analysis of wild-type and mutant plant nitrate reductase expressed

Plant Physiol. (1 997) 11 5: 11 35-1 143

Analysis of Wild-Type and Mutant Plant Nitrate Reductase Expressed in the Methylotrophic Yeast Picha pastoris’

Wenpei Su, Jeffrey A. Mertens, Kengo Kanamaru, Wilbur H. Campbell, and Nigel M. Crawford*

Department of Biology, University of California, San Diego, La Jolla, California 92093-01 16 (W.S., K.K., N.M.C.); and Phytotechnology Research Center and Department of Biological Sciences, Michigan Technological

University, Houghton, Michigan 49931-1 295 (J.A.M., W.H.C.)

Recombinant Arabidopsis thaliana NADH:nitrate reductase (NR; EC 1.6.6.1) was produced in the methylotrophic yeast Pichia pastoris and purified to near-electrophoretic homogeneity. Purified enzyme had the spectral and kinetic properties typical of highly purified NR from natural plant sources. Site-directed mutagenesis altering several key residues and regions was carried out, and the mutant enzyme forms were expressed in P. pastoris. When the invariant cysteine residue, cysteine-191, in the molybdo-pterin region of the A. thaliana NIA2 protein was replaced with serine or alanine, the NR protein was still produced but was inactive, showing that this residue is essential for enzyme activity. Deletions or substitutions of the conserved N terminus of NR retained activity and the ability to be inactivated in vitro when incubated with ATP. Enzyme with a histidine sequence appended to the N terminus was still active and was easily purified using metal-chelate affinity chromatography. These results demonstrate that P. pastoris is a useful and reliable system for producing recombinant holo-NR from plants.

Assimilatory NR (EC 1.6.6.1-3) catalyzes the reduction of nitrate to nitrite, the first committed step in the nitrate assimilation pathway (Solomonson and Barber, 1990; Hoff et al., 1994; Crawford, 1995). NR is a large, metalloflavoen- zyme that exists as a homodimer with a monomer mass of approximately 100 kD. Each subunit consists of three functional regions or domains, with each region associated with a redox center or cofactor (FAD, heme-Fe, and MoCo) that transfers electrons from NAD(P)H to nitrate (Fig. 1; Camp- bell and Kinghorn, 1990; Solomonson and Barber, 1990; Crawford, 1995, 1996). Electron flow is from NAD(P)H to FAD to heme-Fe to MoCo to nitrate (Campbell and King- horn, 1990; Solomonson and Barber, 1990; Campbell, 1996). Artificial electron donors (such as methyl viologen or bromphenol blue) or acceptors (such as Cyt c) can be used to assay several nonphysiological partia1 activities of NR (Solomonson and Barber, 1990). Extensive amino acid sequence similarities exist between each of the three functional regions of NR and other redox enzymes and proteins (Campbell and Kinghorn, 1990; Solomonson and Barber,

This work was supported by the National Institutes of Health (grant no. GM40672 to N.M.C.) and the National Science Founda- tion (grant no. MCB-9420313 to W.H.C.).

* Corresponding author; e-mail [email protected]; fax 1- 619-534-7108.

1135

1990; Hoff et al., 1994; Crawford, 1995). The MoCo-binding region of NR is similar to the MoCo-binding region of sulfite oxidase; the heme-Fe-binding domain is similar to proteins in the Cyt b, superfamily and the Cyt b reductase region, which contains the FAD-binding domain, is most similar to NADH:Cyt b, reductase (Calza et al., 1987; Craw- ford et al., 1988; Campbell, 1996). The Cyt b reductase fragment of NR has been shown by structural analysis to be a member of the Fd NADP+ reductase family of flavoen- zymes (Lu et al., 1994; Campbell, 1996).

Although the biochemistry of NR has been studied for more than four decades, the complete three-dimensional structure is not known, and mechanistic details of enzyme catalysis are lacking because of the limited quantities of pure NR that are available (Campbell, 1996). Until recently, the study of mutant NR forms has been restricted to the analysis of a few plant mutants (Warner and Kleinhofs, 1992; Crawford and Arst, 1993; Hoff et al., 1994). One approach to overcome these limitations is to express NR in a heterologous system. Toward this end, various fragments of NR have been expressed in Esckerickia coli, including the Cyt b reductase, the heme-binding domain, and the com- bination of these two, known as the Cyt c reductase fragment (Hyde and Campbell, 1990; Campbell, 1991; Cannons et al., 1993; Dwivedi et al., 1994; Gonzalez et al., 1995; Quinn et al., 1996). The Cyt c reductase fragment of spinach (Spinacia oleracea L.) NR has also been produced in the methylotrophic yeast Pichia pastoris at much higher levels than could be attained in E . coli (Shiraishi and Campbell, 1997). Biochemical characterization and site-directed mutagenesis were performed on these recombinant proteins. The Cyt b reductase fragment of corn NR has been crystal- lized, its three-dimensional structure determined, and the mechanism of electron transfer analyzed (Lu et al., 1995; Ratnam et al., 1995; Campbell, 1996). Expression of the MoCo-binding fragment of NR alone has not been reported, but sulfite oxidase, which shows striking similarity to the MoCo and heme domains of NR, has been expressed in E. coli (Garrett and Rajagopalan, 1994, 1996). Crystal structures of other MoCo enzymes such as DMSO reductase have been described (Schindelin et al., 1996), but these enzymes have no sequence similarity to NR or sulfite oxidase.

Abbreviations: MoCo, molybdenum molybdo-pterin cofactor; NR, nitrate reductase.

1136 Su et al. Plant Physiol. Vol. 11 5, 1997

An important feature of NR is that it is regulated by a posttranslational mechanism that rapidly and reversibly inhibits the enzyme in response to severa1 signals, including light/dark transitions (Kaiser and Huber, 1994). This mechanism involves phosphorylation of a Ser in the hinge 1 region of NR (Ser-543 in spinach and Ser-534 in Arabi- dopsis) and interaction with an inactivator protein (14-3-3 protein) in the presence of magnesium (Douglas et al., 1995; Bachmann et al., 1996a, 1996b; Huber et al., 1996; Moorhead et al., 1996; Su et al., 1996). NR is reactivated by dephosphorylation by type 2A protein phosphatase (Mack- intosh, 1992). Phosphopeptide analysis of NR has revealed that multiple Ser residues are phosphorylated in NR (Hu- ber et al., 1992; Labrie and Crawford, 1994), but only the phosphorylation of Ser-543 in spinach correlates with the loss of NR activity (Huber et al., 1992; Douglas et al., 1995; Bachmann et al., 1996b). The production of recombinant fragments of spinach NR as targets for the spinach NR protein kinase was instrumental in leading to the identification of hinge 1 as a key regulatory region of NR (Bach- mann et al., 1996b).

