Journal of Experimental Botany, Vol. 52, No. 360, pp. 1401±1408, July 2001
Water deficit inhibits cell division and expression of
transcripts involved in cell proliferation and
endoreduplication in maize endosperm
Tim L. Setter1 and Brian A. Flannigan
Department of Crop and Soil Sciences, Cornell University, Ithaca, NY 14853, USA
Received 29 August 2000; Accepted 18 February 2001
Abstract
Water deficit at the early post-pollination stage in
cereal grains decreases endosperm cell division and,
in turn, decreases the capacity for storage material
accumulation. Post-mitotic replication of nuclear
DNA (endoreduplication) may also play a role in
stress effects. To gain a better understanding of the
extent to which cell proliferation and endoreduplication are affected by water deficit, nuclear numbers
and size were examined in endosperms of maize
(Zea mays L.) by flow cytometry and the transcript
levels of genes which have recognized roles in the
cell cycle were quantified. Water deficit from 5 ±13 d
after pollination (DAP) decreased the rate of endosperm cell division by 90% and inhibited w3Hxthymidine incorporation into DNA from 9 ±13 DAP.
The proportion of nuclei engaging in endoreduplication and nuclear DNA content increased steadily
from 9 ±13 DAP in controls, but water deficit initially
increased the proportion of endoreduplicating nuclei
at 9 DAP, then halted further entry into endoreduplication and S-phase cycling from 9 ±13 DAP. Transcript levels of a-tubulin, and the S-phase gene
products histone H3 and PCNA were not affected by
water deficit until 13 DAP, whereas those of ZmCdc2,
a cyclin dependent kinase (CDK) with regulatory
roles in mitosis, were inhibited substantially from
9 ±13 DAP. Cell proliferation and associated processes were inhibited at initial stages of the stress
episode, whereas endoreduplication and associated
S-phase processes were not inhibited until the stress
was more advanced. It was concluded that endosperm
mitosis has greater sensitivity than endoreduplication
to water deficit.
1
Key words: Cell cycle, drought, endopolyploidy, kernel set.
Introduction
The storage capacity of developing cereal grains is
established during the early stage of endosperm development by the processes of cell division, organelle proliferation and cell enlargement, which create the metabolic
capacity and ®nal volume of this tissue (Jones et al.,
1996; Olsen et al., 1999). Environmental stresses in the
initial period following fertilization inhibit these early
events and, in turn, decrease ®nal grain yield (Artlip
et al., 1995; Commuri and Jones, 1999; Nicolas et al.,
1985; Ober et al., 1991). During early endosperm development, the mitotic phase is followed by a developmental
transition, after which cells engage in endoreduplication,
a process of repeated rounds of S phase in the absence
of mitosis (Gra®, 1998; Kowles et al., 1990). Endoreduplication coincides with the period of organelle proliferation and rapid expression of mRNAs for starch pathway
enzymes and storage protein (Ober et al., 1991). It is
thought to contribute to growth capacity as it is correlated with reproductive-organ and cell size among
contrasting genotypes and plants subjected to various
stress treatments (Artlip et al., 1995; Cavallini et al., 1995;
Cebolla et al., 1999; Lemontey et al., 2000). However,
with respect to kernel growth and yield, the stage in
which the number of endosperm cells is established is
particularly sensitive to stress; whereas the later phases,
when endoreduplication and starch accumulation occur,
are relatively tolerant of stress (Ouattar et al., 1987;
reviewed in Mambelli and Setter, 1998). Hence, there is
a need to understand better the differential response of
these developmental events to stress.
To whom correspondence should be addressed. Fax: q1 607 255 2644. E-mail: [email protected]
ß Society for Experimental Biology 2001
1402
Setter and Flannigan
Knowledge of cell cycle regulation in eukaryotic cells
has advanced considerably in recent years. The eukaryotic cell cycle is controlled by a family of protein
kinases, each with a positive regulatory subunit, termed
a cyclin, and a catalytic subunit, termed a cyclin
dependent kinase (CDK) (reviewed in den Boer and
Murray, 2000). There are several distinct cyclinuCDK
pairs, each functioning at a speci®c phase(s) of the cell
cycle (Roberts, 1999). The activity of each of these
cyclinuCDK complexes oscillates with each turn of the
cell cycle to control progress at each step and respond
to signals arising from the environment, such as those
created by stress. Given the speci®city of these regulatory
components for particular phases of the cell cycle, it is
possible that during a stress episode mitotic cell cycling
might be affected differently from cell-cycle phases
speci®c to endoreduplication.
There is evidence in maize (Zea mays L.) that mitosis
and endoreduplication are differentially regulated.
Auxin stimulates endoreduplication of maize endosperm
while not affecting cell numbers (Lur and Setter, 1993).
