() Loss of Function Mutations in the Arabidopsis Heterotrimeric G protein a Subunit Enhance the Developmental Defects of Brassinosteroid Signaling and Biosynthesis Mutants Yajun Gao 1, 2, Shucai Wang[.]
Trang 1Loss-of-Function Mutations in the Arabidopsis Heterotrimeric G-protein
Signaling and Biosynthesis Mutants
Yajun Gao1, 2, Shucai Wang1, Tadao Asami3and Jin-Gui Chen 1,*
1
Department of Botany, University of British Columbia, Vancouver, British Columbia V6T 1Z4, Canada
2 College of Resources and Environment, Northwest Agriculture and Forestry University, Yangling, Shaanxi 712100, PR China
3 Plant Science Center and Plant Functions Laboratory, RIKEN, 2-1 Hirosawa, Wako-shi, Saitama, 351-0198 Japan
Loss-of-function alleles of the sole heterotrimeric
G-protein a subunit in Arabidopsis, GPA1, display defects
in cell proliferation throughout plant development Previous
studies indicated that GPA1 is involved in brassinosteroid
(BR) response Here we provide genetic evidence that
loss-of-function mutations in GPA1, gpa1-2 and gpa1-4, enhance
the developmental defects of bri1-5, a weak allele of a
BR receptor mutant, and det2-1, a BR-deficient mutant in
Arabidopsis gpa1-2 bri1-5 and gpa1-4 det2-1 double mutants
had shorter hypocotyls, shorter roots and fewer lateral roots,
and displayed more severe dwarfism than bri1-5 and det2-1
single mutants, respectively By using the Arabidopsis
hypo-cotyl as a model system where the parameters of cell division
and cell elongation can be simultaneously measured, we found
that gpa1 can specifically enhance the cell division defects
of bri1-5 and det2-1 mutants Similarly, gpa1 specifically
enhances cell division defects in the primary roots of bri1-5
and det2-1 mutants Furthermore, an additive effect on cell
division between gpa1 and bri1-5 or det2-1 mutations was
observed in the hypocotyls, whereas a synergistic effect was
observed in the roots Taken together, these results provided
the first genetic evidence that G-protein- and BR-mediated
pathways may be converged to modulate cell proliferation in
a cell/tissue-specific manner
Keywords: Arabidopsis — Brassinosteroid (BR) — Cell
proliferation — Heterotrimeric G-protein a subunit
(GPA1) — Hypocotyl — Root
Abbreviations: BR, brassinosteroid; GPA1, Arabidopsis
heterotrimeric G-protein a subunit; GPCR, G-protein-coupled
receptor; G-proteins, heterotrimeric guanine nucleotide-binding
proteins.
Introduction
The heterotrimeric guanine nucleotide-binding proteins
(G-proteins) act as critical molecular switches in diverse
signal transduction pathways in eukaryotes (Gilman 1987,
Hamm 1998, Neubig and Siderovski 2002, Pierce et al 2002) Plants use G-proteins to regulate multiple develop-mental processes and hormone responses (reviewed by Perfus-Barbeoch et al 2004, Chen 2008) Loss-of-function mutations in the G-protein a subunit (Ga) and b subunit (Gb) in plants confer altered sensitivities to multiple hormones including auxin, ABA, gibberellins, brassinoste-roids (BRs) and jasmonic acid (Ueguchi-Tanaka et al 2000, Wang et al 2001, Ullah et al 2002, Ullah et al 2003, Chen
et al 2004, Trusov et al 2006, Wang et al 2006, Trusov
et al 2007) Many morphological phenotypes observed
in the loss-of-function mutants of Arabidopsis Ga and
Gb subunits are attributed to the modulatory role of G-proteins in cell proliferation (Ullah et al 2001, Chen
et al 2003, Ullah et al 2003, Chen et al 2006)
G-proteins couple the recognition of extracellular sig-nals by cell surface G-protein-coupled receptors (GPCRs)
to activation of downstream effectors (Gilman 1987) In the classical signaling paradigm, upon GPCR activation, the
Ga of the G-protein complex undergoes a conformational change, resulting in GDP/GTP exchange and the dissocia-tion of Gbg dimer from the complex Activated Ga (GTP-bound) and liberated Gbg then bind to down-stream effectors In Arabidopsis, the Ga is encoded by a single gene, GPA1 (Ma et al 1990) Although GPA1 plays
a regulatory role in multiple developmental processes and hormonal responses, the upstream (GPCR) and down-stream (effector) components in the G-protein signaling pathway remain largely elusive GPCRs are proteins that typically have seven-transmembrane domains There are several dozen proteins that contain seven-transmembrane domains in Arabidopsis (Moriyama et al 2006, Temple and Jones 2007) However, only two such proteins, GCR1 and AtRGS1, have been shown to bind GPA1 physically (Chen
et al 2003, Pandey and Assmann 2004), but no ligand has been identified for either GCR1 or AtRGS1 Recently, GCR2 has been proposed to be a GPCR for ABA (Liu et al 2007) However, GCR2 was not clearly predicted as a seven-transmembrane domain-containing protein (Gao
et al 2007, Johnston et al 2007, Illingworth et al 2008)
*Corresponding author: E-mail, jingui@interchange.ubc.ca; Fax, þ1-604-822-6089.
