|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
MORPHOLOGICAL SURVEY AND CHARACTERIZATION OF PROGRAMMED
CELL DEATH IN THE PLACENTA-CHALAZA AND ENDOSPERM IN THE
DEVELOPING CARYOPSIS OF MAlZE
A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE
U NIVE RS ITY OF FLORI DA
This work is dedicated to my family and friends who have been tremendous
pillars of aid and support
I would like to acknowledge and thank my chairman, Dr. Prem S. Chourey,
for his support and for giving me a great opportunity to explore research and
learn invaluable research techniques. I also would like to thank my committee,
Dr. Daryl Pring, Dr. Christine Chase, and Dr. Henry Aldrich, for their support and
valuable advice. Special thanks go to AleS Kladnik whose research and efforts
were indispensable in pointing me in the direction that allowed me to pursue my
master's degree. I also thank Maureen Petersen, Donna Williams, and Lorraine
McDowell for their complete technical support in electron microscopy and Tim
Vaught for his aid in helping me with confocal microscopy. Special support and
aid came from close friends whom I would like to give a special thanks to:
Melanie Cash, Shayna Sutherland, Joyce Merritt, Asha Brunings, Wayne Jurick,
Fabricio Rodrigues, and the late Richard Wheeler who always saw the sunny
side. Very special thanks go to Donna Perry, Gail Harris, and Lauretta Rahmes
in Plant Pathology and Melissa Webb and Brandy Burgess in the plant molecular
and cellular biology program for providing administrative help. Lastly, I'd like to
thank my husband, Larry Chamusco; parents, Fred and Cleo Courington; and
sister, Becky Lockridge who have been loving and supportive in countless ways.
TABLE OF CONTENTS
ACKNOWLEDGMENTS ................. ...............iv........_ .....
LIST OF FIGURES ................. ...............vii........ .....
ABSTRACT ................. ...............ix........_ .....
1 INTRODUCTION................ ............. 1
Programmed Cell Death................ ................ 1
Categories of PCD Described in Plants ......... ................. ................. 5
Physiologic and Molecular Factors Involved in PCD ........._.._... .........._..... 7
Comparing PCD in Plants and Animals................ ................ 9
The Role of Calcium in PCD .........._._........____........ ...........1
The Role of Endonucleases ....__ ......_____ .......___ ............1
Seed Development................ .............. 15
2 MATERIALS AND METHODS ............. ...... .__ .....___............2
Plant M materials ............. ...... ._ ............... 20....
Utrastuctural Studies ............. ......_ ............... 20..
Calcium Trapping ............. ...... ._ ............... 22....
Light M icroscopy ............. ...... ._ ............... 23....
Nuclease Activity Assays ............. ...... ._ ....._ _.............2
DNA Extraction................ .............. 27
3 RESULTS ............. ...... ._ ............... 29....
The Endosperm................ ............... 29
Light M icroscopy.......................... ......... 29
Transmission Electron Microscopy ............. ...... .__ ............... 30
The Placenta-Chalaza Region ............. ...... .__ .....___............3
Light M icroscopy................ ... ................... 31
PCD in the PC Region is Fertilization Dependent................ ............... 32
Transmission Electron Microscopy ............. ...... .__ ............... 33
Calcium Trapping................ ............... 37
In-Gel Nuclease Activity Assay ....._._.__ ......._.. ... ...._._......... 39
Genomic DNA Extraction ........._...... .....___......... ............4
4 DISCUSSION............... ............... 82
PCD in the PC Region .........._..... ... ....... .. .... ... ..... ..........8
PCD in the PC May Be Tied to Growth and Development of the Filial
Tissues ........._.... .. ..... ... ................ 84....
Sink Demand and Assimilate Flow.................... ....... ................. 86
Two Separate, Temporal Waves of PCD Occur in the PC and May Be
Linked to the Sink Demand of the Growing Filial Tissues............._._... ..... 87
The Role of Calcium in PC PCD ........._._... ........ .......___..........9
The Role of Endonucleases in the PC ............. _.....___ .......... .... 94
PCD in the endosperm ............. .....__ ............... 96..
5 CONCLUSION ............. ...... ._ ............... 100...
LIST OF REFERENCES ............. .....__ ............... 103..
BIOGRAPHICAL SKETCH ............. ...... ._ ....___ ............10
LIST OF FIGURES
1 Classic apoptotic cellular morphology ....._____ ... .... .___ ............... 18
2 Classic necrotic cellular morphology. ..........._._ ....._.._ .........._..... 19
3 Schematic drawing illustrating the regions of the maize endosperm.......... 41
4 Nuclear morphological changes in the endosperm of maize...................... 43
5 PCD in the endosperm of a 24 DAP caryopsis. ........._.._.. ........._.._..... 44
6 Confocal image of the nuclei in the central crown region of the
endosperm of a 28 DAP caryopsis of maize. ...........__... .......__........ 45
7 Nuclear morphology of the central crown region of the endosperm of a 12
DAP maize caryopsis. ...._ ......_____ .......___ .............4
8 Nuclear Morphology of the central crown region of the endosperm of
a 14 DAP maize caryopsis. ...._.._.._ ... .. ...__. ...._.._ ...........4
9 Nuclear Morphology of the central crown region of the endosperm of a
16 DAP maize caryopsis. ...._ _. ...._.._.._ ......._.. ............4
10 Nuclear Morphology of the central crown region of the endosperm of a
20 DAP maize caryopsis. ...._.._.._ ......_._._ ...._.._ ............4
11 Nuclear morphology of the central crown region of the endosperm of a
22 DAP caryopsis of maize. ........._._.. ....__.._ ...._ ... ...........5
12 Nuclear Morphology of the central crown region of the endosperm of a
25 DAP caryopsis of maize. ........._._.. ....__.._ ...._ ... ...........5
13 Origin of the Placenta-chalaza (PC) region. ........._...... ....._.._............ 52
14 Anatomy of the Placenta-chalaza. ........... ..... ..__ ........... .... 53
15 Placenta-chalaza (PC): left-right borders ................. ....................... 55
16 TEM image of one of the boundaries of the PC region. ........................... 56
17 Placenta-chalaza boundaries. ...._ ......_____ .......___ .............5
18 4'6-diam idino-2-phenylindole-2HCI (DAPI) DNA stain. .........._... ............. 58
19 Crystal Violet Stain showing condensed chromatin in the nuclei of
the cells of the integument-derived PC region from 4 through 10 DAP...... 59
20 Crystal Violet Stain showing condensed chromatin in the nuclei of the
integument-derived PC region from 12 through 16 DAP..............._._......... 60
21 TUNEL assays of PC region showing broken or nicked DNA in the nuclei of
the integument-derived portion of the PC ........._._. .........._._............ 62
22 Crystal violet stain showing condensed chromatin in the integument
derived portion of the PC region of a 24 DAP caryopsis..............._._......... 63
23 TEM micrographs illustrating consecutive developmental stages and
subsequent cell death of the nucellus-derived PC region. ........._.._............ 64
24 TEM micrographs illustrating the plasmodesmata (PD) in the walls of the
nucelluc-derived PC cells. ...._.._.._ ... ... ..._.._......_ .............6
25 Development and subsequent death of upper integument-derived region
of PC. ........._.._.. ...._..._ ..............._ 66...
26 Development and subsequent death of lower integument-derived region of
PC. ........._... ...... ___ ..............._ 68...
27 Lower Integument-derived PC Cells. ........._.._.. ...._.._ ........_....... 69
28 Nuclear body formation in the integument-derived portion of the
PC cells. ........._._. ._......_.._ ..............._ 70...
29 Nuclear morphological changes in integument derived portion of the PC.. 71
30 Ca2+ Trapping: Nucellus-derived PC. ...._._._._ ....... .___ ........_.._.... 72
31 Ca2+ Trapping in the integument-derived PC. ........._ .. ..... ...._._......... 74
32 Ca2+ trapping in the normal appearing, lower integument PC. ........._.._..... 76
33 In-gel nuclease activity assays for Ca2+, Mg2+ and Zn2+-dependent
nucleases and pH controls. ...._ .. ...._._._._ .........__. ............7
34 In-gel nuclease activity assays with EDTA and EGTA and no metals........ 80
35 Genomic DNA isolated from the PC region. ......____ ....... ....__.......... 81
Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science
MORPHOLOGICAL SURVEY AND CHARACTERIZATION OF PROGRAMMED
CELL DEATH IN THE PLACENTA-CHALAZA AND ENDOSPERM IN THE
DEVELOPING CARYOPSIS OF MAlZE
Karen C. Chamusco
Chair: Prem S. Chourey
Major Department: Plant Molecular and Cellular Biology
Programmed cell death (PCD) was examined on a morphological basis
employing both light and electron microscopy techniques. The tissues examined
were the placenta-chalaza (PC) region and maturing endosperm of the maize
caryopsis. The PC region is sporophytic and, in maize, composed of two distinct
types of cells, which undergo two distinct morphological changes as they die.
This cell death process is fertilization dependent and takes place between 6 and
12 days after pollination (DAP). The first type of cells in the PC is derived from
the nucellus and undergoes a rapid cell death leaving behind cell corpses and
the number of detectable plasmodesmata decreases. There is no detectable
DNA cleavage via terminal deoxynucleotidy transferase dUTP nick end label
(TUNEL) assay in this zone at any time.
The second type of cells in the PC is derived from the inner integument and
undergoes a PCD that, morphologically, resembles apoptosis, i.e., condensed
chromatin, nuclear fragmentation, and preservation of organelles and
membranes until the end of the death process. There is no loss of
plasmodesmata in these cells. These cells are TUNEL-positive indicating DNA
Intracellular calcium [Ca2+]i W8S localized in an attempt to determine the
general flux and location of free [Ca2+]i 8n8 geerl [C2+]i aCCUmU tiOn in
storage sites such as the endoplasmic reticulum and the central vacuole does
not appear to take place until after 5 DAP. Mitochondria, while appearing to
collect [Ca2+]i St Some point prior to cell death, do not appear to change
morphologically. The general trend of [Ca2+]i f UX can be summed up as follows:
[Ca2+]i iS collected and stored in cellular compartments and organelles, released
into the cytoplasm and then dispersed or removed from the cell after death.
Nuclease assays revealed three nucleases dependent upon Ca2+ andlor
magnesium (Mg2+) and a fourth nuclease dependent on zinc (Zn2+). The peak of
activity for these nucleases occurred between 8 and 12 DAP which coincide with
the peak of morphological changes and the peak of positive TUNEL activity.
Unlike the nuclear changes seen in the PC cells, the nuclei of the central
crown region of the endosperm cells undergo a series of morphological changes
as the endosperm develops. At 12 DAP the nuclei start out spherical and the
DNA diffuse, by 16 DAP expand in size while chromatin condenses at the
nuclear envelope. Finally, by 25 DAP, the nuclei collapse and condense tightly.
These data indicate the possibility for three unique PCD processes
occurring in the developing caryopsis of maize.
Programmed Cell Death
Programmed cell death, or PCD, is a physiologically controlled, carefully
timed, energy dependent cell death program that organisms use to aid in
defensive processes or depend on for many normal developmental processes
(Solomon et al., 1999; Earnshaw, 1995).
Although cell death as a physiological process has been known to exist for
the last 150 years, apoptosis was the first type of PCD characterized in 1972 in
animal cells by Kerr and was called apoptosis from the Greek meaning "fallen"
and has been extensively studied ever since (Bursch et al., 2000; Earnshaw,
1995; Mllerriam Webster's Dictionary, 2004). Plant biologists have used these
animal studies as spring-boards from which to investigate PCD in plants.
Since the apoptotic form of PCD is the longest studied and best
characterized it will be discussed, here, as a model to illustrate the PCD process
and for use as a comparison against plant PCD. But it must be noted that
apoptosis is but one way PCD can occur. Other types of PCD that have been
described include autophagic PCD, a PCD which involves the degradation of
intracellular organelles prior to nuclear degradation, and endoplasmic reticulum
(ER)-mediated PCD that is independent of apoptosis protease activating factor-1
(apaf 1) and cytochrome c (Bursch et al., 2000; Rao et al., 2004).
