Characterization of an Atypical Superoxide Dismutase from ...

3 downloads 47 Views 1MB Size Report
Jan 11, 1999 - thetaiotaomicron. J. Bacteriol. 166:528–532. 39. Pesci, E. C., D. L. Cottle, and C. L. Pickett. 1994. Genetic, enzymatic, and pathogenic studies of ...
JOURNAL OF BACTERIOLOGY, Aug. 1999, p. 4509–4516 0021-9193/99/$04.00⫹0 Copyright © 1999, American Society for Microbiology. All Rights Reserved.

Vol. 181, No. 15

Characterization of an Atypical Superoxide Dismutase from Sinorhizobium meliloti RENATA SANTOS,1 STEPHANE BOCQUET,2† ALAIN PUPPO,3

AND

` LE TOUATI1* DANIE

1

Laboratoire de Ge´ne´tique Mole´culaire des Re´ponses Adaptatives and Laboratoire d’Embryologie Mole´culaire,2 Institut Jacques Monod, CNRS-Universite´s Paris 6 et 7, 75251 Paris Cedex 05, and Laboratoire de Biologie Ve´ge´tale et Microbiologie, CNRS ERS 590, Universite´ de Nice Sophia-Antipolis, 06108 Nice Cedex 02,3 France Received 11 January 1999/Accepted 24 May 1999

Sinorhizobium meliloti Rm5000 is an aerobic bacterium that can live free in the soil or in symbiosis with the roots of leguminous plants. A single detectable superoxide dismutase (SOD) was found in free-living growth conditions. The corresponding gene was isolated from a genomic library by using a sod fragment amplified by PCR from degenerate primers as a probe. The sodA gene was located in the chromosome. It is transcribed monocistronically and encodes a 200-amino-acid protein with a theoretical Mr of 22,430 and pI of 5.8. S. meliloti SOD complemented a deficient E. coli mutant, restoring aerobic growth of a sodA sodB recA strain, when the gene was expressed from the synthetic tac promoter but not from its own promoter. Amino acid sequence alignment showed great similarity with Fe-containing SODs (FeSODs), but the enzyme was not inactivated by H2O2. The native enzyme was purified and found to be a dimeric protein, with a specific activity of 4,000 U/mg. Despite its Fe-type sequence, atomic absorption spectroscopy showed manganese to be the cofactor (0.75 mol of manganese and 0.24 mol of iron per mol of monomer). The apoenzyme was prepared from crude extracts of S. meliloti. Activity was restored by dialysis against either MnCl2 or Fe(NH4)2(SO4)2, demonstrating the cambialistic nature of the S. meliloti SOD. The recovered activity with manganese was sevenfold higher than with iron. Both reconstituted enzymes were resistant to H2O2. Sequence comparison with 70 FeSODs and MnSODs indicates that S. meliloti SOD contains several atypical residues at specific sites that might account for the activation by manganese and resistance to H2O2 of this unusual Fe-type SOD. Sinorhizobium meliloti is an aerobic gram-negative bacterium that can live free in the soil or establish a symbiotic association with the roots of leguminous plants, leading to the formation of nodules. In these specialized structures, the bacteria differentiate to bacteroids that can fix atmospheric nitrogen, converting it to ammonia due to the activity of the nitrogenase enzyme complex. The nitrogenase reductase is rapidly and irreversibly inactivated by oxygen and free radicals (43). Despite the low level of molecular oxygen in the nodules, there have been several reports that free radicals are produced in great quantities (5, 32, 36). The SOD could be important for protecting the nitrogen fixation process, as suggested by Puppo and Rigaud (41). Early reports on the SOD content of several Rhizobium species are confusing. Stowers and Elkan (52) reported the presence of a single FeSOD in free-living bacteria in several species. In contrast, Becana and Salin (6) found one MnSOD in free-living bacteria and two Mn-containing isoenzymes in the nodule bacteroids. Dimitrijevic et al. (12) found that the SOD activity of free-living Rhizobium phaseoli is due to the presence of two isoenzymes, one Mn-type and another Fe-type inducible, and that the bacteroids contained only the Mn type. Only the sodA gene encoding an MnSOD from Bradyrhizobium sp. (Parasponia) strain ANU289 has been cloned and sequenced to date (55), and little is known about the defenses against oxidative stress in the symbiotic interaction between rhizobia and leguminous plants. This report describes the cloning and sequencing of the S. meliloti sodA gene and characterization of the encoded cambialistic SOD.

The superoxide dismutases (SODs; EC 1.15.1.1) are metalloenzymes that catalyze the dismutation of superoxide (O2.⫺) to hydrogen peroxide (H2O2) and molecular oxygen (O2). They have been found in nearly all organisms examined to date and play a major role in the defense against oxidative stress (reviewed in references 15 and 56). There are three general classes of SODs in bacteria, which differ in their metal cofactors. The manganese-containing (MnSOD) and iron-containing (FeSOD) enzymes are cytoplasmic, while the copper-plus-zinc (CuZnSOD) enzyme is periplasmic. In addition, a new class of nickel-containing SODs has been recently discovered in Streptomyces griseus and S. coelicolor (25, 26, 63). The MnSODs and FeSODs have very similar sequences and structures and are evolutionarily unrelated to CuZnSODs (9, 56). Usually FeSODs and MnSODs require specific metal for activity (8) and can be distinguished on the basis of amino acid sequence (37) and sensitivity to H2O2 (7, 9). However, these criteria can be misleading (53, 64), and the purified protein must be analyzed to correctly determine the metal at the active site (54). A small group of Mn/FeSODs, termed cambialistic, are active with either manganese or iron incorporated into the same active site. They have been found in the anaerobic (aerotolerant) species Propionibacterium shermanii (35), Bacteroides fragilis (20), Bacteroides thetaiotaomicron (38), Streptococcus mutans (31), and Porphyromonas gingivalis (1) and in the aerobic methylotrophic Methylomonas strain J (61). The aerobic hyperthermophilic Aquifex pyrophilus (30) SOD is, presumably, also cambialistic. * Corresponding author. Mailing address: Laboratoire de Ge´ne´tique Mole´culaire des Re´ponses Adaptatives, Institut Jacques Monod, 2 place Jussieu, 75251 Paris Cedex 05, France. Phone: 33 1 44274719. Fax: 33 1 44277667. E-mail: [email protected]. † Present address: Institut de Ge´ne´tique Humaine, 34396 Montpellier Cedex 5, France.

