INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Nitrate reductase (NR) is regulated post-translation-ally by phosphorylation/dephosphorylation and binding of inhibitor proteins, members of the so called 14-3-3 family, to the phosphorylated form of NR (Huber et al., 1996; Moorhead et al., 1996; Su et al., 1996). NR is rapidly inactivated in this way in response to, for instance, a sudden lowering of the light intensity (Kaiser and Brendle-Behnisch, 1991; MacKintosh, 1992; Kojima et al., 1995). To detect this type of regulation, NR activity is assayed in the presence of Mg²⁺, because formation of the NR/14-3-3 complex requires a divalent cation (MacKintosh et al., 1995; Athwal et al., 1998). A deleted NR lacking 56 amino acids in the amino-terminal end (NR) was previously shown not to obey the usual post-translational dark-inactivation in vivo and was not inactivated by ATP in vitro (Nussaume et al., 1995; Lillo et al., 1997). NR was suggested not to be phosphorylated at the regulatory site, and/or not to bind the endogenous 14-3-3 proteins (Nussaume et al., 1995). NR affinity purified from dark-exposed tobacco leaves was, however, clearly inhibited by yeast 14-3-3, and the NR appeared to be phosphorylated and capable of binding yeast 14-3-3 proteins (Lillo et al., 1997). Six isoforms of 14-3-3 proteins have so far been identified in tobacco (Piotrowski and Oecking, 1998). Different 14-3-3 isoforms may have different functions in the cell (Bachmann et al., 1996a; Sehnke and Ferl, 1996), and it is possible that the NR is incapable of binding to the specific 14-3-3 that regulates NR activity. This would explain the fact that only modest inhibition of NR by Mg²⁺ was detected in the crude extracts, although partially purified NR was inhibited by yeast 14-3-3 in the presence of Mg²⁺ (Lillo et al., 1997). However, the work presented here shows that the NR is inhibited by endogenous 14-3-3 proteins. Binding of 14-3-3 proteins to NR has also been demonstrated by immunoprecipitation (Weiner and Kaiser, 1999). We have copurified NR and 14-3-3 proteins by Blue Sepharose affinity purification; and in the presence of phosphatase inhibitors and Mg²⁺, 14-3-3 proteins were found to co-purify with NR and NR based on activity assays and western blots. Under certain conditions, the stability of NR and NR is different (Nussaume et al., 1995). Lability of NR was further confirmed in the present work, and we suggest that stabilization of NR activity is an important function of the N-terminal domain of NR.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Inhibition of Purified NR and NR by Mg²⁺ and Endogenous 14-3-3

We partially purified NR from different sources in the presence of phosphatase inhibitors and in the presence or absence of Mg²⁺ (Table I). In agreement with previous results, the phosphorylation state is preserved when NR is partially purified in the presence of phosphatase inhibitors (Lillo et al., 1997). This was confirmed by the high Mg²⁺ inhibition always observed when yeast 14-3-3 was added to the assay buffer (Table I). Preparations made in the absence of 10 mM Mg²⁺ in all buffers were clearly different from preparations made in the presence of Mg²⁺. When NR was affinity purified in the absence of Mg²⁺, almost no difference was seen between assays with EDTA and assays with Mg²⁺, i.e. Mg²⁺ added to the assay gave only 0% to 6% inhibition (Table I). Apparently, when prepared in the absence of Mg²⁺, endogenous 14-3-3 was not retained in the preparation. However, when Mg²⁺ was present during purification, Mg²⁺ inhibition of partially purified NR was 41% to 66%, reflecting that endogenous 14-3-3 was present. This supports the assumption that one effect of Mg²⁺ is indeed to facilitate binding of 14-3-3 to phosphorylated NR, not just to inhibit activity of the NR/14-3-3 complex. Both tobacco NR (from C1 plants) and NR (from del7 plants) purified in the presence of Mg²⁺ were inhibited by Mg²⁺, showing that endogenous 14-3-3 copurified with NR as well as with NR (Table I). However, less Mg²⁺ inhibition was observed for NR compared with NR, indicating that less endogenous 14-3-3 was present in the NR preparation compared with the NR preparation. Purification of NR and NR in phosphate buffer confirmed that NR and NR were both phosphorylated in dark-exposed tobacco plants. Phosphate is known to inhibit phosphatases, and clearly after purifying tobacco NR and NR in phosphate buffer, strong inhibition by Mg²⁺ in the presence of yeast 14-3-3 was observed. However, Mg²⁺ alone inhibited NR activity by only 6% showing that endogenous 14-3-3 did not copurify with either NR or NR when prepared in this buffer (Table I).

