AKCHIVES

OF RIOC’HEMIS?‘KY

AND

BIOPHYSICS

Vol. 279, No. 1, May 15, pp. 195-201, 1990

Evidence that Chemical Modification of a Positively Charged Residue at Position 189 Causes the Loss of Catalytic Activity of Iron-Containing and ManganeseContaining Superoxide Dismutases’ V. W. F. Ghan,” M. J. Bjerrum,? and C. L. Borders, Jr.*,2 Departments of Chemistry, *The College of Wooster, Wooster, Ohio 44691, and tThe Royal Veterinary and Agricultural

Received

November

University,

1871 Frederiksberg

C, Denmark

14,1989, and in revised form January

11,199O

The Escherichia coli, Bacillus stearothermophilus, and human manganese-containing superoxide dismutases (MnSODs) and the E. coli iron-containing superoxide dismutase (FeSOD) are extensively inactivated by treatment with phenylglyoxal, an arginine-specific reagent. Arg-189, the only conserved arginine in the primary sequences of these four enzymes, is also conserved in the three additional FeSODs and five of the six additional MnSODs sequenced to date. The only exception is Saccharomyces cerevisiae MnSOD, in which it is conservatively replaced by lysine. Treatment of S. cereuisiae MnSOD with phenylglyoxal under the same conditions used for the other SODS gives very little inactivation. However, treatment with low levels of 2,4,6-trinitrobenzenesulfonate (TNBS) or acetic anhydride, two lysine-selective reagents that cause a maximum of 60-80% inactivation of the other four SODS, gives complete inactivation of the yeast enzyme. Total inactivation of yeast MnSOD with TNBS correlates with the modification of approximately five lysines per subunit, whereas six to seven acetyl groups per subunit are incorporated on complete inactivation with [14C]acetic anhydride. It appears that the positive charge contributed by residue 189, lysine in yeast MnSOD and arginine in all other SODS, is critical for the catalytic function of MnSODs and FeSODs. ccl 1990 Academic Press, Inc.

’ This work was supported by grants to C.L.B. from The Petroleum Research Fund, administered by the American Chemical Society, the Research Corporation, the Henry Lute III fund, the Howard Hughes Medical Institute, and to M.J.B. from the Danish Natural Science Research Council (grant nos. 11-7783 and 11.7598). ’ TO whom correspondence should be addressed. ” Abbreviations used: SOD, superoxide dismutase; MnSOD, manganese-containing SOD; FeSOD, iron-containing SOD; Cu,ZnSOD, Copyright I.C 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.

The manganese-containing (MnSOD)3 and iron-containing (FeSOD) superoxide dismutases are highly homologous to each other, but are unrelated to the more thoroughly characterized Cu,ZnSODs at all levels of structure (l-5). Three-dimensional structures have been determined for the FeSODs from Escherichia coli (1, 6) and Pseudomonas oval& (2, 6) and the MnSODs from Thermus thermophilus (7) and Bacillus stearothermophilus (8). The active site metal ligands have been identified as 3 histidines and 1 aspartate in all four structures and these residues are conserved in the 13 members of this class sequenced to date (vide infra). A notable feature of MnSOD and FeSOD structures is a conserved apolar core of approximately 10 or 11 aromatic residues immediately surrounding the metal-ligand cluster (7-9); i.e., in B. stearothermophilus MnSOD, excluding the histidine metal ligands, there are 3 tryptophans, 3 tyrosines, 3 histidines, and 2 phenylalanines within 10 A of the catalytically essential metal ion (8). Chemical modification is an approach frequently used to identify amino acid residues that may be important for enzyme catalytic activity. Inactivation of E. coli FeSOD by Hz02 has been used to confirm the importance of tryptophan for normal catalysis (10). Recent work (11) has suggested that arginyl residues may be important for the catalytic activity of the E. coli MnSOD and E. coli FeSOD. Examination of the aligned amino acid sequences of these two proteins reveals that the only highly conserved arginine is at position 189 (vide infra). This arginine is present in 12 of the 13 MnSOD

copper- and zinc-containing nate.

SOD; TNBS,

2,4,6-trinitrobenzenesulfo-

195

196

CHAN.

