Cloning, Yeast Expression, and Characterization of the Coupling of Two Distantly Related Arabidopsis thaliana NADPH-Cytochrome P450 Reductases with P450 CYP73A5*

(Received for publication, April 9, 1997, and in revised form, May 22, 1997)

Philippe Urban Dagger , Claudia Mignotte Dagger §, Michaël Kazmaier §, Frédéric Delorme Dagger par and Denis Pompon Dagger **

From the Dagger  Centre de Génétique Moléculaire du CNRS, 91198 Gif-sur-Yvette, France, the § Centre Orsan de Recherches en Biotechnologies, 91953 Les Ulis, France, and the  Service de Radiophysiologie Végétale du CEA, 13108 St. Paul-Lès-Durance, France

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES


ABSTRACT

Two NADPH-cytochrome P450 reductase-encoding cDNAs were isolated from an Arabidopsis cDNA library by metabolic interference in a Saccharomyces cerevisiae mutant disrupted for its endogenous cpr1 gene. ATR1 encodes a protein of 692 amino acids, while ATR2 encodes either a 712-residue protein (ATR2-1), or a 702-residue protein (ATR2-2) depending on the choice of the initiation codon. Comparative analysis of ATR1 and ATR2-1 indicates 64% amino acid sequence identity and the absence of conservation in the third base of conserved amino acid codons. The two Arabidopsis reductases are encoded by distinct genes whose divergence is expected an early event in angiosperms evolution. A poly(Ser/Thr) stretch reminiscent of a plant chloroplastic targeting signal is present at the ATR2-1 N-terminal end but absent in ATR1. The cDNA open reading frames were expressed in yeast. The recombinant polypeptides were found present in the yeast endoplasmic reticulum membrane and exhibited a high specific NADPH-cytochrome c reductase activity. To gain more insight into the respective functions of the two reductases, the Arabidopsis cDNA encoding cinnamate 4-hydroxylase (CYP73A5) was cloned and co-expressed with ATR1 or ATR2 in yeast. Biochemical characterization of the Arabidopsis ATR1/CYP73A5 and ATR2-1/CYP73A5 systems demonstrates that the two distantly related Arabidopsis reductases similarly support the first oxidative step of the phenylpropanoid general pathway.


INTRODUCTION

P450 monooxygenases are involved in various biosynthesis pathways and in degradation of a large range of exogenous compounds (1, 2). In plants, besides their roles in herbicide and pesticide detoxication (3, 4) and in promutagen oxidative activations (5), P450s are widely involved in several secondary metabolisms such as the phenylpropanoid pathway (6, 7). This pathway, which leads to important molecules such as lignins, pigments, coumarins, flavonoids, and phytoalexins (6), involves a common oxidative step: the para-hydroxylation of cinnamate catalyzed by a P450 of the CYP73 family.

In animals, the microsomal P450 system associates in the same organism multiple P450 isoenzymes but a single ubiquitous NADPH-P450 reductase (CPR).1 Conclusive evidence for the presence of a single CPR-encoding gene was provided with Chinese hamster cells, mouse spleen cells (8), and rat hepatocytes (9). Unlike animals, higher plants express multiple forms of CPRs as inferred from Western blot analysis of purified material (10, 11). Durst et al. (12) have recently isolated two distinct partial length cDNAs (HTR1 and HTR3) from Helianthus tuberosus, corresponding to distinct (2.4- and 2.6-kb) mRNA species.

This observation questioned the physiological significance of the multiplicity of plant CPRs. Among hypotheses, one could consider the requirement of different CPRs to support the activities of specific plant P450s or the differential CPR expressions in plant tissues or at various development stages. The possibility of multiple subcellular locations for CPRs in plant-specific organelles in addition to the regular endoplasmic reticulum (ER) location is also of particular interest. To address these questions, full-length Arabidopsis CPR-encoding cDNAs were cloned by a function-based approach involving complementation in yeast of an endogenous CPR defect. The selection relies on the reduced growth rate and the strong ketoconazole hypersensitivity of a yeast strain, which has been disrupted for the CPR1 gene, which encodes yeast microsomal CPR. The addition of ketoconazole, a potent inhibitor of yeast P450 lanosterol demethylase, is required for selection due to the presence of alternate electron donors in yeast, making CPR1 deletion non-lethal. In the cpr1 strain, the CPR defect and ketoconazole inhibition contribute together to cause a full depletion of ergosterol biosynthesis, leading to cell growth arrest (13-15). Screening of an Arabidopsis cDNA library in a multicopy plasmid pFL61 under the transcriptional control of the phosphoglycerate kinase (PGK) promoter (16) was carried out in the W(R) strain. This strain is engineered to overexpress the yeast CPR in galactose- but not in glucose-containing medium (17). The PGK promoter is active on both carbon sources, allowing complementation by heterologous CPRs of the conditional mutant phenotype induced by culture of transformed W(R) cells on glucose. The isolated Arabidopsis CPRs were further characterized for a physiologic coupling with Arabidopsis cinnamate 4-hydroxylase P450 (CYP73A5) by coexpression in yeast. For that purpose, the Arabidopsis cDNA encoding CYP73A5 was cloned from the same library using a hybridization probe derived from the corresponding H. tuberosus CYP73A1 gene (18).


EXPERIMENTAL PROCEDURES

Materials

Restriction and DNA modification enzymes were obtained from New England Biolabs. Thermostable Taq and Pfu DNA polymerases were from Boehringer Mannheim and Stratagene, respectively. The SequenaseTM (version 2.0) DNA sequencing kit was purchased from U.S. Biochemical Corp. The pCRScriptTM cloning kit was from Stratagene. NADPH, horse heart cytochrome c, and cinnamic and coumaric acids were from Sigma, and 2-naphthoic acid was from Aldrich.

Yeast Strains and Vectors

The Saccharomyces cerevisiae strain W303-1B (MATa; ade2-1; his3-11,-15; leu2-3,-112; ura3-1; trp1-1) is designated as W(N). Yeast strain W(R) was constructed by substitution of the natural promoter of the CPR1 gene (encoding microsomal CPR) by the galactose-inducible and glucose-repressed promoter GAL10-CYC1. WRDelta derives from W(N) by disruption of the CPR1 gene with a TRP1 selection marker (17).

The pFL61 vector has been previously described (16). The yeast expression vector pYeDP60 (V60) contains both URA3 and ADE2 as selection markers, and an expression cassette constituted by GAL10-CYC1 promoter and PGK terminator sequences surrounding a cDNA insertion polylinker (19). The pGP1 vector contains a 5500-bp-long fragment that encompasses the full CPR1 gene at the HindIII site of pUC19. The pYeDP51 yeast expression vector is identical to pYeDP1/8-2 (20), except that the PvuII site close to the yeast origin of replication is replaced by a BglII site.

Using ORS31 and ORS32 as primers (Table I) and pGP1 as a template, the 428-bp-long CPR1 gene terminator sequence was PCR-amplified and cloned into the SmaI site of pUC19. The 420-bp-long EcoRI-PvuII fragment of the resulting vector was ligated to the EcoRI-PvuII fragment of pYeDP51 encompassing URA3. The 6900-bp-long resulting vector, named pYeDP100, places the inserted cDNA under the transcriptional control of GAL10-CYC1-inducible promoter and CPR1 terminator. By PCR amplification using ORS33 and ORS34 primers and pGP1 as a template, a 631-bp-long HindIII-BglII fragment of the upstream part of the CPR1 promoter has been obtained and cloned in pUC19. The HindIII-BglII fragment of the CPR1 promoter was ligated with the larger HindIII-BglII fragment of pYeDP100, which encompasses URA3, yielding the 5210-bp-long pYeDP110 integrative vector. The pYeDP110 vector bears GAL10-CYC1 promoter and CPR1 terminator sandwiching a cDNA insertion polylinker, and the CPR1 5'-noncoding region placed upstream of the URA3 selection marker. The entire integrative fragment can be removed by a NotI digestion. The respective orientation of the two CPR1 noncoding regions allows targeted integration of the expression cassette at the CPR1 locus by homologous recombination.