doi:10.1093/pcp/pcn078, available online at www.pcp.oxfordjournals.org
All rights reserved For permissions, please email: journals.permissions@oxfordjournals.org
1013
Trang 2There are several proteins that could act as potential
downstream effectors for GPA1, including AtPirin1 (Lapik
and Kaufman 2003), PLDa1 (Zhao and Wang 2004), PD1
(Warpeha et al 2006) and THF1 (Huang et al 2006)
However, a complete cascade in any given developmental
process or hormonal response mediated by the G-proteins is
still lacking, and the mechanism through which G-proteins
regulate phenotypic and developmental plasticity remains
unclear (Assmann 2004)
Previous studies indicated that a loss-of-function allele
of the Ga subunit in Arabidopsis, GPA1, displayed reduced
sensitivities to brassinolide, a biologically active form of
BR, in hypocotyl growth elongation and seed germination
assays (Ullah et al 2002) In rice, a loss-of-function allele
of Ga, d1, displayed slightly reduced sensitivity to BR in
BR-induced OsBLE2 expression (Yang et al 2003) In the
assays of root elongation inhibition, lamina inclination
promotion and coleoptile elongation promotion, rice d1
mutants also displayed reduced sensitivities to BR (Wang
et al 2006) These findings suggested that G-proteins may
have a role in the BR-mediated pathway, which led us to
investigate further the relationship between the G-proteins
and the key components of BR signaling pathways In
particular, we wanted to investigate the relationship
between GPA1, the sole Ga subunit in Arabidopsis, and
BRI1, a receptor for BR, in the modulation of cell
prolif-eration We demonstrate that the loss-of-function
muta-tions in GPA1 specifically enhance the cell division defects
in bri1-5, a weak mutant allele of BRI1 We propose that
GPA1- and BR-mediated pathways may act in concert to
modulate cell proliferation
Results
Loss-of-function mutations in GPA1 enhance the dwarfism
of BR receptor and biosynthesis mutants
The Ga subunit of the heterotrimeric G-protein
com-plex is encoded by a single gene, GPA1, in Arabidopsis
Previous studies indicated that the GPA1 may be involved
in BR responses (Ullah et al 2002, Chen et al 2004)
In order to obtain insight into the relationship between
GPA1- and the BR-mediated pathways, we generated
double mutants between gpa1-2 (Ullah et al 2001), a
loss-of-function allele of GPA1, and bri1-5 (Li and Chory 1997,
Noguchi et al 1999a), a weak mutant allele of BRI1 which
encodes a receptor for BR Both gpa1-2 and bri1-5 are in
the Wassilewskija (WS) ecotypic background Because the
regulation of BR homeostasis is an important step in BR
signaling and because both BR receptor and biosynthesis
mutants are dwarf, we also generated double mutants
between gpa1-4, a loss-of-function allele of GPA1 in the
Columbia (Col) ecotypic background, and det2-1 (in the Col
ecotypic background), a mutant which has a defect in the
early step of BR biosynthesis (Li et al 1996, Fujioka et al
1997, Li et al 1997, Noguchi et al 1999b)
Under normal growth conditions (e.g 14/10 h photo-period at 140 mmol m–2s–1, 238C), gpa1 mutants do not display dwarfism whereas bri1-5 and det2-1 are dwarf, consistent with previous reports (Li et al 1996, Fujioka
et al 1997, Li and Chory 1997, Li et al 1997, Noguchi et al 1999a, Noguchi et al 1999b, Ullah et al 2001, Ullah et al 2003) Interestingly, we found that gpa1 mutations could significantly enhance the dwarfism of bri1-5 and det2-1 mutants (Fig 1), suggestive of a possible genetic interaction between GPA1- and BR-mediated pathways
C
WS
WS
Col
gpa1-4 det2-1
gpa1-4 det2-1
gpa1-2
gpa1-2
bri1-5
bri1-5
gpa1-2 bri1-5
gpa1-2 bri1-5
*
*
# 15
12 9 6 3 0
D
Col
gpa1-4 gpa1-4 det2-1 det2-1
*
*#
15
12 9 6 3 0
Fig 1 Plant heights of gpa1-2 bri1-5 and gpa1-4 det2-1 double mutants (A) Phenotype of a 58-day-old gpa1-2 bri1-5 double mutant (B) Height of a gpa1-2 bri1-5 double mutant (C) Phenotype
of a 58-day-old gpa1-4 det2-1 double mutant (D) Height of a gpa1-4 det2-1 double mutant Wild-type and mutant plants were grown under identical conditions with a 14/10 h photoperiod Scale bars in (A) and (C), 1 cm The heights of plants were mea-sured at maturity Shown are the average heights of at least
10 plants for each genotype SE
significant difference from the wild-type (WS or Col), P50.05 #significant difference from bri1-5
or det2-1, P50.05.