MATERIALS AND METHODS Bacterial strains and plasmids. The bacterial strains and plasmids used are listed in Table 1. To obtain pRS51, the sodA coding sequence was amplified by PCR using two primers, 5⬘TTTTGAATTCCCCGACGAAATCCATGCCA3⬘

4509

4510

J. BACTERIOL.

SANTOS ET AL. TABLE 1. Bacterial strains and plasmids used in this study

Strain or plasmid

Strains S. meliloti Rm5000 A. tumefaciens GMI9023 At125 At128 E. coli DH5␣ QC1799 QC2375 QC2461 Plasmids pUC18 pRS41 pRS41.1 pJF119EH pRS51 pDT1-19 pKOK5 pJQ200sk pRS58

Relevant characteristics

Reference or source

SU47 rif-5

13

C58 cured of pAtC58 and pTiC58 GMI9023 pRmeSU47b GMI9023 pRMeSU47a

44 14 14

F⫺ supE44 ⌬lacU169(␾180dlacZ⌬M15) hsdR17(rK⫺ mK⫹) recA1 endA1 gyrA96 thi-1 relA1 F⫺ ⌬(argF-lac)U169 rpsL ⌬sodA3 ⌽(sodB-kan)⌬2 QC1799 ⌬recA306 srl::Tn10 MG1655 ⌬lacIZ

Laboratory stock

Cloning vector, ColE1 origin, Apr pUC18, 2.7-kb S. meliloti library plasmid with sodA gene, Apr pRS41 with 1.2 kb EcoRI-EcoRI deletion, sodA⫹ Apr Expression vector with tac promoter, Apr pJF119EH with sodA structural gene on a 760-bp EcoRI-HindIII fragment under tac promoter control, Apr Tcr pBR322 with E. coli sodA structural gene under tac promoter control pSUP202 derivative, source of lacZ-Km cassette, Apr Kmr Cloning vector, p15A origin, sacB⫹ Gmr pJQ200sk with sodA-lacZ-Km transcriptional fusion, Gmr Kmr

Laboratory stock This study This study 16 This study

and 5⬘TTTTAAGCTTCCGATTTCCGCTGGTAGAAGC3⬘, which carry 5⬘ EcoRI and HindIII restriction sites, respectively. The amplified fragment was digested with EcoRI and HindIII and inserted into pJF119EH under the control of the tac promoter (16). Two DNA fragments were amplified from pRS41.1 by PCR using the vector polylinker primers and two sodA internal primers, 5⬘GT TTTGGATCCATTGTGGCATCTCCTCTTG3⬘ and 5⬘GAAAAGGATCCGCT CGTCCACGGCGCAAC3⬘, which carry a 5⬘ BamHI site. These fragments were ligated at the BamHI site (creating a deletion of amino acids 2 to 154) and inserted into the ApaI-XbaI sites of the vector pJQ200sk (42). Plasmid pRS58 was obtained by inserting the lacZ-Km (kanamycin resistance) cassette from pKOK5 (28) into the BamHI site of pRS56, creating a transcriptional fusion. All constructs were verified by sequencing. Growth conditions. E. coli was grown aerobically at 37°C in Luria-Bertani medium (LB; yeast extract, 5 g/liter; tryptone, 10 g/liter; NaCl, 10 g/liter). Anaerobic cultures were grown in a Forma Scientific anaerobic chamber in LB medium supplemented with 1% glucose. All media and materials were equilibrated in the anaerobic chamber for 24 h before use. S. meliloti was grown at 30°C in LB containing 2.5 mM CaCl2 and 2.5 mM MgSO4 (18) under aerobic conditions. Ampicillin (500 ␮g/ml in liquid media and 50 ␮g/ml in plates), rifampin (20 ␮g/ml), kanamycin (20 ␮g/ml), gentamicin (20 ␮g/ml), and isopropyl-␤-D-thiogalactopyranoside (IPTG) were added to the medium when needed. Reagent-grade chemicals were purchased from Sigma. Cloning strategy. (i) S. meliloti genomic library construction. Genomic DNA was extracted from stationary-phase S. meliloti Rm5000 by the method of Pitcher et al. (40). The DNA was partially digested with Sau3A (200 ng of DNA to 0.4 U of Sau3A), and 100-␮g DNA fragments were fractionated in a 5 to 30% sucrose gradient (Beckman SW41 rotor, 15°C, 27,500 rpm, 20 h) as described by Sambrook et al. (45). Fragments of 2.3 to 4.3 kb were collected, dialyzed against 1 mM EDTA–10 mM Tris-Cl (pH 8.0), and inserted into the BamHI site of pUC18. Approximately 11,000 transformants were obtained with strain DH5␣. (ii) Nested-PCR amplification of a sodA fragment. The PCR mix contained 100 pmol of each primer, 300 ng of genomic DNA, 0.2 mM deoxynucleoside triphosphate (Pharmacia), 1.25 mM MgCl2, 1⫻ Taq polymerase buffer, and 0.4 U Taq DNA polymerase (Goldstar). The reaction parameters were 4 cycles (2 min at 94°C, 2 min at 40°C, 2 min at 72°C) followed by 30 identical cycles with 45°C as the annealing temperature and a final elongation step of 10 min at 72°C. Degenerate primers were designed according to the conserved amino acid regions 2, 3, and 4 of SOD proteins defined by Heinzen et al. (21): 5⬘CCAYCA YGACAAGCAYC3⬘ (KHH3), 5⬘CCANCCNGANCCRAA3⬘ (FGS1), 5⬘TTYG GNTCNGGNTGGGCNTGG3⬘ (WAW1), 5⬘TARTANGCRTGYTCCCANAC RTC3⬘ (DVWEH), and 5⬘RTAGTASGCARGTYCCC3⬘ (WEH6). Two primer combinations, KHH3-DVWEH and KHH3-WEH6, allowed amplification of a fragment of approximately 400 bp from Rm5000 genomic DNA. Using nested primers (region 3) in both orientations, the expected size fragments were obtained by the pairs WAW1-DVWEH/WEH6 and KHH3-FGS1. The sequence of

