| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
(Received for publication, November 17, 1994; and in revised form, January 30, 1995 ) From the
Expression of plant tetrapyrroles is high in photosynthetic
tissues and in legume root nodules in the form of chlorophyll and heme,
respectively. The universal tetrapyrrole precursor The tetrapyrroles chlorophyll, heme, siroheme, and bilins are
expressed in plants for participation in numerous cellular processes,
and they are synthesized from the universal tetrapyrrole precursor ALA
(reviewed in (1) ). Chlorophyll is the most abundant
tetrapyrrole in plants, and the bulk of ALA synthesized in
photosynthetic tissues is incorporated into the chlorophyll ring.
Accordingly, some evidence shows that ALA formation in those tissues is
controlled by, or coordinated with, factors related to
photosynthesis(2, 3, 4, 5, 6, 7) .
Glutamate-dependent ALA ( Legume root nodules are specialized plant
organs elicited by rhizobia bacteria that contain a large quantity of
heme for the prosthetic group of plant hemoglobin (reviewed in (12) ), but lack chlorophyll. A soybean glutamate-dependent ALA
synthesis activity is induced in nodules(13, 14) , as
well as other genes encoding heme pathway enzymes (15, 16) , but no plant ALA synthase activity is
detectable. In addition, soybean nodule cDNA encoding GSA
aminotransferase was isolated, and both enzyme activity and mRNA are
induced in the symbiotic tissue(10) . These data strongly
support the universality of the C It is not certain whether the
C In the present work, we isolate
soybean gsa1, one of two gsa genes, and provide
evidence that it is involved in the synthesis of cellular tetrapyrroles
in leaves and root nodules, thereby demonstrating the universality of a
C
Figure 1:
Overexpression of nodule gsa cDNA and detection of GSA aminotransferase protein and enzyme
activity in soybean fractions. A, gsa cDNA was
overexpressed in E. coli and a Coomassie-stained SDS-PAGE gel
of the purified product from inclusion bodies is shown (inset). GSA aminotransferase activity was measured in the
inclusion body fraction is shown as ALA formed from GSA/mg of protein
as a function of time. B, Western blot analysis of GSA
aminotransferase and enzyme activity in leaves of etiolated plants (L), uninfected roots (R), nodules (N), and
the nodule bacteroid fraction (B). For the Western blot, 20
µg of protein was loaded per lane and antibodies were raised
against protein purified in A. GSA aminotransferase activity
is expressed as nmol ALA formed in 20 min/mg of
protein.
For gel retardation assays, the
oligonucleotides (dG-dA)
Figure 2:
Northern blot analysis of the expression
of nodule mRNAs as a function of nodule age. Approximately 5 µg of
poly(A)
Figure 3:
Western blot analysis of GSA
aminotransferase (GSA) and leghemoglobin (LB) protein
in nodules as a function of nodule age. Protein corresponding to 2 mg
of tissue was loaded per lane run on a 12% SDS-PAGE gel; protein was
transferred to a filter and analyzed with antibodies raised against the
respective enzyme.
Figure 4:
Southern blot analysis of soybean genomic
DNA probed with a 1-kb NcoI/EcoRI nodule gsa cDNA fragment (A) or a 25-base oligonucleotide probe
(primer 1) (B). 25 mg of DNA was digested with EcoRV (R), HincII (H), or EcoRI (E) and run on a 0.7% agarose gel. DNA from the gel was
transferred to nitrocellulose and probed with
Figure 5:
Gene
structures of soybean gsa1 and A. thaliana gsa1 and gsa2. The striped areas represent exon coding regions
and open areas represent introns.
One
remarkable feature between the gsa genes from soybean and Arabidopsis is the intron variability with respect to number,
size, and relative positions (Fig. 5). The only feature shared
among the three genes is the approximate size of the first exon and
position of the 5` boundary of the first intron. The nucleotide
sequences of the coding regions of the three genes are highly
homologous (73-83%), as are the peptide sequences (80-90%
identical and 88-95% similar), thus it appears that these genes
were derived from a common ancestor. Therefore, the intron variability
indicates that plant gsa gene structure has changed recently
on an evolutionary time scale, subsequent to the establishment of
higher plant lineages.
Figure 6:
Northern blot analysis of gsa1 and total gsa RNA from various soybean tissues.
Poly(A)
Figure 7:
Upstream region of soybean gsa1.
The underlined region denotes the translation start site. Asterisks below nucleotides denote transcription start sites.
The putative TATA box is shown with a broken underline. The GAGA
element is boxed.
Figure 8:
S1 nuclease sensitivity of the GAGA
element. Plasmid-borne DNA containing upstream sequence of gsa1 was used as a template for (A) sequencing reactions or
was (B) treated with S1 nuclease for 0, 16, 32, or 64 min
(lanes 1, 2, 3, 4, respectively) and used as template for synthesis of
a complementary strand. The T7 primer was used resulting in the strand
containing (GA)
A synthetic (dG-dA)
Figure 9:
Retardation of GAGA DNA mobility in
nondenaturing gels.
