Download Interallelic Complementation at the Ubiquitous

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Document related concepts

Proteolysis wikipedia, lookup

Plant virus wikipedia, lookup

Two-hybrid screening wikipedia, lookup

Amino acid synthesis wikipedia, lookup

Plant nutrition wikipedia, lookup

Artificial gene synthesis wikipedia, lookup

Silencer (genetics) wikipedia, lookup

Gartons Agricultural Plant Breeders wikipedia, lookup

Leaf wikipedia, lookup

Specialized pro-resolving mediators wikipedia, lookup

Plant breeding wikipedia, lookup

Point mutation wikipedia, lookup

Transcript
Interallelic Complementation at the Ubiquitous Urease
Coding Locus of Soybean1
Ariel Goldraij, Lesa J. Beamer, and Joe C. Polacco*
Department of Biochemistry (A.G., L.J.B., J.C.P.) and Interdisciplinary Plant Group (A.G., J.C.P.), University
of Missouri, Columbia, Missouri 65211
Soybean (Glycine max [L.] Merrill) mutant aj6 carries a single recessive lesion, aj6, that eliminates ubiquitous urease activity
in leaves and callus while retaining normal embryo-specific urease activity. Consistently, aj6/aj6 plants accumulated urea
in leaves. In crosses of aj6/aj6 by urease mutants at the Eu1, Eu2, and Eu3 loci, F1 individuals exhibited wild-type leaf urease
activity, and the F2 segregated urease-negative individuals, demonstrating that aj6 is not an allele at these loci. F2 of aj6/aj6
crossed with a null mutant lacking the Eu1-encoded embryo-specific urease showed that ubiquitous urease was also inactive
in seeds of aj6/aj6. The cross of aj6/aj6 to eu4/eu4, a mutant previously assigned to the ubiquitous urease structural gene
(R.S. Torisky, J.D. Griffin, R.L. Yenofsky, J.C. Polacco [1994] Mol Gen Genet 242: 404–414), yielded an F1 having 22% ⫾ 11%
of wild-type leaf urease activity. Coding sequences for ubiquitous urease were cloned by reverse transcriptase-polymerase
chain reaction from wild-type, aj6/aj6, and eu4/eu4 leaf RNA. The ubiquitous urease had an 837-amino acid open reading
frame (ORF), 87% identical to the embryo-specific urease. The aj6/aj6 ORF showed an R201C change that cosegregated with
the lack of leaf urease activity in a cross against a urease-positive line, whereas the eu4/eu4 ORF showed a G468E change.
Heteroallelic interaction in F2 progeny of aj6/aj6 ⫻ eu4/eu4 resulted in partially restored leaf urease activity. These results
confirm that aj6/aj6 and eu4/eu4 are mutants affected in the ubiquitous urease structural gene. They also indicate that radical
amino acid changes in distinct domains can be partially compensated in the urease heterotrimer.
The main function of urease in plants is to recycle
N from urea. In soybean (Glycine max [L.] Merrill),
this is particularly important during germination
when storage proteins are mobilized to nourish the
seedling. Most of the large endogenously generated
urea pool comes from Arg (Stebbins and Polacco,
1995), which constitutes 18% of storage protein N
(Micallef and Shelp, 1989) and is actively degraded to
urea and Orn upon germination (Goldraij and Polacco, 1999). Urease catalyzes N reconversion from
urea to ammonia, which is subsequently assimilated
via Gln synthetase (Lam et al., 1996). Urease also
recycles N derived from urea exogenously applied as
a foliar fertilizer (Witte et al., 2002). A potential urease role in maintaining N2 fixation under water stress
was posited in a model by Purcell et al. (2000) and
Vadez and Sinclair (2001), whereby urea is generated
directly from ureides in soybean cultivars (e.g. Maple
arrow) able to fix N2 under water deficit. Ureides are
converted to urea in reactions that are not affected by
water deficit thus avoiding ureide buildup and its
inhibition of N fixation (Serraj et al., 1999). In this
model, urease is essential to N use in soybean cultivars adapted to dry areas where fixed N is the sole or
major N source. Drought-sensitive cultivars, how1
This work was supported by the National Science Foundation
(grant no. IBN–9982213 to J.C.P.).
* Corresponding author; e-mail [email protected]; fax
314 – 882–5635.
Article, publication date, and citation information can be found
at www.plantphysiol.org/cgi/doi/10.1104/pp.103.022699.
ever, are purported to bypass urea production by
generating ammonia directly from ureides.
All plant and bacterial ureases described so far
are metalloenzymes having a nickel metallocenter
(Dixon et al., 1975; Hausinger et al., 2001). In Klebsiella aerogenes, where urease structure and its activation are best understood, the enzyme is a trimer of
trimers (␣, ␤, ␥)3 in which ␣, ␤, and ␥ are encoded by
ureC, ureB, and ureA, respectively. To become functional, apourease is activated by four known accessory proteins (UreD, UreE, UreF, and UreG) to incorporate two nickel atoms per active site (Jabri et al.,
1995; Hausinger et al., 2001).
Plant ureases are multimers (Polacco et al., 1985) of
a single subunit that is colinear with the three bacterial subunits (Mobley et al., 1995). In soybean, there
are two distinct, nonallelic urease isozymes, differing
in both structural and biochemical properties (for a
summary, see Bacanamwo et al., 2002a). The embryospecific urease is synthesized exclusively in developing embryos (Torisky and Polacco, 1990) and is encoded by the Eu1 gene (Meyer-Bothling and Polacco,
1987). Apparent electrophoretic variants are trimers
or hexamers (Buttery and Buzzell, 1971), although
the genetic hexamer can be converted to the trimer by
treatment with heat, Suc, or glycerol (Polacco and
Havir, 1979). The ubiquitous urease is a constitutive
enzyme expressed in all organs and encoded by the
Eu4 gene (Polacco et al., 1989). It is trimeric in all
varieties that we have examined (Polacco et al., 1985).
The absence of ubiquitous urease activity in soybean
Downloaded
from on June
18, 2017 - Published©by
www.plantphysiol.org
Plant Physiology, August 2003, Vol.
132, pp. 1801–1810,
www.plantphysiol.org
2003
American Society of Plant Biologists
Copyright © 2003 American Society of Plant Biologists. All rights reserved.
1801
Goldraij et al.
results in necrotic leaf tips due to urea accumulation
(Stebbins et al., 1991).
