Shoot apical meristem maintenance:
the art of a dynamic balance
Cristel C. Carles and Jennifer C. Fletcher
USDA/UC Berkeley, Plant Gene Expression Center, 800 Buchanan Street, Albany, CA 94710, USA
The aerial structure of higher plants derives from cells
at the tip of the stem, in the shoot apical meristem
(SAM). Throughout the life of a plant, the SAM produces
stem tissues and lateral organs, and also regenerates
itself. For correct growth, the plant must maintain
a constant flow of cells through the meristem, where
the input of dividing pluripotent stem cells offsets the
output of differentiating cells. This flow depends on
extracellular signaling within the SAM, governed by a
spatial regulatory feedback loop that maintains a reservoir
of stem cells, and on factors that prevent meristem
cells from differentiating prematurely. The terminating
floral meristem incorporates the spatial regulation
scheme into a temporal regulation pathway involving
flower patterning factors.
A specialty of plants is that they can modulate their body
plan after embryogenesis, an ability that helps them to
deal with the environmental changes that affect their
growth. This flexibility in development and organogenesis
is provided by pools of dividing, pluripotent cells that
reside in structures called meristems. Two meristematic
cell populations arise during embryogenesis [1] and grow
in polar directions throughout plant life: the shoot apical
meristem (SAM) generates the aerial part of the plant,
whereas the root apical meristem (RAM) generates the
underground part. The SAM continuously produces cells
that will become incorporated into stem tissue, lateral
organs (leaves and flowers) and axillary meristems. The
above-ground part of the plant can be divided into iterative
modules, called phytomers, each containing a leaf
(or leaves) attached at a node, an axillary meristem at
the base of the leaf(s) and a subtending internode (Fig. 1a).
After a phase of vegetative growth during which stem
tissues and leaves are formed, the SAM and each axillary
meristem turns into an inflorescence meristem that bears
flowers [2] (Fig. 1b).
The flowers are generated by specialized types of
meristems, the floral meristems (FMs), which are also
products of the SAM (Fig. 1c). Floral meristems have the
same organization in terms of cellular domains as the
SAMs [3] (see below) but they differ in two important ways.
First, FMs do not generate vegetative structures but only
flowers, which are composed of sepals, petals, stamens and
carpels in four rings (whorls) from the outside to the inside
of the flower (Fig. 1d). Second, FMs are terminating
structures: they have a transient activity that stops after
enough cells have been produced for their recruitment into
the most internal organs, the carpels [2]. Termination of
the FM requires the activity of specific factors that are not
present in the SAM.
Architecture of the SAM: layers and zones
Cytohistological analyses of several angiosperm species
reveal two main architectural aspects of the SAM [2]. The
first and most evident feature is its organization into cell
layers (Fig. 1c). The surface region or tunica consists of one
to five clonal layers – two in most dicots (L1 and L2) and
one in monocots – in which cells divide in a single plane,
whereas the underlying corpus (L3) is an arrangement of
cells that divide in all planes. The different layers
generally maintain distinct cell lineages, each generating
specific derivatives. In Arabidopsis, the L1 contributes to
the epidermis of shoots, leaves and flowers, the L2
produces the ground tissues and germ cells, and the L3
contributes to the vascular tissues of the stem and the
most internal tissues of leaves and flowers. However, the
development of each clonal layer is flexible and can adapt
to cell proliferation changes occurring in other layers, as
shown by studies of genetic mosaics [4].
Juxtaposed over the SAM cell layers are three zones
that have distinct functions (Fig. 1c). The peripheral zone
(PZ) and the rib zone (RZ) contain cells that will become
incorporated into lateral organs and the stem core,
respectively. The central zone (CZ), which is characterized
by a lower mitotic activity [5], constitutes the self-renewing
stem cell reservoir that is the source of cells for the PZ
and RZ. During primordia initiation, sets of cells in the PZ
and RZ become specified as the founder cells of young plant
tissues. These cells are replaced by dividing pluripotent
cells provided by the CZ, which correspondingly replenish
themselves through new cell divisions. In this way, the
SAM sustains itself as a stable structure, in spite of the
constant flow of cells passing through it.
To sustain the modular growth of the shoot and
simultaneously maintain SAM homeostasis, shifts in
gene expression must occur within cells as their
positions within the meristem change. This implies
that SAM cells determine their appropriate gene
expression states based on signal exchanges with
other cells in the meristem rather than through a
strict lineage-based predetermination of cell fate [4].
How groups of cells in the SAM signal one another to
learn their fates and coordinate proper meristemCorresponding author: Jennifer C. Fletcher (fletcher@nature.berkeley.edu).
Review TRENDS in Plant Science Vol.8 No.8 August 2003394
http://plants.trends.com 1360-1385/03/$ - see front matter q 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S1360-1385(03)00164-X
function is the subject of intense study. Here, we
mainly focus on recent data from Arabidopsis, which
have revealed that SAM maintenance involves an
extracellular signaling network, the CLAVATA (CLV)
pathway, that communicates cell fate information
between different SAM domains.
CLAVATA signal transduction pathway maintains stem
cell identity
CLAVATA ligand–receptor complex
The discovery of plants producing broader and distorted
meristems has caught naturalists’ interest for centuries.
Early reports describe these plants as fasciated (from the
Fig. 1. The shoot apical meristem (SAM) is a cell supply system for the aerial part of the plant. (a) The modular growth of the shoot is illustrated in this three-dimensional
model of a growing stem with leaves. The plant body generates from the iterative initiation of units called ‘phytomers’ composed of a leaf (or leaves) attached at a node, a
subtending internode and a bud – an axillary meristem – at the base of the leaf. Depending on their position on the shoot, phytomers differ in leaf size and shape, internode
length, and axillary meristem potential. Axillary meristems differ from the SAM in that they are formed postembryonically but both types of meristems can produce an
indefinite number of structures such as branches, leaves and flowers. (b) A mature Arabidopsis plant. The postembryonic growth of the aerial parts of the plant starts with
the rosette stage, during which the SAM initiates a variable number of leaves in a spiral phyllotaxy with unexpanded internodes. At the end of this stage, in response to
endogenous and environmental cues, the stem elongates and the SAM produces a variable number of cauline leaves that each develop an axillary meristem at its base.
