and chlorophyll synthesis enzyme 6-aminolevulinic acid (ALA) de- hydratase by .... Plus hemin No hemin min ... plant cytosol(4 versus 16 nmol h-' mg-' protein).
Plant Physiol. (1994) 104: 1411-1417
Plant 6-Aminolevulinic Acid Dehydratase’ Expression in Soybean Root Nodules and Evidence for a Bacterial Lineage of the Alad Gene Christine M. Kaczor, Michael W. Smith, lndu Sangwan, and Mark R. O’Brian* Department of Biochemistry, State University of New York at Buffalo, Buffalo, New York 14214 (C.M.K., I.S., M.R.O’B.); and Molecular Genetics Laboratory and Human Genome Center, The Salk Institute, San Diego, California 92186 (M.W.S).
proteins essential for symbiosis. Despite the unique context in which plant hemoglobin is expressed, aspects of heme precursor synthesis in soybean (Glycine max) nodules have features in common with tetrapyrrole formation described in leaves. Plants synthesize ALA from glutamate in leaves by the C5 pathway (reviewed by Beale and Weinstein, 1991; Jahn et al., 1992), and the C5 pathway enzyme GSA aminotransferase is expressed in root nodules (Sangwan and O’Brian, 1993). ALA formation activity is activated in nodules (Sangwan and O’Brian, 1991, 1992), and this is at least in part due to induction of the GSA aminotransferase gene (Gsa) (Sangwan and O’Brian, 1993). The soybean Gsa gene is strongly expressed in leaves of etiolated plantlets and is reduced in older leaves; thus Gsa is regulated in leaves as well. Synthesis of the leghemoglobin heme prosthetic group has been a topic of controversy, and it underscores the lack of information about tetrapyrrole synthesis by the plant host in nodules. The hypothesis that nodule hemoglobin heme is a bacterial product has been put forth (Cutting and Schulman, 1969), and that idea is based primarily on the lack of heme synthesis enzyme activities in plant fractions of nodules (Cutting and Schulman, 1969; Porra, 1975; Nadler and Avissar, 1977) and on the leghemoglobin-defective phenotype of nodules induced by some rhizobial heme synthesis mutants (Leong et al., 1982; O’Brian et al., 1987). The former argument is undermined somewhat by the discovery of a soybean nodule ALA formation activity (Sangwan and O’Brian, 1991, 1992, 1993), and the analysis of mutants must be viewed in the context of data that show that nodules elicited by a Rhizobium meliloti hemA mutant are arrested in development at a stage that precedes hemoglobin expression (Dickstein et al., 1991). Similarly, Bradyrhizobium japonicum mutants defective in hemB (Chauhan and O’Brian, 1993) and hemH (Frustaci and O’Brian, 1992, 1993) incite soybean nodules that are poorly developed; thus, the absence of latenodule proteins such as hemoglobin is expected regardless of the source of prosthetic group. Thus, the symbiont that cames out hemoglobin heme synthesis remains unknown. Evidence suggests that Bradyrhizobium japonicum can take up and utilize soybean-derived ALA (Sangwan and O’Brian,
We isolated a soybean (Clycine max) cDNA encoding the heme and chlorophyll synthesis enzyme 6-aminolevulinic acid (ALA) dehydratase by functional complementation of an Escherichia coli hemB mutant, and we designated the gene Alad. ALA dehydratase was strongly expressed in nodules but not in uninfected roots, although Alad mRNA was only 2- to 3-fold greater in the symbiotic tissue. Light was not essential for expression of Alad in leaves of dark-grown etiolated plantlets as discerned by mRNA, protein, and enzyme activity levels; hence, its expression in subterranean nodules was not unique in that regard. The data show that soybean can metabolize the ALA it synthesizes in nodules, which argues in favor of tetrapyrrole formation by the plant host in that organ. Molecular phylogenetic analysis of ALA dehydratases from 11 organisms indicated that plant and bacterial enzymes have a common lineage not shared by animals and yeast. We suggest that plant ALA dehydratase is descended from the bacterial endosymbiont ancestor of chloroplasts and that the Alad gene was transferred to the nucleus during plant evolution.
Plants synthesize a wide variety of tetrapyrroles that include Chl, hemes, siroheme, and bilins, and these compounds participate in many cellular processes (reviewed by Beale and Weinstein, 1991). Synthesis of hemes and Chl from the universal tetrapyrrole precursor ALA share common enzymic steps and diverge after protoporphyrin IX synthesis for formation of the respective metal porphyrin derivative. Chl is by far the predominant tetrapyrrole found in most plants, and its abundance in photosynthetic tissues has rendered it amenable to experimental analysis. Accordingly, much less is known about the expression of other tetrapyrroles that are functionally important but are usually found in low amounts. One exception is plant hemoglobin expressed in legume root nodules, which makes up approximately 25% of the plant protein mass in the cytosol (Appleby, 1984). Nodules are specialized root organs elicited by nitrogen-fixing rhizobia (reviewed by Nap and Bisseling, 1990; Caetano-Anolles and Gresshoff, 1991; Verma, 1992; Long and Staskawicz, 1993), and hemoglobin is one of numerous nodule-specific plant This work was supported by National Science Foundation grant IBN-9204778 and by the Cooperative State Research Service, U.S. Department of Agriculture, under agreement 91-37305-6750. * Corresponding - author; fax 1-716-829-2725.
