Directory UMM :Data Elmu:jurnal:P:PlantScience:Plant Science_BioMedNet:081-100:

trends in plant science
Reviews

Gibberellin and abscisic acid
signalling in aleurone
Alison Lovegrove and Richard Hooley
The plant hormones gibberellin and abscisic acid regulate gene expression, secretion and cell
death in aleurone. The emerging picture is of gibberellin perception at the plasma membrane
whereas abscisic acid acts at both the plasma membrane and in the cytoplasm – although gibberellin and abscisic acid receptors have yet to be identified. A range of downstream-signalling
components and events has been implicated in gibberellin and abscisic acid signalling in aleurone. These include the Ga subunit of a heterotrimeric G protein, a transient elevation in cGMP,
Ca21-dependent and Ca21-independent events in the cytoplasm, reversible protein phosphorylation, and several promoter cis-elements and transcription factors, including GAMYB. In
parallel, molecular genetic studies on mutants of Arabidopsis that show defects in responses to
these hormones have identified components of gibberellin and abscisic acid signalling. These
two approaches are yielding results that raise the possibility that specific gibberellin and
abscisic acid signalling components perform similar functions in aleurone and other tissues.

I

n the time between germinating and establishing phototrophic
growth, seedlings rely on nutrients stored in the seed. In the
Gramineae, aleurone cells help mobilize these reserves by synthesizing and secreting a variety of hydrolytic enzymes that participate in the breakdown of endosperm starch, protein, lipid and

cell walls. In addition, aleurone cells contain stored phytin, the
seed’s reserves of myo-inositol, phosphorous and mineral cations,
such as K1 and Mg21. These are also released into the endosperm

ABA

after germination. Aleurone cells eventually die a few days after
germination. Reserve mobilization by aleurone cells is strongly
correlated with gibberellin (GA) produced de novo by the embryo,
and with a decline in abscisic acid (ABA) content of the
endosperm1. The relative ease with which aleurone tissue can be
isolated from other cell types, combined with the specificity of
the responses of these cells to GA and ABA, has made aleurone
a model system for studying the cell and molecular biology of
these hormones.
GA and ABA influence a range of other events during growth
and development2,3, and studies of Arabidopsis mutants defective
in GA and ABA signalling4 have led to the identification of some
components of these signalling pathways. These recent developments in GA and ABA signalling in aleurone are reviewed here,
together with insight gained from molecular genetic studies with

Arabidopsis (see also Refs 5–8).
How do aleurone cells sense GA and ABA?

GA

α-Amylase
gene expression

ABA

Em:GUS
expression

Trends in Plant Science

Fig. 1. Simple scheme depicting the perception of gibberellin (GA)
and abscisic acid (ABA) in aleurone cells. GA is perceived by a
receptor (red) at the plasma membrane (green line) and induces
a-amylase gene expression via a signal transduction pathway
(arrows). ABA inhibition of this response is mediated by an ABA

receptor at the plasma membrane (dark blue) and a signal transduction pathway (arrows and lines), which represses GA signalling. ABA stimulation of Em:GUS expression is via an
intracellular ABA receptor (pale blue).

102

March 2000, Vol. 5, No. 3

Evidence from two different experimental approaches suggests
that GAs are perceived at the aleurone plasma membrane. First,
GA4 covalently coupled to Sepharose beads is membrane-impermeant, but nevertheless stimulates high-level a-amylase gene
expression and protein secretion in aleurone protoplasts9. Second,
microinjection of GA into aleurone protoplasts does not stimulate
a-amylase gene expression, and it is only when GA is present in
the protoplast incubation medium that they respond10. Other studies11,12 also support a plasma membrane site of GA action (Fig. 1).
Microinjection of ABA into GA-treated barley aleurone protoplasts does not appreciably inhibit GA-induced expression of an
Amy:GUS reporter or a-amylase production, whereas ABA in the
incubation medium does10. Thus, ABA inhibition of GA-action is
mediated by perception at the plasma membrane. However, induction of the ABA-responsive Em:GUS in barley aleurone protoplasts can be achieved either by microinjecting ABA or by adding
it to the medium6,13, suggesting that in this response ABA is perceived inside the protoplast (Fig. 1). Stomatal guard cells also
appear to have internal and external sites of ABA perception14.

The current challenge is to identify GA and ABA receptors
unequivocally. There has been some progress in the identification
of GA binding proteins in aleurone, although no ABA-binding
proteins have been found in this tissue.

1360 - 1385/00/$ – see front matter © 2000 Elsevier Science Ltd. All rights reserved. PII: S1360-1385(00)01571-5

trends in plant science
Reviews
GA-binding proteins at the plasma membrane

The development of a method for isolating large quantities of
highly purified aleurone plasma membrane, combined with the use
of a high specific-activity 125I-labelled GA-photoaffinity probe, led
to the detection of two plasma membrane polypeptides of ~18 kDa
and 68 kDa (Ref. 15). Photoaffinity labelling of these was competed by biologically active GA4, GA1 and unlabelled photoaffinity probe, whereas biologically inactive GA34 did not compete.
Polypeptides of 18 kDa and 68 kDa were also GA-photoaffinity
labelled in plasma membrane from vegetative tissue of pea, sweet
pea (Lathyrus odoratus) and Arabidopsis. The GA-binding proteins could not be detected on intracellular membranes and 2D-gel
analysis demonstrated their low abundance15. Preliminary N-terminal amino acid sequence from the 18-kDa polypeptide purified

from Arabidopsis has yet to identify a potential function. Nevertheless, the possibility that the polypeptides are involved in GA
action is suggested by the observation that they are less intensely
labelled in the sweet pea semi-dwarf GA-sensitivity mutant lb
compared with wild-type Lb (Ref. 15). On non-reducing SDS
gels, the 18 kDa and 68 kDa polypeptides are both resolved, indicating that they are not disulphide-linked as part of the same protein (A. Lovegrove, unpublished). The fact that both polypeptides
are labelled by the GA-photoaffinity probe suggests that the GA
receptor might be a complex of at least two proteins (Box 1).
Heterotrimeric G proteins transduce GA signals

