Peroxisomal Copper,Zinc Superoxide Dismutase - NCBI

3 downloads 4 Views 2MB Size Report
Jan 17, 1995 - Pablo Bueno, Javier Varela, Cuillermo Ciménez-Gallego, and Luis A. de1 Rio* ... SOD) located in the matrix of peroxisomes from watermelon. (Cifrullus vulgaris ...... hydrogen peroxide the enzyme had a half-life of nearly 60.

Plant Physiol. (1995) 108: 1 1 51-1 160

Peroxisomal Copper,Zinc Superoxide Dismutase’ Characterization of the lsoenzyme from Watermelon Cotyledons Pablo Bueno, Javier Varela, Cuillermo Ciménez-Gallego, and Luis A. de1 Rio* Departamento de Bioquímica, Biologia Celular y Molecular de Plantas, Estación Experimental de1 Zaidín, Consejo Superior de lnvestigaciones Científicas, Granada, Spain (P.B., L.A.d.R.); and Laboratorio de Química de Proteínas, Centro de lnvestigaciones Biológicas, Consejo Superior de lnvestigaciones Científicas, Madrid, Spain (J.V., G.G.-G.)

ents (Tolbert, 1981; Huang et al., 1983). These organelles have an essentially oxidative type of metabolism, and in recent years it has become increasingly clear that peroxisomes carry out essential functions in almost a11 eukaryotic cells (Fahimi and Sies, 1987; Van den Bosch et al., 1992; Mannaerts and Van Veldhoven, 1993). By using immunocytochemical and density-gradient centrifugation methods, a Mn-containing SOD was localized for the first time in peroxisomes from pea leaves (de1Rio et al., 1983; Sandalio et al., 1987). The presence of SOD was also demonstrated in peroxisomes from watermelon (Citrullus vulgaris Schrad.) cotyledons, but in this case two isozymes were detected, a Cu,Zn- and a Mn-containing SOD (Sandalio and de1 Rio, 1987). The peroxisomal CuZnSOD represented about 18%of the total SOD activity of the plant tissue and was the major SOD isozyme in these organelles (Sandalio and de1 Rio, 1987). The study of the intraorganellar distribution of SOD in plant peroxisomes showed that the CuZn-SOD was present in the matrix of these organelles, whereas the Mn-SOD isozyme was apparently bound to the externa1 side of the peroxisomal membrane (Sandalio and de1 Rio, 1988). The occurrence of SOD isozymes in peroxisomes from cucumber, cotton, and sunflower cotyledons also has recently been reported (Corpas et al., 1994). In recent years, the presence of SOD in plant peroxisomes has been extended to human and animal cells. Human fibroblasts and hepatoma cells were found to contain CuZn-SOD activity (Keller et al., 1991; Crapo et al., 19921, and the occurrence of CuZn-SOD was also demonstrated in rat liver peroxisomes (Dhaunsi et al., 1992; Wanders and Denis, 1992). The generation of 0;- radicals, the substrate of SOD, in matrices and membranes of plant peroxisomes has been reported and the generating systems of superoxide have been partially characterized (Sandalio et al., 1988; de1 Rio et al., 1989; López-Huertas et al., 1994). These findings together with other experimental evidence led to the proposal of the existence of new functions for peroxisomes related to activated oxygen species that could

The biochemical and immunochemical characterization of a superoxide dismutase (SOD, EC l .15.1.1) from peroxisomal origin has been carried out. The enzyme is a Cu,Zn-containing SOD (CuZnSOD) located in the matrix of peroxisomes from watermelon (Cifrullus vulgaris Schrad.) cotyledons (L.M. Sandalio and L.A. de1 Rio 119881 Plant Physiol 88: 1215-1218). The amino acid composition of the enzyme was determined. Analysis by reversed-phase high-performance liquid chromatography of the peroxisomal CuZnSOD incubated with 6 M guanidine-HCI indicated that this enzyme contained a noncovalently bound chromophore group that was responsible for the absorbance peak of the native enzyme at 260

nm. The amino acid sequence of the peroxisomal CuZn-SOD was determined by Edman degradation. Comparison of its sequence with those reported for other plant SODs revealed homologies of about 70% with cytosolic CuZn-SODs and of 90% with chloroplastic CuZn-SODs. The peroxisomal SOD has a high thermal stability and resistance t o inactivation by hydrogen peroxide. A polyclonal antibody was raised against peroxisomal CuZn-SOD, and by western blotting the antibody cross-reacted with plant CuZn-SODs but did not recognize either plant Mn-SOD or bacterial Fe-SOD. The antiSOD-immunoglobulin C showed a weak cross-reaction with bovine erythrocytes and liver CuZn-SODs, and also with cell-free extracts from trout liver. The possible function of this CuZn-SOD in the oxidative metabolism of peroxisomes is discussed.

SODs (EC are a family of metalloenzymes that catalyze the disproportionation of superoxide (O;-) radicais, and they play an important role in protecting cells against the toxic effects of superoxide radicals produced in different cellular loci (Fridovich, 1986; Halliwell and Gutteridge, 1989). In recent years the presence of different types of SOD has been demonstrated in peroxisomes from severa1 plant species. Peroxisomes are subcellular respiratory organelles that contain catalase and hydrogen peroxide-producing flavin oxidases as basic enzymatic constituThis work was supported by grants PB87-0404-01 and PB920492-01 from the Dirección General de Investigación Científica y Tetnica and the Junta de Andalucía (Research Group No. 3315),

Spain. * Corresponding author; e-mail;fax 34-58-

Abbreviations: CuZn-SOD, copper, zinc-containing superoxide dismutase; SOD, superoxide dismutase.