57 57 This laboratory

57 28 42 This study

the amplified 422-bp KHH3-DVWEH fragment was determined and found to be very similar to those of known Mn/FeSODs. This fragment was radiolabeled and used as a probe to screen the genomic library by colony hybridization (45). A clone carrying a plasmid with a 2.7-kb insert (pRS41) was isolated, and subcloning located the sodA gene in a 1.5-kb EcoRI-HindIII fragment (pRS41.1). General techniques. The molecular cloning techniques and gel electrophoresis were essentially as described by Sambrook et al. (45). Small-scale preparation of bacterial DNA was by the procedure of Chen and Kuo (11). Pure plasmid DNA was prepared by using a Qiagen kit. DNA fragments were isolated from agarose gels with a QIAEX kit (Qiagen). Southern blotting, hybridization, and detection methods were previously described (48), and the DNA was labeled with [␣-32P] dATP (ICN, Orsay, France), using the Megaprime DNA labeling system (Amersham). Plasmid proteins were labeled in maxicells as previously described (46). Plasmid double-stranded DNA was sequenced by the dye terminator method on an ABI model 377 DNA sequencer. The sequence of 1,196 bp from pRS41.1 was obtained by primer walking on both strands. RNA isolation and analysis. RNA was isolated from a S. meliloti culture at an optical density at 600 nm (OD600) of 1.6 by the method of Babst et al. (2). RNA (10 ␮g) was separated on a 1.0% gel using the RNA Transcripts 9488-363 nt (USB Amersham) as size markers. Northern blotting and hybridization (at 42°C in 50% [vol/vol] formamide) were essentially as described by Sambrook et al. (45). An internal sodA fragment amplified by PCR with primers 5⬘CGGTCTT TCCGATC3⬘ and 5⬘TGCGCCGTGGAC3⬘ was used as the probe. Primer extension was carried as previously described (58), using primer 5⬘GTCATAGG GAAGGTTCGGCA3⬘, complementary to nt 255 to 274. The extended primer was loaded onto a sequencing gel next to sequencing reactions performed with the same primer by the dideoxy-chain termination method with a Sequenase kit version 2.0 (U.S. Biochemical Corp.) with [␣-35S]dATP (ICN). Preparation of cell lysates and SOD activity assay. Saturated cultures of E. coli and S. meliloti were harvested by centrifugation, washed, and disrupted by sonication. The doubling time being much longer for S. meliloti than for E. coli, saturation (OD of 4 to 5) was reached for E. coli and S. meliloti overnight and after 2 days, respectively. The total protein concentration was measured by using the bicinchoninic acid reagent (Pierce Chemical Company) and bovine serum albumin as the standard. Specific activity was determined by using the standard xanthine oxidase/cytochrome c assay at pH 7.8 (34). SOD activity in nondenaturing 10% polyacrylamide gels was visualized by nitroblue tetrazolium negative staining (4). The gels were stained in presence of 5 mM H2O2 or 10 mM potassium cyanide (KCN) for activity inhibition studies. Protein purification and metal cofactor determination. The SodA from S. meliloti was purified essentially as described by Slykhouse and Fee (49). The soluble protein fraction from 15 g of cells was precipitated at 90% ammonium sulfate. The resulting precipitate was suspended in 5 mM potassium acetate (pH 5.5), dialyzed, and run on a CM50 (Pharmacia) column equilibrated with 5 mM potassium acetate. The column was eluted stepwise with 10, 20, 30, 40, and 50

ATYPICAL SOD FROM S. MELILOTI

VOL. 181, 1999

FIG. 1. Detection of SOD in E. coli and S. meliloti on nondenaturing polyacrylamide gels stained for SOD activity. (A) No inhibitors; (B) with 5 mM H2O2. Lanes: 1, E. coli DH5␣; 2, S. meliloti Rm5000; 3, E. coli sodA sodB strain QC1799; 4, QC1799/pRS41.1 with S. meliloti sodA gene. Lanes contain crude extracts of saturated cultures (35 ␮g of total protein). The E. coli MnSOD, hybrid SOD (HySOD), and FeSOD are indicated.

mM potassium acetate. The fractions containing SOD activity (40 and 50 mM) were pooled, dialyzed against 5 mM potassium phosphate (pH 7.4) for 2 days, loaded onto a DEAE column (Pharmacia) equilibrated with the same buffer, and eluted stepwise with 10 to 60 mM potassium phosphate. The SOD was eluted at 50 and 60 mM potassium phosphate (pH 7.4). These fractions were pooled and concentrated with a 10-kDa-cutoff Centricon (Amicon). The mass of the native SOD was estimated by gel filtration–high-pressure liquid chromatography (HPLC) on a Bio-Sil SEC 250 column (Bio-Rad), using the Bio-Rad BioLogic HR system. The molecular mass of the monomer protein was determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The metal content of the purified protein was determined with a GBC 902 atomic absorption spectrophotometer. Metal removal and reconstitution experiments from crude extracts were done by the procedure of Kirby et al. (27) except that 5 M guanidinium chloride was used for metal removal. Nucleotide sequence accession number. The S. meliloti sodA sequence has been registered in GenBank and assigned accession no. AF110770.

RESULTS SOD activity in S. meliloti and sodA gene cloning. Previous results based on the inhibition of SOD activity in two S. meliloti strains by H2O2 and KCN suggested that strain 102F28 had an FeSOD (52) and strain 102F51 an MnSOD (6). The SOD in S. meliloti Rm5000 in free-living growth conditions (Fig. 1A) gave a single activity band on nondenaturing PAGE. The enzyme activity was not inhibited by H2O2 (Fig. 1B) and KCN (data not shown), suggesting an MnSOD. The sodA gene was cloned (pRS41) by using a PCR-based strategy with degenerate primers. The E. coli sodA sodB strain (QC1799) was transformed with pRS41.1 and grown in LB medium to the stationary phase. The protein extract exhibited a single band of SOD activity that migrated like the S. meliloti