Evidence is emerging that functional counterparts of several components of the G protein-signalling pathway exist in plants16.
Recent results from two independent lines of enquiry have
revealed a possible role of heterotrimeric G proteins in GA signal
transduction in wild oat and rice aleurone. The mastoparan analogue Mas7 stimulates GDP–GTP exchange by heterotrimeric G
proteins, and is thought to mimic activated G protein-coupled
receptors. When wild oat aleurone protoplasts are incubated with
Mas7 they produce and secrete a-amylase in a dose-dependent
manner and with an identical time course to GA (Ref. 17). As with
GA, the effect is largely overcome by ABA. Mas7 induces a-amylase mRNA and expression of a-Amy2/54:GUS. This raises the
possibility that Mas7 is activating a heterotrimeric G protein in the
GA-signalling pathway. Further evidence supporting this theory

has come from studying the effects of hydrolysis-resistant guanine
nucleotides on GA-induction of a-Amy2/54:GUS expression17.
The hydrolysis-resistant guanine nucleotide analogues GTP-g-S
and GDP-b-S bind to Ga subunits and hold them in either the activated (GTP-g-S-bound) or inactivated (GDP-b-S-bound) form.
GDP-b-S introduced into aleurone protoplasts during transfection
with reporter gene constructs prevents GA induction of aAmy2/54:GUS expression, whereas GTP-g-S stimulates expression slightly. Taken together the Mas 7 and guanine nucleotide
data strongly suggest that a heterotrimeric G protein(s) is involved
at an early stage of the GA-signalling pathway in wild-oat aleurone. It predicts that aleurone cells should contain G protein subunits, which was confirmed by PCR and northern analysis for a
partial Ga subunit cDNA and two related Gb cDNAs (Ref. 17).
ABA inhibition of GA-induction of a-amylase can partly be overcome by treating wild-oat aleurone protoplasts with cholera or pertussis toxin (H.D. Jones and R. Hooley, unpublished), perhaps
indicating G protein involvement in ABA signalling in aleurone.
Embryo-less half seed of the Dwarf 1 mutant of rice produce
little a-amylase when treated with GA, and Dwarf 1 seedlings
elongate only slightly in response to GA (Ref. 18). These observations are consistent with Dwarf 1 being a GA-sensitivity
mutant18, although we await measurement of the endogenous GAs
in Dwarf 1 to see if they are elevated, as would be expected for

Box 1. Gibberellin and abscisic acid perception:
key points
•

•
•
•
•

Gibberellin (GA) is perceived at the aleurone plasma membrane.
Plasma membrane GA-binding proteins (18 kDa and 68 kDa)
occur in aleurone and other tissues.
Abscisic acid (ABA) inhibition of GA action is mediated by
perception at the aleurone plasma membrane.
ABA induction of Em:GUS expression is mediated by perception
inside the cell.
No ABA-binding proteins have been identified in aleurone.

Box 2. Heterotrimeric G proteins: key points
•
•

G protein agonists and antagonists can mimic and block
gibberellin (GA) action.

Mutations in the rice G protein a-subunit reduce GA-insensitivity
in aleurone and other tissues.

such a mutant. Mutations giving rise to Dwarf 1 are now known to
be in the Ga subunit GPA1 (Refs 19,20). Because these clearly
affect GA sensitivity of rice aleurone, these observations add
further support to the theory that a G protein is involved in GA
signalling in aleurone.
Precedents from well characterized G protein-signalling pathways21 make it tempting to speculate that GA, perceived by a
receptor at the plasma membrane, signals through a heterotrimeric
G protein. Although this hypothesis requires testing experimentally, it is clear that the theory that plant hormone signals might be
transduced by G protein-signalling pathways is further supported
by the discovery of GCR1 – the first plant G protein-coupled
receptor homologue – and the evidence that it might be involved
in cytokinin signal transduction22 (Box 2).
Ca21/CaM-dependent and -independent events

The earliest event following GA treatment of aleurone is an
increase in cytoplasmic calcium concentration ([Ca21]i). In barley
aleurone protoplasts, it occurs after 4–6 h (Ref. 23). In intact wheat

aleurone cells, a faster response (2–5 min) has been observed24,
which does not occur in response to the inactive GA8. Elevation of
[Ca21]i by GA is quicker in some cells than others13,24, perhaps
because of heterogeneity in the sensitivity of individual aleurone
cells and protoplasts25,26. Provided GA is present continuously, it
causes sustained elevations in [Ca21]i that, when averaged over the
whole cell, amount to increases of 100–500 nM above a resting
level of 100–250 nM. Localized [Ca21]i can be much higher. Combined temporal and spatial analysis of [Ca21]i (S. Gilroy, pers.
commun.; www.bio.psu.edu/faculty/gilroy/algal.html) coupled with
a known dependency of these changes on extracellular Ca21,
has led to the suggestion that GA alters Ca21 flux at the plasma
membrane13,23,24. The mechanism is unknown, although one hypothesis is that the GA receptor might, directly or indirectly, regulate a plasma membrane Ca21 channel, possibly through a
heterotrimeric G protein. Release of Ca21 from intracellular stores
is another possibility27. However, it is unlikely to be mediated by
IP3 (Ref. 28), even though there is some evidence that GA causes a
transient elevation in IP3 in rice aleurone27. In barley, ABA prevents the GA-stimulated increases in [Ca21]i and, in the absence of
GA, high concentrations of ABA have been proposed to lower
[Ca21]i by activating a plasma membrane Ca21-ATPase (Ref. 29).
March 2000, Vol. 5, No. 3

103

trends in plant science
Reviews
release. Ca21/CaM are thought to affect
phosphorylation at two sites on the SVtype channel, regulating its activity in a
complex way. CaM-affinity screening of
ABA
Em gene expression
an aleurone cDNA library has been used in
an attempt to identify other targets of
Ca21/CaM (Ref. 34). This led to the isoABA
lation of a plasma membrane-located
CaM-binding transporter that might potentially be regulated by cyclic nucleotide
monophosphates such as cGMP.
CaM and α-amylase
GA
To identify Ca21- and CaM-dependent and
gene expression
-independent signalling events in aleurone,