129600. 1151


Bueno et al.

have important effects on cellular metabolism (de1 Rio et al., 1990, 1992). Recently, the peroxisomal CuZn-SOD from watermelon cotyledons was purified to homogeneity (Bueno and de1 Rio, 1992). Although there is abundant information on SODs from cytosolic, mitochondrial, and chloroplastic origin (Steinman, 1982; Parker et al., 1984; Bannister et al., 1987, 1991; Halliwell and Gutteridge, 1989; Asada, 1992), until now no SOD has been characterized from any type of peroxisome of a eukaryotic organism. The characterization of the peroxisomal CuZn-SOD is important to study its function in these oxidative cell organelles and also to establish comparative biochemical and evolutionary studies with SODs present in other cellular compartments from different organisms, such as mitochondria, cytosol, chloroplasts, and lysosomes. To our knowledge, ours is the first report of the biochemical and immunochemical characterization of a SOD from peroxisomal origin, and the molecular properties of this enzyme are compared with those of other SODs distributed in different cellular compartments. MATERIALS A N D METHODS Plant Material

Seeds of watermelon (Citrullus vulgaris Schrad. cv Sugar Baby) were obtained from Fitó (Barcelona, Spain). Seeds were germinated at 30°C under dark conditions, as described (Bueno and de1 Rio, 1992). From the seedlings obtained, the cotyledons (first leaves) were used for enzyme purification. Purification of Peroxisomal CuZn-SOD

The enzyme was purified from 10-d-old watermelon cotyledons by the procedure previously described (Bueno and de1 Rio, 1992). A yield of 8 p g enzyme per g cotyledon was obtained, and the peroxisomal CuZn-SOD was homogenous by native PAGE and SDS-PAGE (Bueno and de1 Rio, 1992). Amino Acid Composition

The amino acid composition was determined by reversed-phase HPLC in a Waters apparatus (Millipore) using the Pico-Tag method (Cohen et al., 1989). Phenylthiocarbamyl-derivatized amino acids were detected at 254 nm and quantitated using an amino acid standard solution from Pierce. Denaturation and Reversed-Phase HPLC

The enzyme (21-230 pg) was incubated with 6 M guanidine-HC1 for 5 min and loaded onto a Vydac (Separations Group, Hesperia, CA) C-4 reversed-phase column equilibrated with 0.1% TFA, and eluted with a O to 100% linear gradient of acetonitrile:H,O (2:l) in 0.1% TFA over 30 min. Pure CuZn-SODs from bovine liver and erythrocytes and from horseradish (Sigma) were also subjected to the same treatment.

Plant Physiol. Vol. 108, 1995

Amino Acid Sequence

The interna1 sequence was determined by proteolytic digestion of intact protein with Staphylococcus aureus V-8 protease and pancreatic trypsin (Boehinger-Mannheim),as described (Gimenez-Gallego et al., 1985, 1986). Peptides were separated by HPLC on a Vydac C-18 reversed-phase column and were subjected to sequence analysis by automated Edman degradation in an Applied Biosystems pulse-liquid sequenator (model 477A). Phenylthiohydantoin-derivatized amino acids were detected using an online phenylthiohydantoin analyzer from Applied Biosystems (model120A). lnhibition by Hydrogen Peroxide and Cyanide

Potassium cyanide and hydrogen peroxide were purchased from Merck (Darmstadt, Germany). Potassium cyanide was prepared in 50 mM potassium phosphate buffer (pH 7.8), 0.1 mM EDTA, and was added to the SOD assay mixture. Hydrogen peroxide was preincubated at 25°C with the protein for 10 to 60 min, and aliquots were taken for assay. SOD activity was determined by its ability to inhibit the superoxide-mediated reduction of ferricytochrome c by the superoxide radicals produced by xanthine/xanthine oxidase (McCord and Fridovich, 1969). SOD control reactions were carried out in the presence of each concentration of inhibitor, and reactions were initiated by adding an appropriate dilution of xanthine oxidase so that the generation of 0;- radicals was constant. Thermal Stability

The stability of peroxisomal CuZn-SOD was studied at 25,50,70,80, and 100°C. Enzyme solutions of 30 p g mL-' in 50 mM potassium phosphate buffer (pH 7.8) were incubated at those temperatures, and 10-pL aliquots were withdrawn at different times for the SOD activity determination. The ranges of incubation times were 1 h to 4 d at 25"C, 10 to 120 min at 50"C, 5 to 60 min at 80"C, and 15 min at 100°C. Preparation of a Polyclonal Antibody against Peroxisomal CuZn-SOD

Two New Zealand rabbits were immunized by four subcutaneous injections of 100 p g pure CuZn-SOD/rabbit, at 14-d intervals. The first injection contained 100 p g of enzyme in PBS and complete Freunds adjuvant (l:l),and the other three booster injections contained 100 pg of enzyme in PBS and incomplete Freund's adjuvant (1:l). The rabbits were bled by cardiac puncture 1 week after the last injection. Blood was allowed to clot and then was centrifuged at 7509 for 20 min at 4°C. A total volume of serum of 42 mL was obtained. Serum immunoglobulins were isolated by ammonium sulfate precipitation (33%saturation) at pH 7.8, three times, and were dissolved in 19 mL of borate-buffered saline and dialyzed overnight at 4°C against 0.0175 M sodium phosphate buffer (pH 6.3) (buffer A). The antiserum solution was purified by anion-exchange chromatography in a Mono Q column HR 10/10 equilibrated with buffer A, using a fast protein liquid chromatography sys-


Characterization of Peroxisomal Superoxide Dismutase

Table 1. Amino acid composition of peroxisomal CuZn-SOD Amino Acid LY5

His Arg ASP Thr

Ser Glu

Pro GlY Ala

Half-cystineb Val

Met Ile Leu TYr

Phe Trp'