The SAM is now an inflorescence, and initiates floral meristems (FMs) that will produce the flowers. (c) Confocal laser-scanning micrograph through an Arabidopsis inflorescence
SAM and its adjacent floral meristems (FMs). Continuous organogenesis is made possible by the activity of meristems, which are sources of pluripotent dividing
cells. Superimposed on the micrograph are the meristem layers and zones. The colored domains depict the different cell layers: in Arabidopsis, the tunica corresponds to
two layers of cells, the epidermal (L1) and subepidermal layers (L2), whereas the corpus corresponds to the internal layers (L3). The two arrowheads point at an anticlinal
cell division occurring in L2. The black outlines represent the approximate boundaries between the different meristematic zones: the peripheral zone (PZ) contributes cells
to the formation of lateral organs, the rib zone (RZ) contributes cells to stem growth and the central zone (CZ) acts as a ‘factory’ of cells for the PZ and RZ, but also for its
own replenishment with new cells. Although the cells of the PZ and the CZ are histologically distinguishable, there is no sharp boundary between each zone. The PZ and
the CZ contain both tunica and corpus cells, whereas the RZ is ‘buried’ beneath them in the deeper layers of the corpus. (d) Confocal laser-scanning micrograph of an Arabidopsis
FM. After the production of cells that are incorporated into the four whorls of organs – sepals (Se), petals (Pe), stamens (St) and carpels (Ca) – the FM terminates.
TRENDS in Plant Science
Main
inflorescence
SAM
Axillary
inflorescence
meristem
Flower
Rosette
Cauline leaf
(b)(a)
Ca
St
Se
Pe
St
Se
Pe
CaCZ
RZ
PZ
PZ
L1
L2
L3
CZ
RZ
PZ PZ
SAM FMFM
(c) (d)
PhytomerInternode
Node
Leaf
SAM
Organ primordia
Axillary meristem
Review TRENDS in Plant Science Vol.8 No.8 August 2003 395
http://plants.trends.com
Latin fascia, meaning ‘bundle’), because their meristems
grow as a band or a ring instead of as a point [6]. In
Arabidopsis, recessive loss-of-function mutations at three
clv loci cause the formation of fasciated meristems [7–9].
Whereas clv2 alleles display additional phenotypes, clv1
and clv3 mutations specifically affect the development of
above-ground meristems of all types such that they
broaden progressively, starting during embryogenesis.
Consequently, the mutant phenotypes become more
dramatic with time: the enlarged vegetative SAM generates
a fasciated stem, the inflorescence meristem
produces many extra floral meristems and the flower
meristems produce extra floral organs. clv mutants’ FMs
often form many extra carpels, giving a club-like shape to
the gynoecium [7–9]. The clavata mutants obtain their
name from this last characteristic, because clavatus
means ‘shaped like a club’ in Latin. Cell number is greatly
increased in the SAMs and FMs of the clv mutants [7–10]
(Fig. 2), which is more likely to be caused by a defect in the
transition of cells out of the CZ than by higher cell division
rates in the CZ [5]. Genetic studies have demonstrated
that CLV1, CLV2 and CLV3 act in the same pathway to
limit the expansion of the undifferentiated stem cell
population in SAMs and FMs [8,9]. These studies also
revealed that the CLV1 and CLV3 gene products have a
quantitative interdependence [8], which argues for CLV1
and CLV3 physically interacting in the same complex.
The cloning of the CLV genes identified them as
components of an extracellular signaling pathway. CLV1
and CLV2 encode a receptor-like kinase (RLK) and a
receptor-like protein (RLP), respectively, and are both
members of large gene families [11,12]. CLV1 contains an
extracellular region consisting of 21 leucine rich repeats
(LRRs), a transmembrane region and an intracellular
serine/threonine kinase domain. CLV2 contains 20 LRRs
and a transmembrane domain but, unlike CLV1, has a
short cytoplasmic tail lacking any signaling domain. CLV3
encodes a secreted polypeptide of 96 amino acids [13] and is
mainly expressed in the L1 and L2 of the central zone,
whereas CLV1 mRNA is mostly found in the L3 of the
central zone [11,13] (Fig. 3a). RNA blot experiments show
that CLV2 transcripts can be detected in many different
tissues, including SAMs and FMs, consistent with clv2
mutants displaying a pleiotropic phenotype [12].
Altogether, analyses of the CLV sequences, expression
domains and mutant phenotypes have led to the proposal
that the CLV factors act in a cell–cell signaling pathway
that extends between layers in the central zone of the
SAM. Moreover, mutations in a CLV2-like gene from Zea
mays called FASCIATED EAR2 (FEA2) cause fasciated
inflorescence meristem phenotypes similar to those of clv2
mutants [14], suggesting that the CLV pathway might be
conserved throughout the angiosperms.
The CLV cell signaling model is now supported by
genetic and biochemical studies jointly demonstrating that
CLV2 and CLV3 are required for the assembly of CLV1 into
an active 450 kDa signaling complex [12,15]. This complex
cannot be isolated from clv2 mutant protein extracts [12],
meaning that CLV2 probably associates as a heterodimer
with CLV1. Analyses of chimeric RLKs show that the
strong clv1 mutations, which are caused by mis-sense
mutations within the CLV1 extracellular domain, are
dominant negative [16]. The dominant negative activity of
these alleles can be accounted for if the mutant CLV1
proteins interact physically via their extracellular
domains with other receptors, and thus block their
function. These findings suggest that other RLKs overlap
functionally with CLV1 to fine tune meristem development
[16]. A simultaneous study of dominant negative alleles of
the ERECTA RLK arrived at a similar conclusion about
the redundancy of RLK signaling pathways involved in
organ shape regulation [17].
