The Presence of Bound Cyanide in the Naturally Inactivated Form of Nitrate Reductase of ChZoreZZa uulgwis

Abstract In the presence of NADH and cyanide, NADH-nitrate oxidoreductase (EC 1.6.6.1) from Chlorella vulgaris is converted to an inactive form which is readily reactivated by ferricyanide. Experiments with H14CN indicated that the inactivation process is associated with the firm binding to the protein of 0.066 nmole of cyanide per unit of enzyme inactivated. This cyanide binding was linearly proportional to the amount of inactivation. No firm cyanide binding and no inactivation occurred in the absence of NADH or an equivalent reductant. The bound 14C was released when the enzyme was reactivated by ferricyanide. After the cells had been treated with ammonia for several hours prior to disruption, the crude cell extracts contained the nitrate reductase primarily in the inactive form. Critical examination of the methods of cell disruption suggested that the inactivation of the enzyme had truly occurred in vivo. After 300-fold purification, activation of this in vivo inactivated enzyme resulted in the release of 0.066 nmole of HCN per unit of enzyme activated. We conclude that the inactivation of nitrate reductase in vivo involves the formation of a firmly bound complex of reduced enzyme and cyanide. The reaction between reduced enzyme and HCN may be written: Er + HCN (ka)/⇄/(kd) Er-HCN The measured value for ka was 1.25 x 106 m-1 min-1, for kd, 4.5 x 10-4 min-1, giving a dissociation constant, Kd = 3.6 x 10-10 m.


BIRGIT VENNESLAND
From the Forschunysstelle Vennesland der Max-PlankGesellschaft, 1 Berlin 3.9, Germany SUMMARY In the presence of NADH and cyanide, NADH-nitrate oxidoreductase (EC 1.6.6.1) from ChZoreZZa vulgaris is converted to an inactive form which is readily reactivated by ferricyanide.
Experiments with H1*CN indicated that the inactivation process is associated with the firm binding to the protein of 0.066 nmole of cyanide per unit of enzyme inactivated.
This cyanide binding was linearly proportional to the amount of inactivation.
No firm cyanide binding and no inactivation occurred in the absence of NADH or an equivalent reductant. The bound 14C was released when the enzyme was reactivated by ferricyanide. After the cells had been treated with ammonia for several hours prior to disruption, the crude cell extracts contained the nitrate reductase primarily in the inactive form. Critical examination of the methods of cell disruption suggested that the inactivation of the enzyme had truly occurred in vivo. After 300-fold purification, activation of this in vivo inactivated enzyme resulted in the release of 0.066 nmole of HCN per unit of enzyme activated.
We conclude that the inactivation of nitrate reductase in vivo involves the formation of a firmly bound complex of reduced enzyme and cyanide.
The reaction between reduced enzyme and HCN may be written: k ER + HCN & E,-HCN ka The measured value for k, was 1.25 x lo6 M-' min-I, for kd, 4.5 x lop4 min-l, giving a dissociation constant, KD = 3.6 x lo-lo M.
In sonicated extracts of Chlorella vulgaris, grown with nitrate as the sole nitrogen source, NADH-nitrate reduct'ase (EC 1.6.6.1) was found initially in an inactive form, which was slowly activated in vitro in the presence of nitrate and phosphate (1, 2), and rapidly activated by ferricyanide (3). A reversible inactivation of crude preparations of the ferricyanide-activated enzyme could be achieved by the addition of either NADH or NADPH (4). The activity of the associated NADH-cytochrome c reductase was not affected by such treatments.
Thus, control of the activity of nitrate reductase appeared to be exercised by simple oxidation-reduction processesassociated with the nitrat,ereducing moiety of the enzyme Simultaneously and independently, Losada el al. made rather similar studies with Chlorella fusca (5-7). With this organism, special prior treatment of the cells was required to obtain extracts containing reversible inactive enzyme. Among these treatments was an incubation with ammonia (5), a procedure which may have been suggested by the results of earlier studies of Syrret and 1forris (8) 011 the effect of ammonia 011 nitrate assimilation by Chlorella. Subsequent purification of the nitrate reductase from C. vulgaris revealed that reduced pyridine nucleotides alone did not inactivate the enzyme. A second component, by far the most effective of which was cyanide, was required to achieve reversible inactivation (9). Kinetic studies showed that this inactivation reaction involved a stoichiometric reaction of cyanide with the NADH-reduced enzyme. It had already been shown that nitrate reductase, inactivated in this manner, could not be reactivated by removal of the inactivating reagents by gel filtration or dialysis (4, 10). It' could, however, be rapidly activated by ferricyanide.
Thus, the inactive cyanide complex, in this respect, resembled the inactive enzyme found in the algal extracts.
The above observations led us to examine the Chlorella cells and their sonicated extracts for the presence of cyanide.
Acid distillation of whole cell suspensions or sonicatcd extracts yielded sufficient cyanide to account for the presence of the inactivated enzyme in the crude extracts (11). It can bc argued, however, that cyanide may be an artifact produced by the acid distillation, or that the inactivation of the enzyme may occur only after the cell structures have been disrupted.