Table I. Sequence of primers used in the PCR amplifications

Lowercase and uppercase letters indicate sequences without and with homology to the template DNA, respectively. The translation initiation codon or the sequence complementary to the stop codon is underlined.

Primer name Sequence

ATC4H-5 ggatccATGACCTCCTCTTGC
ATC4H-3 ggagatcTTAACAGTTCCTTGG
ORS21 gcggatccTTACCATACATCTCTAAGATATCT
ORS22 ccggatccATGTCCTCTTCTTCTTCTTCGTCAACC
ORS23 ccggatccATGATCGATCTCATGGCAGC
ORS24 gcgagatctATGACTTCTGCTTTGTATGCTTCC
ORS25 gggaattcTCACCAGACATCTCTGAGGTATC
ORS31 cgggaattccagctggcggccgcGTATGTTGTCTTTGAAGATGCACCCGTGGG
ORS32 gcggaattcTAATCAGCCCACTGATCAACGTTCGG
ORS33 ccagatctgcggccgcCAGTAAATTCACTTGGAATCACACCG
ORS34 gcgaagcttGCGACAGTAGATCACTTGAAAACCG

Cell Culture

Transformation of yeast strains was performed by a modified lithium acetate procedure as described (21), except for the transformation of W(R) cells with the pFL61 library (see below). Transformed cells were selected onto glucose-containing SGI plates. Culture media, cell cultures, and galactose induction procedures of individual clones in both SLI and YPGE were as described previously (17, 22).

Screening of the Arabidopsis cDNA Expression Library in Yeast

A sized (inserts >2 kb) cDNA library from Arabidopsis thaliana seedlings including roots (two-leaf stage) in pFL61 yeast expression vector was kindly provided by Dr. M. Minet. In pFL61, cDNAs were inserted as NotI cassettes between the yeast PGK promoter and terminator (16). Yeast W(R) cells were grown on galactose, thus overexpressing yeast CPR, up to a cell density of 2 × 107 cells/ml. Cells are harvested and resuspended at a cell density of 106 cells/ml in a glucose-containing YPGA medium, thus turning off CPR1 gene expression. W(R) cells were further grown on glucose for 5 h until the microsomal CPR activity decreases to the value found in wild-type W(N) cells. About 108 cells were lithium-treated and used for transformation by 10 µg of the Arabidopsis cDNA library, 0.2 mg of heat-denatured herring sperm DNA in 2 ml of 50 mM Tris-HCl buffer, pH 7.4, containing 1 mM EDTA and 0.1 M LiCl. After a 10-min incubation at room temperature, 5 ml of the same buffer containing 50% of poly(ethylene)glycol-4000 was added and gently mixed to the previous 2 ml, and the incubation was continued for 20 min at 37 °C and then 5 min at 42 °C. The cells were sprayed in 0.1-ml aliquots onto 20 plates of SGAI synthetic medium for uracil prototrophy selection.

Two pools of Ura+ transformants were constituted, and cells from each pool were sprayed at 104 cells/plate onto a series of SGAI plates containing increasing concentrations of ketoconazole (0, 5, 10, 20, and 50 µg/ml ketoconazole) and incubated for 36 h at 28 °C. The 52 clones found to be resistant to 10 and 20 µg/ml ketoconazole were striped on SGAI medium containing 10 µg/ml ketoconazole. The plasmidic DNA of 24 of the resistant clones was rescued in Escherichia coli. The BamHI and HindIII digestion patterns revealed five different restriction profiles (profiles A-E).

Isolation of Arabidopsis CYP73A5-encoding cDNA

The Arabidopsis cDNA library in pFL61 was transformed in E. coli by electroporation. About 105 independent clones were recovered and transferred on nitrocellulose. The blots were screened by hybridization in SSC × 6 buffer containing 0.1% SDS and 0.05% milk overnight at 55 °C with the 32P-labeled 1800-bp-long cDNA encoding Helianthus CYP73A1 (18). Of 12 independent clones selected, two strongly hybridize with the same probe after subcloning. The plasmidic DNA of one of the two clones, named pCYP73A5, was isolated and the NotI fragment encompassing the full-length cDNA sequenced.

DNA Sequence Determination

The NotI cDNA inserts encoding ATR1 and ATR2 were extracted from pFL61 and subcloned into the unique NotI site of a derivative of pUC9. For ATR1 and ATR2 cDNAs, sequencing was carried out with M13 universal oligonucleotides and a series of specific oligonucleotides belonging to the sequences. For CYP73A5 cDNA, the sequencing using M13 universal oligonucleotides was performed by using nested deletions with exonuclease III of pCYP73A5 digested by KpnI and HindIII. Nucleotide sequences were determined on both strands by the dideoxynucleotide chain termination procedure.

Southern Blot Analysis

The Arabidopsis genomic DNA (kindly provided by Dr. Clint S. Chapple) was digested with BamHI, EcoRI, or HindIII. The digested aliquots (10 µg) were separated by 0.8% agarose electrophoresis and then transferred to a Nylon membrane (Schleicher and Schüll). The blotted membrane was probed in stringent conditions with 32P-labeled ATR1 or ATR2 open reading frame.

Tailoring of the Selected cDNAs

Oligonucleotides were synthesized and purified as described previously (19). The 2079-bp-long ATR1 coding sequence was PCR-amplified using ORS24 and ORS25 primers with 10 ng of pFL61/ATR1 as template. These primers introduce a BglII site immediately upstream of the ATG codon and a EcoRI site downstream of the stop codon of the PCR-amplified fragment. The 2139-bp-long ATR2-1 coding sequence was PCR-amplified using the combination of ORS21 and ORS22 primers, and the 2109-bp-long ATR2-2 coding sequence was PCR-amplified using ORS21 and ORS23 primers and 10 ng of pFL61/ATR2 as template in both cases. The three ATR2 primers introduce a BamHI site immediately upstream of the ATG initiation codon and downstream of the stop codon of the PCR-amplified fragments. The expected PCR-amplified products were cloned at the SmaI site of pUC19, giving, respectively, pUC19/ATR1, pUC19/ATR2-1, and pUC19/ATR2-2. The ATR1 coding sequence was then excised by BglII-EcoRI digestion and inserted into pYeDP60 linearized at BamHI and EcoRI unique sites, resulting in plasmid pATR1/V60. The ATR2-1 and ATR2-2 coding sequences were excised by BamHI digestion and inserted into pYeDP60 linearized at BamHI site, a clone with the proper orientation was chosen for each insert giving, respectively, pATR2-1/V60 and pATR2-2/V60. Amplification of the 1518-bp-long CYP73A5 coding sequence was performed by PCR using ATC4H-5 and ATC4H-3 primers, and the product was cloned at the SfrI site of pCRScript vector, giving pCRScript/ATC4H. The primers introduce a BamHI site immediately upstream of the ATG codon and a BglII site downstream of the stop codon. The coding sequence was excised by BamHI-BglII digestion and inserted at the BamHI site of pYeDP60, and a clone with the correct orientation was selected and named pATC4H/V60.