Among a set of defective kernel mutants most decreased
the relative extent to which nuclei undergo endoreduplication, but one mutant maintained DNA endoreduplication at wild-type levels (Kowles et al., 1992). Studies using
maize inbreds differing in endoreduplication and their
reciprocal crosses, showed that genotypic effects on the
extent of endoreduplication are largely under maternal
control (Kowles et al., 1997).
The impact of stress on mitosis and endoreduplication
in maize endosperm has been tested with short-term
post-pollination treatments. In studies where plants were
subjected to short-term water deprivation at two stages
of endosperm development, from 1±10 d after pollination
(DAP), when mitotic cell cycling predominates, or from
9±15 DAP, when endoreduplication predominates, water
de®cit drastically inhibited the rate of endosperm cell
division whereas the rate of endoreduplication was
inhibited to a lesser extent (Artlip et al., 1995). Studies
of maize endosperm have also shown that mitosis is
more sensitive than endoreduplication to exogenously
applied abscisic acid (ABA) applied from 5±11 DAP,
coinciding with maximal cell division rates and the onset
of endoreduplication (Mambelli and Setter, 1998). The
distribution of nuclei among DNA-content size-classes
indicated that water de®cit and ABA inhibited both the
rate of transition from mitotic to endoreduplication status,
and the rate of S-phase cycling. Studies of in vitro-cultured
maize kernels have indicated that although high temperature stress (35 8C), imposed from 4±8 or 4±10 DAP,
inhibited both mitosis and endoreduplication, the impact
on endoreduplication was later, during recovery after
stress was relieved (Engelen-Eigles et al., 2000).
Previous studies of stress effects on endosperm
cell cycle involved short-term stress imposition during
maximal cell division or endoreduplication (Artlip et al.,
1995; Mambelli and Setter, 1998). The objective of the
current study was to determine the extent to which mitosis
and endoreduplication in maize endosperm are affected
by water de®cit imposed during a time-frame bracketing
the maximum activity of both of these processes.
In addition to ¯ow cytometry, the impact of stress
on expression of gene products with roles in the cell cycle
was assessed. The results indicate that water de®cit
inhibited both mitosis and endoreduplication, but mitosis
was affected earlier in the stress episode, whereas endoreduplication was affected after stress had advanced
further.
Materials and methods
Plant material
Maize (Pioneer Brand 3925) was grown in a glasshouse with
hourly irrigation and arti®cial illumination as previously
described (Artlip et al., 1995). When solar photon ¯ux density
of photosynthetically active radiation (PAR, 400±700 nm
wavelengths) was less than 250 mmol m 2 s 1, it was supplemented between 06.00 h and 20.00 h with illumination from
1000 W metal halide lamps.
Watering treatments
Water de®cit stress was imposed as previously described (Artlip
et al., 1995), except irrigation water was withheld beginning at
5 DAP. Brie¯y, the mass of each well-watered pot, containing
plants plus soil, was measured at 5 DAP, pots were transferred
to an automatic gravimetric system that allowed water to be
depleted until the mass reached 50% of initial wet mass, and
then this set-point was maintained by hourly addition of
irrigation solution. At about 7 DAP soil water content was
depleted to the set-point which was suf®cient to induce leaf
wilt as indicated by leaf rolling and appearance of glaucous
leaf surfaces. Tests indicated that during days of high transpiration (high light ¯ux density), this set-point corresponded
to midday leaf water potentials of about 2.2 MPa while
well-watered control leaf water potentials were about
1.2 MPa.
Analysis of nuclei
Endosperms from kernels in the apical zone of each ear, the
upper 33% with respect to ear longitudinal length, were dissected free of embryo and nucellus, and endosperms were ®xed
in 3 : 1 (vuv) ethanol : acetic acid. Nuclei in homogenates were
analysed by ¯ow cytometry, as previously described (Artlip
et al., 1995). Brie¯y, ®xative was removed by washing
endosperms in water, endosperms were then incubated with
pectinase, and cell walls were disrupted by gentle passage
through successively smaller hypodermic needles to release individual nuclei. RNA was destroyed during the pectinase reaction. Nuclei in homogeneous suspension were stained with
propidium iodide (a DNA-binding ¯uorochrome), and analysed
on a ¯ow cytometer which measured the propidium ¯uorescence intensity, a measure of DNA content, of each nucleus.
For each sample, between 5000 and 19 000 nuclei were
Water deficit inhibits endosperm cell cycle
analysed. Histograms of frequency (nuclear counts) versus
logarithm of ¯uorescence were then produced, as shown in
Fig. 1.
Tritiated thymidine labelling
Endosperms from the apical ear zone, as described above, were
cut in half lengthwise, dissected free of embryo and nucellus,
and incubated at 22 8C for 2 h in 0.25 ml of 10 mM TrisHCl ( pH 7.5), 1% (wuv) glucose, to which 3.7 MBq of
w3H-methylx-thymidine (740 MBq mmol 1) was added. DNA
was extracted and puri®ed as described earlier (Davis et al.,
1986), then radioactivity was determined by liquid scintillation
spectrometry.