BJERRUM,

AND

BORDERS

and FeSOD sequences reported to date, with the only exception being Saccharomyces cerevisiae (yeast) MnSOD, where it is conservatively replaced by a lysine. We have examined the inactivation of E. coli FeSOD and the MnSODs from S. cereuisiae, E. coli, B. stearothermophilus and human by phenylglyoxal, a reagent specific for the modification of arginyl residues in proteins, and 2,4,6-trinitrobenzenesulfonate (TNBS) and acetic anhydride, two lysine-selective reagents. The results strongly suggest that the positively charged amino acid side chain at position 189 in all MnSODs and FeSODs is critical for catalytic function. EXPERIMENTAL Materials. E. coli MnSOD, E. co/i FeSOD, and B. stearothermophilus MnSOD were purchased from Sigma Chemical Co. and were used without further purification. 8’. cereuisiae MnSOD was purified and handled according to procedures described elsewhere (12), while human MnSOD, a recombinant protein expressed in E. coli, was prepared by BioTechnology General Ltd., Rehovot, Israel and was a generous gift from Dr. Irwin Fridovich. Phenylglyoxal and TNBS were purchased from Aldrich Chemical Co. and acetic anhydride was from J. T. Baker Chemical Co. [l-‘4C]Acetic anhydride (5.0 mCi/mmol) was a product of DuPont NEN Research Products, while Ecolume scintillation fluid was obtained from ICN. SOD assay. SOD activity was determined by the ability of the protein to inhibit pyrogallol autoxidation (13) as described elsewhere (14, 15). Appropriate corrections in the blanks were made for the small changes in the rate of autoxidation caused by phenylglyoxal, TNBS, acetonitrile, and/or buffer salts. Chemical modifications. Modifications were carried out under conditions described in the figure legends. Reactions were initiated by the addition of a stock solution of the enzyme to an aqueous solution of the reagent, or by the addition of a stock solution of the reagent in acetonitrile to an aqueous solution of the enzyme. By following the change in pH on addition of an aliquot of acetic anhydride in acetonitrile to the appropriate buffer, it was found that this reagent is almost completely hydrolyzed in 2 min; thus, the reaction of acetic anhydride with the enzymes is finished in the times used for these modifications. At appropriate times after mixing, aliquots were diluted and assayed for SOD activity. The controls, including those containing 10% (v/v) acetonitrile, were stable throughout all experiments. Modification of amino acids by phenylglyoxal. Correlation of the inactivation of E. coli MnSOD by 20 mM phenylglyoxal and S. cereuisiae MnSOD by 40 mM phenylglyoxal with the loss of arginine and other amino acids was patterned after published procedures (14, 16). Amino acid analyses were carried out by Dr. Eric Eccleston at the University of Minnesota using a Beckman 6300 amino acid analyzer. The protein concentration of each sample was determined from quantitative amino acid analyses and the known amino acid composition of each enzyme. The values of arginine and lysine per SOD subunit in each hydrolyzed sample were normalized to B Leu + Tyr + Phe + His = 47 or 43 per subunit for E. coli MnSOD (17, 18) and S. cerevisiue MnSOD (19, 20), respectively. Stoichiometry for inactiuution of S. cerevisiae MnSOD by TNBS and acetic anhydride. For the modification of lysine by TNBS, S. cereuisiue MnSOD concentration was determined from the absorbance at 282 nm (c = 1.92 X lo5 Mm’. cm-‘) (21) and a molecular weight for the tetramerit enzyme of 90,800 (19, 21). The number of lysines modified was estimated from the increase in AzG7with time, using a molar absorptivity at this wavelength, chosen because it is the isosbestic point in the uv-visible spectrum of t-2,4,6-trinitrophenyllysine and its complex

240

120

360

MINUTES FIG. 1. Inactivation of superoxide dismutases by phenylglyoxal. The enzymes (0.08-0.56 mg/ml) were treated at 25°C with 20 mM phenylglyoxal in 50 mM sodium pyrophosphate, 50 mM NaHCO?, 0.1 mM EDTA, pH 9.0, and at the indicated times aliquots were diluted and assayed for enzymatic activity as described in the text. (A) B. steurothermophilus MnSOD, (0) E. coli MnSOD, (0) E. coli FeSOD, (0) human MnSOD, (w) S. cereuisiue MnSOD.

with sulfite, of 1.05 X lo4 M-i cm-i (22). At appropriate times aliquots were withdrawn from the modified sample and assayed for SOD activity. To correlate the inactivation of S. cereuisiue MnSOD by [i4C]acetic anhydride with loss of activity, the enzyme was treated with acetic anhydride for an appropriate time before gel filtration on a column of Sephadex G-25 to remove excess reagent. Fractions containing protein were pooled and the enzyme concentration was determined from the A 2R2> several aliquots were diluted and assayed for SOD activity, and an aliquot was diluted into 5 ml Ecolume scintillation fluid and counted on a Beckman LS-1OOC scintillation counter. Sequence alignments. Alignment of the MnSOD and FeSOD sequences was carried out using the IBI/Putsell Sequence Alignment Program, version 1.80. Minor adjustments were made by eye to optimize the alignments of the 13 sequences. Viewing of T. thermophilus MnSOLI structure. An Evans and Sutherland Model 390 molecular graphics system was used to view the 1.8-A refined tertiary structural data of T. thermophilus MnSOD kindly provided by Dr. Martha L. Ludwig. RESULTS