Engineering of WAT11, WAT21, and Derived Yeast Strains

The ATR1 and ATR2-1 open reading frames were isolated from pUC19/ATR1 and pUC19/ATR2-1 as BamHI-EcoRI and BamHI-BamHI fragment, respectively. The ATR1 fragment was inserted at the BamHI-EcoRI sites of pYeDP110, yielding pATR1/DP110. Similarly, the ATR2-1 fragment was inserted at the unique BamHI site of pYeDP110, and a clone with the proper orientation was selected and named pATR2-1/DP110. The NotI fragment of each plasmid encompassing URA3 and the ATR coding sequence was isolated and used to transform WRDelta cells. Since in WRDelta the CPR1 locus is disrupted with the TRP1 gene, integration events result in a phenotype shift from Ura- Trp+ to Ura+ Trp-. This procedure yielded two new yeast strains, WAT11 (for ATR1) and WAT21 (for ATR2-1). They express the Arabidopsis CPR instead of the yeast enzyme when grown on galactose. On glucose, WAT cells express no CPR activity. The URA3 selection marker in both WAT strains was mutated to ura3 by selecting 5-fluoroorotate-resistant clones (23), giving WAT11U and WAT21U strains.

Subcellular Fractionation and P450 Quantitation

W(R) and WAT cells were harvested after a 12-h galactose induction and microsomal fractions prepared as described in Ref. 17. P450 content in yeast microsomes was calculated from the reduced carbon monoxide difference using a differential absorption coefficient (450 versus 490 nm) of 91 mM-1·cm-1 (24). Protein concentrations were determined by the Pierce BCA (bicinchoninic acid) protein microassay using bovine serum albumin as a standard.

Enzymatic Activities

Cinnamate 4-hydroxylase activity was assayed by high pressure liquid chromatography-monitoring 4-hydroxy-cinnamate production as described previously (22). For naphthoate hydroxylation reactions, activity was determined using a 1-ml assay mixture containing 120 µM NADPH, various concentrations of 2-naphthoate, and the microsomal aliquot in 50 mM Tris/HCl buffer, 1 mM EDTA, pH 7.4. The time dependence of the fluorescence emission was monitored at 440 nm, the excitation wavelength being set at 290 nm. The rate of the 2-naphthoate 6-hydroxylation reaction was calculated from the slope of the linear graph obtained. Transformed yeast was grown and assayed for cinnamate bioconversion essentially as described in Refs. 17 and 22.

CPR activity was measured on microsomal fractions as described previously (22). The rate of cytochrome c reduction was calculated using a differential absorption coefficient of 21 mM-1·cm-1 at 550 nm. The rates of dichlorophenolindophenol and ferricyanide reductions were monitored, respectively, at 600 nm (22 mM-1·cm-1) and 420 nm (1 mM-1·cm-1).


RESULTS

Expression Cloning of Arabidopsis CPRs

Yeast cells W(R) that carry the CPR1 gene under the transcriptional control of the galactose-inducible GAL10-CYC1 promoter were transformed with the Arabidopsis cDNA expression library in pFL61 (16). Extinction of the modified CPR1 gene, when glucose is used as a carbon source for culture, is not lethal (14, 17). However, yeast CPR deficiency caused a dramatic lowering in transformation efficiency due to changes in the cell wall structure. This led us to the design of an "on the fly" transformation strategy (see "Experimental Procedures"). The ketoconazole resistance criterion was used for the selection of clones expressing constitutively an alternate CPR in the pFL61-based cDNA expression library.

Approximately 105 Ura+ colonies were selected, pooled, and sprayed on a series of SGAI plates containing increasing ketoconazole concentrations. Fifty-two colonies were sorted out as resistant to 10 and 20 µg/ml ketoconazole. Twenty-four of them, selected at random, were further analyzed by restriction digestion. Five types of restriction profile were evidenced: A (7 clones), B (6 clones), C (5 clones), D (1 clone), and E (5 clones). The plasmidic inheritance of the ketoconazole resistance associated with each type of sequence was tested by transforming WRDelta , a yeast strain carrying a permanent CPR1 disruption. Four (A, B, C, and E) of the five plasmid classes were shown to confer ketoconazole resistance at 40 µg/ml to WRDelta . Partial cDNA sequencing was used for further identification and showed that classes A, C, and E (~2300-bp-long cDNA NotI fragment) contain related inserts differing only by their orientation and their 5'- and 3'-noncoding extremities when class B (2199-bp-long NotI insert) appeared containing a different open reading frame. The relative amounts of both ATR mRNAs in the plant were estimated by transforming E. coli with the Arabidopsis cDNA library in pFL61. Out of 104 independent clones, two hybridize ATR1 probe and seven hybridize ATR2 probe (not shown). This result demonstrates that in the cDNA library used, ATR1 cDNA is 3 times less represented than ATR2 cDNA.

Sequence Analyses of Arabidopsis CPRs

The cDNA inserts of class B and C plasmids were sequenced on both strands. The ATR1 cDNA is 2199 bp long and encompasses 69 bp of the 5'-noncoding region followed by a 2079-bp-long open reading frame and 51 bp of the 3'-noncoding region (Fig. 1). A termination TGA codon (from -13 to -15) immediately precedes in frame the first putative initiation codon. The first methionine codon of ATR1 cDNA closely matches the dicot plant initiator codon consensus sequence (25). The ATR1 open reading frame encodes a protein of 692 amino acid residues with an estimated molecular mass of 76,720 Da, similar to that of rat liver CPR (77 kDa) (26).


Fig. 1. Nucleotide sequence and deduced protein sequence of Arabidopsis ATR1. The deduced gene product is indicated in one-letter code below the DNA sequence. The 2199-bp-long cDNA encompasses a single open reading frame beginning at positions 70-72 and encoding a protein of 692 amino acids. A proposed motif for anchoring to the endoplasmic reticulum membrane is underlined. Numbers on the left indicate amino acid positions.
[View Larger Version of this Image (55K GIF file)]

The 2290-bp-long ATR2 cDNA, also obtained as a full-length clone, consists of a 2139-bp-long open reading frame sandwiched by a 5'-noncoding region of 50 bp and a 3'-flanking region of 101 bp (Fig. 2). The ATR2 open reading frame encodes a protein of 712 amino acid residues, which is 64% identical to ATR1, with an estimated molecular mass of 79,077 Da, more similar to the Mr observed for higher plant CPRs (11, 27, 28). The assignment of the starting codon was not straightforward, since the flanking sequence preceding the first ATG of ATR2 cDNA does not contain any stop codon in frame and since two other ATG codons were found in the 15 first codons. Comparison with the consensus sequences around the initiation codon in dicot plant mRNAs (25) indicates that both the first and the second ATG of ATR2 cDNA could be a suitable translation initiation site; it is therefore difficult to predict which of these is the true starting site. The ATR2 protein encoded by the first ATG was designated ATR2-1 and is 712 residues long; the shorter protein, starting at the second ATG, was named ATR2-2 and is 702 residues long (Fig. 3).