A cell may be induced to undergo PCD due to a variety of causes such as
pathogen invasion, if the lifetime and or function of the cell has ended, or normal
developmental processes (Krishnamurthy et al., 2000). Apoptosis, like all
programmed cell death processes, is, by definition, a genetically controlled,
energy dependent and does not induce inflammation (Earnshaw, 1995). In the
case of apoptosis one of the first responses a cell has after perceiving signals to
undergo cell death is to produce an asymmetry in the aminophospholipids across
the plasma membrane. This duty requires the aid of the membrane bound
enzymes floppase, scramblase and the calcium (Ca2+) inhibition of
aminophospholipid translocase (Reutelingsperger 1997). The two former
enzymes induce an aminophospholipid flip, specifically, phosphatidylserine,
whereby the outer leaflet of the membrane has a higher occurrence of serine
residues than it normally would have. Once the serine residues are exposed,
Annexin V, a eukaryotic conserved aminophospholipid binding protein with high
affinity for serine residues, binds them in either a Ca2+ Or low pH dependent
manner. The bound Annexin V proteins then undergo Ca2+ induced structural
changes, ultimately forming an integral transmembrane channel that is thought to
act in an amphipathic or polytopic manner (Reutelingsperger 1997; Li and
Piazza, 2002). The mitochondria play a role at this point which includes a
release of intermembrane space proteins such as procaspase 9, cytochrome c,
apoptosis inducing factor (AlF), Ca2+ and reactive oxygen species (ROS). It has
been shown that mitochondria are intimately involved in regulating both plant and
animal PCD events (Arnoult et al., 2002; Curtis and Wolpert, 2002; Balk et al.,
2003). Mitochondrial inner and outer membrane-spanning pore formation is a
critical part of mitochondrial PCD effectors being released. Studies focusing on
apoptosis in animal cells have revealed two main mechanisms for the formation
of these pores and, consequently, each results in different mitochondrial
morphology (Yu et al., 2002). The first mechanism is a direct one and involves
the voltage-dependent anion channel, or VDAC, located at the mitochondrial
outer membrane. This mechanism is proposed to use the pro-apoptotic, Bcl-2
family protein, Bax, to interact with and tightly regulate release of cytochrome c
via the VDAC channel. The morphological consequence of this is that the
mitochondria remain normal in appearance.
The second mechanism is an indirect mechanism and involves the
permeability transition pore, or PTP (Curtis and Wolpert, 2002; Yu et al., 2002).
The PTP is formed from a complex of proteins, primarily adenine nucleotide
(ANT) located on the inner membrane, VDAC located on the outer membrane,
hexokinase, and cyclophilin D (CpD) in the matrix and when induced to interact
they form a transient, multi-faceted pore that acts as a redox, ion, pH, and Ca2+
gated channel. Upon continual opening solutes up to 1.5 kDa leak out of the
mitochondrial matrix and an uncontrolled influx of solutes, especially Ca2+, into
the matrix is initiated. Calcium increases the inner membrane permeability; the
result is gross over-swelling of the matrix, loss of transmembrane potential, an
uncoupling of the electron transport chain, superoxide production and complete
disruption of the outer membrane, and thus the release of inter-membrane
space, PCD promoting solutes such as apoptosis inducing factor (AlF),
cytochrome c, procaspase 9 and Ca2+ (CUrtiS and Wolpert, 2002; Arnoult et al.,
2002; Balk et al., 2003; Yu et al., 2002).
Once in the cytosol procaspase 9 autoprocesses to caspase 9 in
cooperation with cytochrome c and apoptosis protease activating factor-1, or
apaf-1, to form the apoptosome complex which further aids in activating effector
caspases. Caspases are cysteinyl-aspartate proteases. One such target of
these effector caspases is caspase activated DNase (CAD) which, along with
other Ca2+, C2+/Mg2+ dependent endonucleases, and caspase independent AlF
acts to condense and cleave DNA into fragments of 300 and 50 kilobases (kb)
then into internucleosomal size fragments which occur in multiples of 180 base
pairs (bp). (Earnshaw,1995; Gromova et al., 1995; Krishnamurthy et al., 2000;
Woltering et al., 2002; Balk, 2003).
A cascade of reactions is initiated at the plasma membrane by initiator
caspase 8, which cleaves other cytosolic effector caspases such as 3, 6 and 7
thus activating them. Poly-(ADP ribose)-polymerase, or PARP, which is involved
in DNA repair and maintenance, and the RNA splicing enzyme U1-snRNP are
other substrates of caspases; however these proteins are cleaved for inactivation
purposes so the apoptotic process may be allowed occur (Krishnamurthy et al.,
2000; Sun et al., 1999; Woltering et al., 2002, Chang and Yang, 2000).
On a morphological basis apoptotic animal cells show a unique pattern of
chromatin condensation (Fig 1) delineated by sharp edges, marginalization of
chromatin, retention and granulation of the nucleolus, plasma and nuclear
membrane blebbing (an outward bulging of the membrane) and the formation of
apoptotic bodies containing degraded cytoplasmic and nuclear contents. In
addition, all membranes and organelles are retained until the end of the cell
death program when neighboring cells phagocytose the dying cell (Kerr et al.,
1995). However the mitochondria may exhibit swelling due to high Ca2+ COntent
in the matrix (Virolainen et al., 2002; Curtis and Wolpert, 2002). Additionally,
apoptosis occurs in individual cells and is typically asymmetric with regard to one
cell's stage of PCD compared with that of another (Kerr et al., 1995). What is
described above is but one means by which cells can undergo a genetically
programmed death thus the process of PCD is not a clear cut one but one in
which it occurs in wide array of ways, especially in plants (Krishnamurthy et al.,
2000; Earnshaw, 1995; Rao et al., 2004)
Necrosis, on the other hand, is a form of cell death that has been reported
to be neither genetically controlled nor energy dependent (Fig 2). This form of
death is caused by cell damage, occurs in masses or clusters of cells and, in
animals, causes inflammation. The morphology does not resemble apoptosis type
PCD in that plasma membrane dissociation is first followed by complete
degradation of the cell. There is no organized chromatin condensation in the
nucleus and no preservation of intracellular contents. (Earnshaw, 1995; Kerr et
al.,1995). However, recent studies dispute necrosis as a passive, non-
genetically controlled cell death, rather, a physiologically controlled process
(Proskuryakov et al., 2002).
Categories of PCD Described in Plants
Plants have been shown to display PCD in for up to approximately six
different processes: hypersensitive response (HR), cells that have served their
purpose and are no longer required, cells that must die to serve their purpose,
cells initiated in development but never used, cells present in wrong places and
damaged cells (Krishnamurthy et al., 2000). The hypersensitive response (HR)
is initiated by pathogen attack, abiotic stresses such as ozone exposure creating
oxidative stresses, incompatible temperatures and osmotic, water, or light
stresses. Hormones such as gibberellic and salicylic acids have been shown to
play critical roles in HR (Krishnamurthy et al., 2000; Jones and Dangl, 1996) HR
does not necessarily end in the death of the cell, but when it does HR is often
described as necrotic. Regardless of whether or not the cell death appears
necrotic it does still involve genetic control and is localized to one or a few cells
at the immediate site of the pathogen attack. For example disease resistance
genes, (R) genes, have been implicated in mediating cell death signals during
HR. R genes are a single gene in the host plant encoding a receptor that
recognizes a single elicitor gene product called an avirulence (Avr) protein from a
specific pathogen resulting in resistance. The fundamental R protein architecture
typically includes a nucleotide binding domain (NB), a leucine rich repeat (LRR),
and serinelthreonine kinase unit. There are several subclasses of R genes in
which protein sequence alignments have revealed similarity in architecture to
several other protein receptors in the animal kingdom that are known to be
involved in apoptosis such as interleukin-1, apaf-1, and CED-4. There is,
however, still much that is unknown about the mechanisms underlying HR
mediated cell death (Shirasu and Schulze-Lefert, 2000).
The category of PCD in which cells have served their purpose and are no
longer required include, but are not limited to: senescence, cells involved in
abscission, synergids, antipodal cells, tapetal cells, and endosperm cells. PCD
in this category is part of the normal developmental processes that occur in
plants (Krishnamurthy et al., 2000).
Cells that must die to serve their function include xylem tracheary elements
(TE), cork cells, sclerenchyma, and modified structures such as spines and
thorns (Yu et al., 2002; Obara, 1998; Krishnamurthy et al., 2000; Jones, 1996).
Cells initiated in development but never used are those of non-functioning
megaspore cells, stamen primordia in female flowers and carpel primordia in
male flowers. (Krishnamurthy et al., 2000; Jones 1996).
Examples of cells present in wrong places are cells that die to give rise to
leaflets of compound leaves and holes in simple leaves (Arunika et al., 2004;
Krishnamurthy et al., 2000; Jones, 1996).
Lastly is the damaged cells category. This category is still unresolved as
damaged cells may also undergo necrosis, a non-programmed cell death.
Damaged cell PCD appears to behave much like that of the HR PCD contributing
further to the lack of resolution (Krishnamurthy et al., 2000).
Physiologic and Mlolecular Factors Involved in PCD
Reactive oxygen species (ROS) such as hydrogen peroxide (H202) and
superoxide (02 ) are 8 So involved in mediating signals for PCD. ROS generation
during HR is suspected to be produced by a handful of enzymes such as the
Ca2+ activated plant homologue of the NADPH oxidase found in animals and
oxalate oxidase in extracellular spaces during HR (Neill, et al., 2002). ROS have
been shown to occur in two distinct phases and may act in a dose dependent
manner. The first wave of ROS's occurs within an hour upon virulent or avirulent
pathogen attack whereas the second wave of ROS is specific to the avirulent
pathogen, occurs approximately five hours later and is sustained. High ROS
levels induce PCD but low levels stimulate antioxidant enzymes such as
gutathione S-transferase and glutathione peroxidase. High levels of ROS are
suspected to act as immediate pathogen stopping agents by killing the cell at the
site of infection as the ROS diffuse to surrounding cells. H202 has also been
shown to stimulate mitogen activated protein kinases (MAPK) in addition to
mitochondria. The mitochondria, in turn, generate more H202 al GE 09
mitochondrial activity thus leading to PCD. (Shirasu and Schulze-Lefert, 2000;
Thordal-Christensen et al., 1997; Neill et al., 2002).
The plant hormone salicylic acid (SA) has been shown to play a signaling
role in helping mediate HR PCD and systemic acquired resistance (SAR). It is
thought that SA coordinates with ROS to amplify, sustain and define the
threshold that must be met to induce PCD. Experiments where SA production
was compromised resulted in an inhibition of ROS production and a delay of
several hours in HR mediated cell death (Shirasu and Schulze-Lefert, 2000).
Nitric oxide (NO) is also a signal agent involved in mediating PCD in plants
and animals. Upon pathogen attack and subsequent R gene dependent HR NO
synthase activity is stimulated. The generation of NO stimulates phenylalanine
ammonia lyase (PAL), an enzyme involved in SA synthesis, and pathogen
resistance genes. Like SA, NO appears to help regulate ROS in a coordinated
fashion by dramatically amplifying the effects of H202 On cell death (Shirasu and
Gibberellic acid (GA) also plays a role in plant PCD. Fath et al. (2000)
demonstrated that isolated barley aleurone layers undergo PCD upon GA
stimulation. GA stimulated the release and de novo synthesis of hydrolases and
nucleases in barley aleurone cells. GA also stimulated the coalescing of protein
storage vacuoles (PSV) within the aleurone cells. When protein synthesis and
secretion is finished the PSV's turn into lytic vacuoles and proceed to digest the
cellular contents while glucanases and xylanases digest the cell walls.
Ethylene is another player in PCD in plants. Young et al. (2000) reported a
spike in ethylene production prior to cell death in the endosperm of maize. It is
thought that ethylene is involved in nuclease activation and regulation.
Comparing PCD in Plants and Animals
In doing a side by side comparison of PCD in plants with apoptosis in
animals it can be seen that there are some surprising similarities regardless of
the method of PCD plants use; however, there have been distinct differences
discovered between the two as well.
Physiologic and molecular similarities include, but not necessarily seen in
every instance, a phosphatidylserine flip from the inner to the outer leaflet of the
plasma membrane, Annexin V binding, calcium signaling, mitochondrial release
of cytochrome c, and DNA cleavage resulting in internucleosomal size fragments
occurring in multiples of 180 bp, protein phosphorylation/dephosphorylation, ROS
production and Ca2+ and Zinc ( Zn2+) -dependent endonucleases (Young et al.,
1997; Wojciechowska, 2003; Ning et al, 2002; Krishnamurthy et al., 2000).
Plants do not necessarily use the components they do share with animals for
PCD, such as cytochrome c. For example cytochrome c release is detected
during PCD for TE differentiation but it does not participate in the PCD action
itself (Yu et al., 2002)
There is, however, a notable difference between plant apoptotic-like PCD
and animal apoptosis. The system described, previously, for apoptosis has not
been found in plants. For example, caspases have not been identified in plants
nor has a caspase cascade reaction; however, in some instances, caspase-like
cysteine proteases found in plants have been reported to behave similarly to true
animal caspases during plant PCD. Unlike caspases, plant caspase-like
proteases do not all have a specificity for aspartate residues. Some plant
cysteine proteases have been implicated in plant PCD; such is the case for the
barley vacuole localized aspartate proteinase, Phytepsin D, which is highly
expressed during autolysis of tracheary element differentiation (Runeberg-Roos
and Saarma, 1998). Previous studies have shown these caspase-like proteases
to be inh ibited by the synthetic caspase inh ibitors choloro-m ethylketone (Ac-
YVAD-CMK) andlor aldehyde (Ac-YVAD-CHO). These inhibitors completely
prevented HR mediated PCD in tobacco cells (Krishnamurthy et al., 2000).