Figure 1 shows the rate of inactivation of SODS from five different sources-MnSODs from E. coli, B. stearothermophilus, S. cereuisiae, and human and the FeSOD from E. coli-by 20 mM phenylglyoxal at pH 9.0 and 25°C. It has been reported (11) that E. coli FeSOD is extensively inactivated by phenylglyoxal in a pseudofirst-order process, while the inactivation of E. coli MnSOD is biphasic. The data in Fig. 1 are consistent with

AN ESSENTIAL

POSITIVE

CHARGE

x

0 0

10

5

[ACETIC

ANHYDRIDE],

mM

FIG. 2. Inactivation of superoxide dismutases by acetic anhydride. The enzymes (0.1 l-0.45 mg/ml) were treated for l-2 h at 25°C with the indicated concentration of acetic anhydride in 45580 mM sodium 0.1 mM EDTA, pH 9.0, 10% pyrophosphate, 4580 mM NaHCO,, (v/v) acetonitrile, and then multiple aliquots of each sample were diluted and assayed for residual enzyme activity as described in the text. (0) 13. stearothermophilus MnSOD, (0) E. coli MnSOD, (w) E. coli FeSOD, (0) human MnSOD, (A) S. ccrcuisiae MnSOD.

and show, in addition, that human and MnSODs are inactivated at comparable rates. The times required to obtain 50% inactivation of E. coli MnSOD, E. coli FeSOD, and B. stearothermophilus and human MnSOD are approximately 0.6, 1, 0.2, and 0.5 h, respectively. However, yeast MnSOD is inactivated much more slowly than any of the others, with 50% inactivation occurring only after approximately 7-8 h under these conditions. It has been reported (11) that the E. coli SODS are partially inactivated by treatment with TNBS and acetic anhydride, but even after extensive modification with relatively high concentrations of these reagents, 20-45% residual activity remains. Figure 2 shows the effect of 0.02-9 mM acetic anhydride, at pH 9.0, on the five SODS examined in the current study. E. coli FeSOD, E. coli MnSOD, B. stearothermophilus MnSOD, and human MnSOD all show some inactivation, but all plateau at a significant residual activity at the highest concentrations of reagent. This is in contrast to S. cerevisiae MnSOD, which is >99% inactivated by as little as 0.5 mM acetic anhydride. Complete inactivation of the yeast enzyme is almost certainly due to the modification of lysine, since enzyme modified to 12 or 0.4% residual activity shows no increase in activity on treatment with 0.1 M NH,OH at pH 9.0 for l-2 h (data not shown), a process that would reverse 0-acetylation.

IN FeSOD

AND

197

MnSOD

The time course for the inactivation of the five SODS by 0.5 mM TNBS at pH 9.0 and 25°C is shown in Fig. 3. The activity of yeast MnSOD is by far the most sensitive to treatment with 0.5 mM TNBS under these conditions, as >98% inactivation is observed after approximately 2 h. The other four SODS retain 20-50% residual activity after approximately 4 h of modification. This appears to be a limit residual activity for E. coli FeSOD and E. coli MnSOD, for treatment of either with 5-8 mM TNBS under similar conditions does not reduce the activity further (data not shown). In contrast, human MnSOD, although much less sensitive than the yeast enzyme to treatment with 0.5 mM TNBS, is approximately 98% inactivated after 4 h modification with 5 mM reagent (data not shown). A possible explanation for the observation that human MnSOD is more sensitive than the E. coli SODS toward inactivation by high concentrations of TNBS may lie in the fact that, while the latter two are dimers, human MnSOD is a tetramer. Extensive conversion of positively charged lysines to bulky nonpolar 2,4,6-trinitroaniline derivatives may cause dissociation of the human enzyme into inactive dimers, in line with the observation that human MnSOD is much less stable than the dimeric E. coli SODS under a variety of experimental conditions (W. F. Beyer, Jr., personal communication). An attempt was made to correlate the inactivation of E. coli MnSOD and S. cerevisiae MnSOD by phenylgly-

this observation

B. stearothermophilus

0

120

240

360

MINUTES FIG. 3. Inactivation of superoxide dismutases by TNBS. The enzymes (0.08-0.55 mg/ml) were treated at 25°C with 0.5 mM TNBS in 50 mM sodium pyrophosphate, 50 mM NaHCO,, 0.1 mM EDTA, pH 9.0, and at the indicated times aliquots were diluted and assayed for enzymatic activity as described in the text. (A) R. stearothermophilus MnSOD, (0) E. coli MnSOD, (m) E. coli FeSOD, (0) human MnSOD, (0) S. cereuisiae MnSOD.