Fig. 2. Nucleotide sequence and deduced protein sequence of Arabidopsis ATR2. The deduced gene product is indicated in one-letter code below the DNA sequence. The 2290-bp-long cDNA encompasses a single open reading frame beginning at positions 51-53 or at positions 81-83 and encoding either a protein of 712 amino acids (ATR2-1) or a protein of 702 residues (ATR2-2). A proposed motif for anchoring to the endoplasmic reticulum membrane is underlined. The poly(Ser/Thr) stretch is double-underlined. The two putative ATG translational start codons are boxed. Numbers on the left indicate amino acid positions.
[View Larger Version of this Image (57K GIF file)]


Fig. 3. Comparison of Arabidopsis CPR sequences. A, nucleotide and amino acid sequences of the N-terminal part of ATR1 and ATR2 with, in the latter case, the two ATR2 putative proteins. B, nucleotide sequence of the 3'-end of ATR1 and ATR2 cDNAs and the deduced amino acid sequences.
[View Larger Version of this Image (16K GIF file)]

Fig. 4 shows, based on sequence comparison with flavodoxin (26, 29) and NADP+-ferredoxin reductase (30, 31), that each ATR presents the conserved segments typical of all known CPRs for the binding of FMN, FAD, and NADPH. In this respect, ATR1 and ATR2 are typical NADPH-P450 reductases. In contrast, no clear sequence similarity is found between the 85 first residues of ATR1 and the 106 first residues of ATR2. However, despite the absence of similarity in their N-terminal part, both ATR1 and ATR2 present in this part an hydrophobic segment that could act as a membrane anchor, as observed in other microsomal CPRs. Another distinct difference between the two ATRs is the net charge of the protein; ATR1 is predicted to be more acidic than ATR2 at neutral pH, exhibiting a calculated charge versus pH profile very different from those of ATR2 and human and yeast CPRs (Fig. 5). The differences in charge density appear to be mainly localized in the C-terminal moieties of the ATRs. More unusual is the presence at the N terminus of ATR2 of a poly(Ser/Thr) stretch that is not found in any other fungal, plant, or animal CPR. This particular segment could account for the specific glycosylations found in certain higher plant CPRs. The protein ATR2-2, encoded from the second ATG, corresponds to ATR2-1 depleted of this stretch of hydroxylated amino acid residues. Moreover, a second hydrophobic stretch that could serve as a membrane anchor is found at the N-terminal side of the canonic ER membrane-anchoring element in both ATR2-1 and ATR2-2.


Fig. 4. Conserved regions in the CPR family. Shown are the eight peptide segments known to be involved in cofactors binding in comparison with other flavoenzymes of known three-dimensional structure. The first entry for each enzyme gives the residue number for the first amino acid in each segment. A, sequence alignments in the FMN-binding domain. B, sequence alignments in the FAD- and NADPH-binding domains. The compared sequences are, from top to bottom, S. cerevisiae (S.cer; GenBankTM D13788); three mammalian CPRs (mouse, human (GenBankTM S90469), and rabbit (GenBankTM X04610)); one insect CPR (fly (Musca domestica, GenBankTM L19897); one fish CPR (trout); six plant CPRs (the Arabidopsis ATR1 and ATR2, HTR1 from Jerusalem artichoke (H. tuberosus, GenBankTM Z26250), Catharanthus roseus from Madagascar periwinkle (C.ros, GenBankTM X69791), Vicia sativa from broad bean (V.sat, GenBankTM Z26252), and Vigna radiata from Mung bean (V.rad, GenBankTM L07843)); two flavoprotein domains (that of cytochrome P450 BM3, a soluble bacterial P450 involved in long chain hydrocarbon hydroxylation (BM3, GenBankTM J04832), and that of the murine macrophage nitric-oxide synthase (Mac NOS, GenBankTM M84373)); and spinach ferredoxin:NADP+ reductase (Sp FNR, GenBankTM X07981).
[View Larger Version of this Image (67K GIF file)]


Fig. 5. Computed net charge versus pH profiles of Arabidopsis CPRs. Profiles shown are those of human CPR, ATR1, and ATR2 in the 4-12 pH range. The broken line highlights values at pH 7. The isoelectric program of the UWGCG software package was used.
[View Larger Version of this Image (21K GIF file)]

Cloning and Sequencing of Arabidopsis CYP73A5

Using as a probe the coding sequence of Helianthus CYP73A1 (18), one pFL61 clone of the Arabidopsis library was selected by hybridization. The 1735-bp-long full-length cDNA contains an open reading frame of 1518 bp. The translation initiation site was assigned to the first ATG, since it is preceded in frame by a stop codon (from -37 to -39-positions). The 5'- and 3'-noncoding regions are 48 and 169 bp long, respectively (Fig. 6). This Arabidopsis cDNA encodes a protein of 505 amino acid residues with a calculated molecular mass of 57,751 Da, a value typical of eukaryotic P450s. The high sequence identity found between Arabidopsis and Helianthus CYP73s (82%) confirms that cinnamate 4-hydroxylase enzymes share highly conserved sequence.


Fig. 6. Nucleotide sequence and deduced amino acid sequence of Arabidopsis cinnamate 4-hydroxylase. The deduced gene product is indicated in one-letter code below the DNA sequence. The 1735-bp-long cDNA encompasses a single open reading frame beginning at positions 49-51 and encoding a protein of 505 amino acids. A proposed motif for anchoring to the endoplasmic reticulum membrane is underlined. Numbers on the left indicate amino acid positions.
[View Larger Version of this Image (66K GIF file)]

Characterizations and Evolution of ATR Genes

A Southern blot of Arabidopsis genomic DNA digested with different restriction enzymes was probed with either ATR1 or ATR2 coding sequence (Fig. 7). With ATR1 probe, a single band of 12.5 kb is observed with EcoRI digestion, a site absent in the cDNA, while two bands are observed with BamHI digestion and three distinct bands with HindIII. On the other hand, with ATR2 probe, two bands are observed with EcoRI and HindIII and only one with BamHI digestion (absent from the cDNA). No cross-hybridization is observed. Therefore, the fact that a digestion gives in each case a single band strongly suggests that each ATR cDNA sequence originates from a single copy gene.


Fig. 7. Southern analysis of Arabidopsis genomic DNA. A, 10 µg of DNA were digested with BamHI (B), EcoRI (E), and HindIII (H), transferred onto a nylon filter, and probed in stringent conditions with 150 ng of 32P-labeled ATR coding sequence obtained by NotI digestions from pUC19/ATR1 (left) or pUC19/ATR2 (right). Sizes of marker fragments (lane ss) are indicated in kb. B, restriction map of both ATR cDNAs for BamHI, EcoRI, and HindIII. The numbering refers to the first bp of the cDNA. Open rectangles refer to noncoding regions, and solid rectangles refer to coding sequence.
[View Larger Version of this Image (35K GIF file)]

PCR amplification starting from Arabidopsis genomic DNA and using the primers situated at both extremities of the open reading frames led to the amplification of a 3.9-4.0-kb fragment for ATR1 and of a 3.5-3.6-kb fragment for ATR2. This result confirms the presence of a total of 1.8-1.9 kb and of 1.4-1.5 kb of intronic sequences in the ATR1 and ATR2 genes, respectively. The PCR amplification of CYP73A5 DNA from Arabidopsis genomic DNA as a template using the two primers situated at both extremities of the CYP73A5 coding sequence resulted in amplification of a 1.7-1.8-kb fragment, indicating the presence of a total of about 200 bp of intron sequence, which is a rather small amount.

Dating of the ATR1/ATR2 gene divergence event was attempted by calculation of the level of identity (or ETBI value) between the third base of codons encoding conserved amino acid residues. A reference value (or RTBI value) was also calculated similarly by comparing all possible nucleotide sequences encoding ATR1/ATR2-1 conserved amino acid residues using the observed average codon usage as a ponderation factor. For ATR1/ATR2, ATR1/HSR1, and HSR1/SCR1 couples, Table II shows that ETBI and RTBI values are similar, as expected for distantly related genes (no memory of the third base usage). In contrast, ETBI and RTBI values still significantly differ for H1A1/H1A2 human P450s taken as a control. Such a difference is also detected for the comparison of Helianthus CPR cDNA (HTR1) (12) with ATR2 but not with ATR1. The lower amino acid sequence identity observed between ATR1 and ATR2 (64%) than between ATR2 and HTR1 (73%) and the higher ETBI value for the HTR1/ATR2 couple therefore indicate that the ATR2 gene is probably more distant from the ATR1 gene, although it belongs to the same species, than from the Helianthus HTR1-encoding gene (interspecies comparison).