ALA formation in root nodules is unique among plants in that
none of the ALA produced there is incorporated into chlorophyll. In
addition, ALA synthesis activity is high relative to that found in
other root cell types and presumably in other nonphotosynthetic
tissues. Finally, synthesis is induced in response to interactions with
a bacterium and should be controlled by factors related to symbiosis
and nodule development rather than to photosynthesis. Despite these
unique aspects of the symbiotic organ, previous
studies(10, 13, 14) and the current work
underscore the similarities between leaves and nodules with respect to
ALA synthesis. Herein, we show that gsa1 is a regulated gene
that is strongly expressed in root nodules and leaves and the data
indicate that a putative second gsa gene is either not
expressed or else has a similar expression pattern as that of gsa1. Thus, it is unlikely that soybean has a gsa gene that is specific to a particular tissue, tetrapyrrole, or an
external stimulus such as light, and therefore the evidence strongly
supports the universality of a step of the C Analysis of GSA aminotransferase mRNA,
protein, and enzyme activity showed that gsa is regulated and
that control occurs at the RNA level ( Fig. 1Fig. 2Fig. 3and Fig. 6; (10) ). These observations prompted us to initiate an analysis
of the promoter region of gsa1, which led to the
identification of a DNA element hitherto characterized only in Drosophila. The so-called GAGA element was located immediately
downstream of the putative TATA box, and the pure, plasmid-borne
element was sensitive to S1 nuclease. Because the nuclease-sensitive
GAGA elements in the promoters of Drosophila his3-his4 (44)
and soybean gsa1 have different flanking sequences, the
dinucleotide repeat itself appears to be sufficient for the sensitivity
and does not depend on the context in which the element is found. Gel
mobility shift experiments showed that nuclear extracts from nodules
and from leaves of greening etiolated plantlets contained a factor
which bound to GAGA DNA ((dG-dA) A
group of nodule proteins called nodulins are generally described as
being strictly specific to the symbiotic tissue, such as leghemoglobin
or Enod2(45) . Other proteins such as soybean phenylalanine
ammonia-lyase and chalcone synthase are found throughout the plant, but
they are encoded by gene families of which some are symbiosis-specific (46) . Soybean GSA aminotransferase differs from these other
proteins involved in symbiosis in that its presence in nodules must
result from regulatory factors that alter the spatial expression of a
gene normally expressed strongly only in photosynthetic tissues.
Soybean ALA dehydratase, the enzyme which metabolizes ALA directly for
tetrapyrrole synthesis, is also induced in nodules, but unlike GSA
aminotransferase, the control of the dehydratase is at protein
synthesis or turnover (16) . It is not clear why two enzymes of
the same pathway should be regulated by separate mechanisms, but we
note that porphobilinogen, the product of ALA dehydratase, is committed
to plant tetrapyrrole synthesis in nodules(47) , whereas ALA
may be taken up by B. japonicum for bacterial heme
formation(47) . In developing nodules, gsa expression
preceded that of lb with respect to both mRNA and protein ( Fig. 2and Fig. 3). GSA aminotransferase protein was
strongly expressed in nodules by 13 days, whereas leghemoglobin was
only weakly expressed compared with levels seen in older nodules. These
data are consistent with previous observations which show that
glutamate-dependent ALA formation activity by soybean is high in young
nodules where the leghemoglobin content is low but
discernible(14) . It is plausible that an increased demand for
plant ALA is needed prior to leghemoglobin synthesis for heme-dependent
respiration associated with cell division or for bacterial heme
formation. Transcripts of the early nodulin gene enod2 is
strongly expressed in 10-day-old nodules where gsa message
induction is discernible but weak in those nodules (Fig. 2).
Although no anti-Enod2 antibodies were available to us to follow the
time course of protein expression, other work shows it to be a nodule
structural protein (see (37) ). Thus, it is probably found in
nodules prior to 13 days post-infection, where GSA aminotransferase was
first discerned (Fig. 3). It is likely that gsa expression in nodules is controlled differently than is lb or enod2, which can be inferred by the lack of a GAGA
element in the promoter of the latter genes, and by a greater spatial
expression of gsa throughout the plant. The coding region
of gsa1 is arranged on three exons, separated by two small
introns. It is remarkable that the genomic arrangements of soybean gsa1 and the two gsa genes from A. thaliana are very different from each other with respect to the size,
position, and number of introns. Soybean GSA aminotransferase has a
greater homology to the Arabidopsis enzymes than with any
other known plant or bacterial GSA aminotransferase, hence it is very
likely that the three genes share a common ancestor rather than having
arisen from separate lineages that have converged during evolution. It
follows then that the intron variability resulted either from the
differential loss of introns found in the common ancestral gene or from
the acquisition of introns to each gene. In either case, the events
leading to intron variability are likely to be recent, after the
establishment of modern lineages. It is also possible that all three
genes were present in a common ancestor, and entire genes have been
lost, in which case gene loss would also be subsequent to the
establishment of modern lineages. A corollary to the theory of ancient
introns is that exons represent functional domains which were
differentially spliced to allow protein diversity from a limited genome
in evolving organisms (discussed in (48) ). It is clear,
however, that there can be no correlation between exons and protein
domains that fits all the three known extant plant gsa genes.