Residues/mol of Enzymea

7 5 13 25 28 20 35 19 43 34 5 28 2 12 21 1 10 1

Determined as Based on a molecular mass of 33,000 D. cysteic acid after performic acid oxidation. Determined fluorimetricallv (Paiot, 1976).

pigiana) leaves, and broad bean (Vicia faba L.) leaves, and soluble fraction from soybean (Glycinemax L.) nodules and trout (Salmo trutta L.) liver were used. Except for the soluble fraction of soybean nodules, which was prepared as described (Becana et al., 1989), a11 the other cell-free extracts were prepared by homogenization in 50 mM potassium phosphate buffer (pH 7.8), containing 1 mM DTT, 8% (w/v) polyvinylpolypyrrolidone, 1 mM PMSF, and 0.1% (v/v) Triton X-100 (3:1, w/v). Homogenates were filtered through six layers of nylon cloth and centrifuged at 34,0008 for 30 min. The supernatants obtained were used for the assays. Other Assays

Protein content was determined either by the method of Lowry et al. (1951) or Murphy and Kies (1960), and BSA (fraction V) was used as a standard. RESULTS


tem (Pharmacia). The column was washed with 160 mL of buffer A and then was eluted with 80 mL of a linear gradient from O to 1 M NaCl in buffer A. Fractions of 4 mL were collected at a flow rate of 3 mL min-'. The first peak eluted contained the IgG fraction, as demonstrated by SDSPAGE and immunoblotting. The IgG fraction was dialyzed against borate-buffered saline, concentrated to a volume of 6 mL (approximately 17 mg protein mL-I), and kept frozen until further use. The serum and purified IgG were tested for monospecificity by SDS-PAGE and immunoblotting, and for antibody titer (ELISA) according to Hermoso et al. (1987).

The amino acid composition of peroxisomal CuZn-SOD purified from watermelon cotyledons is shown in Table I. The most abundant amino acids were Gly, Ala, Glu, Thr, Val, and Asp, and there were low contents of His, Cys, and Met. Only one residue of Tyr and Trp was detected in the enzyme. The UV absorption spectrum of purified peroxisomal CuZn-SOD is characterized by a weak absorption maximum at 260 nm and a low absorption at 280 nm (Fig. 1).In preliminary experiments for the determination of the amino acid sequence, the peroxisomal CuZn-SOD was incubated with 6 M guanidine-HC1 and then analyzed by reversed-phase HPLC. Under these conditions, a polypeptide with the molecular mass and amino acid composition of peroxisomal CuZn-SOD and a nonproteinaceous chromophore were separated (Fig. 2A). The chromophore ac-


SDS-PACE and lmmunoblotting

For SDS-PAGE, proteins were heated at 100°C for 5 min in the presence of 2% SDS and 5% mercaptoethanol. Electrophoresis was carried out on 12%acrylamide-SDS gels, as described by Laemmli (1970), using a Bio-Rad Mini-Protean I1 slab cell. For immunoblotting, the proteins from the SDS-PAGE were electrophoretically blotted onto Immobilon nitrocellulose sheets from Millipore using a Bio-Rad Transblot apparatus, and were detected with anti-peroxisomal CuZn-SOD as the primary antibody (1/1000 dilution), goat anti-rabbit IgG with horseradish peroxidase (Bio-Rad) as the second antibody, and 4-chloro-1-naphthol as the stain (Bio-Rad Immunoblot instruction manual). Preparation of Cell-Free Extracts

For antigenic cross-reactivity assays of peroxisomal CuZn-SOD antibody, cell-free extracts from watermelon (C. vulgaris Schrad.) cotyledons, pea (Pisum sativum L.) leaves, spinach (Spinacia oleracea L.) leaves, sorghum (Sorghum vulgare L.) leaves, clover (Trifolium repens L.) leaves, lettuce (Lactuca sativa L.) leaves, erythrina (Evythrina poep-

a, U


0.25 O










Wavelength, nm Figure 1. U V absorption spectrum of native peroxisomal CuZn-SOD. A n enzyme concentration of 69 Fg mL-' in 50 mM potassium phosphate buffer, pH 7.8, was used.

Bueno et al.


counted for most of the absorbance peak at 260 nm, previously detected in the native enzyme, whereas the polypeptide did not present any significant absorption at this wavelength (Fig. 2B). In the light of these results, the CuZn-SODs from bovine erythrocytes and liver and from horseradish were also subjected to denaturation with 6 M guanidine-HC1 followed by reversed-phase HPLC. Results showed only the appearance of a polypeptide band, and no chromophore group absorbing at 260 nm was detected (results not shown). Attempts were made to identify the nature of the noncovalently associated chromophore group detected in the peroxisomal CuZn-SOD from watermelon cotyledons, and using an HPLC method specific for nucleotides, nucleosides, and major purine bases (Wynants and Van Belle, 1985), it was found that the chromophore group apparently was not a nucleotide (results not shown). The sequence of peroxisomal CuZn-SOD was determined by N-terminal Edman degradation of the whole enzyme, and a set of peptides was generated by digestion with several peptide hydrolases. Overlapping of the sequence of the N terminus of the protein and several of the peptides provided an uninterrupted stretch of sequence of 40 amino acids (Fig. 3 ) . Another set of peptides that overlapped produced an additional uninterrupted stretch of 50 amino acids. Comparison of these two fragments and of peptides V-5, T-3, and V-6 with other SODs (Fig. 4) led to the establishment of the overall sequence for peroxisomal CuZn-SOD shown in Figure 4. Comparison of the partia1 amino acid sequence of peroxisomal CuZn-SOD (118 residues) with the sequences of different cytosolic and chloroplastic CuZn-SOD, showed a maximum homology of about 70% with the cytosolic enzymes from pine, maize, and tomato, and of about 90% with chloroplastic CuZn-SODs from spinach, petunia, tomato, and pea. The highest homology (93%) was that found with the chloroplast CuZn'OD from 'pinach' Cyanide at a I-mM concentration completely inhibited the CuZn-SOD activity, as expected for a SOD [email protected]% to this metalloenzyme family (Fridovich, 1986). Hydrogen peroxide inhibited peroxisomal CuZn-SOD activity at 25°C