198

CHAN,

BJERRUM,

AND

BORDERS

of 6 acetyl groups is accompanied by approximately 90% inactivation. These data suggest that acetic anhydride shows only a modest degree of selectivity in the inactivation of yeast MnSOD, i.e., full inactivation is obtained only after the modification of up to 7 lysines per subunit. Correlation of the loss of yeast MnSOD activity caused by TNBS with the formation of t-2,4,6-trinitrophenyllysine shows a linear relationship to approximately 25% residual activity, and extrapolation to complete inactivation suggests that 3.5 lysines are modified in this phase (Fig. 5). The actual data show that full inactivation is approached only when approximately 5 lysines are modified. This experiment contains data obtained at two pH values, 8.0 and 9.0, and there appears to be no difference in selectivity at either pH. DISCUSSION 0

2

4

6

8

ACETYL GROUPS/SUBUNIT FIG. 4. Correlation of the inactivation of S. cereuisiae MnSOD by [“Clacetic anhydride with incorporation of acetyl groups. The enzyme (0.54 mg/ml) was treated in separate experiments for l-2 h at 25°C with 33, 66, 99, 198, 264, or 330 pM [‘“Clacetic anhydride in 72 mM sodium pyrophosphate, 72 mM NaHC03, 0.1 mM EDTA, pH 9.0,10% (v/v) acetonitrile, before gel filtration and subsequent analyses as described in the text.

oxal with the loss of specific amino acids as determined by amino acid analysis (data not shown). Modification of the E. coli enzyme for 1.5 h with 20 mM phenylglyoxal gave a preparation that had lost 73% of the control activity on modification of 4.5 of the 6 arginines per subunit. The yeast enzyme, on treatment for 7 h with 40 mM phenylglyoxal, lost only 57% activity on disappearance of approximately 2 of the 3 arginines per subunit. Phenylglyoxal is known to be highly selective for the modification of arginine in proteins (16, 23-25), but it has also been reported to react with peptide and protein w-amino groups under some conditions, presumably by deamination (23). Yeast MnSOD treated under the strenuous reaction conditions of the above experiment, i.e., a high concentration of phenylglyoxal, high pH, and a long reaction time, also lost 2.8 lysines per subunit, a likely major cause of inactivation of this enzyme. The E. coli MnSOD sample (vide supra) was found to have also lost 0.4 lysines per subunit under the vigorous reaction conditions. Yeast MnSOD is extensively inactivated by low levels of acetic anhydride (vide supra). When inactivation was carried out with [14C]acetic anhydride and the loss of activity correlated with the incorporation of acetyl groups, the results shown in Fig. 4 were obtained. The enzyme is 50% inactivated on covalent attachment of approximately 2 acetyl groups per subunit, while incorporation

It has previously been shown that E. coli FeSOD and E. coli MnSOD are extensively inactivated by phenylglyoxal, the former by a pseudo-first-order process and the latter by a biphasic process (11). Phenylglyoxal is known to be highly selective for the modification of arginyl residues in proteins (16, 23-25). Figure 6 shows the aligned amino acid sequences of the nine MnSODs and four FeSODs reported to date. A comparison of the primary sequences of the two E. coli SODS reveals that Arg-189 is the only conserved arginine in the two polypeptides

0

2

4

6

TNP-LYSINEWSUBUNIT FIG. 5. Correlation of the inactivation of S. cereuisiae MnSOD by TNBS with the formation of c-2,4,6-trinitrophenyllysine. The enzyme (0.88-1.1 mg/ml) was treated for O-100 min at 25°C in 50 mM sodium pyrophosphate, 50 mM NaHC03, 0.1 mM EDTA with (0) 0.1 mM TNBS, pH 9.0; (0) 0.2 mM TNBS, pH 9.0; or (m) 0.5 mM TNBS, pH 8.0. The data were analyzed as described in the text.