View this table: [in this window] [in a new window]   Table I.   Inhibition of NR and NR activity by 5 mM free Mg2+ in the assay Tests were made with crude extracts and after Blue Sepharose affinity purification of NR from spinach, and tobacco C1 and NR from tobacco del7. Extracts were made from plants 30 min after transfer to darkness. After purification NR and NR activity was also tested in the presence of exogenous 14-3-3 (30 nM yeast 14-3-3 protein). Data presented are mean values from n preparations as indicated. SE is given.

During this work, we observed that Mg²⁺ inhibition of the NR activity decreased significantly after centrifugation of the crude extract. Activity was, therefore, tested immediately after extraction and again after centrifugation also for extracts of other plants. However, the decrease in Mg²⁺ inhibition during centrifugation was observed only for NR and only in HEPES buffer [4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid], not in phosphate buffer. We then wanted to determine whether centrifugation or time was crucial for the decrease in Mg²⁺ inhibition of the NR enzyme. A 1-min centrifugation in a microcentrifuge at 12,000g did not lead to more than a 4% decrease in Mg²⁺ inhibition. Furthermore, when part of the extract was left at 0°C, whereas the other part of the extract was centrifuged for 10 min at 30,000g, the decrease of Mg²⁺ inhibition was observed in both cases, i.e. time was the critical factor for activation of NR, not removal of debris by centrifugation. These results point to previous suggestions that some factor in the crude extract appears to activate NR, whereas NR is not activated by the same factor (Pigaglio et al., 1999).

Generally, only small differences in activity were seen between assays with and without Mg²⁺ when testing NR in crude extracts made with HEPES buffer and assayed after centrifugation (Table I; Nussaume et al., 1995; Lejay et al., 1997). To determine whether this difference was due to kinetic characteristics, we tested the concentration of cations necessary for half maximal inhibition of NR and NR partially purified from dark-exposed plants in the presence of phosphatase inhibitors (phosphorylated enzyme). Cations which could be of importance in addition to Mg²⁺ in vivo include the polyamine spermidine, which is known to reach fairly high concentrations in plant tissue (0.3 mM) and is involved in regulation of various processes (Flores and Galston, 1984). Although highly speculative, inhibition by spermidine could represent a possible negative feedback mechanism in NO₃ assimilation by a metabolite rich in nitrogen. The concentrations of Mg²⁺ and spermidine were varied in the range 0 to 5 mM and 0 to 1 mM, respectively, in the presence, as well as absence, of yeast 14-3-3. Mg²⁺ and spermidine did not inhibit NR or NR at these concentrations unless 14-3-3 was present. In the presence of 14-3-3, 1 mM Mg²⁺ or 0.2 mM spermidine gave half maximal inhibition of NR and NR. No difference in IC₅₀ values was seen for NR and NR (Table II). The maximal inhibition of NR and NR activity varied between 70% and 80% in these experiments. At higher concentrations of Mg²⁺ (10 mM) and spermidine (2 mM), some inhibition of NR activity was detected, even in the absence of 14-3-3. The same IC₅₀ values were found for Mg²⁺ and for spermidine in the presence of endogenous tobacco 14-3-3. However, no difference was found between NR and NR with respect to IC₅₀ values.

View this table: [in this window] [in a new window]   Table II.   Concentration of Mg2+ or spermidine necessary for half-maximal inhibition (IC50-values) of purified NR and NR activity in the presence of 30 nM yeast 14-3-3  Three preparations from C1 and del7 plants harvested after 30 min of darkness and prepared in the presence of phosphatase inhibitors were tested. SE is given.

