Nuclear Endosperm Development in Cereals and
Arabidopsis thaliana
Odd-Arne Olsen1
Pioneer Hi-Bred International, A DuPont Company, Johnston, Iowa 50131
INTRODUCTION
The nuclear endosperm of monocots, including the cereal
species maize, rice, barley, and wheat, represents humankinďs
most important renewable source of food, feed, and industrial
raw materials. In addition, the endosperm is an attractive system
for developmental biology studies. Similar to the embryo, the
endosperm is the result of a fertilization process and therefore
may be considered an organism in its own right. Endosperm
development has been the subject of several recent reviews
(DeManson, 1997; Becraft et al., 2000; Becraft, 2001; Olsen,
2001; Brown et al., 2002; Berger, 2003). This review emphasizes
the main developmental aspects of nuclear endosperm
development in cereals and Arabidopsis thaliana, including
evolutionary origin, coenocyte development, endosperm cellularization,
cell fate specification, and differentiation.
After fertilization in what is the most common type of
endosperm development, the nuclear type, the initial endosperm
nucleus divides repeatedly without cell wall formation, resulting
in a characteristic coenocyte-stage endosperm. Cellularization
occurs via the formation of radial microtubule systems (RMS) and
alveolation. Many dicot species, including commercially important
crop plants such as soybean, cotton, and Arabidopsis, also
have a nuclear endosperm, although the endosperm is consumed
in part by the embryo during seed maturation. During its
development, striking similarities exist between the endosperm
in these groups, particularly with respect to the cellularization
phase. Insight into the genetic specification of the different
processes in endosperm development (e.g., coenocyte formation,
alveolation, cell fate specification, and differentiation) may
ultimately reveal how developmental subprograms are integrated
into a master program specifying the entire endosperm
body plan. This knowledge will not only benefit basic science but
also strengthen efforts to improve cereal grain quality.
THE MEGAGAMETOPHYTE AND THE CENTRAL CELL
OF CEREALS AND ARABIDOPSIS
The nuclear endosperm of angiosperms, including the monocot
cereals and the dicot Arabidopsis, develops from the central cell
of the megagametophyte after the process of double fertilization
(Figure 1), resulting in a diploid embryo and a triploid endosperm
(for definition of terms used throughout the text, see Table 1).
There are two main reasons why megagametophyte development
is of importance in understanding endosperm development.
First is the likelihood of developmental cues being
established that affect the path of endosperm development.
Second, the debate on the evolutionary origin of the nuclear
endosperm itself is based on the evolution of the megagametophyte.
The Polygonum type of megagametophyte found in
cereals and Arabidopsis results from a differentiation process in
which the nucleus of the surviving megaspore first undergoes
one mitotic division without cell wall deposition between sister
nuclei, followed by nuclear migration to opposite ends of the
megaspore. Two additional rounds of mitoses without cell wall
deposition result in two groups of four cells near each pole of the
developing megagametophyte (Figures 1A and 1B). One nucleus
from each pole (the polar nuclei) migrates to the center of the
embryo sac, while cell walls are deposited around the remaining
three nuclei at each pole in a process that is believed to involve
RMS (see below) (Russel, 1993) (Figure 1C). This process results
in a seven-cell, eight-nuclei megagametophyte or embryo sac in
which the central cell receives two nuclei that may either remain
separate until fertilization, as in maize (Figure 1D), or fuse before
fertilization, as in Arabidopsis (Figure 1E) (Webb and Gunning,
1990; Mansfield et al., 1991; Schneitz et al., 1995; Christensen
et al., 1997; Drews et al., 1998). The central cell, which develops
into the endosperm after fertilization, fills the majority of the
volume of the embryo sac and consists of a proximal mass of
cytoplasm and a thin line of cytoplasm surrounding a large
central vacuole (Figure 2A).
Not surprisingly, genetic evidence shows that mutants
affecting maternal ovule tissues also affect endosperm development.
A priori, these maternal effects are expected to be of
two types, gametophytic (effects expressed in the megagametophyte
itself) and sporophytic (effects expressed in maternal
plant tissues) (for an overview and references, see Drews et al.,
1998; Grossniklaus and Schneitz, 1998; Chaudhury and Berger,
2001). A special group of Arabidopsis female-gametophytic
maternal genes that affect endosperm (and embryo) development
that have received considerable attention in recent years
include Mea/Fis1 (Medea/Fertilization-independent seed1),
Fis/Fis2, and Fie (Fertilization-independent endosperm)/Fis3
(Ohad et al., 1996; Chaudhury et al., 1997; Drews et al., 1998;
Grossniklaus and Schneitz, 1998; Grossniklaus et al., 1998;
Chaudhury and Berger, 2001). The products of these genes
are similar to Polycomb group proteins, which regulate the
expression of genes through epigenetic silencing in Drosophila
1 To whom correspondence should be addressed. E-mail odd-arne.
olsen@pioneer.com; fax 515-254-2619.
Article, publication date, and citation information can be found at
www.plantcell.org/cgi/doi/10.1105/tpc.017111.
The Plant Cell, Vol. 16, S214­S227, Supplement 2004, www.plantcell.org  2004 American Society of Plant Biologists
melanogaster (Pirrotta, 1998). During early embryogenesis of D.
melanogaster, these genes play a role in maintaining established
patterns of gene expression, and it is likely that they play a similar
role in the endosperm. Recently, the Polycomb group protein
Medea was shown to regulate seed development by controlling
the expression of the MADS box gene Pheres1 (Kohler et al.,
2003). For a detailed discussion, see review by Gehring et al.
(2004) in this volume.
THE EVOLUTIONARY ORIGIN OF NUCLEAR ENDOSPERM
Double fertilization, the process by which one male gamete
fertilizes the egg cell to produce a diploid zygote and the second
gamete fertilizes the diploid central cell to give rise to the triploid
endosperm (Figures 1D and 1E), was discovered more than
100 years ago (Nawaschin, 1898; Guignard, 1901). Since then, it
has been regarded as a defining characteristic of angiosperms
and has set the stage for the debate on the evolutionary origin
of endosperm (Friedman, 1995). Soon after its discovery, two
proposals for the origin of the endosperm were advanced, both
of which are still valid because the issue remains unresolved
(reviewed by Friedman, 1998). The first hypothesis proposes that
the endosperm represents an altruistic, modified second embryo
that during angiosperm evolution provided a selective advantage
by nurturing the embryo proper. The second hypothesis states
that the endosperm represents extended development of the
megagametophyte (i.e., the central cell). Until recently, several
lines of evidence suggested that the Gnetales represent the
extant seed plants most closely related to angiosperms. Based
on this assumption, and the observation that double fertilization
results in a diploid supernumerary embryo in Ephedra and
Gnetum species, Friedman and co-workers supported the twinembryo
hypothesis (Carmichael and Friedman, 1995). However,
recent discoveries in angiosperm phylogeny do not support
Gnetales as an angiosperm predecessor (for details, see
Mathews and Donoghue, 1999; Qiu et al., 1999; Soltis et al.,
1999; reviewed by Friedman and Williams, 2003), lending little
support to the twin-embryo hypothesis.
The revised angiosperm phylogeny has motivated a reexamination
of the basic questions in the debate on the origin of the
angiosperm endosperm, including the structure of the ancestral
angiosperm megagametophyte, the role of double fertilization in
angiosperms, and whether the ancestral endosperm was cellular
or nuclear. Based in part on a reexamination of basal endosperm
gametophytes according to the new angiosperm phylogeny,
Williams and Friedman (2002) resurrect an idea proposed by
Swamy (1946) among others that a four-celled embryo sac
reflects the minimal structure common to all angiosperms and
that the Polygonum embryo sac evolved from a duplication of
this basic module gametophyte (Baroux et al., 2002; Friedman
and Williams, 2003). In light of this information, Friedman and
Williams (2004) favor the second hypothesis, that the triploid
endosperm evolved from a maternal (megagametophyte) endosperm
to a biparental tissue by the addition of the male nucleus at
a later stage in evolution (see also Friedman and Floyd, 2001;
Williams and Friedman, 2002). Finally, based on phylogenetic
studies, the ancestral endosperm is believed to have been
cellular and the nuclear endosperm is believed to have arisen
multiple times, including in the angiosperm lineages of the
cereals and Arabidopsis (reviewed by Geeta, 2003).
In spite of recent progress in understanding angiosperm
phylogeny, all of the main questions regarding the evolutionary
history of the nuclear endosperm remain unresolved. As summarized
above, one likely scenario is that the ancestral endosperm
of the cereals and Arabidopsis evolved from a maternal
cellular megagametophyte tissue to a biparental endosperm
by the addition of a male (pollen) nucleus at a later stage.
Furthermore, the nuclear endosperm of cereals and Arabidopsis
are not homologs in an evolutionary sense and consequently
must have evolved independently.
THE ENDOSPERM COENOCYTE OF CEREALS
The morphogenetic events of the early stages of endosperm
development in cereals were first detailed in a comprehensive
way in barley using confocal microscopy (Brown et al., 1994) and
later confirmed in rice (Brown et al., 1996a, 1996b). The first
division of the triploid endosperm nucleus in the central cell of
barley reveals the hallmark of nuclear endosperm, namely, the
absence of a cell plate between separating daughter nuclei
(Figures 2A and 2B). The continued absence of cell wall formation
in the ensuing mitotic divisions leads to a multinucleate cell (the
endosperm coenocyte) (Figures 2C and 2D) and stands in
Figure 1. Development and Structure of the Polygonum-Type Embryo
Sac of Maize and Arabidopsis.
(A) The embryo sac develops from the surviving haploid megaspore
resulting from female meiosis.
(B) After three rounds of mitosis without cell wall formation between
sister nuclei, two groups of four nuclei form at the micropylar (mp) and
chalazal (cz) poles of the young megagametophyte.
(C) Three of the nuclei at each pole are walled off and become the egg
cell (e), the synergids (sy), and the antipodal cells (ap). One nucleus from
each pole migrates to the center (polar nuclei; pn) of the megagameto-
phyte.
(D) The haploid polar nuclei of maize remain separate until fertilization.
The shaded area of the megagametophyte represents the central cell
that develops into the endosperm after fertilization. One of the haploid
male nuclei (m) from the pollen tube (pt) enters the central cell, and the
three nuclei fuse to form the triploid primary endosperm nucleus.
(E) The two polar nuclei of the Arabidopsis central cell fuse before
fertilization. The antipodal cells are eliminated by programmed cell death.
Figure modified from Drews et al. (1998).