At least two other genes, Eu2 and Eu3, encode
accessory proteins necessary for urease activity
(Meyer-Bothling et al., 1987). A mutation in either of
them is pleiotropic, eliminating the activities of both
ureases. Although Eu3 encodes a nickel-binding protein, the plant ortholog of bacterial UreG (Freyermuth et al., 2000), the identity and specific function
of the Eu2 protein is still unknown. Eu2 and Eu3
proteins interact to activate urease; incubation of
mixed crude extracts from developing embryos of
eu2/eu2 and eu3/eu3 mutants slowly gain urease activity (Polacco et al., 1999). Also, the orthologs of
bacterial urease accessory genes ureD and ureF have
been recently identified in soybean, and at least ureF
has been showed to be functional by complementation of the homologous urease-negative mutant of
fission yeast (Schizosaccharomyces pombe; Bacanamwo
et al., 2002b).
We report here the genetic and molecular characterization of a urease-negative mutant, aj6/aj6, and
complete the previous characterization of the eu4
lesion, which was assigned to the ubiquitous urease
structural gene (Torisky et al., 1994). We demonstrate
that both mutants have lesions in different regions of
the Eu4 locus. Moreover, in the heterozygous progeny derived from a cross of the two urease-negative
mutants, urease activity is partially restored, indicating that the products of both alleles interact, with at
least one heterotrimer regaining catalytic activity.
RESULTS
Soybean aj6 Contains a Homozygous Recessive Lesion
Eliminating Ubiquitous Urease Activity
Soybean mutant aj6 was recovered from a pooled
M2 seed population. It exhibited urease-negative
leaves and seeds with variable activity in the M3 and
M4 generations (Polacco et al., 1989). We quantified
urease activity and urea content of seeds and leaves
in a true breeding line of aj6 advanced for several
generations (Fig. 1). Seed urease activity and total
seed urea content were similar in both aj6 and in its
urease-positive progenitor, soybean cv Williams. In
contrast, leaf urease activity was undetectable in aj6.
Consistently, urea accumulated in leaves of aj6 to
levels about 65-fold over wild type. aj6 adult plants
exhibited necrotic leaf tips likely due to “urea burn,”
a typical trait of plants lacking leaf ubiquitous urease
activity (Eskew et al., 1983; Stebbins et al., 1991).
In an aj6 ⫻ wild type cross, the F2 segregation
pattern of leaf urease-positive to leaf urease-negative
individuals was close to 3:1, suggesting that the aj6
trait is due to a single recessive mutation (Table I).
Henceforth, the lesion is called aj6, and the mutant
genotype is designated aj6/aj6.
To test the activity of ubiquitous urease in aj6/aj6
seeds it was necessary to remove the masking
embryo-specific urease. The eu1-sun/eu1-sun mutant
has normal levels of ubiquitous urease but lacks the
embryo-specific urease whose wild-type-specific activity, being almost 8 ⫻ 103 times higher, masks the
expression of ubiquitous urease in the embryo
(Meyer-Bothling and Polacco, 1987). F2 progeny from
an aj6/aj6 ⫻ eu1-sun/eu1-sun cross were analyzed by
the seed chip assay, a qualitative assay in which the
time for alkalinization of the medium, detected by
cresol red, was used to distinguish the activities of
urease isozymes in soybean seeds (Meyer-Bothling
and Polacco, 1987). The segregation ratio in the F2
generation (Table II) was consistent with the presence of two independent lesions in different
isozymes and confirmed that the ubiquitous urease
isozyme is inactive in aj6/aj6 seeds. The urease phenotype of the double mutant seeds, eu1-sun/eu1-sun,
aj6/aj6, resembled that of eu3-e1/eu3-e1, a pleiotropic
Figure 1. Urease phenotype in the soybean mutant aj6. Urease activity (A) and urea content (B) determined in dry seed and
trifoliate leaf of wild type and aj6. Results are means ⫾ SD of three independent experiments.
1802
Downloaded from on June 18, 2017 - Published by www.plantphysiol.org
Copyright © 2003 American Society of Plant Biologists. All rights reserved.
Plant Physiol. Vol. 132, 2003
Interallelic Complementation at the Urease Locus
Table I. F2 transmission of leaf urease trait in aj6
Crossa
Observed F2 Phenotypic Ratiob
F1 Urease Activity
aj6 (乆) ⫻ soybean cv Kenwood (么)
nmol urea h⫺1leaf disk⫺1
⫹:⫺
27.8 ⫾ 11.3(n⫽4)
14:5
n
19
⫺1
⫺1
␹2
0.017
Pc
⬎0.70
Parental activities: soybean cv Kenwood, 26.2 ⫾ 10 (n ⫽ 9); aj6, 0 (n ⫽ 10) nmol urea h leaf disk .
Urease activity was 23.55 ⫾
c
15 and 0 for urease-positive and urease-negative individuals, respectively.
Probability of random deviation from a 3:1 ratio.
a
mutation in the urease accessory gene UreG that lacks
all urease activities (Freyermuth et al., 2000).
Consistent with the absence of leaf urease activity,
a leaf-derived callus of aj6/aj6 was unable to grow in
2.5 mm urea as sole N source (Fig. 2, A and B).
Conversely, in the presence of a permissible N
source, it was not affected by 50 mm urea, a concentration toxic to urease-positive callus due to the
abundant ammonia generated by urease action (Polacco et al., 1989; Stebbins et al., 1991). Callus urease
activity was almost undetectable in aj6/aj6 (Fig. 2C).
Thus, the pattern of urease activity and urea accumulation in aj6/aj6 was similar to that of the partially
characterized eu4/eu4 mutant (Torisky et al., 1994), in
which a mutation linked to or within the structural
gene of the ubiquitous urease was responsible for the
loss of its enzymatic activity.
Allelism Tests and Molecular Analyses of aj6/aj6
In the crosses of aj6/aj6 to mutants at the Eu1, Eu2,
and Eu3 loci, all F1 individuals were urease positive,
and their activities were close to that of the F1 of
aj6/aj6 ⫻ wild type (Tables I and III). These data
along with the patterns of segregation in the F2 generations (Table III) suggested that the aj6 mutation
was not an allele of these genes. Consistent with this
conclusion was the distinct phenotype of aj6/aj6, in
which only the leaf urease was affected. In contrast,
the levels of leaf urease activities in the F1 generation
derived from the aj6/aj6 ⫻ eu4/eu4 cross were notably lower than those of the F1 of the aj6/aj6 crosses to
the other mutants. Given the similarities of the phenotypic traits between aj6 and eu4, we considered the
possibility that both mutations could be affecting the
same locus, complementing each other to produce a
partially active F1 progeny. If this were the case, both
mutations would be linked, and the phenotypic segregation ratio expected in the F2 generation would be
b
1:1 (negative:partially positive), which was close to
that experimentally obtained (Table III).
