0 1994 by The American Society for Biochemistry and Molecular Biology, Inc. Vol. .... acetylpenicillamine (SNAP) was from BIOMOL Research Laboratoires. Inc.
Vol. 269, No. 47, Issue of November 25, pp, 29409-29415, 1994 Printed in U.S.A.
THEJOURNALOF BIOLOGICAL CHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc
Aconitase Is Readily Inactivated by Peroxynitrite, but Not by Its Precursor, Nitric Oxide* (Received for publication, June 9, and in revised form, August 5 , 1994) Laura Castro, Marianela Rodriguez, and Rafael Radif From the Department of Biochemistry, Facultad de Medicina, Universidad de la Republica, Avenida General Flores, 2125, 11800 Montevideo, Uruguay
Several oxidants includinghydrogen peroxide, ferricyanide, and persulfate can inactivate the enzyme because of the oxidant-mediated lost of the labile iron from the cluster, giving a paramagnetic [3Fe-4SI1+cluster (4, 5). In this regard,superoxide radical (Oi), a toxic by-product of normal aerobic metabolism (6), was shown to be particularly reactive with hydroxyacid dehydratases containing prosthetic iron-sulfur clusters (7-10). Superoxide anion inactivatesEscherichia coli aconitase, with a rate constant above lo7 M - ~s-’, with this inactivation inhibitable by virtual substrates (9). Nitric oxide (‘NO) is anendogenous free radical formed in a diverse variety of cell types by the enzymatic oxidation of Larginine to citrulline (11).In addition to its functions as a signal-transducing molecule (12), nitric oxide has been implicated incytotoxic processes due to its ability to interfere incell energy metabolism (13),protein synthesis(141, and ironhomeostasis (15). Mitochondrial aconitase hasbeen repeatedly indicated asa major target of ‘NO-mediated toxicity, and inactivation of aconitase in tumorcells, rat hepatocytes, Kupffer cells, and Langerghans’ p-cells by ‘NO or ‘NO-generating systems has been shown (13,16-19). Similarly, ‘NO has been implicated in the activationof iron-responsive element-binding protein in macrophages (18). Iron-responsive element-binding protein is recognized t o be cytosolic aconitase, which binds to the ironresponsive element (stemloops of mRNA of the 5’-untranslated repeat of ferritin and the 3“untranslated repeatof the transferrin receptor) whencells are iron-depleted. This activationis accomplished with the lost of enzyme activity and iron from the enzymatic cluster (18, 20-23), and there is agreement that the properties and iron-sulfur centers of cytosolic and mitochondrial aconitases are rather similar(22). However, most of the work on ‘NO-mediated inactivation of aconitase has been carried out in cell systems, and no mechanistic information is available on the reactionsof ‘NO and‘NO-derived specieswith purified aconitase. In addition to thefact that ‘NO can bind to iron-sulfur centers (24,251, we hypothesize that the interactions between ‘NO Thecatalytically activeform of mitochondrial aconitase and 0, are of critical importance to the understanding of the (citrate(is0citrate) hydro-lyase, EC 4.2.1.3) contains a cubane [4Fe-4S12+cluster. In this cluster, only three of the iron atoms role of ‘NO in aconitase inactivation. Indeed, ‘NO itself is a are ligated directly to cysteines of the protein backbone. The weak oxidant, butit turns intoa strong oxidizing intermediate fourth iron(Fecu) is only ligated to inorganic sulfurof the iron- after its fast reaction with 0, (k = 6.7 x lo9M” s-’) (26),yielding sulfur cluster and has a free coordination site that participates peroxynitrite anion (ONOO-; E‘, = +1.4 V) (27). ONOO- has a in the bindingof substrates to theactive site interacting with half-life of lo0p ~ led ) to a moderate inhibition of the enzyme, which could be fully overcomeby ’NO displacement under an argon-saturated atmosphere, in agreement withthe formation of a reversible inhibitory complex between ’NO and the active site of aconitase. Superoxideinactivated mitochondrial aconitase at (3.5 2) x lo5 M - ~s-’, a reaction rate 3 orders of magnitude slower than its reaction rate with ‘NO. 0; could represent the main mechanism of inactivation of the enzyme in systems in which it is formedwithoutsignificantconcomitantproduction of ‘NO. Our results imply thatthe mechanismsby which ‘NO and 0; inactivate aconitase in cell systems may not be simple due to theirdirect reactions with the iron-sulfur cluster, but may rely on the formation of ONOO-.
*
*
29409
29410
Inactivation Aconitase
by Peroxynitrite
which ONOO- is produced at increased rates (37-40). In thiswork, we have studied thereactivity of 'NO, O;, and ONOO- toward isolated pig heart mitochondrial aconitase to determine whether previously ascribed effects of 'NO are in fact solely due t o direct reactions with the enzyme or may rely on the formation of more oxidizing 'NO-derived intermediates. In addition, our observations provide insight into the molecular mechanisms by which peroxynitrite attacks [4Fe-4S]-containing proteins.