Nuclear Endosperm S215
contrast to somatic cells, in which the default mitotic division
cycle includes the formation of an interzonal phragmoplast
between separating sister nuclei. For an overview of the cytoskeletal
components of somatic cells relevant to the discussion
of endosperm development, see Figure 3. The interzonal phragmoplast
of somatic cells consists of two circular arrays of
microtubules of opposing polarity that transport Golgi-derived
vesicles containing glucan polymers to the site of cell wall
deposition (Figures 3E and 3F). The phragmoplast and its cell
plate expand laterally until the cell plate fuses with the parental
plasma membrane and cell wall (Staehelin and Hepler, 1996;
Heese et al., 1998; Sylvester, 2000; Brown and Lemmon, 2001)
(Figures 3G and 3H). The molecular basis for the lack of cell wall
formation in nuclear endosperm is unknown. Interestingly, in
barley, phragmoplast formation is initiated between dividing
sister nuclei (Brown et al., 1994), even forming occasional
rudimentary cell walls in wheat (Tian et al., 1998). These observations
suggest that interzonal phragmoplast function is suppressed
after initiation in the endosperm coenocyte and appear
to support the conclusion that cellular endosperm represents the
basal state for endosperm. Although many details still are lacking
about the regulation of phragmoplast formation and expansion,
recent data suggest that the cytoskeletal apparatus is controlled
by Cdc2-like kinases and mitotic cyclins (reviewed by Calderini
et al., 1998; Pickett-Heaps et al., 1999; Smith, 1999; Sato
and Kawashima, 2001). The most direct evidence for a role of
mitogen-activated protein kinases in phragmoplast formation
comes from tobacco, in which the mitogen-activated protein
kinase kinase kinase NPK1 interacts with a phragmoplastlocalized
kinesin-like protein (Machida et al., 1998). These and
other proteins should be used as probes to identify and compare
the mechanisms for phragmoplast suppression in nuclear
endosperm from different angiosperm lineages.
In maize, the first three mitotic divisions of the endosperm
nuclei occur in predictable planes, resulting in eight nuclei
positioned in a single plane in the basal cytoplasm of the
coenocyte (Figure 2C). From this position, each nucleus divides
and daughter nuclei migrate and divide, producing a population
of nuclei that spreads to a sector corresponding to one-eighth of
the coenocyte surface (Coe, 1978; McClintock, 1984). In maize,
continued division produces 256 to 512 nuclei, marking the end
of the coenocytic stage (Walbot, 1994) (Figure 2D). In barley, this
stage is reached at 3 days after pollination (DAP), and the nuclei
enter a mitotic hiatus that lasts for 2 days. The molecular
control mechanism that causes the arrest in the progression of
the cell cycle in barley is unknown.
Table 1. Definitions of Terms
Term Explanation Figure
Adventitious phragmoplast Phragmoplast formed between canopies of microtubules in alveoli toward the
central vacuole
5D
Alveolus (plural, alveoli) Tube-shaped structure consisting of cell wall encasing one nucleus with one open
end facing the central vacuole
5C
Anticlinal division Mitotic division leading to a new cell wall that is perpendicular to the central cell wall 5B
Central cell Central compartment of the megagametophyte with two nuclei (2n) after the formation
of cell walls around the nuclei at each pole
1C
Chalazal pole Distant pole of the megagametophyte 1B
Coenocyte A cell with multiple nuclei in the same cytoplasm 2D
Cytoplasmic phragmoplast Phragmoplast formed between opposing arrays of radial microtubules from
neighboring coenocyte nuclei
5B
CZE The endosperm in the chalazal end of the seed 2F
Embryo sac (megagametophyte) Structure resulting from female meiosis containing the central cell and the egg cell 1C
ESR Embryo-surrounding region 7A
Hoop-like cortical array Array of circular microtubules close to the cell surface 3A
Interzonal phragmoplast The phragmoplast formed between separating sister nuclei in somatic cells 3E and 3F
MCE Endosperm in the micropylar end of the seed 2F
Megagametophyte (embryo sac) Structure resulting from female meiosis containing the central cell and the egg cell 1C
Micropylar pole The pole of the embryo sac where the pollen tube enters and where the egg
cell is located
1B
Miniphragmoplast Substructure of microtubules forming the cytoplasmic phragmoplast 5M
NCD Nuclear cytoplasmic domain, a portion of the cytoplasm around one nucleus
claimed by the radial microtubule system of that nucleus
4A
PEN Peripheral endosperm in the central chamber of the seed 2F
Periclinal division Mitotic division leading to a new wall that is parallel to the central cell wall 4C
Phragmoplast Array of microtubules with opposite polarity mediating the deposition of a new
cell wall between nuclei
3E and 3F
Polar nuclei Haploid nuclei that migrate from the two poles of the megagametophyte to the
center of the central cell
1C
PPB Preprophase band of microtubules marking the future plane of cell division 3B
RMS Radial microtubule system emanating from the surface of endosperm nuclei 4A
The figure citations are to the first mentions of each term in the text.
S216 The Plant Cell
THE ENDOSPERM COENOCYTE OF ARABIDOPSIS
Dicot nuclear endosperm development has been studied in
several Brassicaceae species, including Arabidopsis (Mansfield
and Briarty, 1990a, 1990b; Webb and Gunning, 1991; Mansfield,
1994; Van Lammeren et al., 1996; Brown et al., 1999). The central
cell containing the endosperm coenocyte of Arabidopsis
has three regions that become distinct as the seed grows: the
embryo-surrounding region or micropylar endosperm (MCE),
the peripheral endosperm (PEN) in the central chamber, and
the chalazal endosperm (CZE) (Brown et al., 1999; BoisnardLorig
et al., 2001; Sorensen et al., 2002) (Figures 2E and 2F). As
the embryo sac expands after fertilization, the central vacuole
enlarges and the cytoplasm of the endosperm syncytium
assumes a peripheral position. At the globular embryo stage,
the syncytial cytoplasm of the MCE surrounds the developing
embryo, and the multinucleate PEN syncytium is a thin peripheral
layer with evenly spaced nuclei (Figure 2H). Using a green
fluorescent protein marker that accumulates in cell plates of
somatic cells, Sorensen and co-workers (2002) detected cell
plate formation between separating sister nuclei in PEN at a low
frequency, suggesting that interzonal phragmoplasts were
sometimes functional. This observation is interesting in light of
the formation of nonfunctional phragmoplasts between sister
nuclei in barley and wheat, suggesting that the coenocytic stage
in Arabidopsis also results from a suppression of phragmoplast
function. Recently, Brown and colleagues (2003) also observed
interzonal phragmoplast formation between sister nuclei in the
early endosperm coenocyte, suggesting that phragmoplast suppression
occurs in a manner similar to that in cereals.
Berger and co-workers divided the syncytial endosperm stage
in Arabidopsis into nine substages, each stage representing one
of the eight rounds of mitosis (Boisnard-Lorig et al., 2001). At the
final stage, the syncytial endosperm contained 200 nuclei. After
the initial three synchronous division cycles, the mitotic activity
of MCE, PEN, and CZE occurred independently, with nuclei
dividing synchronously within domains. Nuclear divisions were
never observed directly in the CZE after the three synchronous
rounds of division, suggesting that these nuclei undergo endoreduplication
(Boisnard-Lorig et al., 2001).
Molecular markers for different endosperm compartments
include the Fis1/Mea and Fis/Fis2 promoters fused to
b-glucuronidase (GUS), which represent specific markers of
early nuclear endosperm development. GUS activity from these
constructs can be detected already in the polar nuclei, the
central cell nucleus of unpollinated ovules, and the syncytial
endosperm (Luo et al., 2000). After cellularization, activity ceases
in the micropylar and peripheral endosperm, being restricted to
the chalazal chamber. The green fluorescent protein reporter of
the enhancer trap line KS117 is expressed in the chalazal cyst
at the heart stage of embryo development, but not in the
endosperm nodule (Sorensen et al., 2001). Recently, a set of
novel marker lines including markers for the chalazal endosperm
as well as the MCE was reported (Stangeland et al., 2003).
ENDOSPERM CELLULARIZATION IN CEREALS
The process of cellularization of the endosperm coenocyte is
initiated by the formation of RMS on all nuclear surfaces (Brown
et al., 1994) (Figures 4A, 5A, and 5E). The portion of the
cytoplasm claimed by these arrays around each nucleus is
referred to as a nuclear cytoplasmic domain (NCD) (Brown and
Lemmon, 1992). Soon, the microtubules from neighboring nuclei
meet, forming interzones in which wall material consisting mainly
of callose is deposited (Figure 5B). In barley, the arrays of
opposing microtubules from adjacent NCDs are termed cytoplasmic
phragmoplasts; these mediate the deposition of the
Figure 2. The Endosperm Coenocyte of Cereals and Arabidopsis.
(A) to (D) Cereals.
(E) to (H) Arabidopsis.
(A) The triploid endosperm nucleus (en) is located in the basal cytoplasm
of the central cell. A large central vacuole (cv) fills up most of the volume,
surrounded by a thin line of cytoplasm (cy).
(B) The central cell nucleus divides without the formation of a functional
interzonal phragmoplast, and no cell wall is formed between sister nuclei.
(C) After three rounds of nuclear divisions, eight endosperm nuclei are
located in a single plane in the basal endosperm coenocyte.
(D) The complete endosperm coenocyte contains evenly spaced nuclei
in the entire peripheral cytoplasm.
(E) The Arabidopsis endosperm coenocyte has nuclei migrating from the
micropylar region (mp) toward the chalazal end (cz), eventually covering
the entire periphery of the coenocyte.
(F) and (G) As development progresses, the endosperm coenocyte
develops three distinct regions: the region surrounding the embryo
(MCE), the central or peripheral endosperm (PEN), and the region of the
chalazal endosperm (CZE), which contains the chalazal cyst (cz).
(H) At the end of the globular embryo stage, the embryo becomes
completely surrounded by cytoplasm.
Figure 3. The Cytoskeletal Cycle of Somatic Cells.
(A) Hoop-like cortical arrays in interphase.
(B) Preprophase band marking the site and orientation of the deposition
of the future cell wall.
(C) Metaphase spindle separating the sister chromosomes (not shown).
(D) Shortened spindles in telophase with connecting microtubules between
the two poles.
(E) Early phragmoplast in the interzone between the two poles.
(F) Complete interzonal phragmoplast.
(G) Expanding phragmoplast depositing the new cell plate (cw) separating
sister nuclei.
(H) Complete cell wall separating the two sister nuclei (cells).
For details, see Staehelin and Hepler (1996), Heese et al. (1998),
Sylvester (2000), and Brown and Lemmon (2001).
Nuclear Endosperm S217
initial cell walls in the endosperm (Brown et al., 1994). The
formation of walls by cytoplasmic phragmoplasts is not unique to
endosperm but also is seen in other plant systems, including
megagametophyte cellularization, sporogenesis in lower plants,
male and female gametophyte development in gymnosperms
and angiosperms, and embryogenesis in gymnosperms (reviewed
by Brown and Lemmon, 2001). Initially, the endosperm
cell walls deposited by the cytoplasmic phragmoplasts form
a tube-like structure, or alveolus, around each nucleus, with the
open end pointing toward the central vacuole (Figures 4B, 5C,
and 5G). In the process that follows, the RMS that encase each
nucleus undergo reorganization, anchoring the nuclei to the
central cell wall while extending toward the central vacuole in
a canopy of microtubules (Figures 5D and 5H). The interzones
between adjacent canopies of microtubules, termed adventitious
phragmoplasts (Brown et al., 1994), function to extend the
alveoli further toward the central vacuole. At the end of the first
round of alveoli formation, the nuclei in each alveolus exit mitotic
arrest synchronously to divide in a periclinal division plane (the
orientation of the new cell wall is parallel with the central cell wall)
(Figures 6A and 6B). Notably, the periclinal cell walls in these
mitotic divisions are formed by functional interzonal phragmoplasts
that appear for the first time in the cell cycle of the
endosperm (Figure 6B). These periclinal cell walls divide the
alveoli into a peripheral cell and a new alveolus with its opening
toward the central vacuole (Figures 4C and 6C). The repetition of
this process four or five times results in a completely cellular
endosperm at 6 to 8 DAP in barley and at 4 DAP in maize, wheat,
and rice (Figures 4D and 6E).