Korthout et al., (2000) have shown caspase 3-like activity in plant cells. They
found that by adding plant mitochondria to the cytosol of Xenopus eggs in a cell
free system they were able to induce caspase 3 activity. Additionally, when a
mixture of protease inhibitors and the caspase 3 specific synthetic substrate N-
acetyl-Asp-Glu-Val-Asp-7-am ino-4-methylcoumarin (Ac-DEVD-AMC) was added
to cytosolic extracts from barely embryonic cell cultures the substrate was
cleaved after the aspartate residue. Only when a caspase 3 specific inhibitor,
Ac-DEVD-fmk, was added to the mix did the caspase 3 substrate not become
Morphological similarities include the following but are not necessarily seen
in every instance: cytoplasmic and nuclear shrinkage, chromatin condensation,
and in this study, nuclear body formation.
One unique difference between animal and plant PCD is that a cell corpse
is usually left behind after plant PCD due to the cell wall and cytosolic
degradation occurring autophagically andlor autolytically rather than by
phagocytosis as it does in animal PCD. However there are some exceptions
such as those seen for the formation of holes in the leaves of the lace plant and
the degeneration of barley aleurone cells (Arunika et al., 2004; Fath et al., 2000).
The Role of Calcium in PCD
It seems that regardless of the kingdom an organism is from, Ca2+ 8ppe8FS
to be a universal role player in PCD. Under normal, or non-stressed, conditions
the basal levels of free cytosolic Ca2+ (C2+]i) in plant cells are kept very low
ranging from 100 to 200 nM (Buchanan et al., 2000). Certain organelles in these
cells can be used as storage sites for Ca2+. Low cytosolic concentrations of Ca2+
are maintained with the aid of membrane bound Ca2+ channels and transporters
such as Ca2+-ATPase and Ca2+/H' antiporters. Such storage organelles may
include the ER and central vacuole and to a lesser degree the mitochondria and
nucleus. The cell wall can also contribute to [Ca2+]i (Buchanan et al., 2000).
Such an example of Ca2+ re 88Se from a storage site in response to plant PCD
stimuli is that of Ca2+ re 88Sed from the central vacuole during TE differentiation
(Yu et al., 2002). Calcium is known to be a major player in signal transduction at
the cell's surface, in the cytoplasm, and in the nucleus as well as an effector of
reactions. Under conditions of stress Ca2+ may be released from these reserves
through such channels as inositol 1,4,5-triphosphate sensitive channels (IP3
channels) and starting a cascade of signal transductions and metabolic reactions
by means of soluble Ca2+ binding proteins such as calmodulins or MAPK's
(Zocchi and Rabotti, 1993; Buchanan et al., 2000).
With regards to apoptotic PCD [Ca2+]i has been observed increasing in
concentration in the cytoplasm of cells coincident with or shortly after any
observable morphological changes in the nucleus, the first site where
morphological changes can be observed for PCD (Sherwood and Schimke,
1995; Kerr et al., 1995).
The Role of Endonucleases
Calcium also plays a role in activating enzymes involved in PCD such as
certain nucleases. However, other nucleases involved in PCD require additional
metals, different metals altogether, andlor specific pHs to be activated. For many
years apoptosis was, more or less, the only type of PCD known and it was
always accompanied by DNA cleavage resulting in multiples of ~180bp size
fragments at the end of the PCD process showing up as an internucleosomal
"DNA ladder" on gels. Only in the more recent past have investigators come
across other types of PCD where DNA is either cleaved into extraordinarily long
fragments of 30, 50, or 300kb or not cleaved at all prior to complete degradation
of the cell (Mittler and Lam, 1995; Balk et al., 2003; Sugiyama et al., 2000;
Sugiyama et al., (2000) reviewed plant endonucleases responsible for DNA
cleavage associated with PCD excluding those that exhibit DNA repair
properties. They were able to categorize these nucleases into two fundamental
classes based on metal dependence and pH optima. The primary division was
based on divalent metal requirements, specifically Ca2+ Or Zn2+. However, more
recent reports have added an additional two classes of PCD associated
nucleases, those requiring magnesium (Mg2+) Only and those requiring both Mg2+
and Ca2+ (Sugiyama et al., 2000; Ito and Fukuda, 2002). Calcium and Mg2+
activated endonucleases operate at a relatively neutral pH whereas Zn2+
activated nucleases, or S1-type nucleases, are activated at mildly acidic pH's,
5.0-6.5 (Sugiyama et al., 2000). This being established, the physical location of
these endonucleases could be identified or extrapolated. For example, Zn2+
dependent nucleases have been found to be located in the central vacuole and
apoplastic areas such as cell walls (Sugiyama et al., 2000; Ito and Fukuda, 2002)
but the exact location of Ca2+ dependent nucleases is still enigmatic to a certain
degree. Because Ca2+-dependent nucleases operate at a relatively neutral pH
Ca2+-dependent nucleases are thought to reside in the nucleus; such an example
includes NUC Ill which was extracted from isolated tobacco nuclei (Mittler and
Lam,1995; Sugiyama et al., 2000). The type of PCD in plants cannot be
determined by one specific class of nuclease alone due to the fact that different
studies have shown cooperative action of nucleases from each class.
During TE differentiation DNA cleavage by Ca2+-dependent endonucleases
appears to result only in large DNA fragments but observations that complete
DNA cleavage does not occur until after the central vacuole has ruptured links
the Zn2+-dependent endonuclease to this phenomenon (Sugiyama et al., 2000;
Ito and Fukuda, 2002). It has also been demonstrated that cereal endosperm
PCD occurs as a cooperative effort among the nucleases and takes place in
three distinct stages employing Ca2+-dependent endonucleases during the first
stage, within the endosperm and Zn2+-dependent endonucleases secreted from
or acting within the aleurone layer at the latter stages (Young et al., 1997, Young
and Gallie, 2000, Sugiyama et al., 2000).
Additionally, unlike certain nucleases associated with animal cell apoptosis
that are constitutively present in the cytoplasm in a dormant state and are
activated only upon PCD signals, CAD, DNase 1 and topoisomerase II, many
plant nucleases are transiently expressed as needed. Ito and Fukuda (2002)
found that ZENV-1, a 43 kDa Zn2+ dependent endonuclease and a 24 kDa
Ca2+/Mg2+-dependent endonuclease were expressed only during induction of
PCD for TE differentiation.
It is this author's opinion that perhaps a third class of endonucleases should
be considered being established; one based on temporal expression and or
activity, those that operate during early stages of DNA cleavage and those in the
last stages of DNA cleavage. The endonucleases already discussed operate at
the late stages of PCD. However, Balk et al., (2003) established an Arabidopsis
cell free system to investigate PCD. They have demonstrated a Mg2+-
dependent, mitochondrial localized G-like endonuclease that cleaves nuclear
DNA into 30 kb size fragments within 3 hours of PCD induction and that further
DNA cleavage does not occur until 12 or more hours after PCD by nucleases
from the cytosolic fraction of the cell.
This study focuses on the morphological changes attributable to PCD in the
developing caryopsis of Zea mays, specifically in the placenta-chalaza (PC)
region of the kernel and the maturing endosperm.
In order to better understand PCD in developing caryopses it is necessary
to describe the process of maize seed development and anatomy. The maize
caryopsis is an indehiscent fruit containing only one seed (Esau, 1977).
Identification of the anatomical parts of the caryopsis will begin from the outer-
most part of the caryopsis and proceed toward the center. The caryopsis
consists of the pericarp, which is the fruit, or ovary. The ovary wall consists of
three layers, the outer, covered by a cuticle, middle, and inner pericarp. The
following two layers are the outer then inner integuments, or the seed coat and
are derived from the ovule. The integuments also retain a cuticular layer due to
their ontogeny having originated from a common epidermis during ovule
development. These cell layers ultimately degenerate during seed development.
Continuing inward is a crassinucleate, or large, nucellus comprised of
unspecialized parenchyma cells, and finally the embryo sac containing the egg,
two synergid cells, two polar nuclei, and three antipodal cells. At the base of the
caryopsis is the chalaza, a region of nucellar tissue fused to integumentary and
funicular tissue; however maize lacks a funiculus, tissue that forms a stalk on
which the ovule develops (Esau, 1977).
The chalaza has become commonly referred to as the placento-chalaza
(PC), a term coined by Katherine Esau in 1944 (Kiesselbach and Walker, 1952).
The tissue subtending this region consists of vascular and thick walled
parenchymatous tissue comprising the pedicel that connects the caryopsis to the
receptacle or cob. All of these tissues are sporophytic, or maternal, in origin.
The physiologic role of the PC is to be the gateway through which all water and
nutrients pass to the growing seed (Wetherwax, 1930; Felker and Shannon,
1980). Upon seed maturity the closing layer forms in this region effectively
sealing the seed off from the maternal tissues and will become the site at which
the seed will abscise (Johann, 1935).
Upon fertilization, the caryopsis undergoes dramatic changes. The egg
fuses with the first of two sperm nuclei from the pollen grain to become the
diploid embryo while the two polar nuclei fuse to each other followed by fusion
with the second sperm nucleus to become the triploid endosperm (Esau, 1977).
It is shortly after this point that several PCD regimens start to take place. Among
the types of PCD observed in the developing caryopsis are nucellar degradation,
PCD within the embryo, i.e. the scutellum, coleoptile, suspensor, root cap, and
starting at about 16 days after pollination (DAP), the endosperm (Giuliani et al.,
2002; Young et al., 1997). In addition to the above mentioned PCD events in the
maize caryopsis, PCD in the PC has been observed for the first time in this study
and by Kladnik et al., (unpublished). Each of these areas of PCD is a normal
part of development and they occur at specific times during development. PCD
was investigated on a cellular level using light and electron microscopy
techniques. Specifically investigated were the PC region during early
development (0-12 DAP) and the endosperm from the starch filling stage to near
physiologic maturity (12-25 DAP) and appears that these PCDs may be, not only
a normal part of development, but a necessary one. Surprisingly, three different
morphological changes were seen between the cells of the PC and the
endosperm, two in the PC and the third in the endosperm. It is, therefore,
possible that there are three different forms of PCD occurring in these tissues.
Figure 1. Classic apoptotic cellular morphology
The earliest morphological change in apoptotic cells occurs in the nucleus.
Organelles maintain good morphology during nuclear changes. (A) Nucleus
shrinks & chromatin condenses at the nuclear periphery. (B) The nucleus starts
to blebb" with chromatin filled pockets. (C) The entire cell blebs & breaks up into
nuclear bodies containing condensed chromatin pieces and cytoplasmic
contents. (D) Cell culture showing cells at various stages of apoptosis.
From Kerr et al., (1995).
Figure 2. Classic necrotic cellular morphology.
Necrotic morphology is illustrated by soft edged, condensed chromatin in various
clumps throughout the nucleus. Additionally, cytoplasmic organelles degenerate
concom itantly with nuclear changes.
From Kerr et al., (1995).
MATERIALS AND METHODS
W22 Maize plants were grown in a greenhouse from March through August
2002 and February through June 2003. The only exceptions were 14 and 16
days after pollination (DAP) caryopses, which were harvested from field grown
plants in the summer of 2003. Caryopses were harvested from the same ear
whenever possible or from ears of sibling plants grown at the same time under
identical conditions. Caryopses were harvested at various stages of
development. For PC examination caryopses from 0 through 12 DAP were used
and for endosperm examination 12 through 28 DAP caryopses were used.
Both distilled glutaraldehyde (8%) sealed in ampules with an inert
atmosphere and paraformaldehyde (PFA) were purchased from Electron
Microscopy Sciences. PFA was prepared by mixing 8% (w/v) PFA into 0.1 M
cacodylate buffer, covered with foil, and, under constant stirring, heated to just
under the boiling point. Sodium hydroxide (NaOH) was added to aid in dissolving
the PFA, maintaining a pH of less than or equal to 8.0. Once dissolved, the PFA
solution was immediately placed on ice to chill, filtered through a number 1
Whatman filter and the pH adjusted to 7.0 with concentrated hydrochloric acid
(HCI). The PFA stock was stored frozen at -200 C for up to 2 weeks.
Some tissues were prepared by fixation in 4% glutaraldehyde (v/v) and 1%
paraformaldehyde (v/v) in 0.1 M phosphate buffer (pH 7.0). Most of the tissues
were prepared using cacodylate buffer (pH 7.0). Fresh caryopses were
harvested and immediately processed for fixation on site by placing a small
aliquot of chilled fixative on a glass plate and hand sectioning out a sagittal
section of the caryopsis approximately 2mm wide with a double edged razor
blade. The tissue was then put into a vial of ice cold fixative and placed under
vacuum for 5 to 10 minutes to aid fixative infiltration. After vacuuming the
samples were placed on a rotator overnight, approximatelyl2 hours at 40 C.