Abbreviations: ALA, 6-aminolevulinic acid; GSA, glutamate 1semialdehyde; PBG, porphobilinogen. 1411
Kaczor et al.
1412
1991), thereby rendering the bacterial hemA gene nonessentia1 for that symbiosis (Guerinot and Chelm, 1986).Therefore, plant ALA is not committed to plant tetrapyrrole synthesis, and induction of ALA formation activity cannot in itself be interpreted as evidence for plant heme synthesis in nodules. Very recently, we found that, unlike the hemA gene, the B. japonicum hemB gene is essential for symbiosis (Chauhan and O‘Brian, 1993), suggesting that ALA is the only plant tetrapyrrole precursor that can be utilized by the bacterial symbiont. It follows then that soybean nodule enzymes involved in ALA metabolism should be committed to plant tetrapyrrole synthesis; so we have initiated a study of the expression of soybean ALA dehydratase, the enzyme that forms PBG from ALA. In addition, we provide molecular phylogenetic evidente for a bacterial origin for plant ALA dehydratases, indicating that the Alad gene was derived from the bacterial ancestor of chloroplasts and was transferred to the nucleus during plant evolution. MATERIALS A N D METHODS
Plant Physiol. Vol. 104, 1994
ALA Dehydratase Protein and Enzyme Activity
The enzyme activity of ALA dehydratase was measured as PBG formation from ALA as described previously (Sangwan and O’Brian, 1991). Extracts of E. coli and of soybean leaves, roots, and nodule cytosol were prepared as demibed previously (Sangwan and O’Brian, 1993). The absence of bacterial contamination in plant nodule fractions is observed as the lack of ALA synthase activity in the plant fractions (Sangwan and O’Brian, 1991). Reactions were camed out for 1 h at 37OC im Tricine buffer (pH 8), 8 m ALA, ancl 20 mM DTT using 0.5 mg of protein per reaction. Control reactions were incubated on ice. PBG was quantitated spectrophotometrically after reaction with the Ehrlich reagent as described by Sangwan and O‘Brian (1992). ALA dehydratase was detected in plant tissue extracts immunologically using antibodies raised against the spinach enzyme (Schaumburg et al., 1992) in sloi blots. Cross-reactive protein on nitrocellulose filters was cliscemed visually using peroxidase-conjugated goat anti-rabbit IgG as described previously (Ausubel et al., 1987).
Plants and Bacteria
Escherichia coli strain RP523 is a hemB mutant of strain C600 and is a hemin auxotroph (10 &mL) (Li et al., 1988). E. coli strains were grown in Luria broth, and ampicillin (50 Fg/mL) was added to media for growth of strains harboring pUC18, pSK Bluescript 11, and their recombinant derivatives. Bradyrhizobium japonicum strain I1 10 was the soybean endosymbiont used in the present work and was grown in GSY media as previously described (Frustaci et al., 1991). Soybeans (Glycine max cv Essex), either inoculated with B. japonicum or uninoculated, were grown in a growth chamber under a 16-h light/8-h dark regimen at 25OC. Nodules, leaves, and roots were harvested from 23-d-old plants for enzyme assays or RNA extraction. lsolation and Analysis of Soybean Leaf c D N A Encoding ALA Dehydratase
A soybean leaf cDNA expression vector library in pUC18 was a gift from Dr. M.L. Kahn (Washington State University, Pullman, WA) and was constructed as described by Udvardi et al. (1991). Library DNA was used to transform the E. coli hemB mutant strain RP523 en masse by electroporation, and cells were subsequently plated onto Luria broth (Ausubel et al., 1987) containing isopropyl p- A-thiogalactopyranoside (0.4 mM) and ampicillin in the absence of hemin. Complemented transformants were scored as ampicillin-resistant hemin prototrophs. Plasmid DNA was isolated from the transformed cells as described previously (Ausubel et al., 1987) and analyzed initially by agarose gel electrophoresis and retransformation of strain RP523 to confirm that the plasmid conferred hemin prototrophy on the mutant. The cDNA insert of one of the complementing clones, pTK1, was cloned into pBluescript IISK+ to construct pRRK1, and deletions were made using an Exo-Mung deletion kit (Stratagene). The nucleotide sequence of both strands of the pRRKl insert was determined by the chain termination method (Sanger et al., 1977) using a Sequence kit (United States Biochemical) according to the manufacturer’s instructions.