AN ESSENTIAL

111

121 [31 [41 I51 [61

171 181 191 [lOI [Ill 1121 1131 [II

121

131 [41 [51 161 (71

[81 191 1101 [Ill

Cl21 [I31 [II [21 [31 [41 [51

[61 171

[81 [91 [lOI [Ill [I21 [I31 [II [21 [31 [41 [51 [61 [71 [81 [91

1101 Cl11 [I21 [I31 [II

[II (31 [41 [51 161 171 [El 191 [lOI [Ill [I21 [I31

POSITIVE

CHARGE

IN FeSOD

AND

199

MnSOD

31

41

81 61 71 51 ENRET-GDH-ASTAGAL-GDVTHNGCGHYLHTMFWEHMSPD--GAASRPVEVEVLLRHLEALPQDIQTAVRNNGGGHLNHSLFWRLLTP--GGAKEP-V TAVSK--GDASAVVQ-LQAAIKFNGGGHVNHSIFWKNLKPISEGGGEPPH -LAKGDVTTQV----ALQPALKFNGGGHINHTIFWTNLSP--KGGGEP-K -LAKGDVTTQV----ALQPALKFNGGGHINHSIFWTNLSP--KGGGEP-K -LAKGDVTAOI---ALOPALKFNGGGHINHSIFWTNLSP--NGGGEP-K LLAKEPSPANARKMIAIaQNIKFHGGGFTNHCLFWENLAPESQGGGEPPT KSLEELLSNLEALPESIRTAVRNNGGGHANHSLFWTILSP--NGGGEP-T LPVEELITKLDQLPADKKTVLRNNAGGHANHSLFWKGL---KK--GTTLQ -AFEG--KSLEEIIRSSEGGVFNNAAQVWNHTFYWNCLAP--NAGGEP-T PEFEG--KTLEEIVKSSSGGIFNNAAQVWNHTFYWNCLSP--DGGGQP-T -ELAE--KSLEEIIKTSTGGVFNNAAQVWNHTFYWNCLAP--NAGGEP-T DGQPIEAVIKAIAGDASKAGLFNNAAQAWNHSFYWNSIKP--NGGGAP-T

91

101 111 121 GRSPTASRRTSAPTRTG-GLNSRWRPARP-SGWALLVYDPVAKQLRNVAV GELKKAIDEQFGGFQALKEKLTQAAMGRFGSGWAWLVKD-PF-GKLHVLS GKLGWAIDEDFGSFEALVKKMNAEGAALQGSGWVWLAL-DKEAKKVSVET GELLEAIKRDFGSFEKFKEKLTAMSVGVQGSGWGWLGFN-KEQGRLQIAA GELLEAIKRDFGSFEKFKEKLTAVSVGVQGSGWGWLGFN-KEQGRLQIAA GELLEAIKRDFGSFDKFKEKLTAASVGVQGSGWGWLGFN-KERGHLQIAA GALAKAIDEQFGSLDELIKLTNTKLAGVQGSGWAFIVKNLSNGGKLDVVQ GELADAINKKFGSFTAFKDEFSKAAAGRFGSGWAWLV---VNNGELEITS GDLKAAIERDFGSVDNFKAEFEKAAASRFGSGWAWLVL---KGDKLAVVS GKVAEAIAASFGSFADFKAQFTDAAIKNFGSGWTWLVKN-SD-GKLAIVS GALADAINAAFGSGDKFKEEFTKTSVGTFGSGWAWLVK--ADGS-LALCS GEVAAAIEKAFGSFAEFKAKFTDSAINNFGSSWTWLVKN-ANGS-LAIVN GALADKIAADFGSFENFVTEFKQAAATQFGSGWAWLVL--IDNGTLKIKT

141

21 11 1 QHELPSLPYDYDALEPHISEQVVTWHHDTHHQSYVDGLNSA-EE-T-L-A PFKLPDLGYPYEALEPHIDAKTMEIHHQKHHGAYVTNLNAALEKYPYLHG TVTLPDLSYDFGALEPAISGEIMRLHHQKHHATYVANYNKALEQ---LEKHSLPDLPYDYGALEPHINAQIMQLHHSKHHAAYVNNLNATEEKYH--EKHSLPDLPYDYGALEPHINAQIMQLHHSKHHATYVNNLNVTEEKYH--EKHSLPDLPYDYGALEPHINAQIMQLHHSKHHAAYVNNLNVTEEKYQ--EKVTLPDLKWDFGALEPYISGQINELHYTKHHQTYVNGFNTAVDQFQELSD PFELPALPYPYDALEPHIDKETMNIHHTKHHNTYVTNLNAALEGHPDLQN SYTLPSLPYAYDALEPHFDKOTMEIHHTKHHQTYVNNANAALESLPEFAN SFELPALPYAKDALAPHISAETIEYHYGKHHQTYVTNLN-NLIKGT---AFELPPLPYAHDALQPHISKETLEYHHDKHHNTYVVNLN-NLVPGT---AFELPALPFAMNALEPHISQETLEYHYGKHHNTYVVKLN-GLVEGT---SYELPALPFDYTALAPYITKETKEFHHDKHHAAYVNNYN-NAVKDTDL--