As in other plant microtubule systems, the molecular basis for
RMS formation in nuclear endosperm is unknown (reviewed
by Canaday et al., 2000). In contrast to animal cells, in which
centrosomes function as microtubule-organizing centers, the
site of the initiation of microtubule polymerization in plants is
unknown. One possibility is that microtubule polymerization in
plants is initiated on nuclear surfaces and that these microtubule
precursors are transported subsequently to their final subcellularFigure 4. Cellularization of the Endosperm Coenocyte in Cereals and
Arabidopsis.
(A) to (D) Cereals.
(E) to (G) Arabidopsis.
(A) RMS form on nuclear membranes in the cereal endosperm
coenocyte. ccw, central cell wall; cv, central vacuole.
(B) Anticlinal cell walls (acw) form tubes or alveoli (alv) around each
nucleus with their open ends toward the central vacuole. For details, see
Figure 5.
(C) Divisions of alveolar nuclei result in a periclinal cell wall (pcw) that
separates the outer layer of cells from a new layer of alveoli.
(D) Repeated periclinal divisions in the innermost layer of alveoli continue
until the endosperm is completely cellular.
(E) The endosperm of Arabidopsis at the globular embryo stage showing
a cellular MCE, a gradient of stages in the alveolation process in PEN,
and endosperm nodules (no) as well as chalazal cyst (cz) formation
in CZE.
(F) Completely cellular endosperm (ce).
(G) The endosperm is consumed during seed maturation, leaving only
the peripheral aleurone-like cell (alc) layer in a mature embryo (me).
Figure modified from Olsen (2001), Brown et al. (1999), and BoisnardLorig
et al. (2001).
Figure 5. Initial Cell Wall Formation (Alveolation) by Cytoplasmic
Phragmoplasts in Cereals and Arabidopsis.
(A) to (D) Cytoplasmic phragmoplasts.
(E) to (H) Cereals.
(I) to (M) Arabidopsis.
(A) Diagram showing RMS on two adjacent cereal endosperm nuclei. For
orientation, see Figure 4A. Ccw, central cell wall.
(B) Cytoplasmic phragmoplasts form in the interzones between
opposing RMS, mediating cell plate deposition (arrow).
(C) Alveoli form around each nucleus by cytoplasmic phragmoplasts.
(D) Alveoli are extended toward the central vacuole by a canopy of
microtubules in a canopy-like fan of microtubules that form adventitious
phragmoplasts (arrow). The nuclei are anchored to the former central cell
wall by microtubules.
(E) to (H) Micrographs from barley depicting the stages diagrammed
above each image.
(I) to (L) Micrographs from Arabidopsis depicting the stages diagrammed
above each image.
(M) Diagram illustrating that cytoplasmic phragmoplasts are composed
of substructures termed miniphragmoplasts (mp).
(A) to (D) are modified from Olsen (2001), and (H) is redrawn from Otegui
and Staehelin (2000).
S218 The Plant Cell
destinations. If this is correct, NCDs could be generated by
a block in the mechanism(s) that transports microtubules from
nuclear surfaces, combined with continued polymerization to
grow microtubules to extend the whole radius of NCDs. A better
understanding of the mechanisms of microtubule formation in
plants should contribute to an understanding of RMS formation
and its regulation.
ENDOSPERM CELLULARIZATION IN ARABIDOPSIS
Similar to the cereal endosperm, cellularization of the Arabidopsis
coenocyte occurs via formation of RMS and alveolation
(Brown et al., 1999; Boisnard-Lorig et al., 2001; Sorensen
et al., 2002). The cellularization process starts as a wave in
MCE, progressing through PEN and CZE at different rates and
with significant variations between the different chambers
(Figure 4E). The process of cellularization in PEN is similar,
if not identical, to the cellularization process in cereals, and
representative images of the different stages in barley and
Arabidopsis are shown in Figures 5E to 5L. As in barley, RMS
form on nuclear surfaces (Figures 4E and 5I), subsequently
forming cytoplasmic phragmoplasts in NCD interzones that
mediate alveolar cell wall formation (Figures 5J and 5K).
Alveolization initiates at the final round of syncytial mitosis
(Sorensen et al., 2002). The leading edge of cytoplasm contains
the adventitious phragmoplast that extends the alveoli inward
(Figure 5L). Synchronous periclinal mitosis of alveolar nuclei,
accompanied by the formation of interzonal phragmoplasts and
periclinal cell wall deposition (data not shown), divides the PEN
alveoli into a peripheral cell and an internal alveolus (Figure 4E).
Completion of endosperm cellularization in Arabidopsis also
occurs by repeated rounds of the RMS-alveolation cycle, leading
to a cellular endosperm except for the specialized chalazal
endosperm cyst (Figure 4F).
The cellularization process in MCE also occurs via RMS and
cytoplasmic phragmoplasts, but typical alveoli do not form as
a result of spatial constraints in this chamber of the central cell
(Figures 2H and 4E). Using labeled cryofixed/freeze-substituted
material and high-resolution electron tomography, Staehelin and
co-workers (Otegui and Staehelin, 2000; Otegui et al., 2001) have
provided a detailed description of cytoplasmic cell plate
formation, showing that cytoplasmic phragmoplasts originate
at NCD boundaries and consist of substructures that they termed
miniphragmoplasts (Figure 5M). On average, six cytoplasmic or
syncytial-type cell plates form one hexagon-shaped alveolus. A
model for the stepwise formation of the cytoplasmic cell plate
was proposed by Otegui et al. (2001). Cellularization in MCE is
completed around the embryo, whereas PEN remains syncytial
(Sorensen et al., 2002) (Figure 4E). The CZE remains syncytial
until late stages of seed maturation (Figure 4F).
How similar are the molecular mechanisms involved in
cytoplasmic and interzonal phragmoplast-mediated cell wall
formation? Structural similarities between the two types of
phragmoplasts include the behavior of the tubular Golgi-derived
networks, the appearance of clathrin-coated buds and vesicles,
and callose deposition (Brown et al., 1997; Otegui et al., 2001). In
addition, Knolle, a protein involved in homotypic fusions of Golgiderived
vesicles, has been shown to play a role both in
endosperm cellularization and in cell division in the embryo. In
both systems, homozygous mutant knolle cells fail to form
complete cell walls (Lukowitz et al., 1996; Lauber et al., 1997;
Sorensen et al., 2002). Several other mutants with knolle-like
embryo phenotypes also show an arrest or delay of endosperm
cellularization at the syncytial stage, including hinkel, which
encodes a kinesin-related protein, open house, runkel, and
pleiade (gene products unknown) (Sorensen et al., 2002;
Strompen et al., 2002). As expected, Arabidopsis mutations that
affect microtubule formation or behavior in embryos also affect
Figure 6. Completion of the Endosperm Cellularization Process and Differentiation in Cereals.
(A) Nuclei in alveoli (see Figure 4B) are surrounded by microtubular networks indicating entry into mitosis. The former central cell wall (ccw) represents
the baseline in all panels.
(B) and (C) For the first time in the endosperm life cycle, functional interzonal phragmoplasts (arrows) appear between separating sister nuclei, resulting
in a periclinal cell wall that divides the alveolus into a peripheral cell (B) and a new layer of alveoli (C). For orientation, see Figure 4C.
(D) Repetitions of the alveolation process result in cell files growing in from the periphery that meet in the middle of the former central vacuole (see
Figure 4D).
(E) After completion of cellularization, the starchy endosperm cell precursor cells (all except the peripheral layer of aleurone cell initials) divide at random
planes.
(F) The cell file pattern in the starchy endosperm is lost soon after several rounds of cell division.
(G) Diagram showing hoop-like cortical arrays in aleurone cells, which represent the first layer of cells formed in the endosperm.
(H) Preprophase bands in an aleurone cell.
Nuclear Endosperm S219
endosperm development, including titan1 and titan5 (McElver
et al., 2000; Steinborn et al., 2002; Tzafrir et al., 2002), titan7, and
titan8 (Liu et al., 2002).
In spite of these striking similarities between cytoplasmic and
interzonal phragmoplasts, functional differences, such as the
mechanism of fusion of the cell plate to form vesicles and the
mode of marginal cell plate growth, have been identified (Otegui
and Staehelin, 2000; Otegui et al., 2001). Compositional differences
between cell plates formed by the two types of
phragmoplasts include a lack of terminal fucose residues in
xyloglucans and the permanent presence of callose in syncytial
cell plates (Liu et al., 2002), possibly making the endosperm walls
more suitable for the storage of polysaccharides (Otegui et al.,
2001). Genetic evidence for differences between somatic and
cytoplasmic cell plates also has been provided in Arabidopsis. In
the spaatzle mutant, the embryo develops normally but PEN
cellularization is perturbed, suggesting that Spaatzle encodes an
endosperm cellularization­specific component (Sorensen et al.,
2002). In spaatzle endosperm, PEN contains regularly organized
NCDs until the initiation of cellularization, when the nuclei
undergo at least one additional mitotic division, resulting in
a syncytium with increased NCD density. Subsequently, the
number of NCDs in the PEN is reduced continuously while the
size of individual nuclei increases, as does the fusion of NCDs,
ultimately resulting in a few giant NCDs with one or more giant
nuclei. Identification of the Spaatzle gene product should
contribute important insight regarding the mechanisms of NCD
formation and cytoplasmic cell plate formation.
ENDOSPERM CELL FATE SPECIFICATION AND
DIFFERENTIATION IN CEREALS
The fully developed cereal endosperm consists of four main cell
types: the starchy endosperm, the aleurone layer, transfer cells,
and cells of the embryo-surrounding region (Figures 7A and 7B).
The cereal endosperm has attracted attention from researchers
because of its economic importance, and much insight has
accumulated about the genes underlying the accumulation of
storage products such as proteins and starch. Considerably less
is known about the genes that regulate the developmental
biology of these cell types, which is the topic of this section. Cell
fate specification in cereal endosperm is believed to occur by
positional signaling at an early developmental stage (Olsen,
2001) (Figures 7C to 7E). For simplicity, each cell type is described
separately below, although cell fate specification occurs
simultaneously with the cellularization process described above.
How this integration occurs is unknown, but elucidation of the
molecular controls for each of the four cell types should lay
the foundation for understanding the genetic specification of the
entire endosperm body plan.