Immunoblot Analysis

Western blots confirmed that in the presence of phosphatase inhibitors and Mg²⁺, 14-3-3 proteins copurified with NR. This was found for both spinach and squash (blots not shown), as well as for full-length and deleted tobacco NR (Fig. 1). Lanes 1 and 3 show that 14-3-3 proteins were present in preparations of Blue Sepharose-purified NR and NR when Mg²⁺ was present in all buffers during purification, whereas 14-3-3 proteins were not detectable when Mg²⁺ was omitted (lanes 2 and 4). The NR/14-3-3 complex was eluted from Blue Sepharose with 0.3 M KNO₃ (lanes 1 and 3) or 100 µM NADH (lanes 5 and 6).

View larger version (45K): [in this window] [in a new window]   Figure 1.   Protein-gel blots showing NR, NR, and 14-3-3 proteins after Blue Sepharose chromatography of extracts from C1, del7 (d7), and del8 (d8) plants. Affinity-purified NR and NR were subjected to SDS-PAGE. NR purified in the presence (lane 1) and absence of Mg2+ (lane 2). NR (d7) purified in the presence (lane 3) and absence of Mg2+ (lane 4). Lanes 5 and 6 show NR from del7 and del8 purified in the presence of Mg2+, and a different assay for detecting 14-3-3 protein was used (see "Materials and Methods"). Samples tested in lanes 1 through 4 were eluted with KNO3 from the Blue Sepharose column and concentrated on Centricon 30. Samples tested in lanes 5 and 6 were eluted with NADH from the Blue Sepharose column and concentrated on Centricon 100. Amounts of protein added to wells 1 through 4 were approximately 6 µg and for wells 5 and 6, approximately 2 µg. Following transfer of the proteins to a nitrocellulose membrane, the membrane was cut in two halves. Top, Upper half of the membrane was probed with antiserum raised against squash NR. Bottom, Lower half of the membrane was probed with antiserum raised against spinach 14-3-3 proteins.

For all species tested, Mg²⁺ in the extraction buffer and buffers used during purification was necessary for detection of 14-3-3 protein on the blot. As a control, a crude (spinach) extract was passed through an antibody column, Mac74 (MacKintosh et al., 1995), which removed 90% of NR activity. This extract, low in NR, was then passed through a Blue Sepharose column, and proteins were eluted with 0.3 M KNO₃. Only negligible amounts of 14-3-3 bound to the column. The experiment confirmed that 14-3-3 elution from the Blue Sepharose column is indeed dependent on the presence of NR.

The results shown in Table I and blots shown in Figure 1 strongly support the assumption that endogenous 14-3-3 proteins bind to NR as well as NR. Densitometer measurements of four different preparations of NR and NR on various blots indicated that, to some degree, 14-3-3 bands were weaker for del7 extracts compared with C1 extracts. However, considerable variations among extracts prevented strict conclusions.

Partial Activities

Different partial activities of affinity-purified tobacco NR and NR were tested with dephosphorylated enzymes (leaves harvested in the light, no phosphatase inhibitors present in extraction buffers) and phosphorylated enzymes (leaves harvested after 30 min of darkness, phosphatase inhibitors in extraction buffers). As expected, dephosphorylated NR and NR activities were not inhibited by Mg²⁺. Like the partial activity of phosphorylated spinach NR (Bachmann et al., 1996b), the reduced methylviologen (MV) to NR partial activity of tobacco NR was inhibited by Mg²⁺ when leaves were harvested from darkness. Also the MV to NR activity was inhibited by Mg²⁺ (Table III). However, NR and NR behaved differently, because for NR the NADH to NADH-cytochrome c reductase partial activity (CR) was inhibited by Mg²⁺. This surprising result was confirmed with four independent preparations of purified NR and NR. Inhibition of NADH to CR by Mg²⁺ was dependent on the phosphorylation state of NR; inhibition was only seen for enzyme prepared from dark-treated leaves (Table III). NADH to ferricyanide-NR partial activity of NR and NR was not inhibited by Mg²⁺ (Table III). The NADH to ferricyanide-NR partial activity involves only the FAD-binding domain, whereas NADH to CR involves also the heme-binding domain (Rouzé and Caboche, 1992). The binding site for 14-3-3 proteins is at the hinge between the heme-binding and Mo-cofactor binding domain (Bachmann et al., 1996b; Moorhead et al., 1996), and it may, therefore, be expected that the heme-binding domain, but not the FAD-binding domain, is influenced by Mg²⁺ in combination with 14-3-3. The surprising result is that the deletion in the N-terminal region of NR is necessary to see any effect of 14-3-3 and Mg²⁺ on the partial NADH to CR activity. The results show that NR and NR have some different characteristics with respect to 14-3-3. 