Trang 3BR receptor mutants and biosynthesis mutants have defects
in both cell elongation and cell division in hypocotyls
Because BR has a prominent role in regulating cell
elongation whereas GPA1 is a critical modulator for cell
division, we wanted to investigate further the cellular basis
of the observed enhanced phenotypes of bri1-5 and det2-1
by gpa1 mutations (Fig 1) However, the plant height
reflects the combined effect of cell division at the shoot
apical meristem and at the intercalary meristem and cell
elongation along the inflorescence stem In addition, the cell
division and cell elongation in the primary inflorescence
stem vary at different developmental stages bri1-5 and
det2-1 mutants are slightly late flowering, whereas gpa1
mutants have wild-type flowering time These factors made
a direct measurement and comparison of cell division and
cell elongation on the inflorescence stems of these single and
double mutants unreliable Therefore, we sought a reliable
and robust system in which cell division and cell
elonga-tion can be simultaneously measured and compared in these
mutants For this purpose, we selected the hypocotyls,
because the number of cells in the epidermis and cortex
is pre-determined during embryogenesis and little cell
division occurs in the epidermis or cortex in dark- or
light-grown Arabidopsis seedlings (Gendreau et al 1997)
A shorter hypocotyl, compared with the wild type, in any
given mutants could be due to a defect in either cell division
(pre-determined during embryogenesis) or cell elongation
(determined post-embryogenesis), or both By counting
the number of cells and measuring the length of cells in
a single cell file longitudinally from the base to the top of
a hypocotyl, defects in cell division and/or cell elongation
can be simultaneously determined More importantly, this
approach had been used previously to determine the cellular
basis of the short hypocotyl phenotype of gpa1 mutants
(Ullah et al 2001), and a loss-of-function allele of the sole
Arabidopsis Gb (Ullah et al 2003)
As reported previously (Ullah et al 2001, Chen et al.,
2003, Jones et al 2003, Chen et al 2004, Chen et al 2006),
etiolated seedlings of the gpa1 mutants, gpa1-2 and gpa1-4,
have short hypocotyls and partially opened hooks
(Fig 2A–D) The phenotype of gpa1-2 is comparable with
that of bri1-5, but bri1-5 mutants are about 30% shorter
than gpa1 mutants (Fig 2A, B) The phenotype of gpa1-4 is
similar to that of det2-1, but det2-1 mutants have even
shorter hypocotyl and are hookless (Fig 2C, D) Consistent
with these observations, gpa1-2 and gpa1-4 mutants are
also similar to seedlings treated with Brz2001, a specific BR
biosynthesis inhibitor (Sekimata et al 2001) On the basis
of these results, we concluded that both GPA1 and BR
have a role in regulating hypocotyl growth and that the
Arabidopsis hypocotyl is a suitable system for dissecting the
role of GPA1 and BR in regulating cell elongation and cell
division
Although it has been recognized for many years that etiolated bri1-5 and det2-1 mutant seedlings have shorter hypocotyls, compared with the wild type, the cellular basis for such shortness was unclear Because BR has a pro-minent role in regulating cell elongation over cell division,
we expected that the reduced lengths of hypocotyls in bri1-5 and det2-1 mutants were due to the defect in cell elongation Indeed, we found that the length of hypocotyl epidermal cells was dramatically reduced in bri1-5 and det2-1 mutants (Fig 2E, F) Interestingly, we found that the number of epidermal cells in bri1-5 and det2-1 mutants was also significantly reduced (Fig 2E, F) From the base (hypo-cotyl/root junction) to the top (lateral to cotyledons) of
a hypocotyl, wild-type (WS or Col) plants have about
21 cells in a single cell file longitudinally However, bri1-5 and det2-1 mutants had only about 15 cells (Fig 2E, F) These results provided direct evidence that BR has a role in regulating hypocotyl cell division and that the short hypo-cotyls of bri1-5 and det2-1 mutants are due to reductions
in both cell elongation and cell division
We extended our analysis to other cell types in the hypocotyl Arabidopsis hypocotyls have two layers of cortex cells, outer and inner cortex cells We counted the number and measured the length of outer cortex cells (located just beneath the epidermal cells) in a single cell file longitudinally from the base to the top of a hypocotyl We found that the length of outer cortex cells was dramatically reduced in the hypocotyls of bri1-5 and det2-1 mutants, compared with that in wild-type (Supplementary Fig S1) Similar to that of epidermal cells, the number of outer cortex cells was also significantly reduced in the hypocotyls
of bri1-5 and det2-1 mutants (Supplementary Fig S1) These results supported the view that the short hypocotyls
of bri1-5 and det2-1 mutants are due to reductions in both cell elongation and cell division
Because epidermal cells on the plant surface are gener-ally believed to have a more important role than cortex cells
in determining the shape of an organ, and epidermal cells
on hypocotyls can be easily recognized due to their location and shape, in our subsequent experiments we specifically focused on the epidermal cells to study cell division and cell elongation in hypocotyls We compared the cell number (indicator of cell division) and cell length (indicator of cell elongation) in gpa1, bri1-5 and det2-1 single and double mutants to determine the relationship between GPA1 and BRI or DET2 in regulating cell division and cell elongation
in hypocotyls
Because the cell division defect in the hypocotyls of
BR signaling or biosynthesis mutants had not been reported previously and because bri1-5 is a weak mutant allele of BRI1, we wanted to examine further if such a defect also occurs in a strong mutant allele of BRI1 Therefore, we used the same assay to examine the number and length of
Trang 4hypocotyl epidermal cells in bri1-4 (Li and Chory 1997.