The mode of CLV3 action has also been investigated.
Genetic and immunological studies demonstrated the
requirement for CLV3 in the extracellular space for its
function in vivo to activate the CLV complex [18].
Furthermore, CLV1 and CLV2 are required to obtain the
gain-of-function phenotypes observed in CLV3-overexpressing
plants, indicating that CLV3 signaling depends
upon CLV1 and CLV2 [19]. These results provide evidence
that the CLV3 polypeptide is secreted from the overlying
L1 and L2 cell layers, and moves into the underlying L3,
where it is likely to act as the ligand for a CLV1/CLV2
signaling complex (Fig. 3). Nevertheless, at this time,
additional evidence supporting the whole scheme is still
missing, such as proof of a physical interaction between
CLV1 and CLV2 or between CLV1/2 and CLV3.
Fig. 2. Phenotypes of Arabidopsis shoot apical meristem (SAM) mutants. Longitudinal sections through the vegetative SAMs of 7-day-old (a) wild type, (b) clavata1-4
(clv1-4), (c) wuschel-1 (wus-1) and (d) shootmeristemless-11 (stm-11) plants. (b) The SAM of clv1-4 plants is both broader and taller than the wild-type SAM (a). (c) In wus-1
mutants, the shoot apex is flat, yet has a tunica–corpus organization. The SAM shown in (c) was probably about to terminate after the initiation of the two leaf primordia.
Eventually, this wus-1 seedling would have repetitively initiated axillary meristems, all terminating prematurely in a similar flat structure. (d) Plants carrying the stm-11
mutation lack a SAM and the two cotyledons are fused at their base. The arrow indicates the boundary between the two cotyledons. Scale bars ? 50 mm.
(a) (b) (c) (d)
Review TRENDS in Plant Science Vol.8 No.8 August 2003396
http://plants.trends.com
CLV–WUS feedback loop maintains the central reservoir
of stem cells
For the SAM to remain in a balanced state, the CLV
pathway, which restricts stem cell accumulation, has to be
counterbalanced by a factor or factors that promote stem
cell production. This function resides with the WUSCHEL
(WUS) homeodomain transcription factor. Nonsense
mutations in the WUS gene result in the mis-specification
of stem cells and the premature termination of the SAM
and FMs after the formation of a few organs [20,21] (Fig. 2).
Eventually, wus mutants undergo iterative processes of
meristem initiation and premature arrest, leading to the
production of disorganized groups of leaves and shoots.
Thus, these mutants were given the name wuschel,
meaning ‘tousled hair’ in German. It has been proposed
that the WUS-expressing domain, which is restricted to
the L3 of the CZ, acts as an organizing center (OC) that
specifies the overlying neighbor cells as stem cells [21].
Because wus mutations are epistatic to clv mutations,
WUS is likely to act downstream of the CLV pathway
[20,22]. WUS produces a non-cell-autonomous signal to
activate cell division, because mosaic expression of WUS
leads to the development of outgrowths that do not express
WUS themselves but appear in the vicinity of WUSexpressing
cells [23].
Several major works have shed considerable light on
how the antagonistic processes driven by the CLV complex
and WUS are coordinated within the SAM. In clv SAMs,
the reduction in CLV signaling leads to the expansion of
the WUS expression domain [22]. Mis-expression of WUS
in an enlarged domain is sufficient to cause a clv-like
phenotype, confirming that the clv meristem phenotype is
caused by the deregulation of WUS expression. Conversely,
ectopic expression of CLV3 is sufficient to repress
WUS expression throughout the meristem and thus leads
to a wus-like phenotype [19]. In wild-type plants, CLV3
protein spreads laterally from the stem cells, signaling to
both the stem cells themselves and their lateral neighbors
to repress transcription from the WUS promoter [24]. The
intercellular movement of CLV3 into the meristem center
is limited by the CLV1 receptor kinase, which is proposed
to sequester the CLV3 protein and to prevent its signal
from causing WUS transcriptional repression in the OC
[24]. Altogether, these results demonstrate that the CLV
pathway negatively regulates stem cell accumulation by
limiting WUS expression. Moreover, WUS expression is
sufficient to induce the expression of CLV3 and to promote
meristem cell identity [22].
Thus, a picture emerges of a feedback loop between the
stem cells and the OC that is mediated by CLV3 and WUS.
WUS-mediated signaling from the OC specifies stem cell
identity in the outermost cell layers, which signal back via
the CLV pathway to limit the size of the WUS-expressing
OC (Fig. 3a). The net result is meristem homeostasis in
which the output of cells into new organs is balanced by the
input of new stem cell derivatives (Box 1). New questions
Fig. 3. Coordination of cell proliferation and cell fate decisions across the shoot
apical meristem (SAM). (a) Stem cell regulation in the SAM via the CLAVATA–
WUSCHEL feedback loop. CLAVATA3 (CLV3) expression is restricted mainly to the
L1 and L2 of the central zone [13], whereas CLV1 mRNA can only be detected in the
L3 of the central zone and rib zone [11]. WUSCHEL (WUS) is expressed in few central
cells of the rib zone [21]. The CLV–WUS feedback loop consists of a WUSmediated
signal from the organizing center (OC) that specifies stem cell identity in
the outermost layers, which signal back via the CLV pathway to limit the size of the
WUS-expressing OC. (b) CLV transduction pathway at the cellular level. The CLV1
and CLV2 leucine-rich repeat (LRR) receptor proteins might homo- or heterodimerize
through the formation of disulfide bridges between the conserved cysteines
pairs (SS) flanking the LRR domain. The binding of CLV3 to CLV1 and/or CLV2
could promote the assembly of the 450 kDa complex, which also contains a protein
phosphatase (KAPP) and a Rho-like GTPase (Rop). KAPP is a negative regulator of
CLV signaling, capable of dephosphorylating CLV1. Rop transduces the signal
from the CLV complex to downstream targets. By analogy with animal systems
[56], the binding of Rop to CLV1 might be mediated by a linker protein, the nature
of which remains to be determined. The signaling cascade from the plasma-membrane-bound
CLV complex leads to the downregulation of WUS transcription in
the nucleus. This intracellular signaling cascade might involve mitogen-activated
protein kinases (MAPKs), based on the example of other LRR receptor-like kinase
CLV3
CLV1
WUS
TRENDS in Plant Science
(a)
(b)
CLV3
CLV3
CLV2CLV1
SS
P
P
P
P
P
P
P
Rop
KAPP
GTP
GDP
MAPKs?