To identify the aj6 and eu4 mutations, the ubiquitous urease open reading frame (ORF) was obtained
by reverse transcriptase (RT)-PCR from leaf RNA
isolated from aj6/aj6, eu4/eu4, and soybean cv Williams, progenitor of both mutants. No significant
differences were seen in the levels of the amplified
products, suggesting that both mutants have normal
levels of ubiquitous urease transcripts (data not
shown). The deduced amino acid sequence of the
ubiquitous urease ORFs (Fig. 3) showed that in aj6/
aj6, Arg was changed to Cys at position 201 (R201C).
In eu4/eu4, Gly at position 468 was replaced by Glu
(G468E), corroborating at the molecular level that the
previously reported mutation in the locus Eu4 (Torisky et al., 1994) is a lesion in the ubiquitous urease
structural gene. The ubiquitous urease nucleotide sequence in the wild type was essentially identical (one
silent nucleotide change) to the GenBank cDNA sequence of soybean urease (accession no. AJ276866)
from an unidentified cultivar and tissue.
The embryo-specific urease ORF from wild type
(Meyer-Bothling et al., 1987), whose complete coding
sequence has not heretofore been reported, was
cloned by RT-PCR from dry seeds. A comparison of
the deduced amino acid sequence between both soybean ureases and the products of ureA, ureB, and ureC
genes (i.e. ␥, ␤, and ␣, respectively) of K. aerogenes is
shown in Figure 3. We term the single subunits for
soybean ureases ␣uu and ␣eu, for the ubiquitous and
embryo-specific ureases, respectively. ␣uu and ␣eu
share 87% identity and 92% similarity. Soybean ␣uu
was found to share 73% and 76% amino acid identity
and 82% and 86% similarity with leaf ureases of
rice (Oryza sativa; accession no. BAB78715) and
potato (Solanum tuberosum; accession no. CAC43859),
respectively.
Table II. F2 segregation of seed urease phenotype determined by the seed chip assay
Cross
eu1-sun/eu1-sunc (乆)
⫻ aj6/aj6 (么)
Purplea after 20 min
(Embryo-Specific ⫹;
Ubiquitous ⫹ or ⫺)
Pale Pink after 48 h
(Embryo-Specific ⫺;
Ubiquitous ⫹)
Yellow after 5 d
(Embryo-Specific ⫺;
Ubiquitous ⫺)
n
␹2
Pb
77 (72)
15 (18)
4 (6)
96
1.52
⬎0.3
a
b
Color due to the pH increase produced by ammonia generated by urease action.
Probability of random deviation from the expected
ratio (12:3:1) with two degrees of freedom assuming non-linkage between the eu1-sun and aj6 lesions. Ideal expectations are in parenc
theses.
Parental phenotypes are pale pink after 48 h (eu1-sun/eu1-sun) and purple after 20 min (aj6/aj6).
Plant Physiol. Vol. 132, 2003
Downloaded from on June 18, 2017 - Published by www.plantphysiol.org
Copyright © 2003 American Society of Plant Biologists. All rights reserved.
1803
Goldraij et al.
Figure 2. Callus yield and urease activity under
different N nutritional regimes in aj6/aj6. A and
B, Callus of wild type and aj6/aj6 were grown
for 21 d in Murashige and Skoog medium (white
bars), 2.5 mM urea (black bars) and Murashige
and Skoog medium supplemented with 50 mM
urea (gray bars) as N sources. C, Callus urease
activity determined on callus grown on Murashige and Skoog medium. Results are means of
two independent experiments. MS, Murashige
and Skoog medium (N ⫽ 20.6 mM NH4NO3 and
18.8 mM KNO3).
As was shown previously for jack bean (Canavalia
ensiformis) urease (Mobley et al., 1995), both ␣uu and
␣eu align with the three urease bacterial subunits
(Fig. 3). K. aerogenes ␥, ␤, and ␣ and the corresponding regions from ␣uu share 56%, 49%, and 69% identity and 71%, 61%, and 80% similarity, respectively. It
is important to point out that the aj6 and eu4 mutations are allelic, but they are in different “domains,”
homologous to bacterial ␤ and ␣, respectively. A
BLASTp using the deduced amino acid sequence of
soybean ubiquitous urease as a query in the GenBank
(Altschul et al., 1997) showed that all urease amino
acid sequences reported conserve both the R201 altered in aj6/aj6 and the G468 altered in eu4/eu4.
The Lack of Ubiquitous Urease Activity in aj6/aj6
Cosegregates with R201C
To confirm the association between the absence of
ubiquitous urease activity and the mutation at position 201 in ␣uu of aj6/aj6, the cDNA sequences of five
urease-negative and six urease-positive F2 individuals from the aj6/aj6 ⫻ wild type cross were analyzed
(Table IV). As expected, leaf urease-positive individuals were either homozygous or heterozygous at
codon 201, encoding either the wild-type Arg or both
Arg and Cys, respectively. All leaf urease-negative
plants, however, were homozygous at position 201,
encoding Cys. The cosegregation of both traits suggests strongly that R201C in the ubiquitous urease
ORF of aj6/aj6 eliminates ubiquitous urease activity.