RNA lsolation and Analysis
Nodules, leaves, and roots were excised, frozen in liquid nitrogen, and homogenized in a blender with homogenization buffer and phenol ( 2 2 3 , w/v/v). The homogenization buffer contained 500 mM Tris (pH 8), 10 mM MgC12, 1 m EDTA, 100 m NaC1, 0.5% (w/v) deoxycholate, and 1 m P-mercaptoethanol. Total RNA was isolated from the homogenate as described previously (Ausubel et iil., 1987), and poly(A)+ RNA was isolated using oligo(dT)-cellulose columns. Northem blot analysis of poly(A)+RNA was camed out as previously described under high-stringency conditions (Ausubel et al., 1987). Ubiquitin cDNA was provided by Dr. D.P.S. Verma (Fortin et al., 1988). Relative amounts of RNA detected in northem blots were estimated by ijcanning densitometry of the autoradiograms. Analysis of ALA Dehydratase Sequences and Construction of Phylogenetic Trees
Deduced amino acid sequences of 11 ALA dehydratases were trimmed to the conserved regions and progressively aligned with the TREE program of Feng and Doolittle (1990), which uses the Dayhoff Mutation Matrix (D‘iyhoff, 1978). Matrix distances are calculated from the number of shared residues in pairwise comparisons, their raxtdomized sequences, and self-comparisons of each (Feng and Doolittle, 1990). The ALA dehydratase sequences analyzed in the present work are from soybean, pea (Boese et al., l991), spinach and Selaginella martensii (Schaumburg et al., 1992), human (Weknur et al., 1986), mouse (Bishop et al., 15189), Saccharomyces cerevisiae (yeast, Myers et al., 1987), E. coli (Li et al., 1989), Bacillus subtilis (Hansson et al., 1991). B. japonicum (Chauhan and O’Brian, 1993), and Methanothtrmus sociabilis (Brocld et al., 1992). The rat ALA dehydratase (Bishop et al., 1986) was not included in the analysis because of its near identity to the mouse sequence.
Expression and Lineage of Plant 6-Aminolevulinic Acid Dehydratase
Table 1. Functional complementation of E. coli hemB mutant strain RP523 with sovbean cDNA clone oTKl ALAD Activity
Strain
Doubling Time of Crowth Plus hemin
nmol PBC h-' mg-' protein
C6OO[pUC18] RP523[pUC18] RP523 [ pTK1 ]
No hemin
min
11.2
41
39
O 6.6
44 43
No growth
62
RESULTS lsolation of a ALA Dehydratase-Encoding c D N A
E. coli strain RP523 is defective in hemB, the gene encoding ALA dehydratase, and behaves as a hemin auxotroph (Li et al., 1988). Strain RP523 was transformed en masse with a soybean leaf cDNA expression vector library, and cells that were functionally complemented were screened as ampicillinresistant hemin prototroph colonies on agar media. Fifteen prototrophic transformants were isolated that bore identical plasmids as judged from restriction digest pattems, and one of them, pTK1, was chosen for subsequent experiments. pTKl conferred on strain RP523 hemin prototrophy and ALA dehydratase activity, which was approximately one-half of that found in parent strain C600 (Table I). The insert of pTKl contained an open reading frame of 412 amino acids beginning with a Met codon (Fig. l), and the deduced protein shared 88% identity with the ALA dehydratase from pea (Boese et al., 1991). This identity, along with the complementation of the E. coli hemB mutant, provides very strong evidence that the cloned cDNA encodes soybean ALA dehydratase, and the corresponding gene is designated Alad. The cDNA sequence shows that Alad mRNA is polyadenylated at the 3' end (Fig. I), indicating that the RNA was processed in the nucleus. Because ALA dehydratases are plastid-localized proteins (Smith, 1988), it must be targeted to that organelle from the cytoplasm, and the amino-terminal region of the soybean Alad product is predicted to contain a chloroplast transit peptide (Fig. 1). The predicted transit peptide is based on the known sequence of the mature spinach ALA dehydratase (Schaumburg et al., 1992) and on features that typify such domains (Gavel and von Heijne, 1990).
1413