the [Ca21]i and CaM levels were modified
Ca2+
by
microinjecting CaCl2, caged Ca21, Ca21
CaM
α-Amylase
CaM
chelators, CaM and CaM antagonists into
[Ca2+]i
barley aleurone protoplasts13. Blocking the
2+
GA-induced increase in [Ca21]i using the
Ca /CaM
2+
Ca21 buffer diazo-2 inhibits GA-induced
ER Ca ATPase
a-amylase secretion but does not appreciably affect Amy:GUS expression. CaM
antagonists also inhibit a-amylase secretion. Attempts to mimic GA-induced [Ca21]i
G protein
and CaM changes by means of microinjec2+
tion do not stimulate Amy:GUS expression.
Ca /CaM
In addition, ABA induction of Em:GUS is
independent of changes in [Ca21]i and CaM
2+
Ca
(Ref. 13). These data therefore exclude
Slow
changes in Ca21/CaM from GA signalling,
vacuolar
which leads to the induction of a-amylase
channel
gene expression and ABA induction of
Em:GUS in barley aleurone13 (Fig. 2).
However, a study with rice aleurone27 conTrends in Plant Science
cludes that Ca21 and CaM are important
21
intermediaries
in the GA induction of gene
Fig. 2. Principal components in Ca and CaM signalling in aleurone. Gibberellin (GA),
expression. Furthermore, the fact that [Ca21]i
perceived by a receptor (red) at the plasma membrane (green line), elevates [Ca21]i and CaM
is elevated within minutes after GA
and this is prevented by abscisic acid (ABA) acting through a receptor (dark blue) at the
plasma membrane. Ca21 and CaM are necessary for hydrolase secretion. CaM stimulates an
treatment of wheat aleurone cells, whereas
endoplasmic reticulum (ER) Ca21ATPase, thus elevating ER [Ca21]. Ca21/CaM regulate the
hydrolase synthesis and secretion does not
activity of a slow vacuolar channel that probably contributes to [Ca21]i maintenance. Ca21
take place until several hours later, argues
stimulates secretory vesicle fusion, probably via a monomeric G protein . ABA stimulation of
that changes in [Ca21]i might be involved
Em gene expression [via an intracellular ABA receptor (pale blue)], and GA stimulation
in events that precede hydrolase secretion.
of a-amylase and CaM gene expression, are Ca21 and CaM-independent.
Further complexity has been uncovered
by the discovery that if the decrease in
[Ca21]i caused by ABA is blocked in GA21
These alterations in [Ca ]i indicate a potential signalling role, and treated barley aleurone protoplasts, ABA is prevented from
the fact that calmodulin (CaM) protein levels are stimulated two- inhibiting GA-induced Amy:GUS expression and a-amylase
to fourfold by GA, and reduced by ABA (Ref. 30), suggest that a secretion13. This presents a paradox; artificially reducing [Ca21]i
Ca21/CaM system might operate. This, and earlier data5,6, strongly in GA-treated protoplasts does not affect Amy:GUS expression,
implicate GA-induced changes in Ca21 and CaM in the secretion although the ABA-driven reduction in [Ca21]i is essential for
of hydrolases that are induced by GA (Fig. 2). Likely targets of ABA-inhibition of GA-induced Amy:GUS expression (Fig. 3).
Ca21/CaM are:
The inhibitory effect of ABA on GA-induced Amy:GUS expres• An ER Ca21-ATPase that can be activated by CaM, and prob- sion and a-amylase secretion can also be overridden by microinably supports the demand for endomembrane Ca21 driven by jecting CaM, although it is not clear what effect the CaM has on
secretion of the a-amylase metalloprotein31.
[Ca21]i. These experiments highlight an additional level of poten21
32
• A tonoplast Ca transporter .
tial complexity that might be explained by temporal or spatial
• Ca21-stimulated secretory vesicle fusion that is modulated by a dynamics of the [Ca21]i signature. A further glimpse of such comG protein33, the identity of which is unknown, although it is plexity might be the observation that overexpression of a putative
probably monomeric.
ER Ca21-ATPase in rice aleurone leads to GA-independent
• A storage vacuole membrane slow-vacuolar (SV)-type cation Osamy-c:luciferase expression, possibly by disrupting spatial or
channel5 (Fig. 2).
temporal aspects of [Ca21]i signals27.
The slow-vacuolar-type cation channel was originally thought to
Taken together these data suggest that for both GA and
be involved in the transport of K1 out of vacuoles, but it is now ABA there are specific Ca21/CaM-dependent and -independent
considered more likely to be involved in Ca21-induced Ca21 signalling events. In the case of ABA, these correlate with responses
104

March 2000, Vol. 5, No. 3

trends in plant science
Reviews

(a)

(b)

(c)

ABA

α -Amy:GUS

GA

[Ca2+]i

diazo-2

α -Amy:GUS

GA

Ca2+ or CaM

[Ca2+]i or CaM

α -Amy:GUS

Ca2+/CaM

[Ca2+]i /CaM

Trends in Plant Science

Fig. 3. Calcium signalling might be more complex than we presently understand. (a) Gibberellin (GA), perceived by a receptor (red) at the
plasma membrane (green line), stimulates an elevation in [Ca21]i and expression of a-Amy:GUS. The Ca21 chelator diazo-2 will prevent the
elevation in [Ca21]i but does not inhibit a-Amy:GUS expression. (b) Abscisic acid (ABA), acting through a receptor (dark blue) at the plasma
membrane, prevents GA stimulation of both [Ca21]i and a-Amy:GUS expression. However, microinjection of Ca21 or CaM restores a-Amy:GUS
expression. (c) Elevating [Ca21]i and CaM in the absence of GA and ABA is not sufficient to stimulate a-Amy:GUS expression.

associated with extracellular and intracellular perception sites, and
therefore further support the concept of dual pathways. For GA, the
simplest interpretation is that signalling downstream from a plasma
membrane receptor is bifurcated, and that ABA antagonism of GAinduced a-amylase gene expression involves more than just a
simple reversal of the GA-signalling events. Nevertheless, the
emerging complexity in Ca21/CaM signalling in aleurone suggests
that additional control factors impart specificity and integration,
and the identification of these is a current challenge (Box 3).
cGMP – a GA second messenger