4511

SOD in a polyacrylamide gel (Fig. 1). This finding indicated that the cloned sodA gene corresponds to the enzyme detected in S. meliloti Rm5000 free-living cultures. Sequence of sodA from S. meliloti. A total of 1,196 bp from pRS41.1 was sequenced on both strands. There was an open reading frame of 600 nucleotides coding for a 200-residue protein with a theoretical Mr of 22,430 and a pI of 5.8. A ribosome binding site similar to the Shine-Dalgarno sequence of E. coli was located 8 bp upstream of the ATG initiation codon. A 10-bp inverted repeat sequence followed by a stretch of T’s was found 25 bp downstream of the stop codon and could function as a rho-independent RNA polymerase terminator. The G⫹C content of the sodA gene was 59%, similar to that of other S. meliloti genes. Northern blot analysis (Fig. 2A) detected a single mRNA of approximately 700 nt that hybridized with a sodA probe, indicating that the sodA gene is transcribed monocistronically. The transcription start site was identified by primer extension to be an adenine 44 bp upstream of the ATG start codon (Fig. 2B). The transcript size indicates that the mRNA terminates in the inverted repeat observed downstream of the stop codon. A putative E. coli ␴70-like promoter was found upstream the transcription start site and matched three of six bases (boldface) with the ⫺35 (TTGACA) consensus sequence and four of six bases with the ⫺10 (TATAAT) consensus sequence (Fig. 2B). Location of the sodA gene in the chromosome. The genome of strain Rm5000 has a chromosome of 3.54 Mb plus two symbiotic megaplasmids, pSym-a (pRmSU47a) of 1.34 Mb and pSym-b (pRmSU47b) of 1.7 Mb (22). These plasmids were mobilized into the Agrobacterium tumefaciens GMI9023, creating the hybrid strains At125 (carrying pSym-b) and At128 (carrying pSym-a) (14). The genomic DNAs of Rm5000, GMI9023, and hybrid strains were probed with a sodA fragment by Southern blotting (data not shown). No specific hybridization signal was obtained with the hybrid A. tumefaciens strains, indicating that the gene lies in the S. meliloti chromosome, unlike the case for the symbiotic plasmid-specific genes. Expression of sodA gene complements SOD-deficient E. coli mutant. E. coli sodA sodB recA strain (QC2375) cannot survive in aerobic conditions, due to unrepaired DNA oxidative damage (57). Transformation with plasmids pRS41 and pRS41.1 did not rescue the aerobic growth (Fig. 3A). This absence of

FIG. 2. S. meliloti sodA transcript analysis. (A) Northern blot analysis of S. meliloti mRNA with an internal sodA fragment as a probe; (B) determination of the transcription start site of sodA. Lanes: P, primer extension product; G, A, T, and C, sequence obtained with the same primer. The ⫺10 promoter region and the transcription start point (ⴱ) are in boldface. rbs, ribosome binding site.

4512

SANTOS ET AL.

J. BACTERIOL. TABLE 2. Anaerobic SOD activity and corresponding survival after a shift from anaerobiosis to aerobiosis of the E. coli sodA sodB recA straina Plasmid

pRS41.1c pDT1-19 pRS51c

IPTG (␮M)

% Aerobic survivalb

SOD activity (U/mg of protein)

0 10 10 500

0.005 ⫾ 0.002 89.4 ⫾ 3.51 7.8 ⫾ 4.15 77.7 ⫾ 2.0

Not detected 1.0 ⫾ 0.42 1.4 ⫾ 0.33 4.8 ⫾ 0.65

a Cultures were grown overnight anaerobically in the presence of IPTG. SOD activity was determined by the xanthine oxidase/cytochrome c assay, and cells were plated under anaerobic and aerobic conditions with IPTG. Colonies were counted after overnight incubation. Values are means of three independent experiments, and standard errors are indicated. b Ratio of the number of colonies under aerobic and anaerobic conditions. c Small colonies.

FIG. 3. (A) Aerobic survival of E. coli QC2375 (sodA sodB recA) transformed with plasmids pUC18, pRS41.1, pJF119EH, and pRS51. Saturated anaerobic cultures were plated under anaerobic conditions (black) and aerobic conditions without (white) and with (shaded) 2 mM IPTG. CFU were counted after incubation overnight. Values are means of at least three independent experiments, and bars represent standard deviations. (B) Expression of S. meliloti sodA-lacZ in E. coli. Strain QC2461 was transformed with pRS58 and assayed for ␤-galactosidase activity in aerobiosis (circles) and anaerobiosis (squares).

complementation of SOD deficiency was surprising, because the SOD was seen in protein extracts from aerobically grown QC1799 (sodA sodB)/pRS41.1 (Fig. 1). To test whether the absence of complementation was due to a defect of expression, the sodA gene was expressed under the control of the synthetic IPTG-inducible tac promoter (pRS51). Production of the corresponding protein (23 kDa) was verified in maxicells (data not shown). Expression of SOD from pRS51 restored aerobic survival of QC2375 (Fig. 3A), indicating that the S. meliloti enzyme complements the E. coli SOD deficiency. Two observations suggested explanations for the failure of pRS41.1 to complement QC2375. Measurements of sodA expression in E. coli from a plasmid containing a transcriptional sodA-lacZ fusion (plasmid pRS58) showed that sodA from S. meliloti was weakly expressed during exponential growth, but protein production increased upon entry into stationary phase (Fig. 3B).

The sodA-lacZ fusion was not expressed in anaerobiosis (Fig. 3B), and no detectable SOD activity was found in anaerobic crude extracts from QC2375/pRS41.1 (Table 2). Thus, a shift from anaerobiosis to aerobiosis results in lethal damage to QC2375 before enough SOD has been produced from pRS41.1. We further questioned whether the S. meliloti SodA was as efficient as the E. coli MnSOD in restoring aerobic viability of the sodA sodB recA strain. We therefore determined the minimum amount of SOD necessary to rescue QC2375 by using constructs in which both SODs were under the control of the inducible tac promoter, using various amounts of IPTG inducer (Table 2). An E. coli MnSOD activity of 1.0 U/mg of protein was sufficient to rescue 85.7% of the bacteria compared, to only 7.8% survival with an SmSodA activity of 1.4 U/mg. A fivefold-higher activity of SmSodA (4.8 U/mg) was necessary to obtain similar rescue (77.7%). The sodA gene encodes a cambialistic SOD. The deduced S. meliloti sodA amino acid sequence was used to search protein databases on the National Center for Biotechnology Information BLAST server. It was very similar to prokaryotic and eukaryotic SODs, being most like FeSODs, with 48% identity to Rhodobacter capsulatus and Chlamydomonas reinhardtii FeSODs. The majority of residues used to distinguish between the Mn- and Fe-type enzymes matched an FeSOD (24, 33, 37). It also had unusual features, with residues not commonly found in SODs such as tyrosine at position 74 (74Tyr), 78His, 82Trp, and 164Ser (Fig. 4). However, the predicted threedimensional structure of the S. meliloti SodA (Swiss-Model; Glaxo Wellcome Experimental Research, Geneva, Switzerland) showed a typical secondary structure of Mn/FeSODs with two domains, the N terminal with two ␣-helices and the C terminal with two ␣-helices, followed by three ␤-strands and two ␣-helices (data not shown) (9, 24, 37). The exact nature of the metal cofactor was found by purifying the S. meliloti SodA from free-living bacteria (Fig. 5A). The molecular mass of the native protein was estimated to be 43 kDa by gel filtration (Fig. 5B) and approximately 23 kDa for the subunit visualized in SDS-PAGE, showing that the active protein is a dimer. The purified enzyme (⬎90% purity) had a specific activity of 4,000 U/mg of protein and contained 0.75 mol of manganese and 0.24 mol of iron per mol monomer, as determined by atomic absorption spectroscopy, indicating that the SOD produced was a MnSOD. Since the amino acid sequence of this protein was closer to that of Fe-type enzymes, we investigated whether it could be active with iron. The enzyme in crude extracts (32.1 U/mg of protein) was depleted of metal (Fig. 6). The resulting inactive apoenzyme was dialyzed against Mn or Fe. Both metals restored an active SOD, al-