Therefore, the gsa genes cannot both be ancient and be
accommodated by the exon theory. The nucleotide
sequence(s) reported in this paper has been submitted to the
GenBank(TM)/EMBL Data Bank with accession number(s)
U20260[GenBank].
Volume 270,
Number 13,
Issue of March 31, 1995 pp. 7387-7393
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
-aminolevulinic
acid (ALA) is synthesized from glutamate 1-semialdehyde (GSA) by GSA
aminotransferase in plants, which is encoded by gsa.
Immunoblot analysis showed that GSA aminotransferase was expressed in
soybean leaves and nodules, but not in roots, and that protein
correlated with enzyme activity. These observations indicate that GSA
aminotransferase expression is controlled in tetrapyrrole formation and
argue against significant activity of an enzyme other than the well
described aminotransferase for GSA-dependent ALA formation. gsa mRNA and protein were induced in soybean nodules, and their
activation was temporally intermediate between those of the respective
early and late genes enod2 and lb. A GSA
aminotransferase gene, designated gsa1, was isolated and
appears to be one of two gsa genes in the soybean genome. gsa1 mRNA accumulated to high levels in leaves and nodules,
but not in uninfected roots as discerned with a gsa1-specific
probe. Message levels were higher in leaves from etiolated plantlets
than in mature plants, and expression in the former was slightly
elevated by light. The expression pattern of gsa1 mRNA was
qualitatively similar to that of total gsa. The data strongly
suggest that gsa1 is a universal tetrapyrrole synthesis gene
and that a gsa gene specific for a tissue, tetrapyrrole, or
light condition is unlikely. The gsa1 promoter contained a
genetic element found in numerous Drosophila melanogaster genes; the so-called GAGA element displayed single-stranded
character in vitro and formed a complex with nuclear factors
from nodules and leaves but not from roots. From these observations we
infer that the GAGA element is involved in the transcriptional control
of gsa1.
)formation occurs in plants by a
three-step mechanism termed the C
pathway; the latter two
steps are committed to ALA synthesis and are catalyzed by glutamyl-tRNA
reductase and glutamate 1-semialdehyde (GSA) aminotransferase,
respectively(1, 8) . Plant cDNA or genes encoding
C pathway enzymes have been isolated from several sources (7, 9, 10) , (
)(
)and
from a green alga(6) . Radiolabel from
[1-
C]glutamate is incorporated into
mitochondrial heme a as well as into plastid chlorophyll in etiolated
seedlings of maize which, along with an absence of ALA synthase
activity, indicates that higher plants use the C pathway
for synthesis of all tetrapyrroles, at least in photosynthetic
tissue(11) . pathway in higher plants
with respect to the tetrapyrrole formed and its distribution in tissues
where ALA formation can be discerned. pathway itself is heterogeneous with regard to the
enzymes that catalyze a given step or to the number of functional genes
that encode an enzyme. Evidence for two enzymes with glutamyl-tRNA
reductase activity are described in two bacterial
species(17, 18, 19) , and although no similar
situation has been reported in plants, separate ALA pools for heme and
chlorophyll synthesis have been proposed in plant
chloroplasts(20) . A plant gene and cDNA encoding glutamyl-tRNA
reductase (hemA) have been isolated from Arabidopsis
thaliana(7) , and multiple copies of hemA are
inferred. Two genes encoding GSA aminotransferase have also been
isolated from A. thaliana with high homology in their
exons(7) . Thus, enzymes of the C pathway may be encoded by gene families in plants, but whether a
particular gene within a family has specificity for a tissue,
tetrapyrrole, or developmental state is not known. This question is
difficult to address in green tissues because heme, although
functionally important, is quantitatively a minor tetrapyrrole in most
plants. However, symbiotic root nodules are not only unique organs
within the plant kingdom, they have high ALA synthetic activity that is
not fated to chlorophyll formation. pathway gene in a higher plant. In addition, we show that gsa1 is regulated, and present evidence for a cis-acting
regulatory element in the gsa1 promoter that has heretofore
been described only in Drosophila. Finally, we argue that the
primary structure of extant plant gsa genes results from
recent evolutionary events.