Analysis of mRNA
Maize endosperms were dissected as above, immediately frozen
in liquid N2, then RNA was extracted and puri®ed as previously described (Wadsworth et al., 1988). Brie¯y, thawing
tissue was homogenized in 2.5 vols of RNA extraction buffer
(100 mM Tris-HCl, pH 8.5, 20 mM aurin tricarboxylic acid,
200 mM LiCl, 100 mM EDTA, 100 mM b-mecaptoethanol).
The mixture was centrifuged at 10 000 g for 10 min and the
supernatant was extracted with phenoluchloroform. RNA in
the aqueous phase was precipitated with 3 M LiCl, redissolved
and reprecipitated with ethanol. Poly(A)q RNA was isolated
by oligo(dT) cellulose column chromatography (Sambrook
et al., 1989).
RNA was quanti®ed by UV spectroscopy and electrophoresed on 1% agaroseuformaldehyde gels. Gels were blotted
downward onto charged nylon membranes (Nytran, Schleicher
& Schuell) utilizing a neutral solution of 10 3 SSC (Sambrook
et al., 1989), and crosslinked with ultraviolet radiation
Fig. 1. Histogram illustrating the distribution of nuclei into DNAcontent size-classes, as analysed by ¯ow cytometry. Nuclei from control
endosperms at 9 DAP (a) and 13 DAP (b) are shown. Endosperms were
treated with pectinase to release nuclei into homogeneous suspension,
treated with propidium iodide ¯uorochrome, and analysed by ¯ow
cytometry. Each peak is labelled with its DNA copy number, where 1C
is the haploid nuclear DNA content.
1403
(Stratalinker, Stratagene Cloning Sys., La Jolla, CA). Radiolabelled probes (indicated in ®gure legends) were synthesized
using the Multiprime DNA labelling system (Amersham) and
w32PxdCTP. Membranes were prehybridized for 0.5 h at 65 8C
in blocking buffer (Boehringer-Mannheim), modi®ed to include
10% SDS, and then hybridized with w32Px-labelled probe at
68 8C for 16 ±20 h. Membranes were washed twice at room
temperature in 2 3 SSC for 30 s, once at 68 8C in 1 3 SSC
for 10 min, and once at 68 8C in 0.1 3 SSC for 20 min. The
hybridized membranes were exposed to X-ray ®lm at 80 8C
using intensifying screens. Each blot was subsequently washed
free of probe and rehybridized with ribosomal RNA probe,
pGMR from soybean, to establish that lanes were equally
loaded (data not shown).
Results and discussion
Water deficit inhibition of cell division and 3H-thymidine
incorporation into DNA
The rate of cell division in maize endosperm is maximal
from about 8±11 d after pollination, preceding rapid
starch accumulation (Artlip et al., 1995; Lur and Setter,
1993; Ober et al., 1991). In the current studies, maize
plants were subjected to water de®cit from 5±13 DAP
encompassing the period of maximal cell division activity.
Water de®cit substantially inhibited endosperm cell division (Fig. 2): in the period from 9±13 DAP the control
endosperms had a nuclear doubling time of about 2 d,
while cell division was essentially halted in endosperms
of water de®cit plants. Plants in which water was withheld beginning at 5 DAP depleted soil water and did not
reach the gravimetric set point until about 7 DAP, corresponding to the ®rst appearance of leaf rolling and
glaucous leaf surface. Nuclear counts are a measure of
treatment effects exerted over the time-frame of treatment
imposition, so treatment differences in nuclear counts
gradually became apparent between 9 and 13 DAP. But
3
H-thymidine incorporation into DNA, a measure of cell
DNA synthesis activity at a speci®c point in time,
Fig. 2. Effect of water de®cit on the proliferation of endosperm nuclei
in maize kernels. Samples were obtained from well-watered control
plants (j) and from plants subjected to water de®cit from 5±13 DAP
(h). Nuclear counts were obtained by ¯ow cytometry. Means "SE of
six replicates are indicated.
1404
Setter and Flannigan
revealed that at 9 DAP, as well as later samplings, cell
cycling was substantially inhibited by water de®cit
(Fig. 3). In controls, DNA synthesis activity mg 1 DNA
was maximal at 9 DAP and decreased at 11 and 13 DAP,
while water de®cit inhibited activity 60, 80 and 88% at
9, 11 and 13 DAP, respectively. The decline from 9±11
and 13 DAP of DNA synthesis activity mg 1 DNA in
controls is an expected result of gradual exit from mitotic
cell cycling of a portion of the cell population.