131

151 161 171 181 -DNHD----EGALWGSH--PILALDVWEHSYYYDYGPDRGTFVDAFFE-V TPNQDNPVME----GFT--PIVGIDVWEHAYYLKYQNRRADYLQAIW-NV TANQD-PLV--T-KGASLVPLLGIDVWEHAYYLQYKNVRPDYLNNIW-KV CSNQD-PLQG-T-TGL--IPLLGIDVWEHAYYLQYKNVRPDYLKAIW-NV CSNQD-PLQG-T-TGL--IPLLGIDVWEHAYYLQYKNVRPDYLKAIW-NV CPNQD-PLQG-T-TGL--IPLLGIDVWEHAYYLQYKNVRPDYLKAIW-NV TYNQD------TVTGPL-VPLVAIDAWEHAYYLQYQNKKADYFKAIW-NV TPNQDSPIME----G-K-TPILGLDVWEHAYYLKYQNRRPEYIAAFW-NV TANQDSPLMGEAISGASGFPIMGLDVWEHAYYLKFQNRRPDYIKEFW-NV TSNAGTPL--T-TDAT--PLLTVDVWEHAYYIDYRNARPGYLEHFW-AL TIGAGAPL---T-SGDT--P LLTCDVWEHAYYIDYRNLRPKYVEAFW-NL TSNAGCPITE---EGVT--PLLTVDLWEHAYYIDYRNLRPSYMDGFW-AL TGNADTPI---A-HGQT--PLLTIDVWEHAYYLDYQNRRPDYISTFVEKL 201 IDWDPIAANYDDVVSL LNWD-VAEEFFK-KAMNWKYAGEVYENVLAINWENVTERYTACK-K INWENVSQRYIVCK-K INWENVTERYMACK-K VNWKEASRRFDAGKIVNWDEVAKRYSEAKAK VNWDEAAARF-AAK-K VNWEFVAKNLAA---VNWAFVAEEGKTFKAVNWDFVSKNLAA---ANWDFASANYAAAIA-

191

211

Cl1 [21 t31 [41 [51 [61 [71 [El [91 [lo] [ll] 1121 [13]

Mn Mn Mn Mn Mn Mn Mn Mn Mn Fe Fe Fe Fe

H. halobium T. thermophilus Maize Mouse Rat Human S. cerevisiae B. stearothermophilus E. coli E. coli P. ovalis P. leioynathi A. nidulans R2

FIG. 6. Aligned amino acid sequences of the nine MnSODs and four FeSODs reported to date. The numbering system is similar to that used by Ditlow et al. (19) and Parker and Blake (26), but has been expanded to accommodate recently published sequences. The metal ligands are indicated by an asterisk and the conserved positively charged residue at position 189 is boxed. The references (metal ion utilized and host) are Mn Halobacterium halobium (27), Mn Thermus fhermophilus (28), Mn maize (29), Mn mouse (30), Mn rat (31), Mn human (32-34), Mn S. cereuisiae (19,20), Mn I?. stearothermophilus (35), Mn E. coli (17,18), Fe E. coli (9,36), Fe Pseudomonas ovalis (37), Fe Photobacterium leiognathi (4), and Fe Amqystis nidulans (38).

200

CHAN.

BJERRUM.

and is thus a prime candidate for the critical arginine. Arg-189 is also conserved in the three additional FeSODS and seven of the eight additional MnSODs, with the only exception being the MnSOD from S. cerevisiae, in which it is replaced by a lysine (Fig. 6). Treatment of four SODS which contain arginine at position 189, i.e., E. coli, B. stearothermophilus, and human MnSODs and E. coli FeSOD, with 20 mM phenylglyoxal at pH 9.0 causes nearly complete inactivation (Fig. l), with the time required for 50% inactivation in the range of 0.2-l h. However, yeast MnSOD, with a lysine at, position 189, is inactivated very slowly under these conditions, with a 50% inactivation time approximately one order of magnitude higher. The four Arg-189 SODS examined in this study are partially inactivated by high levels of acetic anhydride (Fig. 2) or by extended treatment with 0.5 mM TNBS (Fig. 3), in line with a proposal that catalysis by MnSODS and FeSODs involves electrostatic facilitation by lysine residues (39). However, the lo-50% residual activity that remains after extensive modification suggests that a lysine is not absolutely essential for catalysis by these enzymes. In contrast, yeast MnSOD is much more sensitive to inactivation by acetic anhydride and TNBS, with >98% inactivation at acetic anhydride concentrations >0.5 mM or after 2 h modification with 0.5 mM TNBS. The stoichiometry of inactivation of yeast MnSOD with [l”C]acetic anhydride and with TNBS was determined, but in neither instance was a high degree of selectivity obtained (Figs. 4 and 5). This lack of selectivity, along with the inability to protect the critical lysine(s) with substrate or a competitive inhibitor, makes it very difficult or impossible to confirm the critical lysine unequivocally by conventional protein chemistry. However, the different behaviors of yeast MnSOD, on one hand, and the Arg-189 SODS, on the other, toward arginine and lysine reagents strongly suggests that modification of residue 189 is the probable cause of the complete loss of activity. An examination of the recent high resolution (1.8 A) structural data on T. thermophilus MnSOD suggests that Arg-189 is uniquely situated to play an important role in catalysis. This enzyme is a dimer of dimers, and the major dimer interface is nearly identical to that of the three dimeric SODS that have had their crystal structures determined to date (1, 2, 8). Arg-189 in one subunit of T. thermophilus MnSOD is approximately equidistant from the two metals in the major dimer, 11.4 A from the Mn in the same polypeptide (subunit A) and 12.1 A from the Mn in the other subunit (subunit B). While the positively charged side chains of the other five arginines in the T. thermophilus enzyme fit into the almost universal “surface-exposed” mode that is found in most other proteins, Arg-189 forms five strong hydrogen