Two recent findings have opened up new avenues for the efficient site-directed mutagenesis and structural/ functional characterization of holo-NR. In fungi, mutant forms of the species Aspergillus NADPH:NR have been expressed in an niaD deletion mutant (Garde et al., 1995). In this study a Cys (Cys-150) in the MoCo-binding region and a His (His-547) in the heme-Fe domain were identified as essen- tia1 residues for funga1 NR activity. Cys-150 (equivalent to Cys-191 in Arabidopsis NR) corresponds to an invariant Cys residue among all NR and sulfite oxidase sequences, is part of the signature sequence CAGNRR for this group of enzymes (Barber and Neame, 1990; Garrett and Rajago- palan, 1994), and has been shown to be a ligand to the molybdenum in rat liver sulfite oxidase (Garrett and Ra- jagopalan, 1996). His-547 is thought to provide an axial ligand to the heme-Fe domain (Meyer et al., 1991; Garde et al., 1995). For plants, a functional NADH:NR has been expressed in P. pastoris (Su et al., 1996). Arabidopsis Nia2 NR was expressed under the control of the P. pastoris alcohol oxidase promoter, which is inducible by methanol. Mutant forms of NR with substitutions of conserved Ser residues were produced in the variable sequence region between the MoCo-binding region and heme-Fe domains, known as hinge 1. Analysis of these mutants showed that Ser-534 is required for the phosphorylation-dependent inhibition of NR in vitro. This Ser corresponds to Ser-543 in spinach NR, which was identified as the target for spinach NR kinase (Douglas et al., 1995; Bachmann et al., 1996b).

The expression of functional holo-NR and the high leve1 of expression of Cyt c reductase fragments indicate that P. pastoris provides an excellent system for further studies of the structure and function of plant NR (Su et al., 1996; Shiraishi and Campbell, 1997). In this study we demonstrated that plant NR purified from P. pastoris i5 similar in its biochemical properties to natural forms isolated from plant sources. We also carried out site-directed mutagenesis to show that the invariant Cys-191 of plant NR is essential for enzyme functionality and that the conserved amino acids at the N terminus are not essential. Further-

more, we showed that the N-terminal amino acids (residues 1-11) are not required for the ATP-dependent inhibition of NR and can be replaced with a hexahistidine sequence, which facilitates purification of NR for in vitro regulatory studies.

MATERIALS A N D METHODS

Recombinant Arabidopsis NADH:NR Production and Purification

NR was expressed in the methylotrophic yeast Pickia pastoris strain GS115 from the Arabidopsis NIA2 NR cDNA pAtc-46 (Crawford et al., 1988) under the control of the P. pastoris alcohol oxidase 1 promoter as previously described (Su et al., 1996). P. pastoris was initially grown in 500-mL cultures containing YPD media (1% yeast extract, 2% pep- tone, and 2% Glc, a11 w / v ) to an A,, of 10. The log-phase cells were washed and then resuspended to an A600 of 5 in 1 L of modified MM media (2.8% [w / V I yeast nitrogen base with ammonium sulfate, 100 mM potassium phosphate, pH 6.0, 4 X 10p5% biotin, 0.2 mM sodium molybdate, and 1% methanol [v/v]) to induce NR expression. Methanol (l'h [v /VI) and sodium molybdate (0.2 mM) were added to the culture every 24 h. After a 72-h induction period the cells were collected by centrifugation, washed, and then resuspended in extraction buffer containing 50 mM sodium phosphate, pH 7.3, 1 mM EDTA, and 5% (v/v) glycerol to an A600 of 50 to 100. The cells were broken with a Bead Beater (Biospec Products, Bartlesville, OK) using 0.5-pm glass beads by processing for 5 min using 30-s bursts followed by 1-min cooling periods. The crude extract was then centrifuged at 15,0008 for 20 min at 4°C.

NR assays of crude extracts and purified samples were performed as previously described (Campbell and Smarrelli, 1978; Redinbaugh and Campbell, 1985). One unit of NR activity is defined as 1 pmol of nitrite produced/ min. For the purification of wild-type NR, crude cell extracts were subjected to a 30 to 45% ammonium sulfate precipitation. The resulting pellet was dissolved in 50 mM sodium phosphate, pH 7.3, and 1 mM EDTA (buffer A) and batch-bound to Blue-Sepharose (Campbell and Smarrelli, 1978). The Blue-Sepharose with bound NR was collected by vacuum filtration, thoroughly washed with buffer A, and packed in a 1.5-cm column. The column was eluted with 0.1 mM NADH in buffer A. Fractions containing NR activity were pooled, buffer exchanged, and concentrated with 25 mM Mops (ultrapure grade, Calbiochem), pH 7.0, using a Centriprep 30 concentrator (Amicon, Beverly, MA). Next, the partially purified sample was applied to a 1.5-mL column of 5'-AMP Sepharose (Pharmacia). The column was washed with 5 column volumes of 50 mM Mops, pH 7.0, and 1 mM EDTA (buffer B). NR was then eluted with 0.1 mM NADH in buffer B. Fractions containing more than 2 units/mL of NR activity were pooled and concentrated, with the buffer exchanged to 25 mM Mops, pH 7.0.

NR Kinetics and Spectra

NR samples purified as described above were used. Ki- netic experiments were performed at 25'C in 1 mL of buffer

Recombinant Arabidopsis Nitrate Reductase 1137

containing 50 mM Mops, pH 7.0, by following the decrease in A,,, using a 1-cm cuvette and a UV/Vis spectrophotom- eter (model no. 1201, Shimadzu, Kyoto, Japan). The concentration of the substrates varied from 0.5 to 20 p~ for NADH (Sigma) and 2.5 to 60 p~ for potassium nitrate. The results of three initial velocity experiments were analyzed using Enzpack software (Elsevier Biosoft, Cambridge, UK). Spectra of oxidized and NADH-reduced samples were determined at 25°C using a UV/Vis diode array spectropho- tometer (model no. 8453, Hewlett-Packard).