Water deficit inhibition of endoreduplication
Although 3H-thymidine incorporation into DNA is
conventionally interpreted as a measure of cell mitotic
activity, in maize endosperm, DNA synthesis associated
with endoreduplication may also substantially contribute
to the observed 3H-thymidine incorporation. To distinguish treatment effects on mitosis versus those on endoreduplication, DNA content of nuclei was measured by
¯ow cytometry and nuclei were grouped based on DNAcontent. Nuclei with 3C and 6C DNA contents (where 1C
is the haploid DNA content) were considered mitotic
while those with G12C as endoreduplicated. This conservatively estimates the proportion of cells that have
advanced to an endoreduplication developmental status,
since some of the 6C nuclei might also have advanced
to an endoreduplicative state. In controls, the proportion
of nuclei with G12C DNA content steadily increased
from 9±13 DAP (Fig. 4). Water de®cit initially (at
9 DAP) increased the proportion of nuclei classi®ed as
endoreduplicated, then halted further increase in this
proportion so that at 13 DAP it was less than controls.
Hence, at 9 DAP, the water de®cit inhibition of
3
H-thymidine incorporation (Fig. 3) was apparently not
due to a decrease in the rate of endoreduplicative S-phase
cycling (Fig. 4). Instead, it was probably due to an initial
Fig. 3. Effect of water de®cit on DNA synthetic activity, estimated
with 3H-thymidine incorporation into DNA, of endosperms in developing maize kernels. Samples were obtained from well-watered control
plants (j) and from plants subjected to water de®cit from 5±13 DAP
(h). Data are expressed as the counts min 1 (CPM) of 3H-thymidine
incorporated mg 1 of total DNA. Means"SE of six replicates are
indicated.
inhibition of S phase in mitotic cells, before endoreduplication rate was affected. This allowed endoreduplication to advance the existing pool of nuclei to higher DNA
contents at 9 DAP. However, as the duration of stress
increased, both cell proliferation (Fig. 2) and endoreduplication (Fig. 4) were inhibited. Thus, the inhibition of
3
H-thymidine incorporation later, at 11 and 13 DAP
(Fig. 3), was from both decreased mitotic cell cycling, as
well as decreased endoreduplication.
Another possible effect of water de®cit on endoreduplication is the rate of progressive cycling between
successive rounds of S phase. To assess this, the proportion of nuclei in each of the DNA-content size classes
of endoreduplicating (G12C) nuclei were examined.
Controls had a steady, progressive increase in the proportion of nuclei in larger size classes (24C, 48C, and
96C), re¯ecting S-phase cycling of the nuclei in endoreduplication. If water de®cit inhibited S-phase cycling
of endoreduplicating nuclei, the proportion of nuclei in
the large size classes (24C, 48C, 96C) would decrease
relative to controls, whereas that in the smallest class
(12C) would increase relative to controls. The relative
proportions (Fig. 5) indicate that initially (at 9 DAP)
water de®cit did not inhibit S-phase cycling, permitting
the average size of the existing endoreduplicated nuclei
to increase. As the stress became progressively more
severe from 9±13 DAP, stress inhibited further S-phase
cycling so the proportion of large nuclei remained about
the same.
The current study is consistent with a previous study
of developing maize endosperm where water de®cit was
timed to coincide with either an early phase where the
majority of cells are mitotically cycling (1±10 DAP) or a
later phase where endoreduplication predominates (9±15
DAP) (Artlip et al., 1995). In that work, water stress at
Fig. 4. Effect of water de®cit treatments on the proportion of nuclei that
have undergone endoreduplication. Nuclei with DNA contents G12C
were summed to calculate the percent of nuclei in endoreduplication.
Samples were obtained from well-watered control plants (j) and from
plants subjected to water de®cit from 5±13 DAP (h). The number of
nuclei was determined using ¯ow cytometry and nuclear counts were
summed for each DNA-content size class, where 1C is the haploid
nuclear DNA content (approximately 2.7 pg for this maize genotype).
Means"SEM of six replicates are indicated.
Water deficit inhibits endosperm cell cycle
1405
from apical zones of the ear (Ober et al., 1991). Also,
although only apical-zone kernels were used in the present
work, in studies where apical and basal kernels were compared, inhibition of cell division and endoreduplication
was greatest in the apical zone (Artlip et al., 1995).
Water deficit inhibition of cell-cycle gene expression
Fig. 5. Effect of water de®cit on the proportional distribution of
endoreduplicated nuclei among each of the DNA-content size-classes
G12C, representing various stages of endoreduplication S-phase cycling.