Bacteria and Plants
Escherichia coli strain BL21(DE3) (pLysS) expresses T7 RNA polymerase (21) and was used for overexpression of gsa1 cDNA in
the present work; it was grown in LB medium (22) supplemented
with 25 µg/ml chloramphenicol, and 100 µg/ml ampicillin was
added to maintain pBluescript-derived plasmids in that strain. Bradyrhizobium japonicum strain I110 was the soybean
endosymbiont used in the present work, and it was grown in GSY
medium(23) . Soybeans (Glycinemax cv Essex),
either inoculated with B. japonicum or not inoculated, were
grown in a growth chamber under a 16 h light/8 h dark regime at 25
°C. Nodules, leaves, and roots were harvested for enzyme assays or
RNA extraction; DNA was extracted from leaves only. Etiolated soybean
plants were grown in total darkness for 9 days and then exposed to
direct light for the final 24 h before harvesting the leaves.GSA Aminotransferase Activity
Enzymatic formation
of ALA from GSA (provided by Dr. C. G. Kannangara) by soybean nodule,
root, leaf, and bacteroid extracts and by purified GSA aminotransferase
obtained by overexpression of gsa cDNA in E. coli was
carried out as described previously(24) . Reactions containing
either 0.5 mg of cell extract or 0.03 mg of partially purified
recombinant GSA aminotransferase were incubated in 1 ml containing 50
mM MOPS, pH 6.8, 1 mM dithiothreitol, 20 µM pyridoxal phosphate, and 5 mM levulinic acid. Controls
contained heat-inactivated protein samples. Reactions were started by
addition of GSA to 50 µM final concentration and proceeded
at 30 °C for times indicated in Fig. 1and its legend.
Reactions were then terminated, and ALA was isolated from reaction
samples by Dowex ion exchange chromatography and solvent extraction and
then quantified as described previously(14) .
Analysis of Genomic DNA and Isolation of a gsa Gene
Fragment
To isolate DNA from soybean, 3 g of leaves were
pulverized to powder in liquid N, then DNA extracted as
described previously(22) . The DNA was purified by subsequent
precipitations with NaCl(25) , cetyl trimethylammonium bromide (26) , and polyethylene glycol(27) . The final DNA
pellet was dissolved in 10 mM Tris, pH 8, 1 mM EDTA.
Southern blot analysis of the genomic DNA was carried out as described
previously (22) . A 2-kb EcoRI fragment of the soybean
genome that hybridized to gsa cDNA was isolated by inverse PCR
using oligonucleotides 5`-GACGACAGATAAGACTCTCTCACTC-3` (hereafter
referred to as primer 1) and 5`-GATCTCTGATCGCATCT-3`. Genomic DNA
isolated from leaves was digested with EcoRI and
size-fractionated by centrifugation in a 1-5 M Nacl step
gradient as described previously(28) . Fractions enriched for
DNA homologous to gsa cDNA were pooled, ligated with T4 ligase
to produce circular DNA, and then used as template DNA for inverse PCR
as described previously(29) . The PCR product was cloned into
pBluescript and identified as the gsa1 gene by comparing its
nucleotide sequence to the corresponding cDNA.Sensitivity of gsa Promoter to S1 Nuclease
A
191-bp DNA fragment, including the gsa1 promoter region, was
ligated into pBluescript SKII, and supercoiled recombinant plasmid was
isolated by cesium chloride centrifugation(22) . DNA (50
µg) was added to 300 µl final volume reaction mixture
containing 30 mM sodium acetate, pH 4.6, 30 mM NaCl,
1 mM ZnCl. S1 nuclease (16 units) was added, and
the digestion was carried out at 37 °C. Aliquots containing 5
µg of DNA were removed at various time points and immediately
extracted with a saturated acid-phenol solution, followed by a
chloroform extraction. The DNA was used as a template for synthesis of
a radiolabeled complementary strand from a T7 primer using a Sequenase
kit (U. S. Biochemical Corp.) and 
-
S-dCTP, except
that dideoxynucleotides were omitted from the reactions. In these
experiments, termination of polymerization results from S1
nuclease-dependent breaks in the template DNA. Sequencing reactions
were carried out in parallel using the T7 primer, undigested DNA, and
dideoxynucleotides according to the manufacturer's instructions,
and all samples were loaded onto a 6% polyacrylamide sequencing gel. In
our hands, the fragments resulting from the sequencing reactions were
more heavily radiolabeled than those using S1 nuclease-treated DNA as
template, thus the data are presented as different exposures of the
same gel.RNA Isolation and Analysis
Nodules, leaves, and
roots were excised, frozen in liquid N, and homogenized in
a blender with buffer and phenol (2:2:3, w/v/v). The homogenization
buffer contained 500 mM Tris, pH 8, 10 mM MgCl, 1 mM EDTA, 100 mM NaCl, 0.5%
(w/v) deoxycholate, and 1 mM 
-mercaptoethanol. Total RNA
was isolated from the homogenate as described previously (22) ,
and poly(A)
RNA was isolated using oligo(dT) cellulose
columns. Northern blot analysis of poly(A) RNA was
carried out as described previously under high stringency conditions (22) using either oligonucleotide or cDNA probes. gsa cDNA used in Northern blots was obtained previously (10) .