Radioimmunoassay has been used to measure the cGMP in barley
aleurone layers35. Controls and layers treated with 5 mM ABA contained the same amounts of cGMP over a 6 h period. GA-treated
layers showed an approximately threefold rise in cGMP 2 h after
GA addition, which returned to control levels within a further 2 h.
It is not possible to accurately estimate the timing of this rise in
cGMP relative to the elevation of [Ca21]i because the elevation of
[Ca21]i was determined in barley aleurone protoplasts. However,
based on measurements of [Ca21]i in wheat aleurone layers, the
alterations in cGMP probably occur after the rise in [Ca21]i.
Nevertheless, they certainly precede the increase in transcription
of a-amylase genes. An inhibitor of guanylyl cyclase, LY 83583
(LY), reduces the increase in cGMP and inhibits the induction of
a-amylase and GAMYB mRNAs, as well as a-amylase protein
secretion. LY also prevents DNA degradation and cell death36.
Diutyryl-cGMP can largely, and specifically, overcome the
inhibitory effects of LY, although on its own it cannot elicit a GA
response. Therefore it is likely that the transient rise in cGMP is
an important component in the signalling necessary for GA-regulated gene expression, a-amylase secretion and programmed cell
death35,36. These data provide further support for a prominent signalling role of cGMP in plant cells, and raise the questions of how
guanylyl cyclase and cGMP phosphodiesterase are regulated in
aleurone, and the identity of downstream targets of cGMP.

cascades in GA and ABA action. ABA treatment of wheat and barley aleurone induces the expression of a serine/threonine protein
kinase, PKABA1 (Ref. 37). A possible role for this kinase in ABA
inhibition of GA-regulated gene expression is suggested from
experiments in which constitutive overexpression of PKABA1,
strongly inhibits GA induction of a-Amy1, a-Amy2 and cysteine
proteinase promoter:GUS constructs (Fig. 4). Overexpression of a
null mutant of PKABA1, or of a different CDPK, has no effect on
GA induction of the a-Amy2 promoter:GUS construct. Overexpressing the kinase also has a minor effect on ABA induction of
HVA1 gene expression37. This could indicate a broader signalling
role for PKABA1, or might be an artifact of overexpression.
In an ingenious assay, protein kinase substrate peptides have
been microinjected into barley aleurone protoplasts in the expectation that they would compete with the endogenous protein phosphorylation targets and block kinase cascades38. One peptide,
syntide-2, selectively inhibits GA-induction of a-Amy2:GUS, aamylase secretion and protoplast vacuolation. The peptide does not
prevent GA-stimulation of [Ca21]i, indicating that it is acting downstream of this early event in GA-signalling. Syntide-2, is a specific
substrate for mammalian CaM-kinase II. To date, there is no direct
evidence for a CaM-kinase II in plants therefore the identity of the
aleurone kinase, or kinases, blocked by syntide-2 is not clear.
Nevertheless, on the basis of a series of in vitro phosphorylation
and protein-labelling assays, a putative Ca21-activated CaM-like
domain protein kinase (CDPK) that interacts with syntide-2 has
been identified in barley aleurone cytosol38. At this stage it is not
known if syntide-2 blocks one or more events in the GA-signalling

Box 3. Calcium and calmodulin: key points
•
•
•

Protein kinase cascades

Reversible protein phosphorylation is a universal signal-transducing mechanism and, not unexpectedly, there is both molecular and
pharmacological evidence for the involvement of protein kinase

•

Gibberellin (GA) stimulates [Ca21]i and CaM, abscisic acid
(ABA) prevents this.
Ca21/CaM maintain secretion in GA-treated protoplasts.
Up-regulation of gene expression by GA and ABA is Ca21/CaMindependent.
Down-regulation of a-amylase gene expression by ABA is
Ca21/CaM-dependent.

March 2000, Vol. 5, No. 3

105

trends in plant science
Reviews
Pharmacological studies have provided
further evidence for kinases and phosphatases in GA and ABA signalling. Okadaic
α -Amy1/α -Amy2:GUS
acid (OA) inhibits GA-induced increases in
GA
[Ca21]i, a-amylase mRNA, a-amylase
secretion and aleurone cell death at conCysteine proteinase:GUS
centrations that inhibit the activity of
serine/threonine protein phosphatases (PP)
PP1 and PP2B (Ref. 39). The effect of OA
PKABA1
on the elevation of [Ca21]i caused by GA
suggests that a reversible phosphorylation
event is an essential part of this particular
ABA
Ca21 signature, although not of the [Ca21]i
PKABA1
elevation caused by hypoxia39. Although
?
the target of this is not currently known it
might prove useful to search for protein
ABA
phosphorylation events that occur within
minutes of GA treatment. Because OA also
Trends in Plant Science
inhibits GA-induction of a-amylase gene
expression and cell death – events indepenFig. 4. The abscisic acid (ABA)-inducible protein kinase PKABA1 negatively regulates
dent of the elevation of [Ca21]i – OA could
gibberellin (GA)-inducible gene expression. GA is perceived by a receptor (red) at the plasma
be acting in several ways (Fig. 5).
membrane (green line). GA-induction of a-Amy1, a-Amy2 or cysteine proteinase:GUS
In barley aleurone protoplasts, OA partly
constructs is strongly suppressed by constitutive overexpression of PKABA1. ABA induction
inhibits
ABA-induced PHVA1 gene exof PKABA1 gene expression might be mediated by ABA receptors at either the plasma
pression39, and tyrosine and serine/threomembrane (dark blue) or inside the cell (pale blue).
nine phosphatase inhibitors prevent ABAinduced RAB gene expression, leading to
tyrosine hyperphosphorylation of two
40 kDa polypeptides40. A current hypothesis is that ABA regulation of rab16 gene expression in barley aleurone involves rapid
OA Syntide-2
stimulation of a putative mitogen-activated protein kinase (MAP
kinase) via a tyrosine phosphatase41.
In summary, a range of molecular and pharmacological data
1
supports
the hypothesis that protein kinases and phosphatases are
α-Amylase
gene
GA
involved in GA and ABA signalling in aleurone. The identity of
expression and
3
OA 2
the enzymes, their targets and their precise relationships to GAPCD
stimulated events are poorly defined. For ABA signalling, the
OA
apparent role of reversible protein phosphorylation in aleurone is
supported by molecular genetic evidence that two OA-insensitive
serine/threonine phosphatases are involved in ABA signalling in
2+
[Ca ]i
Arabidopsis3. In the case of GA, investigations have highlighted
that the role of [Ca21]i in GA-regulation of gene expression is not
Syntide-2
fully understood (Box 4).
α-Amylase secretion