Site-Directed Mutagenesis

Site-directed mutagenesis was performed as previously described (Higuchi, 1990) using the Arabidopsis NR cDNA clone pAtc-46 (Crawford et al., 1988). For the point mutations, a 1.8-kb SmaI fragment from pAtc-46 was subcloned into the Bluescript (KS-) vector (Stratagene) to generate a template for PCR. In the first round of PCR T7 and T3 primers were used as outside primers, whereas inside primers were oligonucleotides with designed mutations. One strand of the oligonucleotides for each mutation is shown; the reverse-strand primers were complementary: S216A, 5’-TCCGCCGGAGTTEACCTCCGT- GTGG-3‘; S261A, 5‘-GGAACTGCTGGT=AAATACG- GAACG-3‘; S266A, 5’-AAATACGGAACG=ATCAA- GAAGGAA-3’; S324A, 5I-ACAACTAAAGAAaGA- CAATTTCTAC-3‘; S365A, 5I-CTAAACATAAACa- GTGATTACGACG-3’; S395A, 5’-GGTTACGCATAT= GGAGGTGGAAAA-3’; S438A, 5‘-TGGTGTTTTTGG=- CTTGAGGTTGAG-3’; C191S, 5’-GCGACGCTAGTC- EGCGGGGAAGCGC-3‘; C191A, 5’-GCGACGCTAGT- CaGCGGGGAACCGC-3’.

The final PCR products were gel-purified and then digested with AccI and XhoI to produce a 1-kb (112-1098) fragment. This fragment was subcloned back into pAtc-46 to replace the wild-type sequence and then sequenced to verify the mutation.

Two mutants with modifications in the N-terminal leader sequence were prepared (Fig. 1). For the ”ntd3aa” deletion mutant (a deletion of amino acid residues Ala-Ala- Ser at positions 2 4 ) , the following oligonucleotides were used as primers: 5’-ACACTTTCCAACATGGTAGATA- ATCGC-3‘ and 5‘-GCGATTATCTACCATGTTGGAAAG- TGT-3.‘ The final PCR products were gel-purified and then digested with XhoI and PstI to generate a 125-bp fragment that was used to replace the corresponding wild-type fragment at the 5‘ end of the NR cDNA in pAtc-46. For the “ntdllaa” mutant (a substitution of amino acid residues 2-11 with the sequence His-His-His-His-His-His-

wt MAASVDNRQYARLEPG ntd3aa M---WNRQYARLBPG

ntdl laa cn”Ea----LEPQ-

\ / Hinge I Hinge II

Ala), the 5’ end of the NR cDNA in pAtc-46 (a 120-bp EcoRI-XhoI fragment) was replaced with the following double-stranded oligonucleotide: 5’-AATTCATGCATCAC- CATCACCATCACGCAC-3‘ and 3’-GTACGTAGTGGTAG-

As a control for the ntdllaa mutant and to provide for a convenient way to purify NR for native PAGE gels, a ”His-tag” NR was constructed in which the 125-bp XhoI- PstI fragment of pAtc-46 was replaced with two double- stranded oligonucleotides so that the amino acid sequence MMHHHHHHASK replaced the N-terminal Met of NR and the EcoRI site was restored. The mutated NR cDNAs were directly subcloned into the EcoRI site of the P. pastoris expression vector pHILD2 (Invitrogen, San Diego, CA). Standard cloning techniques were used for DNA manipu- lation (Sambrook et al., 1989).

TGGTAGTGCGTGAGCT-5’.

Expression of Mutant NR Proteins

Plasmids containing the desired mutations were transformed into P. pastoris strains GS115 and SMD1168 by electroporation as previously described (Becker and Guar- ente, 1990; Su et al., 1996). Growth and preparation of protein extracts from the transformants were as described previously, except no Triton X-100 was used in the extraction buffer (Su et al., 1996). For each mutant, at least 10 transformants were initially screened for NR expression, and the transformants showing the highest leve1 of NR activity were used for further characterization. The NR assay condition for the mutant NRs was as described previously (Su et al., 1996).

The partia1 purification of NR with a His-tag at the N terminus was performed as follows. Frozen P. pastoris cells (10 g) were suspended in 10 mL of 50 mM sodium phosphate, pH 7.5, 1 mM EDTA, 5% glycerol (v/v), and 1 mM PMSF. The suspended cells were lysed with a French press. The crude cell lysate was centrifuged at 16,000 rpm for 10 min. The supernatant was incubated with 5 mL of 80% (v/v) CL-6B Sepharose (Pharmacia) at 4°C for 30 min and then centrifuged for 5 min at 12,000 rpm. The supernatant was then incubated at 4°C for 30 min with 2 mL of a 50% slurry of nickel-nitrilotriacetate gel agarose (Qiagen, Chats- worth, CA) in nickel-nitrilotriacetate gel buffer (50 mM sodium phosphate, pH 7.5, 20 mM imidazole, 10 mM 2- mercaptoethanol, and 5% glycerol [v/v]) and then centrifuged at 4000 rpm for 30 s. The column resin was washed with 20 mL of 50 mM sodium phosphate, pH 7.0, 1 mM EDTA, 5% (v/v) glycerol, 300 mM NaCl, and 10 miv mercaptoethanol, resuspended in 5 mL of Mops buffer (50 mM Mops, pH 7.5, 1 mM EDTA, and 5% [v/v] glycerol), and

Figure 1. The diagram shows the location of the three cofactor-binding regions or domains, the N-terminal leader, and the hinge regions along the linear seauence of NR. Amino acids that were mutated in this study are shown along with

I I Heme 1 1 FAD I c the phosphorylated Ser-534. wt, Wild type. t t

MoCo

1138 Su et al. Plant Physiol. Vol. 11 5, 1997

then transferred to a 10-mL column. The column was washed with 5 mL of Mops buffer, and then NR was eluted with 4 mL of Mops buffer with 200 mM imidazole. Column fractions with maximal NR activity were dialyzed in Mops buffer and stored at -80°C. Overall recovery was 50% of the starting material with a final NR specific activity of 4 units/mg (180 pg of protein).