Table II. Sequence comparisons

The comparison was carried out on the amino acid sequences and on the open reading frames. The difference value is measured by subtracting the RTBI value from the ETBI. The higher this value is, the more recent is the divergence of the two compared genes. H1A1 and H1A2 are human CYP1A1 and CYP1A2.

Compared sequences ATR1 ATR2 ATR1 HTR1 ATR2 HTR1 HSR1 SCR1 HSR1 ATR1 or ATR2 SCR1 ATR1 or ATR2 H1A1 H1A2

Amino acid identity (%) 64 69 73 33 36 30 72
ETBI 46 47 54 44 34 38 77
RTBI 42 43 44 41 32 37 52
Difference 4 4 10 3 2 1 25

To confirm these findings, a phylogenetic analysis was performed on CPR protein sequences using the CLUSTER algorithm and the BLOSUM62 sequence comparison matrix (32). The phylogenetic tree (not shown) revealed that plant CPRs clearly partition into two divergent subfamilies, one including ATR1 and Mung bean CPR and the second clustering ATR2 and all other plant CPR sequences available today. Interestingly, the two known Helianthus CPR sequences, the full-length HTR1 and the partial length HTR3 (which are 85% similar), cluster in the same subfamily together with ATR2. Thus, Arabidopsis ATR1 and ATR2 sequences constitute the first characterized presence of two distantly related CPR subfamilies in a single organism. These data highly suggest that the corresponding gene divergence probably occurred early during angiosperms evolution.

High Level Expression of Arabidopsis CPRs in Yeast

To optimize expression, the cDNAs encoding ATR1 and ATR2 were reformatted by deletion of their flanking regions and cloned in pYeDP60, placing the coding sequences under the transcriptional control of a GAL10-CYC1 promoter. WRDelta cells were transformed by the resulting vectors, namely pATR1/V60, pATR2-1/V60, and pATR2-2/V60. Upon expression in strain WRDelta following galactose induction, the three Arabidopsis enzymes confer a resistance to ketoconazole at least equal to that observed with the original pFL61 clones. This result demonstrates that each ATR is fully competent to substitute for endogenous yeast CPR and particularly that both ATR2-1 and ATR2-2 are functional in coupling with yeast P450s. Thus, deletion of the poly(Ser/Thr) stretch in ATR2-2 has no detectable consequence on the subcellular location or function in yeast.

Fig. 8 shows that microsomes prepared from ATR-expressing yeasts present a dramatically reduced endogenous P450 content as compared with the WRDelta control. The high level of P450s found in the ER of WRDelta cells was shown previously to be induced by CPR deficiency and decreased upon yeast cytochrome b5 overexpression (15). The expression of ATR1 or ATR2-1 in WRDelta thus reestablishes a wild type transcriptional regulation of endogenous P450s in CPR-deficient cells, indicating that both Arabidopsis CPRs efficiently substitute physiologically for the endogenous yeast enzyme.


Fig. 8. Carbon monoxide-induced difference spectra exhibited by dithionite-reduced microsomal fractions of transformed yeast cells. WRDelta cells were transformed by a cDNA-free pYeDP60 vector (control yeast), by pATR1/V60 (ATR1-expressing), or by pATR2-1/V60 (ATR2-expressing). Cell culture was performed in SLI up to a density of 3 OD at 600 nm. Microsomal fractions were isolated, and the difference spectra were recorded using 1 mg of microsomal proteins/ml of 0.1 M Tris-HCl buffer, pH 7, containing an excess of sodium dithionite.
[View Larger Version of this Image (23K GIF file)]

Microsomal fractions prepared from transformed WRDelta cells were prepared and assayed for several NADH- and NADPH-dependent acceptor reduction reactions typical of the CPR enzymes (Table III). As a control, microsomes prepared from WRDelta cells transformed by a cDNA-free pYeDP60 exhibited very low NADPH-cytochrome c reductase activity. NADH does not support any ATR-dependent reductase activity. In contrast, a strong microsomal NADPH-cytochrome c reductase activity is observed for the three ATRs, with ATR2-1 exhibiting an unusually high activity (2840 nmol/min/mg) as compared with ATR1 and ATR2-2 (100 and 280 nmol/min/mg, respectively). These values are to be compared with a microsomal specific activity of 100 nmol/min/mg for human CPR overexpressed in WRDelta cells using the same vector system. This result indicates that the N-terminal poly(Ser/Thr) stretch of ATR2 is not essential for its CPR activity or ER targeting in yeast but significantly enhances the expressed ATR2-1 reductase activity as compared with ATR2-2. The Km value of ATR1 and ATR2-1 are, respectively, 12 ± 2 and 8 ± 1 µM for NADPH and 3 ± 1 and 3 ± 1 µM for cytochrome c. These values compare well with those reported for other plant CPRs (10, 27, 33, 34).

Table III. Reductase activities catalyzed by Arabidopsis CPRs in microsomal fractions from transformed WRDelta cells

Control, ATR1, ATR2-1, and ATR2-2, respectively, stand for cDNA-free pYeDP60-, pATR1/V60-, pATR2-1/V60-, and pATR2-2/V60-transformed WRDelta cultured in YPGE/Gal medium. The reductase activities were measured spectrophotometrically in 50 mM Tris HCl buffer, pH 7.4, containing 1 mM EDTA at room temperature. Data represent the average of values measured on microsomal fractions from five independent transformed clones.

Donor Acceptor CPR activity
Control ATR1 ATR2-1 ATR2-2

nmol/min/mg microsomal protein
NADPH Cytochrome c 4 100 2840 280
DCPIPa 20 NDb 1800 75
Ferricyanide 100 ND b 3940 220
NADH Cytochrome c 350 60 270 290
DCPIP 320 330 220 310
Ferricyanide 600 ND b 900 700

a DCPIP, dichlorophenolindophenol.
b ND, not determined.

Analysis of the coupling between any ATR and a plant P450 requires coexpression in yeast. Genomic integration of the ATR1 and ATR2-1 expression cassettes at the CPR1 locus was considered. For this purpose, ATR1 and ATR2-1 coding sequences were cloned into pYeDP110 integration vector, yielding pATR1/V110 and pATR2-1/V110. The pYeDP110 vector places the heterologous coding sequence under the transcriptional control of GAL10-CYC1 promoter and CPR1 terminator and provides sequences from both flanking regions of the CPR1 gene that direct the integration at this locus, resulting in the deletion of most of the CPR1 gene. The ATR1 and ATR2-1 coding sequences were integrated at the CPR1 locus, resulting in two new yeast strains, WAT11 and WAT21, respectively. PCR analysis and Southern blotting confirmed that the full ATR expression cassette has been integrated within the CPR1 locus of each selected clone (not shown).

For each WAT strain, a uracil auxotrophic mutant was selected on 5-fluoroorotate (23) and named WAT11U and WAT21U. When cultivated on glucose, WAT cells exhibit no microsomal CPR activity and are phenotypically equivalent to WRDelta cells. But when grown on galactose, WAT11U and WAT21U cells express, respectively, microsomal ATR1 and ATR2-1 from the integrated expression cassette. Whether on glucose or in galactose, WAT cells express no yeast CPR. The generation time on galactose of the engineered strains was compared with that of the W(R) strain, which overexpresses yeast CPR, and no difference was observed. On the other hand, WAT11U and W(R) cells grown on glucose have identical generation times, significantly longer than upon growth on galactose.

