Products of the reaction of cytosolic and mitochondrial aconitases with nitric oxide

M. Claire Kennedy1, William E. Antholine2 and Helmut Beinert3

1 Department of Chemistry, Gannon University, Erie, PA 16561, USA 2Biophysics Research Institute, Medical College of Wisconsin, Milwaukee, WI53226, USA 3Institute for Enzyme Research, University of Wisconsin—Madison, 1710 University Avenue, Madison,

WI 53726-4087, USA

Aconitases are enzymes that catalyze the stereospecific isomerization of citrate to isocitrate in the tricarboxylic cycle [1]. They are iron-sulfur proteins that contain [4Fe—4S] clusters. In most Fe-S proteins, the cluster is involved in electron transfer without itself undergoing significant structural changes [2,3]. Aconitases, however, are unique in that their Fe-S cluster directly interacts with its substrate [4,5]. Mammalian tissues encode two different aconitases, mitochondrial, m-acon and cytosolic, c-acon. Cytosolic aconitase is a bifunctional protein, which upon the loss of its Fe-S cluster is converted from an enzyme to iron-regulatory protein-1 (IRP1) [4,6,7]. This may occur when the intracellular concentration of iron drops to suboptimal levels. Therefore, the structural integrity of the Fe-S cluster is crucial for understanding the physiological role of aconitases. Interest in aconitases greatly increased after it was recognized that these enzymes play an important role in apoptosis [8] and in addition, that nitric oxide (NO) plays a critical role in the interconversion of c-acon and IRP1 [9]. It is, therefore, not surprising that intensive research efforts were made concerning the action of the catalytic cluster and the mechanisms by which this cluster could be modified or disrupted.

The geometrical structure of the 4Fe cluster in aconitases is that of a distorted cube with alternating Fe and S atoms at its corners. This structure is typical of the cubane type [4Fe—4S] cluster found in a wide range of different proteins [2,3]. In most cubane clusters, the iron atoms are ligated to cysteine thiolates. Aconitases are unusual in that one of the iron atoms, Fea, is coordinated to a hydroxyl group in the absence of substrate. During catalysis, the coordination of Fea changes from tetrahedral to octahedral when substrate and water are bound [4]. Another important feature of the aconitase cluster is the ease with which Fea is lost upon oxidation, resulting in inactivation of the enzyme and formation of a [3Fe—4S] cluster. This arrangement illustrates how the enzymatic activity of aconitase is directly related to the structural geometry of the cluster. The reader interested in more detail on the reaction and the use of electron paramagnetic resonance (EPR) and electron nuclear double resonance (ENDOR)

spectroscopy in the exploration of the mechanisms involved in the functioning of the enzyme may consult the Refs. [2,4,5,10].

There are contradictory reports in the existing literature concerning the reaction of NO with the enzyme aconitase [11-14]. While together at the National Biomedical EPR Center at Milwaukee, the authors decided to seize the opportunity and combine their experience with the enzyme aconitase and EPR, respectively, to determine the sensitivity of aconitase toward NO [14]. As expected, the result was that the active forms of both m-acon and c-acon readily form dinitrosyl-iron complexes (DNICs) containing protein cysteinyl ligands and are thus inactivated. The finding that under some conditions m-acon can also form an analogous dinitrosyl-iron-histidyl complex with NO was unexpected.

In order to pin down the reasons for the contradictory results found in the literature, we will consider in detail the properties of all the reagents and procedures used in the relevant experiments. The most important ingredient is, of course, the protein aconitase. Preparations having "aconitase activity" have been commercially available from Sigma Chemical Company for many years. Out of interest, we acquired a sample of this material that was most likely quite similar to the Sigma aconitase used by other authors. On analysis, we found extremely low enzyme activity in the sample we had received, even after subjecting it to the same activation procedure routinely used with aconitase prepared in our laboratory [15]. Furthermore, the low-temperature EPR spectrum indicated very minor amounts of aconitase present in relation to the high protein content of the commercial sample. In addition, the light absorption spectrum showed features typical of heme compounds, as was also shown in the article by Castro et al. [11], in which no sensitivity of "aconitase" toward NO was found. It is most likely that in such samples NO reacts with the large amounts of heme compounds and other impurities present, such that any aconitase would have been spared.