ALA dehydratase (Schaumburg et al., 1992) were used to detect the enzyme in soybean tissues; cross-reactive protein was found in leaves and nodules but not in uninfected roots (Fig. 3). ALA dehydratase enzyme activities in the respective tissues were consistent with the antibody data, showing activity in leaves and nodules but no discemible activity in uninfected roots (Fig. 3). It is unlikely that the ALA dehydratase found in nodule cytosol was due to B. japonicum contamination during fractionation because bacterial extracts expressed only one-fourth of the activity of that found in the plant cytosol(4 versus 16 nmol h-' mg-' protein). In addition, ALA synthase activity, a bacterial marker enzyme, is absent in the plant fraction of nodules. These observations indicate that Alad gene expression in nodules is controlled primarily at the synthesis or tumover of the enzyme and not of the mRNA; Boese et al. (1991) reached a similar conclusion for control of ALA dehydratase in pea leaves. Although we
1 CAATTCGGCACGAGATGGCTTCTTCAATCCCTAATGCGCCTTCTGCGTTCAATTCTCAGA 60 i
r
h
e
M
A
S
S
l
P
N
A
P
S
A
F
N
S
O
S
61 GCTACGTTGGTCTCAGAGCGCCACTGAGGACCTTCAACTTTTCTTCTCCTCAAGCTGCCA 120 Y
V
G
L
R
A
P
L
R
T
F
N
F
S
S
P
Q
A
A
K
121 AAATTCCTCGCTCCCAACGCCTTTTCGTCGTCAGAGCCTCCGATTCGGAGTTCGAAGCCG 180 I
P
R
S
O
R
L
F
V
V
R
A
S
D
S
E
F
E
A
A
181 CCGTTGTCGCCGGTAAGGTTCCGCCGGCGCCTCCCGTGCGGCCCAGACCGGCGGCTCCGG 240 V
V
A
G
K
V
P
P
A
P
P
V
R
P
R
P
A
A
P
V
241 TTGGAACACCGGTGGTTCCTTCACTTCCACTTCACCGGCGTCCTCGTCGGAACCGGAAGT 300 G
T
P
V
V
P
S
L
P
L
H
R
R
P
R
R
N
R
K
S
301 CCCCGGCGCTTCCGTCGGCTTTTCAGGAAACGAGCATTTCGCCGGCGAATTTCGTGTATC 360 P
A
L
R
S
A
F
O
E
T
S
I
S
P
A
N
F
V
Y
P
361 CGCTTTTCATTCACGAACGTGAAGAGGATACTCCAATTGGGGCTATGCCTGGATGCTACA 420 L
F
l
H
E
G
E
E
D
T
P
l
G
A
M
P
G
C
Y
R
421 GGCTTGGGTGGACGCATGGACTTGTAGAAGAGGTTGCAAAAGCACGGGATGTTGGTGTTA 480 L
G
U
R
H
G
L
V
E
E
V
A
K
A
R
D
V
G
V
N
481 ATAGTGTTGTGCTCTTCCCCAAAATTCCAGATGCTTTGAAGTCTCCCACAGGAGATGAAG S
V
V
L
F
P
K
I
P
D
A
L
K
S
P
T
G
D
E
540
A
541 CATACAATGAAAATGGTTTAGTGCCTCGAACAATACGTTTGCTCAAGGATAAGTACCCAG Y
N
E
N
G
L
V
P
R
T
I
R
L
L
K
D
K
Y
P
600
D
601 ACCTTCTTATCTATACAGATGTTGCATTAGATCCTTATTCATCAGATGGGCATGATGGCA660 L
V
I
Y
T
D
V
A
L
D
P
Y
S
S
D
G
H
D
G
I
661 TAGTTAGAGAAGATGGAGTTATTATGAATGATGAGACAGTTCATCAGCTATGTAAACAAG V
R
E
D
G
V
I
M
N
D
E
T
V
H
O
L
C
K
O
720
A
721 CTGTAGCCCAGGCCCAAGCTGGAGCAGATGTTGTCAGTCCTAGTGATATGATGGATGGTC 780 V
A
Q
A
Q
A
G
A
D
V
V
S
P
S
D
M
M
D
G
R
781 GGGTAGGAGCACTGCGTGCAGCTCTGGATGCTGAAGGCTTTCAGCATGTTTCTATAATGT 840 V
G
A
L
R
A
A
L
D
A
E
G
F
O
H
V
S
I
M
S
841 CCTATACAGCAAAGTATGCAAGTTCTTTTTATGGTCCATTTAGAGAGGCACTAGACTCAA Y
T
A
K
Y
A
S
S
F
Y
G
P
F
R
E
A
L
D
900
S
N
901 ACCCCCGGTTTGGAGACAAGAAAACTTATCAGATGAACCCAGCTAATTACAGAGAGGCTC 960 P
R
F
C
D
K
K
T
Y
Q
M
N
P
A
N
Y
R
E
A
L
961 TGACTGAGATGCGGGAGGATGAATCTGAAGGAGCTGACATTCTCTTGGTGAAGCCTGGTC 1020 T
E
M
R
E
D
E
S
E
G
A
D
I
L
L
V
K
P
G
L
1020 TTCCTTACTTGGATATCATAAGGCTGCTCAGGGATAATTCTCCTTTGCCAATTGCAGCAT
Expression
of the AIad Cene in Soybean Tissues
P
Y
L
D
I
I
R
L
L
R
D
N
S
P
L
P
I
A
1080
A
Y
1081 ACCAGGTTTCTGGTCAATATGCAATGATAAAGGCTGCCGGTGCTCTCAAAATGATAGACG
Poly(A)+RNA from leaves, nodules, and uninfected (asymbiotic) roots of 23-d-old plants was probed with ALA dehydratase cDNA in northem blots to discern Alad transcripts in those tissues (Fig. 2 ) . Alad mRNA was expressed in root nodules as discerned by a single band, and the steady-state leve1 was 2- to 3-fold greater than was found in uninfected roots (Fig. 2). The modest increase of Alad message in nodules compared to roots differs from that of Gsa, which is much greater in the symbiotic tissue than in uninfected roots (Sangwan and O'Brian, 1993; Fig. 2). The single band at 1.6 kb in a11 tissues examined suggested that the message represents a single gene or gene family. Antibodies raised against spinach
Q
V
S
G
E
Y
A
M
I
K
A
A
G
A
L
K
M
I
D
1140
E
1141 AAGAAAAGGTTATGATGGAGTCACTGATGTGCCTCCGAAGGGCCGGTGCTGATATCATCC 1200 E
K
V
M
H
E
S
L
M
C
L
R
R
A
G
A
D
I
I
L
1201 TCACATATTCTGCTCTCCAAGCTGCCAGATGTTTGTGTGGAGAGAAGAGGTGAAGTTCTC T
1261 1321 1381 1441 1501
Y
S
A
L
O
A
A
R
C
L
C
G
E
K
R
1260
"
1320 1380 TCATAATAAAATATTATGGCCTCGTTTGATTTAATATATTTAAGGACATAATAGAAAAAG 1440 ACTTGGCCCTTGTTTTTGGTGTTAATTGTAACTTTACGTGTGGATTGTCAACCCATTGTA 1500 TGAATGAAAGTGTGCACTCACCTGCAGAATATTTCGTACTGTAAAAAAAAAAAAAAAAAA 1560 1595 1561 AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA TGATTATGTAGGGCGTTGTTCATGTAGAAGGTTGAAGAGTTTATAATACCAGTATCTGCT
GGATTTTGGTTATTGTAAATTGTTTAAGAGGGACATGGAGGTTTGTGTATAGAGAGACAT
Figure 1. Nucleotide sequence and deduced product of t h e cloned cDNA insert of pTK1 encoding soybean ALA dehydratase. The underlined region represents the putative leader peptide.