Noguchi et al 1999a), a strong and null allele for BRI1
in the WS ecotypic background Because bri1-4 mutants
rarely produce seeds, we picked up bri1-4 mutants from
the progeny of a plant heterozygous for the bri1-4 locus
In 4-day-old, light-grown seedlings, bri1-4 mutants, evident
by their extreme shortness of hypocotyl and dark green
cotyledons, were readily picked up from this segregating
population We found that similar to that of the bri1-5
mutant, the short hypocotyl of bri1-4 was due to reductions
in both cell elongation and cell division (Supplementary
Fig S2)
We also examined the cell number and length of
hypocotyl epidermal cells in another BR-deficient mutant,
dwf4-102(Azpiroz et al 1998, Choe et al 1998, Nakamoto
et al 2006), which is a T-DNA insertion mutant in the Col
ecotypic background and has defects in the key steps of
BR biosynthesis (Choe et al 1998) Again, we found that
the short hypocotyl of dwf4-102 was due to reductions
in both cell elongation and cell division (Supplementary Fig S2) Therefore, we concluded that BR regulates both cell elongation and cell division in hypocotyls
Loss-of-function mutations in GPA1 enhance the cell division defect in the hypocotyls of bri1-5 and det2-1 mutants Having determined that both GPA1 and BR have a role in regulating cell division in the hypocotyl epidermal cells, we wanted to investigate the relationship between GPA1 and BR in the modulation of cell proliferation by examining cell elongation and cell division defects in gpa1-2 bri1-5 and gpa1-4 det2-1 double mutants and comparing them with those in single mutants We found that the hypocotyls of gpa1-2 bri1-5 double mutants were signifi-cantly shorter than those of the bri1-5 single mutant when grown in the dark (Fig 3A, B) Similarly, in etiolated seedlings, the hypocotyls of gpa1-4 det2-1 double mutants
700 600
500 400 300 200 100 0
1000 Col
det2-1
800 600
400 200 0
Cell number
13 16 19 21
Cell number
13 16 19 21
WS
bri1-5
*
*
* 8.0
6.0
4.0
2.0
0 WS
WS + Brz2001
gpa1-2bri1-5
*
*
* 6.0
5.0 4.0
2.0 1.0 3.0
0
C ol
Col + Brz2001
gpa1-4det2-1
2 mm
Fig 2 gpa1, bri1-5 and det2-1 mutants share the similar short hypocotyl phenotype in etiolated seedlings (A) Phenotype of 3-day-old, dark-grown mutant seedlings in the Wassilewskija (WS) ecotypic background (B) The hypocotyl length of 3-day-old, dark-grown mutant seedlings (C) Phenotype of 3-day-old, dark-grown mutant seedlings in the Columbia-0 (Col) ecotypic background (D) The hypocotyl length of 3-day-old, dark-grown mutant seedlings Brz2001, a specific biosynthesis inhibitor of BR, was applied at 2 mM Scale bars in (A) and (C), 2 mm
significant difference from WS or Col, P50.05 Shown in (B) and (D) are the average lengths of hypocotyls from at least
20 seedlings for each genotype SE (E) The number and length of hypocotyl epidermal cells in 3-day-old, dark-grown WS and bri1-5 mutant seedlings (F) The number and length of hypocotyl epidermal cells in 3-day-old, dark-grown Col and det2-1 mutant seedlings Shown in (E) and (F) are the average cell lengths of 10 seedlings for each genotype SE.
Trang 5were significantly shorter than those of the det2-1 single
mutant (Fig 3C, D)
Light plays an important role in regulating hypocotyl
growth The growth characteristics of hypocotyl epidermal
cells in light-grown Arabidopsis seedlings are different
from those in etiolated seedlings (Gendreau et al 1997)
Therefore, we also measured the hypocotyl lengths of
light-grown gpa1-2 bri1-5 and gpa1-4 det2-1 mutant seedlings
and compared them with those of bri1-5 and det2-1 single
mutants, respectively Because bri1-5 and det2-1 mutants
have very short hypocotyls under normal light conditions
(e.g 140 mmol m–2s–1), we adopted low-light conditions
(30 mmol m–2s–1) under which the hypocotyls of bri1-5 and
det2-1mutants are longer, making it easier and more
accu-rate to quantify the number and length of epidermal cells
We found that in low-light-grown seedlings, the hypocotyl lengths of bri1-5 and det2-1 were also further reduced by gpa1-2 and gpa1-4 mutations, respectively (Figs 4A, B, and 5A, B) Taken together, these results suggested that gpa1 mutations can enhance the hypocotyl phenotype of dark- or low-light-grown bri1-5 and det2-1 seedlings Subsequently, we counted the number and measured the length of hypocotyl epidermal cells in a single cell file longitudinally in gpa1-2 bri1-5 and gpa1-4 det2-1 double mutants, and compared them with those in bri1-5 and det2-1single mutants, respectively Interestingly, we found that the number of epidermal cells was further reduced in gpa1-2 bri1-5 double mutants, compared with gpa1-2 or bri1-5 single mutants (Fig 4C, F) gpa1-2 bri1-5 double mutants only had about 11 epidermal cells, compared with approximately 15 cells in gpa1-2 or bri1-5 single mutants and 21 cells in wild-type plants (Fig 4C, F) However, the cell elongation in the gpa1-2 bri1-5 double mutant, measured by the average length or maximal length of epidermal cells (Fig 4D–F), was similar to that in the bri1-5 single mutant Similarly, we found that that the number of epidermal cells in gpa1-4 det2-1 double mutants was significantly reduced, compared with that in gpa1-4 and det2-1 single mutants (Fig 5C, F), whereas the average length and maximal length of hypocotyl epidermal cells in gpa1-4 det2-1double mutants were similar to those in det2-1 single mutants (Fig 5D–F) Taken together, these results suggested that gpa1 mutations can specifically enhance the cell division defects in the hypocotyls of bri1-5 and det2-1 mutants
Loss-of-function mutations in GPA1 enhance the defect in root development of bri1-5 and det2-1 mutants
We extended our analysis of the relationship between gpa1-2and bri1-5, and between gpa1-4 and det2-1, to non-aerial organs, specifically the roots We measured the length
of the primary root and counted the number of lateral roots
to assess the impact on and relationship of these mutations
to root development gpa1 mutants have primary roots of normal length and produced fewer lateral roots (Ullah et al
2003, Chen et al 2006), whereas both bri1-5 and det2-1 mutants have short primary roots and produced fewer lateral roots (Fig 6A, B, D, E) We found that gpa1 mutations significantly enhanced the short primary root phenotype of the bri1-5 and det2-1 mutants (Fig 6A, B, D, E) The number of lateral roots was also significantly reduced in gpa1-2 bri1-5 and gpa1-4 det2-1 double mutants, compared with bri1-5 and det2-1 single mutants, respec-tively (Fig 6A, C, D, F) These results suggested that similar to the situation of aerial organs, gpa1 mutations can also enhance the defects in root development of bri1-5 and det2-1mutants
B
-5
*
*
*
#
8.0
6.0
4.0
2.0
0
WS
gpa1-2
gpa1-2 bri1-5
bri1-5
D
*
*
*
#
8.0
6.0
4.0
2.0
0
C ol
gpa1-4 gpa1-4
det2-1
det2-1
Fig 3 Phenotype of dark-grown seedlings of gpa1-2 bri1-5
and gpa1-4 det2-1 double mutants (A) Phenotype of 3-day-old,
dark-grown gpa1-2 bri1-5 double mutants Scale bar, 2 mm.