Another possible source of ^reproducibility concerned the mode of addition of NO. We found that adding NO in a solution after gassing under anaerobic conditions is less reproducible than evolving NO from compounds referred to as NONOates and monitoring the NO content with a specific NO electrode. Addition of substrate delayed the decomposition of aconitase to some extent, but did not prevent reaction of the enzyme with NO. We did, however, find that as little as 10 mM citrate had an influence on the rate of decomposition of the NONOate, which was an unexpected complication.

Although the native [4Fe—4S]2+ clusters found in active aconitases are in the S = 0, EPR-silent state, the oxidative loss of Fea from their clusters leads to inactivation with the formation of [3Fe—4S]1+ clusters which can be observed by EPR at temperatures below 30 K at g ~ 2.02 [16]. In contrast, the EPR signals for the DNIC complexes formed upon addition of NO to the enzyme can be observed at all temperatures. Thus, in order to obtain maximal information from the reaction of the various forms of aconitase and NO, the products of the reaction were monitored by EPR at room (RT) and at low temperatures, i.e. 77 K and < 20 K, which allowed us to decide whether the DNIC signals observed originated from a macromolecular (protein bound) source or from a rapidly tumbling species of low molecular weight. Consideration was also given to the various species of DNIC which might be formed and to their electronic states such as the d9 state obtained on reduction of the EPR detectable d7 species as well as the other oxidation states, d6 or d8, which are not detected by EPR. Mossbauer spectroscopy would be a suitable method for recognizing these latter species.

Fig. 1. EPR spectra of the d7 and d9 forms of DNI-acon. (Upper curve) EPR spectrum of product formed upon the anaerobic incubation of 120 ^M [4Fe—4S]2+ active m-acon with 8.3 mM spermine NONOate, pH 6.5, at 23°C for 30 min (no enzyme activity remaining). (Lower curve) EPR spectrum of (a) following the addition of excess dithionite. Conditions of spectroscopy: microwave power and frequency, 0.1 mW and 9.229 GHz; modulation amplitude and frequency, 0.5 mT and 100 kHz; time constant, 0.128 s; scanning time, 20 mT/min; temperature, 15 K.

Fig. 1. EPR spectra of the d7 and d9 forms of DNI-acon. (Upper curve) EPR spectrum of product formed upon the anaerobic incubation of 120 ^M [4Fe—4S]2+ active m-acon with 8.3 mM spermine NONOate, pH 6.5, at 23°C for 30 min (no enzyme activity remaining). (Lower curve) EPR spectrum of (a) following the addition of excess dithionite. Conditions of spectroscopy: microwave power and frequency, 0.1 mW and 9.229 GHz; modulation amplitude and frequency, 0.5 mT and 100 kHz; time constant, 0.128 s; scanning time, 20 mT/min; temperature, 15 K.

We shall, in the following, present some illustrations from the experiments that convinced us that both m-acon and c-acon are inactivated by NO through the formation of DNIC complexes. The only EPR signal observed at RT or 77 K upon the bolus injection of a solution of NO at pH 7.5 to active [4Fe—4S]2+ m-acon is that of the d7 DNIC form at g ~ 2.04 (Fig. 1). This same signal is observed at pH 6.5 when the NO source is a NONOate solution. These samples, upon reduction with dithionite, yield the d9 species which upon analysis by EPR gives a spectrum with a signal at g ~ 2.006 (Fig. 1). When examined at temperatures below 15 K, the spectrum of these samples is a composite of the g ~ 2.02 DNIC and g ~ 2.04 [3Fe—4S]1+ signals. Quantitation of these signals does not correlate with loss of activity nor does formation of [3Fe—4S]1+ aconitase parallel loss of activity (Fig. 2). This may indicate the presence of EPR-silent d6 or d8 DNIC species or further disassembly of the Fe—S cluster. Results similar to those obtained for active m-acon were also observed for c-acon. A notable difference was that a higher ratio of NO to enzyme was needed to obtain a similar loss of activity of c-acon and also that a transient EPR signal, tentatively assigned to a thiyl radical was seen on occasion.