Other DNAs were obtained from T. Bisseling (enod2; (30) ), K. Marcker (lba; (31) ), and D. P. S.
Verma (ubi; (32) ). The 5` end of gsa1 mRNA
from nodules and from leaves of etiolated plants was determined by S1
nuclease protection and primer extension as described previously (22) . Primer 1 (see above) was used to prime RNA-dependent DNA
synthesis in the primer extension analysis.Preparation of Nuclear Extracts and Gel Retardation
Assays
Preparation of nuclear extracts from leaves, nodules, or
roots was carried out using a modified protocol of Jensen et
al.(33) , made available to us by Dr. Frans de Bruijn. 20
g of tissue was pulverized in liquid N and resuspended in 5
ml of buffer A/g of tissue (buffer A is 10 mM MES, pH 6, 10
mM NaCl, 5 mM EDTA, 0.15 mM spermine, 0.5
mM spermidine, 10 mM -mercaptoethanol, 1 mM phenylmethylsulfonyl fluoride, 0.6% Triton X-100, and 0.25 M sucrose). The homogenate was filtered successively through one and
two layers of Miracloth (CalBiochem), and the filtrate was centrifuged
for 5 min at 2000 
g. The pellet, which contained the
crude nuclei, was washed once with 5 ml of buffer A and then
resuspended in 5 ml of buffer B (1 part 5 buffer A to 7.5 parts
Percoll, pH 6 (w/w)). The homogenate was centrifuged at 5000 g for 5 min, and the fraction floating on top of the gradient
was collected and washed twice with buffer A by centrifugation at 2000
g for 5 min. The pellet was resuspended in buffer C
(20 mM Hepes, pH 7.9, 420 mM NaCl, 12% glycerol, 1.5
mM MgCl, 0.2 mM EDTA, 0.5 mM dithiothreitol, and 0.5 mM phenylmethylsulfonyl fluoride)
and passed through a small French pressure cell (4 ml capacity) at 900
p.s.i. The extracts were incubated at 4 °C for 1 h with slow
shaking on an Orbitron shaker and then nuclear debris was removed by
centrifugation in a microcentrifuge at 15,000 g for 30
min. Supernatant fractions were aliquoted in Eppendorf tubes and stored
at -70 °C.
and (dT-dC) were
synthesized at the core facility at the State University of New York at
Buffalo and annealed to form double-stranded DNA. The DNA was then
end-labeled with 
P using
[
-[P]dATP and polynucleotide kinase.
Gel retardation assays were carried out as described
previously(34) . 0.5 ng (50 fmol) of radiolabeled probe (
6
10
bequerels) and 1500 ng of the unlabeled
nonspecific competitor DNA (dI-dC)
(dI-dC) were used per reaction, along with 5 µg of nuclear extract
from leaves, roots, or nodules. The samples were run on nondenaturing
12% polyacrylamide gels and subsequently developed by autoradiography.Overexpression of Soybean gsa cDNA in E. coli and
Antibody Production
Soybean gsa cDNA was isolated and
described previously(10) . Modified 5` and 3` ends of a
subcloned fragment were constructed such that it could be cloned into
the NdeI-BamHI sites of pET3c (21) and
translated from the second methionine codon in the open reading frame.
To do this, the upstream primer 5`-CCGCATATGGCCGTATCTATCGACCC-3` and
the downstream primer 5`-CCTGGATCCAACCATCAGATCTCCCT-3` were employed in
a PCR reaction using pKN4 (10) as template. The PCR product was
blunt-ended with T4 polymerase and ligated into the EcoRV site
of pBluescript SKII to construct pKN4C1. The gsa coding region
was removed from pKN4C1 by digestion with BamHI and partial
digestion with NdeI and ligated into the NdeI/BamHI sites of pET3c to construct pET3CGSAA. The
product encoded by pET3CGSAA lacks the putative plastid leader peptide
and should be nearly identical to the mature peptide. The modified gsa cDNA in pET3CGSAA was expressed in E. coli strain
BL21(DE3) (pLysS) as described previously(35) , and the nearly
pure 45-kDa protein was recovered from the inclusion body fraction of
those cells catalyzed GSA-dependent ALA formation (Fig. 1). For
antibody production, 50 µg of the inclusion body extract was loaded
onto a preparative 10% SDS-PAGE gel, and the GSA aminotransferase was
excised from the gel with a razor blade. The excised fragment was
homogenized by freezing and thawing, then forcing through a 18-guage
needle several times. The sample was then used to raise antibodies in
rabbits as described previously(22) . Crude antiserum was
affinity-purified using GSA aminotransferase bound to nitrocellulose as
described previously(36) .Western Blotting
Protein extracts were run on
10-12% SDS-PAGE gels, transferred to nitrocellulose or Immobilon
(Millipore) filters, and screened with antibodies raised against
soybean GSA aminotransferase or leghemoglobin as described
previously(22) . Cross-reactive material bound to the filter
was discerned with peroxidase-conjugated goat anti-rabbit IgG and
visualized either by color development using 4-chloro-1-naphthol or by
chemiluminescence using the Renaissance kit (DuPont NEN) according to
the manufacturer's instructions.