Trends in Plant Science

Fig. 5. Scheme depicting possible protein phosphorylation events
revealed by okadaic acid (OA) and syntide-2. Gibberellin (GA) is
perceived by a receptor (red) at the plasma membrane (green line).
Because OA inhibits the GA stimulation of [Ca21]i, a-amylase
gene expression and programmed cell death (PCD) it might be
inhibiting PP1 or PP2B at a point in GA signalling upstream of
these events (1), or at multiple sites downstream of this (2 and 3).
Syntide-2 identifies a protein phosphorylation event downstream
of the effect of GA on [Ca21]i that could be in the pathway to gene
expression or a-amylase secretion.

and response pathway (Fig. 5). Syntide-2 could potentially
interfere with CDPK regulation of the SV-type cation
channel5 although what effect this would have on signalling is
not known. Syntide-2 does not affect ABA action, suggesting
that the kinase acting on it does not crosstalk with ABA
signalling.
106

March 2000, Vol. 5, No. 3

Putative serine/threonine-O-linked N-acetylglucosamine
transferase is a negative regulator of GA signalling

Recessive mutations at the spindly (spy) locus of Arabidopsis confer
a largely GA-independent phenotype characterized by a reduced GA
requirement for germination, increased internode length and altered
flowering time. All spy alleles tested partially suppress the phenotype
of the GA-deficient mutant ga1-1. SPY is therefore thought to act as
a negative regulator of GA signalling in Arabidopsis42. The barley
homologue of SPY is expressed at a low level in aleurone cells.
When it is over expressed behind the CaMV35S promoter in a transient assay, SPY inhibits GA-induced expression of an a-amylase
promoter-reporter construct43. Intriguingly, overexpression of SPY
in aleurone stimulates the expression of an ABA-regulated dehydrin
promoter. These data suggest that SPY has a conserved function as a
negative regulator of GA responses in vegetative tissues and aleurone cells, and that it might be a positive regulator of ABA signalling
in aleurone. Nevertheless, overexpression of signalling proteins
might not reveal their true function in vivo, or accommodate subtleties of post-transcriptional regulation that might be essential to
their action. Further studies will no doubt address such issues.

trends in plant science
Reviews
SPY has significant amino acid sequence identity to a family of
serine and threonine-O-linked N-acetylglucosamine (O-GlcNAc)
transferases (OGTs)42 that extends over both the characteristic
tetratricopeptide repeat (TPR) and catalytic domains. Although
OGT activity has yet to be demonstrated for SPY, it is likely that
this signalling protein interacts with other proteins via the TPR
domain and that O-GlcNAc modification affects their signalling
potential42. However, there are numerous potential targets of
OGTs: two possible targets of SPY O-GlcNAc modification are
GAI and RGA.

Box 4. Protein kinase cascades: key points
•
•
•
•
•

GAI and RGA

To date there is no direct evidence for a role for GAI or RGA in
GA signalling in aleurone cells. GAI and RGA are members of a
small gene family in Arabidopsis, encoding nuclear-localized
putative transcription factors44. GAI and RGA are GA-derepressible repressors of GA-mediated growth responses. The N-terminal
region of these proteins contains a so-called DELLA region,
which is thought to play a critical role in GA-derepression. It is
now known that the maize dwarf-8 (d8) gene and two Reduced
height loci of wheat, Rht-B1b (formerly Rht1) and Rht-D1b
(formerly Rht2), are orthologs of GAI, which contain nucleotide
substitutions resulting in the mutant genes encoding N-terminally
truncated proteins45. The related Reduced height locus Rht-B1c of
wheat is known to affect GA sensitivity in aleurone46, raising
the possibility that this class of proteins might have a role in GA
signalling in aleurone (Box 5).

Protein kinase peptide substrate syntide-2 blocks some gibberellin
(GA) responses.
Ca21-activated CaM-like domain protein kinase might be
involved in GA-induction of a-amylase production.
GA stimulation of [Ca21]i is inhibited by the protein phosphatase
inhibitor okadaic acid.
Abscisic acid (ABA) suppression of GA-induced gene expression
might involve protein kinase PKABA1.
ABA-induced gene expression might involve a tyrosine-kinase and
MAP-kinase.

Box 5. Arabidopsis gibberellin signalling genes:
key points
•
•
•

GA signalling genes identified in Arabidopsis might have
functional counterparts in aleurone.
SPY is a negative regulator of GA-induced a-amylase gene
expression.
SPY might be a positive regulator of ABA-induced dehydrin
gene expression.

led to the isolation of a barley aleurone cDNA, encoding a MYB
transcription factor that binds specifically to the TAACAA/GA
element and is able to transactivate an a-Amy1 gene promoter in
a transient assay51. A rice GAMYB appears to have a similar
function52 (Fig. 7).
GAMYB expression is stimulated by GA in advance of
a-amylase genes and, unlike a-amylase, is not dependent on de
novo protein synthesis. These observations strongly suggest that
GA signalling stimulates the expression of GAMYB, which in turn

Regulation of gene transcription

The expression of a variety of aleurone hydrolase genes is induced
or stimulated by GA, and this is overcome by ABA. Investigations
have concentrated on the transcriptional regulation by GA and ABA
of barley, wheat, rice and wild oat a-amylase gene promoters.
Functional analysis of a-Amy1 and a-Amy2 gene promoters by
transient expression and gel-retardation assays in combination
with DNAse-1 footprinting, have led to the identification of
several cis-elements and sites at which
nuclear proteins bind. In some cases, the
trans-acting factors that bind to specific
sequences have been cloned and characterized. It appears that the regulation of
a-Amy1 and a-Amy2 gene promoters
Box 2
α -Amy1
involves the interplay of several classes
of transcription factor (Fig. 6). As yet,
relatively little is known about comparable
elements in other GA-regulated genes.

HRT GAMYB
TATCCAC

GARE
TAACAAA
GARC

GA response element GAMYB and HRT

The GA response element (GARE) has
been identified as a 21-nucleotide region of
the barley Amy 1/6-4 promoter that, when
present as six copies in a chimeric promoter driving a reporter gene, confers GAresponsive expression that could be
overcome by ABA (Ref. 47). This region
of the promoter in both a-Amy1 and
a-Amy2 genes contains the sequence element TAACAA/GA, which has been
shown in functional studies to be an important component of the GARE (Refs 48,49).
The GARE appears to bind GA-inducible
nuclear proteins50. A potential similarity
between the TAACAA/GA element and
c-MYB and v-MYB recognition sequences

α -Amy2

ABF1/2

GAMYB

Box 2

GARE

TATCCAC

TAACAGA
GARC

Trends in Plant Science

Fig. 6. Principal cis-elements in a-Amy1 and a-Amy2 promoters and factors that bind to
them. GAMYB binds to the sequence TAACAA/GA in both a-Amy1 and a-Amy2 promoters. GAMYB functions within the GA-response complex (GARC) to transactivate
a-Amy1 and a-Amy2 promoters. The GA response element (GARE) of a-Amy1 promoters
binds the repressor HRT. Box 2 is a binding site for the WRKY proteins ABF1 and ABF2
in a-Amy2 promoters, although the regulation of ABF1 and ABF2 is not understood.