The Embryo-Surrounding Region
The embryo-surrounding region (ESR) lines the cavity of the
endosperm in which the embryo develops and has been studied
most extensively in maize (Figure 7A). The exact role of the ESR
is unknown, but possible functions include a role in embryo
nutrition, the establishment of a physical barrier between the
embryo and the endosperm during seed development, and
providing a zone for communication between the embryo and the
endosperm. In maize, ESR cells are characterized by their dense
cytoplasmic contents (Schel et al., 1984; Kowles and Phillips,
1988) and by the expression of the Esr1, Esr2, Esr3 (OpsahlFerstad
et al., 1997), ZmAE1 (Zea mays androgenic embryo1),
and ZmAE3 (Magnard et al., 2000) genes between 5 and 20 DAP.
Transgenic maize lines expressing the GUS reporter under the
control of Esr promoters confirm the ESR-preferred pattern of
expression for these genes (Bonello et al., 2000). Esr proteins
localize to ESR cell walls (Bonello et al., 2000, 2002). Esr3
belongs to a family of small hydrophilic proteins that share
a conserved motif with Clavata3 (Clv3), a protein that has been
reported to interacts with the receptor-like kinases Clv1 and Clv2
in Arabidopsis and that functions in regulating meristem size
(Fletcher et al., 1999). The functional significance of this similarity
is strengthened by the observation that it is limited to a highly
conserved region of 15 amino acids in Clv3 that contains two
point mutations that result in the clv3 phenotype (Fletcher et al.,
1999), although the proposed role for Clv3 as a ligand for clv1 has
not been proven (Nishihama et al., 2003). This region of 15 amino
acid residues also is shared by >40 predicted proteins called Cle
(Clv3/Esr-related) that are all small hydrophilic proteins with
a signal peptide (Cock and McCormick, 2001). Future studies will
show whether, and how, ESR proteins may be involved in ESR
signaling. In spontaneously occurring embryoless endosperm,
Esr expression is lacking, suggesting a dependence of Esr
Figure 7. Cell Fate Specification and Development in Wild-Type and
Mutant Cereal Endosperm.
(A) Cell types of maize endosperm: starchy endosperm (se), aleurone
layer (al), transfer cells (tc), and ESR cells. e, embryo.
(B) Cell types of the barley endosperm.
(C) to (E) Proposed developmental domains in the young cereal
endosperm at the syncytial stage (C), the alveolar stage (D), and the
complete cellular stage (E). Color coding is the same as in (A) and (B). cv,
central vacuole.
(F) Peripheral section of wild-type maize endosperm with one layer of
aleurone cells. The arrow points to the maternal cell layer adjacent to the
aleurone cells. P, maternal tissues.
(G) cr4 endosperm lacks aleurone cells in discrete areas.
(H) dek1 endosperm with a complete lack of aleurone cells. The arrow
points to the corresponding layer shown in (F).
(I) sal1 endosperm with up to seven layers of aleurone cells.
S220 The Plant Cell
expression on signaling from the embryo (Opsahl-Ferstad et al.,
1997). The promoters of Esr genes should provide useful tools to
investigate the underlying mechanism of transcriptional activation
in these genes (Bonello et al., 2000).
The mechanism underlying cell fate specification of the ESR
is unknown. Based on the observation in maize that cell walls
appear to form in the endosperm coenocyte around the embryo
during the coenocytic stage (R.C. Brown, B.E. Lemmon, and
O.-A. Olsen, unpublished data), it is possible that the ESR forms
through a mechanism that permits functional phragmoplasts to
form near the embryo. Also in barley, cellularization occurs early
in the immediate vicinity of the embryo (Engell, 1989). Further
studies are needed to confirm whether or not these cells
represent ESR precursor cells. The observation that the
endosperm of embryoless mutant grains forms a normal-sized
embryo cavity suggests that the endosperm has an intrinsic
program for the formation of the ESR domain (Heckel et al., 1999).
Transfer Cells
Transfer cells develop in the basal endosperm over the main
vascular tissue of the maternal plant (Figures 7A and 7B), where
they facilitate solute transfer, mainly of amino acids, sucrose, and
monosaccharides, across the plasmalemma between the symplastic
(maternal plant) and apoplastic (endosperm) compartments
(Thompson et al., 2001). In maize, the miniature1 mutant
has reduced grain size and lacks normal levels of type 2 cell wall
invertase in transfer cells, strongly suggesting that invertase
contributes to the establishment of a sucrose concentration
gradient in the apoplastic gap between the pedicel and the
endosperm by hydrolyzing sucrose to glucose and fructose
(Miller and Chourey, 1992; Cheng et al., 1996). In maize, two to
three layers of transfer cells have wall ingrowths in a gradient
decreasing toward the interior of the endosperm (Schel et al.,
1984; Gao et al., 1998; Thompson et al., 2001). Several groups of
transcripts have been shown to be expressed preferentially in
transfer cells, including maize Betl1 (Basal endosperm transfer
cell layer1) (Hueros et al., 1995), Betl2, Betl3, and Betl4 (Hueros
et al., 1999), Bap1 (Basal layer-type antifungal protein1), Bap2,
and Bap3 (Serna et al., 2001). Many of these proteins are similar
to antimicrobial proteins, suggesting a role in defense against
invading pathogens. Betl1 is synthesized and located in basal
endosperm cells, where it is tightly bound to the cell wall (Hueros
et al., 1995), whereas Bap2 is secreted into the intercellular matrix
of the basal endosperm and accumulates predominantly in the
adjacent, thick-walled cell layer of the pedicel (Serna et al., 2001).
Genes expressed at early developmental stages in transfer
cells are of special interest as developmental markers for
investigating the mechanisms of transfer cell fate specification.
In barley, Endosperm1 (End1) is present in the basal transfer cell
domain of the endosperm coenocyte, which gives rise to the cells
that differentiate into transfer cells (Doan et al., 1996) (Figure 7C).
The function of this transcript has yet to be determined. A similar
pattern of expression is seen for the maize Zea mays MYB-related
protein-1 (ZmMRP-1) transcript, which encodes a single Mybrepeat
protein (Gó mez et al., 2002). Interestingly, ZmMRP-1
expression precedes that of other Betl-specific genes and has
been shown to activate Betl transcription in transient assays
(Gó mez et al., 2002). Although the mechanisms for early transfer
cell domain transcription are unknown, differential transcription
of End1 and ZmMRP-1 in the nuclei of this region of the
coenocyte is a plausible explanation (Gó mez et al., 2002). Such
differential transcription could be triggered by either maternal
factors deposited in the basal region of the central cell before
fertilization or maternal factors from the pedicel in developing
grains. The observations that xylem transfer cells are induced by
high CO2 concentration in lettuce and that transfer cells are
induced in Vicia faba (broad bean) cotyledons as a response to
glucose and fructose, but not by sucrose (reviewed by Thompson
et al., 2001), make maternal factors from the pedicel attractive
candidates for transfer cell gene-specific activators. During the
cellularization process, two to three basal cells in cell files derived
from the transfer cell domain of the endosperm coenocyte
assume transfer cell identity (Figure 7E). This is different from the
aleurone layer, where the border between the single layer of
aleurone cells and the starchy endosperm is sharply defined (see
below for details), suggesting that different mechanisms are
involved in specifying the two cell types. One mechanism that
could explain the gradient of transfer cell morphology in the basal
endosperm is a dilution of transcription factor(s) present in the
transfer cell region of the endosperm coenocyte as the cell files
form and grow toward the central vacuole. Interestingly, kernels
of the defective kernel1 (dek1) mutant lack aleurone cells but
contain normal transfer cells (Lid et al., 2002), supporting the
notion that aleurone and transfer cell fates are specified by
different mechanisms. Recently, based on the phenotype of the
globby1-1 mutant in maize, Costa and co-workers (2003)
proposed that specification of transfer cells is an irreversible
event that occurs during syncytial development and that transfer
cell fate is inherited in a cell lineage­dependent manner. In
addition to mutants that are impaired in transfer cell development
(Maitz et al., 2000), mutants that lack transfer cells would be
invaluable in elucidating the mechanisms underlying transfer cell
fate specification.
Starchy Endosperm
Starchy endosperm cells represent the largest body of cells in
the endosperm (Figures 7A and 7B). Starchy endosperm cells
accumulate starch and prolamin storage proteins encoded by
transcripts that are expressed differentially in these cells. Starchy
endosperm cells are derived from two sources. The first, and
most important, is the inner cells of cell files that are present
at the completion of endosperm cellularization (Figure 7E, red
zone). Soon after the completion of the cellularization phase, cell
division resumes in the inner cell files (Figure 6E). Similar to the
first periclinal divisions in alveoli, preprophase bands (PPBs) are
absent, but unlike the alveolar divisions, which are strictly
periclinal, the division planes are oriented randomly and the cell
file pattern is soon lost (Figure 6F). The second source of starchy
endosperm cells is the inner daughter cells of aleurone cells that
divide periclinally (Figure 7E, blue zone). These cells redifferentiate
to become starchy endosperm cells and likely are the
source of the so-called subaleurone cells found adjacent to the
aleurone layer in the starchy endosperm in all cereals.
Nuclear Endosperm S221
Several collections of chemically induced mutants have led to
the isolation of mutants broadly referred to as dek (defective
kernel) in both maize and barley (Neuffer and Sheridan, 1980;
Felker et al., 1985, 1987; Bosnes et al., 1987; Scanlon et al.,
1994b). More recently, collections of maize mutant genes have
been created in which the Mutator (Mu) transposon facilitates the
identification and cloning of the mutant genes (Bensen et al.,
1995). In the majority of these mutants, all tissues form normally,
but the degree of filling of the starchy endosperm is reduced
severely (Lid et al., 2002). Two such maize mutant genes have
been cloned, dsc1 (discolored1) (Scanlon and Myers, 1998) and
emp2 (empty pericarp2) (Fu et al., 2002). The Dsc1 mRNA is
detected specifically in kernels at 5 to 7 DAP, but no function has
been assigned to the cloned genomic region. emp2 is an embryolethal
dek mutant that encodes a predicted protein with high
similarity to Heat-shock binding protein1 (Fu et al., 2002). In
addition to a predicted role in the heat-shock response, the mutant
phenotype suggests that Emp2 also performs an important
function(s) in seed development that has yet to be identified.
Pending the isolation of mutants that specifically affect starchy
endosperm cell fate specification, little progress has been made
in understanding the underlying mechanism of cell fate specification.
It is interesting that in maize mutants that lack aleurone
cells, crinkly4 (cr4) (Becraft et al., 1996) and dek1 (Becraft and
Asuncion-Crabb, 2000; Lid et al., 2002) starchy endosperm cells
are formed in place of aleurone cells (Figures 7G and 7H). Thus,
signaling leading to aleurone cell formation appears to override
signaling leading to starchy endosperm cell formation. Two
important aspects of starchy endosperm development that are
not discussed here are endoreduplication (reviewed by Larkins
et al., 2001) and programmed cell death (Young et al., 1997).