After fixation the tissue was rinsed three times for 30 minutes each in
chilled cacodylate buffer then placed in 2% aqueous osmium tetroxide overnight
at 40 C. Once the osmication was completed the samples were rinsed 3 times,
one hour each, in distilled water then dehydrated in a chilled acetone series at
20% increments, each for one hour. Once the samples reached 100% acetone
the acetone was exchanged three times. On the third change the acetone was
mixed with propylene oxide as a transition solvent. The ratios of acetone to
propylene oxide were 2: 1, 1:1, and finally 100% propylene oxide. Each transition
lasted 20 minutes at room temperature. Following the 100% propylene oxide
incubation, fresh propylene oxide was mixed with Spurr's epoxy resin (including
the catalyst) in ratios of 2: 1, 1:1, 1:2, propylene oxide to resin and infiltrated in
each for 12 hours. Upon completing the 1:2 propylene oxide to resin step the
samples were placed in 100% resin and infiltrated for 24 hours. This step was
repeated 3 times with fresh resin. After the third exchange of resin at 100% the
samples were placed in flat molds with fresh resin and polymerized at 560 C for
Samples were sectioned on a Sorvall MT-2B ultramicrotome. The sections
were cut to approximately 60 nm in thickness, indicated by a pale gold color, then
transferred to formvar coated copper grids (50 or 100 mesh). The sections were
post stained 20 minutes in filtered 2% aqueous uranyl acetate (UA), rinsed three
times, one minute each in distilled water then were allowed to dry. However, this
UA step was later modified to 25 minutes in 10% (saturated) UA. This was found
to give much better contrast to the sections. The sections were then stained six
minutes in Reynolds' lead citrate, rinsed once in 0.02 M NaOH for one minute,
then three times more, 1 minute each, in distilled water. The sections were
examined and photographed in a Zeiss 109 or a Zeiss EM-10 CA transmission
electron microscope (TEM). Negatives were scanned at 300dpi using Epson
Perfectionl650 scanner, digitally processed using Adobe Photoshop version 7.0
and saved to compact disc.
Calcium (Ca2+) W8S localized by using the oxalo-acetatelantimonate
method as described by VanReempts et al. (1982) and Borgers et al. (1977). All
steps were carried out at 40C. The fixative consisted of 2.8% glutaraldehyde, 2%
PFA, 90 mM potassium oxalate, and 1.4% sucrose in distilled water then
adjusted to pH 7.4 with 1M potassium hydroxide (KOH). This step is for the
purpose of allowing oxalate ions to combine with Ca2+ iOns to form the insoluble
salt calcium oxalate. After fixation the samples were rinsed for two hours in a
solution of 7.5% sucrose and 90mM potassium oxalate at pH 7.4. Next the
samples were post fixed overnight in 1% osmium tetroxide mixed with 2%
potassium pyroantimonate in 0.1M acetic acid and adjusted to a pH of 7.4 with
1M KOH. This step is for the exchange of oxalate ions for antimonate ions in the
calcium salt to form an insoluble, electron dense calcium antimonate precipitate.
Samples were then rinsed for half an hour in distilled water adjusted to pH to 10
with 1M KOH. Subsequent steps for dehydration, infiltration and embedding
were carried out as described for ultrastructure.
Caryopses were harvested and fixed as described for ultrastructural
studies. Dehydration was carried out in an ethanol series consisting of 20%
increments, one hour each, up to 100% ethanol at 40 C. After two incubations in
100% ethanol for one hour each the samples were allowed to come to room
temperature and were placed in a 1:1 mixture of ethanol to tertiary butanol (TBA)
then twice in 100% TBA, again, for one hour each. After the second change of
TBA, a third change of TBA was mixed with Paraplast Plus paraffin chips (Fisher
Scientific) and placed in an oven at 560 C overnight (12 hours). which allowed the
paraffin to slowly melt and infiltrate while the TBA evaporated, facilitating a slow
controlled exchange. Following this step the TBA paraffin mixture was removed
and fresh molten paraffin poured on the samples to continue infiltration in the
oven for another 12 hours. This step was repeated two more times. After the
final infiltration step the samples were embedded using disposable boats and
embedding rings (Fisher Scientific) and sectioned to a thickness of 10 pm on a
Microm-325 rotary microtome. The sections were adhered to Fisher Scientific
Probon poly-L-lysine coated slides. To stretch the sections, a bubble of water
was placed on each slide, warmed on the slide warmer at approximately 400C
and the sections floated on it until the water evaporated. The slides and sections
dried on the warmer for 24 hours to assure proper adherence.
Terminal deoxynucleotidyl transferase (TdT)-mediated dUTP-fluorescent
nick end-labeling or, TUNEL, assays were performed using the TUNEL assay kit
purchased from Roche Diagnostics. The assay was done according to the
manufacturer's directions except, the tissue permeablization step using
Proteinase-K was reduced to one minute at 370 C and 14 minutes at room temp
to help preserve tissue integrity. Briefly, paraffin sections were de-waxed in two
changes of xylene and re-hydrated through an ethanol series (100, 70, 50, 30%)
for five minutes each and finishing in distilled water. The sections were treated,
as described, with 200g/ml Proteinase-K for 15 minutes, rinsed in PBS twice for
five minutes each. The sections were then treated with the TUNEL label plus
TdT enzyme for the experimental or simply just label for the negative control.
The slides were then placed in a petri dish lined with moist Whatman paper and
floated in a 370 C water bath in the dark for one hour. Once the incubation was
complete the slides were rinsed three times in PBS in the dark for five minutes
each, then counter stained with 600 pM (v/v) 4'6-diamidino-2-phenylindole-2HCI
(DAPI) DNA stain. The DAPI was mixed into the aqueous mounting medium,
Gelmount, (Fisher Scientific). The sections were examined with a Nikon Optiphot
fluorescent microscope and photographed using a digital CCD camera and Spot
software version 3.52 (Diagnostic Instruments). The TUNEL was examined with
an excitation wavelength of 450nm and emission at 520nm while the DAPI was
examined with an excitation of 380 and emission at 450nm.
For confocal imaging of nuclei in de-waxed sections propidium iodide DNA
stain (PI), purchased from Molecular Probes, was used. Pl DNA staining was
done according to manufacturer's directions exactly. Re-hydrated paraffin
sections were equilibrated in 2X sodium citrate, sodium chloride (SSC) buffer
(0.3 M NaCI and 0.03 M sodium citrate) for 3 minutes then treated with 100 pg/ml
RNAse A in 2X SSC in a water bath at 370 C for 20 minutes. The slides were
then rinsed in 3X SSC three times for one minute each. A solution of 500 nM Pl
was placed on the sections and incubated at room temperature in the dark for
five minutes. Following the incubation the slides were rinsed two times in SSC
for three minutes each and the sections mounted in Gelmount. A Bio-Rad 1024
ES Confocal Microscope was used to view the samples and images were
captured using LaserSharp Software.
Toluidine blue-O stain (TBO) was made tol% (w/v) in 1% sodium borate
solution (w/v) and filtered through a 0.2 pm filter before use. Resin sections were
TBO stained by covering the sections on the slides with a few drops of the stain,
gently heating the slide then rinsing the stain away with water. The slides were
dried over low heat on a hot plate and viewed by light microscopy.
Nuclease Activity Assays
In-gel assays were performed for endonuclease activity according to Young
et al. (1997), with minor alterations regarding the extraction buffer, discussed
below. Caryopses were harvested fresh from the field, flash frozen in liquid
nitrogen and stored at -800 C until used for protein extraction. Caryopses of 6, 8,
10, 12, and 28 DAP were tested. The caryopses were cut such that the top two-
thirds of the caryopses were separated from the bottom one-third. This was
done so nuclease activity of the PC region could be checked free of endosperm
tissue. Approximately 500 mg of tissue was used for the protein extraction. The
tissue was ground to a fine powder in liquid nitrogen for 15 minutes using a
mortar and pestle. The powder was then decanted into a sterile Corex tube and
diluted to 5 ml with ice-cold extraction buffer. The protein extraction buffer used
for these assays differed from that used by Young et al. (1997) in that the buffer
was 50 mM Tris Maleate pH 7.0. This change was made to eliminate any metals
that might interfere with or inadvertently activate the nuclease enzymes. The
samples were spun in a Sorvall RC-5B refrigerated rotor superspeed centrifuge
at 8000 rpm for 10 minutes at 40C. The supernatant/ protein was then
transferred to 1 .5 mlI Eppindorf tubes and kept on ice until quantified. P rotei n
quantification was carried out in a BioRad Smartspec 3000 using the Bradford
assay (Biorad) according to the manufacturer's specifications. Once quantified
the samples were stored in aliquots of one milliliter, flash frozen in liquid nitrogen
and stored at -80 until needed.
Two micrograms of protein from each sample was mixed with sample
dilution buffer (1% SDS w/v, 62.5 mM Tris-HCI pH 6.8, 10% glycerol (w/v), and
0.5% bromophenol blue (w/v)) and run on a 12% SDS-PAGE gel at 200 V for one
hour. Once completed the gels were rinsed in 25% isopropanol in 0. 1 M Tris
buffer pH 7.4 three times for ten minutes each to rinse out the SDS detergent.
The gels were then washed another three times in 0.1M Tris pH 7.4 to rinse out
the isopropanol. Afterwards the gels were incubated in washes containing either
0. 1 M CaCl2 Or MgCl2, Or both in 0. 1 M MOPS buffer pH 6.8 at 500C for 30
minutes or 1 mM ZnSO4 in 0.025 M acetate buffer (1.76 ml 1 M acetic acid, 8.24
ml 1 M sodium acetate and 40 ml H20) pH 5.5 at 470C overnight while gently
rotating in water baths. Several types of controls were employed: 1) gels were
either incubated with EDTA and EGTA in a 1:1 stoichiometric ratio with that of
the metals added to the incubation medium, 2) the pHs were changed such that
the Ca2+/Mg2+ buffers were lowered to that required for Zn2+ and the pH of the
Zn2+ buffer raised to that required for Ca2+ and Mg2+ Or 3) the metals were left out
of the incubation medium altogether. After incubation the gels were bathed in
200 pg/ml ethidium bromide for 10 minutes, rinsed in distilled water for 10
minutes then inspected in a transilluminator and photographed using BioRad
Genomic DNA was isolated from 5 grams of the same tissues cut in the
same manner as was used for the protein extraction. The developmental stages
examined were 8, 10, 12, 16, 20, and 25 DAP. Tissue was ground with a mortar
and pestle in liquid nitrogen for 20 to 30 minutes to a fine powder then
immediately put in sterile culture tubes containing 15 ml extraction buffer (100
mM Tris pH 8, 50 mM EDTA pH 8, 100 mM NaCI, 1% SDS, 0.2% PVP, and 10
mM. -mercaptoethanol) and incubated at 650C for 10 minutes. After incubation 5
ml 5 M ammonium acetate was added, mixed, and then incubated on ice for 20
minutes. The samples were then poured through Miracloth into a new sterile
tube and 1 ml 5 M ammonium acetate and 1 volume isopropanol added. The
samples were then kept at -200C for a minimum of 30 minutes, centrifuged at
12,000 rpm at 40C in a Sorvall RC-5B refrigerated superspeed centrifuge for 15
minutes and the supernatant poured off. The pellet was washed with 5 ml 70%
ethanol after which the tube was inverted on a paper towel for 5 minutes then
placed under vacuum for an additional 5 minutes. The pellet was resuspended in
500 pl TE (10 mM Tris-HCI, 1 mM EDTA pH 8.0) and placed on ice for 30
minutes with intermittent mixing every 5 to 10 minutes. Next, 250 pl 5 M
ammonium acetate and 2 pl of 10 mg/ml RNAse A were added and incubated at
370C 30 minutes. An equal volume of phenol, chloroform, isoamylalcohol (IAA)
in a ratio of 25:24:1, respectively was added to the tube, spun for 10 minutes at
8,000 rpm and the supernatent pulled off and placed in a new tube. One volume
chloroform and IAA in a ratio of 24: 1 respectively was added to the tube and
spun again at 8,000 rpm for 10 minutes. The supernatant was removed, placed
in a fresh tube, two volumes absolute ethanol added then incubated for 1 hour.
The sample was spun for 20 minutes at 8,000 rpm, the ethanol poured off, and
the pellet dried under vacuum. Once the pellet dried it was resuspended in 50 to
100 pl sterile water and stored at -200C.
The DNA was quantified in a BioRad SmartSpec 3000 spectrophotometer.