The reaction of NO with inactive [3Fe—4S]1+ m-acon yields spectra at RT and 77 K that exhibited a signal not found in experiments with the active [4Fe—4S]2+ form. This signal appeared early in the reaction and disappeared with time. As shown in Fig. 3, subtraction

0 40 80 120

Time (min)

Fig. 2. Time course of inactivation of m-acon and formation of DNI-acon in the presence of spermine NONOate. Spermine NONOate, final concentration 4.5 mM, was added anaerobically to a solution of 200 ^M [4Fe—4S]2+ m-acon in 0.1 M MES at pH 6.6 and at 23°C. (A) Continuous monitoring of the EPR signal at g = 2.04 was made using a flat cell at a microwave power of 100 mW and at a modulation amplitude and frequency of 0.5 mT and 100 kHz. (B) Activity measurements of the same sample as in (A) were performed simultaneously as described in Kennedy et al. [15].

0 40 80 120

Time (min)

Fig. 2. Time course of inactivation of m-acon and formation of DNI-acon in the presence of spermine NONOate. Spermine NONOate, final concentration 4.5 mM, was added anaerobically to a solution of 200 ^M [4Fe—4S]2+ m-acon in 0.1 M MES at pH 6.6 and at 23°C. (A) Continuous monitoring of the EPR signal at g = 2.04 was made using a flat cell at a microwave power of 100 mW and at a modulation amplitude and frequency of 0.5 mT and 100 kHz. (B) Activity measurements of the same sample as in (A) were performed simultaneously as described in Kennedy et al. [15].

of the g ~ 2.04 signal obtained at 75 min at RT from the spectrum of the earlier sample at 15 min yields a new signal with g'x = 2.050, g'y = 2.032 and g'z = 2.004. Similar samples examined at 77 K gave identical results. Characteristics of the EPR spectra suggested that the signal might be due to a histidyl-iron-nitrosyl complex which had been observed by others [17]. The spectrum of a solution containing Fe2+, imidazole and NONOate shown in Fig. 4, curve A, exhibited a mixture of two signals which was converted to a single species upon addition of dithionite to the solution (Fig. 4, curve B). Subtraction of the signal of

Fig. 3. EPR spectra obtained at 23°C, during the reaction of [3Fe—4S]2+ m-acon and spermine NONOate. Spectrum (curve A) is the scan at 15 min and spectrum (curve B) is the scan at 75 min of a mixture of 95 ^M [3Fe—4S]1+ m-acon with 2.6 mM spermine NONOate. Conditions of spectroscopy as in Fig. 2.

Fig. 3. EPR spectra obtained at 23°C, during the reaction of [3Fe—4S]2+ m-acon and spermine NONOate. Spectrum (curve A) is the scan at 15 min and spectrum (curve B) is the scan at 75 min of a mixture of 95 ^M [3Fe—4S]1+ m-acon with 2.6 mM spermine NONOate. Conditions of spectroscopy as in Fig. 2.

Fig. 4. EPR spectra of the d7 and d9 forms of the dinitrosyl-iron-imidazole complex. A spectrum of a solution prepared anaerobically in 0.1 M HEPES, pH 7.5, containing 10 mM imidazole, 0.5 mM ferrous ammonium sulfate and 5.0 mM diethylamine NONOate and incubated 30 min at room temperature. (curve B) Spectrum of sample as in (curve A) to which an excess of dithionite has been added. (Curve A — B) is the difference spectrum of (curve A) and (curve B). The conditions of spectroscopy are as described in the legend of Fig. 1 except that the microwave frequency is 9.090 GHz and microwave power 2 mW.

Fig. 4. EPR spectra of the d7 and d9 forms of the dinitrosyl-iron-imidazole complex. A spectrum of a solution prepared anaerobically in 0.1 M HEPES, pH 7.5, containing 10 mM imidazole, 0.5 mM ferrous ammonium sulfate and 5.0 mM diethylamine NONOate and incubated 30 min at room temperature. (curve B) Spectrum of sample as in (curve A) to which an excess of dithionite has been added. (Curve A — B) is the difference spectrum of (curve A) and (curve B). The conditions of spectroscopy are as described in the legend of Fig. 1 except that the microwave frequency is 9.090 GHz and microwave power 2 mW.

0 0

Post a comment