AND

BORDERS

bonds and appears to play a more critical structural role. Four of the 0 hydrogen bonds, of distance 2.61, 2.74, 3.22, and 3.40 A, are to the backbone carbonyl oxygens of residues l-25, 128, 130, and 176, respectively, while the fifth (3.09 A) is to the OH of the fully conserved Tyr-181. The highly ordered Arg-189 in T. thermophilus MnSOD lies near the major dimer interface. It has been proposed (7) that superoxide anion approaches the Mn ion through a channel formed by the two dimer subunits from a direction that is trans to the Asp-175 ligand. One wall of this channel to the metal ion in subunit A is built around the side chain of Arg-189 from subunit B. If, as seems likely, this highly hydrogen-bonded arginine is positively charged, it would provide a diffuse electrostatic attractive force to guide superoxide to the solventfilled cone that leads to the catalytic metal center. It has been noted (11) that the rates of inactivation of E. coli FeSOD and E. coli MnSOD by phenylglyoxal are very slow compared to the rates of modification of essential arginines in most proteins, and the relative inaccessibility of the guanidinium group of Arg-189 is consistent with this finding. Moreover, it is likely that covalent modification of Arg-189 would block the channel and/or dramatically alter the major dimer interface. The data reported here suggest that residue 189 is a prime target for the preparation of site-specific mutants of MnSODs and FeSODs. This approach would avoid the multiple modifications observed with each of the chemical modifications carried out in our study and could be of invaluable use in the demonstration of a role for a positive charge at position 189. Nature has, in essence, carried out the first mutagenesis experiment by exchange of a lysine for an arginine at this position in S. cereuisiae MnSOD. However, given the multiple hydrogen bonding interactions of Arg-189 in T. thermophilus MnSOD, it appears that other compensatory interactions in Lys-189 SODS may be needed to generate a viable enzyme. ACKNOWLEDGMENTS We thank Drs. Eric Eccleston and James B. Howard for the amino acid analyses, Dr. Irwin Fridovich for the human MnSOD, Dr. Wayne F. Beyer, Jr. for helpful discussions, Drs. Martha A. Ludwig and William C. Stallings for the tape containing the 1.8-A X-ray crystallographic data on T. thermophilus MnSOD and for helpful discussions, Dr. Clague Hodgson for help with the sequence alignment program, and Dr. Virginia B. P&t for help with the use of the Evans and Sutherland 390 molecular graphics terminal.

REFERENCES 1. Stallings, W. C., Powers, T. B., Pattridge, K. A., Fee, J. A., and Ludwig, M. L. (1983) Proc. Natl. Acad. Sci. USA 80,3884-3888. 2. Ringe, D., Petsko, G. A., Yamakura, F., Suzuki, K., and Ohmori, D. (1983) Proc. Natl. Acad. Sci. USA 80,3879-3883. 3. Stallings, W., Bull, C., Pattridge, K. A., Powers, T. B., Fee, J. A.,

AN ESSENTIAL

4. 5. 6.

7.

POSITIVE

CHARGE

Ludwig, M. L., Ringe, D., and Petsko, G. A. (1984) in Oxygen Radicals in Chemistry and Biology (Bors, W., Saran, M., and Tate, D., Eds.) pp. 779-792, de Gruyter, Berlin. Barra, I).. Schinini, M. E., Bannister, W. H., Bannister, J. V., and Bossa, F. (1987) J. Riol. Chem. 262,1001~1009. Bannister, J. V., Bannister, W. H., and Rotilio, G. (1987) CKC Crit. Reu. Biochem. 22,111-180. Ponzi, D. R., Petsko, G. A., Yamakura, F., Ohmori, D., Suzuki, K., Stallings, W. C., Pattridge, K. A., Powers, T. B., Fee, J. A., and Ludwig, M. L. (1983) in Oxy Radicals and Their Scavenger Systems (Cohen, G., and Greenwald, R. A., Eds.) Vol. 1, pp 340-343, Elsevier, New York. Stallings, W. C., Pattridge, K. A., Strong, R. K., and Ludwig, M. L. (1985) J. Biol. Chcm. 260, 16,424-16,43X.