21.5 14.4



;J 1.5 aJ









a Q

0.5 1

O 1.5 c O (D







2a O.! O

m Ll Q : (



15 20

25 30

min Figure 2. Reversed-phase HPLC of peroxisomal CuZn-SOD incubated with 6 M guanidine-HCI. A total of 70 pg of CuZn-SOD was processed. The absorption of eluted samples was recorded at 214 nm (A) and 260 nm (B). Inset, Coomassie blue-stained SDS-PACE of isolated polypeptide. Molecular mass standards are expressed in kD.

Figure 3. Amino acid sequence of peroxisomal CuZn-SoD' The amino Of lapping peptide fragments generated hy enzymatic cleavage of peroxisomal CuZn-SOD were aligned. V and T numhers refer to S. aureus V-8 protease- and pancreatic trypsin-generated peptides, respectively. X, Unidentified residue.

Plant Physiol. Vol. 108, 1995















I- T - 2 i






+ V-4















130 14 O E L E D D L G X G G H E L S L T X G N A G G X L A X G V

+ - T-4





--1 - 1










Characterization of Peroxisomal Superoxide Dismutase






pea pine CYTOSOLIC spinach cabbage tomato maize I1 pine 10.


pine 90






pea pine CYTOSOLIC spinach



Figure 4. Alignment of amino acid residues for peroxisomal CuZn-SOD, chloroplastic CuZn-SODs from spinach (Kitagawa et al., 1986), petunia (Tepperman et al.,1988), tomato (Perl-Treves et al., 19881, pea (Scioli and Zilinskas, 1988), and pine (Karpinski et al., 1992), and cytosolic CuZn-SODs from spinach (Kanematsu and Asada, 1990), cabbage (Steffens et al., 1986), tomato (Perl-Treves et al., 1988), maize (Cannon et al., 1987), and pine (Karpinski et al., 1992). Proteins were visually aligned. Since homologies were very high, no statistical methods for alignment and assessment of significance were used. Alignments were made to maximize homology. Residues that are identical to peroxisomal CuZn-SOD are boxed, and X indicates an unidentified residue. Residues coordinating Cu and Zn are indicated with asterisks.

Plant Physiol. Vol. 108, 1995

Bueno et al.


in a concentration- and time-dependent manner (Fig. 5). At a 0.5-mM hydrogen peroxide concentration the enzyme retained about 45% of its initial activity after 1 h of incubation. A similar remaining activity was observed after incubation with 1 mM hydrogen peroxide for about 20 min, and after 30 min at this concentration the CuZn-SOD activity was completely inhibited. Hydrogen peroxide concentrations of 2.5 and 5 mM produced the total loss of activity after 10 min. The thermal stability of peroxisomal CuZn-SOD was studied at 25, 50, 70, 80, and 100°C. After 6 d at 25°C the CuZn-SOD retained 65% of its initial activity, and after 4 h of incubation at 50°C the enzyme still kept more than 60% of its activity (results not shown). The minimum time at which the CuZn-SOD activity was reduced to 50% after heating at 70 and 100°C was about 15 and 3 min, respectively (Fig. 6). A polyclonal antibody was raised in rabbits against peroxisomal CuZn-SOD. The titration of the antibody in the immune serum by the ELISA method indicated an antibody titer of about 1/2,000,000. The specificity of purified IgG to peroxisomal CuZn-SOD was assayed by western blotting against watermelon cotyledon crude extracts and homogenous peroxisomal CuZn-SOD from watermelon cotyledons. The incubation of the nitrocellulose sheets with 1/1,000 dilution of anti-peroxisomal CuZn-SOD revealed a single 16.5-kD band that corresponded to the peroxisomal CuZn-SOD subunit (Bueno and del Rio, 1992) (Fig. 7). In crude extracts when the incubation times and antibody dilutions in the western blot assays were forced, occasionally a weak band of high Mr was detected. This band, however, was also produced under the same conditions by the preimmune serum (results not shown). The cross-reactivity of anti-peroxisomal CuZn-SOD against pure SODs



30 40 Time (min)


Figure 6. Temperature stability of peroxisomal CuZn-SOD. The purified enzyme (30 ^g ml"1) in 50 mM potassium phosphate buffer, pH 7.8, was incubated at the temperatures indicated. Aliquots were withdrawn at time intervals and assayed for residual SOD activity.

from different origins was studied (Fig. 8). The antibody did not recognize either the Mn-SOD from pea leaves nor the Escherichia coli Fe-SOD, but showed a strong crossreactivity against the 15-kD subunit of Sigma horseradish CuZn-SOD. In this case, different bands also appeared in the high Mr region of nitrocellulose sheets, probably due to the presence of abundant cross-reactive impurities in the commercial enzyme, which was overloaded (10 jxg) in the SDS-PAGE gels. The incubation of the anti-peroxisomal CuZn-SOD with animal CuZn-SODs showed a weak crossreaction with the 16-kD polypeptide band of this CuZnSOD subunit (Steinman, 1982), apart from other impurity bands. Different cell-free extracts from plant and animal











Time (min) Figure 5. Effect of hydrogen peroxide on the activity of peroxisomal CuZn-SOD. The enzyme (30 /xg ml"') was preincubated at 25°C in 50 mM potassium phosphate buffer, pH 7.8, containing different hydrogen peroxide concentrations (0.5, 1.0, 2.5, and 5.0 mM). At different incubation times, aliquots were withdrawn and their SOD activity was determined.

