(Received for publication, February 13, 1997, and in revised form, March 31, 1997)
From the § Frontier Research Program and the
Primary structural information of a plant
aldehyde oxidase (AO), which was purified from maize coleoptiles using
indole-3-acetaldehyde as a substrate, was obtained by sequencing a
series of cleavage peptides, permitting the cloning of the
corresponding cDNA (zmAO-1). The complete nucleotide sequence was
determined; the deduced amino acid sequence encodes a protein of 1358 amino acid residues of Mr 146,681, which is
consistent with the size of the AO monomeric subunit. There is a
significant similarity with AO from mammals and xanthine dehydrogenases
from various sources. The maize AO polypeptide contains consensus
sequences for iron-sulfur centers and a putative molybdopterin
cofactor-binding domain. In addition, another cDNA (zmAO-2), highly
homologous to zmAO-1 at both the nucleotide and amino acid sequence
levels, was cloned. zmAO-2 would encode a protein of 1349 amino acid
residues of Mr 145,173 and has molecular
characteristics similar to those of zmAO-1. zmAO-1 was expressed at a
high level in roots, which was closely correlated with immunoblotting
results using antiserum raised against the purified maize AO protein,
whereas zmAO-2 was expressed at a higher level in coleoptiles than in
roots. We propose each zmAO may have a unique function during plant
development.
Aldehyde oxidases (aldehyde-oxygen oxidoreductase; EC 1.2.3.1;
AO)1 are ubiquitous enzymes that have been
extensively investigated in animals and microorganisms. They catalyze
the oxidation of a variety of aldehydes and N2-containing
heterocyclic compounds in the presence of O2 or certain
redox dyes (1-3). AOs belong to the family of molybdenum-containing
protein like xanthine oxidoreductase (xanthine:NAD+
oxidoreductase; EC 1.1.1.204; XD) and sulfite oxidase (4), and they
possess a short oxidoreductive chain characterized by four oxidation
centers, such as two iron-sulfur clusters, a flavin cofactor, and a
molybdopterin cofactor (5). AOs are known to be a homodimer, consisting
of two 148-kDa subunits (see Ref. 6 and references therein).
Recently, an AO was purified from bovine liver and the corresponding
cDNA was cloned (7). Consensus sequences for iron-sulfur centers
and a molybdopterin-binding site in the deduced polypeptide were
identified. The deduced amino acid sequence shows significant similarity with that of XD from various sources. Bovine AO is expressed
at high levels in the liver and lung and has been implicated in the
detoxification of certain types of environmental pollutants and
xenobiotic. The animal AO may also play a role in retinoic acid
synthesis (7-9). In addition, a human liver cDNA that had been
thought to encode XD is now thought to encode an AO (10).
In plants, only a limited amount of information has been published
concerning AO, being focused on its possible involvement in plant
hormone biosynthesis, such as indole-3-acetic acid (IAA) (11-14) and
abscisic acid (ABA) (15, 16). The last step in their biosynthetic
pathways has been thought to be the oxidation of corresponding
aldehyde: indole-3-acetaldehyde (IAAld) for IAA and abscisic aldehyde
(ABAld) for ABA. However, there have been almost no detailed analyses
of the molecular properties of plant AO, and no sequence data of either
protein or DNA have been reported. Therefore, the actual function of
the plant AO(s) is still obscure. In a previous paper, we described the
purification of an AO that could oxidize IAAld to IAA from coleoptiles
of maize (Zea mays L.) seedlings (17). The maize AO has an
apparent molecular mass of about 300 kDa as estimated by gel filtration
and was composed of two 150-kDa subunits. It contains flavin adenine
dinucleotide (FAD), iron, and molybdenum as prosthetic groups. This was
the first report showing the presence of these cofactors in a plant AO.
In the present paper, we describe the sequencing of peptides obtained
by proteolytic cleavage of the purified maize AO and the molecular
cloning of the respective cDNA. In addition, another cDNA
showing significant similarity with cDNA of the AO was cloned. The
structures of both deduced polypeptides were analyzed and compared with
mammalian AOs and various XDs. These cDNA probes and a monospecific
antiserum raised against the purified enzyme were used to study the
tissue distribution of maize AO mRNA and protein.