?
POL
2
1
1
WUS
signaling pathways [57]. The protein phosphatase 2C POLTERGEIST (POL) acts
either (1) as a positive regulator of WUS transcription that is antagonized by CLV
signaling or (2) as a direct inhibitor of CLV signal transduction, analogous to KAPP
but acting at a different point in the pathway.
Review TRENDS in Plant Science Vol.8 No.8 August 2003 397
http://plants.trends.com
arise from these results. What is the intracellular pathway
leading to the inactivation of WUS? How does WUS send a
signal from the OC to the overlying stem cells? Does WUS
itself migrate between the cell layers or is the signal
relayed by other factors? If a WUS-induced target gene
encodes the signal, how is it perceived?
The group of CLAVATA interactors has recently enlarged
Two immediate downstream components of the CLV
signaling complex are a kinase-associated protein phosphatase
(KAPP) [25] and a Rho-like GTPase (Rop) [26,27].
These proteins interact directly with CLV1 in the 450 kDa
active signaling complex [15] (Fig. 3b). KAPP functions in
vivo as a negative regulator of the CLV pathway through
direct dephosphorylation of CLV1, preventing downstream
components from receiving the CLV1 signal
[28,29]. The Rop GTPase is proposed to transduce the
signal from the receptor complex toward the nucleus [15].
However, the question of how the signal reaches the
nucleus from the plasma-membrane-bound signaling
complex is as yet unresolved.
Within the past year, several new factors involved in
CLV signaling have been identified. The SHEPHERD
(SHD) protein might interact directly with the CLAVATA
complex, assisting in its assembly [30]. The shd mutation
causes the same meristem phenotypes as weak and
intermediate clv alleles, although it also has more
pleiotropic effects. Mutations in CLV and WUS are
epistatic to shd, suggesting that SHD functions in the
CLV pathway. Moreover, the CLV3 and WUS expression
domains are greatly enlarged in shd mutants and the effect
of CLV3 overexpression is abolished. Thus, the function of
the CLV pathway in suppressing WUS expression depends
on SHD activity. SHD encodes an ortholog of the
mammalian GRP94 chaperone protein and is expressed
throughout the SAM and FMs. By analogy with GRP94,
SHD is likely to be involved in achieving the proper
conformation or association of the CLV proteins. SHD is
also expressed in the root and the RAM of shd mutants is
disorganized, suggesting that SHD might promote the
assembly of CLV1-like RLKs that are expressed in roots.
POLTERGEIST (POL) is a newly identified intracellular
component of the CLV signaling pathway. Mutations in
POl are suppressors of clv mutant phenotypes and genetic
evidence places POL as a downstream regulator of CLV
signal transduction [31]. In a pol background, wus
mutations are semidominant and are no longer epistatic
to clv1 mutations [32], suggesting that POL might function
downstream of CLV through two pathways, one WUS
dependent and the other WUS independent. Because pol
reveals its effect only when associated with other
mutations, it has been named ‘the noisy ghost’. POL
encodes a nuclear-localized protein phosphatase 2C
(PP2C) that is expressed broadly throughout the plant
and is predicted to act in multiple signaling pathways [32].
It is not clear yet whether POL functions in the SAM as a
direct inhibitor of CLV signal transduction or whether it is
a positive regulator of WUS that is repressed by the CLV
pathway (Fig. 3b). The identification of POL as a new
player in CLV signaling emphasizes the idea that
meristem balance is achieved by complex mechanisms
that often involve negative interactions between factors.
Combined effects of homeobox genes prevent stem cell
differentiation
The prevention of stem cell differentiation involves several
KNOX (KNOTTED1-like homeobox) genes. KNOTTED1
(KN1), the first homeodomain transcription factor identified
in plants [33], was isolated from a maize (Zea mays)
gain-of-function mutant that produced outgrowths
(‘knots’) of undifferentiated tissue on the leaves [34].
SHOOTMERISTEMLESS (STM) is the Arabidopsis ortholog
of KN1 [35] and was isolated from loss-of-function
mutants that fail to establish and maintain the SAM such
that no lateral organs are produced once the cotyledons
have formed [36] (Fig. 2). Thus, STM is required to
maintain the indeterminate cell fate and to prevent cell
differentiation in the meristem, where it acts independently
of CLV3 and WUS [23,37,38]. STM, which is
expressed throughout the SAM, is not absolutely required
to activate CLV3 transcription [39]. However, its activity is
necessary in addition to that of WUS to maintain an
appropriately high level of CLV3 expression at the
meristem apex.