Aleurone Cells
The aleurone layer covers the perimeter of the endosperm with
the exception of the transfer cell region (Figures 7A and 7B). The
aleurone layer functions in seed germination by mobilizing starch
and storage protein reserves in the starchy endosperm through
the production of hydrolases, glucanases, and proteinases after
hormone (gibberellic acid) stimulation from the embryo. Maize
(Figures 7A and 7F) and wheat have one layer of aleurone cells,
rice has one to several layers, and barley has three layers (Figure
7B) (Buttrose, 1963; Hoshikawa, 1993; Walbot, 1994). Barley
aleurone cells are highly polyploid (Keown et al., 1977). In the
mature grain of maize, the aleurone layer consists of an
estimated 250,000 aleurone cells derived by an estimated 17
rounds of anticlinal divisions (Levy and Walbot, 1990; Walbot,
1994). Toward the end of seed maturation, a specialized
developmental program confers desiccation tolerance to the
aleurone cells, allowing them to survive the maturation process
(Hoecker et al., 1995; Kao et al., 1996, and references therein).
Molecular markers for aleurone cells include Ltp2, B22E, pZE40,
ole-1, ole-2, per-1, and chi33 in barley (Klemsdal et al., 1991;
Madrid, 1991; Smith et al., 1992; Kalla et al., 1994; Leah et al.,
1994; Stacy et al., 1999) and C1 in maize (Neuffer et al., 1997).
Aleurone cells become morphologically distinct in barley
endosperm at 8 DAP (Bosnes et al., 1992), comparable to the
other cereals (Morrison et al., 1975; Brown et al., 1999). GUS
expression driven by the barley Ltp2 promoter in transgenic rice
grains is detectable at 9 DAP, closely matching the morphological
differentiation of aleurone cells (Kalla et al., 1994). How early does
aleurone cell fate specification occur? The analysis of barley
endosperm development described above suggests that aleurone
cell fate specification occurs after the first periclinal division
of the alveolar nuclei, with the outer sister nuclei assuming an
aleurone cell fate (Figures 4C and 7E) (Brown et al., 1994). The
basis for this conclusion is the observation that after the completion
of the cellularization process, these aleurone cell initials
(Figures 6G and 6H) display the full complement of cytoskeletal
arrays, including hoop-like cortical arrays and PPBs (Figures 6G
and 6H). By contrast, the inner daughter cells of this periclinal
division (giving rise to starchy endosperm cells) divide without
cortical arrays and PPBs (Figures 6E and 6F). Anticlinal divisions
in the aleurone layer expand the surface area of the aleurone
layer, whereas periclinal divisions contribute to the inner starchy
endosperm cells (see above). After 20 DAP, maize aleurone cell
mitotic divisions are predominantly anticlinal (Kiesselbach,
1949). Because of the role of PPBs in mitotic division plane
control in somatic cells, it is tempting to interpret the presence of
PPBs as the first structural manifestation of aleurone cell fate
specification. Three mutants in maize support the existence of
a genetic control mechanisms for division plane control in the
aleurone: xcl1 (extra cell layer1), in which the aleurone layer
possesses one extra cell layer as a result of aberrant periclinal
divisions (Kessler and Sinha, 2000); and dal1 and dal2 (disorganized
aleurone layer1 and 2), which have relaxed control over
aleurone division plane determination (Lid et al., 2004).
In wild-type grains, aleurone cells develop in close contact
with nucellus cells, which are part of the maternal plant. Recently,
during a microscopy screen of 12,000 maize mutant lines from
a collection of Mu transposon insertion lines (Lid et al., 2004),
we identified several hundred lines in which the endosperm
displayed developmental defects (Olsen, 2004). In many of
these, the mutant endosperm contained crevasses or indentations
from the surface penetrating into the starchy endosperm.
In all such cases, the crevasses were lined with aleurone cells. In
other lines, the endosperm consisted of small bodies of endosperm
cells that did not have direct contact with the maternal
tissues surrounding the endosperm. The surfaces of these
small bodies always were covered by one layer of aleurone
cells on top of an inner mass of starchy endosperm cells. From
these studies, we conclude that the endosperm of maize is
programmed to develop a layered structure with aleurone cells
on external surfaces.
What is the molecular basis of aleurone cell fate specification
and maintenance? In light of the ability of the endosperm to
develop a layer of aleurone cells on the surface of a body of
starchy endosperm cells, it is interesting that the three genes
known to affect aleurone cell fate specification and development
in maize, Cr4 (Becraft et al., 1996) (Figure 7G), Dek1 (Lid et al.,
2002) (Figure 7H), and Sal1 (Supernumerary aleurone layers1)
(Shen et al., 2003) (Figure 7I), encode proteins with similarity to
proteins implicated in cell-to-cell signaling in animals.
Cr4 encodes a protein receptor kinase­like molecule with
similarity to tumor necrosis factor receptors, prototypes of a large
family of cell surface receptors that are critical for lymphocyte
S222 The Plant Cell
development and function in mammals (Chan et al., 2000). The
similarity to tumor necrosis factor receptors is limited to three
Cys-rich domains of the extracellular domain that form the ligand
binding pocket for tumor necrosis factor.
Dek1 encodes a predicted 2159­amino acid protein with
a membrane-targeting signal in its N terminus followed by 21
transmembrane regions interrupted by an extracellular loop
region (Lid et al., 2002). The cytosolic C terminus encodes
a calpain-like Cys proteinase domain (Lid et al., 2002). Recently,
Wang et al. (2003) showed that the calpain-like domain of Dek1,
which is structurally very similar to animal calpains, has Cys
proteinase activity in vitro. Although stimulated by Ca21, the
Dek1 calpain is active in the absence of Ca21, suggesting that
the regulatory properties may be different from those of typical
animal calpains (Wang et al., 2003). The Dek1 transcript is
expressed ubiquitously in maize (Lid et al., 2002). The dependence
of aleurone cell identity on Dek1 throughout grain
development was investigated by revertant sector analysis
(Becraft and Asuncion-Crabb, 2000). In this analysis, aleurone
cells in heterozygous dek1/1 endosperm that lost the wild-type
allele as a result of Ds-induced chromosome breakage and loss
of the resulting acentric fragment carrying the Dek1 allele were
reported to revert to the starchy endosperm cell fate, even late
in grain development. Conversely, starchy endosperm cells in
the periphery of homozygous mutant dek1:Mu/dek1 grains that
gained Dek1 function as a result of Mu transposition from
a dek1:Mu insertion allele (restoring Dek1 wild-type function)
gained the aleurone cell fate. These data suggest that neither
aleurone nor starchy endosperm cell fate is a terminal state
of differentiation and that whatever cues are necessary for
the specification of aleurone identity are present throughout
development.
The third gene that affects aleurone cell specification, Sal1,
encodes a predicted 204­amino acid protein that is a homolog of
the human Charged vesicular body protein1/Chromatin modulating
protein1 gene, a member of the conserved family of the
class-E vacuolar protein­sorting genes implicated in membrane
vesicle trafficking (Shen et al., 2003). This gene also is expressed
ubiquitously in maize.
In addition to genes that have already been cloned, other
mutant genes have been described that result in multiple layers
of aleurone cells, including xcl1. Also, mutants in which aleurone
cell differentiation is disrupted have been reported, such as
collapsed2, opaque12, which has thin walled, flattened aleurone
cells with numerous vacuoles, paleface, with unusually rounded
cells and sporadically more than one cell layer (Becraft and
Asuncion-Crabb, 2000), and dappled mutants, with abnormal
aleurone cell morphology (Stinard and Robertson, 1987; Gavazzi
et al., 1997). Mutants in the etched loci (Scanlon et al., 1994a) and
the newly isolated mutants dal1 and dal2 (Lid et al., 2004) also
affect aleurone cell development.
Although significant progress has been made in identifying
genes implicated in aleurone cell development, the current level
of insight into signal transduction mechanisms in plants makes it
difficult to propose a model integrating the functions of Cr4,
Dek1, and Sal1. Based on analogy with animal signal transduction
mechanisms, we have proposed that aleurone cell
identity is specified by a ligand (unknown) in the periphery of the
endosperm that activates the Cr4 protein receptor kinase (Olsen
et al., 1998). As suggested by the identity of the Sal1 and Dek1
proteins, endosome trafficking and a calpain-like Cys proteinase
also play roles in aleurone signaling. The sal1 loss-of-function
mutant endosperm has multiple layers of aleurone cells, suggesting
that Sal1 functions to limit aleurone cell identity to the
outer cell layer in wild-type endosperm. Obviously, additional
research is needed before a complete model of aleurone cell fate
specification can be proposed.
ENDOSPERM CELL FATE SPECIFICATION AND
DIFFERENTIATION IN ARABIDOPSIS
The cellularization process for Arabidopsis described above
results in a completely cellular endosperm except for a small area
in the CZE adjacent to the chalazal cyst (Figure 4F). In contrast to
the persistent endosperm of the cereals, the cellular endosperm
of Arabidopsis is depleted gradually as the embryo grows. It
is generally assumed that the purpose of the nonpersistent
endosperm is to support the developing and growing embryo
and that the support function for the germinating embryo is taken
over by the cotyledons. In mature seeds not yet released from the
silique, a massive embryo fills the ovule and a single peripheral
layer sometimes referred to as the aleurone layer persists in the
mature ovule (Vaughn and Whitehouse, 1971; Chamberlain and
Horner, 1990; Groot and Van Caeseele Lawrence, 1993) (Figure
4G). In Arabidopsis, these cells exhibit few storage products and
thin cell walls (Keith et al., 1994), and their function remains
unknown. In the chalazal chamber, nodules of multinucleate
endosperm line the wall and a large coenocytic cyst of multinucleate
cytoplasm is positioned in the tip of the chalazal
chamber atop a specialized pad of maternal tissue known as the
chalazal proliferating tissue (data not shown), which has been
suggested to serve a role similar to the transfer cells in cereal
endosperm (for more details, see Brown et al., 1999).
MONOCOT AND DICOT NUCLEAR ENDOSPERM EVOLVED
INDEPENDENTLY BUT MAY HAVE RECRUITED THE SAME
ANCIENT SUBDEVELOPMENTAL PROGRAM FOR NCD
FORMATION AND ALVEOLATION
A comparison of the cellularization processes leading to the
cellular endosperm in cereals and the PEN in Arabidopsis
reveals striking similarities. Other aspects, including cell cycle
regulation during the cellularization process and the identity of
the different cell types, obviously are different. Therefore, it is
unclear whether knowledge about differentiation mechanisms
beyond the alveolation process applies to both types of
endosperm. Current phylogenetic data suggest that monocots
represent a monophyletic group that shares a common ancestor.
In agreement with this finding, the monocot nuclear
endosperm is highly conserved in all species investigated to
date. As described above, the monocot and dicot endosperm
are believed to have evolved independently. In spite of this,
nuclear endosperm in these groups show striking similarities
with respect to the cellularization process (Figures 4 and 5).