Seeds of maize (Z. mays L. cv
Golden Cross Bantam 70) were germinated, and coleoptiles (about 1.5 cm
long) were harvested from 4-day-old seedlings (18).
AO was purified from maize coleoptile tips as described
previously (17) and subjected to SDS-polyacrylamide gel electrophoresis for final purification. Coomassie-stained bands of 150 kDa were excised
and treated with 0.3 µg of Achromobacter protease I (API; a gift from Dr. Masaki, Ibaraki University (19)) at 37 °C for 12 h in 0.1 M Tris-HCl (pH 9.0) containing 0.1% SDS.
Peptides generated were extracted from the gel and separated on columns of DEAE-5PW (2 × 20 mm; Tosoh, Tokyo) and Supersphere RP-select B
(2 × 120 mm; Tosoh) connected in series with a model 1090M
(Hewlett Packard) liquid chromatography system. Peptides were eluted at a flow rate of 0.2 ml min Table I.
Amino acid sequence and mass values of selected API peptides of maize
AO
Division of Biomolecular Characterization,
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
Addendum
REFERENCES
Plant Material
1 using a linear gradient of
0-60% solvent B, where solvents A and B were 0.09% (v/v) aqueous
trifluoroacetic acid and 0.075% (v/v) trifluoroacetic acid in 80%
(v/v) acetonitrile, respectively. Selected peptides were subjected to
Edman degradation using a model 477A automated protein sequencer
(Applied Biosystems, Inc.) connected on-line to a model 120A
PTH-Analyzer (Perkin Elmer) and to matrix-assisted laser desorption
ionization time of flight mass spectrometry on a Reflex MALD-TOF
(Bruker-Franzen Analytik, Bremen, Germany) in linear mode using
-cyano-4-hydrocinnamic acid as a matrix. Amino acid sequences of
four peptides and mass values of four peptides (peptides 1-4) are
shown in Table I.
Peptide sequence
Observed
mass (H+)
Calculated mass (H+)
1.
SIEELHRLFDSSWFDDSSVK
2400.0
2398.6
2.
QVNDVPIAASGDGWYHPK
1956.2
1955.1
3.
TNSDGLVIHDGTWTYK
1809.8
1807.9
4.
VGAEIQASGEAVYVDDIPAPK
2131.9
2130.3
The amino acid sequence information from
API peptide 4 was used for synthesis of the sense oligonucleotide
primer (5-GGIGA(A/G)GCIGTITA(T/C)GTIGA(T/C)GA-3
) corresponding to
amino acids 9-16. The antisense primer
(5
-GTCCAIGTICC(A/G)TC(A/G)TGIATIAC-3
) was derived from positions
7-14 of peptide 3. RNA was extracted from maize coleoptile tips
according to standard procedures, and the latter primer served to
reverse transcribe the total RNA. The resulting cDNA was 100-fold
diluted and amplified by polymerase chain reaction (94 °C for 1 min,
48 °C for 2 min, and 72 °C for 3 min) in the presence of both the
sense and antisense primers. Fifty cycles of amplification produced
cDNA of 1959 bp that was cloned into the pCRII vector using the TA
cloning kit (Invitrogen). The inserts were sequenced in both directions
using double-stranded DNA as templates with a Taq dye primer
cycle sequencing kit (Perkin Elmer). Nucleotide sequences of inserts
from 11 independent clones demonstrated the presence of at least two
kinds of cDNA fragments ("first PCR" in Fig. 1) for
maize AO, both encoding amino acid sequences that match those of the
peptides.
