of PQ2’, where growth inhibition was ap- and G-6-PD were measured.

142

LIOCHEV

AND FRIDOVICH

induction of the proteins of the oxyR (13) and the soxR (4, 5) regulons and thus delay attainment of a balanced defense against the oxidative stress imposed by PQ”.‘ In the case of the MnSOD-overproducer, SOD accumulated to -30% of the extractable protein, while in the case of the FeSOD-overproducer, SOD accumulated to - 10%. To achieve such great accumulations of SOD a large fraction of protein synthesis must have been devoted to SOD, inescapably diminishing the biosynthesis of other proteins. It should be anticipated that gross overproduction of any one protein should limit the ability of the cells to achieve adaptive responses,by induction of other proteins. Moreover, this effect should be greater with the stationary phase, than with log phase, inocula. Touati (27) has examined the induction of /3-galactosidase in strains of Escherichia coli bearing the sodAlacZ protein fusion. She noted marked inductions of @galactosidase in response to dioxygen, paraquat, or quinones, to which the sodA operator is responsive. Incorporation of plasmids carrying the sodA gene caused overproduction of MnSOD but inhibited induction of P-galatosidase. This inhibitory effect did not depend upon the catalytic activity of MnSOD. Thus, it was seen whether or not the medium was enriched with Mn(II), even though this enrichment strongly increased the level of active SOD. Moreover, it was also seen when the plasmid carried truncated sodA, such that its protein product was inactive. This was interpreted (27) in terms of autogenous repression of the sodA operator by the sodA gene product. It can now be seen as the result of a competition for amino acids and other cellular resources needed for biosynthesis of P-galactosidase on the one hand, and the complete MnSOD, or fragments thereof, on the other hand. Laudenbach et al. (28) examined the effects of overproduction of MnSOD or of FeSOD, due to multicopy plasmids bearing the corresponding genes, on the paraquat-resistance of E. coli. When examining rates of growth in liquid medium, they used log phase inocula and did not see significant effects. In contrast, when examining colony-forming ability on solidified medium they used stationary phase inocula and saw a diminution in colony counts due to the plasmids. This is another case of overproduction of one protein being more detrimental to stationary phase cells, which must up-regulate the protein synthetic apparatus, than to log phase cells, in which this apparatus is already in place. Log phase cells reacted to PQ’+ by prompt induction of G-6-PD, and a FeSOD-overproducer was as responsive as a control strain. This indicates that induction of G-6PD, and by extension of the soxR regulon, need not be a direct response to 0; per se. PQ”,‘ while increasing production first of 0; and then of HzOz, takes electrons from NAD(P)H. In the presence of the normal levels of SOD and of the hydroperoxidases, the net effect would be a decrease in NAD(P)H/NAD(P)+ which would lead to a

decrease in GSH/GSSG. Rather than being a response to 0,) it seems likely that soxR is responsive to the redox status of the cell as reflected in these ratios. This likelihood is supported by the abilities of diverse oxidants to cause the anaerobic inductions of MnSOD (14-19) and of G-6-PD (20). It should be recalled, in this context, that the in uitro transcription of the E. coli MnSOD gene was suppressed by thiols and by NADPH (14). The micromolar steady-state levels of 0; , estimated to be present in SOD-nulls (21), is sufficient to directly inactivate susceptible enzymes, such as the [4Fe-4S]-containing dehydratases (22-24). It is probably also enough to directly or indirectly influence the cellular redox status, However, in SOD-normal E. coli the steady state [O;] is approximately 10-l’ M (21) and this is probably too low to significantly inactivate enzymes or to modulate the activities of regulatory proteins. It is thus informative, vis-a-vis the physiological significance of SOD, to compare normal with SOD-null strains, but it is not useful to compare SOD-normal with SOD-overproducing strains. In the latter case the negative effect of wasting cellular resources on production of a surplus of SOD is not offset by the marginal benefit of lowering 0; below 10-l’ M. Failure to recognize this trade-off has led several groups to erroneous conclusions (1, 2, 25). The FeSOD-overproducer responded to 0.1 mM PQ”’ by inducing G-6-PD and did so as well as the control strain (Fig. 4). Yet the FeSOD-overproducer did not appear to respond to PQ”’ with a net increase in SOD (Fig. 7). Since G-6-PD and MnSOD are controlled by soxR this seems paradoxical. PQ”’ has been reported to induce MnSOD in an FeSOD-overproducing strain of E. coli (26) and this increase in MnSOD was accompanied by a decrease in FeSOD such that the net SOD was not significantly changed. This explains our results but also indicates that FeSOD, although considered to be constitutive, is somehow regulated. REFERENCES 1. Bloch, C. A., and Ausubel, F. M. (1986) J. Bacterial. 168,795-798. 2. Scott, M. D., Meshnick, S. R., and Eaton, J. W. (1987) J. Biol. Chem. 262,3640-3645.