Partial Interallelic Complementation between
aj6 and eu4
The F1 of the aj6/aj6 ⫻ eu4/eu4 cross was urease
positive but its activity was clearly less than either
the wild-type or the F1 levels in crosses of aj6/aj6
with wild type and with the mutants eu1-sun/eu1sun, eu2/eu2, and eu3-e1/eu3-e1 (Table III). F2 from
the aj6/aj6 ⫻ eu4/eu4 cross were analyzed for leaf
urease activities and the segregation of the aj6 and
eu4 mutations (Table V). Only those individuals heterozygous for both the aj6 and eu4 alterations were
urease positive and their activity levels ranged from
14% to 47% of wild-type activity. Thus, in the F2
Table III. Allelism tests of aj6/aj6a
Allele
eu1-sun
eu2
eu3-e1
eu4
Locus
Eu1
Eu2
Eu3
Eu4
Leaf Urease Activity
⫹
⫺
⫺
⫺
F1 Urease Activity
Expected F2 Phenotypic Ratiob
F2 Phenotypic Distribution
nmol urea h⫺1 leaf disk⫺1
urease ⫹:urease ⫺
urease ⫹:urease ⫺
26.9 ⫾ 5.4(n⫽4)
21.5 ⫾ 5.9(n⫽6)
25.7 ⫾ 5.7(n⫽4)
5.8 ⫾ 3.1(n⫽7)
3:1
9:7
9:7
9:7(non-linked)
1:1(allelic)
16:4
20:8
13:7
9:7
9:7
n
␹2
Pc
20
28
20
16
16
0.27
2.61
0.62
0
0.25
⬎0.5
⬎0.1
⬎0.3
⬎0.95
⬎0.5
b
aj6 plants were 乆 in all crosses except when crossed with eu1-sun/eu1-sun.
Expected F2 ratios assuming that the lesion in aj6/aj6 is
c
at a locus unlinked to Eu1, Eu2, or Eu3 and either unlinked to (9:7) or allelic with (1:1) Eu4.
Probabilities of random deviation from the
expected values with 1 degree of freedom.
a
1804
Downloaded from on June 18, 2017 - Published by www.plantphysiol.org
Copyright © 2003 American Society of Plant Biologists. All rights reserved.
Plant Physiol. Vol. 132, 2003
Interallelic Complementation at the Urease Locus
Figure 3. Deduced amino acid sequences of soybean ureases aligned with K. aerogenes urease subunits. UreA (␥), UreB (␤),
and UreC (␣) are the subunits of the urease heterotrimer in most bacteria. These subunits are roughly colinear with urease
amino acid sequence of plants. Identities are indicated in black boxes, similarities in gray boxes. The arrows indicate the
R201C and G468E alterations in the ubiquitous ureases of aj6/aj6 and of eu4/eu4, respectively. Accession numbers: soybean
ubiquitous urease, AY230156; soybean embryo-specific urease, AY230157; and K. aerogenes urease subunits UreA,
P18316; UreB, P18315; and UreC, P18314.
generation the two alleles partially complement to
restore some leaf urease activity. The activity in these
heterozygous individuals showed high instability after prolonged tissue storage at ⫺80°C or tissue extraction with buffer (data not shown). Despite this,
urease activity was detected in native gels (Fig. 4). The
identical migration of both ureases suggests strongly
that urease subunits in the heterozygous progeny are
also assembled into a trimer. A trimeric conformation
Plant Physiol. Vol. 132, 2003
was previously reported for soybean ubiquitous urease (Polacco et al., 1985). Urease activity in the F2
progeny also demonstrates that the lack of activity in
aj6/aj6 is not due to the absence of ubiquitous urease
protein. Ubiquitous urease protein was reported to be
present in eu4/eu4 (Polacco et al., 1989). Both genotypic and phenotypic segregation ratios were experimentally close to those expected in the F2 for two
allelic or tightly linked mutations (1:2:1).
Downloaded from on June 18, 2017 - Published by www.plantphysiol.org
Copyright © 2003 American Society of Plant Biologists. All rights reserved.
1805
Goldraij et al.
Table IV. F2 cosegregation of the aj6 lesion with the leaf ureasenegative phenotype
Plant
Ureasea Activity
Phenotype
Urease Locus
Genotype
Urease Amino Acid at
Position 201
5
9
11
14
16
19
8
10
13
17
D
⫹
⫹
⫹
⫹
⫹
⫹
⫺
⫺
⫺
⫺
⫺
aj6/⫹
⫹/⫹
aj6/⫹
aj6/⫹
⫹/⫹
⫹/⫹
aj6/aj6
aj6/aj6
aj6/aj6
aj6/aj6
aj6/aj6
Arg/Cys
Arg
Arg/Cys
Arg/Cys
Arg
Arg
Cys
Cys
Cys
Cys
Cys
Urease activity was 25.7 ⫾ 19.3 and 0 nmol urea h⫺1 leaf disk⫺1
in urease-positive and urease-negative individuals, respectively.
a
DISCUSSION
The aj6 lesion was demonstrated to be a single
recessive mutation (Table I). Like other soybean mutants lacking leaf urease activity, aj6/aj6 accumulated
considerable leaf urea (Fig. 1). The analysis of F2
segregants in the cross of aj6/aj6 to eu1-sun/eu1-sun,
the latter a null mutant in the Eu1 locus that encodes
the embryo-specific urease (Meyer-Bothling and Polacco, 1987), confirmed that the aj6 mutation also
affects the ubiquitous urease activity in seeds (Table
II). Thus, the aj6 mutation effects were similar to
those of the eu4 mutation, which was assigned to the
ubiquitous urease structural gene (Torisky et al.,
1994). Allelism tests of aj6 with eu1-sun and with the
pleiotropic mutations eu2 and eu3-e1, the latter two
involved in the activation of the two soybean urease
isozymes (Meyer-Bothling et al., 1987; Polacco et al.,
1999), demonstrated that the aj6 mutation is not in
Eu1, Eu2, or Eu3 (Table III). The aj6/aj6 ⫻ eu4/eu4 F1
was urease positive but its activity was clearly lower
than wild-type activity, unlike the full F1 activity
levels in crosses of aj6/aj6 with mutants in the Eu1,
Eu2, and Eu3 loci (Table III). These results indicate
that aj6 and eu4 are different but related mutations.
The leaf urease ORFs of wild type, aj6/aj6, and
eu4/eu4 suggested that the lack of ubiquitous urease
activity in each mutant was due to a single nucleotide
missense alteration. In aj6/aj6, the change was Arg to
Cys at position 201 whereas in eu4/eu4, a Gly was
replaced by Glu at position 468. Both changes are
radical and all bacterial, fungal, and plant urease
sequences producing significant alignments with
soybean ubiquitous urease in the GenBank retain
wild-type Arg (201) and Gly (468) at these positions.
Thus both amino acids are likely important for maintaining urease structure and/or function. eu4/eu4
callus was partially corrected with a 15-kD genomic
DNA fragment containing the wild-type ubiquitous
urease coding sequence (Torisky et al., 1994). The
G468E change reported in this work confirms that eu4
is a lesion in the ubiquitous urease structural gene.