Full-length cDNAs were obtained using the Marathon cDNA
amplification kit (CLONTECH) in accordance with the
supplier's instructions (Fig. 1). In addition, oligonucleotide
5-GAGTATAGCACAGAAAATCTGCAGCCA-3
served to reverse transcribe the
poly(A)+ RNA. The first 5
extended cDNAs
("RACE-PCR(1)" for zmAO-1 and -2) were obtained by the
amplification with the oligonucleotides 5
-TGCTTTGCAGCCATATTAGCATATCTT-3
and 5
-CTTTGCCGCCATGTAGGCATACTTC-3
identified in the nucleotide sequences of first PCR
for zmAO-1 and -2, respectively. The second cDNAs
("RACE-PCR(2)" for zmAO-1 and -2) were obtained with the
oligonucleotides 5
-ACAGCCTTTTGGAAGCCACCTGGA-3
and
5
-TTCCACCTATGGTTGCAGTGTTCC-3
synthesized on the basis of the
nucleotide sequences of RACE-PCR(1) for zmAO-1 and -2, respectively. The final extended cDNA ("RACE-PCR(3)"
for zmAO-1) were obtained with the oligonucleotide
5
-ATCGGACTTGTTGTCGGCCTTGAC-3
on the basis of the sequence of
RACE-PCR(2) for zmAO-1. The 3
region of the cDNAs
(zmAO1-5 and zmAO2-4) were obtained by the amplification with
oligonucleotides 5
-GATTGCTGAAACACAAAGATATGCTAAT-3
and
5
-GATTGCTCAAACACAGAAGTATGCCTAC-3
identified in the sequences of
first PCR for zmAO-1 and -2, respectively. Each clone was
sequenced in both directions using either vector primers or specific
oligonucleotides synthesized based on the basis of information obtained
from previous sequence analysis. Full-length cDNA sequences were
obtained after fusion of corresponding 5
- and 3
-RACE fragments in
accordance with the supplier's instructions. Computer analysis of the
DNA sequence was performed with the Genetyx software package (Software
Development. Co., Tokyo).
Genomic DNA was prepared from maize seedlings according to the method of Shure et al. (20). The DNA was digested with EcoRI or HindIII, fractionated on a 0.7% (w/v) agarose gel, denatured, and transferred to a Hybond-N+ (Amersham) nylon membrane. Hybridization was carried out according to standard conditions (21), using 32P-radiolabeled full length zmAO-1 or zmAO-2 cDNA as probes. Northern blot analysis was also performed (21), using same probes as Southern hybridization.
Native Polyacrylamide Gel Electrophoresis and Immunoblot AnalysisFreshly isolated maize organs were homogenized in 2 volumes of 50 mM Tris-HCl (pH 7.5) buffer containing 1 mM EDTA, 2 mM dithiothreitol, 10 µM FAD, 1 µM sodium molybdate, and 5 µM leupeptin, and the homogenates were centrifuged at 12,000 × g for 20 min. The supernatants were collected and subjected to native polyacrylamide gel electrophoresis followed by immunoblotting. Native polyacrylamide gel electrophoresis was performed with a 7.5% acrylamide gel in Laemmli buffers (22) in the absence of SDS at 4 °C. Immunoblotting was performed using anti-AO mouse serum (17).
Amino-terminal sequencing of the maize AO protein failed.
This might be due to blocking as a result of unknown post-translational modifications, as in the case of bovine AO (7). Partial sequence information for maize AO was obtained from protein subjected to cleavage with API. Of many peptides separated by reversed phase chromatography, the four most prominent fragments were sequenced, allowing the cloning of cDNA fragments corresponding to maize AO
mRNA. Polymerase chain reaction amplification with a combination of
two degenerate oligonucleotides as primers produced two kinds of
1959-bp cDNA fragments (Fig. 1). Comparison of the
deduced amino acid sequences of these cDNAs to the amino acid
sequences of the corresponding peptides indicated that both cDNA
fragments might encode part of AO subunits. The sequences of the
remaining 5- and 3
-stretches were established by the RACE-polymerase
chain reaction technique using nucleotide sequence information from the
respective fragments (Fig. 1). The complete sequence of a cDNA was
determined from the sequence of several independent clones obtained by
5
- and 3
-RACE procedures with oligonucleotide primers to complete the
sequences as required.
One nearly full-length cDNA (zmAO-1) exhibits a 45-bp 5
untranslated region, followed by a 4074-nucleotide open reading frame and a 263-nucleotide 3
-untranslated sequence (Fig. 2).
Three presumptive polyadenylation site sequences (AATAA) are observed (double underlined) in the 3
-untranslated region. The open
reading frame of the cDNA predicts a protein of 1358 amino acids of
Mr 146,681 and a translation product that
contains the sequence of the four peptides obtained by API cleavage of
the AO protein (see also Table I).
The other nearly full-length cDNA (zmAO-2) consists of a 90-bp
5-untranslated region, followed by a 4047-nucleotide open reading
frame and a 196-nucleotide 3
-untranslated sequence (data not shown).