(B) Hypocotyl length of gpa1-2 bri1-5 double mutants
significant difference from WS, P50.05 #significant difference from the
bri1-5 mutant, P50.05 (C) Phenotype of 3-day-old, dark-grown
gpa1-4 det2-1 double mutants Scale bar, 2 mm (D) Hypocotyl
length of gpa1-4 det2-1 double mutants
significant difference from Col, P50.05 #significant difference from det2-1 mutant,
P50.05 Shown in (B) and (D) are average lengths of hypocotyls
of 20 seedlings for each genotype SE.
Trang 6Subsequently, we wanted to determine if such
enhanced defects in root development caused by gpa1
mutations were also due to the enhanced defects in cell
division Unlike hypocotyls in which the number of cells
in the epidermis and cortex is pre-determined during embryogenesis and little cell division occurs in the epidermis
or cortex, cell division occurs in various stages of root development Cell division in the root apical meristem and
gpa1-2 bri1-5 gpa1-2 bri1-5
*
*
*
#
6.0
5.0 4.0 3.0 2.0 1.0 0
WS
gpa1-2bri1-5
gpa1-2 bri1-5
* *
*#
25
C
20
15
10
5
0
WS
gpa1-2bri1-5 gpa1-2 bri1-5
* *
350 D
m) 300 250 200 150 100 50 0
WS
gpa1-2bri1-5 gpa1-2 bri1-5
* *
500 E
400 300 200 100 0
600
500
F
400
300
200
100
0
Cell number
WS
gpa1-2
gpa1-2 bri1-5 bri1-5
Fig 4 The number and length of hypocotyl epidermal cells in
low-light-grown gpa1-2 bri1-5 double mutants (A) Phenotype of
4-day-old, low-light-grown seedlings Scale bar, 2 mm Arrowheads
point to the base and the top of hypocotyls (B) Hypocotyl length of
4-day-old, low-light-grown seedlings (C) Number of hypocotyl
epidermal cells (D) The average lengths of hypocotyl epidermal
cells in a single cell file (from base to top) (E) The maximal length
of hypocotyl epidermal cells The three longest hypocotyl
epider-mal cells in each seedling were examined
significant difference from WS, P50.05 #significant difference from bri1-5 mutant,
P50.05 (F) The number and length of hypocotyl epidermal cells in
a single cell file of gpa1-2 bri1-5 double mutants, compared with
that in single mutants Shown in (B) are the averages of hypocotyl
length from 20 seedlings for each genotype SE Shown in (C–F)
are the averages of 10 seedlings for each genotype SE.
Col
gpa1-4 det2-1 gpa1-4 de
gpa1-4 det2-1 gpa1-4 det2-1
*
*
*
#
6.0 B A
5.0 4.0 3.0 2.0 1.0 0
Col
gpa1-4 det2-1
gpa1-4 det2-1
* *
*#
25 C
20 15 10 5 0
Col
gpa1-4 det2-1 gpa1-4 det2-1
* *
350 D
m) 300 250 200 150 100 50 0
Col
gpa1-4 det2-1 gpa1-4 det2-1
* *
600 500 E
400 300 200 100 0
600 500 F
400 300 200 100 0
Cell number
Col
gpa1-4 gpa1-4 det2-1 det2-1
Fig 5 The number and length of hypocotyl epidermal cells in low-light-grown gpa1-4 det2-1 double mutants (A) Phenotype of 4-day-old, low-light-grown seedlings Scale bar, 2 mm Arrowheads point to the base and the top of hypocotyls (B) Hypocotyl length of 4-day-old, low-light-grown seedlings (C) Number of hypocotyl epidermal cells (D) The average lengths of hypocotyl epidermal cells in a single cell file (from base to top) (E) The maximal length
of hypocotyl epidermal cells The three longest hypocotyl epi-dermal cells in each seedling were examined (F) The number and length of hypocotyl epidermal cells in a single cell file of gpa1-4 det2-1 double mutants, compared with that in single mutants.
significant difference from Col, P50.05 #significant difference from the det2-1 mutant, P50.05 Shown in (B) are the averages of hypocotyl length from 20 seedlings for each genotype SE Shown
in (C–F) are the averages of 10 seedlings for each genotype SE.
Trang 7cell elongation in the elongation zone determine the length
of the primary root, whereas the formation of the lateral
root requires the activation of or re-entry into the cell
cycle in pericycle founder cells Therefore, a short root
could be due to defects in cell division or cell elongation,
or both The number of lateral roots is generally regarded
as an indirect measurement of cell division in the pericycle
founder cells, although initiation of cell division in the
pericycle does not always lead to lateral root initiation
(Vanneste et al 2005) However, because both bri1-5 and
det2-1mutants also have short roots, a direct comparison
of the absolute number of lateral roots between single
and double mutants may not be a reliable method for determining the relationship between gpa1 mutations and bri1-5 or det2-1 mutations in cell division during the pro-cess of lateral root formation Therefore, in this study,
we specifically focused on the primary root to determine the cellular basis of the enhanced short root phenotypes of gpa1-2 bri1-5and gpa1-4 det2-1 double mutants
We used a similar assay to that used previously to determine the cellular basis of altered length of primary roots in G-protein subunit and signaling component mutants (Chen et al 2003, Chen et al 2006) We measured the length of cortex cells in the root hair zone in which the
D
WS gpa1-2 bri1-5 gpa1-2 bri1-5
Col gpa1-4 det2-1 gpa1-4 det2-1
*
*
*
#
80
60
40
20
0
W S
gpa1-2
gpa1-2 bri1-5 bri1-5
*
*
#
30 35
25
15 20
10 5 0 WS
gpa1-2
gpa1-2 bri1-5 bri1-5
F
*
*
30 25
15 20
10 5 0 Col
gpa1-4 g a1-4 det2-1 det2-1
E
*
*
#
60 70
50
30 40
20 10 0 Col
gpa1-4 gpa1-4 det2-1 det2-1
Fig 6 Root phenotype of gpa1-2 bri1-5 and gpa1-4 det2-1 double mutants (A) Phenotype of 7-day-old, light-grown gpa1-2 bri1-5 seedlings Scale bar, 5 mm (B) Length of the primary root of gpa1-2 bri1-5 mutants (C) The number of lateral roots of gpa1-2 bri1-5 mutants (D) Phenotype of 7-day-old, light-grown gpa1-4 det2-1 seedlings (E) Length of the primary root of gpa1-4 det2-1 mutants (F) The number of lateral roots of gpa1-4 det2-1 mutants Shown in (B), (C), (E) and (F) are the averages of 10 seedlings for each genotype SE Data were collected from 10-day-old, light-grown seedlings
significant difference from WS or Col, P50.05 #significant difference from the bri1-5 or det2-1 mutant, P50.05.
Trang 8cortex cells have finished elongation The length of cortex
cells in this region is generally used to reflect their ability
for cell elongation (occurring in the cell elongation zone)
The rate of cell production (number of cells h–1), which is
generally used to reflect cell division activity in the root
apical meristem, was calculated by using the rate of primary
root growth (mm h–1) divided by the average length of
mature cortex cells (mm) As reported previously (Chen
et al 2006b), gpa1 mutants had the wild-type length of
cortex cells and the wild-type rate of root cell production
(Fig 7) While both bri1-5 and det2-1 mutants had shorter
cortex cells (an indicator of cell elongation defect),
com-pared with the wild type, they also had a reduced rate of
root cell production (an indicator of cell division defect) (Fig 7) Interestingly, the reduced rate of root cell pro-duction in the bri1-5 and det2-1 mutants was significantly enhanced by gpa1 mutations (Fig 7B, D) These results suggested that gpa1 mutations enhance the shortness of primary root of bri1-5 and det2-1 mutants by specifically enhancing their cell division defects
Discussion
Heterotrimeric G-protein complexes may serve as
a nexus for the signal regulation of multiple develop-mental processes and hormonal responses in plants Loss-of-function mutations in G-protein subunits resulted
in multiple morphological and conditional phenotypes (reviewed by Perfus-Barbeoch et al 2004, Chen 2008), and conferred defects in cell division throughout plant develop-ment (Ullah et al 2001, Chen et al 2003, Ullah et al
2003, Chen et al 2006) Loss-of-function alleles of Ga in Arabidopsis and in rice displayed altered sensitivities to
BR (Ullah et al 2002, Wang et al 2006), suggesting that G-proteins are involved in the BR-mediated pathway Here
we provide evidence that loss-of-function mutations in the sole Ga in Arabidopsis, GPA1, can specifically enhance the cell division defects of BR signaling and biosynthesis mutants We propose that GPA1- and BR-mediated path-ways may act in concert to modulate cell proliferation in
a cell/tissue-specific manner
Role of tthe heterotrimeric G-protein a subunit in cell elongation and cell division
GPA1 has been known as a critical modulator of cell division (Ullah et al 2001, Chen et al 2003, Ullah et al
2003, Chen et al 2006) A role for GPA1 in cell elongation has not been unequivocally established in Arabidopsis Cell division in the hypocotyl epidermal cells of gpa1 mutants is reduced, but cell elongation is unaffected (Figs 4, 5), sup-porting a prominent role for GPA1 in modulating cell division over cell elongation Because the length of epi-dermal cells along a hypocotyl appears to have a positional effect, with the longest cell located near the middle of
a hypocotyl in both wild-type and gpa1 mutants under our assay conditions (Figs 4, 5), and because gpa1 mutants have fewer cells than the wild type, a direct cell to cell comparison may not reflect the overall difference in cell elongation between wild-type and gpa1 mutants Therefore,
we chose to use the parameters of average cell length of all cells in a single cell file and the maximal cell length
of the three longest cells in a single cell file to compare the difference in cell elongation among different genotypes We found that the average cell length and maximal cell length
of hypocotyl epidermal cells in gpa1 mutants do not differ significantly from those of wild-type seedlings (Figs 4, 5)
A
*
*
#
#
140
120
100
60
80
40
20
0
W S
gpa1-2
gpa1-2 bri1-5 bri1-5
B
*
*
1.0
0.8
0.6
0.4
productiion (cell/h) 0.2
0
WS
gpa1-2
gpa1-2 bri1-5 bri1-5
C
*
*
140 120 100
60 80
40 20 0 Col
gpa1-4 gpa1-4 det2-1 det2-1
D
*
*
1.2
0.9
0.6
0.3 Rate of root cell production (cell/h)
0 Col
gpa1-4 gpa1-4 det2-1 det2-1
Fig 7 Rate of cell production in the primary roots of gpa1-2
bri1-5 and gpa1-4 det2-1 double mutants (A) Average length of
mature root cortex cells in the gpa1-2 bri1-5 double mutants.
(B) Rate of cell production in the primary root of gpa1-2 bri1-5
double mutants (C) Average length of mature root cortex cells in
the gpa1-4 det2-1 double mutants (D) Rate of cell production in
the primary root of gpa1-4 det2-1 double mutants The rate of cell
production was calculated as the rate of primary root elongation
divided by the average length of mature cortical cells in the root
hair zone Shown in (C) and (D) are the averages SE of 10
seed-lings for each genotype
significant difference from the wild type (WS or Col), P50.05 #significant difference from the bri1-5
or det2-1 mutant, P50.05.