1806
Genetic evidence that the R201C alteration is the
cause of the lack of ubiquitous urease activity is
derived from the F2 of two crosses: aj6/aj6 ⫻ wild
type and the aj6/aj6 ⫻ eu4/eu4. In the first case, there
was complete cosegregation of the loss of urease
activity and homozygosity for R201C (Table IV). In
the second, only those individuals heterozygous for
both mutations showed urease activity, albeit at levels that ranged between 14% and 47% of the wildtype activity (Table V). The urease activity detected
only in the heterozygous F2 population means that
there is a functional interaction between polypeptides bearing different mutations that promotes partial restoration of the activity. This constitutes clear
evidence that urease transcripts in aj6/aj6 are translated into protein arguing against alteration(s) in
transcription or translation. Similarly, the hypothesis
of a mutation in a separate gene tightly linked to Eu4
and controlling a posttranslational activation step
specific for the ubiquitous urease is very unlikely. In
this model, the natural candidates to be considered
are the urease accessory genes involved in nickel
transport and emplacement into apourease. Eu3
(UreG) and Eu2 products activate both ureases in
soybean, and mutations in either of them are pleiotropic and eliminate both embryo-specific and ubiquitous ureases (Meyer-Bothling et al., 1987), a trait
not exhibited by aj6/aj6. Further, aj6/aj6 accumulated apparently normal levels of UreF and UreD
transcripts, each with a wild-type ORF (A. Goldraij
and J.C. Polacco, unpublished data). The G468E alteration in the ubiquitous urease of eu4/eu4 confirms
the strong circumstantial evidence we presented earlier that Eu4 is the ubiquitous urease structural gene,
namely cosegregation of a ubiquitous urease RFLP
with the eu4 lesion and correction of eu4/eu4 by a
Table V. Complementation of aj6 and eu4 alleles in the F2 of aj6/
aj6 ⫻ eu4/eu4
Plant
Leaf Urease
Activity
Genotype at aj6
Site (Residue 201)
Genotype at eu4
Site (Residue 468)
Wild-Typea
Activity
(n ⫽ 3)
⫹
⫹
⫺
⫺
⫹
⫹
⫺
⫺
⫺
⫺
⫹
⫺
⫹
⫹
⫺
⫺
aj6/⫹
aj6/⫹
aj6/aj6
⫹/⫹
aj6/⫹
aj6/⫹
aj6/aj6
⫹/⫹
⫹/⫹
aj6/aj6
aj6/⫹
⫹/⫹
aj6/⫹
aj6/⫹
aj6/aj6
⫹/⫹
eu4/⫹
eu4/⫹
⫹/⫹
eu4/eu4
eu4/⫹
eu4/⫹
⫹/⫹
eu4/eu4
eu4/eu4
⫹/⫹
eu4/⫹
eu4/eu4
eu4/⫹
eu4/⫹
⫹/⫹
eu4/eu4
21 ⫾ 10
14 ⫾ 4
0
0
39 ⫾ 11
25 ⫾ 6
0
0
0
0
47 ⫾ 4
0
20 ⫾ 20
27 ⫾ 13
0
0
%
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
a
Wild-type activity was 25 ⫾ 8 nmol urea h⫺1 leaf disk⫺1 (n ⫽ 9).
Downloaded from on June 18, 2017 - Published by www.plantphysiol.org
Copyright © 2003 American Society of Plant Biologists. All rights reserved.
Plant Physiol. Vol. 132, 2003
Interallelic Complementation at the Urease Locus
Figure 4. In-gel detection of urease activity.
Leaf crude extracts (35 ␮L) from wild type and
from aj6/eu4 (plant 5; Table V) were electrophoresed on native polyacrylamide gels. Urease activity was detected by activity staining (A) and by
incubation of gel slices with [14C]urea (B).
genomic clone containing the ubiquitous urease
structural gene (Torisky et al., 1994). This, coupled
with phenotypic and genotypic segregation patterns
in the eu4/eu4 ⫻ aj6/aj6 cross, shows convincingly
that aj6 is a second mutant allele in the Eu4 locus.
The simplest explanation for the interallelic
complementation between the aj6 and eu4 alleles is
the functional interaction of polypeptides mutated in
different sites. In the cross of aj6/aj6 to eu4/eu4, both
mutations are linked, consistent with an experimental phenotypic ratio of leaf urease positive to leaf
urease negative in the F2 close to 1:1 (Tables III and
V). The heterozygous progeny produce two types of
polypeptides each having a mutation in the aj6 or eu4
sites. Association of these polypeptides forms two
types of heterotrimers viz. ␣uu(R201C)2,␣uu(G468E)
and ␣uu(R201C),␣uu(G468E)2, at least one of which
has activity. Considering the clear sequence colinearity of ␣uu with the three (or two) subunits of bacterial
ureases (Mobley et al., 1995), it is reasonable to assign
R201C and G468E to different domains. Therefore,
eu4/aj6 heterozygotes have at least one intact copy of
each domain per heteromeric urease molecule.
Interallelic complementation has been extensively
reported in a wide variety of organisms. A familiar
example is the ␣-complementation in ␤ galactosidase
Plant Physiol. Vol. 132, 2003
of E. coli, which is the basis of identifying bacterial
colonies carrying recombinant DNA (Ullman et al.,
1967). Early reports in plants include interallelic
complementation in maize at the alcohol dehydrogenase (Schwartz, 1975) and Suc synthase (Chourey
and Nelson, 1979) loci. More recently, interallelic
complementation has been reported in a protein required for coordination of cell polarity in Arabidopsis (Grebe et al., 2000). Examples are found in other
organisms such as fungi (Sanz et al., 2002), fruitfly
(Drosophila melanogaster; Ponomareff et al., 2001), and
mouse (Steingrimsson et al., 2003).
Given the similarities between ␣uu and the urease
of K. aerogenes (Fig. 3), we sought to localize the plant
mutations on the three-dimensional structure of the
bacterial enzyme. In the nonameric (␣, ␤, ␥)3 structure of the K. aerogenes urease protein (Jabri et al.,
1995), the homologous positions for the ubiquitous
urease R201 (␤ R70) and G468 (␥ G197) residues are
shown in Figure 5. Neither residue is involved in the
active site of the enzyme. ␤ R70 is very close to the
interface of two trimers, equivalent to the ␣uu:␣uu
interaction. Although not directly involved in contacting the neighboring trimer, the side chain of ␤
R70 forms multiple hydrogen bonds, including one
to ␤ Q35, which in turn interacts with the neighbor-
Downloaded from on June 18, 2017 - Published by www.plantphysiol.org
Copyright © 2003 American Society of Plant Biologists. All rights reserved.