Hamilton syringe to active aconitase solutions 100 in mM Tris-HC1, pH 7.6, containedin sealed rubber-capped tubes and incubated for 5 min at 25 "C. Aconitase activity was measured immediately after. Under our experimentalconditions,allperoxynitrite-dependentreactionswere completed in 10-20 s as the half-life of ONOO- is 10 m ~ . When GSNO was used as a 'NO-generating system, aconitase showed a rapid and progressive loss of activity (Fig. 5). In In our system,F, represents thefraction of protection of aconicontrast, there was no significant effect on aconitase activity tase ([Dl) in thepresence of methionine or GSH ([SI), k, is the when the enzyme was incubated with SNAP at similar 'NO rate constant for the reaction of ONOO- with the scavenger, and k, is the rate constant for the reaction of ONOO- with exposure levels as with GSNO (data not shown). Estimated Rate Constantsfor ONOO- and 0, Reaction with aconitase. The k, values were taken as150 and 740 M - ~s-l for according to Refs. 56 and 27, Aconitase-Simple competition kinetics for oxidation of a de- methionineandglutathione
+
Aconitase Inactivation by Peroxynitrite 80 r
29413
100 h
"
0
0
2
4
6
8
5
10
15
20
25
10
[scavenger] (mM) FIG.6. Rate constant determination for the reaction between peroxynitrite and aconitase. Aconitase (30 p ~ was ) preincubated with different concentrations of methionine (m) or GSH (0)for 5 min before the addition of 100 PM peroxynitrite.Aconitase activity was measured immediately after, and Equation3 was applied for each concentration of scavenger.Theslopeof the plots represents 1/k,, and averaging both slopes results in a k, of (1.42 2 0.09)x lo5 M - ~s-'.
respectively. If a competitive mechanism holds, plotting Fi[Dl/ (1 - Fi)k, against [SI results in a straight line with the slope 1kd,which allows the determination of k,. Indeed, a linear relationship was found for both scavengers (Fig. 61, and the estimated rate constant for the reaction of peroxynitrite with mitochondrial pig heart aconitase is (1.42 0.09) x lo5 s-' at pH 7.6 and 25 "C. As in previous works (4, lo), aconitase activity decayed in an air-saturated atmosphere (Fig. 7, line 2). Incubation of aconitase with a superoxide- plus hydrogen peroxide-generating system caused a further progressive loss of enzyme activity (line 61, which was dependenton both of the reactive oxygen species (0; and H,O,) since catalase (line 3) or superoxide dismutase (lines 4 and 5) alone partially prevented the inactivation of aconitase. Full protection from xanthine oxidase-derived oxidants was obtained when catalase and superoxide dismutase were simultaneously present as the rate of inactivation under this condition was the same as the background rate of air inactivation (line 1 ) . This latter observation indicates that the amount of catalase added was sufficient to cope with all of the hydrogen peroxide yielded either directly from xanthine oxidase or from the enzymatic dismutation of 0,. Since there was no further protection at superoxide dismutase concentrations >1.92 1"(lines 1 and 4 ) , we can assume that the decrease in aconitase activity between lines 4 and 6 represents 100% superoxide-mediated inactivation. Theseallows us toapply Equation 2 t o the different points on line 5 (for which a partially ) inhibitory concentration of superoxide dismutase (0.77 p ~ was used) using a rate constant of 2 x lo9 M - ~s-l between 0; and CuZn superoxide dimutase (6). The rate constant calculated for the reaction of 0, with mitochondrialpig heart aconitase is (3.5 k 2) x lo6 M - ~s-' at pH 8.4 and 25 "C.
time (mid FIG.7. Inactivation of mitochondrial aconitaseby superoxide. Aconitase (76 p ~ was ) incubated with hypoxanthine (150 p ~ and ) xanthine oxidase (5.5 milliunits/ml) as describedunder"Experimental Procedures." Aliquots were withdrawnat different times, and activity was measuredwith the coupled assay at 340 nm in the presence of 100 PM allopurinol to inhibit the remaining xanthine oxidase activity.Line 2 represents the incubation of aconitase under an air-saturated atmosphere with no xanthine oxidase present. All other conditions included xanthine oxidaseplus 1.92 superoxide dismutase and 0.5 PM catalase (line 1 ), 0.5 p~ catalase (line 3 ) , 1.92 PMsuperoxide dismutase (line 4 ) , 0.77 PM superoxide dismutase (line 51,and no additions (line 6 ) .
Nitric oxide is a highly diffusible free radical known to have affinity for iron (15). Indeed, various authors haveshown that after several stimuli (Le. y-interferon, lipopolysaccharide, and tumor necrosis factor-a) trigger 'NO synthesis inmacrophages, the formation of nitrosyl-iron complexes becomes detectable in macrophages as well as in their targetcells (24,25).Mitochondrial aconitase has been repeatedly implicated as one of the major intracellular targets of 'NO, and the lost of aconitase activity observed has been attributed to thedirect reactions of 'NO with the iron-sulfur cluster (13, 17, 18, 60). However, herein we have shown a rather marginal effect of authentic 'NO on isolated mitochondrial aconitase activity (Fig. 2). The lack of effect observed at physiologically relevant concentra) mitochondrial aconitase activityis in tions of 'NO (