Figure 7. Western blot of watermelon peroxisomal CuZn-SOD and crude extracts from watermelon cotyledons. After SDS-PAGE on 12% gels and transfer to nitrocellulose sheets, the peroxisomal CuZn-SOD subunit was detected by incubation with a 1/1,000 dilution of antiperoxisomal CuZn-SOD, followed by the second antibody and staining with 4-chloro-1-naphthol. A, Watermelon cotyledons extract. B, Homogenous peroxisomal CuZn-SOD from watermelon cotyledons (Bueno and del Rfo, 1992). C, Molecular mass standards.

Characterization of Peroxisomal Superoxide Dismutase

kOa — 97.4 — 66.2 -. 45

-» 31

— 21.5

-« 14.4

Figure 8. Western blot analysis of Mn-SOD, Fe-SOD, and CuZnSODs from different sources with the anti-peroxisomal CuZn-SOD. SDS-PAGE and immunoblotting conditions were as described in Figure 7. Lane 1, Homogenous Mn-SOD from pea leaves (Sevilla et al., 1980) (1 jig). Lane 2, Fe-SOD from E. coli (Sigma) (10 jug). Lane 3, Bovine erythrocyte CuZn-SOD (Sigma) (10 jug). Lane 4, Bovine liver CuZn-SOD (Sigma) (10 jug). Lane 5, Horseradish CuZn-SOD (Sigma) (10 fig). Lane 6, Homogenous peroxisomal CuZn-SOD from watermelon (0.4 /ug). Lane 7, Molecular mass standards.

origin were examined by western blotting for reactivity to purified IgG to peroxisomal CuZn-SOD. Results showed varying degrees of cross-reactivity (Fig. 9). The intensity of the CuZn-SOD band in the immunoblots was greatest for pea leaves, followed by clover leaves, broad bean leaves, and erythrina leaves, whereas cell-free extracts from soybean nodules and spinach and lettuce leaves did not show any reaction (Fig. 9). Cell-free extracts of trout liver were recognized by the peroxisomal CuZn-SOD antibody, with a weak band of about 16 kD being detected.


horseradish also show a UV band around 260 nm, but treatment of these SODs with 6 M guanidine-HCl followed by reversed-phase HPLC showed that the chromophore responsible for the absorption at 260 nm was covalently bound to the protein. The only data known thus far on the peroxisomal CuZn-SOD chromophore is that it has an absorption maximum at 260 nm, it has a low molecular mass, probably lower than 1000 D, and that it is not a nucleotide. The full characterization of the peroxisomal CuZn-SOD chromophore group using MS and ^-NMR techniques is now under way in our laboratory. Attempts were made to compare the amino acid sequence of peroxisomal CuZn-SOD with that reported for CuZn-SOD from P. leiognathi, which has molecular mass and copper content similar to those of peroxisomal CuZnSOD (Puget and Michelson, 1974). It is well established that the CuZn-SOD from P. leiognathi has a low degree of amino acid sequence homology with most of the CuZn-SODs known (Steffens et al., 1983; Steinman, 1987; Bannister et al., 1991). The comparison with peroxisomal CuZn-SOD showed the existence in the P. leiognathi enzyme of extended regions of nonhomologous amino acid sequence and also an insert of 12 amino acids, and provided evidence for the absence of evolutionary relationships between peroxisomal CuZn-SOD and the P. leiognathi CuZnSOD. However, comparison of the partial amino acid sequence determined for peroxisomal CuZn-SOD with those reported for chloroplastic and cytosolic CuZn-SODs showed high similarities with respect to length and alignment of the homologous regions, and evidenced a high homology (about 90%) with chloroplastic CuZn-SODs. However, there are several amino acid replacements that differentiate peroxisomal CuZn-SOD from most plant SODs. Pro (P) at position 40 is conserved in all cytosolic and chloroplastic CuZn-SODs (Fig. 4), but in the peroxisomal enzyme it is substituted for by Glu (E), as in the bovine, human, sheep, pig, rat, and mouse CuZn-SOD (Bannister et al., 1991). Other amino acids that are nearly specific for


As we have previously reported (Bueno and del Rio, 1992), peroxisomal CuZn-SOD from watermelon cotyledons has a molecular mass of 33,000 D and is composed of two equal subunits of 16.5 kD. The peroxisomal enzyme has a pi of 4.0 (Sandalio and del Rio, 1987) and, unlike other CuZn-SODs, contains only 1 g-atom Cu and 1 g-atom Zn/mol dimer (Bueno and del Rio, 1992). Most CuZnSODs from animal and plant origin contain 2 g-atom Cu and 2 g-atom Zn/mol dimer (Steinman, 1982; Fridovich, 1986; Bannister et al., 1987, 1991), and apparently the only exception is the CuZn-SOD from Photobacterium leiognathi, which has a molecular mass of 33,100 D and contains 1 g-atom Cu and 2 g-atom Zn/mol (Puget and Michelson, 1974). Analysis by reversed-phase HPLC of the peroxisomal CuZn-SOD incubated with 6 M guanidine-HCl indicated that this peroxisomal enzyme, unlike other CuZn-SODs, contains a noncovalently bound chromophore group that is responsible for the A260 peak of the native enzyme. Pure CuZn-SODs from bovine erythrocytes and liver and from

2 3 4

5 6 7

— 21.5

— 14-4

Figure 9. Western blot analysis of cell-free extracts from plant and animal sources with the peroxisomal CuZn-SOD antibody. SDSPAGE and immunoblotting conditions were as described in Figure 7. The cell-free extracts used were: lane 1, pea leaves; lane 2, spinach leaves; lane 3, sorghum leaves; lane 4, clover leaves; lane 5, lettuce leaves; lane 6, erythrina leaves; lane 7, broad bean leaves; lane 8, soybean nodules; lane 9, trout liver. Lane 10, Molecular mass standards.