ATYPICAL SOD FROM S. MELILOTI

VOL. 181, 1999

4513

FIG. 4. Comparison of residues from S. meliloti SodA, cambialistic SODs, and three atypical FeSODs with corresponding residues that distinguish between FeSODs and MnSODs (S. meliloti SodA amino acid sequence numbering). Typical FeSOD and MnSOD consensus residues were obtained from comparison of 19 and 42 sequences, respectively. ⌬, ligands to metal cofactor; ⴱ, ⫹, and ¥, no typical residue of alternative metal was found. Signs above the typical FeSOD sequence: ⴱ, zero to one mismatch; ⫹, two to five mismatches. Signs below typical MnSOD sequence: ⴱ, zero to one mismatch; ⫹, two to six mismatches. X, variable; boldface, typical FeSOD residues; bold italic, typical MnSOD residues; underline, residues found rarely among the 70 sequences. Amino acid sequences of SODs from Bacteroides fragilis (P53638), Porphyromonas gingivalis (P19665), Aquifex pyrophilus (AE000743), Tetrahymena pyriformis (P19666), Methanobacterium thermoautotrophicum (Q60036), Mycobacterium tuberculosis (P17670), Propionibacterium shermanii (P80293), Methylomonas strain J (P23744), and Streptococcus mutans (P09738) are from SWISSPROT and GenBank databases.

though the activity recovered with manganese was sevenfold higher than that obtained with iron. SOD activities of Mnreconstituted (12.2 U/mg of protein) and Fe-reconstituted (1.8 U/mg) enzymes were determined by the xanthine oxidase/cytochrome c assay described previously. This result demonstrated the cambialistic nature of the S. meliloti SodA. The two reconstituted enzymes were not inhibited by 5 mM H2O2 (Fig. 6B). DISCUSSION The SOD from S. meliloti Rm5000, the only cytoplasmic SOD detectable in free-living bacteria, belongs to the family of

cambialistic Mn/FeSODs. Cambialistic SODs generally function efficiently with either manganese or iron at their active sites. The SODs from Propionibacterium shermanii (35) and P. gingivalis (1) have similar activities with both metals. In contrast, Methylomonas strain J (61) is more active with manganese, both native and reconstituted FeSOD showing less than 10% of the activity of the MnSOD. Similarly, in S. meliloti, the Mn-reconstituted SOD from apo-SOD is more active than the Fe-reconstituted SOD (15% activity with Fe compared to Mn). Comparison of sequences of FeSODs and MnSODs and structural data led several authors to identify amino acid res-

FIG. 5. (A) Purification of S. meliloti SOD as visualized by Coomassie brilliant blue staining of an SDS-polyacrylamide gel. Lanes: 1, Rainbow molecular weight marker (Amersham); 2, ammonium sulfate 90% precipitate; 3, flowthrough from a CM50 column; 4, elution from a CM50 column; 5, sixfold-concentrated eluate from DEAE column. (B) Molecular mass of native SodA determinated by HPLC-gel filtration. The standards used were bovine gamma globulin (158,000 Da), chicken ovalbumin (44,000 Da), horse myoglobin (17,000 Da), and vitamin B12 (1,350 Da).

4514

SANTOS ET AL.

FIG. 6. Activity of reconstituted SodA from S. meliloti Rm5000 with manganese and iron. (A) No inhibitors; (B) with 5 mM H2O2. Lanes: 1, E. coli DH5␣; 2, crude extract; 3, apoenzyme; 4, Mn-reconstituted SOD; 5, Fe-reconstituted SOD.

idues in the metal ligand environment that might account for metal specificity (24, 33, 37). We aligned (by MaxHom [47] at the Predict Protein server) the S. meliloti SodA amino acid sequence with 70 complete published sequences of SODs including 42 MnSODs, 19 FeSODs, 6 cambialistic SODs, and 3 atypical FeSODs. Alignment revealed several unusual features. While the S. meliloti native enzyme, at least in free-living growth conditions, is essentially a manganese-containing SOD, sequence comparison shows higher similarity to FeSODs (Fig. 4). Among residues that highly discriminate between the typical FeSODs and MnSODs, seven are of Fe type and only two are of Mn type in S. meliloti SodA, while others are atypical. This clearly classifies the S. meliloti sequence among those of atypical FeSODs. In typical FeSODs, the solvent interacts with the 72Gln residue (S. meliloti SodA numbering), and with the 144Gln in typical MnSODs. The glutamine residues occupy very similar positions in the structure (29). The presence of a 72Gln and a 144Gly in the S. meliloti SodA suggests that 72Gln interacts with the solvent, as in FeSODs. Further, several highly conserved residues are not found in S. meliloti SodA. The 85Pro conserved in 58 of 70 sequences, 104Phe in 64 of 70, 106Ser in 59 of 70, and 164Ala in 68 of 70 are replaced by Lys, Leu, Gly, and Ser, respectively. A 78His residue nearby 76His metal ligand is unique among all sequences analyzed. The high activity of S. meliloti SOD when it incorporates manganese is puzzling. This is a unique example of an Fe-type protein, according to its specific residues, that is more active with manganese. The Mycobacterium tuberculosis FeSOD, in contrast, with an Mn-type sequence and a 72Gly, binds iron in an Mntype way with a 144His acting as the metal solvent ligand (24). The many unusual residues in the environment of the active site in S. meliloti SodA may contribute to a subtle change that favors activation with manganese. The Fe-reconstituted SOD from S. meliloti is not sensitive to 5 mM H2O2. The inactivation of E. coli FeSOD by H2O2 depends on the presence of iron at the active site and is correlated with oxidation of tryptophan residues (7). The tryptophan at position 74, replaces by a valine in the H2O2-resistant FeSOD from Methanobacterium thermoautotrophicum, has been proposed to be responsible for inactivation (37, 54). However, some findings were incompatible with this hypothesis. For example, a valine in this position does not make the FeSODs from Tetrahymena pyriformis (3) and Mycobacterium tuberculosis (10) resistant to H2O2, and the FeSOD from Campylobacter jejuni, which has a tyrosine instead of tryptophan, is sensitive (39). The cambialistic SODs all lack a 74Trp (Fig. 4). However, the reconstituted SODs from B. fragilis (20) and P. gingivalis (1) are sensitive with iron at the active site and resistant with manganese. The Fe-substituted form of cambialistic SOD from

