International Review of Cytology A Survey of Cell Biology

Visit the link below to download the full version of this book:

https://medidownload.com/product/international-review-of-cytology-a-survey-of-ce
ll-biology-3/

Click Download Now

CONTRIBUTORS

Numbers in parentheses indicate the pages on which the authors' contributions begin.

FrantiSek BaluSka (91), Botanisches lnsfitut der Universitdf Bonn, 0-53115 Bonn,
Germany; and lnstitute of Botany, Slovak Academy of Sciences, SK-84223 Bra-
tuskava, Slovakia
Peter W. Barlow (91),IACR-LongAshton Research Station, Department ofAgriculfural
Sciences, University of Bristol, Long Ashton, Bristol BS 189AF, United Kingdom
Cesira Batini (241), Laborafoire de Physiologie de la Motricite, CNRS, Universite
Pierre-et-Marie Curie, CHU Pitie-Salpitriere, 75634 Paris Cedex, France
Martin Catala (241), lnstifut d'Embiyologie Cellulaire et Moleculaire du CNRS et
du College de France, 94736 Nogent-Sur-Marne Cedex, France; and Service
d'Hisfologie-Embiyologie et Cytogenefique, URA CNRS21 15, GroupeHospifalier
Pifie-Salpitriere, 75651 Paris Cedex 13, France
Marcus Fechheimer (29), Department of Cellular Biology, University of Georgia,
Athens, Georgia 30602
Ruth Furukawa (29), Department of Cellular Biology, University of Georgia, Athens,
Georgia 30602
Serrge Kelm (1 37), Biochemisches lnstituf, University of Kiel, 24098 Kiel, Germany
Nicole M. Le Douarin (241),lnsfifut d'Embryologie Cellulaire et Moleculaire du CNRS
et du College de France, 94736 Nogent-Sur-Marne Cedex, France
Roland Schauer (137), Biochemisches lnsfifuf, University of Kiel, 24098 Kiel, Germany
Seiji Sonobe ( l ) , Department of Life Science, Himeji lnstitute of Technology, Hyogo
678-12, Japan
Dieter Volkmann (91), Botanisches lnstifuf der Universifdf Bonn, 0-531 15 Bonn,
Germany

vii
Cell Model Systems in Plant
Cytoskeleton Studies
Seiji Sonobe
Department of Life Science, Faculty of Science, Himeji Institute of Technology,
Harima Science Park City, Hyogo 678-12, Japan

Cytoskeletonsplay an essential role in cellular functions in both animal and plant cells.
In studies of the molecular mechanisms of their functions, a variety of cell model systems,
mainly of animal cells, have yielded much information. With plant cells, cell model
systems have mostly been restricted to studies on the mechanism of cytoplasmic
streaming. Recently, however, there have been several reports of studies employing plant
cell model systems to investigate plant cytoskeletonsthat have revealed new concepts
about their structure and functions. To promote and support a general understandingof
cell model systems, this review attempts to categorize them, present currently known
information on the structure and function of plant cytoskeletons, and offer a possible role
of cell model systems in future studies of plant cytoskeletons.
KEY WORDS: Cytoskeleton, Cell model, Plant cell, Tobacco BY-2 cell, microtubule,
actin filament.

1. Introduction

Since Szent-Gyorgi (1949) introduced the first cell model system, a glycerin-
ated model of skeletal muscle, many types of systems have been used to
analyze the organization and function of cytoskeletons. In the case of
animal cells, many types of cell models have been introduced to study
cytoskeletons, whereas model systems for plant cells have been almost
restricted to studies on the mechanism of cytoplasmic streaming. Recently,
however, several types of plant cell models have been developed and have
contributed new concepts to studies of plant cytoskeletons, which form the
structures characteristic to plant cells.

~nrl,r,,~,liullolHuvleh, of C'vrolog~.Vol. 17.5 1 Copyright 0 1997 by Acadcmic P r c s .

o(174-76Y6197 $ZS.OO All rights of reproduction in any form reserved.
2 SElJl SONOBE

To date, only one review (on model systems of tobacco BY-2 cells by
Shibaoka el al., 1996) has focused on the cell model systems of plant cells.
In this review, I will survey the model systems of plant cells that have
been reported and present a view of cell models in future studies of plant
cytoskeletons.

II. Categories of Cell Model Systems

There are many model systems for studying functions of cytoskeletons, and
it is not possible to refer to all of them here. Therefore, I will attempt to
categorize the model systems with some examples based on their structural
aspects. This categorization should help readers understand general con-
cepts of cell model systems and enable them to prepare such systems in
their own work.