Two pg DNA was electrophoresed in a 2% agarose gel infused with 0.5 pg/ml
ethidium bromide with a low mass DNA ladder and run at 120V for approximately
1 hour. The gel was photographed in a transilluminator using Biorad Geldoc
TUNEL and DAPI stains were used to investigate DNA conditions and
nuclear morphology, respectively, in developing endosperm cells of maize. The
endosperm was divided into four regions for inspection: the crown, central crown,
mid endosperm and lower endosperm (Fig. 3). As illustrated in Figure 4, the
nuclei at each stage of development showed distinct morphological changes
taking place up through 24 DAP. At 12 DAP the cells in the crown of the
endosperm were starting to fill with starch and nuclei appeared healthy and
maintained an approximate central position in each cell. The nuclei were
characterized by diffuse DNA which appeared uniform in texture throughout the
nucleus. By 16 DAP the crown of the endosperm was filled with greater
quantities of starch and the DNA of each nucleus had condensed at the
periphery giving the nuclei a beaded appearance. By 24 DAP the TUNEL assay
showed positive results as indicated by bright green fluorescence in the nuclei in
the central and crown region of the endosperm, indicating that DNA cleavage in
the endosperm does not take place until after starch is abundant and the nuclei
lose their apparent integrity (Fig. 5). It is at this stage when the nuclei condense
and shift closer to the cell wall.
The 28 DAP kernels proved to be too opaque for any TUNEL activity or
DAP I stain to be seen due to the autofluorescence of the starch grains.
However, confocal imaging of propidium iodide stained sections was used to look
at the nuclear morphology. By 28 DAP, as seen in Figure 6, the nuclei of the
crown of the endosperm appeared completely condensed and appressed against
the cell wall.
Transmission Electron Mlicroscopy
Caryopses of the same developmental stages used for LM investigation
were selected for nuclear investigation from the central crown region of the
endosperm by TEM.
The pattern of morphological changes of endosperm nuclei seen by DAPI
staining at the LM level compared favorably to those seen at the EM level. At 12
DAP, the nuclei appeared rounded, the DNA diffuse and the cell appeared
normal as indicated by a normal appearing nucleus with diffuse DNA and an
intact nuclear envelope (Fig. 7). At 14 DAP the nuclei appear slightly larger and
the chromatin less diffuse (Fig. 8). Interestingly, the rough endoplasmic
reticulum (ER) appeared more prevalent and displayed intermittent areas of
bloating while in other areas the bloated bodies become filled with electron-
dense substances then appeared to bud off, presumably forming protein bodies.
By 16 DAP the nuclei displayed condensed chromatin that collected at the
nuclear envelope (Fig. 9) supporting the beaded appearance as seen by DAPI
stain at the LM level. The ER became less pronounced and the membrane
bound, electron dense bodies took on a more oval shaped appearance and mini-
vacuoles contained within them could be seen. By 20 DAP the nuclei appeared
bloated, appeared to have much less chromatin and, in some cases, appeared to
be almost devoid of contents leaving only an intact nuclear envelope (Fig. 10).
However, some of the nuclei displayed enlarged condensed chromatin spheres
that looked like oversized nucleoli; but, as stated previously, the rest of the
nucleus is empty except for very small, condensed chromatin pieces sparsely
distributed at the envelope. The nuclei at 22 DAP (Fig. 11) became very difficult
to discern as the membranous structures of the cell, including the plasma
membrane, appeared to be degenerating and cytoplasmic contents started to
coalesce and were not readily identifiable. By 25 DAP the nuclei and cytoplasm
of the cells became indistinguishable (Fig. 12). All the cellular contents
condensed, became amorphous in appearance and, for the most part, moved to
the cell periphery. At higher magnification the amorphous material acquired the
appearance of being comprised of many small spheres that appeared to be fatty
The Placenta-Chalaza Region
Both LM and TEM were also employed to investigate morphological
changes occurring the PC region during PCD. In order to investigate the
changes associated with the PC region it was necessary to anatomically define
this region more clearly. The ontogeny of the PC region, (Fig.13), showed two
cell lineages from each the nucellus epidermis and the inner integument, the
nucellus epidermis contributing to the upper region of the PC and the inner
integument contributing to a region of the PC subtending the nucellar derived
region. These two regions will be referred to as "zones" for the sake of clarity.
The anatomy of the PC region clearly shows two distinct zones, the upper from
the nucellus epidermis and the lower from the inner integument, as described
above (Fig. 14). This observation was in agreement with the literature in that the
PC region of the maize caryopsis is where the integumentary and nucellar
tissues merge (Esau, 1977). After pollination, as described by Esau, (1977), two
semipermeable, suberized andlor cutinized, extra-cellular membranes became
detectable in the seed coat and by approximately 7 DAP extended around the
entire seed but ended abruptly on either side of the PC region (Fig. 15). The
thicker of the two membranes appeared bi-layered (Fig. 16). This membrane
was derived, primarily, from the nucellus epidermis and, secondarily, the inner
integument. These membranes ended abruptly on either side of the PC thus
defining the boundaries of the PC region (Fig. 17).
PCD in the PC Region is Fertilization Dependent
DAPI DNA staining of a 12 DAP caryopsis and an unpollinated ovule of 12
DAP equivalency revealed enucleation of the PC cells only in the fertilized kernel
(Fig 18). Additionally crystal violet staining (Figs. 19-20) revealed condensed
chromatin in cells of the integumentary PC between 6 and 14 DAP which
coincided with the TUNEL positive reactions (Fig. 21) seen in this region (Kladnik
et al., unpublished). While there were no detectable condensed nuclei between
16 and 22 DAP condensed chromatin was seen, again, at about 24 DAP (Fig.
22) in the region of the integumentary PC that was most compressed by the
Transmission Electron Mlicroscopy
Prior to pollination both layers, the nucellus-derived and integument-derived
layers, of PC cells appeared normal and only a few morphological differences
aided in distinguishing the two cell types from one another. The nucellar PC cells
bordering the integument-derived cells in the PC region of the 0 DAP caryopsis
were large relative to the integument-derived cells. The nucellar PC cells were
characterized by thin, primary cell walls and very large, single vacuoles and a
small nucleus relative to the cell's size (Fig. 23A). These cells contained healthy
organelles and abundant plasmodesmata
Up through approximately 5 DAP all the PC cells remained relatively
unchanged in appearance from those of the PC in the 0 DAP caryopsis. At 5
DAP the only significant change in the cells of the PC region from that of the 0
DAP caryopsis was that the nuclei of the PC cells had become more rounded. At
this stage of development the growing endosperm started to crush the nucellar
tissue immediately in contact with it thus beginning the initial steps of forming a
multi-walled barrier between the basal endosperm cells and the PC region. The
nucellar-PC cells just below (Fig. 23B) still contained healthy cytoplasm and
nuclei and the average number of plasmodesmata per cell per section was 7. 1.
After pollination a dramatic series of changes took place in the PC region as
the caryopsis developed. The first observable changes that occurred were at
about 7 DAP. The enlarging endosperm had further crushed the nucellar PC
cells creating a more compacted barrier between the basal endosperm and PC
region. These cells had substantially emptied and contained remnants of what
appeared to be plasmolysed cytoplasm (Fig. 23C). These cells also appeared a
bit more flattened than previously seen in younger stages. The average number
of distinguishable plasmodesmata per cell remained the same as it was for 5
DAP. By 9 DAP, the border between the basal endosperm and the nucellar-PC
had become very thick and solid and had completely isolated the endosperm
from any symplastic connections to the PC. The nucellar-PC cells had only very
sparse remnants of unidentifiable degenerated cytoplasmic contents (Fig. 23D).
The number of plasmodesmata from 9 to 12 DAP varied from 1.6 to 3.3 per cell
per section for an average of 2.4 per cell per section (Fig. 24). This was a
reduction of 33% from the 7.1 seen from 5 to 7 DAP. However, it is important to
note that this apparent reduction in the number of plasmodesmata was seen only
in the nucellus-derived portion of the PC and not the integument-derived portion,
which appeared to maintain an abundance of plasmodesmata.
The layer of cells below the nucellar-PC cells are those derived from the
inner integument. At 0 DAP the integument-derived cells were characterized by
a more rectangular shape, thicker walls than the nucellus-derived cells, had a
more dense cytoplasm (Fig. 25A). Additionally, the nuclei in these cells were
small, elongated and looked like flattened footballs. At 5 DAP these cells
appeared healthy and, generally, contained one vacuole; the only significant
change from 0 DAP was a slightly more dense cytoplasm (Figs. 26 A, B). The
next cell layer below the nucellar PC cells, the upper integumentary PC, was,
approximately, two cells wide; these cells will eventually become part of the
closing layer. By 7 DAP (Fig. 25C) the nuclei of these cells showed the
beginning stages of condensing chromatin; however the nuclei were still intact
and appeared spherical. By 9 DAP the cells of the closing layer had become
laterally elongated, compacted and more cytoplasmically dense but some cells
still contained crushed nuclei with condensed chromatin. Generally, the
cytoplasm of these cells, at this stage of development, had become very dense,
the vacuoles were absent or were in the process of disappearing and there were
no identifiable organelles (Fig. 25D). The cells subtending the closing layer, the
lower integumentary PC cells (Fig. 26A-C) contained normal nuclei and
cytoplasm and averaged one vacuole per cell. At 9 DAP this cell layer (Fig. 26D)
was characterized by cells with a total lack of nuclei, unidentifiable organelles
and no vacuoles but retained all their disintegrated cytoplasmic contents. These
cells had become flattened and elongated. This layer of cells was peculiar in that
it showed variable positioning around the closing layer depending on which part
of the PC region was being viewed. On the lateral edges this third cell layer was
above the closing layer for a short distance but when moving toward the center of
the PC the third layer appeared below the closing layer and held this position for
the greater portion of its span. The fourth layer of cells down (Fig. 26E)
contained early stages of PCD. The cytoplasm of these cells contained multiple,
small vacuoles, but the nuclei were starting to display condensing chromatin.
Additionally, the endoplasmic reticulum (ER) became much more abundant and
formed concentric ring-like andlor layered, laminar patterns (Fig. 27). Continuing
basipetally, the next layer of cells appeared healthy as well with the only one
distinguishing feature from those of the previous layer. These cells, on average,
contained several smaller vacuoles. Below these cells was the vascular tissue
and starch-containing parenchyma.
The closing layer and a few cells beneath the closing layer, by 8 DAP,
appeared to be the first set of cells at which gross morphological changes related
to PCD were seen. This was the first stage at which completely condensed
nuclei were detected. Interestingly, in the cells that harbored completely
condensed nuclei, small spheres filled with condensed chromatin remains were
seen throughout the cytoplasm corresponding to similar TUNEL positive bodies
seen within PC cells at a light microscopy level (Fig. 28). For those cells at such
an advanced state of degradation the only cytoplasmic feature that could be
clearly identified with certainty was the rough ER. However, the ER was very
swollen. By 11 DAP, as seen in Figure 31E and F, the cells of the closing layer
looked extremely chaotic; the contents of most of the cells in this layer looked
disjointed. For example, the membranes were collected in tight whirls and
seemingly disconnected from anything else in the cell, lipid bodies and
unidentifiable cellular debris constituted the rest of the cytoplasm. In some cases
the cytoplasm was dominated by many very tiny vacuoles. The cells furthest
down in the PC appeared normal with completely normal looking nuclei and
organelles. By 12 DAP the first set of cells with any cellular debris to be seen
was the top most layer of the closing layer, which coincides with the chaotic cells
seen in the 11 DAP. By this stage all that was left was condensed cytoplasm.
Further down into the closing layer and the rest of the integumentary PC cells
with condensed nuclei became more abundant. Some cells were seen at such a
stage where the mitochondria were found to be dramatically swollen and losing,
or had lost, inner membrane integrity. Moving further down into the PC nuclei
had mild to no condensed chromatin and there were identifiable organelles.
A general trend was seen in the morphology of the nuclei as the PCD
process ensued. Initially the chromatin in the nuclei appears diffuse and the
organelles normal in appearance (Fig. 29A). As in animal apoptosis the
chromatin condenses and is delineated by sharp edges. However, unlike animal
apoptosis, the chromatin does not coalesce at the nuclear periphery (Fig. 29B).
Eventually the chromatin in the nucleus completely condenses but the nuclear
membranes remain intact (Fig. 29C). At the end of the process the nucleus has
broken up into nuclear bodies (Fig. 29D).
The oxalatelantimonate calcium ion (Ca2+) trapping method was used in an
attempt to determine the flux and location of free Ca2+ iOns within the cell during
this cell death process. At 5 DAP the nucellar-PC cells contained only trace
amounts of Ca2+ in the cytoplasm and organelles (Fig. 30A). The cells did not
contain Ca2+ in the central vacuoles at this time, but rough ER had light Ca2+
deposits associated with the membrane. The mitochondrial matrix did not
appear to be associated with appreciable deposits of Ca2+ While the nucleus was
very lightly peppered with Ca2+. At 8 DAP the nucellar-PC contained moderate
background Ca2+ (Fig. 30C) but the remnants of cellular contents did not contain
Ca2+. The next cell layer down, still within the nucellar PC (Fig. 30C-inset), was
intermediate in its cytoplasmic degeneration between that seen in the nucellar-
PC and the closing layer. This cell layer showed Ca2+ tO be abundant in the cell
corpse background but still did not appear to associate with the degenerated
cytoplasmic remains. By 11 DAP the nucellar-PC appears similar to that of the 8
DAP except at this point there is no Ca2+ in the cell corpse background any
longer (Fig. 30E).