WAT11U and WAT21U cells were grown on galactose to induce ATR expressions. The microsomal NADPH-cytochrome c reductase activity was measured (Table IV). A single integrated copy of the ATR1 expression cassette in WAT11 allows a CPR expression level similar to the value observed with a multicopy plasmid. In contrast, the reductase activity is reduced about 10-fold in WAT21, as compared with the plasmid-based expression, reaching a figure similar to WAT11. This suggests that the observed difference between ATR2-1 and ATR2-2 expression levels on multicopy plasmid might be related to different plasmid copy numbers and not to intrinsic properties of these two CPRs. All of these strains are fully isogenic, with the exception of the modified CPR1 locus. They can thus be used for comparing the effect of the CPR nature on any P450 catalytic properties.

Table IV. Reconstitution of Arabidopsis cinnamate 4-hydroxylase activity in yeast

The yeast strains were constructed as described under "Experimental Procedures." For bioconversions, the rate is expressed in molecules of the substrate transformed per min per cell. The microsomal fractions used for the in vitro assays were isolated from cells grown at high (YPGE/Gal) or low (SLI) cell density. Experiments were carried out in 50 mM Tris-HCl buffer, pH 7.4, and 37 °C. Protein concentration of pATCA4HV60-transformed yeast microsomes was 15-30 µg/ml depending on the assay. Data represent mean calculated from three independent clones. Turnover numbers for CYP73A5-catalyzed reactions are expressed in nmol/min/nmol P450.

Strain Type of CPR Culture medium CPR specific activity Bioconversion rates of cinnamate Kinetic parameters
2-Naphtoate (Km) Cinnamate
km kcat

µM µM min-1
W(R) Yeast SLI 3000 3700 1.5  ± 0.3 8  ± 2 470  ± 30
Yeast YPGE/Gal 2500 3100 NDa ND 420  ± 30
WAT11 ATR1 SLI 200 2400 1.1  ± 0.2 8  ± 2 200  ± 20
ATR1 YPGE/Gal 150 3100 ND a ND 160  ± 30
WAT21 ATR2-1 SLI 250 3300 1.1  ± 0.3 10  ± 3 250  ± 20
ATR2-1 YPGE/Gal 190 4100 ND a ND 150  ± 30

a ND, not determined.

Coupling Arabidopsis CPR and Cinnamate 4-Hydroxylase Activities in Yeast

The CYP73A5 coding sequence was PCR-amplified and cloned in pYeDP60. The resulting expression vector, pCYP73A5/V60, was used to transform W(R), WAT11U, and WAT21U. Coexpression of cinnamate 4-hydroxylase and Arabidopsis or yeast CPRs was initiated by galactose induction, and bioconversion was assayed by adding cinnamic acid directly into the culture medium. Rapid accumulation of p-coumarate was observed (see Table IV). In contrast, no detectable product formation was observed with the same strains transformed by a cDNA-free pYeDP60 vector. The rate of bioconversion was found to be roughly identical in the three tested strains, suggesting that the three CPRs similarly support CYP73A5 in vivo activity in yeast.

Microsomal fractions of pCYP73A5/V60-transformed W(R), WAT11U, and WAT21U cells were collected, and carbon monoxide-induced difference spectra were recorded on dithionite-reduced yeast microsomes suspensions. Spectrally detectable P450 content was found to be approximately 200 pmol of CYP73A5/mg of microsomal protein in W(R), 100 pmol/mg for WAT11 and WAT21 strains, and undetectable (<10 pmol/mg) in control. These values compare well with those usually observed in microsomes of yeast expressing mammalian P450s. The addition of cinnamate to microsomes from CYP73A5-expressing W(R), WAT11, and WAT21 cells causes a spectral shift with a differential absorption peak centered at 389 nm and a trough at 423 nm (not shown), typical of a ligand-induced low spin to high spin transition of the P450 heme ferric iron.

Microsomes from CYP73A5-expressing yeast cells were assayed for NADPH-dependent CYP73A5-catalyzed cinnamate and naphthoate hydroxylations. These reactions were found to be strictly NADPH-dependent and were not supported by NADH. Table IV lists the apparent kinetic parameters at steady state for the two substrates of the recombinant CYP73A5 in each of the three microsomal redox environments. Whatever the nature of the CPR, yeast or Arabidopsis, the turnover of the recombinant CYP73A5 was very high, reaching 470 min-1 in W(R). Although the level of NADPH cytochrome c reductase activity was 15-fold lower in WAT11 as compared with W(R), the cinnamate hydroxylase turnover number was only reduced 2.3-fold. Intermediate results between these two cases were obtained with WAT21. Km values of the P450 for the two substrates tested (cinnamate and naphthoate) were found to be identical whatever the strain (W(R), WAT11, or WAT21). These results suggest that ATR2-1 and ATR1 are similarly efficient in supporting Arabidopsis CYP73A5 activity when coexpressed in yeast.


DISCUSSION

In this work, two Arabidopsis cDNAs, whose products increase the resistance of CPR-deficient yeast cells to the antifungal drug ketoconazole and reestablish a physiological expression level of yeast endogenous P450s, have been isolated. These cDNAs encode two distantly related (64% amino acid residue identity) NADPH-P450 reductases, ATR1 and ATR2, which are similarly found in the ER membrane when expressed in yeast. The two ATRs efficiently support the activity of P450 cinnamate 4-hydroxylase from the same plant when coexpressed in yeast. The use of a cloning procedure by metabolic interference allowed us to overcome the strict requirement of sequence similarity, which is critical for cDNA cloning by standard hybridization strategies. Southern blot analysis and sequence comparison in flanking regions of the cDNAs demonstrated that ATR1 and ATR2 correspond to two distinct genes, each existing as a single copy in the Arabidopsis genome. These CPRs are not close allelic variants but are encoded by distantly related genes as shown by the amino acid sequence divergence and the third base randomization in conserved codons. Similar cases have been found in several other Arabidopsis gene families and in particular with phenylalanine ammonia lyases, phytochromes, and chalcone synthases (35-39).

The presence of multiple CPR genes in plants was also suggested by the multiplicity of bands detected in Western blots of CPR purified to homogeneity from Helianthus (11), sweet potato (10), and mung bean (27) and by the multiplicity of CPR peaks eluting from the 2',5'-ADP-Sepharose column when purifying CPRs from periwinkle (28), tulip bulbs (33), and petunia flowers (34) with, in the latter case, an absence of antigenic cross-reaction between the proteins eluted in two peaks. In Helianthus, two partial length cDNAs probably encoding distinct CPRs have recently been cloned (12), but the phylogenetic comparison carried out in this work demonstrates that they cluster in the same plant CPR subfamily together with ATR2. In contrast, ATR1 belongs to the other plant CPR subfamily, clustering with the unique mung bean CPR sequence characterized to date. The fact that ATR2 appears to be more similar to Helianthus HTR1 and HTR3 than to ATR1 suggests that a parent common to both ATR genes would have existed before the separation of Arabidopsis (Brassicaceae) and Helianthus (Asteraceae). Arabidopsis is thus the first organism in which the cloning and function of CPR genes of two distinct subfamilies are achieved. In contrast, a single CPR-encoding gene has been systematically reported in lower eukaryotes, like fungi, and in mammals. This raises the question of the physiological role for two distant CPRs in higher plants. A hypothesis would be a differential expression of each ATR either during various stages of plant development or in different plant tissues. There are several known cases of multiple gene families that encode protein isoforms with different tissue localizations in Arabidopsis such as potassium channels isoforms (40), zinc finger proteins (41), and the myosin gene family (42). There are also examples of plant protein isoforms encoded by multiple gene families, which present differential expressions during development of the plant seedling such as maize manganese superoxide dismutases (43) and, of course, the Arabidopsis floral homeotic MADS-box regulatory gene family (44). In other cases, the members of a gene family exhibit both differential temporal and tissue-specific regulations, such as the Arabidopsis gibberellin aldehyde-induced transcripts (45).