1414
Kaczor et al.
L
R
N
A lad
Gsa
Ubi Figure 2. Northern blot analysis poly(A)+ RNA from leaves (L), roots (R), and nodules (N) of 23-d-old plants. RNA (5 Mg) was loaded onto each lane, and the figure represents a composite of a single filter probed separately with cDNA encoding ALA dehydratase (A/ad), CSA aminotransferase (Gsa), or ubiquitin (Ubi). Ubiquitin was used as a constitutively expressed gene.
cannot strictly rule out a nodule-specific dehydratase as an alternative explanation, such a protein would need to be homologous enough to the ubiquitous enzyme to be detected by the anti-ALA dehydratase antibodies, but the mRNA would have to be sufficiently disparate to be undetected in northern blots. Regardless, expression of soybean ALA dehydratase in root nodules indicates that the plant can utilize the ALA it synthesizes there, thus inferring tetrapyrrole formation by the host in the symbiotic organ. In addition, the induction of the Alad gene in nodules suggests an increased capacity for plant tetrapyrrole formation compared with uninfected roots. Effect of Light on Alad Expression
Full expression of Chl in soybean and in other angiosperms requires light, and evidence for light regulation of tetrapyrrole enzymes, including ALA dehydratase, in photosynthetic tissues has been put forth (Kasemir and Masoner, 1975). Direct light exposure is an unlikely positive effector of gene expression in subterranean root nodules; hence, we examined
Plant Physiol. Vol. 104, 1994
whether light was necessary for expression of the Alad gene in leaves. Poly(A)+ RNA was isolated from leaves of etiolated soybean plants grown completely in darkness and from greening etiolated plant leaves made so by exposure to light for 24 h prior to harvesting. The soybean cab gene was used as a control for a light-induced gene (Chang and Walling, 1992), and it can be seen in northern blots that expression of that gene is strongly dependent on light (Fig. 4). However, Alad mRNA levels did not require light for high expression, as seen by the pronounced hybridization signal in RNA from leaves of dark-grown plantlets. The Alad message was enhanced only 2-fold by light, which is similar to what we
previously found for the Gsa gene (Sangwan and O'Brian, 1993; Fig. 4). Because leaves of etiolated plants expand considerably during the 24-h greening period, it is plausible that the modest differences in Alad and Gsa messages expressed in the respective tissues reflect developmental control of those genes. The protein and enzyme activity of ALA dehydratase were expressed in leaves of dark-grown plants at levels at least as high as was found in light-exposed leaves (Fig. 3). From this, we conclude that full expression of ALA dehydratase in leaves does not require light. Evidence for a Bacterial Lineage for Plant ALA Dehydratases
The putative metal-binding domain of the B. japonicum ALA dehydratase contains some features characteristic of the the corresponding domain in the plant enzymes (Chauhan and O'Brian, 1993). In the present work, we wanted to determine whether the anomalous B. japonicum protein could be the result of an evolutionary horizontal gene transfer from the plant host to the bacterium, in which case the B. japonicum
Alad
Cab 23 days
Etiolated N
Gsa
in*
Protein
Ubi Activity
16
42
26
Figure 3. Immunoblot and enzyme activities of ALA dehydratase from soybean tissue extracts. Protein (25 ^g Per slot) was loaded and probed with antibodies raised against spinach ALA dehydratase. The enzyme activity of ALA dehydratase was carried out as described in the text, and the data are expressed as nmol PBG formed h~' mg~' protein and are averages of three trials. Tissue extracts were prepared from leaves (L), roots (R), and nodules (N) of 23-d-old plants and from the leaves of etiolated plantlets that were either subjected to complete darkness (D) or illuminated for 24 h prior to harvesting (I), in which time they greened.
Figure 4. Effect of light on Alad mRNA levels in leaves of etiolated plantlets. Poly(A)+ RNA was isolated from leaves of plantlets grown completely in darkness (D) or illuminated for 24 h prior to harvesting (I), in which time they greened. RNA (5 Mg) was loaded onto each lane, and the figure represents a composite of a single filter hybridized separately with radiolabeled DNA probes. The cDNA probes used encode ALA dehydratase (Alad), Chl a/b-binding protein (Cab), CSA aminotransferase (Csa), and ubiquitin (Ubi). Cab and Ubi were used as controls for light-induced and constitutive genes, respectively.