3. Liochev, S. I., and Fridovich, I. (1991) J. Biol. Chem. 266, 87478750. 4. Greenberg, J. T., Monach, P., Chou, J. H., Josephy, P. D., and Dempie, B. (1990) Proc. Natl. Acud. Sci. USA 87, 6181-6185. 5. Tsaneva, I. R., and Weiss, B. (1990) J. Bacterial. 172, 4197-4205. 6. Davis, B. D. (1989) Proc. N&l. Acad. Sci. USA 86, 5005-5009. 7. Maniatis, T., Fritsch, E. F., and Sanbrook, J. Molecular Cloning, A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 1982. 8. Imlay, J. A., and Linn, S. (1987) J. Bacterial. 169, 2967-2976. 9. Waud, W. R., Brady, F. O., Wiley, R. D., and Rajagopalan, K. V. (1975) Arch. Biochem. Biophys. 169, 695-701. 10. McCord, J. M., and Fridovich, I. (1969) J. Biol. &em. 244,6049-

6055.

EFFECTS

OF OVERPRODUCTION

OF SUPEROXIDE

11. Kao, S. M., and Hassan, H. M. (1985) J. Bid. &em. 260,10,47810,481. 12. Minakami, H., and Fridovich, I. (1990) FASEB J. 4,3239-3244. 13. &on, G., Tart&i, L. A., Farr, S. B., andAmes, B. N. (1999) Trends Genet.6.363-368. 14. Gardner, P. R., andFridovich, I. (1987) J. Biol. C/tern. 262,17,59117,595. 15. Smith, M. W., and Neidhardt, F. C. (1983) J. Bade&L 164,344350. 16. Hassan, H. M., andMoody, C. S. (1987) J. Bid. Chem. 262,17,17317,177. 17. Miyake, K. (1986) J. Gen. Appl. Micrabid. 22,527-533. 18. Schiavone, J. R., and Hassan, H. M. (1988) J. Bid. Chem. 263, 4269-4273. 19. Privalle, 4279.

C. T., and Fridovich,

I. (1988) J. Bid. Chem. 263, 4274-

DISMUTASE

IN E. COLI

143

20. Privahe, C. T., and Fridovich, I. (1990) J. BioL Chem. 265,21,96621,970. 21. Imlay, J. A., and Fridovich, I. (1991) J. BioL Chem. 266, 69576965. 22. Kuo, C. F., Mashino, T., and Fridovich, I. (1987) J. Bid. &em. 262.4724-4727. 23. Gardner, P. R., and Fridovich, I. (1991) J. Bid. Chem. 266,14781483. 24. Gardner, P. R., andFridovich, I. (1991) J. Bid. Chem. 266,1932819333. 25. Siwecki, G., and Brown, D. R. (1996) Biochem. Znt. 20, 191-266. 26. Nettleton, C. J., Bull, C., Baldwin, T. O., and Fee, J. A. (1984) Proc. Natl. Ad. Sci. USA 61.4970-4973. 27. Touati, D. (1988) J. BacterioL 170,2511-2520. 28. Laudenbach, D. E., Trick, C. G., and Straus, N. A. (1989) Mol. Gen. Genet. 216,455-461.

Effects of overproduction of superoxide dismutases in Escherichia coli on inhibition of growth and on induction of glucose-6-phosphate dehydrogenase by paraquat.

Stationary phase inocula were more susceptible to the growth inhibitory effect of paraquat than were log phase inocula and this difference was exacerb...
614KB Sizes 0 Downloads 0 Views