1807
Goldraij et al.
of their ureases can be deduced. Eu3 is the plant
UreG ortholog, but unlike the bacterial protein, it
contains a His-rich N terminus and binds nickel
(Freyermuth et al., 2000). Two eu2/eu2 mutants produce wild-type UreD and UreF proteins (A. Goldraij
and J.C. Polacco, unpublished data), so Eu2 may be
unique in plants. Likewise, the different structural
and biochemical properties of embryo-specific and
ubiquitous ureases (i.e. embryo-specific is 1,000 times
more abundant, whereas its Km is 20 to 600 times
higher than that of ubiquitous [Polacco and Holland,
1994]) raise questions about the activation of the
soybean urease isozymes. Are they activated by the
same pool of accessory proteins? Are these accessory
proteins similarly regulated for activating each
isozyme? Focusing on these topics will contribute to
further elucidation of the complex urease reaction
mechanism in plants.
MATERIALS AND METHODS
Plant Material and Growth Conditions
Figure 5. The three-dimensional structure of urease from K. aerogenes. The three chains (␣, ␤, and ␥) of the bacterial protein are
shown in the same color (yellow, green, or lavender) for each heterotrimer, and each corresponds to the ␣uu monomer of the soybean
trimer. The two nickel ions in each active site are shown as blue
spheres. Residues corresponding to those mutated in soybean ubiquitous urease are highlighted in red (␤ R70 and ␥ G197). Both
residues are near the homotrimeric interfaces, suggesting that they
may interfere with correct folding or assembly of the oligomer. The
urease figure was produced using MOLSCRIPT program (Kraulis,
1991).
ing subunit. This hydrogen-bonding pattern cannot
be maintained in the aj6/aj6 Cys mutant. ␥ G197 is
highly buried, and introduction of a Glu, as in the
G468E alteration of eu4/eu4, would introduce a negatively charged residue into the hydrophobic core of
the protein and, in addition, would likely impose
unfavorable steric interactions. Although G197 is not
immediately adjacent to a subunit interface, it is also
relatively close. Although there is no clear structural
explanation for why R201C and G468E complement
each other, one possibility would be offsetting effects
at the ␣uu:␣uu interface.
We do not know whether urease levels are altered
in aj6/aj6. In a previous work, it was determined that
despite the total lack of activity, the amount of urease
in eu4/eu4 plants was 40% of the wild-type value,
suggesting a possible instability in the inactive urease polypeptides. In any case, the activity in the
heterozygous F2 progeny (Table V) could be reduced
by changes imposed by mutations in urease structure, by reduction in urease protein levels or by the
random association of polypeptides in which only
some combinations produce an active trimer.
Despite generally similar urease structural and activation genes between plants and bacteria (Bacanamwo et al., 2002b), several differences in activation
1808
Unless stated otherwise, wild-type soybean (Glycine max [L.] Merrill) was
cv Williams 82. In some experiments, wild-type soybean was cv Kenwood.
Both cultivars showed the same deduced amino acid sequence and similar
levels of activity for the ubiquitous urease. Soybean mutants aj6/aj6, eu4/
eu4 (Polacco et al., 1989), eu2/eu2, eu3-e1/eu3-e1 (Meyer-Bothling et al.,
1987), and eu1-sun/eu1-sun (Meyer-Bothling and Polacco, 1987) were in the
soybean cv Williams/Williams 82 backgrounds. Seeds were germinated in
the dark at 27°C in rolls of germination paper (Anchor Paper, St. Paul)
moistened with deionized water. Plants were grown in pots on regular soil
in a controlled-environment greenhouse at 25°C with a 16-h/8-h light/dark
regime.
F2 seed phenotypes of the eu1-sun/eu1-sun ⫻ aj6/aj6 cross were distinguished by the seed-chip assay (Meyer-Bothling and Polacco, 1987). Seeds
with low (embryo-specific negative, ubiquitous positive) versus null
(embryo-specific negative, ubiquitous negative) urease levels were distinguished by incubating seed chips 5 d at 37°C in capped tubes.
Callus was induced and cultured on R3 medium as reported previously
(Polacco et al., 1989). In the growth experiment of Figure 2B, callus was
induced on medium identical to R3 except that hormones were 4.86 mg L⫺1
naphthalene acetic acid and 1.8 mg L⫺1 kinetin. Callus was maintained on
S3 medium in which hormones were 5.5 ⫻ 10⫺4 mg L⫺1 each of benzylaminopurine and 2,4-dichlorophenoxyacetic acid.
Plant Crosses and Allelism Tests
aj6/aj6 plants were crossed to wild type and to individuals homozygous
for recessive urease-negative alleles at the Eu1, Eu2, and Eu3 loci at the Iowa
State Soybean Breeding Nursery (Isabela Substation, University of Puerto
Rico, Isabela, Puerto Rico). The eu1-sun/eu1-sun ⫻ aj6/aj6 cross was performed at Bradford Farm (University of Missouri). F1 seeds were tested
non-destructively by the seed chip urease assay. Leaf urease activity was
measured in leaves of F1 plants that were selfed to obtain F2 seeds. Segregation patterns of F2 progeny were statistically analyzed by t tests (P ⬎ 0.05).
Urea Content and Urease Activity Analyses
Urea content was quantified by the phenol-hypochlorite method (Weatherburn, 1967), determining the ammonia generated following urease treatment. Seed and leaf (10th trifoliate) tissue extraction and urea determination
were performed as described by Goldraij and Polacco (1999) and Stebbins et
al. (1991), respectively.
To determine seed urease activity, the seed coat was removed, and a seed
sliver was taken from the side opposite the hilum. The tissue was homog-
Downloaded from on June 18, 2017 - Published by www.plantphysiol.org
Copyright © 2003 American Society of Plant Biologists. All rights reserved.