Importantly, the process of cellularization by RMS formation
also occurs during the cellularization process in a number of
Nuclear Endosperm S223
other systems, including sporogenesis in lower plants, male and
female gametophyte development in gymnosperms and angiosperms,
and embryogenesis in gymnosperms (Brown and
Lemmon, 2001). By contrast, the alveolation process as seen
in nuclear endosperm is found only in megagametophyte
development in gymnosperms and not in other existing
angiosperm systems. One possible explanation for the highly
conserved cellularization process of nuclear endosperm in
cereals and Arabidopsis is that the nuclear endosperm evolved
independently from a cellular endosperm by the same two steps
in these two angiosperm lineages. First came a mutation(s) that
suppressed phragmoplast formation in the mitotic divisions of
the central cell nucleus after fertilization, creating the endosperm
coenocyte. Further investigation is needed to determine
whether or not the same mechanisms for phragmoplast
suppression occur in both groups. Second came recruitment
of the same RMS-alveolation ``subprogram'' in both cases. In
this scenario, insight into the process of endosperm cellularization
is valid for both monocots and dicots.
CONCLUDING REMARKS AND PROSPECTS
Insight into the mechanisms of nuclear endosperm development
has advanced considerably during the last decade. Currently,
advances in nuclear endosperm research come from two of the
most powerful plant experimental systems available, maize and
Arabidopsis. Because of the assumed independent origin of
monocot and dicot nuclear endosperm, efforts to solve questions
related to nuclear endosperm evolution, coenocyte development,
cell cycle regulation, cell fate specification, and
differentiation need to continue with equal strength in both
cereals and Arabidopsis.
ACKNOWLEDGMENTS
Roy Brown and Betty Lemmon are gratefully acknowledged for introducing
me to the fascinating world of the cytoskeleton and for continued
collaboration and discussions. They also are thanked for providing the
micrographs in Figures 5 and 6 and for comments and suggestions on
the manuscript. Phil Becraft is acknowledged for helpful comments.
Kaisa Stenberg Olsen is acknowledged for the artwork in the figures,
and Changjiang Li is thanked for Figure 7G. Fred Gruis is acknowledged
for fruitful discussions during the course of the writing of this article, and
Paul Anderson is thanked for continued support. Thanks also are due to
many other colleagues who read and commented on the manuscript.
Received September 8, 2003; accepted January 7, 2004.
REFERENCES
Baroux, C., Spillane, C., and Grossniklaus, U. (2002). Evolutionary
origins of the endosperm in flowering plants. Genome Biol. 30, 1026.
Becraft, P.W. (2001). Cell fate specification in the cereal endosperm.
Semin. Cell Dev. Biol. 12, 387­394.
Becraft, P.W., and Asuncion-Crabb, Y. (2000). Positional cues specify
and maintain aleurone cell fate in maize endosperm development.
Development 127, 4039­4048.
Becraft, P.W., Brown, R.C., Lemmon, B.E., Olsen, O.-A., and
Opsahl-Ferstad, H.G. (2000). Endosperm development. In Current
Trends in the Embryology of Angiosperms, S.S. Bhojwani and W.-Y.
Soh, eds (Dordrecht, The Netherlands: Kluwer Academic Publishers),
Chapter 14.
Becraft, P.W., Stinard, P.S., and McCarty, D. (1996). CRINCKLY4: A
TNFR-like receptor kinase involved in maize epidermal differentiation.
Science 273, 1406­1409.
Bensen, R.J., Johal, G.S., Crane, V.C., Tossberg, J.T., Schnable,
P.S., Meeley, R.B., and Briggs, S.P. (1995). Cloning and characterization
of the maize An1 gene. Plant Cell 7, 75­84.
Berger, F. (2003). Endosperm: The crossroad of seed development.
Curr. Opin. Plant Biol. 6, 42­50.
Boisnard-Lorig, C., Colon-Carmona, A., Bauch, M., Hodge, S.,
Doerner, P., Bancharel, E., Dumas, C., Haseloff, J., and Berger, F.
(2001). Dynamic analyses of the expression of the HISTONE::YFP
fusion protein in Arabidopsis show that syncytial endosperm is
divided in mitotic domains. Plant Cell 13, 495­509.
Bonello, J.-F., Opsahl-Ferstad, H.-G., Perez, P., Dumas, C., and
Rogowsky, P.M. (2000). Esr genes show different levels of expression
in the same region of maize endosperm. Gene 246, 219­227.
Bonello, J.-F., Sevilla-Lecoq, S., Berne, A., Risueno, M.-C., Dumas,
C., and Rogowsky, P.M. (2002). Esr proteins are secreted by the
cells of the embryo surrounding region. J. Exp. Bot. 53, 1559­1568.
Bosnes, M., Harris, E., Ailgeltiger, L., and Olsen, O.-A. (1987).
Morphology and ultrastructure of 11 barley shrunken endosperm
mutants. Theor. Appl. Genet. 74, 177­187.
Bosnes, M., Weideman, F., and Olsen, O.-A. (1992). Endosperm differentiation
in barley wild-type and sex mutants. Plant J. 2, 661­674.
Brown, R., Lemmon, B.E., Nguyen, H., and Olsen, O.-A. (1999).
Development of endosperm in Arabidopsis thaliana. Sex. Plant
Reprod. 12, 32­42.
Brown, R.C., and Lemmon, B.E. (1992). Cytoplasmic domain: A model
for spatial control of cytokinesis in reproductive cells of plants. EMSA
Bull. 22, 48­53.
Brown, R.C., and Lemmon, B.E. (2001). The cytoskeleton and spatial
control of cytokinesis in the plant life cycle. Protoplasma 215, 35­49.
Brown, R.C., Lemmon, B.E., and Nguyen, H. (2002). Endosperm
development. In Annual Plant Review: Plant Reproduction, S.D.
O'Neill and J.A. Roberts, eds (Sheffield, UK: Sheffield University
Press), pp. 193­220.
Brown, R.C., Lemmon, B.E., and Nguyen, H. (2003). Events during the
first four rounds of mitosis establish three developmental domains in
the syncytial endosperm of Arabidopsis. Protoplasma, in press.
Brown, R.C., Lemmon, B.E., and Olsen, O.-A. (1994). Endosperm
development in barley: microtubule involvement in the morphogenetic
pathway. Plant Cell 6, 1241­1252.
Brown, R.C., Lemmon, B.E., and Olsen, O.-A. (1996a). Development
of the endosperm in rice (Oryza sativa L.): Cellularization. J. Plant Res.
109, 301­313.
Brown, R.C., Lemmon, B.E., and Olsen, O.-A. (1996b). Polarization
predicts the pattern of cellularization in cereal endosperm. Protoplasma
192, 168­177.
Brown, R.C., Lemmon, B.E., Stone, B., and Olsen, O.-A. (1997). Cell
wall (1!3)- and (1!3, 1!4)-b-glucans during early grain development
in rice (Oryza sativa L.). Planta 202, 414­426.
Buttrose, M. (1963). Ultrastructure of the developing aleurone cells of
wheat grain. Aust. J. Biol. Sci. 16, 768­774.
Calderini, O., Bo¨ gre, L., Vicente, O., Binarova, P., Herble-Bors, E., and
Wilson, C. (1998). A cell cycle regulated MAP kinase with a possible
role in cytokinesis in tobacco cells. J. Cell Sci. 111, 3091­3100.
Canaday, J., Stoppin-Mellet, V., Mutterer, J., Lambert, A.-M., and
Schmit, A.-C. (2000). Higher plant cells: g-Tubulin and microtubule
S224 The Plant Cell
nucleation in the absence of centrosomes. Microsc. Res. Tech. 49,
487­495.
Carmichael, J.S., and Friedman, W.E. (1995). Double fertilization in
Gnetum gnemon: The relationship between the cell cycle and sexual
reproduction. Plant Cell 7, 1975­1988.
Chamberlain, M.A., and Horner, H.T. (1990). An ultrastructural study of
endosperm wall formation and degeneration in Glycine max (L.) Merr.
Am. J. Bot. 77, 33.
Chan, F.K.-M., Chun, H.J., Zheng, L., Siegel, R.M., Bui, K.L., and
Lenardo, M.J. (2000). A domain in TNF receptors that mediates
ligand-independent receptor assembly and signaling. Science 288,
2351­2354.
Chaudhury, A.M., and Berger, F. (2001). Maternal control of seed
development. Semin. Cell Dev. Biol. 12, 381­386.
Chaudhury, A.M., Ming, L., Miller, C., Craig, S., Dennis, E.S., and
Peacock, W.J. (1997). Fertilization-independent seed development in
Arabidopsis thaliana. Proc. Natl. Acad. Sci. USA 94, 4223­4228.
Cheng, W.H., Taliercio, E.W., and Chourey, P.S. (1996). The
Miniature1 seed locus of maize encodes a cell wall invertase required
for normal development of endosperm and maternal cells in the
pedicel. Plant Cell 8, 971­983.
Christensen, C.A., King, E.J., Jordan, J.R., and Drews, G.N. (1997).
Megagametogenesis in Arabidopsis wild type and the Gf mutant. Sex.
Plant Reprod. 10, 49­64.
Cock, J.M., and McCormick, S. (2001). A large family of genes that
share homology with CLAVATA3. Plant Physiol. 126, 939­942.
Coe, E.H.J. (1978). The aleurone tissue of maize as a genetic tool. In
Maize Breeding and Genetics, D.B. Walden, ed (New York: John Wiley
& Sons), pp. 447­459.
Costa, L.M., Gutierrez-Marcos, J.F., Brutnell, T.P., Greenland, A.J.,
and Dickinson, H.G. (2003). The globby1-1 (glo1-1) mutation disrupts
nuclear and cell division in the developing maize seed causing
alterations in endosperm cell fate and tissue differentiation. Development
130, 5009­5017.
DeManson, D.A. (1997). Endosperm structure and development. In
Cellular and Molecular Biology of Plant Seed Development, B.A.
Larkins and I.K. Vasil, eds (Dordrecht, The Netherlands: Kluwer
Academic Publishers), pp. 73­115.
Doan, D.N., Linnestad, C., and Olsen, O.-A. (1996). Isolation of
molecular markers from the barley endosperm coenocyte and the
surrounding nucellus cell layers. Plant Mol. Biol. 31, 877­886.
Drews, G.N., Lee, D., and Christensen, C.A. (1998). Genetic analysis
of female gametophyte development and function. Plant Cell 10,
5­17.
Engell, K. (1989). Embryology of barley: Time course and analysis of
controlled fertilization and early embryo formation based on serial
sections. Nord. J. Bot. 9, 265­280.
Felker, F.C., Peterson, D.M., and Nelson, O.E. (1985). Anatomy of
immature grains of eight maternal effect shrunken endosperm barley
mutants. Am. J. Bot. 72, 248­256.
Felker, F.C., Peterson, D.M., and Nelson, O.E. (1987). Early grain
development of the seg2 maternal-effect shrunken-endosperm
mutant of barley. Can. J. Bot. 65, 943­948.