March 2000, Vol. 5, No. 3

107

trends in plant science
Reviews
is not clear whether VP1 interacts directly
with the a-amylase promoter or with other
transcription factors bound to it.
GA-response complex

GA

Although the GARE clearly plays a central
role in GA activation of a-amylase gene
expression, additional cis-elements are associated with the GARE and act as enhancers
within a GA-response complex (GARC) that
mediates GA- and ABA-responsive expression in both high and low pI a-amylase promoters47,49,57–59. In barley a-Amy1 genes, it is
thought that the GARE and TATCCAT box
comprise a GARC (Ref. 49). In a-Amy2
genes the GARC is thought to comprise the
GARE and Box 2 (Ref. 59).

GAMYB

Post-transcriptional
regulation

GAMYB
GARE

α-Amy1/α-Amy2

TAACAA/GA

TATCCAC box
GAMYB

Cathepsin B
GAMYB

β1-3,1-4 Glucanase
Trends in Plant Science

Fig. 7. Model of gibberellin (GA) regulation of GAMYB. GA is perceived by a receptor
(red) at the plasma membrane (green line). GA stimulates the expression of GAMYB, possibly
by acting on a rapidly turning-over repressor (black). GAMYB binds to the TAACAA/GA
element in the GA response element (GARE) of a-Amy1 and a-Amy2 promoters, and to
other sequences in cathepsin B-like protease and EII(1-3,1-4)-b-glucanase promoters. GA
signalling might also act on GAMYB post-transcriptionally.

induces the expression of a-Amy1 genes51. Because GAMYB
expression is super-induced by cycloheximide51 one theory is that
GA signalling might act on a rapidly turned-over repressor of
GAMYB expression although the identity of this is unknown.
Nevertheless, post-transcriptional regulation of GAMYB activity
by phosphorylation, or interaction with other proteins, is also probably based on models of how this class of transcription factors is
regulated in other signalling systems53. Is GAMYB involved in
stimulating the expression of other GA-regulated genes in aleurone? A recent study54 has demonstrated that GAMYB can transactivate the expression of GUS fused to promoter fragments of
three other GA-regulated genes; an a-Amy2 gene, an EII(1-3,1-4)b-glucanase gene and a cathepsin B-like protease gene. Analysis
of these promoters reveals that the a-Amy2 gene contains a
TAACAGA element, which is probably the GAMYB-binding site.
However, neither the EII(1-3,1-4)-b-glucanase nor the cathepsin
B-like protease promoters contain a TAACAA/GA element, and it
is likely that GAMYB transactivates these promoters through
closely related, although as yet not unambiguously defined,
elements54.
GAMYB is not the only transcription factor that binds to the 21
nucleotide GARE (Ref. 47). Southwestern screens of a barley
aleurone cDNA library using fragments of the Amy 1/6-4 promoter have identified a novel, nuclear-located, zinc-finger protein, HRT, which binds the 21 nucleotide GARE and can repress
GA-induced expression from a-Amy1 and a-Amy2 promoters55.
This HRT is therefore another control element acting on the
GARE, although its regulation and relationships to GAMYB are
not understood. The maize transcriptional activator VP1 can
repress GA-induction of an a-Amy1 gene promoter56, although it
108

March 2000, Vol. 5, No. 3

The TATCCAC box is present, and binds
as yet uncloned nuclear protein, in both
Amy1 and Amy2 genes51,59. Functional studies suggest that this box49, and sequence
elements immediately downstream from
it60 play an important role in driving GAregulated expression.
Box 2 and WRKY proteins

Functional analysis has demonstrated that
for a single copy of the GARE to function
in a-Amy2 promoters, Box 2 must be present as a coupling element57,58. Box 2 binds
two members, ABF1 and ABF2, of the
WRKY family of DNA-binding proteins61
although how these are regulated and whether or not they interact
with GAMYB or HRT is not clear (Box 6).

Box 6. Gene transcription: key points
•
•
•
•

The gibberellin response element (GARE) of a-amylase promoters
plays a central role in gibberellin (GA)- and abscisic acid
(ABA)-regulated gene expression.
GAMYB is induced by GA and binds to the TAACAA/GA element
within the GARE.
The repressor HRT also binds to elements within the GARE.
Other cis-elements and trans-acting factors are involved in
GA-regulated gene expression.

Perspectives

Aleurone cells and protoplasts are an excellent system for studying GA regulation of gene expression, protein secretion, cell death
and the antagonistic effect of ABA. The use of effector and
reporter gene constructs, in combination with a range of pharmacological agents, has helped to define elements of signalling pathways that encompass events at the plasma membrane, in the
cytoplasm and in the nucleus. Understanding the order and integration of signalling events in aleurone is a challenge that should
become accessible through the generation and use of mutants.
Arabidopsis and cereal genomics are poised to identify a range of
genes that might be involved – the function of these can be examined in aleurone in the context of the role and regulation of this
highly specialized secretory tissue.