Nuclei were analysed as in Fig. 4. Data are expressed relative to the total
number of endoreduplicating nuclei (G12C). Samples were obtained
from well-watered control plants (top panel) and plants subjected to
water de®cit from 5±13 DAP (bottom panel). Means"SE of six
replicates are indicated.
both the early and late phases substantially decreased
cell division in apical-kernel endosperms, whereas these
stresses decreased endoreduplication to a lesser extent,
and only in the stress from 1±10 DAP (Artlip et al.,
1995). Indeed, in that study, water stress from 9±15 DAP
increased the proportion of endoreduplicated nuclei
relative to well-watered controls due to a larger inhibition of mitotic cycling than endoreduplication. This is
similar to the transient increase in the proportion of
endoreduplicated nuclei at 9 DAP in the current study
(Fig. 4).
Other stresses also appear to have effects that change
temporally. Studies of in vitro-cultured maize kernels
indicated that although high temperature stress (35 8C
from 4±8 or 4 ±10 DAP) decreased the extent of both
endosperm cell division and endoreduplication in the
recovery phase; on the ®nal date of stress imposition, only
cell division was inhibited (Engelen-Eigles et al., 2000).
The present data are also consistent with studies
involving exogenous application of ABA to maize endosperm from 5±11 DAP (Mambelli and Setter, 1998). In
that work, mitotic cycling was inhibited 50% by 100 mM
ABA whereas transition to endoreduplication and endoreduplicative cycling were not inhibited until ABA concentrations were G300 mM. Such differential sensitivity
to ABA may play a role in the water de®cit effects
observed in the present study. Water de®cit increases
ABA levels in maize endosperm, particularly in kernels
To gain insight into the component processes of the cell
cycle that are affected by water de®cit, the expression of
genes which have recognized roles in various phases of the
cell cycle was examined. Four cDNAs were used as
hybridization probes in RNA gel blots of endosperm
samples: (1) maize a-tubulin (Montoliu et al., 1989),
(2) maize histone H3 (Chaubet et al., 1986), (3) a maize
CDK, ZmCdc2 (Colasanti et al., 1991), and (4) rice
(Oryza sativa L.) PCNA, a highly conserved component
of DNA polymerase complexes (Suzuka et al., 1991). As
shown in Fig. 6, at the ®nal sampling date (13 DAP)
water de®cit decreased the abundance, relative to 9 DAP
controls, of all four of the probed mRNAs. However, at
the earlier sampling dates (9 and 11 DAP), when cell
division was ®rst affected in this material, only ZmCdc2
abundance was signi®cantly (PF0.05) decreased by water
de®cit. The extent to which stress decreased ZmCdc2
mRNA abundance was similar to that for cell division
(Fig. 2). Furthermore, in controls, even though cell
division rates declined during the period from 9±13
DAP, the mRNA abundance of a-tubulin, histone H3
and PCNA remained high whereas ZmCdc2 decreased
substantially at 13 DAP, as expected for a gene product
whose expression is speci®c for cells engaged in cell
division.
Given that a- and b-tubulin subunits assemble to form
microtubules and that a large quantity of microtubules
are needed to form spindle ®bres and phragmoplasts
in dividing cells, the observed high level of a-tubulin
expression at 9 DAP (Fig. 6), when cells were predominantly mitotic (Fig. 4), was expected (Montoliu et al.,
1990). However, microtubules are also used in a variety
of other processes, such as in directing the deposition of
cellulose. This lack of speci®city to mitosis was re¯ected
in the continued high expression of a-tubulin in controls
from 9±13 DAP and the relative insensitivity of its
expression in response to stress at 9 and 11 DAP (Fig. 5).
Histone H3 is a structural component of chromatin
and PCNA has multiple functions in S phase as an
enhancer of DNA polymerase processivity and as a
component of regulatory complexes with cyclin and
S-phase CDK (Laquel et al., 1993; Tsurimoto, 1999).
Hence, these gene products are good markers for cells
actively synthesizing DNA (Fobert et al., 1994; Shimizu
and Mori, 1998). In well-watered endosperms, histone
H3 and PCNA expression remained high throughout
the observed period (Fig. 6), and, in agreement with
1406
Setter and Flannigan
Fig. 6. Effect of water de®cit on relative abundance of transcripts
encoding gene products involved in cell proliferation and endoreduplication. Samples were obtained from well-watered control plants (j) and
from plants subjected to water de®cit from 5±13 DAP (h). Levels of
RNA encoding ZmCdc2 (Colasanti et al., 1991), a-tubulin (Montoliu
et al., 1989), histone H3 (Chaubet et al., 1986), and PCNA (Suzuka
et al., 1991) in 20 mg of total RNA were determined by hybridization
of RNA gel blots with 32P-labelled cDNA probes. Each blot was
subsequently washed free of probe and rehybridized with ribosomal
RNA probe, pGMR from soybean, to establish that lanes were equally
loaded (data not shown). Signals on autoradiograms were quanti®ed by
laser densitometry and normalized with respect to the 9 DAP control
sample for each probe. Means"SEM of four replicates are indicated.
¯ow cytometry (Figs 2, 4, 5), water de®cit did not affect
their expression until later stages of stress when S phase
of both mitosis and endoreduplication were inhibited
(Fig. 6).