Expression and Lineage of Plant 6-Aminolevulinic Acid Dehydratase ALA dehydratase would be unique in its similarity to the plant enzymes over its entire length. Conserved regions of 11 ALA dehydratases from plants, bacteria, yeast, and mammals were progressively aligned as described by Feng and Doolittle (1990), and phylogenetic trees were constructed (Fig. 5 ) . The analysis as depicted in the unrooted matrix tree did not indicate a plant source for the B. japonicum ALA dehydratase, but rather it infers a bacterial origin of plant ALA dehydratases. The sequence similarities between each of the plant-bacteria comparisons (50-58%) were higher than any of those between plants and other eukaryotes (39-47%). Grouping of the archaebacterium M. sociabilis with the eubacteria is an unexplained anomaly of the phylogenetic analysis (Fig. 5 ) . The endosymbiotic theory of organelle origination proposes that chloroplasts (and mitochondria) are derived from once free-living bacteria (Gray, 1989). Thus, the deduced phylogeny of ALA dehydratases, along with the known plastid location of the nuclear-encoded enzyme, leads us to postulate that the plant Alad gene is descended from the bacterial ancestor of chloroplasts and was transferred to the nucleus during plant evolution. DISCUSSION
In the present work, we isolated a cDNA encoding soybean ALA dehydratase and examined the expression of the Alad
1 O matrix units
Aea soybean
L s p i n a c h
S.martensii 6. japonicum
B. subtilis
1L__
M. sociabilis
human mouse yeast
Figure 5. Molecular phylogenetic analysis of ALA dehydratase from various organisms. Deduced amino acid sequences of 11 ALA dehydratases were trimmed to the conserved regions and progressively aligned with the TREE program of Feng and Doolittle (1990), which uses the Dayhoff Mutation Matrix (Dayhoff, 1978). The ALA dehydratase sequences analyzed in the present work are from soybean (Fig. Z), pea (Boese et al., 1991), spinach and S. martensii (Schaumburg et al., 1992), human (Wetmur et al., 1986), mouse (Bishop et al., 1989), 5.cerevisiae (yeast, Myers et al., 1987), E. coli (Li et al., 1989), 5. subtdis (Hansson et ai., 19911, 6 . japonicum (Chauhan and O’Brian, 1993), and M. sociabilis (Brockl et al., 1992). The rat ALA dehydratase (Bishop et al., 1986) was not included in the analysis because of its near identity to the mouse sequence.
1415
gene in root nodules as a step toward assessing whether the plant host can synthesize significant quantities of heme in that organ. Soybean ALA dehydratase was highly expressed in nodules but not in uninfected roots, and thus, the Alad gene was induced by symbiosis. Alad mRNA was only modestly higher in nodules than in uninfected roots; hence, the large differences in enzyme expression between the two tissues are likely to be due primarily to protein synthesis or stability. This level of control is unlike that of the GSA aminotransferase gene in nodules, in which the transcript level contributes significantly to differential expression (Sangwan and O’Brian, 1993). Although Chl synthesis requires light, expression of Alad did not require light in leaves of etiolated plantlets; thus, there is no need to invoke a regulatory factor to compensate for light to account for the expression of ALA dehydratase in subterranean root nodules. Our studies with etiolated plants do not rule out light regulation of the Alad gene in leaves of plants grown under a normal diumal light/dark cycle. The expression of ALA dehydratase by soybean in nodules strongly suggests that plant-synthesized ALA can be metabolized by the host as well as by the bacterial symbiont. This conclusion is significant because plant PBG cannot support B. japonicum heme formation in nodules (Chauhan and O’Brian, 1993); hence, the soybean ALA dehydratase is committed to plant tetrapyrrole synthesis. Expressions of soybean Alad and Gsa are comparable to those found in leaves in which copious amounts of Chl are synthesized (Figs. 2 and 3; Sangwan and O’Brian, 1993); thus, it is clear that the activities of those enzymes are sufficient to support heme formation in high quantities. A diminution of the induced ALA formation and ALA dehydratase activities as a function of nodule age (Nadler and Avissar, 1977; Sangwan and O’Brian, 1992) infers that rates of synthesis are highest in young nitrogen-fixing nodules. The data do not prove that hemoglobin heme is provided, in part or in toto, by the eukaryotic symbiont, but it alters the premises on which past arguments have been made (Cutting and Schulman, 1969). In addition, we reiterate that soybean nodules elicited by B. japonicum hemB or hemH mutants are underdeveloped (Frustaci and O’Brian, 1992; Chauhan and OBrian, 1993); thus, the lack of leghemoglobin, a late-nodule protein, is expected regardless of the source of the prosthetic group. Molecular phylogenetic analysis suggests that plant and bacterial ALA dehydratases have a common lineage not shared by animals or yeast; we interpret this as evidence that plant ALA dehydratase is derived from the bacterial chloroplast ancestor and was transferred to the nucleus during plant evolution. Paleobotanicaland molecular evidence identifies the lycopsids, the group of primitive tracheophytes to which Selaginella belongs, as the basal lineage in vascular land plants (Raubeson and Jansen, 1992, and refs. therein). Thus, if one assumes that the nuclear location of the bacteriatype Alad genes analyzed herein is the result of a single transfer, the event would have occurred no less than 400 million years ago, and the nuclear-encoded gene should be prevalent in vascular land plants. Transfer of the Alad gene from the chloroplast to the nucleus may be considerably more ancient, and the identification and localization of the
Plant Physiol. Vol. 104, 1994
Kaczor e t al.