Trang 9Interestingly, the rice Ga (RGA1) has been best known
for its role in response to gibberellins (Fujisawa et al 1999,
Ueguchi-Tanaka et al 2000) The dwarfism of the rice
d1 mutant was caused by a loss-of-function mutation in
the Ga gene (Fujisawa et al 1999) The d1 mutant also
displayed reduced sensitivity to BR in the BR-mediated
inhibition of root elongation, promotion of lamina
inclina-tion and promoinclina-tion of coleoptile elongainclina-tion (Wang et al
2006) More strikingly, d1 mutants are dwarf whereas gpa1
mutants have wild-type height It remains unclear why
the loss of function of Ga in one species (e.g rice) results
in dwarfism whereas it does not do so in another species
(e.g Arabidopsis)
Role of BR in regulating cell elongation and cell division
BR regulates both cell division and cell elongation,
but is generally believed to have a prominent role in
regu-lating cell elongation over cell division Both BR signaling
mutants, such as bri1-5 (Li and Chory 1997, Noguchi et al
1999a) and bin2 (Li and Nam 2002), and BR biosynthesis
mutants, such as det2-1 (Li et al 1996, Li et al 1997) and
dwf4-102(Azpiroz et al 1998, Choe et al 1998, Nakamoto
et al 2006), displayed dwarfism Although a role for BR
in cell division has been established in tissue culture and
cell culture systems (Sala and Sala 1985, Nakajima et al
1996, Oh and Clouse 1998, Hu et al 2000, Miyazawa et al
2003), the genetic evidence had been lacking By using the
Arabidopsis hypocotyl as a model system in which the
number of epidermal cells is established during
embryogen-esis, we demonstrated that BR has a role in regulating both
cell elongation and cell division We found that the length
of hypocotyl epidermal cells was dramatically reduced
in the BR receptor mutants, bri1-5 and bri1-4, and the BR
biosynthesis mutants, det2-1 and dwf4-102 (Fig 2,
Supple-mentary Fig S2), supporting the view that BR is a major
regulator of cell elongation However, we found that the
number of hypocotyl epidermal cells was also significantly
reduced in these mutants (Fig 2, Supplementary Fig S2)
These results provided direct evidence that BR regulates
both cell elongation and cell division We extended our
analysis to the roots Consistent with previous reports
of BR mutants (Mouchel et al 2006), bri1-5 and det2-1
mutants had shorter primary roots than the wild type
(Fig 6) By analyzing the length of root cortex cells and the
rate of root cell production in bri1-5 and det2-1 mutants
(Fig 7), we showed that BR also regulates both cell
elongation and cell division in the roots
Additive effect of GPA1 and BR in regulating cell division
in hypocotyls
We used the Arabidopsis hypocotyl as a model system
to analyze the cellular basis of enhanced dwarfism
pheno-types of gpa1-2 bri1-5 and gpa1-4 det2-1 double mutants,
originally observed in adult plants (Fig 1) Both gpa1 and
BR mutants, bri1-5 and det2-1, share the short hypocotyl phenotype (Fig 2) However, the cellular basis of such shortness is different The short hypocotyl phenotype
of gpa1 mutants was due to a reduction in cell division, whereas the short hypocotyl phenotype of bri1-5 and det2-1 mutants was due to a reduction in both cell elongation and cell division (Fig 2) Further, we found that gpa1 mutations could specifically enhance the cell division defects of bri1-5 and det2-1 mutants (Figs 4, 5) Because the cell division defects between gpa1-2 and bri1-5 and between gpa1-4 and det2-1were additive, these results suggested that GPA1 and
BR probably function in parallel pathways to regulate cell division in hypocotyls Similarly, GPA1 and BR may also function in independent pathways to regulate lateral root formation, because an additive effect was observed between gpa1-2and bri1-5 and between gpa1-4 and det2-1 (Fig 6)
Synergistic effect of GPA1 and BR in regulating cell division in roots
As discussed above, GPA1 and BR probably function independently to regulate cell division in hypocotyls Inter-estingly, we found that GPA1 and BR may act in concert
to regulate cell division in the primary roots (Figs 6, 7) gpa1mutants have primary roots of normal length, whereas bri1-5and det2-1 mutants have short primary roots (Fig 6)
We found that the short root phenotype of bri1-5 and det2-1 mutants was due to a defect in both cell elongation and cell division (Fig 7) Interestingly, we found that although gpa1mutants do not display defects in cell elongation or cell division in the primary roots, gpa1 mutations can specifi-cally enhance the cell division defects in the primary roots
of bri1-5 and det2-1 mutants (Fig 7) Such s synergistic effect suggested that GPA1 may interact genetically with the BR-mediated pathways to regulate cell division in the primary roots Similarly, GPA1- and BR-mediated path-ways may work together to regulate plant height because gpa1 mutants have wild-type height but can significantly enhance the dwarfism of bri1-5 and det2-1 mutants (Fig 1) These findings raise the possibility that the G-protein-mediated pathway may cross-talk with the receptor kinase-mediated pathway to regulate a specific cellular process (e.g cell division) in plants
How can GPA1- and BR-mediated pathways function independently in one tissue (e.g hypocotyl) whereas they work together in other tissues (e.g primary root) to regulate cell division? One crucial concept of heterotri-meric G-protein action in plants is that G-proteins play regulatory roles in diverse developmental processes and function in a cell type- or developmental stage-specific manner (Perfus-Barbeoch et al 2004) For example, gpa1 mutants are hypersensitive to ABA during seed germina-tion and early seedling development (Ullah et al 2002,