ZmCdc2 is a member of the family of CDKs that
contain the conserved PSTAIRE motif. It is homologous to Arabidopsis (Arabidopsis thaliana L.) CDC2aAt
and alfalfa (Medicago sativa L.) Cdc2MsA, and to rice
Cdc2Os1, with which it shares 94% amino acid sequence
identity (Dudits et al., 1998; Umeda et al., 1999). Studies
of the expression of various CDKs with respect to cell
cycle phases have indicated that PSTAIRE-containing
CDKs in rice (Cdc2Os1), Arabidopsis (CDC2aAt) and
alfalfa (Cdc2MsA) are predominantly expressed in actively dividing cells, but their expression is not speci®c to
a particular phase of the cell cycle (Magyar et al., 1997;
Segers et al., 1996; Umeda et al., 1999). Nevertheless,
in situ hybridization and ¯ow cytometric analyses of
vegetative shoot apices in Arabidopsis (Jacqmard et al.,
1999) and young tomato (Lycopersicum esculentum L.)
fruit (JoubeÁs et al., 1999) have shown that PSTAIREcontaining CDKs are restricted to mitotically dividing
cells and are not expressed in endoreduplicating cells.
The current data are consistent with this developmental
pattern of expression. In controls, ZmCdc2 expression
progressively decreased from 9±13 DAP (Fig. 6) while an
increasing proportion of cells became engaged in endoreduplication (Fig. 4). Expression of histone H3 and
PCNA, which are expressed in S phase of endoreduplicating cells, remained at high levels (Fig. 6). Furthermore,
at 9 DAP, water de®cit decreased ZmCdc2 expression
while the percentage of endoreduplicating cells increased.
As discussed above, water de®cit inhibited mitotic
cell cycling beginning at 9 DAP, and concomitantly
inhibited ZmCdc2 expression from 9±13 DAP (Fig. 6).
However, such inhibition was only partial, whereas cell
division was nearly halted (Fig. 2). This indicates that
decreases in ZmCdc2 transcript levels were not solely
responsible for decreased rates of mitotic cell cycling
during water de®cit. Additional regulation may be due
to post-transcriptional inhibition of CDK activity in
response to water de®cit. Studies have shown that
decreases in wheat leaf cell division in response to water
de®cit are associated with decreased activation state of
mitotic CDK, which is correlated with an increased
extent of its tyrosine phosphorylation (Schuppler et al.,
1998), a recognized mode of post-transcriptional CDK
regulation (Sun et al., 1999a).
Another possible mode of post-transcriptional regulation of CDK is ABA-induced expression of an inhibitor
of CDK, ICK1, whose interaction with Cdc2aAt and
cyclin-D3 decreases CDK activity (Wang et al., 1998).
ABA levels in maize endosperm increase substantially in
response to water de®cit (Ober et al., 1991), consistent
with the possibility such an inhibitor might play a role
in the current system.
The observed temporal separation of water de®cit
in¯uence on mitosis and endoreduplication suggests that
water de®cit down-regulates mitotic cell cycling and
endoreduplication via different mechanisms. In maize
endosperm, the transition of mitotic cells to an endoreduplicating status is accompanied by several changes
that might play a role in the developmental transition.
Endoreduplication is accompanied by (1) increases in
S-phase-related CDK activity (Gra® and Larkins, 1995)
associated with an increased phosphorylation of the G1uS
regulatory protein retinoblastoma (ZmRb) (Gra® et al.,
1996), (2) down-regulation of the G2uM cyclin transcript
CycZme1 (Sun et al., 1999b), (3) an increase in the level
of a mitotic CDK inhibitor (Gra® and Larkins, 1995),
and (4) an increased level of ZmWee1, a protein kinase
responsible for inhibitory tyrosine phosphorylation of
mitotic CDK (Sun et al., 1999a). Thus, the mechanisms
cited above as possible contributors to cell cycle arrest
during water de®cit and concomitant increase in ABA
levels (Schuppler et al., 1998; Wang et al., 1998), overlap
with those associated with endoreduplication. Further
studies are needed to elucidate the interplay of these
and other regulatory factors responsible for developmental down-regulation of G2uM during the transition to
endoreduplication and to distinguish them from those
Water deficit inhibits endosperm cell cycle
involved in stress-mediated inhibition of mitosis and
endoreduplication.
References
Artlip TS, Madison JT, Setter TL. 1995. Water de®cit in
developing endosperm of maize: cell division and nuclear
DNA endoreduplication. Plant, Cell and Environment 18,
1034±1040.
Cavallini A, Natali L, Balconi C, Rizzi E, Motto M, Cionini G,
D'Amato D. 1995. Chromosome endoreduplication in endosperm cells of two maize genotypes and their progenies.
Protoplasma 189, 156±162.