Three other Arabidopsis KNOX genes (KNAT1, KNAT2
and KNAT6) are, like STM, expressed in the shoot apex
and function redundantly with it [40–42]. The role of these
KNOX genes is to restrict the expression of the ASYMMETRIC
LEAVES1 (AS1) and AS2 genes to organ
primordia, and thus to prevent inappropriate leaf development
at the shoot apex. In turn, AS1 and AS2 negatively
regulate KNOX gene expression, excluding them from
Box 1. Factors that maintain the organization of the shoot
apical meristem
Independent of the CLAVATA (CLV) pathway are other factors that
regulate the cellular organization of the shoot apical meristem
(SAM). Two such factors are the FASCIATED1 (FAS1) and FAS2
proteins, which were identified from the fas mutants, in which the
cellular and functional organization of both the SAM and the root
apical meristem (RAM) are disturbed [10,58]. The fas1 and fas2 SAM
phenotypes are linked to the distortion of WUSCHEL (WUS)
expression pattern in a broader and more random manner than in
clv mutants, regardless of the cell layers and zones. The FAS proteins
are components of Arabidopsis chromatin assembly factor 1 and
thus are likely to ensure the appropriate state of gene expression,
including WUS expression, by facilitating reformation of the correct
chromatin structure after replication [58]. A third factor is an
APETALA2/ETHYLENE RESPONSE FACTOR (AP2/ERF) type transcription
factor that was recently isolated [59] from the activation
tagged mutant dornro¨ schen (drn). Ectopic expression of DRN in
drn-D mutants causes a dramatic increase in SAM size, followed by
premature meristem arrest. The terminated drn-D meristems display
perturbations in cell layering and expanded and reorganized CLV3
and WUS expression domains [59]. DRN is expressed in the
meristem central zone at all stages of development and also in
lateral organ primordia. Double mutant analysis shows that ectopic
DRN can function independently of SHOOTMERISTEMLESS (STM),
WUS and CLV activities. A drn loss offunction mutant is aphenotypic,
giving DORNRO¨ SCHEN its name (‘sleeping beauty’). A second,
closely related gene is also present in the genome. Thus, DRN is likely
to be a redundant component of a new SAM regulatory pathway,
independent of the CLV–WUS signaling system. In this scheme, DRN
would act to repress stem cell fate, counteracting WUS function.
Review TRENDS in Plant Science Vol.8 No.8 August 2003398
http://plants.trends.com
organ primordia [41–43]. AS1 encodes a Myb domain
transcription factor [40], and AS2 encodes a putative
transcription factor containing a leucine zipper domain
[44]. Reciprocal negative interactions between KNOX
transcription factors expressed in meristems and Myb
transcription factors expressed in organ primordia have
also been reported in Zea mays [45,46] and Antirrhinum
majus [47], where they are likewise required to distinguish
between stem cells and organ founder cells.
Ectopic expression of STM is sufficient to induce the cell
division machinery as well as KNAT gene expression
[23,38]. However, ectopic STM expression does not
activate CLV3 expression [38], showing that STM is not
capable of inducing stem cell identity on its own. Both STM
and WUS functions are in fact required and sufficient to
activate CLV1 and CLV3 expression in non-meristematic
cells [23,38,39]. Altogether, these observations suggest
that STM and WUS serve distinct yet complementary
functions in the SAM. STM prevents meristem cells from
differentiating prematurely, whereas WUS specifies a
subset of cells at the apex of the meristem as stem cells.
In conclusion, the joint functions of these two factors are
essential for a continuous formation of organ primordia and
stem tissues from cells provided by the self-maintaining
meristem (Box 2).
Particular case of the terminating FM
The FM constitutes a particular case in which stem cell
activity must be turned off for morphogenesis to be
completed. The FM sequentially produces sepal, petal
and stamen primordia from stem cell daughters on its
flanks. The remaining meristematic cells in the center of
the FM then differentiate and form the carpel organs. The
AGAMOUS (AG) MADS domain transcription factor, one
function of which is to specify flower organ identity, is also
required for FM termination [48]. Early in flower development,
WUS activates AG transcription, thus leading to
its own transcriptional demise: AG later switches off OC
activity by repressing WUS expression, resulting in the
differentiation of the last stem cells [20,49,50]. Interestingly,
AG seems to repress WUS independently of the CLV
pathway [50].
The discovery of this ‘suicidal’ temporal feedback
mechanism raises two questions. First, why does this
phenomenon occur specifically in the FM and not in the
SAM? At least part of the answer is that the activation of
AG by WUS requires the presence of a third element [49],
the LEAFY (LFY) transcription factor, which specifies FM
identity [51] and is not expressed in the SAM [52]. WUS
and LFY bind separately to adjacent enhancer sequences
in the AG second intron [50]. Second, what makes the
arrest of meristem activity such a time-specific event, so
that enough cells have been produced to ensure the correct
formation of each whorl? This question seems to find its
answer in the increasing level of AG expression observed
during flower development [53]. There might be a
threshold amount of AG necessary for efficient repression
of WUS, so only at the time of carpel initiation has enough
AG accumulated in the FM to repress WUS expression.
Another possibility is the participation of an additional
factor, the activity of which begins at the time of carpel
differentiation. The study of plants overexpressing AG
favors this last hypothesis [54]. In addition, because the
AG expression domain is larger than the WUS domain and
because LFY is expressed throughout the FM, there might
be other factors involved in the spatial activation of AG,
unless WUS protein itself moves from cell to cell. Thus, FM
development involves two negative feedback loops, one the
spatial CLV–WUS feedback loop shared with the SAM and
the other a WUS–AG temporal feedback loop specific to
the FM (Fig. 4).