Trang 10Pandey et al 2006) whereas they are insensitive to ABA in
guard cells by abolishing the inhibition of the inward Kþ
channels by ABA (Wang et al 2001) Another example
is that the sole Arabidopsis heterotrimeric G-protein b
subunit, AGB1, has been shown to be a positive regulator
of axial cell division in the hypocotyls, whereas it functions
as a negative regulator of cell division in the roots (Ullah
et al 2003, Chen et al 2006) It is likely that G-proteins may
utilize a signaling cascade to regulate cell division in the
primary roots that is distinct from that in the hypocotyls
The exact molecular mechanisms of GPA1 and BR
in regulating cell division are presently unknown gpa1
mutants displayed altered sensitivities to multiple
hor-mones, such as auxin (Ullah et al 2003), ABA (Wang
et al 2001, Ullah et al 2002, Pandey et al 2006) and
gibberellins (Ullah et al 2002, Chen et al 2004), in addition
to BR (Ullah et al 2002) On the other hand, BR also
cross-talks with other hormones, particularly auxin (Nemhauser
et al 2004, Hardtke 2007) For example, BR can activate
the expression of some auxin-induced genes, as revealed by
microarray analyses (Goda et al 2004, Nemhauser et al
2004), and act synergistically with auxin to regulate
hypo-cotyl growth (Nakamura et al 2003, Nemhauser et al 2004)
and lateral root formation (Bao et al 2004) in Arabidopsis
Recently, BREVIS RADIX (BRX) has been shown to act
at the nexus of a feedback loop that maintains threshold
BR levels to permit optimal auxin action in roots (Mouchel
et al 2006) These findings raise the possibility that
GPA1-and BR-mediated pathways may interact through
interac-tion with other plant hormones However, because a
canon-ical GPCR together with its ligand has yet to be identified
in plants and the G-protein-mediated signal transduction
pathway in plants is far from complete, the elucidation of
the interaction between G-proteins and the BR signaling
pathway at the mechanistic level is a challenging yet exciting
topic for future research
Materials and Methods
Plant materials
gpa1-2 (Ullah et al 2001), bri1-5 (Li and Chory 1997,
Noguchi et al 1999a) and bri1-4 (Li and Chory 1997, Noguchi
et al 1999a) mutants are in the WS ecotypic background gpa1-4
(Jones et al 2003), det2-1 (Li and Chory 1996, Li et al 1997) and
dwf4-102 (Azpiroz et al 1998, Choe et al 1998, Nakamoto et al.
2006) are in the Col ecotypic background gpa1-2 bri1-5 double
mutants were generated by crossing gpa1-2 (pollen donor) into
bri1-5 and identified from the F2 progeny by PCR genotyping
and phenotyping Similarly, gpa1-4 det2-1 double mutants were
generated by crossing gpa1-4 into det2-1 and identified from the
F2 progeny.
Plant growth conditions and phenotypic analyses
For Petri dish-based phenotypic analyses, wild-type and
mutant seeds were sterilized, sown on MS/G Petri dishes
contain-ing 1/2 strength Murashige and Skoog (MS) basal medium with
vitamins (plantmedia, Dublin, OH, USA), 1% sucrose and 0.5% phytoagar (plantmedia), adjusted to pH 5.7 with 1 N KOH, and treated at 48C in the dark for 3 d, then moved to a growth chamber at 238C For the phenotypic analysis of 3-day-old, dark-grown seedlings, the Petri dishes were wrapped in double layers of aluminum foil and placed in the dark For light-grown seedlings, Petri dishes were placed under low-light conditions (30 mmol m–2s–1) with a 14/10 h photoperiod for easy examination
of hypocotyl epidermal and cortex cells The hypocotyl lengths were measured from at least 20 seedlings for each genotype For soil-based phenotypic analysis, wild-type and mutant plants were grown under identical conditions with a 14/10 h photo-period at 140 mmol m–2s–1 Plant heights were measured from at least 10 plants for each genotype.
Measurement of hypocotyl cells For the examination of hypocotyl epidermal cells and outer cortex cells, 3-day-old, dark-grown, or 4-day-old, low-light-grown seedlings of the wild type and mutants were fixed and cleared in chloral hydrate solution (chloral hydrate : glycerol : water ¼ 8 : 1 : 3) A single file of epidermal or outer cortex cells from the base (root–hypocotyl junction) to the top of the hypo-cotyl (lateral to hypo-cotyledons) of each seedling was examined under
a Leica DM-6000B upright microscope with phase and DIC equipped with a Leica FW4000 digital image acquisition and processing system [Leica Microsystems (Canada) Inc., Richmond Hill, Ontario, Canada].
Root assays Seedlings grown on MS/G plates were placed under a 14/10 h photoperiod with approximately 120 mmol m –2 s –1 at 238C with a vertical orientation for monitoring root growth The length
of primary roots and the number of lateral roots were determined
on 10-day-old seedlings, and at least 10 seedlings were used for each genotype.
The procedure for the measurement of the rate of root cell production has been described previously (Chen et al 2003, Chen et al 2006) The root growth was monitored daily by marking the positions of the root tips Rates of primary root growth were calculated over 3 d periods from day 3 to day 6 Seedlings were sampled at day 7, fixed, and cleared in chloral hydrate solution The lengths of at least 20 cortex cells in the root hair zone of each root were measured under a Leica DM-6000B upright microscope with phase and DIC equipped with a Leica FW4000 digital image acquisition and processing system Cell production was calculated as the rate of root growth divided by the average length of mature cortex cells.
Supplementary material Supplementary material mentioned in the article is available
to online subscribers at the journal website www.pcp.oxford journals.org.
Funding
Natural Sciences and Engineering Research Council
of Canada and Canada Foundation for Innovation (to J.-G.C.)