Fletcher, J.C., Brand, U., Running, M.P., Simon, R., and Meyerowitz,
E.M. (1999). Signaling of cell fate decisions by CLAVATA3 in
Arabidopsis shoot meristems. Science 283, 1911­1914.
Friedman, W.E. (1995). Organismal duplication, inclusive fitness theory,
and altruism: Understanding the evolution of endosperm and the
angiosperm reproductive syndrome. Proc. Natl. Acad. Sci. USA 92,
3913­3917.
Friedman, W.E. (1998). The evolution of double fertilization and
endosperm: An ``historicaĺ' perspective. Sex. Plant Reprod. 11, 6­16.
Friedman, W.E., and Floyd, S.K. (2001). The origin of flowering plants
and their reproductive biology: A tale of two phylogenies. Evolution
Int. J. Org. Evolution 55, 217­231.
Friedman, W.E., and Williams, J.H. (2003). Modularity of the angiosperm
female gametophyte and its bearing on the early evolution of
endosperm in flowering plants. Evolution Int. J. Org. Evolution 57,
216­230.
Friedman, W.D., and Williams, J.H. (2004). Developmental evolution of
the sexual process in ancient flowering plant lineages. Plant Cell
16 (suppl.), S119­S132.
Fu, S., Meeley, R., and Scanlon, M.J. (2002). empty pericarp2 encodes
a negative regulator of the heat shock response and is required for
maize embryogenesis. Plant Cell 14, 3119­3132.
Gao, R., Dong, S., Fan, J., and Hu, C. (1998). Relationship between
development of endosperm transfer cells and grain mass in maize.
Biol. Plant. 41, 539­546.
Gavazzi, G., Dolfini, S., Allegra, D., Castiglioni, P., Todesco, G., and
Hoxha, M. (1997). Dap (defective aleurone pigmentation) mutations
affect maize aleurone development. Mol. Gen. Genet. 256, 223­230.
Geeta, R. (2003). The origin and maintenance of nuclear endosperm:
Viewing development through a phylogenetic lens. Proc. R. Soc.
Lond. B Biol. Sci. 270, 29­35.
Gehring, M., Choi, Y., and Fischer, R.L. (2004). Imprinting and seed
development. Plant Cell 16 (suppl.), S203­S213.
Gó mez, E., Royo, J., Guo, Y., Thompson, R., and Hueros, G. (2002).
Establishment of cereal endosperm expression domains: Identification
and properties of a maize transfer cell­specific transcription
factor, ZmMRP-1. Plant Cell 14, 599­610.
Groot, E.P., and Van Caeseele Lawrence, A. (1993). The development
of the aleurone layer in canola (Brassica napus). Can. J. Bot. 71, 1193­
1201.
Grossniklaus, U., and Schneitz, K. (1998). The molecular and genetic
basis of ovule and megagametophyte development. Semin. Cell Dev.
Biol. 9, 227­238.
Grossniklaus, U., Vielle-Calzada, J.P., Hoeppner, M.A., and Gagliano,
W.B. (1998). Maternal control of embryogenesis by MEDEA, a
Polycomb group gene in Arabidopsis. Science 280, 446­450.
Guignard, L. (1901). La double fe` condation dans le mais. J. Bot. 15,
37­50.
Heckel, T., Werner, K., Sheridan, W.F., Dumas, C., and Rogowsky,
P.M. (1999). Novel phenotypes and developmental arrest in early
embryo specific mutants of maize. Planta 210, 1­8.
Heese, M., Mayer, U., and Jurgens, G. (1998). Cytokinesis in higher
plants: Cellular process and developmental integration. Curr. Opin.
Plant Biol. 1, 486­491.
Hoecker, U., Vasil, I.K., and McCarty, D.R. (1995). Integrated control
of seed maturation and germination programs by activator and
repressor functions of Viviparous-1 of maize. Genes Dev. 9, 2459­
2469.
Hoshikawa, K. (1993). Anthesis, fertilization and development of
caryopsis. In Science of the Rice Plant. Vol. I. Morphology, T. Matsuo
and K. Hoshikawa, eds (Tokyo: Nobunkyo), pp. 339­376.
Hueros, G., Royo, J., Maitz, M., Salamini, F., and Thompson, R.D.
(1999). Evidence for factors regulating transfer cell-specific expression
in maize endosperm. Plant Mol. Biol. 41, 403­414.
Hueros, G., Varotto, S., Salamini, F., and Thompson, R.D. (1995).
Molecular characterization of BET1, a gene expressed in the
endosperm transfer cells of maize. Plant Cell 7, 747­757.
Kalla, R., Shimamoto, K., Potter, R., Nielsen, P.S., Linnestad, C., and
Olsen, O.A. (1994). The promoter of the barley aleurone-specific gene
encoding a putative 7 kDa lipid transfer protein confers aleurone cellspecific
expression in transgenic rice. Plant J. 6, 849­860.
Nuclear Endosperm S225
Kao, C.Y., Cocciolone, S.M., Vasil, I.K., and McCarty, D.R. (1996).
Localization and interaction of the cis-acting element for abscisic
acid, Viviparous 1, and light activation of the C1 gene of maize. Plant
Cell 8, 1171­1179.
Keith, K., Krami, M., Dengler, N.G., and McCourt, P. (1994). Fusca3: A
heterochronic mutation affecting late embryo development in Arabidopsis.
Plant Cell 6, 589­600.
Keown, A.C., Taiz, L., and Jones, R.L. (1977). The nuclear content of
developing barley aleurone cells. Am. J. Bot. 64, 1248­1253.
Kessler, S.A., and Sinha, N.R. (2000). Characterization of xc1 (extra cell
layers), a mutation affecting plane of cell division during maize
development. In Plant Biology 2000. (Rockville, MD: American Society
of Plant Physiologists), p. 48.
Kiesselbach, T.A. (1949). The structure and reproduction of corn.
Research Bulletin 161, University of Nebraska College of Agriculture.
Klemsdal, S.S., Hughes, W., Lonneborg, A., Aalen, R.B., and Olsen,
O.-A. (1991). Primary structure of a novel barley gene differentially
expressed in immature aleurone layers. Mol. Gen. Genet. 228, 9­16.
Kohler, C., Hennig, L., Spillane, C., Pien, S., Gruissem, W., and
Grossniklaus, U. (2003). The Polycomb-group protein MEDEA
regulates seed development by controlling expression of the MADSbox
gene PHERES1. Genes Dev. 17, 1540­1553.
Kowles, R.V., and Phillips, R.L. (1988). Endosperm development in
maize. Int. Rev. Cytol. 112, 97­136.
Larkins, B.A., Dilkes, B.P., Dante, R.A., Coelho, C.M., Woo, Y.M.,
and Liu, Y. (2001). Investigating the hows and whys of DNA
endoreduplication. J. Exp. Bot. 52, 183­192.
Lauber, M.H., Waizenegger, I., Steinmann, T., Schwarz, H., Mayer,
U., Hwang, I., Lukowitz, W., and Jurgens, G. (1997). The Arabidopsis
KNOLLE protein is a cytokinesis-specific syntaxin. J. Cell Biol.
139, 1485­1493.
Leah, R., Skriver, K., Knudsen, S., Ruud-Hansen, J., Raikhel, N.V.,
and Mundy, J. (1994). Identification of an enhancer/silencer sequence
directing the aleurone specific expression of a barley
chitinase gene. Plant J. 6, 579­589.
Levy, A.A., and Walbot, V. (1990). Regulation of the timing of
transposable element excision during maize endosperm development.
Science 248, 1534­1537.
Lid, S.E., Al, R.H., Krekling, T., Meeley, R.B., Ranch, J., OpsahlFerstad,
H.-G., and Olsen, O.-A. (2004). The maize disorganized
aleurone layer 1 and 2 (dil1, dil2) mutants lack control of mitotic
division plane in the aleurone layer of developing endosperm. Planta
218, 370­378.
Lid, S.E., Gruis, D., Jung, R., Lorentzen, J.A., Ananiev, E.,
Chamberlain, M., Niu, X., Meeley, R., Nichols, S., and Olsen,
O.-A. (2002). The defective kernel 1 (dek1) gene required for aleurone
cell development in the endosperm of maize grains encodes a membrane
protein of the calpain gene superfamily. Proc. Natl. Acad. Sci.
USA 99, 5460­5465.
Liu, C.M., McElver, J., Tzafrir, I., Joosen, R., Wittich, P., Patton, D.,
Van Lammeren, A.A., and Meinke, D. (2002). Condensin and
cohesin knockouts in Arabidopsis exhibit a titan seed phenotype.
Plant J. 29, 405­415.
Lukowitz, W., Mayer, U., and Jurgens, G. (1996). Cytokinesis in the
Arabidopsis embryo involves the syntaxin-related KNOLLE gene
product. Cell 84, 61­71.
Luo, M., Bilodeau, P., Dennis, E.S., Peacock, W.J., and Chaudhury,
A. (2000). Expression and parent-of-origin effects for FIS2, MEA, and
FIE in the endosperm and embryo of developing Arabidopsis seeds.
Proc. Natl. Acad. Sci. USA 97, 10637­10642.
Machida, Y., Nakashima, M., Morikiyo, K., Banno, H., Ishikawa, M.,
Soyano, T., and Nishihama, R. (1998). MAPKKK-related protein
kinase NPK1: Regulation of the M phase plant cycle. J. Plant Cell Res.
111, 243­246.
Madrid, S.M. (1991). The barley lipid transfer protein is targeted into
the lumen of the endoplasmic reticulum. Plant Physiol. Biochem. 29,
695­704.
Magnard, J.L., Le Deunff, E., Domenech, J., Rogowsky, P.M.,
Testillano, P.S., Rougier, M., Risueno, M.C., Vergne, P., and
Dumas, C. (2000). Endosperm-specific features at the onset of
microspore embryogenesis in maize. Plant Mol. Biol. 44, 559­574.
Maitz, M., Santandrea, G., Zhang, Z., Lal, S., Hannah, C., Salamini,
F., and Thompson, R.D. (2000). Rgf1, a mutation reducing grain filling
in maize through effects on basal endosperm and pedicel development.
Plant J. 23, 29­42.
Mansfield, S.G. (1994). Endosperm development. In Arabidopsis: An
Atlas of Morphology and Development, J.L. Bowman, ed (Berlin:
Springer), pp. 385­397.
Mansfield, S.G., and Briarty, L.G. (1990a). Development of the freenuclear
endosperm in Arabidopsis thaliana L. Arabidopsis Inf. Serv.
27, 53­64.
Mansfield, S.G., and Briarty, L.G. (1990b). Endosperm cellularization in
Arabidopsis thaliana L. Arabidopsis Inf. Serv. 27, 65­72.
Mansfield, S.G., Briarty, L.G., and Erni, S. (1991). Early embryogenesis
in Arabidopsis thaliana. I. The mature embryo sac. Can. J. Bot. 69,
447­460.
Mathews, S., and Donoghue, M.J. (1999). The root of angiosperm
phylogeny inferred from duplicate phytochrome genes. Science 286,
947­950.