trends in plant science
Reviews
Acknowledgements

IACR receives grant-aided support from the Biotechnology and
Biological Sciences Research Council of the UK. This work was
supported by the BBSRC Intracellular Signalling Programme and
EC Biotechnology FPIV. We apologize to colleagues whose
research we were not able to cite because of space restrictions, and
thank Drs John Lenton and Mike Holdsworth for useful comments
on the manuscript.
References
1 Appleford, N.E.J. and Lenton, J.R. (1997) Hormonal regulation of a-amylase
gene expression in germinating wheat (Triticum aestivum) grains. Physiol.
Plant. 100, 534–542
2 Hooley, R. (1994) Gibberellins: perception, transduction and responses.
Plant Mol. Biol. 26, 1529–1556
3 Giraudat, J. et al. (1994) Current advances in abscisic acid action and
signalling. Plant Mol. Biol. 26, 1557–1577
4 McCourt, P. (1999) Genetic analysis of hormone signaling. Annu. Rev. Plant
Physiol. Plant Mol. Biol. 50, 219–243
5 Bethke, P.C. et al. (1997) Hormonal signalling in cereal aleurone. J. Exp. Bot.
48, 1337–1356
6 Ritchie, S. and Gilroy, S. (1998) Tansley Review No 100. Gibberellins:
regulating genes and germination. New Phytol. 140, 363–383
7 Skadsen, R.W. (1998) Physiological and molecular genetic mechanisms regulating
hydrolytic enzyme gene expression in cereal grains. Physiol. Plant. 104, 486–502
8 Bethke, P.C. and Jones, R.L. (1998) Gibberellin signaling. Curr. Opin. Plant
Biol. 1, 440–446
9 Hooley, R. et al. (1991) Gibberellin perception at the plasma membrane of
Avena fatua aleurone protoplasts. Planta 183, 274–280
10 Gilroy, S. and Jones, R.L. (1994) Perception of gibberellin and abscisic acid at
the external face of the plasma membrane of barley (Hordeum vulgare L.)
aleurone protoplasts. Plant Physiol. 104, 1185–1192
11 Beale, M.H. et al. (1992) A new approach to gibberellin perception in
aleurone: novel hydrophylic, membrane-impermeant, GA-sulphonic acid
derivatives induce a-amylase formation. Physiol. Plant. 85, A136
12 Hooley, R. et al. (1992) Gibberellin perception and the Avena fatua aleurone:
do our molecular keys fit the correct locks? Biochem. Soc. Proc. 20, 85–89
13 Gilroy, S. (1996) Signal transduction in barley aleurone protoplasts is calcium
dependent and independent. Plant Cell 8, 2193–2209
14 Assmann, S.M. (1994) Ins and outs of guard cell ABA receptors. Plant Cell
6, 1187–1190
15 Lovegrove, A. et al. (1998) Gibberellin-photoaffinity labelling of two
polypeptides in plant plasma membrane. Plant J. 15, 311–320
16 Hooley, R. (1999) A role for G proteins in plant hormone signalling?
Plant Physiol. Biochem. 37, 393–402
17 Jones, H.D. et al. (1998) Heterotrimeric G proteins are implicated in
gibberellin induction of a-amylase gene expression in wild oat aleurone. Plant
Cell 10, 245–253
18 Mitsunaga, S. et al. (1994) Identification and characterization of gibberellininsensitive mutants selected from among dwarf mutants of rice. Theor. Appl.
Genet. 87, 705–712
19 Fujisawa, Y. et al. (1999) Suppression of the heterotrimeric G protein causes
abnormal morphology, including dwarfism, in rice. Proc. Natl. Acad. Sci.
U. S. A. 96, 7575–7580
20 Ashikari, M. et al. (1999) Rice gibberellin-insensitive dwarf mutant gene
Dwarf 1 encodes the a-subunit of GTP-binding protein. Proc. Natl. Acad. Sci.
U. S. A. 96, 10284–10289
21 Neer, E.J. (1995) Heterotrimeric G proteins: organizers of transmembrane
signals. Cell 80, 249–257
22 Plakidou-Dymock, S. et al. (1998) A higher plant seven-transmembrane
receptor that influences sensitivity to cytokinins. Curr. Biol. 8, 315–324
23 Gilroy, S. and Jones, R.L. (1992) Gibberellic acid and abscisic acid
coordinately regulate cytoplasmic calcium and secretory activity in barley
aleurone protoplasts. Proc. Natl. Acad. Sci. U. S. A. 89, 3591–3595

24 Bush, D.S. (1996) Effects of gibberellic acid and environmental factors on
cytosolic calcium in wheat aleurone cells. Planta 199, 89–99
25 Hillmer, S. et al. (1992) Visualizing enzyme secretion from individual barley
(Hordeum vulgare) aleurone protoplasts. Plant Physiol. 102, 279–286
26 Ritchie, S. et al. (1999) The sensitivity of barley aleurone tissue to gibberellins
is heterogeneous and may be spatially determined. Plant Physiol. 120, 361–370
27 Chen, X. et al. (1997) Cloning of a Ca21-ATPase gene and the role of
cytosolic Ca21 in the gibberellin-dependent signaling pathway in aleurone
cells. Plant J. 11, 363–371
28 Ritchie, S. and Gilroy, S. (1998) Abscisic acid signal transduction in the
barley aleurone is mediated by phospholipase D activity. Proc. Natl. Acad.
Sci. U. S. A. 95, 2697–2702
29 Wang, M. et al. (1991) Abscisic acid induces a cytosolic calcium decrease in
barley aleurone protoplasts. FEBS Lett. 278, 69–74
30 Schuurink, R.C. et al. (1996) Modulation of calmodulin mRNA and protein
levels in barley aleurone. Plant Physiol. 111, 371–380
31 Bush, D.S. (1995) Calcium regulation in plant cells and its role in signalling.
Annu. Rev. Plant Physiol. Plant Mol. Biol. 46, 95–122
32 Bush, D.S. and Wang, T. (1995) Diversity of calcium-efflux transporters in
wheat aleurone cells. Planta 197, 19–30
33 Homann, U. and Tester, M. (1997) Ca21-independent and Ca21/GTP-binding
protein-controlled exocytosis in a plant cell. Proc. Natl. Acad. Sci. U. S. A.
94, 6565–6570
34 Schuurink, R.C. et al. (1998) Characterization of a calmodulin-binding
transporter from the plasma membrane of barley aleurone. Proc. Natl. Acad.
Sci. U. S. A. 95, 1944–1949
35 Penson, S.P. et al. (1996) cGMP is required for gibberellic acid-induced gene
expression in barley aleurone. Plant Cell 8, 2325–2333
36 Bethke, P.C. et al. (1999) Hormonally regulated programmed cell death in
barley aleurone cells. Plant Cell 11, 1033–1045
37 Gomez-Cadenas, A. et al. (1999) An abscisic acid-inducible protein kinase,
PKABA1, mediates abscisic acid-suppressed gene expression in barley
aleurone layers. Proc. Natl. Acad. Sci. U. S. A. 96, 1767–1772
38 Ritchie, S. and Gilroy, S. (1998) Calcium-dependent protein phosphorylation may
mediate the gibberellic acid response in barley aleurone. Plant Physiol. 116, 765–776
39 Kuo, A. et al. (1996) Okadaic acid, a protein phosphatase inhibitor, blocks
calcium changes, gene expression, and cell death induced by gibberellin in
wheat aleurone cells. Plant Cell 8, 259–269
40 Heimovaara-Dijkstra, S. et al. (1996) Abscisic acid-induced gene-expression
requires the activity of protein(s) sensitive to the protein-tyrosine phosphatase
inhibitor phenylarsine oxide. Plant Growth Regul. 18, 115–123
41 Knetsch, M.L.W. et al. (1996) Abscisic acid induces mitogen-activated protein
kinase activation in barley aleurone protoplasts. Plant Cell 8, 1061–1067
42 Thornton, T.M. et al. (1999) Gibberellin signal transduction presents ...the
SPY who O-GlcNAc’d me. Trends Plant Sci. 4, 424–428
43 Robertson, M. et al. (1998) Identification of a negative regulator of gibberellin
action, HvSPY, in barley. Plant Cell 10, 995–1007
44 Harberd, N.P. et al. (1998) Gibberellin: inhibitor of an inhibitor of...?
BioEssays 20, 1001–1008
45 Peng, J. et al. (1999) ‘Green revolution’ genes encode mutant gibberellin
response modulators. Nature 400, 256–261
46 Flintham, J.E. and Gale, M.D. (1982) The Tom Thumb dwarfing gene, Rht3 in
wheat, I. Reduced pre-harvest damage to breadmaking quality. Theor. Appl.
Genet. 62, 121–126
47 Skriver, K. et al. (1991) cis-acting DNA elements responsive to gibberellin
and its antagonist abscisic acid. Proc. Natl. Acad. Sci. U. S. A. 88,
7266–7270
48 Huttly, A.K. et al. (1992) Localization of cis elements in the promoter of a
wheat a-Amy2 gene. Plant Mol. Biol. 19, 903–911
49 Gubler, F. and Jacobsen, J.V. (1992) Gibberellin-responsive elements in the
promoter of a barley high-pI a-amylase gene. Plant Cell 4, 1435–1441
50 Sutliff, T.D. et al. (1993) Gibberellin treatment stimulates nuclear factor
binding to the gibberellin response complex in a barley a-amylase promoter.
Plant Cell 5, 1681–1692