Expression of GSA Aminotransferase Protein and Enzyme
Activity
The soybean gsa nodule cDNA isolated
previously (10) was overexpressed in E. coli such that
the product was missing the putative plastid leader peptide. The
purified protein expressed GSA aminotransferase activity (Fig. 1A), thereby formally demonstrating that the gsa cDNA encodes that enzyme. Expression of GSA
aminotransferase was examined in soybean tissues in Western blots using
antibodies raised against the purified protein. Immunoreactive protein
was detected in the plant fraction of root nodules elicited by B.
japonicum and in leaves of etiolated plantlets, but not in
uninfected roots (Fig. 1B). In addition, GSA
aminotransferase enzyme activity was also observed only in leaves and
nodules (Fig. 1B; (10) ), with protein and
enzyme activity being about 2-fold greater in the etiolated leaves than
in nodules. The correlation between protein and enzyme activity argues
against significant expression of an enzyme other than the well
described aminotransferase that has GSA-dependent ALA formation
activity. A high level of GSA aminotransferase was observed only in
tissues that synthesize a large quantity of tetrapyrrole, showing that
expression of gsa is a control point for ALA and tetrapyrrole
synthesis in soybean. Furthermore, no GSA aminotransferase protein or
enzyme activity was found in extracts of B. japonicum bacteroids isolated from nodules, showing that the protein
observed in the plant fraction was not a bacterial contaminant, nor
does the plant enzyme compartmentalize into a bacterial space. This
latter conclusion is relevant because bacterial heme expression is
rescued in a B. japonicum ALA auxotroph in nodules, which we
have attributed to provision of ALA itself rather than ALA synthetic
enzymes to the bacterial endosymbiont by the plant host(13) .Temporal Expression of gsa in Root Nodules
Root
nodule ontogeny is broadly divided into early and late development,
with the latter stage commencing with the onset of nitrogen fixation.
In addition, developmental stages of cells of so-called determinate
nodules of soybean are approximately uniform at a given time, allowing
temporal information to be obtained from analysis of whole nodules
(reviewed in (37) and (38) ). We compared the temporal
expression of gsa with those of the nodule-specific genes enod2 and lb, which are well described markers of
early and late development,
respectively(37, 39, 40) . Northern blot
experiments showed that enod2 mRNA was not detected in
uninfected root, but was easily discerned by 10 days post-infection,
whereas lb mRNA was not observed in nodules until 13 days (Fig. 2). A weak but discernible induction of gsa mRNA
was observed at 10 days post-infection in Northern blot experiments,
and it was strongly expressed by 13 days and maintained thereafter to
at least 25 days (Fig. 2). GSA aminotransferase protein was also
expressed in developing nodules by Western blot analysis. GSA
aminotransferase protein was not detected in extracts from uninfected
roots or from 10-day-old nodules, but was strongly expressed in nodules
at 13 days and older (Fig. 3), as was gsa message (Fig. 2). Unlike GSA aminotransferase, leghemoglobin was only
weakly expressed in 13-day-old nodules compared with the amount
observed in older nodules (Fig. 3). The data show that full
expression of gsa was temporally intermediate between the
expressions of enod2 and lb and is likely to be
controlled at the RNA level.
RNA from uninfected roots (U) and
from nodules 10, 13, and 25 days post-infection were loaded onto each
lane. A single filter was hybridized with each radiolabeled cDNA
separately, and the filter was stripped after each hybridization and
exposure. Ubiquitin (Ubi) was used as a control for a
constitutively expressed gene. Exposure times varied with different
probes, so they cannot be directly compared with each
other.
Isolation and Characterization of gsa Genomic
DNA
Southern blot analysis showed that a 1-kb NcoI/EcoRI 3` fragment of gsa cDNA
hybridized to two fragments of genomic DNA when digested with EcoRV, HincII, or EcoRI, but only one
fragment was observed for each digestion when a 25-bp oligonucleotide
probe (primer 1) was used that corresponded to the 5`-untranslated
region of the cDNA (Fig. 4). Primer 1 was one of two primers
used to isolate the unique 2-kb EcoRI genomic fragment by
inverse PCR (see ``Materials and Methods''), and the
nucleotide sequence was subsequently determined. The cloned fragment
contained the entire gsa coding region on three exons that
were separated by two small introns (Fig. 5). We designate this
gene gsa1. The nucleotide sequence of the gsa cDNA
shared 100% identity with gsa1 exon sequence, hence we
conclude that the cDNA was derived from gsa1 transcript. The gsa1 gene contained no internal restriction sites of EcoRV, HincII, or EcoRI within the region
delimited by the probe used in the Southern blot, thus the
hybridization to two genomic fragments (Fig. 4) indicates that gsa1 is one of two gsa genes in the soybean genome.