View this table: [in this window] [in a new window]   Table III.   Inhibition of partial activities of NR and NR by 5 mM Mg2+ in the presence of 30 nM yeast 14-3-3 protein Assays were run in triplicate, and SE  5%.

Labile NR Activity

When using NADH as an electron donor and NO₃ as an acceptor, electron transfer includes all three cofactor binding domains of NR (FAD-, heme-, and molybdopterin-binding domains), whereas Cytochrome c (cyt c) as substrate needs only active FAD- and heme-binding domains. The ratio of CR to NR and NR activities varied between preparations, but was always higher for NR. An average of six different preparations after affinity purification gave a ratio of 3 for NR and 7 for NR (Table IV). The number of units of purified NR and NR was clearly related to the number of units in the crude extracts (Table IV), and the yield varied from 34% to 48%. The NR activity was approximately three times higher in crude extracts from C1 plants compared with del7 plants (data derived from Table IV). However, CR activity was similar in preparations from del7 and C1 plants. This shows that del7 plants contained a high proportion of inactive NR, i.e. only the partial CR activity was functional. Alternatively, the NR was immediately inactivated when extracted and assayed. Clearly, NR is more labile than NR in vitro. For instance, (NH₄)₂SO₄ precipitation of NR resulted in unstable enzyme (Nussaume et al., 1995). Following Blue Sepharose purification, NR activity was unstable if not immediately desalted. For instance, after incubation of the enzyme at 25°C for 60 min, 80% of the NR activity was lost, whereas NR activity was unchanged under these conditions (Fig. 2). Inactivation of NR was not caused by proteolysis because a western blot was performed with the NR sample at time 0, 60, and 120 min of incubation at 25°C, and showed no changes in NR protein (blot not shown). CR was stable during such experiments, which confirms that inactivation of activity was not due to proteolysis, but rather, caused by inactivation of the Mo-containing domain. NR and NR were eluted with a high salt concentration (0.3 M KNO₃), and after desalting, both NR and NR were stable during 1 h at 25°C (Fig. 2). However, attempts to ascribe the lability of NR to a salt effect were not successful because, although NR was much more stable after desalting of the Blue Sepharose fractions, adding back KNO₃ or other salts (KCl, K₂SO₄) up to 0.3 M did not render the enzyme unstable. Apparently, NR is unstable after the conformation change that probably takes place during elution from Blue Sepharose. However, this behavior is still not understood.

View this table: [in this window] [in a new window]   Table IV.   Total activity in units (nmol NO2 formed per NADH consumed or Cyt c reduced per min) in crude extract and affinity purified NR and NR preparations using 10 g of leaf tissue There were n repeats for each type of extraction buffer and species. CR was measured in three of the preparations in each case. SE is given.

View larger version (17K): [in this window] [in a new window]   Figure 2.   Stability of NR and NR in fractions of freshly prepared affinity-purified enzyme eluted with 0.3 M KNO3. NR not desalted (), NR not desalted (), NR desalted (), and 0.3 M KNO3 was added to desalted NR (). Storage temperature was 25°C. The experiment was repeated three times with different enzyme preparations. Data presented represent () NR prepared in phosphate buffer and in the absence of Mg2+, however, the same results were obtained for preparations made in HEPES buffer in the presence or absence of Mg2+ and phosphatase inhibitors. SE is indicated when exceeding the size of the symbol.