In Vitro lnhibition Assay and lmmunoblot Analysis

The inhibition assay was performed as previously described (Su et al., 1996). Briefly, P. pastoris extracts containing NR (0.001-0.003 unit) were mixed with 150 FL (approximately 300 pg of protein) of desalted protein extracts from G’4-3, a NR double mutant of Arabidopsis (Wilkinson and Crawford, 1993). The mixture was incubated at 22°C for 20 min in the presence or absence of 5 mM ATP, and the extent of inhibition was determined with an NR assay.

The immunoblotting procedure was as previously described (Su et al., 1996), except that an Immobilon-P mem- brane (Millipore), instead of nitrocellulose, and a Protean I1 gel apparatus (Bio-Rad) were used. Anti-NR antibodies prepared against Arabidopsis NIA2 NR purified from Esch- erichia coli (Wilkinson and Crawford, 1991) were used. Bound antibody was visualized by secondary antibodies and horseradish peroxidase. For the native polyacrylamide gel analysis, P. pastoris was lysed in extraction buffer (50 mM Mops, pH 7.5,1 mM EDTA, and 5% [v/v] glycerol with 0.4 pg/mL leupeptin and 1 pg/mL pepstatin). Tris-HC1 (pH 6.8) and glycerol were added to the samples to a final concentration of 50 mM and 10%, respectively. The samples were loaded onto a 5 to 20% gradient of Ready Gels for the Bio-Rad Mini-Protean I1 cell apparatus and electropho- resed according to the manufacturer’s instruction manual (Bio-Rad), and then the gels were immunoblotted as described above. For quantifying immunoblot results, the blots were scanned with an 1s-1000 digital imaging system (Alpha Innotech, San Leandro, CA).

RESULTS

Spectral and Kinetic Properties of Recombinant NADH:NR

To establish the utility of the P. pastoris expression system for producing plant NRs, the enzyme expressed in P. pastoris must be shown to be biochemically similar to NR isolated from plant tissues. To achieve this goal Arabidop- sis NR from the NIA2 gene was expressed in liter-shake cultures of P. pastoris, partially purified, and analyzed. The time course of NR expression in P. pastoris shake cultures showed that the highest levels of enzyme activity were attained after 48 h of methanol induction (Fig. 2) . The maximum NR leve1 achieved after 48 h of induction was 200 units/L P. pastoris culture. NR was extracted from these cultures and purified in three steps (Table I). Proteins were precipitated with ammonium sulfate (45% AS pellet in Table I) and then purified by affinity chromatography on Blue-Sepharose, with NADH elution followed by binding to AMP Sepharose and elution with NADH as described in “Materials and Methods.” The tight binding of NR to AMP

06 C Q ._ c

. c v)

c 3

c 2 0.2 u a

O Units N W m g Units N W L

z“ O 0 v,I, O 24 48 72

Methanol induction Time (hours)

Figure 2. Liquid cultures of P. pastoris expressing the wild-type Arabidopsis N R were induced with methanol at time O. At the indicated times, a sample of the culture was removed and assayed for N R activity and total protein. Details are provided in “Materials and Methods.” The average values and SDS are from three independent cultures. NRA, N R activity.

Sepharose is interesting, since we have observed that neither the recombinant Cyt b reductase nor the Cyt c reductase fragments of NR, which contain the NADH active site, binds tightly to this affinity gel (Shiraishi and Campbell, 1997). These results suggest that holo-NR may have an- other binding site for AMP in addition to the NADH active site in the Cyt b reductase fragment of the enzyme. Re- cently, some evidence for the influence of AMP on regulation of NR was reported (Huber et al., 1996).

NR was purified to a specific activity of 36 units/mg protein, which represents more than a 550-fold enrichment compared with the crude extract (Table I). Analysis by SDS-PAGE of the purified NR revealed that it had a major protein band at approximately 110 kD and was substan- tially free of contaminating proteins (data not shown). Western-blot analysis using antibodies against Arabidopsis and maize NADH:NR forms demonstrated that the purified NR was derived from the plant cDNA clone and not from a P. pastoris gene, and at the same time revealed the presence of partially degraded forms of the NR subunits as minor components of the purified samples (data not shown).

The purified NR samples were characterized to determine the spectral and kinetic properties of the enzyme. Spectral analysis demonstrated the presence of absorbance peaks typical of plant NR (Fig. 3). The oxidized enzyme possesses a peak A,,, that is characteristic of heme-Fe (Redinbaugh and Campbell, 1985; Campbell, 1991). Fur- thermore, reduction of the enzyme by the addition of solid NADH produced major peaks at 424 and 557 nm, which are characteristic of the interna1 Cyt b (heme-Fe) of NR (Red- inbaugh and Campbell, 1985). Enzyme reduced with di- thionite gave an identical spectrum (data not shown). These spectra compare favorably with the spectra previously reported for plant-derived NR (Solomonson et al., 1975; Redinbaugh and Campbell, 1985; Campbell, 1991). The A,,,:A,,, ratio of 1.4 for the P. pastoris-derived NR is nearly identical to both squash and Chlorella vulgaris NR,

Recombi nant Arabidopsis N itrate Reductase 1139

Table I . Purification table for wild-type NR expressed in P. pastoris Fraction Volume Protein Activitv Specific Activitv Total Units Recoverv Fold Purification

mL mg"L units/mL units/mg 0%

Crude extract 500 16.1 0.91 9 0.057 460 1 O0 1 45% AS Pellet 400 5.1 0.735 0.1 44 294 64 2.5 B I ue-Sepharose 3 7.7 36.3 4.7 1 o9 23.6 82.5 5'-AMP Sepharose 1.25 1.3 42.8 32 53.5 11.6 561

suggesting that the heme-Fe of these enzymes are structur- ally and functionally equivalent. There are some minor differences in the ratios of the UV absorbance to the Soret peaks at 413 and 424 nm. The P. pastoris-derived NR has ratios of 2.16 for A,,,:A,,, and 1.52 for A,,,:A,,,, which are slightly higher than the ratios reported for squash and C. vulgaris NR (Solomonson et al., 1975; Redinbaugh and Campbell, 1985; Campbell, 1991). The ratios are approximately 15% higher and may be due to differences in the amino Bcid sequence (Campbell, 1991) or perhaps differences in sample preparation.