At 5 DAP the integument PC cells did not appear significantly different than
those in the nucellar PC except that the mitochondria appeared to have a light
scattering of Ca2+ deposits throughout giving them a "peppered" appearance
(Fig. 31A). By 8 DAP the closing layer cells in the upper integumentary PC
showed Ca2+ Sparsely in the cytoplasm but small amounts were present in what
was left of the ER (Fig. 31C). By 11 DAP the closing layer showed chaotic,
condensed cytoplasmic contents with highly compacted swirls of membranous
structures but with little Ca2+ preSent (Fig. 31E).
The cells immediately subtending the closing layer varied subtly in their
localization of Ca2+. These cells were examined in order to check where Ca2+
might localize in normal appearing PC cells of more mature tissues. Some of the
cells displayed mitochondria with little to no Ca2+ lOcalized within (Fig. 32A).
Additionally, the ER did not appear to contain much Ca2+ sS indicated by few
electron-dense precipitates. The other cells of this region looked similar to those
discussed above except the ER was abundant with Ca2+ Which were heavy in
electron-dense precipitates (Fig. 32B), the mitochondria seemed to have more
Ca2+ (Fig. 32C), and the nuclei were peppered with Ca2+; however, the vacuoles
of both cell types contained abundant Ca2+ and the cytoplasm did not have much
Ca2+ in either cell type relative to the rest of the cell. In cells of moderate to
advanced stages of PCD the nuclei contained more Ca2+ than the cytoplasm and
the mitochondria showing morphological changes, i.e. swelling, contained fewer
Ca2+ deposits than those not yet showing morphological changes in other cells
In-Gel Nuclease Activity Assay
The lower one-third of the wild type W22 kernels, 6, 8, 10, 12, and 28 DAP
were examined for Ca2+, Mg2+, both Ca2+ and + Mg2+, Or Zn2+- dependent
nuclease activities. The Ca2+ and Mg2+ treatments, whether tested together or
independently resulted in the same pattern (Fig. 33). As was reported by Young,
et al. (1997) three nucleases of 33.5, 36.0 and 38.5 kiloDaltons (kDa) were seen
when tested using these metals. The 33.5 kDa nuclease was highly active at all
developmental stages examined; however, only the 38.5 and 36.5 kDa nucleases
were most active at 8 DAP as indicated by a lack of DNA detection in the gel
(Fig. 33A). The Zn2+-dependent nucleases (Fig. 33B) showed a similar pattern of
banding with the addition of a fourth band at 35 kDa and a slight decrease in
activation of the 36.5 kDa and 38.5 kDa nucleases. The 35 kDa nuclease
appeared to increase slightly at 10 DAP and persisted through 12 DAP. As a
control the Ca2+ and Zn2+ ge S were incubated at acidic (5.5) and neutral (6.8)
pH, respectively, to test the effect of improper pH on each nuclease (Fig. 33C
and D). The results were a complete inactivation of all nucleases except the
smallest one. EGTA and EDTA were then added to the Ca2+/Mg2+ and Zn2+
assays. Each was then incubated at its proper pH. The Ca2+/Mg2+ gel Showed a
loss of activity in all nucleases except the smallest one; however, the Zn2+ gel did
not show a reduction in activation most likely due to the acidic conditions of the
incubation medium which inhibits the ability of EGTA and EDTA to act as
chelating agents (Fig. 34A and C). Lastly, the gels were incubated at their
respective pHs without the metals included in the incubation media. The gels
showed the same pattern of nuclease banding as with their respective metals
added indicating that these metals must be present and inherent in the tissue
itself (Fig. 34B and D).
Genomic DNA Extraction
Genomic DNA was extracted to try to identify inter-nucleosomal laddering in
8-24 DAP PC tissues (Fig. 35). No inter-nucleosomal banding was detected for
the DAP's tested.
Top of caryopsis
Base of caryopsis
Figure 3. Schematic drawing illustrating the regions of the maize endosperm.
The crown is the upper-most region of the endosperm. The central crown is the
interior portion of the crown. The mid-endosperm lies parallel to the embryo and
the lower endosperm subtends the mid-endosperm and extends to the basal
endosperm transfer cells, the first layer of cells at the base of the endosperm just
above the placenta-chalaza region (PC).
Figure 4 Nuclear morphological changes in the endosperm of maize
Nuclear morphological changes in the endosperm of maize at 12, 14, 16, and 24
days after pollination (DAP) visualized via DAPI DNA stain. Crown: through 14
DAP the chromatin in the nuclei appears round and diffuse. By 16 DAP the
nuclei and chromatin have started to condense. By 24 DAP the nuclei are so
condensed they have become difficult to discern. Central crown: the nuclei show
the greatest morphological changes in this region of the endosperm through
development. At 12 DAP the nuclei are round and diffuse. At 14 DAP the nuclei
have enlarged and the chromatin is starting to appear more granular. By 16 DAP
the nuclei have substantially enlarged and the chromatin has condensed at the
nuclear periphery. By 24 DAP the nuclei have condensed appearing like those in
the crown. Mid endosperm: through 14 DAP the chromatin in the nuclei is diffuse
and round. By 16 DAP the nuclei have enlarged slightly and appear similar to
those of the central crown at 14 DAP in that the chromatin has started to become
granular in appearance. By 24 DAP the chromatin has condensed at the nuclear
periphery and appears like the nuclei in the central crown at 16 DAP. Lower
endosperm: the nuclei remain relatively unchanged up through 24 DAP. The
nuclei retain their size and diffuse chromatin. Magnification: 20x.
Figure 5 PCD in the endosperm of a 24 DAP caryopsis.
(A-C) DAPI DNA stain was used to identify the location of nuclei in the cells.
(D-F) Terminal deoxynucleotidyltransferase mediated dUTP nick end labeling
(TUNEL) assay was used to show broken and nicked DNA in the nuclei denoted
by green fluorescence. (G) The negative TUNEL control shows no green
fluorescently labeled nuclei. The nuclei in the upper endosperm are condensed
while those of the mid endosperm are ring-like in appearance due to chromatin
condensation at the nuclear envelope. The nuclei in the lower endosperm are
rounded and the chromatin is diffuse throughout each nucleus. Only the mid and
upper endosperm are showing TUNEL positive activity. Magnification, 20x.
Figure 6 Confocal image of the nuclei in the central crown region of the
endosperm of a 28 DAP caryopsis of maize.
The nuclei (orange color) appear completely condensed at maturity, have shifted
to the cell periphery and are appressed against the cell wall (green). Nuclei
stained with propidium iodide. Magnification, 20x.
Figure 7 Nuclear morphology of the central crown region of the endosperm of a
12 DAP maize caryopsis.
At 12 DAP the nuclei of the central upper endosperm appear normal; the
chromatin is diffuse and the nuclei are, generally, centrally located. (A) A light
micrograph of a TBO stained resin section of endosperm used as reference for
(B). Arrows point to nuclei. (B) TEM image of endosperm nucleus N, nucleus; s,
starch grain. Magnification for A, 20x; B, bar 6Cim.
Figure 8. Nuclear Morphology of the central crown region of the endosperm of a
14 DAP maize caryopsis.
(A) TEM image. The nucleus is enlarged and, while not yet condensed, the DNA
has collected at the periphery of the nucleus. The nucleolus is still prominent.
(B, C) TEM images. Relative to all other developmental stages examined, the
endosperm, at 14 DAP, appears to be it's most active in forming protein bodies.
(C) The protein bodies (box) bud directly from the rough ER.
N, nucleus; n, nucleolus; rER, rough endoplasmic reticulum.
Bars: A, B, 6Cim; C, 0.5Cim.
Figure 9 Nuclear Morphology of the central crown region of the endosperm of a
16 DAP maize caryopsis.
Nuclei are enlarged and are starting to display condensed chromatin at the
nuclear envelope. (A) TBO stained resin section of endosperm. Boxes enclose
nuclei. (B) TEM image of a nucleus. (C) TEM composite of a nucleus.
N, nucleus; cc, condensed chromatin. Magnification; A, 20x; Bars: B, C, 6Cim.
Figure 10 Nuclear Morphology of the central crown region of the endosperm of a
20 DAP maize caryopsis.
Similar to 16 DAP the chromatin is condensed and adhered to the nuclear
envelope but the nuclei are slightly more enlarged. The nucleoli are still present.
The nuclei have shifted to the periphery of each cell. (A): TBO stained resin
section. Boxes enclose nuclei. (B, C): TEM images of nuclei.
N, nucleus; n, nucleolus. Magnification; A, 20x; Bars: B, C, 5Cim.
Figure 11 Nuclear morphology of the central crown region of the endosperm of a
22 DAP caryopsis of maize.
TEM image of the remnants of a collapsing nucleus. Some condensed
chromatin, denoted by black masses, is associated with the nuclear envelope.
N, nucleus; cc, condensed chromatin. Bar: 5Cim.
Figure 12 Nuclear Morphology of the central crown region of the endosperm of a
25 DAP caryopsis of maize.
By 25 DAP the nuclei have collapsed and shifted to the periphery of the cell. The
nuclei have a disorganized, amorphous appearance. Close inspection reveals
no organization of the nucleus but shows lipid or fatty appearing spheres
surrounded by an electron dense material. These may be left over membranous
and chromatin materials from the dissociated nucleus. (A) TBO stained resin
section. Box encloses a nucleus. (B) TEM image of nucleus. (C) TEM image,
close up of nucleus. N, nucleus. Magnification: A, 20x; Bars: B, 5Cim; C, 0.5Cim.
Figure 13 Origin of the Placenta-chalaza (PC) region.
This 9 DAP caryopsis illustrates the PC region containing two distinct sets of
cells comprising two zones. The upper zone is contiguous with the nucellar
epidermis and the lower zone is contiguous with the inner integument. 1)
Endosperm, 2) Basal endosperm transfer cells (BETC), 3) Nucellar epidermis, 4)
Semi-permeable membrane (gray line) subtended by inner integument, 5) Outer
integument, 6) Inner pericarp, 7) Mid pericarp. Magnification: 20x.
Figure 14 Anatomy of the Placenta-chalaza.
The PC is comprised of two different types of cells creating two distinct zones,
the nucellus-derived PC and the integument-derived PC. The PC is sandwiched
between the basal endosperm transfer cells (BETC) and the pedicel
parenchyma. 1) Endosperm, 2) (BETC), 3) Nucellus derived PC, 4) Closing layer
(part of integument derived PC), 5) Integument derived PC, 6) Pedicel
parenchyma. Magnification: 20x.
Figure 15 Placenta-chalaza (PC): left-right borders.
The boundaries of the PC region are denoted by the developing semi-permeable
membrane. The PC region lies between the ends of the semi-permeable
membrane at the base of the caryopsis thus becoming the only port through
which water and nutrients can pass. Column A represents the adaxial side of
caryopsis while column B represents the abaxial side of caryopsis. (2 DAP) The
membrane starts its development below the developing endosperm first followed
by rapid development at the abaxial face of the lower kernel opposite the
developing endosperm. (6 DAP) The membrane has developed fully around the
entire seed except at the PC where it ends abruptly. (8-10 DAP) The membrane
matures and continues to thicken. Magnification: 20x.
Figure 16 TEM image of one of the boundaries of the PC region.
A semi-permeable membrane develops from the nucellus epidermis and
surrounds the entire ovule but ends abruptly where the PC region begins at the
base of the caryopsis. 1) Nucellus epidermis, 2) Semipermeable membrane, 3)
Fatty substances, 4) Inner integument, 5) Membrane terminus, 6) Start of PC.
Horizontal Ilmits of PC
Vertical Ilmits ofPC
Figure 17 Placenta-chalaza boundaries.
The horizontal limits of the PC region are defined by the start points of the
semipermeable membrane and the vertical limits are defined by the nucellus-
derived and integument-derived cell layers sandwiched between the basal
endosperm transfer cells and the pedicel parenchyma cells. (A) Abaxial side of
the caryopsis, (B) Adaxial side of the caryopsis. Magnification, 20x.
Figure 18 4'6-diam idino-2-phenylindole-2HCI (DAPI) DNA stain.
Nuclei appear as blue, glowing spheres or spots. Enucleation of PC is
fertilization dependent. Enucleation is denoted by a lack of blue glowing nuclei in
the cells (black area). (A) Equivalent of a 12 DAP caryopsis, unpollinated ovule,
(B) 12 DAP caryopsis. Magnification, 20x.