Plant Physiol. Vol. 132, 2003
Interallelic Complementation at the Urease Locus
enized with a plastic pestle in 300 to 400 ␮L of 0.1 m Tris-maleate and 1 mm
EDTA, pH 7, in a microcentrifuge tube at 4°C. After centrifugation, the
supernatant was mixed with 55 mm urea in a total volume of 250 ␮L and
was incubated at 37°C for 10 min. Aliquots of 0.1 mL were taken, and the
reaction was stopped by the addition of 40 ␮L of 0.64 n H2SO4. After
dilution with 2 mL of water, 0.1 mL of Nessler’s reagent (Fisher Scientific,
Fair Lawn, NJ) was added, and the absorbance read at 425 nm. Routinely, 1
␮mol of NH4Cl gave an A425 of 1.0. To calculate seed urease-specific activity
(␮mol NH4⫹ mg⫺1 protein min⫺1), protein was determined by the Bradford
assay according to the manufacturer (Bio-Rad Laboratories, Hercules, CA).
Leaf urease activity was determined in discs cut from mid-sections of
fully expanded eighth to 10th trifoliate leaves with a number 8 cork borer.
Discs were incubated 2 h at 37°C with 1 mL of 0.1 m Tris-maleate, pH 7,
containing 1 mm EDTA, 5% (v/v) n-propanol, and 10 mm [14C]urea (5.5 mCi
mmol⫺1). Reactions were stopped by the addition of 0.5 mL of 1 n H2SO4.
Acid-released 14CO2 was trapped and quantified as described by MeyerBothling and Polacco (1987). A similar procedure was followed to determine
urease activity in lyophilized callus cultured on R3 medium (Polacco et al.,
1989).
To determine urease activity in native gels, leaf tissue was frozen in
liquid N and ground in a mortar to a fine powder. Leaf powder (0.1 g) was
extracted with 0.5 mL of 50 mm phosphate buffer, pH 7.5, containing 50 mm
NaCl, 1 mm EDTA, and 50% (v/v) glycerol. Immediately before extraction,
dithiothreitol and phenylmethylsulfonyl fluoride were added to 20 mm and
0.5 mm, respectively. Samples were homogenized in a 1.5 polypropylene
tube with a plastic pestle, vortexed for 30 s, and centrifuged at 12,000g for
10 min at 4°C. Supernatants were transferred to new tubes and centrifuged
again for 20 min under the same conditions. Aliquots of clarified crude
extracts were subjected to electrophoresis under non-denaturing conditions
as indicated in Witte and Medina-Escobar (2001) except that the gels were
run in an ice bucket. After electrophoresis, the gel was washed five times
with 50 mL of 5 mm sodium acetate, pH 6, and once with 50 mL of water at
4°C. Urease activity was detected either by staining (Witte and MedinaEscobar, 2001) or by incubating gel slices (0.58 mm) with 10 mm [14C]urea
following the same procedure stated above for the leaf disc urease activity
assay, except that samples were incubated for 20 h.
Cloning and Sequencing of Soybean Ureases
Leaf ubiquitous urease ORFs were cloned from two soybean ureasepositive wild-type cvs Williams and Kenwood and from the aj6/aj6 and
eu4/eu4 mutants, which were induced in soybean cv Williams. The soybean
embryo-specific urease ORF was cloned from dry soybean cv Williams 82
seeds. Total RNA (2–4 ␮g) isolated according to Murfett et al. (1994) was
used as template for cDNA synthesis by the RT-PCR in the ImProm-II
Reverse Transcription System (Promega, Madison, WI), according to the
manufacturer’s instructions. Subsequent amplification of the ORF regions
used primers designed from the chemically determined N-terminal amino
acid sequences and from genomic DNA sequences containing the
C-terminal regions of both soybean ureases (Torisky et al., 1994). For ubiquitous urease, oligonucleotide primers (Integrated DNA Technologies Inc,
Coralville, IA) were 5⬘-ATG AAACTGAGTCCAAGGGAGAT-3⬘ and 5⬘TAAAATTATCTATTGAA CACATAGCTG-3⬘. For embryo-specific urease,
they were 5⬘-ATGAAACTGAGTCCAAGGGAG-3⬘ and 5⬘-GCTTTATTTATGG GAGCTAGTA-3⬘. PCR was performed using 2 ␮L of RT reaction, using
TaKaRa Ex TAQ (Takara Biomedicals, Kyoto) at 59°C annealing temperature
for 35 cycles. All amplification products obtained had the expected size and
were cloned into pGEM-Teasy (Promega). For all ureases, both strands of
cDNA from the PCR products were sequenced using oligonucleotides based
on internal sequences. Point mutations in aj6/aj6 and eu4/eu4 were consistently detected in five clones from different sequencing reactions. Sequencing was done in the DNA Core of the University of Missouri (Columbia).
Identity of sequenced products was confirmed through BLASTn and
BLASTp searches (Altschul et al., 1997).
ACKNOWLEDGMENTS
We thank Elizabeth Hoyos, who performed the callus culture in Figure
2B, and Silvia Cianzio and Dave Sleper for the crosses and provision of F1
seed. Ed Coe critically read the manuscript.
Plant Physiol. Vol. 132, 2003
Received February 25, 2003; returned for revision April 18, 2003; accepted
May 10, 2003.