Similarly, two different genes from A. thaliana encoding GSA
aminotransferase have been reported ((7) ).
P-labeled
DNA.
gsa1 Is a Universal Tetrapyrrole Synthesis Gene in
Soybean
Expression of GSA aminotransferase in root nodules for
heme synthesis shows that the enzyme, and most likely the C pathway, is not confined to chlorophyll synthesis or to
photosynthetic tissues. In addition, enzyme activity and message is
found in leaves of dark-grown etiolated plantlets (10) which,
along with the observations of nodules, shows that gsa can be
strongly expressed in the absence of light. The presence of multiple GSA genes in plants led us to ask whether a given gene has
specificity for a tissue, for light, or for the tetrapyrrole in which
the ALA is incorporated. RNA from several soybean tissues was analyzed
by Northern blotting, using either a 1-kb cDNA fragment that hybridizes
to two genomic regions (Fig. 4) or a 25-bp oligonucleotide
specific to gsa1 (primer 1). The data show clearly that gsa1 mRNA was strongly expressed in leaves and nodules from
23-day-old plants, but very little message was observed in uninfected
roots (Fig. 6). Therefore, gsa1 is induced in nodules
for heme synthesis, and the same gene is expressed in leaves, where
chlorophyll is the predominant tetrapyrrole. To assess the light
requirement for gsa1 mRNA expression, message was analyzed in
leaves of dark-grown etiolated plantlets, where the level was higher
than in any tissue from 23-day-old plants grown under a light/dark
regime (Fig. 6). gsa1 mRNA increased about 2-fold in
the etiolated plantlet leaves upon illumination for 24 h prior to
harvest (Fig. 6). The data infer strongly that soybean gsa1 is a universal tetrapyrrole synthesis gene that is expressed
significantly in tissues where ALA is synthesized for heme or
chlorophyll formation. In addition, the pattern of gsa1 mRNA
expression and that of total gsa were qualitatively the same,
indicating either that the putative second gsa gene is not
expressed in the tissues examined or else its pattern of expression is
similar to that of gsa1. Again, this conclusion indicates that
the same gene or genes are required for ALA synthesis from GSA
irrespective of the fate of the precursor or the tissue where it is
formed. Finally, the data described herein, along with previous
work(10) , show that ALA formation in soybean is controlled, at
least in part, by gsa expression, and that this regulation is
at the RNA level.
RNA was analyzed from leaves (L),
roots (R), and nodules (N) from 23-day-old plants and
from leaves of dark-grown etiolated plantlets that were either
illuminated (I) or kept in the dark (D) for 24 h
prior to harvest. The RNA was probed either with gsa cDNA to
assess total gsa mRNA or with a 25-base oligonucleotide
(primer 1) specific to gsa1. Exposure times varied with
different probes, thus they cannot be compared with each other in a
quantitative way.
Identification of a GAGA Element in the gsa1 Promoter
Region
The isolated EcoRI genomic fragment encoding GSA
aminotransferase included 223 bp of DNA upstream of the translation
start site (Fig. 7). Two transcription start sites were found 90
and 114 bp upstream of the initiation codon as determined by S1
nuclease and by primer extension analysis using RNA either from nodules
or from leaves of light-exposed etiolated plantlets (Fig. 7,
data not shown). Because primer 1 (see above) was used for the primer
extension experiments and is specific for the gsa1 gene, we
interpret those data as two transcription start sites of a single gene
rather than a start site for each of two genes. A striking feature of
the 5` upstream region of gsa is a perfect dinucleotide repeat
of (GA) found between the 5`-most transcription start site
and a putative TATA element (Fig. 7). This motif has been
observed in the promoter of several Drosophila genes, where it
binds to the GAGA transcription factor and affects transcription
positively ( (41) and references therein). For Drosophila
hsp26 and hsp70, binding to the GAGA element results in
localized nucleosome disruption, thereby making the promoters
accessible for transcription(42, 43) . Pure plasmid
DNA containing a GAGA element has been shown be sensitive to S1
nuclease in the dinucleotide repeat region at low pH(44) ,
indicating that the double stranded DNA has some single stranded
character. A 191-bp fragment of the soybean gsa1 promoter
region was cloned into pBluescript, treated with S1 nuclease, and the
treated DNA was then used as as a template for DNA synthesis (Fig. 8). The data show clearly that the GAGA element of the
plasmid-borne gsa1 promoter region was sensitive to S1
nuclease as seen by chain termination within the dinucleotide repeat
region.
as the template for the reactions. A and B were exposed for 12 and 48 h, respectively, and are
different portions of the same gel (see ``Materials and
Methods'').