The kinetic properties of the P. pastoris-derived enzyme were also determined. Apparent K , values of 4 and 15 PM for NADH and nitrate, respectively, were determined. These kinetic constants are very similar to those previously reported for spinach NR (Barber and Notton, 1990). The P. pastoris-derived NR displays a pH optimum of 7.0, which is similar to previously characterized NR forms (Campbell and Smarrelli, 1978; Barber and Notton, 1990). Thus, by the criteria tested, NR produced in P. pastoris behaves like NR isolated from plants.

Generation and Analysis of NR Mutants

With the previous results giving us confidence that P. pastoris-expressed NR is similar to authentic plant NR, we next examined the function of severa1 highly conserved amino acid residues through site-directed mutagenesis. One of the most important residues is the highly conserved Cys at position 191. This is the only invariant Cys residue

'.' t 424

1.2

0.8

0.4

0.0

I

400 450 500 550 600

Wavelength (nm) Figure 3. Visible spectra of recombinant Arabidopsis N R from P. pastoris. Spectra were determined at 25°C using N R purified as described in "Materials and Methods." The reduced spectrum was produced by adding solid NADH.

in the MoCo-binding regions of both NR and sulfite oxidase, is essential for funga1 NR activity, and is a proposed ligand to the molybdenum in NR (Barber and Neame, 1990; Solomonson and Barber, 1990; Garrett and Rajagopalan, 1994; Garde et al., 1995). Mutant forms of NR with alter- ations at position 191 would be invaluable for studying the role Cys-191 plays in binding molybdenum and catalyzing nitrate reduction in NR, as has been done for sulfite oxidase (Garrett and Rajagopalan, 1996). Mutations were constructed in the NR cDNA clone so that the Cys-191 was replaced with Ala or Ser.

Twenty P. pastoris transformants were generated for each mutation and then were tested for NR activity. NR activity was undetectable in extracts of P. pastoris expressing the C191S or C191A mutant forms of NR (Table 11). In addition, bromphenol blue- and methyl viologen-dependent NR activities, which require a functional MoCo-binding region, were not detected in these extracts (data not shown). One explanation for the absence of NR activity in these extracts is that the altered NRs were not stable and did not accu- mulate in P. pastoris. To examine this possibility, immunoblot analyses were performed. Both C191A and C191S NR proteins were detectable in the P. pastoris extracts, although at a lower leve1 than for wild-type NR (Fig. 4; Table 11). Thus, the Cys-191 mutant NRs are present but inactive, indicating that Cys-191 is essential for nitrate-reducing activity.

Because native NR forms dimers (Solomonson and Bar- ber, 1990), we wanted to determine whether the Cys-191 replacements affected dimerization. We investigated this by analyzing protein extracts on native gradient PAGE gels. Immunoblotting of the gels revealed that the mutant NR in the crude extracts forms both the apparent enzyme dimer and a dimer of dimers, as does wild-type NR in both crude extracts and purified form (Fig. 5). Therefore, Cys- 191 is not required for NR dimerization, and its replacement with Ser or Ala does not perturb the protein structure of the monomer enough to impair dimer formation. It will be interesting to determine whether the MoCo is still bound in these altered forms of NR after they are purified and analyzed in more detail. In the case of sulfite oxidase, the MoCo is still bound to the modified enzyme in which the corresponding invariant Cys is replaced with a Ser (Garrett and Rajagopalan, 1996).

We next examined the importance of the N terminus on NR activity by altering the most N-terminal amino acids. Six residues (MAASVD / E) are highly conserved in plant NRs. Two mutants were generated that lacked this sequence: ntd3aa with a deletion of residues 2 4 (AAS) and ntdllaa with a substitution of residues 2 to 12 (AASVD-

1140 Su et al. Plant Physiol. Vol. 115, 1997

Table II. Comparison of wild-type (wt) and mutant NR expressed in P. pastoris

Enzyme Form

wtNRC191AC191Sntd 3aantd! laaS216AS261AS266AS324AS365AS395AS438A

NR Crude ExtractSpecific Activity

units/mg0.0250.0000.0000.0180.0230.0150.0200.0120.0300.0050.0040.005

NR Protein Loadedon PACE Gel

fj. units

125000

9001150750

1000600

1500250200250

NR Protein fromWestern Blot'1

"812.56.35.6

11171212.512.5142.521.6

NR Specific Activity6

units/mg NR

10000

8268628048

107150100156

a Amount of NR protein for the wild-type was converted from a blot result to nanograms of NR proteinusing 100 units/mg as the specific activity for pure NR (see Redinbaugh and Campbell, 1985). b Cal-culated by dividing units of NR loaded onto gel by amount of NR protein determined by western blot.

NRQYAR) with HHHHHHA (Fig. 1). The specific NR ac-tivities of these mutants were similar to wild-type NR(Table II). This result shows that these residues are notrequired for activity and suggests that it may be possible toadd a His-tag sequence at the N terminus of NR to assist inpurification. The N-terminal Met of Arabidopsis NR wasreplaced with MMHHHHHHASK and transformed into P.pastoris. After induction, these strains produced functionalNR at the same level as for wild-type NR (data not shown).NR containing the His-tag sequence could be partiallypurified to 4 units/mg specific activity in a single stepusing a bound-nickel column as described in "Materialsand Methods."

In only one known NR sequence from higher plants isthe N-terminal sequence not conserved. This sequence isMSTCVEQ. . . from a maize NR genomic clone (GenBankaccession no. U20450). This NR sequence also lacks theconserved hinge 1 sequence that is involved in thephosphorylation-dependent regulation of NR. Perhaps,the N-terminal sequence has been conserved along withthe hinge 1 site for the proper regulation of NR. It has beensuggested that the N-terminal 45 amino acids of spinachNR are needed for inhibition (Douglas et al., 1995). Weassessed the role of the most N-terminal residues by assay-ing the ntdSaa, ntdllaa, and His-tag NR forms in our invitro inhibition assay developed previously for Arabidop-sis NR (Su et al., 1996). In this assay NR kinase and inhib-itor protein are provided by an extract from an NR- mu-

Figure 4. Wild-type and mutant protein extracts were subjected toSDS-PACE using a Protean II gel apparatus and then to immunoblotanalysis with anti-NR antibody as described in "Materials and Meth-ods." Lane 1, Wild-type NR; lane 2, C191A mutant NR; and lane 3,C191S mutant NR. In lane 1, 40 jug of total protein with an NRspecific activity of 0.03 unit/mg was loaded onto the gel. In lanes 2and 3, 120 and 180 jug of total protein were loaded onto the gel.

tant of Arabidopsis called G'4-3 (Wilkinson and Crawford,1993). After incubation with ATP recombinant NR pro-vided by the P. pastoris extracts is inhibited. As can be seenin Table II and Figure 6, the two mutant forms of NR with3 or 11 N-terminal amino acids missing were inhibited tothe same extent as wild-type NR. Thus, the very N termi-nus of Arabidopsis NR is not essential for enzyme activityor the phosphorylation-dependent regulation.