J. BACTERIOL.

Propionibacterium shermanii is only partially inactivated (60%) when exposed to 5 mM H2O2, and a mutant in which the tryptophan has been substituted for valine is completely inactivated (17). Conversely, mutation of 74Trp to Val in the sensitive FeSOD from Plasmodium falciparum did not reverse H2O2 sensitivity, although it was shown that iron was more stable in the mutant during inactivation (19). Recently, Yamakura et al. (62) demonstrated that oxidation of the conserved 161Trp residue was responsible for inactivation of the Fe form of P. gingivalis cambialistic SOD. Altogether, these results suggest that the difference in H2O2 sensitivity is caused by the fine-tuning of amino acid environment determining the redox activity of Fe center with regard to H2O2, rather than by the position of tryptophan H2O2-sensitive residues (62). The unusual active-site environment in S. meliloti SOD may explain the resistance of the Fe-reconstituted enzyme. It has been demonstrated that the E. coli MnSOD and FeSOD are not physiologically equivalent; the MnSOD associates with DNA (51) and is more efficient in preventing DNA damage, while the FeSOD is more effective in protecting cytoplasmic superoxide-sensitive enzymes (23). The reason why the S. meliloti SodA does not rescue the E. coli sodA sodB recA strain as efficiently as the E. coli MnSOD is unclear. Nonetheless, the atypical nature of the S. meliloti SodA might account for the reduced protection of DNA oxidative damage. It was shown that incorporation of metal in cambialistic SODs depends on its availability in the medium (31, 33, 35) and is oxygen dependent, iron being preferentially used under anaerobiosis and manganese under aerobiosis (1). S. meliloti encounters two completely different environments during its life cycle, the soil and the nodules. The free oxygen concentration in the nodules is low (50), and iron is abundant (5). Also, manganese is not available in high concentration in the cytoplasm of eukaryotic cells, since it is actively pumped into organelles like the lysosomes and Golgi apparatus (14a). Moreover, acidic conditions within the nodule should favor higher activity of an iron-substituted SOD (59, 60). The transition from the aerobic cycle to the microaerobic nodule environment, together with increased iron and presumably reduced manganese availability, might have encouraged the evolution of a cambialistic SOD that is active with manganese when free-living and active with iron in nodules. ACKNOWLEDGMENTS This work was supported by the EC Human Capital and Mobility Program and ACC SV no. 6 from the MENRT (France). R.S. acknowledges grant Praxis XXI BPD/9917-96 from Fundac¸˜ao para a Cieˆncia e a Tecnologia (Portugal). We are grateful to D. Lavergne (Paris, France) for atomic absorption spectrometry and P. Rodrigues (Paris, France) for help with the HPLC experiments. We thank J. Batut (Toulouse, France) and D. He´rouart (Nice, France) for sending strains and J. M. Camadro (Paris) for helpful discussions. REFERENCES 1. Amano, A., S. Shizukuishi, H. Tamagawa, K. Iwakura, S. Tsunasawa, and A. Tsunemitsu. 1990. Characterization of superoxide dismutase purified from either anaerobically maintained or aerated Bacteroides gingivalis. J. Bacteriol. 172:1457–1463. 2. Babst, M., H. Hennecke, and H.-M. Fischer. 1996. Two different mechanisms are involved in the heat-shock regulation of chaperonin gene expression in Bradyrhizobium japonicum. Mol. Microbiol. 19:827–839. 3. Barra, D., M. E. Schinina, F. Bossa, K. Puget, P. Durosay, A. Guissani, and A. M. Michelson. 1990. A tetrameric iron superoxide dismutase from the eucaryote Tetrahymena pyriformis. J. Biol. Chem. 265:17680–17687. 4. Beauchamp, C., and I. Fridovich. 1971. Superoxide dismutase: improved assays and an assay applicable to acrylamide ges. Anal. Biochem. 44:276–287. 5. Becana, M., and R. V. Klucas. 1992. Transition metals in legume root nodules: iron-dependent free radical production increases during senes-