Recently, the histone acetyltransferase AtGCN5 has
been shown to be required for proper FM activity and to act
by regulating the WUS–AG pathway [55]. Plants carrying
a T-DNA insertion into the bromodomain of AtGCN5
display homeotic transformations offlower organs, and the
conversion of the inflorescence meristem into a terminal
flower. These phenotypes correlate with upregulation of
WUS and AG transcription in FMs, expansion of the WUS
and AG expression domains within FMs, and ectopic
induction of AG in the inflorescence apex. These data
Box 2. HAM from Petunia reveals a function for
differentiating primordial cells in maintaining the
uncommitted state of meristematic cells
The HAIRY MERISTEM (HAM) factor from Petunia hybrida constitutes
a relay between the differentiating cells in new organ primordia
and the uncommitted cells at the shoot apex. Shoot apical meristems
(SAMs) of ham mutants terminate prematurely in differentiated stem
tissues that fail to respond to Petunia hybrida SHOOTMERISTEMLESS
(PhSTM) and WUSCHEL (PhWUS) activities [60]. HAM encodes
a putative transcription factor of the GRAS family [61], and is
expressed in the L3 cells of organ primordia and the provasculature
of the stem. Because it is expressed in differentiating cells but affects
the SAM, HAM functions in a non-cell-autonomous manner, in
parallel with PhWUS. These data reveal that differentiating cells, as
they leave the meristem field, actively maintain a pool of uncommitted
cells in the SAM through HAM activity. Such a mechanism,
revealing a new level in the maintenance of the meristem cells in an
undifferentiated state, is likely to be shared by other plant species.
Fig. 4. Floral meristem (FM) termination via a WUSCHEL–AGAMOUS temporal
feedback loop. (a) In the early stages of flower development, initiation of AGAMOUS
(AG) expression in the center of the FM is mediated by WUSCHEL (WUS)
and LEAFY (LFY). Because the AG expression domain (blue) is larger than the
WUS expression domain (orange) and because LFY is expressed throughout the
meristem (yellow and blue domains), another factor (X) or factors is likely to be
involved in the spatial activation of AG. (b) At the time of carpel initiation, AG and
at least one other factor (Y) represses WUS expression to terminate stem cell
activity. The last meristematic cells are then consumed in the formation of the
gynoecium. Repression of WUS by AG alone is not sufficient to lead to FM termination,
as observed in plants overexpressing AG [54].
(a) (b)
LFY
AG
WUS+
AG +Y
WUS
X
Review TRENDS in Plant Science Vol.8 No.8 August 2003 399
http://plants.trends.com
reveal an important role for chromatin modification in
regulating the level and domain of expression of key
components of the temporal feedback loop.
Prospects
The continuous production of new lateral organs by the
SAM of growing plants depends upon the presence of
mechanisms that permit a constant supply of new cells for
organogenesis while preventing those cells from differentiating
prematurely. Several recent studies have provided
exciting new insights into the molecular
mechanisms of Arabidopsis SAM maintenance. These
analyses have revealed that SAM homeostasis, rather
than relying upon genetically predetermined cell fate
specification, requires the active exchange of signals
between cells to maintain meristem organization and
function as the shoot tip grows.
Even though key genes have been identified that are
involved in several elegant spatial and temporal meristem
feedback loops, our understanding of meristem maintenance
mechanisms has not reached its apogee yet (Box 3).
A major unanswered question is how, or indeed if, the
CLV1, CLV2 and CLV3 gene products physically interact
in a protein signaling complex. The most parsimonious
hypothesis is that CLV3 acts as a ligand that directly binds
to the extracellular LRR domains of CLV1, CLV2 or a
heterodimer of both receptor proteins. However, other
models cannot be ruled out, such as CLV3 facilitating
ligand binding to CLV1 and/or CLV2, or being involved in
producing the ligand, perhaps by proteolytic processing of
a ligand precursor protein. Further biochemical analysis
will be required to resolve this crucial issue.
In addition, many other areas of plant stem cell
signaling research will continue to yield exciting results.
Identification of the complete set of CLV–WUS and
WUS–AG signal transduction components is ongoing,
and will require creative experimental designs to find
factors that have not been uncovered in traditional
forward genetic screens. The nature of the signals driven
by the CLV–WUS factors is still not completely elucidated,
and neither are the modes of regulation to which these
factors are subjected. The recent characterization of
several new factors in the CLV pathway opens the door
to the isolation of many more that will fit into the current
picture.
Acknowledgements
We thank Dan Choffnes, Leor Williams, Vijay Sharma, Giovanni Mele and
Robert Blanvillain for critical reading of this manuscript. Work in J.C.F.’s
laboratory is supported by grants from the US Department of Agriculture
and the National Science Foundation.
References
1 Jurgens, G. (2001) Apical–basal pattern formation in Arabidopsis
embryogenesis. EMBO J. 20, 3609–3616
2 Steeves, T.A. and Sussex, I.M. (1989) Patterns in Plant Development,
Cambridge University Press
3 Milksche, J.P. and Brown, J.A.M. (1965) Development of vegetative
and floral meristems of Arabidopsis thaliana. Am. J. Bot. 52, 533–537
4 Szymkowiak, E.J. and Sussex, I.M. (1996) What chimeras can tell us
about plant development. Annu. Rev. Plant Physiol. Plant Mol. Biol.