1 2 3

669 —

440 —

4 5

I29 —

Figure 5. Native gel analysis of wild type and C191 mutant NRs.Sample preparation and native gel electrophoresis were as describedin "Materials and Methods." Bio-Rad 5-20% polyacrylamide gradi-ent gels were used. After electrophoresis, NR protein was visualizedby immunoblot analysis as described in "Experimental Procedures."In all lanes NR was obtained from P. pastoris extracts. Lanes 1 to 4,Wild-type NR; lane 5, C191A NR; and lane 6, C191S NR. Thefollowing amounts of protein and NR activity were added to eachlane: lane 1, 100 jug of total protein (0.004 unit NR) of crude proteinextract; lanes 2 and 3, 2 /u.g of protein (0.004 unit NR) and 1 jug ofprotein (0.002 unit NR) of partially purified wild-type NR (with aHis-tag to aid purification, see "Materials and Methods"); lane 4, 50/u.g of protein (0.002 unit NR) of crude protein extract; and lanes 5and 6, 100 u.% of protein of crude protein extract. The proteinmarkers were thyroglobulin (669 kD), ferritin (440 kD), /3-amylase(200 kD), alcohol dehydrogenase (150 kD), and BSA (66 kD). Lanes1 to 3 were produced on a gel different from that used for lanes 4 to6. Arrows indicate corresponding bands on each gel.


i ~~

.NR S216A 1 166A I 324A

r T

0 no ATP ATP

d3aa

Our last analysis focused on highly conserved Ser residues in the MoCo-binding region of NR: Ser-216, Ser-261, Ser-266, Ser-324, Ser-365, Ser-395, and Ser-438. These Ser residues could be essential for nitrate reduction or could be sites for NR phosphorylation, because it has been shown that NR is phosphorylated on multiple sites (Huber et al., 1992; Labrie and Crawford, 1994). These Ser residues were changed to Ala, and the mutant NR forms were expressed in P. pastoris, as described in "Materials and Methods." Crude protein extracts from the P. pastoris transformants were assayed for NR activity and then separated by SDS- PAGE and visualized by protein blotting (Table 11). All of the Ser-to-Ala-substituted NRs were functional and had no less than 45% of the wild-type leve1 of NR activity after normalizing for the amount of NR protein in the extracts (Table 11, last column). We conclude that none of these Ser residues are essential for NR activity; however, some (especially S266A, S365A, S395A, and S438A) may have an influence on NR function or stability. The Ser mutants were also tested in the vitro inhibition assay. As shown in Figure 6, all mutant NRs showed Mg-dependent ATP inhibition similar to wild-type NR. Thus, none of these Ser residues are essential for the in vitro inhibition.

D I SCU SSI ON

In this study we demonstrated that Arabidopsis NR produced in P. pastoris and purified to near homogeneity behaves like NR isolated from plants. Since NR has not been purified from Arabidopsis, we took advantage of the fact that NR has characteristic properties, regardless of the plant source from which it was isolated (Campbell and Kinghorn, 1990; Solomonson and Barber, 1990; Hoff et al., 1994; Campbell, 1996). We found that the P. pasforis- derived NR has the same biochemical characteristics as natural plant NR. The visible spectra of oxidized and NADH-reduced enzyme and its kinetic characteristics are similar to those previously reported for NADH:NR from

I j l l aa His-wtNR

Figure 6. In vitro inactivation of wild-type (wt) and mutant NRs. P. pastoris extracts containing N R were mixed with desalted G'4-3 extracts and then incubated with or without 5 mM ATP; an N R assay was then performed as described in "Materials and Methods." The specific activity for each sample is shown in Table I . The values in the histogram are percentages of N R activity (NRA) relative to the O mM ATP control. Means and SES were calculated from duplicate experiments.

squash, maize leaf, and C. vulgaris (Solomonson et al., 1975; Redinbaugh and Campbell, 1985; Hyde et al., 1989; Camp- bell, 1991). The maximal absorbances (at 413 nm in the oxidized and 424 and 557 nm in reduced enzyme) are virtually identical to those found for squash and corn holo-NR and the recombinant Cyt c reductase fragments of corn and spinach NR, all of which are highly similar to eukaryotic Cyt b, (Redinbaugh and Campbell, 1985; So- lomonson and Barber, 1990; Campbell, 1991, 1996). The P. pastoris-derived NR also has similar K , values for both nitrate and NADH compared with plant-derived NR, being closest to those of spinach NR (Barber and Notton, 1990; Kleinhofs and Warner, 1990). The biochemical characteristics of the P. pastoris-derived NR described here, in addition to previous work (Su et al., 1996), make the P. pastoris expression system an attractive means to produce NR for more detailed biochemical characterization. Especially promising is the observation that under optimal methanol induction conditions in shake cultures (Fig. 2), P. pastoris produced more than 200 units/L active NR (estimated to be 2 mg/L of active NR based on a specific activity of 100 units / mg for pure NR [Redinbaugh and Campbell, 19851).

Large quantities of pure enzyme will be needed for detailed biochemical and structural analysis of holo-NR, such as those that have been carried out with the recombinant fragments of NR (reviewed in Campbell, 1996). Whereas it is clear from our results that more NR can be obtained from P. pastoris than from plant sources, it is very likely that the expression of NR in P. pastoris can be im- proved even more. Even higher levels of NR production should be possible when P. pastoris is grown in a fermentor in which much higher cell densities can be achieved (Ha- genson, 1991). We have also found that at least 5 times more spinach Cyt c reductase fragment than Arabidopsis holo-NR can be produced in P. pastoris (Shiraishi and Campbell, 1997). It is possible that MoCo production is limiting in P. pastoris, because we have found that supple- menting the P. pastoris growth medium with molybdate

1142 Su et al. Plant Physiol. Vol. 1 1 5, 1997

increases the production of active NR i n P. pastoris (data not shown). These and other improvements are currently being investigated to increase the leve1 of NR production i n P. pastoris.