Two presumptive polyadenylation site sequences (AATAA) are observed in
the 3
-untranslated region. The open reading frame of the cDNA
predicts a protein of 1349 amino acids of Mr
145,173 and a translation product that contains a sequences identical to peptide 3, and similar but not identical to peptides 1, 2, and 4 obtained by API cleavage of the AO protein. These cDNAs are highly
homologous in respect to both nucleotide (83.8% identity) and amino
acid (83.6% identity), excepting the nucleotide sequences of their
presumptive 5
- and 3
-untranslated regions.
zmAO
proteins have no sequences that encode presumptive hydrophobic signal
peptides, indicating they are cytosolic enzymes. One of two [2Fe-2S]
centers is found between amino acid residues 50 and 80 for zmAO-1
protein and between 46 and 76 for zmAO-2 protein, respectively (Fig.
3). This iron-sulfur center is of the same type observed
in ferredoxin from a number of photosynthetic organisms, in bacterial
fumarate reductase and in eukaryotic succinate dehydrogenase (23, 24).
The second putative iron-sulfur center is probably located between
amino acid residues 120-123 for zmAO-1 and 116-119 for zmAO-2,
respectively, since they are conserved in two other classes of
molybdoflavoproteins, such as bovine AO and several XDs (7, 10, 25).
Sequences conforming to the consensus for the molybdopterin cofactor
binding site reside between amino acids 760-853 (zmAO-1) and 752-845
(zmAO-2), respectively (Fig. 3). An FAD-binding consensus sequence of
the type described by Correll et al. (26) cannot be
precisely identified in either amino acid sequence. Consensus sequences
for binding of NAD are observed between amino acids 49-54 and 816-821
for zmAO-1 and 45-50 and 808-813 for zmAO-2, respectively (Fig. 3).
Despite their high homology overall, some regions that are difficult to
align occur between amino acids 451-453, 532-549, and 897-902 for
zmAO-1 and 447-454, 533-540, and 889-894 for zmAO-2, respectively
(Fig. 3).
A homology search in the SwissProt and EMBL data banks determined that the two zmAOs have a significant level of similarity with animal AOs and XDs from various sources (Fig. 3). The overall level of respective identities with zmAO-1 and zmAO-2 are 30.9 and 31.2% to bovine AO (7), 29.9 and 30.1% to human AO (initially reported as human XD) (10), 31.2 and 32.1% to human XD (27), 31.5 and 30.3% to Drosophila pseudoobscura XD (28), and 30.3 and 30.8% to Aspergillus nidulans XD (25).
Southern Blotting, Immunoblotting, and Northern BlottingGenomic Southern blotting showed different hybridization
patterns with the two cDNA probes employed for Fig.
4 under conditions of very high stringency with no
detectable cross-hybridization. Immunoblot analysis was performed on
crude extracts of various maize organs with anti-AO mouse antiserum
(Fig. 5). The highest levels of AO protein were observed
in the root tissue followed by the coleoptile and mesocotyl. Only a
faint immunoreactive band was detected in leaf tissues. Northern blot
experiments with RNA extracted from the same tissue preparations used
for immunoblotting were performed. zmAO-1 was expressed at high level
in the roots, whereas zmAO-2 was expressed at higher levels in
coleoptiles than in roots (Fig. 6).
We cloned two cDNAs for plant AO. The molecular mass values of predicted proteins for these cDNAs are coincident with that of the purified protein, and one of them (zmAO-1) contains the sequence of four peptides obtained by API cleavage of the purified AO protein. The pattern of tissue-specific expression of zmAO-1 (Fig. 6) is similar to that of the protein as determined by immunoblotting (Fig. 5). In addition, maize AO activity was immunoprecipitated with a combination of the antiserum raised against purified AO and protein A-Sepharose (17). We conclude that zmAO-1 encodes maize AO that had been purified previously. In the case of human AO, the predicted amino acid sequence shows a relatively lower level of homology to their XD (49.3%) than to bovine AO (85.7%). Since the deduced amino acid sequence for the other cDNA (zmAO-2) shows high similarity to zmAO-1 (83.6%), we considered that both encode maize AOs that may have related functions.