Alignments reveal the presence in the two ATRs of all expected conserved FMN-, FAD- and NADPH-binding domains and thus are new members of the ferredoxin-NADP+ reductase family of proteins (30, 31, 46). A very efficient cinnamate bioconversion is observed by co-expression in yeast of ATR1 or ATR2 and Arabidopsis CYP73A5. ATR1, ATR2, and yeast CPR each support rather similar CYP73A5 turnover numbers for cinnamate. Based on these data, the two ATRs would have virtually exchangeable roles. Nevertheless, the relative molar ratio between CPR and P450 expressed in yeast is 10-fold lower in the case of ATR expressions than in the case of yeast CPR in W(R) cells. The apparent equivalence between ATR1 and ATR2 to support cinnamate hydroxylase activity in yeast microsomes might thus be related to a saturation of CYP73A5 activity by high level of CPR. The W(R) or WAT engineered strains permit us to obtain turnover numbers for Arabidopsis and Helianthus CYP73s about 1000-fold higher than the value reported for Medicago CYP73 expressed in wild-type yeast (47). Similarly, expression in E. coli of a functional fusion protein consisting of the Catharanthus CYP73 fused at the N terminus of Catharanthus CPR, was recently reported (48). However, the CYP73 specific activity in this case was found 40-fold lower than in the present work. Functional analysis of the large number of plant P450-encoding cDNAs reported to date is thus expected to take advantage of the engineered WAT strains.

Almost no amino acid sequence similarity is found between the 100 first residues of ATR1 and ATR2-1, except for the presence of the hydrophobic stretch that is critical for the binding of CPRs to microsomal membrane (49). In ATR2-1, this stretch is preceded by a poly(Ser/Thr) N-terminal extension and an unusually long amphipatic sequence segment. This motif, which has no equivalent in any other fungal, plant, or animal CPRs, is highly reminiscent of a chloroplastic targeting sequence (50, 51), thus questioning the authentic subcellular location of ATR2-1 in plants. The observed addressing in yeast could thus be artifactual due to the absence of the plant-specific organites. The presence of P450s in chloroplast still remains debatable, but very recently it has been shown that Linum allene oxide synthase, CYP74, presents a typical N-terminal chloroplastic transit peptide (52). Moreover, it is highly suggested that the gibberellin biosynthesis pathway, which produces phytohormones, is localized at least partially in the chloroplast (Ref. 53 and references therein). ent-kaurene synthesis, from which all gibberellins derive, has been clearly established to take place in the chloroplast (54, 55). Nevertheless, the subcellular location in plants of the P450-catalyzed steps of ent-kaurene and gibberellin oxidations is still debated (53, 56, 57). This leaves open the hypothesis that in Arabidopsis chloroplasts could house some P450-catalyzed reactions with ATR2-1 as a chloroplastic CPR. In this hypothesis, ATR1 would be the Arabidopsis microsomal counterpart. A similar, but clearly established situation, has been found with Arabidopsis signal recognition particles. Distinct genes have been shown to encode two SRP54p isoforms, one cytosolic (58) and the other chloroplastic (59). This raises the question of the occurrence in plants of chloroplast-specific CPR and P450s whose physiological roles remain to be fully investigated.


FOOTNOTES

*   The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) X66016 (for ATR1), X66017 (for ATR2), and U37235 (for CYP73A5).


par    A fellow from Rhône-Poulenc-Rorer (Bio Avenir).
**   To whom correspondence should be addressed: Centre de Génétique Moléculaire du CNRS, UPR 9061, Avenue de la Terrasse, 91198 Gif-sur-Yvette, France. Tel.: 33-1-6982-3181; Fax: 33-1-6907-5539; E-mail: pompon{at}cgm.cnrs-gif.fr.
1   The abbreviations used are: CPR, NADPH-cytochrome P450 reductase; kb, kilobase pair; bp, base pair; CPR1, yeast gene encoding microsomal CPR; HTR1, H. tuberosus CPR1; SCR1, S. cerevisiae CPR; HSR1, Homo sapiens CPR; ER, endoplasmic reticulum; P450, hemethiolate protein P450; PGK, phosphoglycerate kinase; PCR, polymerase chain reaction; ETBI, experimental third base identity; RTBI, reference third base identity.

ACKNOWLEDGEMENTS

We thank Philippe Fusier and Cécile Le Dû for excellent technical assistance. We especially thank Drs. Danièle Werck-Reichhart and Francis Durst for fruitful discussions.