VOL. 181, 1999 cence. Proc. Natl. Acad. Sci. USA 89:8958–8962. 6. Becana, M., and M. L. Salin. 1989. Superoxide dismutases in nodules of leguminous plants. Can. J. Bot. 67:415–421. 7. Beyer, J., W. F., and I. Fridovich. 1987. Effect of hydrogen peroxide on the iron-containing superoxide dismutase of Escherichia coli. Biochemistry 26: 1251–1257. 8. Beyer, J., W. F., and I. Fridovich. 1991. In vivo competition between iron and manganese for occupancy of the active site region of the manganese-superoxide dismutase of Escherichia coli. J. Biol. Chem. 266:303–308. 9. Beyer, W., J. Imlay, and I. Fridovich. 1991. Superoxide dismutases. Progress Nucleic Acid Res. Mol. Biol. 40:221–253. 10. Bunting, K., J. B. Cooper, M. O. Badasso, I. J. Tickle, M. Newton, S. P. Wood, Y. Zhang, and D. Young. 1998. Engineering a change in metal-ion specificity of the iron-dependent superoxide dismutase from Mycobacterium tuberculosis X-ray structure analysis of site-directed mutants. Eur. J. Biochem. 251:795–803. 11. Chen, W. P., and T. T. Kuo. 1993. A simple and rapid method for the preparation of gram-negative bacterial genomic DNA. Nucleic Acids Res. 21:2260. 12. Dimitrijevic, L., A. Puppo, and J. Rigaud. 1984. Superoxide dismutase activities in Rhizobium phaseoli bacteria and bacteroids. Arch. Microbiol. 139: 174–178. 13. Finan, T. M., E. Hartwieg, K. LeMieux, K. Bergman, G. C. Walker, and E. R. Signer. 1984. General transduction in Rhizobium meliloti. J. Bacteriol. 159: 120–124. 14. Finan, T. M., B. Kunkel, G. F. De Vos, and E. R. Signer. 1986. Second symbiotic megaplasmid in Rhizobium meliloti carrying exopolysaccharide and thiamine synthesis genes. J. Bacteriol. 167:66–72. 14a.Frausta da Silva, J. J. R., and R. J. P. Williams. 1991. The biological chemistry of the elements: the inorganic chemistry of life. Clarendon Press, Oxford, England. 15. Fridovich, I. 1995. Superoxide radical and superoxide dismutases. Annu. Rev. Biochem. 64:97–112. 16. Fu ¨rste, J. P., W. Pansegrau, R. Frank, H. Blo¨cker, P. Scholz, M. Bagdassarian, and E. Lanka. 1986. Molecular cloning of the plasmid RP4 primase region in a multi-host-range tacP expression vector. Gene 48:119–131. 17. Gabbianelli, R., A. Battistoni, C. Capo, F. Polticelli, G. Rotilio, B. Meier, and A. Desideri. 1997. Effect of Val 733Trp mutation on the reaction of “cambialistic” superoxide dismutase from Propionibacterium shermanii with hydrogen peroxide. Arch. Biochem. Biophys. 345:156–159. 18. Glazebrook, J., and G. C. Walker. 1991. Genetic techniques in Rhizobium meliloti. Methods Enzymol. 204:398–418. 19. Gratepanche, S., S. Me´nage, D. Touati, C. Odberg-Ferragut, M. Fontecave, A. Masset, C. Slomianny, D. Camus, and D. Dive. 1999. Unpublished data. 20. Gregory, E. M., and C. H. Dapper. 1983. Isolation of iron-containing superoxide dismutase from Bacteroides fragilis: reconstitution as a Mn-containing enzyme. Arch. Biochem. Biophys. 220:293–300. 21. Heinzen, R. A., M. E. Frazier, and L. P. Mallavia. 1992. Coxiella burnetii superoxide dismutase gene: cloning, sequencing, and expression in Escherichia coli. Infect. Immun. 60:3814–3823. 22. Honeycutt, R. J., M. McClelland, and B. W. S. Sobral. 1993. Physical map of the genome of Rhizobium meliloti 1021. J. Bacteriol. 175:6945–6952. 23. Hopkin, K. A., M. A. Papazian, and H. M. Steinman. 1992. Functional differences between manganese and iron superoxide dismutases in Escherichia coli K-12. J. Biol. Chem. 267:24253–24258. 24. Jackson, S. M. J., and J. B. Cooper. 1998. An analysis of structural similarity in the iron and manganese superoxide dismutases based on known structures and sequences. Biometals 11:159–173. 25. Kim, E.-J., H.-J. Chung, B. Suh, Y. C. Hah, and J.-H. Roe. 1998. Transcriptional and post-transcriptional regulation by nickel of sodN gene encoding nickel-containing superoxide dismutase from Streptomyces coelicolor Mu ¨ller. Mol. Microbiol. 27:187–195. 26. Kim, E.-J., H.-P. Kim, Y. C. Hah, and J.-H. Roe. 1996. Differential expression of superoxide dismutases containing Ni and Fe/Zn in Streptomyces coelicolor. Eur. J. Biochem. 241:178–185. 27. Kirby, T., J. Blum, I. Kahane, and I. Fridovich. 1980. Distinguishing between Mn-containing and Fe-containing superoxide dismutase in crude extracts of cells. Arch. Biochem. Biophys. 201:551–555. 28. Kokotek, W., and W. Lotz. 1989. Construction of a lacZ-kanamycin-resistance cassette, useful for site directed mutagenesis and as promoter probe. Gene 84:467–471. 29. Lah, M. S., M. M. Dixon, K. A. Pattridge, W. C. Stallings, J. A. Fee, and M. L. Ludwig. 1995. Structure-function in Escherichia coli iron superoxide dismutase: comparisons with the manganese enzyme from Thermus thermophilus. Biochemistry 34:1646–1660. 30. Lim, J.-H., Y. G. Yu, I.-G. Choi, J.-R. Ryu, B.-Y. Ahn, S.-H. Kim, and Y. S. Han. 1997. Cloning and expression of superoxide dismutase from Aquifex pyrophilus, a hyperthermophilic bacterium. FEBS Lett. 406:142–146. 31. Martin, M. E., B. R. Byers, M. O. J. Olson, M. L. Salin, J. E. L. Arceneaux, and C. Tolbert. 1986. A Streptococcus mutans superoxide dismutase that is active with either manganese or iron as a cofactor. J. Biol. Chem. 261:9361– 9367.