8. Parker, M. W., and Blake, C. C. F. (1988) J. Mol. Biol. 661.

199, 649-

9. Carlioz, A., Ludwig, M. L., Stallings, W. C., Fee, J. A., Steinman, H. M., and Touati, D. (1988) J. Biol. Chem. 263, 1555-1562. 10. Beyer, W. F., -Jr., and Fridovich 1257. 11. Borders, C. I,., Jr., Horton, Biochrm.

Biophys.

268,

I. (1987) Biochemistry

26,

1251-

P. J., and Beyer, W. F., ?Jr. (1989) Arch.

74-80.

12. Bjerrum.

M. J. (1987) Aiochim.

13. Marklund,

S., and Marklund,

Biophys.

Acta

915,225-237.

(:. (1974) Eur. J. Biochem.

47,469-

474. 14. Borders, C. I~., .Jr., and ,Johansen, .I. 7’. (1980) Cnrlsberg mun. 45.185194. 15. Blech, D. M., and Borders, ph,ys. 224, ,579-586. 16. Borders,

C. L., Jr. (1983) Arch.

C. L., Jr., and Riordan,

RPS. Com-

Biochem.

J. F. (1975) Biochemistry

Bio-

14,

4699-3704. 17. Steinman, 18. Takeda. 4589. 19. Ditlow,

H. M. (1978) J. Riol. Chem. 253,8708M720. Y.. and Avila,

H. (1986) Nuckic

Acids

Rrs.

14, 4577-

(:., Johansen, J. T., Martin, B. M., and Svendsen, I. (198’2) C’nrlsbrrg Rw (‘ommun. 47, 81-91.

IN FeSOD

AND

201

MnSOD

20. Marres, C. A. M., Van Loon, A. P. G. M., Oudshoorn, P., Van Steeg, H., Grivell, L. A., and Slater, E. C. (1985) Eur. J. Biochem. 147,153-161. 21. Ditlow, C., and Johansen, 47,71-79. 22. 23. 24. 25.

,J. T. (1982) Curlsberg

Res. Commun.

Goldfarh, A. R. (1966) Biochemistry 5, 2570-2574. Takahashi, K. (1968) J. Biol. Chem. 243,6171-6179. Riordan, J. F. (1979) Mol. Cell. Biochem. 26, 71-92. Bjerrum, P. J., Wieth, J. O., and Borders, C. I~., Jr. (1983) J. Gen. Ph.ysiol.

81, 453-484.

26. Parker, M. W., and Blake, C. C. F. (1988) FEBS Lett. 229, X77382. 27. Salin, M. L., Duke, M. V., Oesterhelt, I)., and Ma, D.-P. (1988) Gene 70,153-159. 28. Sato, S., Nakada, Y., and Nakazawa-Tomizawa, K. (1987) Bio~ chim. Biophys. Acfa 912, 178-184. 29. White, J. A., and Scandalios, J. G. (1988) Biochim. Biophys. Acta

951,61-70. 30. Hallewell, R. A., Mullenhach, G. T., Stempien, G. I. (1986) Nucleic Acids Res. 14,9539.

M. M., and Bell,

31. Ho, Y.-S., and Crapo, J. D. (1987) Nucleic Acids Kes. 15, 10,070. 32. Barra, D., Schinia, M. E., Simmaco, M., Bannister, .J. V., Bannister, W. H., Rotilio, G., and Bossa, F. (1984) J. Biol. Chem. 259, 12,595512,601. 33. Beck, Y., Oren, R., Amit, B., Levanon, A., Gorecki, M., and Hartman, J. R. (1987) Nucleic Acids Res. 15,9076. 34. Ho, Y.-S., and Crapo, ,J. D. (1988) FE’BS Lett. 229, 256-260. 35. Brock, C. J., and Walker, J. E. (1980) Biochemistry 19, 2873G 2882. 36. Schinina, M. E., Maffey, L., Barra, D., Bossa, F., Puget, K., and Michelson, A. M. (1987) FEES Lett. 221,87-90. 37. Isobe, ‘I’., Fang, Y.-I., Muno, D., Okuyama, T., Ohmori, I)., and Yamakura, F. (1987) FEBS L&t. 223, 92-96. 38. Laudenbach, D. IX., Trick, C. G., and Straus, N. A. (1989) Mol.

Gcn. Gcnrt. 216,455-461. 39. Benovic, Biochem.

.J., Tillman,

T., Cudd, A., and Fridovich,

Rioph,ys.

1, 329KD2.

22

I. (1983) Arch.