47, 351–376
5 Laufs, P. et al. (1998) Cellular parameters of the shoot apical meristem
in Arabidopsis. Plant Cell 10, 1375–1389
6 Gorter, C.J. (1965) Origin of fasciation. In Encyclopedia of Plant
Physiology (Vol. 15, Part II) (Ruhland, W., ed.), pp. 330–351, Springer-
Verlag
7 Clark, S.E. et al. (1993) CLAVATA1, a regulator of meristem and flower
development in Arabidopsis. Development 119, 397–418
8 Clark, S.E. et al. (1995) CLAVATA3 is a specific regulator of shoot
and floral meristem development affecting the same processes as
CLAVATA1. Development 121, 2057–2067
9 Kayes, J.M. and Clark, S.E. (1998) CLAVATA2, a regulator of meristem
and organ development in Arabidopsis. Development 125, 3843–3851
10 Leyser, H.M.O. and Furner, I.J. (1992) Characterisation of three shoot
apical meristem mutants of Arabidopsis thaliana. Development 116,
397–403
11 Clark, S.E. et al. (1997) The CLAVATA1 gene encodes a putative
receptor kinase that controls shoot and floral meristem size in
Arabidopsis. Cell 89, 575–585
12 Jeong, S. et al. (1999) The Arabidopsis CLAVATA2 gene encodes a
receptor-like protein required for the stability of the CLAVATA1
receptor-like kinase. Plant Cell 11, 1925–1933
13 Fletcher, J.C. et al. (1999) Signaling of cell fate decisions by CLAVATA3
in Arabidopsis shoot meristems. Science 283, 1911–1914
14 Taguchi-Shiobara, F. et al. (2001) The fasciated ear2 gene encodes a
leucine-rich repeat receptor-like protein that regulates shoot meristem
proliferation in maize. Genes Dev. 15, 2755–2766
15 Trotochaud, A.E. et al. (1999) The CLAVATA1 receptor-like kinase
requires CLAVATA3 for its assembly into a signaling complex that
includes KAPP and a Rho-related protein. Plant Cell 11, 393–405
16 Dievart, A. et al. (2003) CLAVATA1 dominant-negative alleles reveal
functional overlap between multiple receptor kinases that regulate
meristem and organ development. Plant Cell 15, 1198–1211
17 Shpak, E.D. et al. (2003) Dominant-negative receptor uncovers
redundancy in the Arabidopsis ERECTA leucine-rich repeat receptor-like
kinase signaling pathway that regulates organ shape. Plant
Cell 15, 1095–1110
18 Rojo, E. et al. (2002) CLV3 is localized to the extracellular space, where
it activates the Arabidopsis CLAVATA stem cell signaling pathway.
Plant Cell 14, 969–977
19 Brand, U. et al. (2000) Dependence of stem cell fate in Arabidopsis on a
feedback loop regulated by CLV3 activity. Science 289, 617–619
20 Laux, T. et al. (1996) The WUSCHEL gene is required for shoot and
floral meristem integrity in Arabidopsis. Development 122, 87–96
Box 3. Are common maintenance mechanisms shared by
the shoot apical meristem and the root apical meristem?
Shoot apical meristems (SAMs) do not share their set of regulatory
factors with root apical meristems (RAMs), yet they adjust their cell
populations according to the same basic mechanisms, such as
intercellular signaling. In both SAMs and RAMs, these mechanisms
involve interactions between two groups of cell populations, the
pluripotent undifferentiated cells –in the organizing center of
the SAM and in the quiescent center (QC) in the RAM – and the
differentiating cells that will be incorporated into the plant body [62].
A recent elegant work shows that in the Arabidopsis root, the stem
cell population depends on the activity of the adjacent QC, which
positions the stem cell niche properly in the RAM. This activity
depends on the function of the SCARECROW putative transcription
factor [63]. The SAM and the RAM might thus be derivative forms of a
common meristem organization that depending on the developmental
context would produce shoot or root organs. Some SAM and
RAM factors belong to the same genes families, which is consistent
with such a hypothesis. For example, the Arabidopsis SHORTROOT
factor, which is expressed in roots and controls cell fate by
intercellular movement [64], is a homolog of HAIRY MERISTEM
[60] (Box 2). Moreover, the FASCIATED (Box 1) and SHEPHERD
(see text) protein examples suggest that, even if the SAM and the
RAM have converged on independent but parallel mechanisms for
regulating their cell populations, they seem to share common
upstream regulators.
Review TRENDS in Plant Science Vol.8 No.8 August 2003400
http://plants.trends.com
21 Mayer, K.F.X. et al. (1998) Role of WUSCHEL in regulating stem cell
fate in the Arabidopsis shoot meristem. Cell 95, 805–815
22 Schoof, H. et al. (2000) The stem cell population of Arabidopsis shoot
meristems is maintained by a regulatory loop between the CLAVATA
and WUSCHEL genes. Cell 100, 635–644
23 Gallois, J-L. et al. (2002) Combined SHOOT MERISTEMLESS and
WUSCHEL trigger ectopic organogenesis in Arabidopsis. Development
129, 3207–3217
24 Lenhard, M. and Laux, T. (2003) Stem cell homeostasis in the
Arabidopsis shoot meristem is regulated by intercellular movement
of CLAVATA3 and its sequestration by CLAVATA1. Development 130,
3163–3173
25 Stone, J.M. et al. (1994) Interaction of a protein phosphatase with an
Arabidopsis serine-threonine receptor kinase. Science 266, 793–795
26 Li, H. et al. (1998) Arabidopsis Rho-related GTPases: differential gene
expression in pollen and polar localization in fission yeast. Plant
Physiol. 118, 407–417
27 Li, H. et al. (2001) The Rop GTPase switch controls multiple
developmental processes in Arabidopsis. Plant Physiol. 126, 670–684
28 Williams, R.W. et al. (1997) A possible role for kinase-associated
protein phosphatase in the Arabidopsis CLAVATA1 signaling pathway.
Proc. Natl. Acad. Sci. U. S. A. 94, 10467–10472
29 Stone, J.M. et al. (1998) Control of meristem development by
CLAVATA1 receptor kinase and kinase-associated phosphatase interactions.
Plant Physiol. 117, 1217–1225
30 Ishiguro, S. et al. (2002) SHEPHERD is the Arabidopsis GRP94
responsible for the formation of functional CLAVATA proteins. EMBO
J. 21, 898–908
31 Yu, L.P. et al. (2000) POLTERGEIST functions to regulate meristem
development downstream of the CLAVATA loci. Development 127,
1661–1670
32 Yu, L.P. et al. (2003) POLTERGEISTencodes a protein phosphatase 2C
that regulates the CLAVATA pathways controlling stem cell identity at
Arabidopsis shoot and floral meristems. Curr. Biol. 13, 179–188
33 Vollbrecht, E. et al. (1991) The developmental gene Knotted-1 is a
member of a maize homeobox gene family. Nature 350, 241–243
34 Hake, S. and Freeling, S. (1985) Developmental genetics of mutants
that specify Knotted leaves in maize. Genetics 111, 617–634
35 Long, J.A. et al. (1996) A member of the KNOTTED class of
homeodomain proteins encoded by the STM gene of Arabidopsis.