March 2000, Vol. 5, No. 3

109

trends in plant science
Reviews
51 Gubler, F. et al. (1995) Gibberellin-regulated expression of a myb gene in
barley aleurone cells: evidence for Myb transactivation of a high-pI a-amylase
gene promoter. Plant Cell 7, 1879–1891
52 Gubler, F. et al. (1997) Cloning of a rice cDNA encoding a transcription
factor homologous to barley GAMyb. Plant Cell Physiol. 38,
362–365
53 Weston, K. (1998) Myb proteins in life, death and differentiation. Curr. Opin.
Genet. Dev. 8, 76–81
54 Gubler, F. et al. (1999) Target genes and regulatory domains of the
GAMYB transcriptional activator in cereal aleurone. Plant Cell 17, 1–9
55 Raventos, D. et al. (1998) HRT, a novel zinc finger, transcriptional repressor
from barley. J. Biol. Chem. 273, 23313–23320
56 Hoecker, U. et al. (1995) Integrated control of seed maturation and
germination programs by activator and repressor functions of Viviparous-1 of
maize. Genes Dev. 9, 2459–2469
57 Rogers, J.C. and Rogers, S.W. (1992) Definition and functional implications
of gibberellin and abscisic acid cis-acting hormone response complexes.
Plant Cell 4, 1443–1451

58 Rogers, J.C. et al. (1994) The cis-acting gibberellin response complex in highpI a-amylase gene promoters. Plant Physiol. 105, 151–158
59 Willmott, R.L. et al. (1998) DNase1 footprints suggest the involvement of at
least three types of transcription factors in the regulation of a-Amy2/A by
gibberellin. Plant Mol. Biol. 38, 817–825
60 Tregear, J.W. et al. (1995) Functional analysis of linker insertions and point
mutations in the a-Amy2/54 GA-regulated promoter. Plant Mol. Biol. 29, 749–758
61 Rushton, P.J. et al. (1995) Members of a new family of DNA-binding proteins bind
to a conserved cis-element in the promoters of a-Amy2 genes. Plant Mol. Biol.
29, 691–702

Alison Lovegrove and Richard Hooley* are at IACR-Long Ashton
Research Station, Dept of Agricultural Sciences,
University of Bristol, Long Ashton, Bristol, UK BS41 9AF.
*Author for correspondence (tel 144 1275 392181;
fax 144 1275 394281; e-mail richard.hooley@bbsrc.ac.uk).

Formation and maintenance
of the shoot apical meristem
John L. Bowman and Yuval Eshed
Development in higher plants is characterized by the reiterative formation of lateral organs
from the flanks of shoot apical meristems. Because organs are produced continuously
throughout the life cycle, the shoot apical meristem must maintain a pluripotent stem cell
population. These two tasks are accomplished within separate functional domains of the
apical meristem. These functional domains develop gradually during embryogenesis.
Subsequently, communication among cells within the shoot apical meristem and between the
shoot apical meristem and the incipient lateral organs is needed to maintain the functional
domains within the shoot apical meristem.

P

ost-embryonic development in higher plants is characterized by the reiterative formation of lateral organs from the
flanks of apical meristems1. A shoot apical meristem (SAM)
is initially formed during embryogenesis, and derivatives of this
meristem give rise to the above-ground portion of the plant. The
SAM contains a population of pluripotent stem cells, which serve
three primary functions1–4:
(1) Lateral organs, such as leaves, are produced from the
peripheral regions of the SAM.
(2) The basal regions of the SAM contribute to the formation of
the stem.
(3) The stem cells of the SAM must replenish those regions
from which cells have been recruited and maintain the pool
of stem cells required for further growth.
In general, we focus on SAMs in this review, although extrapolation of concepts to other shoot meristems, such as flower
meristems, will be discussed when pertinent.
As a result of histological analyses the SAM has been subdivided in two different manners. First, three distinct zones of the
SAM are defined by cytoplasmic densities and cell division rates:
the peripheral zone, the central zone and the rib zone1–4 (Fig. 1).
These three zones might represent a functional subdivision of the
SAM although direct evidence for this is lacking. Lateral organs
are produced from cells recruited from the peripheral zone
1