(dT-dC) double-stranded DNA was used in gel retardation experiments to
discern nuclear factors from soybean tissues that bind to the GAGA
element. The mobility of GAGA DNA was retarded on polyacrylamide gels
when treated with nuclear extracts from nodules or from leaves of
illuminated etiolated plantlets, but no retarded species resulted from
treatment with root extracts (Fig. 9). The retarded DNA was
discerned as a doublet for both nodule and etiolated leaf extracts (Fig. 9), which was also observed for pure Drosophila GAGA protein complexed with GAGA DNA(41) . In addition,
leaf extracts yielded another feature on the gel that appeared to be a
diffuse shadow rather than a sharp band, and we have not attempted to
interpret that feature. In conclusion, formation of complexes with DNA
were observed only with nuclear extracts from tissues where gsa1 is strongly expressed, which infers that binding of a nuclear
factor to the GAGA element positively affects the transcription of gsa1.
P-End-labeled
(dG-dA)(dT-dC) (50 fmol) was incubated
with 5 µg of nuclear extract from nodules (N), roots (R), or leaves from illuminated etiolated plantlets (L) or was incubated as free probe (F). The samples
were run on a 12% nondenaturing polyacrylamide gel and then assayed by
autoradiography.
pathway at the
genetic level in a higher plant. This implies that a pool of ALA
committed to either chlorophyll or heme as proposed by Huang and
Castelfranco (20) does not have a genetic basis, but may rely
on spatial separation of common enzymes. Multiple hemA genes,
which encodes the first committed enzyme glutamyl-tRNA reductase, have
been inferred in A. thaliana (7), but the specificity of a
given gene has not been assessed. In addition, the induction of gsa1 in root nodules infers that vigorous ALA synthesis can be
uncoupled from chloroplast development (see (5) ), hence gsa1 may be affected by separate and independent signal
transduction pathways.(dT-dC)),
but a GAGA binding factor was not discerned in root extracts. These
data indicate that binding of a nuclear factor to the GAGA element has
a positive effect on transcription, and they underscore similarities in
the regulation of gsa1 in leaves and nodules despite the
specialization of those tissues for different metabolic processes.
)-aminolevulinic acid; GSA, glutamate 1-semialdehyde; MOPS,
4-morpholinepropanesulfonic acid; kb, kilobase pair(s); PCR, polymerase
chain reaction; bp, base pair(s); MES, 4-morpholineethanesulfonic acid;
PAGE, polyacrylamide gel electrophoresis.))
We thank Drs. Ton Bisseling, Kjeld Marcker, Desh Pal
Verma, and Gary Stacey for cDNAs and antibodies. We also thank
Christine Kaczor for construction of pET3CGSAA.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  L. A. Nogaj, A. Srivastava, R. van Lis, and S. I. Beale Cellular Levels of Glutamyl-tRNA Reductase and Glutamate-1-Semialdehyde Aminotransferase Do Not Control Chlorophyll Synthesis in Chlamydomonas reinhardtii Plant Physiology, September 1, 2005; 139(1): 389 - 396. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Sangwan and M. R. O'Brian Identification of a Soybean Protein That Interacts with GAGA Element Dinucleotide Repeat DNA Plant Physiology, August 1, 2002; 129(4): 1788 - 1794. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Sangwan and M. R. O'Brian Expression of a Soybean Gene Encoding the Tetrapyrrole-Synthesis Enzyme Glutamyl-tRNA Reductase in Symbiotic Root Nodules Plant Physiology, February 1, 1999; 119(2): 593 - 598. [Abstract] [Full Text]
|
|
|
|
 |
|
 I. Hamza, S. Chauhan, R. Hassett, and M. R. O'Brian The Bacterial Irr Protein Is Required for Coordination of Heme Biosynthesis with Iron Availability J. Biol. Chem., August 21, 1998; 273(34): 21669 - 21674. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. A. Santana, K. Pihakaski-Maunsbach, N. Sandal, K. A. Marcker, and A. G. Smith Evidence that the Plant Host Synthesizes the Heme Moiety of Leghemoglobin in Root Nodules Plant Physiology, April 1, 1998; 116(4): 1259 - 1269. [Abstract] [Full Text]
|
|
|
|
|
|
S. Chauhan and M. R. O'Brian A Mutant Bradyrhizobium japonicum [IMAGE]-Aminolevulinic Acid Dehydratase with an Altered Metal Requirement Functions in Situ for Tetrapyrrole Synthesis in Soybean Root Nodules J. Biol. Chem., August 25, 1995; 270(34): 19823 - 19827. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