Bueno et al.

peroxisomal CuZn-SOD are Ile’s (I) at positions 58 and 78; Ser (S) at position 102 is present only in cytosolic CuZnSODs and not in the chloroplastic enzymes. Some of the ligands coordinating the Cu and Zn that were sequenced in the peroxisomal enzyme (His71, Hisso, Asps3, and His’”) appear to be conserved, as they are among a11 known CuZn-SODs. It is known that the chloroplastic CuZn-SOD is much more conserved than the cytosolic isozyme (Kwiatowski and Kaniuga, 1986; Perl-Treves et al., 1988; Scioli and Zilinskas, 1988; Kanematsu and Asada, 1990; Karpinski et al., 1992), and this indicates that the rate of mutation of the cytosolic CuZn-SOD is higher than that of the chloroplastic CuZn-SOD (Kwiatowski and Kaniuga, 1986; Kanematsu and Asada, 1990). It has been suggested that the high rate of production of 0;- radicals in chloroplasts would have a lethal effect if mutations in CuZn-SOD bring about a decrease or loss of their activity to scavenge superoxide radicals (Kanematsu and Asada, 1990). From a functional viewpoint, plant peroxisomes are closely related to chloroplasts, and both organelles participate in conjunction in the oxidative carbon cycle of photorespiration (Huang et al., 1983). Moreover, in peroxisomes as in chloroplasts, the production of superoxide radicals takes place as a result of their oxidative metabolism (de1 Rio et al., 1992). The high homology (90%)of the peroxisomal CuZn-SOD amino acid sequence with the chloroplastic CuZn-SOD could be explained by the close functional affinity between these two organelles, and strongly suggests that chloroplast and peroxisomal CuZn-SODs could be encoded by the same or closely related genes. Other remarkable features of peroxisomal CuZn-SOD are its thermal stability and resistance to hydrogen peroxide inactivation. The peroxisomal enzyme has half-life values at 50 and 70°C of more than 4 h and 15 min, respectively, which are higher than those reported for most CuZn-SODs (Steinman, 1982; Kwiatowski and Kaniuga, 1986). Peroxisoma1 CuZn-SOD was inhibited by hydrogen peroxide, as expected for a Cu,Zn-SOD, but in the presence of 0.5 mM hydrogen peroxide the enzyme had a half-life of nearly 60 min, a value that is higher than that shown by many CuZn-SODs (Kwiatowski and Kaniuga, 1986; Kanematsu and Asada, 1990). The higher resistance of the peroxisomal enzyme to hydrogen peroxide could be due to its lower metal content, since hydrogen peroxide apparently inactivates CuZn-SODs by destroying the His ligands of Cu and Zn in the protein (Kwiatowski and Kaniuga, 1986). Considering that peroxisomes are the main cellular site of hydrogen peroxide production (Huang et al., 1983; Fahimi and Sies, 1987; Van den Bosch et al., 1992), the increased resistance to hydrogen peroxide inactivation of peroxisomal CuZn-SOD could represent a protective adaptation mechanism of the enzyme to the peroxisomal environment rich in hydrogen peroxide. An interesting property of the purified antibody to plant peroxisomal CuZn-SOD is its cross-reactivity with the phylogenetically distant animal CuZn-SODs. The antibody to spinach CuZn-SOD does not cross-react with bovine erythrocyte CuZn-SOD, and the antibodies to the cytosolic CuZn-SODs do not recognize the chloroplastic enzymes

Plant Physiol. Vol. 108, 1995

and vice versa (Asada et al., 1977; Kanematsu and Asada, 1990). Human and animal CuZn-SODs were traditionally considered as exclusive cytosolic enzymes, but recent reports have demonstrated the presence of CuZn-SOD in peroxisomes from human and animal cells (Keller et al., 1991; Crapo et al., 1992; Dhaunsi et al., 1992; Wanders and Denis, 1992). The weak cross-reactivity observed between the antibody to plant peroxisomal CuZn-SOD and animal CuZn-SODs could be due to the antibody recognition of the animal CuZn-SOD isozyme localized in peroxisomes. Research carried out on plant peroxisomes in our laboratory in recent years has allowed us to postulate that oxyradical metabolism may be a common metabolic property of peroxisomes of eukaryotes (de1 No et al., 1990, 1992). The main function of CuZn-SOD in peroxisomes could be the protection of these cell organelles against toxic effects derived from superoxide radicals generated therein. An example could be the protection of catalase from its inhibition by superoxide radicals (Kono and Fridovich, 1982). Under certain conditions of depressed catalase activity induced by different plant stress situations (de1Rio et al., 1992), 0;- radicals produced in peroxisomes could react with hydrogen peroxide and give rise to the vastly more reactive hydroxyl radicals (.OH) (Halliwell and Gutteridge, 1989; Elstner, 1990). This strong oxidizing species damages biological membranes and reacts with most of the compounds present in biological systems, including DNA (Halliwell and Gutteridge, 1989; Elstner, 1990; Packer and Glazer, 1990). The availability of the antibody against peroxisomal CuZn-SOD will allow us to carry out immunolocalization studies of this enzyme in plant peroxisomes. Additionally, molecular biology studies will be conducted in our laboratory to isolate a cDNA clone for this SOD and to study the mechanism of import of the enzyme into peroxisomes.