Tobacco NR and NR were routinely eluted with NO₃ from the Blue Sepharose column because NR activity was hardly detectable after elution with NADH. However, CR was still present in these NADH-eluted preparations. Following NADH elution, the ratio of CR to NR was 2 and the ratio of CR to NR was 20. Further tests of NADH-eluted NR showed that the very terminal activity, i.e. the reduced bromphenol blue to NR that involves the molybdopterin cofactor binding domain only (Rouzé and Caboche, 1992), was also inactivated (data not shown). Reversible inactivation of NR by NADH has previously been reported for Chlorella and wheat and was shown to depend on FAD and formation of superoxide, and involve the molybdopterin-binding domain (Moreno et al., 1972; Aryan and Wallace, 1985). Attempts to regain activity by incubation with an oxidizing agent like ferricyanide, 0.5 mM, as described by Aryan and Wallace (1985), increased NR activity slightly. However, NR activity was still very low compared with NO₃ eluted NR or NR. Since most of the NR protein was still inactive with respect to the terminal activity after treatment with ferricyanide, this argues against (reversible) inactivation by superoxide, but supports the assumption that the molybdopterin cofactor is released or non-functional (Nussaume et al., 1995).

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Following purification of phosphorylated NR and NR, IC₅₀ values for yeast 14-3-3 (Lillo et al., 1997), Mg²⁺, and spermidine (present work) were the same for NR and NR. No differences of kinetic constants were found that would explain the modest Mg²⁺-inhibition (Nussaume et al., 1995; present paper) often found for NR in crude extracts.

Following centrifugation, i.e. 15 min after extraction, Mg²⁺ inhibition of NR in HEPES buffer was only about 25% (Table I); in agreement with values found previously under similar extraction and assay conditions (Nussaume et al., 1995). However, in the present work, we also show that when NR activity was tested in extracts immediately after homogenization of leaves harvested from darkness, stronger Mg²⁺ inhibition was observed (45%), but still not as strong as for NR (81%). The decrease in Mg²⁺ inhibition after extraction was not observed for NR. This implies that modest Mg²⁺ inhibition of NR in crude extracts may be caused by reactions taking place both in the plant and after extraction. When extracts were made in phosphate buffer, less inhibition of NR was observed compared with HEPES buffer; Mg²⁺ inhibition was almost the same for NR and NR (Table I). High concentration of phosphate is likely to promote release of 14-3-3 protein from NR and may therefore result in NR appearing more similar to NR. Phosphate is known to stimulate NR activity by binding to the molybdopterin-binding domain (Solomonson and Barber, 1990) and, possibly, phosphate stabilizes NR and also makes NR and NR more similar in activity assays.

The fact that the NR protein was inactivated by yeast 14-3-3 after purification led us to propose previously a model where a putative activating factor would remain bound to the NR protein in crude extracts and would impede 14-3-3 binding (Pigaglio et al., 1999). We show here that endogenous 14-3-3 copurified with NR through the purification process, as shown by activity measurements in the presence of Mg²⁺ (Table I), as well as western blots (Fig. 1). Thus, our previous model appears to be no longer valid, as the NR protein seems to bind 14-3-3 proteins to almost the same extent as NR. An explanation for this observation could be that 14-3-3 proteins bind the NR protein to a different site and/or in a different manner. Although no alternative binding site for 14-3-3 proteins has so far been described for NR, other 14-3-3 interacting proteins, like the Raf kinase, present multiple sites of interaction with 14-3-3 (Tzivion et al., 1998). To reconcile our previous model with the present data, we propose that the factor which remains bound to the NR protein is in fact the interacting 14-3-3 protein(s). In this case, 14-3-3 binding would inactivate NR less, either because the 14-3-3 proteins are also bound to another site on NR or because the NR/14-3-3 complex adopts a different conformation due to the removal of the NR N-terminal region. This hypothesis is supported by the fact that 14-3-3 proteins inhibited CR partial activity to a different extent in NR and NR proteins (Table III).