The present study addresses itself to these questions and establishes that the in &JO inactivated nitrate reductase contains bound cyanide.

MATERIALS AND METHODS
Growth of Chlorella-The cells were grown in continuous white light on a mineral salts medium, with nitrate as the only source of nitrogen, in a stream of 57, (v/v) CO, in air, at 2(f22", as previously described (1,12). For the present experiments, round bottom flasks containing 500 ml of medium were inoculated with 0.4 ml of cells per flask. After 24 hours, the yield, on the average, was about 2 ml of packed cells per flask. The cells were washed with >< volume of distilled water and susnended in water (250 ~1 of cells per ml) prior to disruption. Subsequent gel filtration of the incubation mixture revealed that radioactivity was incorporated into the fractions which co-eluted with the NADH-cytochrome c reductase activity (Fig. 1A).
When the incubation was carried out in the absence of KADH, there was a slight loss of activity (3.87,) and practically no incorporation of radioactivity into the protein fraction (Fig. 1B).
In the experiment of Fig. lA, 56.5 units of cytochrome c reductase, 8.7 units of inactive nitrate reductase, and 6.86 x lo4 dpm of 14C were recovered in the protein fractions.
From this data and a measured specific radioactivity of 1.2 x lo5 dpm per nmole-1 for the H14CN used, we calculate that 1 unit of inactive enzyme binds 0.066 nmole of HCN.
The same stoichiometry holds when the enzyme is only partially inactivated, as shown in Fig. 2. Intermediary levels of inactivation were achieved by incubation of the enzyme with limiting concentrations of NADH in the presence of excess H14CN, or with limiting concentrations of HVN in the presence of excess KADH. The binding of the HYX in the inactive enzyme complex is very firm. When 0.1 nmole of the inactive complex was incubated for 2 hours at 20" in the presence of a 500.fold molar excess of unlabeled HCN and NADH, only about IO y0 of the radioactivity exchanged out of the inactive enzyme (Fig. 3A).
In contrast to the slow release of H14CN from the inactive enzyme complex, all of the 14C was released within a few minutes upon activation of the enzyme by ferricyanide (Fig. 3B) As shown in Fig. 3A, there is a slow but measurable exchange reaction between H14CN of the E&*CN complex and unlabeled HCN. This reaction reflects the rate of dissociation of the ER-14CN complex since unlabeled HCN is present in great excess. In an experiment similar in design to that shown in Fig. 3A, ER-14CN was incubated at 20" with an excess of unlabeled HCN, and the loss in 14C from the ER-'~CN complex was determined as described in the legend for Fig. 4. As shown in Fig. 4, the dissociation of the ER-~~CN complex is a first order reaction. From the slope of this plot, a value of 4.50 x lo-4 min+ way calculated for the first order rate constant, kd, for the dissociation of the ER-CN complex. The second order rate constant, k,, for the reaction of HCN with the reduced enzyme is 1.25 X lo6 M-l min-l. This value was calculated from the data given in Fig. 3 of Ref. 9, on the assumption that one enzyme unit equals 0.066 nmole of enzyme. From these two values a dissociation constant, KD = kd/k, = 3.6 x lo-*O M, was calculated. At the end of the experiment shown in Fig. 4, there was a 70% loss in NADH-nitrate reductase activity but only a 12% loss in reduced methyl viologen-nitrate reductase activity, which indicates that the nitrate-reducing moiety remains largely catalytically active under these conditions. lt has previously been shown that regulation occurs on the nitrate-reducing moiety (9) It was, therefore, reasoned that if nitrate was added to the cell suspension immediately prior to cell disruption, then its presence should prevent any inactivation which might occur during or after the disruption process.
In earlier studies from this laboratory, water-washed cells were disrupted by sonication and consistently yielded extracts in which most of the enzyme was inactive.
We found, however, that when nitrate is added to the cell suspension immediately prior to sonication, the extracts contain a considerable proportion of the enzyme in the active form (Table II, Experiment  1). When the extracts are disrupted with a Ribi cell fractionator, on the other hand, the extracts contain more active enzyme than inactive enzyme, and addition of nitrate to the medium in which the cells are disrupted has less effect on the proportion of the enzyme present in the inactive form (Table II, Experiment 2). In two other similar experiments, there was no significant effect of added nitrate.
About 30% of the enzyme was inactive, on the average, in Ribi extracts of normal cells. To obtain Ribi extracts containing most of the nitrate reductase in the inactive form, we used an ammonia treatment similar to that which Losada et al. employed to obtain inactive nitrate reductase from Chlorella jusca (5) and Chlamyobmonas reinhardii (15,16).
Experiment 3 of Table II shows that addition of nitrate prior to cell disruption had no effect on the activation state of the enzyme from such cells. Because nitrate added prior to cell disruption did not decrease the amount of inactive enzyme found in the Ribi extracts of ammonium-treated cells, we conclude that enzyme inactivation had, in fact, occurred in tivo. This conclusion is supported also by the fact that these Ribi extracts do not cause rapid inactivation of added, active enzyme, as do the sonicated extracts (Fig. 5).