McClintock, B. (1984). The significance of responses of the genome to
challenge. Science 226, 792­801.
McElver, J., Patton, D., Rumbaugh, M., Liu, C., Yang, L., and
Meinke, D. (2000). The TITAN5 gene of Arabidopsis encodes a protein
related to the ADP ribosylation factor family of GTP binding proteins.
Plant Cell 12, 1379­1392.
Miller, M.E., and Chourey, P.S. (1992). The maize invertase-deficient
miniature-1 seed mutation is associated with aberrant pedicel and
endosperm development. Plant Cell 4, 297­305.
Morrison, I.N., Kuo, J., and O'Brien, T.P. (1975). Histochemistry
and fine structure of developing wheat aleurone cells. Planta 123,
105­116.
Nawaschin, S. (1898). Resulte eine revison der Befruchtungsvorgange
bei Lilium martagon und Ritillaria tenella. Bulletin of the Academy of
Imperial Sciences St. Petersburg 9, 377­382.
Neuffer, M.G., Coe, E.H., and Wessler, S.R. (1997). Mutants of Maize.
(Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press).
Neuffer, M.G., and Sheridan, W.F. (1980). Defective kernels of maize.
I. Genetic and lethality studies. Genetics 95, 929­944.
Nishihama, R., Jeong, S., DeYoung, B., and Clark, S.E. (2003).
Retraction. Science 300, 1370.
Ohad, N., Margossian, L., Hsu Yung, C., Williams, C., Repetti, P.,
and Fischer, R.L. (1996). A mutation that allows endosperm
development without fertilization. Proc. Natl. Acad. Sci. USA 93,
5319­5324.
Olsen, O.-A. (2001). Endosperm development: Cellularization and
cell fate specification. Annu. Rev. Plant Physiol. Plant Mol. Biol. 52,
233­267.
Olsen, O.-A. (2004). Dynamics of maize aleurone cell formation: the
``surface'' rule. Maydica, in press.
Olsen, O.-A., Lemmon, B.E., and Brown, R.C. (1998). A model for
aleurone cell development. Trends Plant Sci. 3, 168­169.
Opsahl-Ferstad, H.G., Le Deunff, E., Dumas, C., and Rogowsky,
P.M. (1997). ZmEsr, a novel endosperm-specific gene expressed in
a restricted region around the maize embryo. Plant J. 12, 235­246.
S226 The Plant Cell
Otegui, M., and Staehelin, A. (2000). Syncytial-type cell plates: A novel
kind of cell plate involved in endosperm cellularization in Arabidopsis.
Plant Cell 12, 933­947.
Otegui, M.S., Mastronarde, D.N., Kang, B.-H., Bednarek, S.Y., and
Staehelin, L.A. (2001). Three-dimensional analysis of syncytial-type
cell plates during endosperm cellularization visualized by high
resolution electron tomography. Plant Cell 13, 2033­2051.
Pickett-Heaps, J.D., Gunning, B.E.S., Brown, R.C., Lemmon, B.E.,
and Cleary, A. (1999). The cytoplast concept in dividing plant cells:
Cytoplasmic domains and the evolution of spatially organized cell
division. Am. J. Bot. 86, 153­172.
Pirrotta, V. (1998). Polycombing the genome: PcG, trxG, and chromatin
silencing. Cell 93, 333­336.
Qiu, Y.L., Lee, J.S., Bernasconi-Quadroni, F., Soltis, D.E., Soltis,
P.S., Zanis, M., Zimmer, E.A., Chen, Z., Savolainen, V., and Chase,
M.W. (1999). The earliest angiosperms: Evidence from mitochondrial,
plastid and nuclear genomes. Nature 402, 404­407.
Russel, S.D. (1993). The egg cell: Development and the role in
fertilization and early embryogenesis. Plant Cell 5, 1349­1359.
Sato, K., and Kawashima, S. (2001). Calpain function in the modulation
of signal transduction molecules. Biol. Chem. 382, 743­751.
Scanlon, M.C., James, M.G., Stinard, P.S., Myers, A.M., and
Roberston, D.S. (1994a). Characterization of ten new mutations of
the maize Etched-1 locus. Maydica 39, 301­308.
Scanlon, M.C., Stinard, P.S., James, M.G., Myers, A.M., and
Roberston, D.S. (1994b). Genetic analysis of 63 mutations affecting
maize kernel development isolated from Mutator stocks. Genetics
136, 281­294.
Scanlon, M.J., and Myers, A.M. (1998). Phenotypic analysis and
molecular cloning of discolored-1 (dsc1), a maize gene required for
early kernel development. Plant Mol. Biol. 37, 483­493.
Schel, J.H.N., Kieft, H., and Van Lammeren, A.A.M. (1984). Interactions
between embryo and endosperm during early developmental
stages of maize caryopses (Zea mays). Can. J. Bot. 62, 2842­2853.
Schneitz, K., Hu¨ lskamp, M., and Pruitt, R.E. (1995). Wild-type ovule
development in Arabidopsis thaliana: A light microscope study of
cleared whole-mount tissue. Plant J. 7, 731­774.
Serna, A., Maitz, M., O'Connell, T., Santandrea, G., Thevissen, K.,
Tienens, K., Hueros, G., Faleri, C., Cai, G., Lottspeich, F., and
Thompson, R.D. (2001). Maize endosperm secretes a novel antifungal
protein into adjacent maternal tissue. Plant J. 25, 687­698.
Shen, B., Li, C., Min, Z., Meeley, R.B., Tarczynski, M.C., and Olsen,
O.-A. (2003). Sal1 determines the number of aleurone cell layers in
maize endosperm and encodes a class E vacuolar sorting protein.
Proc. Natl. Acad. Sci. USA 100, 6552­6557.
Smith, L.G. (1999). Divide and conquer: Cytokinesis in plant cells. Curr.
Opin. Plant Biol. 2, 447­453.
Smith, L.M., Handley, J., Li, Y., Martin, H., Donovan, L., and Bowles,
D.J. (1992). Temporal and spatial regulation of a novel gene in barley
embryos. Plant Mol. Biol. 20, 255­266.
Soltis, P.S., Soltis, D.E., and Chase, M.W. (1999). Angiosperm
phylogeny inferred from multiple genes as a research tool for
comparative biology. Nature 402, 402­404.
Sorensen, M.B., Chaudhury, A.M., Robert, H., Bancharel, E., and
Berger, F. (2001). Polycomb group genes control pattern formation in
plant seed. Curr. Biol. 11, 277­281.
Sorensen, M.B., Mayer, U., Lukowitz, W., Robert, H., Chambrier, P.,
Jurgens, G., Somerville, C., Lepiniec, L., and Berger, F. (2002).
Cellularisation in the endosperm of Arabidopsis thaliana is coupled to
mitosis and shares multiple components with cytokinesis. Development
129, 5567­5576.
Stacy, R.A., Nordeng, T.W., Culianez-Macia, F.A., and Aalen, R.B.
(1999). The dormancy-related peroxiredoxin anti-oxidant, PER1, is
localized to the nucleus of barley embryo and aleurone cells. Plant J.
19, 1­8.
Staehelin, L.A., and Hepler, P.K. (1996). Cytokinesis in higher plants.
Cell 84, 821­824.
Stangeland, B., Salehian, Z., Aalen, R., Mandal, A., and Olsen, O.-A.
(2003). Isolation of GUS marker lines for genes expressed in Arabidopsis
endosperm, embryo and maternal tissues. J. Exp. Bot. 54,
279­290.
Steinborn, K., Maulbetsch, C., Priester, B., Trautmann, S., Pacher,
T., Geiges, B., Ku¨ ttner, F., Lepiniec, L., Stierhof, Y., Schwarz, H.,
Ju¨ rgens, G., and Mayer, U. (2002). The Arabidopsis PILZ group
genes encode tubulin-folding cofactor orthologs required for cell
division but not cell growth. Genet. Dev. 16, 959­971.
Stinard, P., and Robertson, D. (1987). Dappled: a putative Mu-induced
aleurone developmental mutant. Maize Genet. Coop. Newsl. 61, 7­9.
Strompen, G., El Kasmi, F., Richter, S., Lukowitz, W., Assaad, F.F.,
Jurgens, G., and Mayer, U. (2002). The Arabidopsis HINKEL gene
encodes a kinesin-related protein involved in cytokinesis and is
expressed in a cell cycle-dependent manner. Curr. Biol. 12, 153­158.
Swamy, B.G.L. (1946). Inverted polarity of the embryo sac of
angiosperms and its relation to the archegonium theory. Ann. Bot.
10, 171­183.
Sylvester, A.W. (2000). Division decisions and the spatial regulation of
cytokinesis. Curr. Opin. Plant Biol. 3, 58­66.
Thompson, R.D., Hueros, G., Becker, H.A., and Maitz, M. (2001).
Development and functions of seed transfer cells. Plant Sci. 160,
775­783.
Tian, G.W., You, R.L., Guo, F.L., and Wang, X.C. (1998). Microtubular
cytoskeleton of free endosperm nuclei during division in wheat.
Cytologia 63, 427­433.
Tzafrir, I., McElver, J.A., Liu, C.-M., Yang, L.J., Wu, J.Q., Martinez,
A., Patton, D.A., and Meinke, D.W. (2002). Diversity of TITAN
functions in Arabidopsis seed development. Plant Physiol. 128,
38­51.
Van Lammeren, A., Ageeva, M., Kieft, H., Ma, F., and Van
Veenendaal, W.L.H. (1996). Light microscopical study of endosperm
formation in Brassica napus L. Acta Soc. Bot. Pol. 65, 267­272.
Vaughn, J.G., and Whitehouse, J.M. (1971). Seed structure and the
taxonomy of the Cruciferae. Bot. J. Linn. Soc. 64, 383­409.
Walbot, V. (1994). Overview of key steps in aleurone development. In
The Maize Handbook, M. Freeling and V. Walbot, eds (New York:
Springer-Verlag), pp. 78­80.
Wang, C., Barry, J.K., Min, Z., Tordsen, G., Rao, A.G., and Olsen,
O.-A. (2003). The calpain domain of the maize DEK1 protein contains
the conserved catalytic triad and functions as a cysteine proteinase.
J. Biol. Chem. 278, 34467­34474.
Webb, M.C., and Gunning, B.E.S. (1990). Embryo sac development in
Arabidopsis thaliana. I. Megasporogenesis, including the microtubular
cytoskeleton. Sex. Plant Reprod. 3, 244­256.
Webb, M.C., and Gunning, B.E.S. (1991). The microtubular cytoskeleton
during development of the zygote, proembryo and freenuclear
endosperm in Arabidopsis thaliana (L.) Heynh. Planta 184,
187­195.
Williams, J.H., and Friedman, W.E. (2002). Identification of diploid
endosperm in an early angiosperm lineage. Nature 425, 522­526.
Young, T.E., Gallie, D.R., and DeMason, D.A. (1997). Ethylenemediated
programmed cell death during maize endosperm development
of wild type and shrunken2 genotypes. Plant Physiol.
115, 737­751.
Nuclear Endosperm S227