Clearly, the deletion in the N-terminal domain influenced characteristics of the NR/14-3-3 complex because electron transport in the heme-binding domain of NR was inhibited by Mg²⁺/14-3-3, whereas this was not the case for NR (Table III). A striking effect of the deletion was the influence on stability (Fig. 2; Table IV); NADH to CR activity was retained, but the terminal activity (associated with molybdopterin-binding domain) was inactivated. Generally, when NADH and NO₃ were used as substrates, del7 plants had only 30% of NR activity, compared to C1 plants. However, when the partial CR activity was tested, activities in C1 and del7 plants were similar (Table IV). This shows that, for the NR, a large proportion of the terminal activity was inactive or less active than for NR. This deactivation took place during extraction or also in the plant. It is therefore likely that an important function of the N-terminal domain is stabilization of the molybdo-pterin-binding domain.

The question that still remains open is the actual activation state of the NR protein in the plant. If regulation of NR is seriously impaired in vivo, accumulation of nitrite in the leaves in darkness would be expected (Riens and Heldt, 1992). Accumulation of nitrite was, however, never observed in del7 plants, since extracts did not have any background level of nitrite. Moreover, as reported by Lejay et al. (1997), almost no stimulation of NO₃ reduction in the dark was detected in del7 plants compared with control plants, despite of the high activation state of the NR protein. Since accumulation of nitrite could have serious negative effects on plants, redundant strategies may have evolved to assure that this toxic compound is not allowed to accumulate. It has been proposed that the limited availability of reducing power in darkness would prevent or at least decrease any effect of the NR higher-activation state. Indeed, when the del7 and del8 plants were grown in the light in very low CO₂ concentrations, where reducing power is probably not limiting, a higher NO₃ reduction rate was observed (Lejay et al., 1997). Other evidence also supports a possible higher activation state of the NR protein in vivo. It has been proposed that the stability of NR is decreased upon inactivation (Weiner and Kaiser, 1999), and we have previously shown that the NR protein is more stable in darkened plants than the NR protein (Pigaglio et al., 1999).

Taken together, these results indicate that the NR protein is probably less inactivated by 14-3-3 proteins than NR in planta and in vitro, although 14-3-3 proteins seem to remain bound to it, and that the Mg²⁺ inhibition of this protein is decreasing with time after extraction. Thus, we propose that the absence of the NR N-terminal domain affects in some way the enzymatic activities of the NR/14-3-3 complex.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Plant Material

Plants tested were: squash (Cucurbita maxima L. cv Buttercup), spinach (Spinacea oleracea L. cv Merkur), Nicotiana plumbaginifolia var. Viviani C1 with the full-length tobacco NR (Vincentz and Caboche, 1991), and N. plumbaginifolia del7 and del8 with a NR lacking 56 amino acids in the N-terminal domain (Nussaume et al., 1995). del7 and del8 are independent transformants carrying the same transgene, and there was no difference in the level of NR expression between them. NR was constitutively expressed using the cauliflower mosaic virus 35 S promoter in the Nicotiana. All the NR activity detected in these transgenic plants is derived from the transgene expression, as the endogenous NR gene is inactivated by a retrotransposon insertion. Plants were grown at 20°C with a 12-h photoperiod at 80 µmol m² s¹, and were watered with Hoagland solution containing 15 mM KNO₃ three times a week and always the day before harvesting (Lillo, 1994).