Presence of Bound Cyanide in in Vivo-inactivated Enzyme-The procedure developed for the purification of activated nitrate reductase can be applied equally well to the purification of the Starting with a Ribi extract of ammoniumtreated cells, two preparations were carried through a protamine sulfate precipit,ation, and one preparation was further purified by gel filtration and density gradient centrifugation in a zonal rotor to give a product estimated to be about 30% pure (XC under "Xaterials and Methods"). Aliquots of these preparations were partially activated by overnight incubation with added nitrate and phosphate buffer in the main compartment of Warburg vessels containing base in the center well. HCN (pK, = 9.15) is completely undissociated at neutral pH and will readily distill into the alkali trap. The use of ferricyanide as activator was avoided, since it is a potential source of cyanide. The enzyme preparation was assayed before and after the ac tivation process, and the contents of the center well were analyzed for cyanide.
The results are summarized in Table III. In all three experiments, the amount of cyanide obtained per unit of enzyme activated was virtually the same, 0.066 to 0.067 nmole per unit activated, in agreement with the previous meas; urements of the amount of Hi4CN bound during enzyme inactivation in vitro.

DISCUSSION
The reaction of HCN with the reduced form of nitrate reductase to give an inactive enzyme complex which can be reactivated by ferricyanide proceeds as shown in Equation 1. This was first surmised from Solomonson's demonstration that the inactivation of the reduced enzyme has the kinetics of a second order reaction between enzyme and HCN.
The present paper establishes the stoichiometry of the reaction. Pending an accurate determination of molecular weight by other methods, the results have been expressed as 0.066 nmole per unit of enzyme. From the firm binding of cyanide to. the reduced enzyme one may infer t'hat the concentration of HCN required for enzyme inactivation is several orders of magnitude smaller than t'hat which would inhibit other vital processes such as respiration. .-it such low concentrations, HCN is probably a rather specific inhibitor for nitrate reductase.
The present study also shows that the inactive nitrate reductasc present in Ribi extracts of ammonia-treated C'hZoreUa releases the same stoichiometric amount of HCN upon activation as the inactive labeled cyanide cornples formed in vitro. This establishes t'hat the inactive enzyrnc present in the extracts is indeed the En-CN complex.
Preliminary studies had shown that the inactive nitrate reductase of sonicated estracts also releases HCN on activation. These extracts contain much more HC?; than is necessary to inactivate the enzyme (II), and the possibility that the formation of HCN might be an artifact caused us some concern.
The Ribi cell fractionation procedure gives, on the average, more nitrate reduct,ase than the sonication process. Values above 20 units of total enzyme per ml of packed cells have been observed. Most important, with the Ribi procedure the activation state of the nitrate reductase is not strongly influenced by addition of &rate prior to cell disruption, so that the activation state of t'hc enzyme in the extracts most probably reflects its condition in vivo. Finally, the ammonia treatment was chosen to obtain in z&o-inactivated enzyme, because it is firmly established that ammonia plays a role as a regulator of nitrate reductase in intact algae (5,8).
At least two different mechanisms exist whereby ammonium controls the utilization of nitrate. One of these is associated with the repression of nitrate reductase synthesis by ammonium, an effect which requires many hours to become fully evident (18,19). The other mechanism is associated with the rapid cessation of nitrate assimilation upon the addition of ammoniurn to the culture medium.
Nitrate utilization does not resume until nearly all the ammonium is assimilated. Syrett and Morris have described this effect in detail (8). They showed further that ammonium did not inhibit the reduction of nitrate to ammonium by carbonstarved cells, which assimilated little ammonium.
They concluded that nitrate reduction was probably inhibited by products of ammonium assimilation.
Thacker and Syrett (20) have more recently examined the control of nitrate reduction by ammonium in Chlamydomonas reinhardii and come to similar conclusions.
It is tempting to speculate that the product of arnmonium assimilation which inhibits nitrate reduction is HCX.
\Ye have no evidence, however, that cyanide is synthesized in response to ammonium, except the fact that an increased quantity of cyanide is bound to nitrate reductase after the cells have been incubated with ammonium.
Nor do we yet understand the relationships between ammonium assimilation and cyanide synthesis, if such exist. It seems relevant, however, that the products of ammonium assimilation, the amino acids, are the precursors of the cyanogenic glucosides and cyanide, in all of the cyanogenic fungal, bacterial, and higher plant systems that have been examined to date (21-23).
A rapid, in vivo activation of i n &o-inactivated nitrate reductase, independent of de novo enzyme synthesis, has yet to be demonstrated unequivocally. Furthermore, the rapid i n vitro activation of the inactive enzyme has until now only been achieved with artificial oxidants.
The activation with nitrate is slow.
It is possible that some other substance is required in addition to nitrate.
The possibility that the inactivation is irreversible i n vivo has not been excluded. It might, for example, be a part of a mechanism involving the continual turnover of the enzyme.
We have touched on only a few of the problems raised by our results. Experiments are in progress to distinguish between 11. the many models to which the current findings can be fitted. rn