Cebolla A, Vinardell JM, Kiss E, OlaÂh B, Roudier F, Kondorosi
A, Kondorosi E. 1999. The mitotic inhibitor ccs52 is required
for endoreduplication and ploidy-dependent cell enlargement
in plants. The EMBO Journal 18, 4476±4484.
Chaubet N, Phillipps G, Chaboute ME, Ehling M, Gigot C. 1986.
Nucleotide sequences of two corn (Zea mays) histone H-3
genes. Genomic organization of the corn histone H-3 and H-4
genes. Plant Molecular Biology 6, 253±264.
Colasanti J, Tyers M, Sundaresan V. 1991. Isolation and
characterization of complementary DNA clones encoding a
functional p34-cdc-2 homologue from Zea mays. Proceedings
of the National Academy of Sciences, USA 88, 3377±3381.
Commuri PD, Jones RJ. 1999. Ultrastructural characterization
of maize (Zea mays L.) kernels exposed to high temperature
during endosperm cell division. Plant, Cell and Environment
22, 375±385.
Davis LG, Dibner MD, Battey JD. 1986. Rapid DNA
preparation. In: Basic methods in molecular biology.
Amsterdam: Elsevier Science Publishing, 42± 43.
den Boer BGW, Murray JAH. 2000. Triggering the cell cycle in
plants. Trends in Cell Biology 10, 245±250.
Dudits D, Magyar Z, DeaÂk M, MeÂszaÂros T, Miskolczi P, FeheÂr
 , Athanasiadis A, Pongor S, Bako L,
A, Brown S, Kondorosi E
Koncz Cs, GyoÈrgyey J. 1998. Cyclin-dependent and calciumdependent kinase families: response of cell division cycle to
hormone and stress signals. In: Francis D, Dudits D, Inze D,
eds. Plant cell division. London: Portland Press Research
Monograph, 21±45.
Engelen-Eigles G, Jones RJ, Phillips RL. 2000. DNA endoreduplication in maize endosperm cells: the effect of exposure
to short-term high temperature. Plant, Cell and Environment
23, 657±663.
Fobert PR, Coen ES, Murphy GJP, Doonan JH. 1994. Patterns
of cell division revealed by transcriptional regulation of genes
during the cell cycle in plants. The EMBO Journal 13,
616±624.
Gra® G. 1998. Cell cycle regulation of DNA replication: the
endoreduplication perspective. Experimental Cell Research
244, 372±378.
Gra® G, Burnett RJ, Helentjaris T, Larkins BA, DeCaprio JA,
Sellers WR, Kaelin Jr WG. 1996. A maize cDNA encoding
a member of the retinoblastoma protein family: Involvement
in endoreduplication. Proceedings of the National Academy
of Sciences, USA 93, 8962±8967.
Gra® G, Larkins BA. 1995. Endoreduplication in maize
endosperm: involvement of M phase-promoting factor inhibition and induction of S phase-related kinases. Science 269,
1262±1264.
Jacqmard A, De Veylder L, Segers G, de Almeida Engler J,
Bernier G, Van Montagu M, Inze D. 1999. Expression of
CKS1At in Arabidopsis thaliana indicates a role for the
1407
protein in both the mitotic and the endoreduplication cycle.
Planta 207, 496±504.
Jones RJ, Schreiber BMN, Roessler JA. 1996. Kernel sink
capacity in maize: genotypic and maternal regulation. Crop
Science 36, 301±306.
JoubeÁs J, Phan T-H, Just D, Rothan C, Bergounioux C,
Raymond P, Chevalier C. 1999. Molecular and biochemical
characterization of the involvement of cyclin-dependent
kinase A during the early development of tomato fruit.
Plant Physiology 121, 857±869.
Kowles RV, Srienc F, Phillips RL. 1990. Endoreduplication of
nuclear DNA in the developing maize endosperm.
Developmental Genetics 11, 125±132.
Kowles RV, McMullen MD, Yerk G, Phillips RL, Kraemer S,
Srienc F. 1992. Endosperm mitotic activity and endoreduplication in maize affected by defective kernel mutations.
Genome 35, 68±77.
Kowles RV, Yerk GL, Haas KM, Phillips RL. 1997. Maternal
effects in¯uencing DNA endoreduplication in developing
endosperm of Zea mays. Genome 40, 798±805.
Laquel P, Litrak S, Castroviejo M. 1993. Mammalian proliferating cell nuclear antigen stimulates the processivity of two
wheat embryo DNA polymerases. Plant Physiology 102,
107±114.
Lemontey C, Mousset-DeÂclas C, Munier-Jolain N, Boutin J-P.
2000. Maternal genotype in¯uences pea seed size by controlling both mitotic activity during early embryogenesis and ®nal
endoreduplication levelucotyledon cell size in mature seed.
Journal of Experimental Botany 51, 167±175.