LITERATURE CITED
Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W,
Lipman DJ (1997) Gapped BLAST and PSI-BLAST: a new generation of
protein database research programs. Nucleic Acids Res 25: 3389–3402
Bacanamwo M, Freyermuth SK, Palacios JM, Witte CP, Polacco JC (2002a)
Nitrification, denitrification and the N cycle. In TM Finan, MR O’Brian,
DB Layzell, JK Vessey, W Newton, eds, Nitrogen Fixation: Global Perspectives. CABI Publishing, Hamilton, Ontario, UK, pp 314–318
Bacanamwo M, Witte CP, Lubbers MW, Polacco JC (2002b) Activation of
the urease of Schizosaccharomyces pombe by accessory proteins from soybean. Mol Genet Genom 268: 525–534
Buttery BR, Buzzell RI (1971) Properties and inheritance of urease isozymes
in soybean seeds. Can J Bot 49: 1101–1105
Chourey PS, Nelson OE (1979) Interallelic complementation at the sh locus
in maize at the enzyme level. Genetics 91: 317–325
Dixon NE, Gazzola TC, Blakeley RL, Zerner B (1975) Jack bean urease (EC
3.5.1.5): a metalloenzyme. A simple biological role for nickel? J Am Chem
Soc 97: 4131–4133
Eskew DL, Welch RM, Carey EE (1983) Nickel: an essential micronutrient
for legumes and possibly all higher plants. Science 222: 621–623
Freyermuth SK, Bacanamwo M, Polacco JC (2000) The soybean Eu3 gene
encodes a N-binding protein necessary for urease activity. Plant J 21: 53–60
Goldraij A, Polacco JC (1999) Arginase is inoperative in developing soybean embryos. Plant Physiol 119: 297–303
Grebe M, Gadea J, Steinmann T, Kientz M, Rahfeld JU, Salchert K, Koncz
C, Jurgens G (2000) A conserved domain of the Arabidopsis GNOM
protein mediates subunit interaction and cyclophilin 5 binding. Plant Cell
12: 343–356
Hausinger RP, Colpas GJ, Soriano A (2001) Urease: a paradigm for proteinassisted metallocenter assembly. American Society for Microbiology
News 67: 79–84
Jabri E, Carr MB, Hausinger RP, Karplus PA (1995) The crystal structure of
urease from Klebsiella aerogenes. Science 268: 998–1004
Kraulis P (1991) MOLSCRIPT: a program to produce both detailed and
schematic plots of protein structures. J Appl Crystallogr 24: 946–950
Lam H-M, Coschigano KT, Oliveira IC, Melo-Oliveira R, Coruzzi GM
(1996) The molecular-genetics of nitrogen assimilation into amino acids
in higher plants. Plant Physiol Plant Mol Biol 47: 569–593
Meyer-Bothling LE, Polacco JC (1987) Mutational analysis of the embryospecific urease locus of soybean. Mol Gen Genet 209: 439–444
Meyer-Bothling LE, Polacco JC, Cianzio SR (1987) Pleiotropic soybean
mutants defective in both urease isozymes. Mol Gen Genet 209:
432–438
Micallef BJ, Shelp BJ (1989) Arginine metabolism in developing soybean
cotyledons: I. Relationship to nitrogen nutrition. Plant Physiol 90: 624–630
Mobley HLT, Island MD, Hausinger RP (1995) Molecular biology of microbial ureases. Microbiol Rev 59: 451–480
Murfett J, Atherton TL, Mou B, Gasser CS, McClure BA (1994) S-RNAse
expressed in transgenic Nicotiana causes S-allele-specific pollen rejection.
Nature 367: 563–566
Polacco JC, Freyermuth SK, Gerendas J, Cianzio SR (1999) Soybean genes
involved in nickel insertion into urease. J Exp Bot 50: 1149–1156
Polacco JC, Havir EA (1979) Comparison of soybean ureas isolated from
seed and tissue culture. J Biol Chem 254: 1707–1715
Polacco JC, Holland MA (1994) Genetic control of plant ureases. In JK
Setlow, ed, Genetic Engineering, 16. Plenum Press, New York, pp 33–48
Polacco JC, Judd AK, Dybing JK, Cianzio SR (1989) A new mutant class of
soybean lacks urease in leaves but not in leaf-derived callus or in roots.
Mol Gen Genet 217: 257–262
Polacco JC, Krueger RW, Winkler RG (1985) Structure and possible ureide
degrading function of the ubiquitous urease of soybean. Plant Physiol 79:
794–800
Ponomareff G, Giordano H, DeLotto Y, DeLotto R (2001) Interallelic
complementation at the Drosophila melanogaster gastrulation defective
locus defines discrete functional domains of the protein. Genetics 159:
635–645
Downloaded from on June 18, 2017 - Published by www.plantphysiol.org
Copyright © 2003 American Society of Plant Biologists. All rights reserved.
1809
Goldraij et al.
Purcell LC, King CA, Ball RA (2000) Soybean cultivar differences in ureides
and the relationship to drought-tolerant nitrogen fixation and manganese
nutrition. Crop Sci 40: 1062–1070
Sanz C, Alvarez MI, Orejas M, Velayos A, Eslava AP, Benito EP (2002)
Interallelic complementation provides genetic evidence for the multimeric organization of the Phycomyces blakesleeanus phytoene dehydrogenase. Eur J Biochem 269: 902–908
Schwartz D (1975) The molecular basis for allelic complementation of
alcohol dehydrogenase mutants of maize. Genetics 79: 207–212
Serraj R, Valdez VV, Denison RF, Sinclair TR (1999) Involvement of ureides
in nitrogen fixation inhibition in soybean. Plant Physiol 119: 289–296
Stebbins NE, Holland MA, Cianzio SR, Polacco JC (1991) Genetic tests of the
roles of the embryonic ureases of soybean. Plant Physiol 97: 1004–1010
Stebbins NE, Polacco JC (1995) Urease is not essential for ureide degradation in soybean. Plant Physiol 109: 169–175
Steingrimsson E, Arnheiter H, Hallson JH, Lamoreux ML, Copeland NG,
Jenkins NA (2003) Interallelic complementation at the mouse mitf locus.
Genetics 163: 267–276
1810
Torisky RS, Griffin JD, Yenofsky RL, Polacco JC (1994) A single gene (Eu4)
encodes the tissue-ubiquitous urease of soybean. Mol Gen Genet 242:
404–414
Torisky RS, Polacco JC (1990) Soybean roots retain the seed urease isozyme
synthesized during embryo development. Plant Physiol 94: 681–689
Ullman A, Jakob F, Monod J (1967) Characterization by in vitro complementation of a peptide corresponding to an operator-proximal segment
of the ␤-galactosidase structural gene of E. coli. J Mol Biol 24: 339–343
Vadez VV, Sinclair TR (2001) Leaf ureide degradation and N2 fixation
tolerance to water deficit in soybean. J Exp Bot 52: 153–159
Weatherburn MB (1967) Phenol-hypochlorite reaction for determination of
ammonia. Anal Chem 39: 971–974
Witte C-P, Medina-Escobar N (2001) In-gel detection of urease with nitroblue tetrazolium and quantification of the enzyme from different crop
plants using the indophenol reaction. Anal Biochem 290: 102–107
Witte C-P, Tiller SA, Taylor MA, Davies HV (2002) Leaf urea metabolism
in potato: urease activity profile and patterns of recovery and distribution
of 15N after foliar urea application in wild type and urease-antisense
transgenics. Plant Physiol 128: 1129–1136
Downloaded from on June 18, 2017 - Published by www.plantphysiol.org
Copyright © 2003 American Society of Plant Biologists. All rights reserved.
Plant Physiol. Vol. 132, 2003