Availability of the P. pastoris expression system allowed us to investigate NR functionality using site-directed mutagenesis. A key residue of NR is Cys-191, which is the only invariant Cys among NRs and sulfite oxidase and h a s been proposed to provide a ligand to molybdenum in the MoCo- binding region (Barber a n d Neame, 1990; Solomonson a n d Barber, 1990; Garrett and Rajagopalan, 1994). Changing this residue to an Ala i n Aspergillus NR completely abol- ishes NADPH:NR activity (Garde e t al., 1995). Changing the corresponding Cys i n rat and h u m a n sulfite oxidase to a Ser reduces oxidase activity by more than 2000-fold and changes the spectra of the MoCo-binding region (Garrett a n d Rajagopalan, 1996). Mutating the corresponding Cys (Cys-191) in Arabidopsis NR abolished activity, showing that this residue also plays a n essential role in plant NRs (Table 11). Because this mutation d id not prevent the accu- mulation of NR in P. pastoris (Fig. 4), we then asked whether it affected the quaternary structure of NR. It has been shown that NR forms dimers a n d dimers of dimers (Redinbaugh and Campbell, 1985; Hyde e t al., 1989; So- lomonson a n d Barber, 1990; Truong et al., 1991). We found that NR altered a t position Cys-191 apparently still forms dimers and dimers of dimers i n native polyacrylamide gels similar to wild-type recombinant Arabidopsis NR (Fig. 5). The C191S a n d C191A mutant NR forms will be useful for investigating the ligands to MoCo, if these forms are found to retain the cofactor after purification.

Our site-directed mutagenesis studies also included the remova1 and replacement of amino acids a t the very N terminus of NR. The first six amino acids are highly conserved, and yet they have no known function. W e examined mutant forms of NR i n which residues 2 to 4 (ntd3aa) were deleted and residues 2 to 12 were substituted with a His-tag sequence (ntdl laa) . Neither mutant w a s impaired i n NR activity or ATP-dependent inhibition. These results encouraged us to alter the N terminus of NR by adding a His-tag sequence directly t o the N terminus of NR, provid- ing a rapid purification of NR using metal-affinity chromatography. W e also replaced a11 of the conserved Ser residues i n the MoCo-binding region of Arabidopsis NR and found that none were required for NR activity or ATP- dependent inhibition. Taken together with our previous work, these results refine our understanding of NR structure a n d regulation a n d open up new avenues for the study of this important enzyme i n plant nitrogen metabolism.

Received May 29, 1997; accepted July 15, 1997. Copyright Clearance Center: 0032-0889/97/ 115/ 1135/09.

LITERATURE ClTED

Bachmann M, Huber JL, Liao P-C, Gage DA, Huber SC (1996a) The inhibitor of phosphorylated nitrate reductase from spinach is a 14-3-3 protein. FEBS Lett 387: 127-131

Bachmann M, Shiraishi N, Campbell WH, Yoo BC, Harmon AC, Huber SC (1996b) Identification of Ser-543 as the major regula-

tory phosphorylation site in spinach leaf nitrate reductase. Plant Cell 8: 505-517

Barber MJ, Neame PJ (1990) A conserved cysteine in molybdenum oxotransferases. J Biol Chem 265: 20912-20915

Barber MJ, Notton BA (1990) Spinach nitrate reductase. Effects of ionic strength and pH on the full and partia1 enzyme activities. Plant Physiol 93: 537-540

Becker DM, Guarente L (1990) High efficiency of transformation of yeast by electroproration. Methods Enzymol 194: 182

Calza R, Huttner E, Vincentz M, Touze P, Galangau F, Vaucheret H, Cherel I, Meyer C, Kronenberger J, Caboche M (1987) Cloning of DNA fragments complementary to tobacco nitrate reductase mRNA and encoding epitopes common to the nitrate reductases from higher plants. Mo1 Gen Genet 209: 552-562

Campbell WH (1991) Expression in Escherichia coli of cytochrome c reductase activity from a maize NADH:nitrate reductase complementary DNA. Plant Physiol 99: 693-699

Campbell WH (1996) Nitrate reductase biochemistry comes of age. Plant Physiol 111: 355-361

Campbell WH, Kinghorn JR (1990) Functional domains of assimilatory nitrate reductases and nitrite reductases. Trends Biochem Sci 15: 315-319

Campbell WH, Smarrelli J (1978) Purification and kinetics of higher plant NADH:nitrate reductase. Plant Physiol 61: 611-616

Cannons AC, Barber MJ, Solomonson LP (1993) Expression and characterization of the heme-binding domain of Chlorellu nitrate reductase. J Biol Chem 268: 3268-3271

Crawford NM (1995) Nitrate: nutrient and signal for plant growth. Plant Cell 7: 859-868

Crawford NM, Arst HN (1993) The molecular genetics of nitrate assimilation in fungi and plants. Annu Rev Genet 27: 115-146

Crawford NM, Smith M, Bellissimo D, Davis RW (1988) Se- quence and nitrate regulation of the Arabidopsis thaliuna mRNA encoding nitrate reductase, a metalloflavoprotein with three functional domains. Proc Natl Acad Sci USA 85: 5006-5010

Douglas P, Morrice P, MacKintosh C (1995) Identification of a regulatory phosphorylation site in the hinge 1 region of nitrate reductase from spinach (Spinacea oleraceu) leaves. FEBS Lett 377: 113-1 17

Dwivedi UN, Shiraishi N, Campbell WH (1994) Identification of an "essential" cysteine of nitrate reductase via mutagenesis of its recombinant cytochrome b reductase domain. J Biol.Chem 269:

Garde J, Kinghorn JR, Tomsett AB (1995) Site-directed mutagenesis of nitrate reductase from Aspergillus nidulans. J Biol Chem 2 7 0 6644-6650

Garrett RM, Rajagopalan KV (1994) Molecular cloning of rat liver sulfite oxidase. J Biol Chem 269: 272-276

Garrett RM, Rajagopalan KV (1996) Site-directed mutagenesis of recombinant sulfite oxidase. J Biol Chem 271: 7387-7391