Figure 19 Crystal Violet Stain showing condensed chromatin in the nuclei of the
cells of the integument-derived PC region from 4 through 10 DAP.
Figures on the right hand side are enlargements of the corresponding figures to
the left Regions that are enlarged are denoted by the red lines. (4 DAP) There
are no condensing nuclei in the PC at this stage of development. (6 DAP)
Condensed nuclei start appearing at this stage in the integument-derived portion
of the PC in what will be the closing layer (cl) and in the cells just below the
closing layer. (10 DAP) Condensed nuclei remain prevalent at this stage.
Magnification; Left side 10x; right side 40x.
Figure 20 Crystal Violet Stain showing condensed chromatin in the nuclei of the
integument-derived PC region from 12 through 16 DAP.
Figures on the right hand side are enlargements of the corresponding figures to
the left. Regions that are enlarged are denoted by the red lines. (12 DAP)
Condensed nuclei (arrows) are limited to just below the closing layer. (14 DAP)
Condensed nuclei have become sparse and are a little further down beneath the
closing layer relative to 12 DAP. (16 DAP) There are no more condensed nuclei
in the PC at this stage. Magnification; Left side 10x; right side 40x.
Figure 21 TUNEL assays of PC region showing broken or nicked DNA in the
nuclei of the integument-derived portion of the PC
Terminal deoxynucleotidyl transferase mediatd dUTP nick end labeling (TUNEL)
assays of PC region showing broken or nicked DNA in the nuclei of the
integument-derived portion of the PC (red brackets). Nicked or broken DNA is
denoted by bright green fluorescence. The nucellus-derived PC cells test
TUNEL negative as indicated by a lack of green fluorescent nuclei. (A, C, E, G)
experimental sections. (B, D, F, H) negative controls. (4 DAP) TUNEL activity is
not yet detected. (8 DAP) TUNEL activity is well established. (10 DAP) TUNEL
activity peaks. (12 DAP) TUNEL activity is still seen but starts to taper off after
12 DAP. Magnification; A-F, 20x; G,H, 10x. From Kladnik et al., unpublished.
Figue 2 Crstlvoestishwncodnechoaiintenegm t
derived~~~~~ poto f h Creino a2 A cross
noncomre 2 rssed reion of tin he wn P.()Condensed nclemaini appea in thegm
compressed region of t PC. iNuclar oies (arrow)l apear nd one of thecels C
There are no condensed nuclei in the non-compressed region. Magnification;
A, 10x; B, 40x.
Figure 23 TEM micrographs illustrating consecutive developmental stages and
subsequent cell death of the nucellus-derived PC region.
This is the region that tests TUNEL negative. (A) At 0 DAP cells appear normal,
contain all organelles and large vacuoles. (B) By 5 DAP nuclei and all organelles
are present and the cells appear normal. (C) By 7 DAP the cells have lost all
organelles and nuclei and have plasmolyzed. (D) By 9 DAP the cells have
almost emptied completely and only degraded cytoplasmic contents remain.
Arrow heads, mitochondria; arrows, plasmolyzed cellular remains; (N), nuclei;
(DC), degenerated cytoplasmic contents. Bars, 5Cim.
7 DAP 9 DAP
Figure 24 TEM micrographs illustrating the plasmodesmata (PD) in the walls of
the nucelluc-derived PC cells.
(A) At 5 DAP the average number of identifiable PD in the walls of the cells per
section is 7.2. (B) The average number of identifiable PD in the walls starts to
decrease. (C) By 9 DAP the average number identifiable PD decreases to 2.4
per cell per section. Bars: A, 0.5Cim; B, 1Cim; C, 0.5Cim.
Figure 25 Development and subsequent death of upper integument-derived
region of PC.
The upper integument-derived region forms the closing layer of the kernel. This
is the region of the PC that tests TUNEL positive. (A) At 0 DAP the cells appear
normal, contain large vacuoles and all organelles are present. (B) Close up of
cells in A showing normal appearing organelles. (C) By 7 DAP the cells have
thickened walls, more dense cytoplasm, smaller vacuoles and some of the nuclei
start to display condensing chromatin (arrows). (D) By 9 DAP the cytoplasm
becomes very dense, the vacuoles break up into mini-vacuoles, there are no
identifiable organelles, and chromatin in nuclei continues to condense. Arrow
heads, mitochondria; arrows, nuclei with condensed chromatin; N, nucleus; C,
cytoplasm; V, vacuoles. Bars, 5[im.
Figure 26 Development and subsequent death of lower integument-derived
region of PC.
(A) At 0 DAP cells appear normal and contain all organelles and large vacuoles.
(B) By 5 DAP cells are still normal but the vacuoles are smaller and the
cytoplasm more dense. (C) By 7 DAP the cells still retain their organelles and
the nuclei still appear normal but, in some cells, the vacuoles are starting to
break into smaller vacuoles. The cytoplasm takes on a granular appearance.
(D) At 9 DAP the cells have lost all identifiable organelles, vacuoles, and nuclei.
The cytoplasm still appears granular. This layer starts to become compressed by
the growing endosperm and the cells appear flattened. (E) The cell layer below
D. These cells test TUNEL positive and contain nuclei that have condensed
chromatin. There are many mini-vacuoles present. V, vacuoles; arrows, nuclei.
9 DAP 11 DAP
Figure 27 Lower Integument-derived PC Cells.
Starting at approximately 9 DAP cells in the lowest portion of the PC
display abundant ER in the form of concentric rings or layered lamina. This
is characteristic of this cell layer up through at least 11 DAP.
V, vacuoles; ER, endoplasmic reticulum; N, nucleus. Bars, 2[tm.
Figure 28 Nuclear body formation in the integument-derived portion of the PC
A) TUNEL assay showing nuclear bodies (boxed) at the cell periphery. Arrow
shows one of the highly condensed TUNEL positive nuclei. (Micrograph by
Shayna Sutherland). (B) TEM micrograph showing the same type of nuclear
morphological pattern seen in A. The nuclei have completely condensed and
have started breaking off into nuclear bodies. Note the nuclear envelope is still
present and intact.
Magnification; A, 40x; B, bar 0.5[tm.
A L ,* 1-'
Figure 29 Nuclear morphological changes in integument derived portion of the
Nuclear morphological changes in the integument-portion of the PC
resemble apoptotic-like morphology. The micrographs of each progressive stage
of each changed nucleus are not from consecutive developmental stages as the
PCD taking place in one cell is independent of PCD occurring in all other cells
and the rate of progression between cells will vary. However, this series of
nuclear changes is progressive from when a cell starts showing morphological
changes through the end of its PCD process. (A) Normal appearing nucleus.
The DNA has not started showing any abnormal chromatin condensation. (B) A
nucleus showing first stages of chromatin condensation. (C) Highly shrunken
nucleus with completely condensed chromatin. (D) Nuclear body-remnants of the
nucleus filled with condensed chromatin (arrows). N, nucleus; m, mitochondria.
Bars: A, 5itm; B, 1.8[tm; C, 1.2[tm; D, 1.8[tm.
Figure 30 Ca2+ Trapping: Nucellus-derived PC.
Free Ca2+ was localized in an attempt to define a flux and organelle interaction of
this ion during PCD of the cell. (A) At 5 DAP Ca2+ is sparse and is associated,
primarily, with membranous structures and the nucleus but is not seen in the
cytoplasm or mitochondria. (B) At 8 DAP Ca2+ is abundant in the cell corpse but
does not appear to be associated with the degenerated cellular remains.
(C) By 11 DAP Ca2+ has disappeared from the cell corpse and is only lightly
associated with degenerated cellular remains. N, nucleus; m, mitochondria; DC,
degenerated cytoplasmic contents. Bars: A-D, 0.5[tm; E-F, 0.8[tm.
Figure 31 Ca2+ Trapping in the integument-derived PC.
Free Ca2+ was localized in an attempt to define a flux and organelle interaction of
this ion during PCD of the cell. (A) At 5DAP Ca2+ is very low in the cytosol but is
associated with membranes of the ER and vacuole. In addition, Ca is
peppered throughout the mitochondria and is very light in the nucleus. (B) By 8
DAP Ca2+ is low but present in the cytolsol. It is also seen in some of the ER
under the nucleus (bracket). However, other rough ER is still present but no Ca2+
is seen within it (arrows). (C) By 11 DAP Ca2+ is extremely low in all places in the
cell. The cell is in an extraordinary chaos and no nucleus or organelles are left.
V, vacuoles; m, mitochondria; N, nucleus; n, nucleolus. Bars, 5[tm.
A __ M
V ~: ~:= ~9
Figure 32 Ca2+ trapping in the normal appearing, lower integument PC.
Free Ca2+ was localized in an attempt to define a flux and organelle interaction of
this ion during PCD of the cell. (A) A normal appearing cell shows low Ca2+ in
the cytosol and mitochondria but higher concentrations in the nucleus and very
high concentrations sequestered in the vacuole. (Inset: enlarged image of
mitochondria). (B) Ca2+ localizes to the ER lumen and nucleus. (C) Ca2+ has
collected in the mitochondrial matrices and vacuoles. However, Ca2+ is in low
concentrations in the cytosol. (D) A cell showing a nucleus with mostly
condensed chromatin. The nucleus still shows moderate amounts of Ca2+ and
the mitochondria have small amounts of Ca2+ in their matrices; one displays great
swelling at this stage._(E) Negative control. N, nucleus; V, vacuole; m,
mitochondria; ER, endoplasmic reticulum. Bars: A, D, 0.8[tm; B, C, E, 0.5[m.
Figure 33 In-gel nuclease activity assays for Ca2+, Mg2+ and Zn2+-dependent nucleases and pH controls.
PC tissue was isolated from the rest of the caryopsis and proteins from thr PC were extracted for nuclease activity.
Whether tested together or separately, the Ca +, Mg2+ -dependent nucleases resulted in the same banding pattern.
(A) The Ca2+-dependent assay shows greatest nuclease activity with a 36.5kD nuclease (arrow) first appearing at
and peeking in activity at 8 DAP. (B) In the Zn2+-dependent assay, nuclease activity is greatest at 8 DAP The
same pattern of banding is seen as in the Ca2+- dependent nuclease assay but unlike the Ca2+-dependent nuclease
activity assay a fourth nuclease is present at approximately the 35kD mark (red arrow). However, this nuclease is
seen starting at 6 DAP, persists throughl2 DAP and is still seen faintly at 28 DAP. (C) Most of the Ca2+-dependent
nuclease activity was eliminated when incubated at an acidic pH of 5.5, that which is required for Zn2+-dependent
nuclease activation. Ca2+-dependent nuclease activity requires a neutral pH of 6.8. (D) Most of the Zn2+-
dependent nuclease activity was eliminated when incubated at a neutral pH of 6.8, that which is required for Ca2-
dependent nuclease activation. Zn2+-dependent nuclease activity requires an acidic pH of 5.5.
M, size marker.
8DAP 12DAP 28DAP Un-P. M
m1-- U -
M 6DAP 8DAP 10DAP 12DAP 28DAP
M 6DAP DAP AP 12DAP 28DAP
M 6DAP BDAP 10DAP 12DAP 2ODAP
... m a
32 5 kD
M 6DAP BDAP 10DAP 12DAP 28DAP
Figure 34 In-gel nuclease activity assays with EDTA and EGTA and no metals.
PC tissue was isolated from the rest of the caryopsis and proteins from the PC were extracted for nuclease activity.
(A) EDTA and EGTA effectively reduced or eliminated nuclease activity. Only the smallest nuclease showed any
activity. However when the Zn2+ gel (not shown) was incubated with these chelating agents there was no inhibitory
effect. This may have been due to the acidic conditions required for Zn2+ activation limiting the chelating abilities of
the EDTA and EGTA. (B and C) When metals were eliminated from the incubation medium the same typical
pattern for each of the nucleases still appeared. This is, presumably,due to metals inherent in the tissue.
M, size marker.
6DAP 8DAP 10DAP 12DAP 28DAP
C Zn2* + EDTA & EGTA
6DAP 8DAP 10DAP 12DAP 28DAP
6DAP 8DAP 10DAP 12DAP 28DAP
6DAP 8DAP 10DAP 12DAP 28DAP
A Ca2+/Mg2++ EDTA & EGTA
Figure 35 Genomic DNA isolated from the PC region.
Genomic DNA was isolated from PC tissues of 8, 10, 12, 16, 20, and 24 DAP
caryopses and electrophoresed on a two per-cent gel to test for internucleosomal
DNA laddering. No nucleosomal DNA laddering was detected in the PC. LML:
Low mass ladder.
PCD in the PC Region
Plants undergo PCD as a normal developmental process in a wide variety
of cell types and exhibit PCD in a variety of different ways, morphologically.
Examples include aerenchyma formation, tracheary element differetiation,
compound or hole developing leaves, and senescence (Jones and Dangl, 1996;
Gietl and Schmid, 2001; Obara et al., 2001). PCD occurs within the different
parts of a seed as a part of normal growth and development (Domingues et al.,
2001; Schmid et al., 1999; Wojciechowska and Olszewska, 2003).