Extraction and Purification of NR

Leaves, 10 g, were harvested in the morning, 2 h after the onset of light, followed by 30 min of darkness. Samples were homogenized with 30 mL of 0.1 M HEPES-KOH (pH 7.5), 10 mM MgCl₂, 3% (w/v) polyvinylpolypyrrolidone, 1 mM EDTA, 7 mM Cys, 7.5 µM leupeptin, 0.1 mM PMSF, 1 µM FAD, 1 µM Na₂MoO₄, 25 mM NaF, and 0.1 µM okadaic acid (stock solution made up in dimethylsulfoxide, Calbiochem, San Deigo), filtered through cheesecloth, and centrifuged for 10 min at 31,000g. In some cases, MgCl₂ was omitted, see "Results." The supernatant was mixed with 3 g of Blue Sepharose equilibrated in homogenization buffer (omitting polyvinylpolypyrrolidone), and shaken for 40 min, then washed with 100 mL of one-half-strength extraction buffer (omitting polyvinylpolypyrrolidone), packed into a syringe, and NR was then eluted with 100 µM NADH or 0.3 M KNO₃ (in case of tobacco) in the one-half-strength extraction buffer. The most active Blue Sepharose fractions were desalted on a Sephadex G-25 PD-10 column (Pharmacia Biotech, Piscataway, NJ) with 25 mM HEPES (pH 7.5), 7 mM Cys, 0.2 mM EDTA, and 0.1 µM okadaic acid, and concentrated in centricon 30 (Millipore, Bedford, MA), or, when mentioned, in centricon 100. For tobacco, NaF strongly inhibited NR activity in the presence of HEPES buffer and had to be avoided. When NaF was included, potassium phosphate buffer was used, otherwise additions were as for HEPES buffer. For efficient binding of tobacco NR to Blue Sepharose, the crude extract was first desalted on Sephadex G-25 columns equilibrated with one-half-strength extraction buffer. Recovery was about 30%, and the specific NADH to NR activity was 10 µmol NO₂ min¹ mg¹ protein for spinach and squash, and 2 µmol NO₂ min¹ mg¹ protein for tobacco NR and NR. Protein was measured with a protein assay (Bio-Rad, Hercules, CA) using -globulin as a standard. NR was purified 1,000-fold when eluted with NADH (as for spinach and squash) and about 200-fold when eluted with 0.3 M KNO₃ (tobacco).

Assay of NR

The assay mixture contained 50 mM HEPES-KOH (pH 7.5), 100 µM NADH, and 5 mM KNO₃ with 2 mM EDTA or 5 mM MgCl₂. The assay volume was 0.75 mL. Activity was measured in crude extracts by determining NO₂ formation, and, in (partially) purified NR preparation also, by continuously following the decrease of NADH at 340 nm (Lillo and Ruoff, 1992). NADH oxidation in the absence of KNO₃ was negligible.

To measure MV to NR, NADH was substituted by 100 µM MV and 0.8 mg Na₂S₂O₄. To measure NADH to CR, KNO₃ was substituted by 0.5 mg Cyt c. Controls with addition of an anti-NR monoclonal antibody (96[9]25), which inhibits CR activity, assured that the CR measured was indeed linked to NR (Chérel et al., 1990). To measure NADH to ferricyanide reductase, KNO₃ was substituted by 0.5 mM potassium ferricyanide. (for details on assays, see Wray and Fido, 1990). One unit of enzyme activity is defined as the rate of 1 µmol of nitrite produced per NADH reduced or 1 µmol Cyt c reduced in 1 min. All assays were performed at 25°C. NADH to NR, MV to NR, and reduced bromphenol blue to NR were assayed by incubating for 5 min and measuring the nitrite produced as the diazo-compound formed from sulfanilamide (Lillo, 1983). NADH to CR was measured by continuously following the production of reduced Cyt c in a spectrophotometric cell at 550 nm.

Purification of Yeast 14-3-3 Proteins

The 6-His tagged BMH1 14-3-3 isoform from Saccharomyces cerevisiae was expressed in Escherichia coli DH5 from the trc promoter in plasmid pTrcHisA, and purified to homogeneity (Moorhead et al., 1996).

Immunoblot Analysis

Proteins were separated using a discontinuous buffer system SDS-PAGE (12% gels) electrophoresis, according to the method of Laemmli (1970). Protein preparations, 2 to 6 µg, were applied to each well. Proteins were transferred to a nitrocellulose membrane (0.45 µm) using a mini trans-blot electrophoretic cell (Bio-Rad). The transfer buffer used was 25 mM Tris, 192 mM Gly, and 20% (v/v) methanol, pH 8.3. The upper section of the membrane was incubated with rabbit polyclonal antibodies raised against purified squash NR, and the lower section was treated with polyclonal antibodies raised against spinach 14-3-3 in sheep. The immunocomplexes were recognized using protein G conjugated to horseradish peroxidase. The membranes were developed using a color development method (Fig. 1, lanes 1-4; Bio-Rad) and the ECL chemiluminescent detection method (Fig. 1, lanes 5 and 6; Amersham Pharmacia, Uppsala). The 14-3-3 antibodies were kindly provided by Drs. Moorhead and MacKintosh (Department of Biochemistry, University of Dundee, Dundee, UK).