Nature 379, 66–69
36 Barton, M.K. and Poethig, R.S. (1993) Formation of the shoot apical
meristem in Arabidopsis thaliana: an analysis of development in the
wild type and in the shoot meristemless mutant. Development 119,
823–831
37 Clark, S.E. et al. (1996) The CLAVATA and SHOOT MERISTEMLESS
loci competitively regulate meristem activity in Arabidopsis. Development
122, 1567–1575
38 Lenhard, M. et al. (2002) The WUSCHEL and SHOOTMERISTEMLESS
genes fulfill complementary roles in Arabidopsis shoot meristem
regulation. Development 129, 3195–3206
39 Brand, U. et al. (2002) Regulation of CLV3 expression by two homeobox
genes in Arabidopsis. Plant Physiol. 129, 565–575
40 Byrne, M.E. et al. (2000) Asymmetric leaves1 mediates leaf patterning
and stem cell function in Arabidopsis. Nature 408, 967–971
41 Byrne, M.E. et al. (2002) ASYMMETRIC LEAVES1 reveals knox gene
redundancy in Arabidopsis. Development 129, 1957–1965
42 Semiarti, E. et al. (2001) The ASYMMETRIC LEAVES2 gene of
Arabidopsis thaliana regulates formation of a symmetric lamina,
establishment of venation and repression of meristem-related homeobox
genes in leaves. Development 128, 1771–1783
43 Ori, N. et al. (2000) Mechanisms that control knox gene expression in
the Arabidopsis shoot. Development 127, 5523–5532
44 Iwakawa, H. et al. (2002) The ASYMMETRIC LEAVES2 gene of
Arabidopsis thaliana, required for formation of a symmetric flat leaf
lamina, encodes a member of a novel family of proteins characterized
by cysteine repeats and a leucine zipper. Plant Cell Physiol. 43,
467–478
45 Timmermans, M.C.P. et al. (1999) ROUGH SHEATH2: a Myb protein
that represses knox homeobox genes in maize lateral organ primordia.
Nature 284, 151–153
46 Tsiantis, M. et al. (1999) The maize rough sheath2 gene and leaf
development programs in monocot and dicot plants. Nature 284,
154–156
47 Waites, R. et al. (1998) The PHANTASTICA gene encodes a MYB
transcription factor involved in growth and dorsoventrality of lateral
organs in Antirrhinum. Cell 93, 779–789
48 Yanofsky, M.F. et al. (1990) The protein encoded by the Arabidopsis
homeotic gene AGAMOUS resembles transcription factors. Nature
346, 35–39
49 Lenhard, M. et al. (2001) Termination of stem cell maintenance in
Arabidopsis floral meristems by interactions between WUSCHEL and
AGAMOUS. Cell 105, 805–814
50 Lohmann, J.U. et al. (2001) A molecular link between stem cell
regulation and floral patterning in Arabidopsis. Cell 105, 793–803
51 Weigel, D. et al. (1992) LEAFY controls floral meristem identity in
Arabidopsis. Cell 69, 843–859
52 Blazquez, M.A. et al. (1997) LEAFYexpression and flower initiation in
Arabidopsis. Development 124, 3835–3844
53 Sieburth, L.E. et al. (1995) Genetic separation of third and fourth
whorl functions of AGAMOUS. Plant Cell 7, 1249–1258
54 Mizukami, Y. and Ma, H. (1997) Determination of Arabidopsis floral
meristem identity by AGAMOUS. Plant Cell 9, 393–408
55 Bertrand, C. (2003) Arabidopsis histone acetyltransferase AtGCN5
regulates the floral meristem activity through the WUSCHEL/
AGAMOUS pathway. J. Biol. Chem. M302787200 (www.jbc.org)
56 Simon, R.A. et al. (1993) An SH3–SH2–SH3 protein is required for
p21–Ras1 activation and binds to sevenless and Sos proteins in vitro.
Cell 73, 169–177
57 Asai, T. et al. (2002) MAP kinase signalling cascade in Arabidopsis
innate immunity. Nature 415, 977–983
58 Kaya, H. et al. (2000) FASCIATA genes for chromatin assembly factor-
1 in Arabidopsis maintain the cellular organization of apical
meristems. Cell 104, 131–142
59 Kirch, T. et al. (2003) The DORNROSCHEN/ENHANCER OF
SHOOT REGENERATION1 gene of Arabidopsis acts in the control
of meristem cell fate and lateral organ development. Plant Cell 15,
694–705
60 Stuurman, J. et al. (2002) Shoot meristem maintenance is controlled by
a GRAS-gene mediated signal from differentiating cells. Genes Dev. 16,
2213–2218
61 Pysh, L.D. et al. (1999) The GRAS gene family in Arabidopsis:
sequence characterization and basic expression analysis of the
SCARECROW-LIKE genes. Plant J. 8, 111–119
62 Nakajima, K. and Benfey, P. (2002) Signalling in and out: control of cell
division and differentiation in the shoot and root. Plant Cell 14
(Suppl.), S265–S276
63 Sabatini, S. et al. (2003) SCARECROW is involved in positioning the
stem cell niche in the Arabidopsis root meristem. Genes Dev. 17,
354–358
64 Nakajima, K. et al. (2001) Intercellular movement of the putative
transcription factor SHR in root patterning. Nature 413, 307–311
Review TRENDS in Plant Science Vol.8 No.8 August 2003 401
http://plants.trends.com