REFERENCES

  1. Porter, T. D., and Coon, M. J. (1991) J. Biol. Chem. 266, 13469-13472 [Free Full Text]
  2. Gonzalez, F. J., and Gelboin, H. V. (1994) Drug Metab. Rev. 26, 165-183 [Medline] [Order article via Infotrieve]
  3. Sanderman, J. H., Jr. (1994) Pharmacogenet. 4, 225-241 [Medline] [Order article via Infotrieve]
  4. Hallahan, D. L., Cheriton, A. K., Hyde, R., and Forde, B. G. (1993) Biochem. Soc. Trans. 21, 1068-1073 [Medline] [Order article via Infotrieve]
  5. Plewa, M. J., and Wagner, E. D. (1994) Annu. Rev. Genet. 27, 93-113 [CrossRef][Medline] [Order article via Infotrieve]
  6. Durst, F., and Benveniste, I. (1993) in Handbook of Experimental Pharmacology (Shenckman, J. B., and Greim, H., eds), Vol. 105, pp. 293-310, Spinger-Verlag, Berlin
  7. Bolwell, G. P., Bozak, K., and Zimmerlin, A. (1994) Phytochem. 37, 1491-1506 [CrossRef]
  8. Simmons, D. L., Lalley, P. A., and Kasper, C. B. (1985) J. Biol. Chem. 260, 515-521 [Abstract/Free Full Text]
  9. Porter, T. D., Beck, T. W., and Kasper, C. B. (1990) Biochemistry 29, 9814-9818 [Medline] [Order article via Infotrieve]
  10. Fujita, M., and Asahi, T. (1985) Plant Cell Physiol. 26, 397-405
  11. Benveniste, I., Lesot, A., Hasenfratz, M. P., Kochs, G., and Durst, F. (1991) Biochem. Biophys. Res. Commun. 177, 105-112 [Medline] [Order article via Infotrieve]
  12. Lesot, A., Hasenfratz, M. P., Batard, Y., Durst, F., and Benveniste, I. (1995) Plant Physiol. Biochem. 33, 751-757
  13. Yoshida, Y., and Aoyama, Y. (1991) Biochem. Soc. Trans. 19, 778-782 [Medline] [Order article via Infotrieve]
  14. Sutter, T. R., and Loper, J. C. (1989) Biochem. Biophys. Res. Commun. 160, 1257-1266 [Medline] [Order article via Infotrieve]
  15. Truan, G., Epinat, J. C., Rougeulle, C., and Pompon, D. (1994) Gene (Amst.) 149, 123-127 [CrossRef][Medline] [Order article via Infotrieve]
  16. Minet, M., Dufour, M. E., and Lacroute, F. (1992) Plant J. 2, 417-422 [CrossRef][Medline] [Order article via Infotrieve]
  17. Pompon, D., Louerat, B., Bronine, A., and Urban, P. (1996) Methods Enzymol. 272, 51-64 [CrossRef][Medline] [Order article via Infotrieve]
  18. Teutsch, H. G., Hasenfratz, M. P., Lesot, A., Stoltz, C., Garnier, J. M., Jeltsch, J. M., Durst, F., and Werck-Reichhart, D. (1993) Proc. Natl. Acad Sci. U. S. A. 90, 8519-8523 [Abstract/Free Full Text]
  19. Urban, P., Cullin, C., and Pompon, D. (1990) Biochimie 72, 463-472 [CrossRef][Medline] [Order article via Infotrieve]
  20. Cullin, C., and Pompon, D. (1988) Gene (Amst.) 65, 203-217 [CrossRef][Medline] [Order article via Infotrieve]
  21. Gietz, D., St. Jean, A., Woods, R. A., and Schiestl, R. H. (1992) Nucleic Acids Res. 20, 1425 [Medline] [Order article via Infotrieve]
  22. Urban, P., Werck-Reichhart, D., Teutsch, H. G., Durst, F., Regnier, S., Kazmaier, M., and Pompon, D. (1994) Eur. J. Biochem. 222, 843-850 [Abstract]
  23. Boeke, J., Lacroute, F., and Fink, G. R. (1984) Mol. Gen. Genet. 197, 345-354 [Medline] [Order article via Infotrieve]
  24. Omura, T., and Sato, R. (1964) J. Biol. Chem. 239, 2370-2378 [Free Full Text]
  25. Cavener, D. R., and Ray, S. C. (1991) Nucleic Acid Res. 19, 3185-3192 [Abstract]
  26. Porter, T. D., and Kasper, C. B. (1985) Proc. Natl. Acad. Sci. U. S. A. 82, 973-977 [Abstract]
  27. Shet, M. S., Sathasivan, K., Arlotto, M. A., Mehdy, M. C., and Estabrook, R. W. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 2890-2894 [Abstract]
  28. Meijer, A. H., Cardoso, M. I. L., Voskuilen, J. T., de Waal, A., Verpoorte, R., and Hoge, J. H. C. (1993) Plant J. 4, 47-60 [CrossRef][Medline] [Order article via Infotrieve]
  29. Porter, T. D., and Kasper, C. B. (1986) Biochemistry 25, 1682-1687 [Medline] [Order article via Infotrieve]
  30. Karplus, P. A., Daniels, M. J., and Herriott, J. R. (1991) Science 251, 60-66 [Medline] [Order article via Infotrieve]
  31. Karplus, P. A., and Bruns, C. M. (1994) J. Bioenerg. Biomembr. 26, 89-99 [Medline] [Order article via Infotrieve]
  32. Henikoff, S., and Henikoff, J. G. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 10915-10919 [Abstract]
  33. Hansikova, H., Frei, E., Anzenbacher, P., and Stiborova, M. (1994) Gen. Physiol. Biophys. 13, 149-169 [Medline] [Order article via Infotrieve]
  34. Menting, J. G. T., Cornish, E., and Scopes, R. K. (1994) Plant Physiol. 106, 643-650 [Abstract/Free Full Text]
  35. Huang, S., McDowell, J. M., Weise, M. J., and Meagher, R. B. (1996) Plant Physiol. 111, 115-126 [Abstract/Free Full Text]
  36. Sharrock, R. A., and Quail, P. H. (1989) Genes & Dev. 3, 1745-1757 [Abstract]
  37. Shirley, B. W. (1996) Trends Plant Sci. 1, 377-382 [CrossRef]
  38. Akada, S., and Dube, S. K. (1995) Plant Mol. Biol. 29, 189-199 [Medline] [Order article via Infotrieve]
  39. Durbin, M. L., Learn, G. H., Huttley, G. A., and Clegg, M. T. (1995) Proc. Natl. Acad. Sci. U. S. A. 92, 3338-3342 [Abstract]
  40. Cao, Y., Ward, J. M., Kelly, W. B., Ichida, A. M., Gaber, R. F., Anderson, J. A., Uozumi, N., Schroeder, J. I., and Crawford, N. M. (1995) Plant Physiol. 109, 1093-1106 [Abstract/Free Full Text]
  41. Tague, B. W., and Goodman, H. M. (1995) Plant Mol. Biol. 28, 267-279 [Medline] [Order article via Infotrieve]
  42. Kinkema, M., Wang, H., and Schiefelbein, J. (1994) Plant Mol. Biol. 26, 1139-1153 [Medline] [Order article via Infotrieve]
  43. Zhu, D., and Scandalios, J. G. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 9310-9314 [Abstract]
  44. Purugganan, M. D., Rounsley, S. D., Schmidt, R. J., and Yanofsky, M. F. (1995) Genetics 140, 345-356 [Abstract/Free Full Text]
  45. Herzog, M., Dorne, A. M., and Grellet, F. (1995) Plant Mol. Biol. 27, 743-752 [Medline] [Order article via Infotrieve]
  46. Shen, A. L., and Kasper, C. B. (1993) in Handbook of Experimental Pharmacology (Schenckman, J. B., and Greim, H., eds), Vol. 105, pp. 35-59, Springer-Verlag, Berlin
  47. Fahrendorf, T., and Dixon, R. A. (1993) Arch. Biochem. Biophys. 305, 509-515 [CrossRef][Medline] [Order article via Infotrieve]
  48. Hotze, M., Schröder, G., and Schröder, J. (1995) FEBS Lett. 374, 345-350 [CrossRef][Medline] [Order article via Infotrieve]
  49. Black, S. D., and Coon, M. J. (1982) J. Biol. Chem. 257, 5929-5938 [Abstract/Free Full Text]
  50. von Heijne, G., Steppuhn, J., and Herrmann, R. G. (1989) Eur. J. Biochem. 180, 535-545 [Abstract]
  51. Douwe de Boer, A., and Weisbeek, P. J. (1991) Biochim. Biophys. Acta 1071, 221-253 [Medline] [Order article via Infotrieve]
  52. Song, W. C., Funk, C. D., and Brash, A. R. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 8519-8523 [Abstract/Free Full Text]
  53. Hedden, P., MacMillan, J., and Phinney, B. U. (1978) Annu. Rev. Plant Physiol. 29, 149-192
  54. Moore, T. C., and Coolbaugh, R. C. (1976) Phytochemistry 15, 1241-1247 [CrossRef]
  55. Bensen, R. J., Johal, G. S., Crane, V. C., Tossberg, J. T., Schnable, P. S., Meeley, R. B., and Briggs, S. P. (1995) Plant Cell 7, 75-84 [Abstract/Free Full Text]
  56. Murphy, G. J. P., and Briggs, D. E. (1975) Phytochemistry 14, 429-433 [CrossRef]
  57. Graebe, J. E. (1987) Annu. Rev. Plant Physiol. 38, 419-465 [CrossRef]
  58. Lindstrom, J. T., Chu, B., and Belanger, F. C. (1993) Plant Mol. Biol. 23, 1265-1272 [Medline] [Order article via Infotrieve]
  59. Franklin, A. E., and Hoffman, N. E. (1993) J. Biol. Chem. 268, 22175-22180 [Abstract/Free Full Text]
  60. Pompon, D., Urban, P., Truan, G., Cullin, C., and Slonimski, P. P. (1993) Patent 93-058796/07
  61. Mignotte, C., Kazmaier, M., Lacroute, F., Minet, F., and Pompon, D. (1992) Patent D13969/233445
  62. Mignotte, C., Kazmaier, M., and Pompon, D. (1993) Patent FR92082454

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.