Gonzalez C, Brito N, Marzluf GA (1995) Functional analysis by site-directed mutagenesis of individual amino acid residues in the flavin domain of Neurospora crassa nitrate reductase. Mo1 Gen Genet 249: 456464

Hagenson MJS (1991) Production of recombinant proteins in the methylotrophic yeast Pichiu pastoris. In R Seetharam, SK Sharma, eds, Purification and Analysis of Recombinant Proteins. Marcel Dekker, New York, pp 193-212

Higuchi R (1990) Recombinant PCR. In MA Innis, DH Gelfand, JJ Sninsky, TJ White, eds, PCR Protocols: A Guide to Method and Application. Academic Press, San Diego, CA, pp 177-183

Hoff JL, Truong HN, Caboche M (1994) The use of mutants and transgenic plants to study nitrate assimilation. Plant Cell Envi- ron 17: 489-506

Huber JL, Huber SC, Campbell WH, Redinbaugh MG (1992) Reversible light dark modulation of spinach leaf nitrate reductase activity involves protein phosphorylation. Arch Biochem Biophys 296: 58-65

Huber SC, Bachmann M, Huber JL (1996) Post-translational regulation of nitrate reductase activity: a role for Cazf and 14-3-3 proteins. Trends Plant Sci 1: 432-438

Hyde GE, Campbell WH (1990) High-leve1 expression in Esche- richia coli of the catalytically active flavin domain of corn leaf

13785-13791


NADH:nitrate reductase and its comparison to human NADH: cytochrome b5 reductase. Biochem Biophys Res Commun 1 6 8

Hyde GE, Wilberding JA, Meyer AL, Campbell ER, Campbell WH (1989) Monoclonal antibody-based immunoaffinity chromatography for purifying corn and squash NADH:nitrate reductase. Evidence for an interchain disulfide bond in nitrate reductase. Plant Mo1 Biol 13: 233-246

Kaiser WM, Huber SC (1994) Posttranslational regulation of nitrate reductase in higher plants. Plant Physiol 1 0 6 817-821

Kleinhofs A, Warner RL (1990) Advances in nitrate assimilation. In BJ Miflin, PJ Lea, eds, The Biochemistry of Plants. Academic Press, San Diego, CA, pp 89-120

Labrie ST, Crawford NM (1994) A glycine to aspartic acid change in the MoCo domain of nitrate reductase reduces both activity and phosphorylation levels in Arabidopsis. J Biol Chem 269: 14497-14501

Lu G, Campbell WH, Schneider G, Lindqvist Y (1994) Crystal structure of the FAD-containing fragment of corn nitrate reductase at 2.5 8, resolution: relationship to other flavoprotein reductases. Structure 2: 809-821

Lu G, Lindqvist Y, Schneider G , Dwivedi UN, Campbell WH (1995) Structural studies on corn nitrate reductase: refined structure of the cytochrome b reductase fragment at 2.5 A, its ADP complex and an active site mutant and modeling of the cytochrome b domain. J Mo1 Biol 248: 931-948

Mackintosh C (1992) Regulation of spinach-leaf nitrate reductase by reversible phosphorylation. Biochim Biophys Acta 1137 121-126

Meyer C, Levin JM, Roussel JM, Rouze P (1991) Mutational and structural analysis of the nitrate reductase heme domain of Nicotiana plumbaginifolia. J Biol Chem 266 20561-20566

Moorhead G, Douglas P, Morrice N, Scarable M, Aitken A, MacKintosh C (1996) Phosphorylated nitrate reductase from spinach leaves is inhibited by 14-3-3 proteins and activated by fusicoccin. Curr Biol 6: 1104-1113

Quinn GB, Trimboli AJ, Prosser IM, Barber MJ (1996) Spectro- scopic and kinetic properties of a recombinant form of the flavin domain of spinach NADH:nitrate reductase. Arch Biochem Bio-

1285-1291

phys 327: 151-160

Ratnam K, Shiraishi N, Campbell WH, Hille R (1995) Spectro- scopic and kinetic characterization of the recombinant wild-type and C242S mutant of the cytochrome b reductase fragment of nitrate reductase. J Biol Chem 270: 24067-24072

Redinbaugh MG, Campbell WH (1985) Quaternary structure and composition of squash NADH:nitrate reductase. J Biol Chem

Sambrook J, Fritsch EF, Maniatis T (1989) Molecular Cloning: A Laboratory Manual, Ed 2. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY

Schindelin H, Kisker C, Hilton J, Rajagopalan KV, Rees DC (1996) Crystal structure of DMSO reductase: redox-linked changes in molybdopterin coordination. Science 272: 1615-1620

Shiraishi N, Campbell WH (1997) Expression of nitrate reductase FAD-containing fragments in Pichia. In KJ Stevenson, ed, Fla- vins and Flavoproteins. University of Calgary Press, Calgary, Alberta, Canada, pp 931-934

Solomonson LP, Barber MJ (1990) Assimilatory nitrate reductase-functional properties and regulation. Annu Rev Plant Physiol Plant Mo1 Biol 41: 225-253

Solomonson LP, Lorimer GH, Hall RL, Borchers R, Bailey JL (1975) Reduced nicotinamide adenine dinucleotide-nitrate reductase of Chlorella vulgaris. J Biol Chem 250: 4120-4127

Su W, Huber SC, Crawford NM (1996) Identification in vitro of a post-translational regulatory site in the hinge 1 region of Ara- bidopsis nitrate reductase. Plant Cell 8 519-527

Truong FN, Meyer C, Daniel-Vedele F (1991) Characterization of Nicotiana tabacum nitrate reductase protein produced in Saccha- romyces cerevisiae. Biochem J 278: 393-397

Warner RL, Kleinhofs A (1992) Genetics and molecular biology of nitrate metabolism. Physiol Plant 85: 245-252

Wilkinson JQ, Crawford NM (1991) Identification of the Arabi- dopsis Chl3 gene as the nitrate reductase structural gene Nia. Plant Cell 3: 461471

Wilkinson JQ, Crawford NM (1993) Identification and characterization of a chlorate-resistant mutant of Arabidopsis thaliana with mutations in both nitrate reductase structural genes Nial and Nia2. Mo1 Gen Genet 239: 289-297

260: 3380-3385

analysis of wild-type and mutant plant nitrate reductase expressed

Documents