As shown in Fig. 3, predicted proteins for zmAOs have a significant level of similarity to those animal AOs and XDs and can be aligned along the whole length of the proteins. zmAOs exhibit unusual clustering of cysteines in the amino-terminal 200 amino acid residues. Nine of these 12 cysteines are conserved in both animal and plant AOs, and eight of them are conserved in both AOs and XDs. Many of them must contribute to the structure of the two Fe-S centers. Li Calzi et al. (7) suggested that one (or both) of the two cysteines, amino acids 149 and 151 for bovine AO located in a strictly conserved stretch of amino acids, were involved in iron ligation in AO and XDs. These cysteines and the conserved region are found in both zmAOs (amino acids 165-176 for zmAO-1), supporting this possibility. In addition, zmAOs also contain sequences that conform to the consensus for the molybdopterin cofactor binding site (amino acids 760-853 for zmAO-1). These results are helpful for confirmation of the previous report concerning the biochemical characters of purified maize AO (17). As shown in Fig. 3, there are some additional regions conserved to some extent among plant AOs, mammalian AOs, and XDs (amino acids 597-608, 814-822, 1053-1064, and 1173-1230 for zmAO-1). Thus, plant AOs are similar to related enzymes from phylogenically different organisms, indicating they evolved from a common ancestral gene. The most primitive animal species in which AO activity has been found is the coelenterate Segartia luciae (29).
Two consensus sequences for NAD binding are found in both zmAOs; however, AO does not require the cofactor for its catalytic activity and in fact purified maize AO is devoid of dehydrogenase activity. It is thus unlikely that these structural elements have functional significance. This is also the case for bovine AO (7).
Glatigny and Scazzocchio (25) suggested that the ERXXXH motif (amino acid residues 910-915 of A. nidulans XD) is involved in determining substrate specificity. These regions are not conserved in animal AO proteins and plant AOs. In addition, no significant similarities are observed in those corresponding regions among animal AO proteins and zmAO-1 and zmAO-2 proteins. This may reflect the different substrate specificities of plant and animal AOs, since the former have affinity for aromatic aldehydes while the latter do not. To analyze the substrate specificity of the two zmAOs, we tried to produce recombinant proteins, but they were almost completely degraded when the cDNA were expressed in Escherichia coli. This problem also occurs with another molybdopterin cofactor-containing protein, nitrate reductase.2 Further research will be required for elucidation of the different substrate specificities of the zmAOs.
In plants, AOs are thought to be involved in plant hormone biosynthesis. Molybdopterin cofactor-deficient mutants of barley and tobacco are deficient in AO and XD activities and have impaired ABA production (15, 16). This indicates that ABAld oxidase is a molybdenum-containing enzyme which is indispensable for ABA biosynthesis. The ABAld oxidase oxidizes ABAld to ABA, but such an enzyme has not been purified. In the case of IAA biosynthesis, the pathway has been extensively investigated (18, 30-32), but is still poorly understood. IAAld oxidase could catalyze the final step in the pathway where IAA is produced from tryptophan (11, 14, 17). However, this has only been studied in a few plants. Recently, a sur mutant that forms an abnormal number of roots has been isolated from Arabidopsis thaliana (33). It was thought that the phenotype was caused by overproduction of IAA, and, in fact, IAA levels are much higher in seedlings of the mutant. We have checked AO activity in the wild-type and mutant seedlings after native polyacrylamide gel electrophoresis, revealing that at least three AO activity bands. One band, showing a substrate preference for indole-3-aldehyde, is much more intense in the mutant seedlings. Three independent cDNA clones have also been isolated from A. thaliana using degenerate primers, and one of these is highly expressed in the sur mutant.3
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) D88451[GenBank] (zmAO-1) and D88452[GenBank] (zmAO-2).
We thank Prof. Russell Jones (University of California, Berkeley) and Prof. Dick Kendrick (Frontier Research Program, RIKEN) for their critical reviews of the manuscript. We also thank Y. Tachiyama for her technical assistance in nucleotide sequencing. We are grateful to Prof. Michel Caboche (INRA, Versailles) for useful discussions.
While this article was being submitted and reviewed, we found a paper describing a gene for a putative molybdenum cofactor-containing plant hydroxylase (Ori, N., Ested, Y., Pinto, P., Paran, I, Zamir, D., and Fluhr, R. (1997) J. Biol. Chem. 272, 1019-1025).