Lur HS, Setter TL. 1993. Role of auxin in maize endosperm
development. Timing of nuclear DNA endoreduplication, zein
expression, and cytokinin. Plant Physiology 103, 273±280.
Magyar Z, MeÂszaÂros T, Miskolczi P, DeaÂk M, FeheÂr A,
Brown S, Kondorosi EÂ, Athanasiadis A, Pongor S, Bilgin M,
Bako L, Koncz C, Dudits D. 1997. Cell cycle phase speci®city
of putative cyclin-dependent kinase variants in synchronized
alfalfa cells. The Plant Cell 9, 223±235.
Mambelli S, Setter TL. 1998. Inhibition of maize endosperm cell
division and endoreduplication by exogenously applied
abscisic acid. Physiologia Plantarum 104, 266±277.
Montoliu L, PuigdomeÁnech P, Rigau J. 1990. The Tuba3 gene
from Zea mays: structure and expression in dividing plant
tissues. Gene 94, 201±207.
Montoliu L, Rigau J, PuigdomeÁnech P. 1989. A tandem of
a-tubulin genes preferentially expressed in radicular tissues
from Zea mays. Plant Molecular Biology 14, 1±15.
Nicolas ME, Gleadow RM, Dalling MJ. 1985. Effect of
post-anthesis drought on cell division and starch accumulation in developing wheat grains. Annals of Botany 55,
433±444.
Ober ES, Setter TL, Madison JT, Thompson JF, Shapiro PS.
1991. In¯uence of water de®cit on maize endosperm development. Enzyme activities and RNA transcripts of starch
and zein synthesis, abscisic acid, and cell division. Plant
Physiology 97, 154±164.
Olsen O-A, Linnestad C, Nichols SE. 1999. Developmental
biology of the cereal endosperm. Trends in Plant Science 4,
253±257.
Ouattar S, Jones RJ, Crookston RK 1987. Effects of moisture
stress during grain ®lling on the pattern maize kernel growth
and development. Crop Science 27, 726±730.
Roberts JM. 1999. Evolving ideas about cyclins. Cell 98,
129±132.
Sambrook J, Fritsch EF, Maniatis T. 1989. Molecular cloning: a
laboratory manual, 2nd edn. Plainview, New York: Cold
Spring Harbor Laboratory Press.
1408
Setter and Flannigan
Schuppler U, He P-H, John PCL, Munns R. 1998. Effect of
water stress on cell division and cell-division-cycle 2-like
cell-cycle kinase activity in wheat leaves. Plant Physiology
117, 667±678.
Segers G, Gadisseur I, Bergounioux C, de Almeida Engler J,
Jacqmard A, Van Montagu M, Inze D. 1996. The Arabidopsis
cyclin-dependent kinase gene cdc2bAt is preferentially
expressed during S and G2 phases of the cell cycle. The
Plant Journal 10, 601±612.
Shimizu S, Mori H. 1998. Analysis of cycles of dormancy and
growth in pea axillary buds based on mRNA accumulation
patterns of cell cycle-related genes. Plant Cell Physiology
39, 255±262.
Sun Y, Dilkes BP, Zhang C, Dante RA, Carneiro NP,
Lowe KS, Jung R, Gordon-Kamm WJ, Larkins BA. 1999a.
Characterization of maize (Zea mays L.) Wee1 and its
activity in developing endosperm. Proceedings of the
National Academy of Sciences, USA 96, 4180±4185.
Sun Y, Flannigan BA, Setter TL. 1999b. Regulation of endoreduplication in maize (Zea mays L.) endosperm. Isolation of
a novel B1-type cyclin and its quantitative analysis. Plant
Molecular Biology 41, 245±258.
Suzuka I, Hata S, Matsuoka M, Kosugi S, Hashimoto J. 1991.
Highly conserved structure of proliferating cell nuclear
antigen DNA polymerase delta auxiliary protein gene in
plants. European Journal of Biochemistry 195, 571±576.
Tsurimoto T. 1999. PCNA binding proteins. Frontiers in
Bioscience 4, 849±858.
Umeda M, Umeda-Hara C, Yamaguchi M, Hashimoto J,
Uchimiya H. 1999. Differential expression of genes for
cyclin-dependent protein kinases in rice plants. Plant
Physiology 119, 31±40.
Wadsworth GJ, Redinbaugh MG, Scandalios JG. 1988. A
procedure for the small-scale isolation of plant RNA suitable
for RNA blot analysis. Analytical Biochemistry 172, 279±283.
Wang H, Qi Q, Schorr P, Cutler AJ, Crosby WL, Fowke LC.
1998. ICK1, a cyclin-dependent protein kinase inhibitor from
Arabidopsis thaliana interacts with both Cdc2a and CycD3,
and its expression is induced by abscisic acid. The Plant
Journal 15, 501±510.