The valuable help of Drs. R. Hermoso-Yáííez,J. Fonollá-Joya, and J. Boza, Jr., in the antibody preparation is gratefully acknowledged. The authors are thankful to Dr. C. Prieto-Palomino and Mr. L. Pérez-Martínez (Estación Experimental de1 Zaidín) and Dr. J.M. Romera-Gómez (Unión Industrial Agroganadera, Sociedad Anonima, Granada) for conducting the amino acid analysis. Received January 17,1995; accepted March 24, 1995. Copyright Clearance Center: 0032-0889/95/108/1151/10.

LITERATURE ClTED Asada K (1992) Production and scavenging of active oxygen in chloroplasts. In JG Scandalios, ed, Current Communications in Cell and Molecular Biology. Vol 5. Molecular Biology of Free

Radical Scavenging Systems. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp 173-192 Asada K, Kanematsu S, Uchida K (1977) Superoxide dismutases in photosynthetic organisms: absence of the cuprozinc enzyme in eukaryotic algae. Arch Biochem Biophys 179 243-256 Bannister JV, Bannister WH, Rotilio G (1987) Aspects of the structure, function, and applications of superoxide dismutase. CRC Crit Rev Biochem 2 2 111-180

Characterization of Peroxisomal Superoxide Dismutase

Bannister WH, Bannister JV, Barra D, Bond J, Bossa F (1991) Evolutionary aspects of superoxide dismutase: the copper/zinc enzyme. Free Radical Res Commun 12-13: 349-361 Becana M, Paris FJ, Sandalio LM, de1 Rio LA (1989) Isoenzymes of superoxide dismutase in nodules of Phaseolus vulgaris L., Pisum sativum L., and Vigna unguiculata (L) Walp. Plant Physiol 9 0 1286-1292 Bueno P, de1 Rio LA (1992) Purification and properties of glyoxysoma1 cuprozinc superoxide dismutase from watermelon cotyledons (Citrullus vulgaris Schrad.). Plant Physiol 98: 331-336 Cannon RE, White JA, Scandalios JG (1987) Cloning of cDNA for maize superoxide dismutase 2 (SOD 2). Proc Natl Acad Sci USA 84: 179-183 Cohen SA, Mays M, Tarvin TL (1989) The Pico-Tag Method: A Manual of Advanced Techniques for Amino Acid Analysis. Millipore Corp, Bedford, MA Corpas FJ, Bunkelmann J, Jiang L, Trelease RN (1994) Identification of superoxide dismutase isoforms in glyoxysomes of oilseeds (abstract No. 804). Plant Physiol 105: S-146 Crapo JD, Oury T, Rabouille C, Slot JW, Chang L-Y (1992) Copper,zinc superoxide dismutase is primarily a cytosolic protein in human cells. Proc Natl Acad Sci USA 89: 10405-10409 de1 Rio LA, Fernández VM, Rupérez FL, Sandalio LM, Palma JM (1989) NADH induces the generation of superoxide radicals in leaf peroxisomes. Plant Physiol 89: 728-731 de1 Rio LA, Lyon DS, Olah I, Glick B, Salin ML (1983) Immunocytochemical evidence for a peroxisomal localization of manganese superoxide dismutase in leaf protoplasts from a higher plant. Planta 1 5 8 216-224 de1 Rio LA, Sandalio LM, Palma JM (1990) A new cellular function for peroxisomes related to oxygen free radicals? Experientia 46 989-992 de1 Rio LA, Sandalio LM, Palma JM, Bueno P, Corpas FJ (1992) Metabolism of oxygen radicals in peroxisomes and cellular implications. Free Radical Biol Med 13:557-580 Dhaunsi GS, Gulati S , Singh AK, Orak JK, Asayama K, Singh I (1992) Demonstration of Cu-Zn superoxide dismutase in rat liver peroxisomes: biochemical and immunochemical evidence. J Biol Chem 267: 6870-6873 Elstner EF (1990) Der Sauerstoff Biochemie, Biologie, Medizin. BI-Wissenschafts-Verlag, Mannheim, Germany Fahimi HD, Sies H (1987) Peroxisomes in Biology and Medicine. Spinger-Verlag, Berlin Fridovich I (1986) Superoxide dismutases. In A Meister, ed, Advances in Enzymology and Related Areas of Molecular Biology. John Wiley & Sons, New York, pp 61-97 Giménez-Gallego G, Conn G, Hatcher VB, Thomas KA (1986) The complete amino acid sequence of human brain-derived acidic fibroblast growth factor. Biochem Biophys Res Commun 1 3 8 611-617 Giménez-Gallego G, Rodkey J, Bennett C, Rios Candelore M, DiSalvo J, Thomas K (1985) Brain-derived acid fibroblast growth factor: complete amino acid sequence and homologies. Science 230: 1385-1388 Halliwell B, Gutteridge JMC (1989) Free Radicals in Biology and Medicine, Ed 2. Clarendon Press, Oxford, UK Hermoso R, Chueca A, Lázaro JJ, López-Gorgé J (1987) An immunological method for quantitative determination of photosynthetic fructose-1,6-bisphosphatase in leaf crude extracts. Photosynth Res 1 4 269-278 Huang AHC, Trelease RN, Moore TS (1983) Plant Peroxisomes. Academic Press, New York Kanematsu S , Asada K (1990)Characteristic amino acid sequences of chloroplast and cytosol isozymes of CuZn-superoxide dismutase in spinach, rice and horsetail. Plant Cell Physiol 31: 99-112 Karpinski S , Wingsle G, Olsson O, Hallgren JE (1992) Characterization of cDNAs encoding CuZn-superoxide dismutases in Scots pine. Plant Mo1 Biol 18: 545-555 Keller GA, Warner TG, Steimer KS, Hallewell RA (1991) Cu,Zn superoxide dismutase is a peroxisomal enzyme in human fibroblasts and hepatoma cells. Proc Natl Acad Sci USA 88: 7381-7385