Within the maize caryopsis, PCD has been reported in cells of the nucellus,
suspensor, scutellum, and endosperm (Giuliani et al., 2002; Young and Gallie,
2000; Young et al., 1997). All these processes are highly controlled and critically
timed. However, within a temporal context the first of these processes to be
observed is PCD in the PC region. The PC region of the maize caryopsis is the
only port of entry through which water and nutrients pass from the maternal
sporophytic tissue to the filial tissue. While the PC is, itself, a maternal tissue, it
goes through a fertilization dependent programmed cell death (PCD) as part of a
normal developmental regimen (Kladnik et al., unpublished). As detailed in the
results section there are two distinct zones in the PC that undergo two distinct
PCD processes. The PC in maize is comprised of only nucellar and integument-
derived tissues as maize seeds do not possess a funiculus. The first of these
two zones is that portion of the PC that is derived from the nucellus epidermis.
The nucellar-PC zone undergoes an extremely rapid death starting at
approximately 6 DAP and ending at about 8 DAP. This process does not
resemble an apoptotic-like cell death. These cells do not test TUNEL positive at
any point tested indicating a lack of detectable DNA cleavage nor do the nuclei
ever appear to contain condensed chromatin. In fact, no intermediate stages of
cellular degradation were seen between the completely normal appearing cells
and those cells with degraded cytoplasmic remains with the exception of an
occasional plasmolysed cell. Since these cells start out with a very large vacuole
at the time that they appear intact, but have no trace of a vacuole when they are
plasmolysed or degraded, perhaps it is a death mediated by vacuolar rupture and
autolysis similar to that of tracheary elements (TE) (Obara et al., 2001). It is
through this process in TE differentiation that proteolytic and digestive enzymes
are released into the cytoplasm (Yu et al., 2002). Obara et al. (2001) reported a
complete degradation of the nucleus within 15 minutes and a complete
degradation of the cytoplasm within two hours after the rupture of the central
vacuole. The DNA degradation appeared to start in the center of the nucleus
while the condensed chromatin left at the periphery was digested last leaving
behind only the nuclear envelope. While TUNEL assays showed positive results
in the nuclei of these cells there was no oligonucleosomal ladder detected on a
The second zone of cells of the PC is that portion which is derived from the
inner integument. As with the nucellar-PC, PCD in the integument-PC is
detected starting at approximately 6 DAP, peaks at about 10 DAP, then subsides
after 12 DAP. This PCD process morphologically resembles apoptosis in animal
cells (Kerr et al., 1995) in that both undergo extensive chromatin condensation
delineated by sharp clean edges and retention of the nucleolus, nuclear
shrinkage, nuclear body formation, cytoplasmic condensation and retain all
organelles and membranes until these are the only cellular components left. The
cell then completely degenerates and the cellular contents become mostly
absorbed. While these similarities are shared, the exact pattern of chromatin
condensation in the PC nuclei is a little different than that in animals in that it
does not appear to amass at the nuclear periphery.
PCD in the PC May Be Tied to Growth and Development of the Filial
Since the maternal tissue of the pedicel and PC is the only region through
which nutrients can pass to the developing filial tissues of the endosperm and
embryo it is apparent these three tissues are intimately linked, physiologically.
Schel et al. (1984) examined early developmental stages of endosperm and
embryo interaction in maize. They noted that the endosperm precedes the
embryo in development and that there is a 4X increase in size of the endosperm
relative to the embryo from 4 through 8 DAP, creating a high sink demand. From
8 DAP through 14 DAP the endosperm suddenly increases its relative growth
rate; within this period at approximately 9 DAP, the embryo starts a rapid
increase in growth rate and size thus further increasing the sink demand. After
14 DAP the growth rates of each, the endosperm and the embryo, starts to slow;
the endosperm's rate of growth slows prior to that of the embryo's (Schel et al.,
1984). It has been noted in past anatomical studies of the developing maize
caryopsis that as the tissues develop, the nucellus tissue is destroyed by the
growing endosperm, which is in turn destroyed by the growing embryo
(Weatherwax, 1930). Weatherwax (1930) points out that the bordering cells at
the interface of these tissues display a regular pattern in that the cells of the
tissues being destroyed, the nucellus, for example, plasmolyze before
destruction while the growing tissue border cells, the endosperm, for example,
remain turgid. While it was not known what caused this phenomenon it appeared
that the majority of nutrition was still delivered via the PC (Weatherwax, 1930).
However, it must be pointed out that recent findings have shown that these
tissues undergo PCD as the caryopsis develops and that it is not, simply, the
mechanical pressure of the growing tissues that is destroying the extant tissues,
i.e. the destruction of the nucellus as the endosperm grows (Linnestad et al.,
1998). The osmoticum of the tissues in the caryopsis has been shown to be a
critical factor for normal maize seed development (Zinselmeier et al., 1999). For
example, it has been shown that there must be a critical level of sucrose
available to the developing tissues in the seed immediately after pollination or the
development of the embryo ceases and it is aborted. This was accompanied by
the observation that when the maize seeds were kept at very low water potentials
at the time of fertilization the starch reserves in the fruit wall were depleted and
were most likely an attempt, by the seed, to maintain proper sugar levels in order
to maintain adequate water levels for development (Zinselmeier et al., 1999). In
the opinion of the author of this present study, perhaps one explanation for the
ability of the growing border tissues to maintain turgidity is that they maintain a
higher water potential than their surrounding tissues. For example, the nucellus
is not a starchy tissue nor is it a physiological sink, as is the endosperm and
embryo, and the cells contain very large central vacuoles. Perhaps the nucellus
acts more as a water reserve for the growing endosperm than a source of
Sink Demand and Assimilate Flow
The growing endosperm and embryo require sugars and other nutrients that
are delivered to the pedicel via the phloem (Weatherwax, 1930; Felker and
Shannon, 1980). The vasculature terminates at the base of the pedicel and the
assimilates must be unloaded from the phloem and pass through the pedicel
parenchyma cells, PC cells, the basal endosperm transfer cells (BETC), then into
the endosperm and growing embryo. However, it is still not clear exactly what
path the assimilates flow through once they have unloaded from the phloem.
Felker and Shannon (1980) proposed a path of flow for the sugars based on
microautoradiographs of the pedicel after the plant was exposed to 14C02 and
was allowed to assimilate the radioactive carbon into sugars. They concluded
the assimilates pass through the pedicel cells symplastically from the phloem
through the PC until they reach the border between the PC and BETC cells at
which point flow must turn apoplastic due to a complete lack of plasmodesmatal
connections (Felker and Shannon, 1980) Felker and Shannon (1980) point out
a division in radiolabeled sucrose accumulation between the central pedicel
parenchyma, PC and the endosperm within the same tissue sample, there being
a noticeable accumulation of radiolabeled sucrose in the pedicel parenchyma
and endosperm but a lack of accumulation through the PC. However, they
stated that there are appreciable limitations in this method. The limitations
discussed mention that there is no way to determine the presence of radiolabeled
sucrose in the "free space" or apoplast. Therefore, this study could not
conclusively determine where sugars could not be found but did shed light on
where the sugars could be found. Additionally this study focused on relatively
mature caryopsiss of 22 DAP which rules out the possible changes in the path
assimilates might take during varying developmental stages, especially the
earliest most stages.
Based on the morphological changes seen in the PC and pedicel during the
this present study, coupled with data from the aforementioned studies, it would
seem that the path of assimilate flow from the phloem to the endosperm must
vary depending on the stage being examined, i.e. symplastic at early stages (0-6
DAP) versus apoplastic at more mature stages (7 DAP and later). This may also
hold true for certain conclusions that can be drawn about the physiological
processes of the growing seed that may be contributing to the initiation of PCD in
Two Separate, Temporal Waves of PCD Occur in the PC and May Be Linked
to the Sink Demand of the Growing Filial Tissues
The endosperm in maize begins cellularization at approximately 4 DAP
during the R1 stage, 0 to 9 DAP, (Esau 1977; Ritchie, 1993). The first wave of
PCD in the PC follows shortly after the beginning of cellularization of the
endosperm at approximately 5 to 6 DAP. The second wave of PCD in the PC
begins during the R3 or "milk" stage, 18 to 22 DAP, when endosperm cellular
division has completed and the cells are maximizing starch filling (Ritchie,1993).
Therefore, there are two physiological processes that may be contributing to the
PCD occurring in the PC cells during the two waves of cell death.
The first physiological contributor during the first wave, starting after 5 DAP,
may be an osmotically induced process due to the strong sink of the developing
endosperm and embryo. Schel et al. (1984) concluded growth rate is slow
through approximately 4 DAP; therefore it can be deduced that nutrient demand
is at a minimum. Since there are abundant plasmodesmata in all cells of the
pedicel at this stage, an average of 7.1 per cell per section, in the nucellus-
derived PC cells, and all cells are normal in appearance, assimilates most likely
travel symplastically throughout the entire pedicel and PC region. Addtionally,
since there is not yet a symplastically discontinuous border between the PC and
the endosperm, assimilates most likely flow symplastically into the growing
endosperm. Schel et al. (1984) also concluded that the endosperm starts an
extremely rapid rate of growth from 5 through 8 DAP. Thus it can be deduced
that assimilate demand increases. This coincides with the rapid PCD in the
nucellus-derived region of the PC at the end of which the cells have almost
entirely emptied and the identifiable plasmodesmata are staring to decrease in
number. Towards the latter end of this stage the endosperm has grown large
enough in size to have formed a loosely packed, symplastically discontinuous,
multi-cell walled border with the PC. In addition, during this period the
integument-derived region of the PC is starting to display TUNEL positive nuclei
(Kladnik et al., unpublished). With a reduction in the number of plasmodesmata
and only hollow corpses of PC cells left in the nucellus-derived PC it can be
deduced that assimilates flow symplastically from the phloem through the
integument-PC. Upon encountering the nucellus-PC, the assimilates would
encounter a disruption in the symplast and change to an apoplastic means of
flow. The flow may still travel through the cell corpses in the nucellus-derived
PC. Since the apoplast is defined as the pathway for water and solute transport
that lies outside the plasma membrane, including inter cellular spaces and xylem
transport (Buchanan et al., 2000), the flow of assimilates through the cell corpses
would still be considered apoplastic since there are no plasma membranes or
cytoplasmic connections. The disruption and conversion of the flow from the
symplast to the apoplast might cause a transient "backup" of hyper-osmotic fluid
in the integument-derived PC cells immediately subtending the nucellus-PC cells
creating an osmotic shock. Studies examining cell wall invertase (INCW2), a cell
wall enzyme that irreversibly cleaves sucrose to glucose and fructose, showed
specific localization to the BETC cells of maize caryopses and more specifically
to the cell wall ingrowths of the BETC cells (Cheng et al., 1996). The cell wall
ingrowths in the BETC cells start developing at approximately 7 DAP and are
mature by 12 DAP (data not shown). It is immediately prior to the development
of the cell wall ingrowths that the PCD process in the PC is initiated and not until
the wall ingrowths, and thus the INCW2 expression, are mature that the PCD
process ceases. In other words, it is concomitant with, and seemingly not until
the sucrose can be readily passed into the endosperm that the PCD process
ends. This shows a correlation between the sink strength of the endosperm, the
sucrose passing through the PC and the possibility of a transient backup of
assimilates. The invertase deficient mutant in maize, mnl-1, is incapable of
cleaving sucrose in the BETC and shows anatomical features of an
underdeveloped endosperm leading to the development of a large gap
separating the pedicel from the endosperm (Cheng et al., 1996). Cheng et al.
(1996) reference a previous study in which a higher concentration of sucrose was
found to exist in the pedicel of the mnl-1 mutant. Because the mnl-1
endosperm is not applying mechanical pressure on the PC and the PC still
shows PCD (data not shown) as does the wild type caryopsis it is reasonable to
speculate that PCD in the PC region of maize is osmitcally influenced.
Schel et al. (1984) noted that both the endosperm and embryo increase
rapidly in size from approximately 9 to 14 DAP This coincides with the highest
peak of TUNEL positive nuclei, i.e. about 10 DAP, in the integument-derived
region of the PC (Kladnik et al., unpublished). Schel et al. (1984) also show a
decrease in the growth rate at 14 DAP which is coincidental with the last signs of
any TUNEL activity seen (Kladnik et al., unpublished). At this point in
development the BETC cells have developed extensive cell wall in-growths, the
PC has formed a symplastically discontinuous, compacted, multi-cell walled
barrier from crushed nucellar-PC cells with the BETC, and the number of
identifiable plasmodesmata have decreased appreciably in the nucellar-PC cells
from an average of 7.1 per cell per section to 2.4.
The second possible contributing factor to the PCD in these cells may play
a role during both waves of PCD but may be the primary cause of PCD during