ATYPICAL SOD FROM S. MELILOTI

4515

32. Mathieu, C., S. Moreau, P. Frendo, A. Puppo, and M. J. Davies. 1998. Direct detection of radicals in intact soybean nodules: presence of nitric oxideleghemoglobin complexes. Free Radical Biol. Med. 24:1242–1249. 33. Matsumoto, T., K. Terauchi, T. Isobe, K. Matsuoka, and F. Yamakura. 1991. Iron- and manganese-containing superoxide dismutases from Methylomonas J: identity of the protein moiety and amino acid sequence. Biochemistry 30:3210–3216. 34. McCord, J. M., and I. Fridovich. 1969. Superoxide dismutase: an enzymic function for erythrocuprein (hemocuprein). J. Biol. Chem. 244:6049–6055. 35. Meier, B., D. Barra, F. Bossa, L. Calabrese, and G. Rotilio. 1982. Synthesis of either Fe- or Mn-superoxide dismutase with an apparently identical moiety by an anaerobic bacterium dependent on the metal supplied. J. Biol. Chem. 257:13977–13980. 36. Moreau, S., M. J. Davies, C. Mathieu, D. He´rouart, and A. Puppo. 1996. Leghemoglobin-derived radicals: evidence for multiple protein-derived radicals and the initiation of peribacteroid membrane damage. J. Biol. Chem. 271:32557–32562. 37. Parker, M. W., and C. C. F. Blake. 1988. Iron- and manganese-containing superoxide dismutases can be distinguished by analysis of their primary structures. FEBS Lett. 229:377–382. 38. Pennington, C. D., and E. M. Gregory. 1986. Isolation and reconstitution of iron- and manganese-containing superoxide dismutases from Bacteroides thetaiotaomicron. J. Bacteriol. 166:528–532. 39. Pesci, E. C., D. L. Cottle, and C. L. Pickett. 1994. Genetic, enzymatic, and pathogenic studies of the iron superoxide dismutase of Campylobacter jejuni. Infect. Immun. 62:2687–2694. 40. Pitcher, D. G., N. A. Saunders, and R. J. Owen. 1989. Rapid extraction of bacterial genomic DNA with guanidium thiocyanate. Lett. Appl. Microbiol. 8:369–384. 41. Puppo, A., and J. Rigaud. 1986. Superoxide dismutase: an essential role in the protection of the nitrogen fixation process? FEBS Lett. 201:187–189. 42. Quandt, J., and M. F. Hynes. 1993. Versatile suicide vectors which allow direct selection for gene replacement in Gram-negative bacteria. Gene 127: 15–21. 43. Robson, R. L., and J. R. Postgate. 1980. Oxygen and hydrogen in biological nitrogen fixation. Annu. Rev. Microbiol. 34:183–207. 44. Rosenberg, C., and T. Huguet. 1984. The pAtC58 plasmid of Agrobacterium tumefaciens is not essential for tumor induction. Mol. Gen. Genet. 196:533– 536. 45. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory, Cold Spring, N.Y. 46. Sancar, A., A. M. Hack, and W. D. Rupp. 1979. Simple method for identification of plasmid-coded proteins. J. Bacteriol. 137:692–693. 47. Sander, C., and R. Schneider. 1991. Database of homology-derived protein structures and the structural meaning of sequence alignment. Proteins 9:56– 68. 48. Santos, R., G. Vieira, M. A. Santos, and H. Paveia. 1996. Characterization of temperate bacteriophages of Leuconostoc oenos and evidence for two prophage attachment sites in the genome of starter strain PSU-1. J. Appl. Bacteriol. 81:383–392. 49. Slykhouse, T. O., and J. A. Fee. 1976. Physical and chemical studies on bacterial superoxide dismutases: purification and some anion binding properties of the iron-containing protein of Escherichia coli B. J. Biol. Chem. 251:5472–5477. 50. Soupe`ne, E., M. Foussard, P. Boistard, G. Truchet, and J. Batut. 1995. Oxygen as a key developmental regulator of Rhizobium meliloti N2-fixation gene expression within the alfalfa root nodule. Proc. Natl. Acad. Sci. USA 92:3759–3763. 51. Steinman, H. M., L. Weinstein, and M. Brenowitz. 1994. The manganese superoxide dismutase of Escherichia coli K-12 associates with DNA. J. Biol. Chem. 269:28629–28634. 52. Stowers, M. D., and G. H. Elkan. 1981. An inducible iron-containing superoxide dismutase in Rhizobium japonicum. Can. J. Microbiol. 27:1202–1208. 53. Takao, M., A. Oikawa, and A. Yasui. 1990. Characterization of a superoxide dismutase gene from the archaebacterium Methanobacterium thermoautotrophicum. Arch. Biochem. Biophys. 283:210–216. 54. Takao, M., A. Yasui, and A. Oikawa. 1991. Unique characteristics of superoxide dismutase of a strictly anaerobic archaebacterium Methanobacterium thermoautotrophicum. J. Biol. Chem. 266:14151–14154. 55. Thygesen, P. W., J. Whelan, M. K. Morell, and D. A. Day. 1995. The isolation and characterisation of a gene encoding superoxide dismutase from Bradyrhizobium sp. (Parasponia) strain ANU289. Biochem. Mol. Biol. Int. 37:401– 412. 56. Touati, D. 1997. Superoxide dismutases in bacteria and pathogen protists, p. 447–493. In J. G. Scandalios (ed.), Oxidative stress and the molecular biology of antioxidant defenses. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 57. Touati, D., M. Jacques, B. Tardat, L. Bouchard, and S. Despied. 1995. Lethal oxidative damage and mutagenesis are generated by iron in ⌬fur mutants of Escherichia coli: protective role of superoxide dismutase. J. Bacteriol. 177: 2305–2314.

4516

SANTOS ET AL.

58. Uzan, M., R. Favre, and E. Brody. 1988. A nuclease that cuts specifically in the ribosome binding site of some T4 mRNAs. Proc. Natl. Acad. Sci. USA 85:8895–8899. 59. Vance, C. K., and A.-F. Miller. 1998. Spectroscopic comparisons of the pH dependencies of Fe-substituted (Mn)superoxide dismutase and Fe-superoxide dismutase. Biochemistry 37:5518–5527. 60. Yamakura, F., K. Kobayashi, H. Ue, and M. Konno. 1995. The pH-dependent changes of the enzymic activity and spectroscopic properties of ironsubstituted manganese superoxide dismutase: a study on the metal-specific activity of Mn-containing superoxide dismutase. Eur. J. Biochem. 227:700– 706. 61. Yamakura, F., T. Matsumoto, and K. Terauchi. 1991. Isolation of Mn-SOD

J. BACTERIOL. and low active Fe-SOD from Methylomonas J; consisting of identical proteins. Free Radical Res. Commun. 12–13:329–334. 62. Yamakura, F., R. L. Rardin, G. A. Petsko, D. Ringe, B. H. Hiraoka, K. Nakayama, T. Fujimura, H. Taka, and K. Murayama. 1998. Inactivation and destruction of conserved Trp159 of Fe-superoxide dismutase from Porphyromonas gingivalis by hydrogen peroxide. Eur. J. Biochem. 253:49–56. 63. Youn, H.-D., E.-J. Kim, J.-H. Roe, Y. C. Hah, and S.-O. Kang. 1996. A novel nickel-containing superoxide dismutase from Streptomyces spp. Biochem. J. 318:889–896. 64. Zhang, Y., R. Lathigra, T. Garbe, D. Catty, and D. Young. 1991. Genetic analysis of superoxide dismutase, the 23 kilodalton antigen of Mycobacterium tuberculosis. Mol. Microbiol. 5:381–391.