Kitagawa Y, Tsunasawa S, Tanaka N, Katsube Y, Sakiyama F, Asada K (1986) Amino acid sequence of copper,zinc-superoxide dismutase from spinach leaves. J Biochem 99: 1289-1298 Kono Y, Fridovich I (1982) Superoxide radical inhibits catalase. J Biol Chem 257: 5751-5754 Kwiatowski J, Kaniuga Z (1986) Isolation and characterization of cytosolic and chloroplast isoenzymes of Cu,Zn-superoxide dismutase from tomato leaves and their relationships to other Cu,Zn-superoxide dismutases. Biochim Biophys Acta 874 99-1 15 Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685 López-Huertas E, Sandalio LM, de1 Rio LA (1994) Preliminary study of the superoxide-generating system of membranes from plant peroxisomes. In K Asada, T Yoshikawa, eds, Frontiers of Reactive Oxygen Species in Biology and Medicine. Elsevier, Amsterdam, pp 37-38 Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951) Protein measurement with the Folin phenol reagent. J Biol Chem 193: 265-275 Mannaerts GP, Van Veldhoven PP (1993) Metabolic pathways in mammalian peroxisomes. Biochimie 7 5 147-158 McCord JM, Fridovich I (1969) Superoxide dismutase: an enzymic function for erythrocuprein. J Biol Chem 244 6049-6055 Murphy JB, Kies MW (1960) Note on spectrophotometric determination of proteins in dilute solutions. Biochim Biophys Acta 45: 382-384 Packer L, Glazer AN (1990)Oxygen radicals in biological systems. Part B. Oxygen radicals and antioxidants. In Methods in Enzymology 186. Academic Press, San Diego, CA Pajot P (1976) Fluorescence of proteins in 6-M guanidine hydrochloride. A method for the quantitative determination of tryptophan. Eur J Biochem 6 3 263-269 Parker MW, Schnini ME, Bossa F, Bannister JV (1984) Chemical aspects of the structure, function and evolution of superoxide dismutases. Inorg Chim Acta 91: 307-317 Perl-Treves R, Nacmias B, Aviv D, Zeelon EP, Galun E (1988) Isolation of two cDNA clones from tomato containing two different superoxide dismutase sequences. Plant Mo1 Biol 11: 609-624 Puget K, Michelson AM (1974) Isolation of a new copper-containing superoxide dismutase, bacteriocuprein. Biochem Biophys Res Commun 5 8 830-838 Sandalio LM, de1 Rio LA (1987) Localization of superoxide dismutase in glyoxysomes from Citrullus vulgaris. Functional implications in cellular metabolism. J Plant Physiol 127: 395409 Sandalio LM, de1 Rio LA (1988) Intraorganellar distribution of superoxide dismutase in plant peroxisomes (glyoxysomes and leaf peroxisomes). Plant Physiol 88: 1215-1218 Sandalio LM, Fernández VM, Rupérez FL, de1 Rio LA (1988) Superoxide free radicals are produced in glyoxysomes. Plant Physiol 87: 1-4 Sandalio LM, Palma JM, de1 Rio LA (1987) Localization of manganese superoxide dismutase in peroxisomes isolated from Pisum sativum L. Plant Sci 51: 1-8 Scioli JR, Zilinskas BA (1988) Cloning and characterization of a cDNA encoding the chloroplastic copper/zinc-superoxide dismutase from pea. Proc Natl Acad Sci USA 8 5 7661-7665 Sevilla F, López-Gorgé J, GÓmez M, de1 Rio LA (1980) Manganese superoxide dismutase from a higher plant. Purification of a new Mn-containing enzyme. Planta 1 5 0 153-157 Steffens GJ, Bannister JV, Bannister WH, Flohé L, Gunzler WA, Kim SMA, Otting F (1983) The primary structure of Cu-Zn superoxide dismutase from Photobacterium leiognathi: evidence for a separate evolution of Cu-Zn superoxide dismutase in bacteria. Hoppe-Seyler’s Z Physiol Chem 364 675-690 Steffens GJ, Michelson AM, Otting F, Puget K, Strassburger W, Flohé L (1986) Primary structure of Cu-Zn superoxide dismutase of Brassica oleracea proves homology with corresponding enzymes of animals, fungi and prokaryotes. Biol Chem HoppeSeyler 367: 1007-1016


Bueno et

Steinman HM (1982) Superoxide dismutases: protein chemistry and structure-function relationship. In LW Oberley, ed, Superoxide Dismutase, Vol 1. CRC Press, Boca Raton, FL, pp 11-68 Steinman HM (1987) Bacteriocuprein superoxide dismutase of Photobacterium leiognathi: isolation and sequence of the gene and evidence for a precursor form. J Biol Chem 262: 1882-1887 Tepperman J, Katayama C, Dunsmuir P (1988) Cloning and nucleotide sequence of a petunia gene encoding a chloroplastlocalized superoxide dismutase. Plant Mo1 Biol 11:871-872 Tolbert NE (1981) Metabolic pathways in peroxisomes and


Plant Physiol. Vol. 108, 1995

glyoxysomes. Annu Rev Biochem 5 0 133-157 Van den Bosch H, Schutgens RBH, Wanders JA, Tager JM (1992) Biochemistry of peroxisomes. Annu Rev Biochem 61: 157-197 Wanders RJA, Denis S (1992) Identification of superoxide dismutase in rat liver peroxisomes. Biochim Biophys Acta 1115 259-262 Wynants J, Van Belle H (1985)Single-run high performance liquid chromatography of nucleotides, nucleosides, and major purine bases and its application to different tissue extracts. Ana1 Biochem 144 258-266