1416
gene from nonvascular plants (bryophytes) a n d algae should clarify this. Separate origins for ALA dehydratase between plants and animals indicates that the Chl pathway in plants is not likely the result of a n evolutionary addition of a unique branch t o the heme pathway enzymes of the ancestor from which plants and animals diverged approximately 1 billion years ago. Also, t h e absence of a cytosolic ALA dehydratase i n plants (Smith, 1988) suggests that the eukaryotic type Alad gene predicted to be inherited from that ancestor has been lost or inactivated. Evidence suggests that Chl-specific plant enzymes are descended from t h e chloroplast ancestor (Fujita e t al., 1992; Bauer et al., 1993; Burke e t al., 1993), a n d the present work points t o t h e same origin for a universal tetrapyrrole synthesis enzyme. NOTE ADDED IN PROOF
A soybean gene encoding coproporphyrinogen oxidase, which catalyzes t h e antepenultimate step i n heme synthesis, is highly expressed i n root nodules (Madsen et al., 1993). This observation supports t h e conclusion that the plant host can form heme in nodules. ACKNOWLEDCMENTS
The authors thank Sharon Cosloy for E. coli strain RP523, Michael Kahn for the soybean leaf cDNA library, H.A.W. Schneider-Poetsch for anti-spinach ALA dehydratase antibodies, Desh Pal Verma for the ubiquitin cDNA, and Linda Walling for cab3 cDNA. Received October 8, 1993; accepted December 13, 1993. Copyright Clearance Center: 0032-0889/94/104/1411/07. The GenBank accession number for the sequence reported in this article is U04525. LITERATURE ClTED
Appleby CA (1984) Leghemoglobin and Rhizobium respiration. Annu Rev Plant Physiol35 443-478 Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA, Struhl K (1987) Current Protocols in Molecular Biology. Wiley Interscience, New York Bauer CE, Bollivar DW, Suzuki JY (1993) Genetic analyses of photopigment biosynthesis in eubacteria: a guiding light for algae and plants. J Bacterioll75 3919-3925 Beale SI, Weinstein JD (1991) Biochemistry and regulation of photosynthetic pigment formation in plants and algae. In PM Jordan, ed, Biosynthesis of Tetrapyrroles. Elsevier Scientific, New York, pp 155-235 Bishop TR, Frelin LP, Boyer SH (1986) Nucleotide sequence of rat liver 8-aminolevulinicacid dehydratase cDNA. Nucleic Acids Res 1 4 10115 Bishop TR, Hodes ZI, Frelin LP, Boyer SH (1989) Cloning and sequence of mouse erythroid 6-aminolevulinate dehydrogenase cDNA. Nucleic Acids Res 17: 1775 Boese QF, Spano AJ, Li J, Timko MP (1991) Aminolevulinic acid dehydratase in pea (Pisum satiuum L.). Identification of an unusual metal binding domain in the plant enzyme. J Biol Chem 266 17060-17066 Brockl G, Berchtold M, Behr M, Konig H (1992) Sequence of the 5-aminolevulinic acid dehydratase-encoding gene from the hyperthermophilic methanogen, Methanothermus socia bilis. Gene 119 151-152 Burke DH, Alberti M, Hearst JE (1993) bchFNBH bacteriochlorophyll synthesis genes of Rhodobacter capsulatus and identification of the third subunit of light-independent protochlorophyllide reductase in bacteria and plants. J Bacterioll75 2414-2422
Caetan'o-AnollesG, Gresshoff PM (1991) Plant genetic control of nodulation. Annu Rev Microbiol45 345-382 Chang YC, Walling LL (1992) Spatial and temporal expression of Cab mRNAs in cotyledons of the developing soybean seedling. Planta 186 262-272 Chauhian S, OBrian MR (1993) Bradyrhizobium japonicum 6-aminolevulinic acid dehydratase is essential for symbiosis with soybean and contains a nove1 metal-binding domain. J Bacteriol 1 7 5 7222-7227 Cutting JA, Schulman HM (1969) The site of heme synthesis in soybean root nodules. Biochim Biophys Acta 192 4 86-493 Dayhoff MO (1978) Atlas of Protein Sequence and Structure, Vol5, Suppl 3. National Biomedical Research Foundation, Washington, DC Dickstctin R, Scheirer DC, Fowle WH, Ausubel FM (1991) Nodules elicited by Rhizobium meliloti heme mutants are arrested at an early stage in development. Mo1 Gen Genet 230: 423-432 Feng DF, Doolittle RF (1990) Progressive alignment and phylogenetic tree construction of protein sequences. Methods Enzymol 183 375-387 Fortin MG, Purohit SK, Verma DPS (1988) The pnmary structure of soybean (Glycine max) is identical to other plant ubiquitins. Nucleic Acids Res 16: 11377 Frustaci JM, OBrian MR (1992) Characterization of a Bradyrhizobium japonicum ferrochelatase mutant and isolation of the hemH gene. J Bacterioll74 4223-4229 Frustaci JM, OBrian MR (1993) Analysis of the Bradyrhizobium japonicum hemH gene and its expression in Eschei,ichia coli. Appl Environ Microbiol59 2347-2351 Frustaci JM, Sangwan I, O'Brian MR (1991) Aerobic growth and respiration of a 6-aminolevulinic acid synthase (ht!mA)mutant of Brad,yrhizobium japonicum. J Bacterioll73 1145-1150 Fujita Y, Takahashi Y, Chuganji M, Matsubara H (1992) The nifHlike (1frxC)gene is involved with the biosynthesis of chlorophyll in the l'ilamentous cyanobacterium Plectonema boryanum. Plant Cell Physiol 3 3 81-92 Gavel Y, von Heijne G (1990) A conserved cleavage-site motif in chloroplast transit peptides. FEBS Lett 261: 455-458 Gray MW (1989) The evolutionary origins of organelles. Trends Genet 5 294-299 Guerinot ML, Chelm BK (1986) Bacterial 6-aminolevulinic acid synthase is not essential for leghemoglobin formation in the soybeanlBradyrhizobium japonicum symbiosis. Proc Natl Acad Sci USA 8 3 1837-1841 Hansson M, Rutberg L, Schroder I, Hederstedt L (1991) The Bacillus subtilis hemAXCDBL gene cluster, which encodes enzymes of the biosynthetic pathway from glutamate to uroporphyrinogen 111. J Bacterioll73 2590-2599 Jahn D, Verkamp E, Sol1 D (1992) Glutamyl-transfer RNA: a precursor of heme and chlorophvll biosynthesis. 'Trends Biochem &i 1.7: 215-218 Kasemir H. Masoner M 11975) Control of chlorouhvll svnthesis bv phytochrome. 11. The effect 'of phytochrome on aminolevulinate dehydratase in mustard seedlings. Planta 126 119-126 Leong SA, Ditta GS, Helinski DR (1982) Heme synthesis in Rhizobium. Identification of a cloned gene coding for I#-aminolevulinic acid synthetase from Rhizobium meliloti. J Biol Chem 257: 8724-8730 Li J-M,Russell CS, Cosloy SD (1989) The structure cif the Escherichia coli hemB gene. Gene 7 5 177-184 Li J-M, Umanoff H, Proenca R, Russell CS, Cosloy SD (1988) Cloning of the Escherichia coli K-12 hemB gene. J Bacteriol 170: 1021-1025 Long SR, Staskawicz BJ (1993) Prokaryotic plant parasites. Cell73 921 -935 Madsen O, Sandal L, Sandal NN, Marcker KA (1993) A soybean coproporphyrinogen oxidase gene is highly expressed in root nodules. Plant Mo1 Biol 2 3 35-43 Myers AM, Crivellone MD, Koerner TJ, Tzagoloff A (1987) Characterization of the yeast HEM2 gene and transcriptional regulation of COX5 and COR1 by heme. J Eiol Chem 262 16822-16829
-~
I
,
2
Expression and Lineage of Plant 8-Aminolevulinic Acid Dehydratase Nadler KD, Avissar YJ (1977) Heme synthesis in soybean root nodules. I. On the role of bacteroid 6-aminolevulinic acid synthase and 6-aminolevulinicacid dehydratase in the synthesis of the heme of leghemoglobin. Plant Physiol 6 0 433-436 Nap JP,Bisseling T (1990) Developmental biology of a plantprokaryote symbiosis: the legume root nodule. Science 250 948-954 OBrian MR, Kirshbom PM, Maier RJ (1987) Bacterial heme synthesis is required for expression of the leghemoglobin holoprotein but not the apoprotein in soybean root nodule. Proc Natl Acad Sci USA 8 4 8390-8393 Porra RJ (1975) A rapid spectrophotometric assay for ferrochelatase activity in preparations containing much endogenous hemoglobin and its application to soybean root-nodule preparations. Ana1 Biochem 68: 289-298 Raubeson LA, Jansen RK (1992) Chloroplast DNA evidence on the andent evolutionary split in vascular land plants. Science 255 1697-1699 Sanger F, Nicklen S, Coulson AR (1977) DNA sequencing with chain termination inhibitors. Proc Natl Acad Sci USA 7 4 5463-5467 Sangwan I, OBrian MR (1991) Evidence for an inter-organismic
1417
heme biosynthetic pathway in symbiotic soybean root nodules. Science 251: 1220-1222 Sangwan I, OBrian MR (1992) Characterization of 6-aminolevulinic formation in soybean root nodules. Plant Physiol 98: 1074-1079 Sangwan I, OBrian MR (1993)Expression of the soybean glutamate I-semialdehyde aminotransferase gene in symbiotic root nodules. Plant Physiol102: 829-834 Schaumburg A, Schneider-Poetsch HAW, Eckerskorn C (1992) Characterization of plastid 5-aminolevulinate dehydratase (ALAD; EC4.2.1.24) from spinach (Spinacia oleracea L.) by sequencing and comparison with non-plant ALAD enzymes. 2 Naturforsch 47c: 77-84 Smith AG (1988)Subcellularlocalization of two porphyrin-synthesis enzymes in Pisum sntivum (pea) and Arum (cuckoo-pint) species. Biochem J 249 423-428 Udvardi MK, Kahn ML (1991) Isolation and analysis of a cDNA clone that encodes an alfalfa (Medicago sativa) aspartate aminotransferase. Mo1 Gen Genet 231: 97-105 Verma DPS (1992) Signals in root organogenesis and endocytdsis of Rhizobium. Plant Cell4 373-382 Wetmur JG, Bishop DF, Cantelmo C, Desnick RJ (1986) Human 6aminolevulinate dehydratase: nucleotide sequence of a full-length cDNA clone. Proc Natl Acad Sci USA 8 3 7703-7707