A. Cell Models

In order to analyze the function of cytoskeletons inside cells, the plasma

membrane, which constitutes a barrier against external conditions, should
be permeabilized. However, for understanding the function of cytoskeletons
of whole cells, cells should retain their intrinsic morphologies even after
permeabilization.
Permeabilized cells that retain their intrinsic morphologies sometimes
show remarkable clontraction upon addition of ATP (Hoffmann-Berling,
1954a). In some cases, movements or contractions are restricted to a portion
of a cell, such as chromosome movement (Cande and Wolniak, 1978),
cytokinesis (Hoffmann-Berling, 1954b), cytoplasmic streaming (Shimmen
and Tazawa, 1983a1, and axonemal movements of ciliate or flagellate cells
(Gibbons and Gibbons, 1972). In order to study the organization of cytoskel-
etons throughout a cell, cells are permeabilized and purified cytoskeletal
proteins are introduced into them to study the protein distribution and/or
the regulatory mechanisms of the cytoskeletal organization (Asada et al.,
1991; Brinkley et al., 1981; Snyder and McIntosh, 1975;Vantard et al., 1990).

B. Isolation of Cytoskeletal Structures

Since Mazia and Dan (1952) succeeded in isolating spindles from sea urchin
eggs, the isolation of functional and structural units consisting of cytoskele-
tons has become a powerful method. Spindles and centrosomes (Mitchison
CELL MODEL SYSTEMS IN CYTOSKELETON STUDIES 3
and Kirschner, 1984) have been isolated from animal cells and spindles
(Cande and McDonald, 1985; Yasuhara et al., 1992), phragmoplasts (Kaki-
mot0 and Shibaoka, 1988), and cortical microtubules (Sonobe et al., 1994)
from plant cells. In some cases, the functions of the isolated organelles
could be reactivated in vifro,for example, nucleation of microtubules (MTs)
in isolated centrosomes (Mitchison and Kirschner, 1984) and spindle elon-
gation (Cande and McDonald, 1985). These studies helped clarify some of
the molecular mechanisms of these functions. Isolation of such organelles
often facilitated identification of the protein(s) involved in their functions
(Asada and Shibaoka, 1994; Toriyama et al., 1988; Yasuhara et ai., 1992).
Interesting phenomena not observed in intact cells have often been ob-
served in artificially modified cells. For example, in a cytoplasmic droplet
isolated from Characean cells that was still “alive,” motile fibrils and rotat-
ing chloroplasts were observed (Jarosh, 1956; Kamiya and Kuroda, 1957).
Although these movements apparently differed from cytoplasmic stream-
ing, the basic mechanism responsible for them was thought to be a mutually
common one, namely, the motive force was generated by interactions be-
tween actin and myosin. Analyses of these “artificial” movements have
yielded much information on the mechanism of cytoplasmic streaming
(Kuroda, 1990). Analyses using isolated cytoplasm were also carried out
for Physarum (Kuroda, 1979) and ameba (Taylor et al., 1973). The motility
of isolated (or naked) cytoplasm showed clearly that contractile elements
existed in the endoplasm (sol) and that they could generate a motive force.
Caffeine drops isolated from Physarum plasmodium were also used to study
the regulatory mechanisms of cytoplasmic streaming (Sato et al., 1981).
Miniprotoplasts that are prepared from protoplasts of plant cells by elimi-
nating vacuoles show cytokinetic cleavage and are classified in this category
(see Section 111,F).

C. In Vitro Reconstruction

A series of experiments using isolated chromosomes clarified the presence

of motor proteins at the kinetochore (Hyman and Mitchison, 1991) and
the significant roles of MT dynamics during mitosis (Coue et al., 1991;
Koshland et al., 1988). Thus, in vitro reconstruction of cytoskeletal functions
may allow us to explain the mechanism in “molecular” terms. Moreover,
if a protein(s) responsible for the function of cytoskeletons can be purified,
we may be able to reconstruct the function in vitro using only purified
proteins . A typical example is the so-called “in vitro motility assay,” in
which sliding between actin and myosin or a MT and its motor proteins
can be visualized (Kron and Spudich, 1986; Sheetz and Spudich, 1983a;
Vale and Yano-Toyoshima, 1988). The reconstruction of centrosome-like
4 SElJl SONOBE

structures from which MTs radiated has been done using purified proteins
(Toriyama et al., 1988).

111. Cell Models of Plant Cells

A. Cytoplasmic Streaming

A subject that has been well documented among the roles of cytoskeletons
of plant cells is cytoplasmic streaming. Because the details of the mechanism
have been presented in many original papers and reviews (Kamiya, 1981;
Kuroda, 1990; Shimmen and Yokota, 1994), I will only refer to the studies
that employed cell model systems for examining cytoplasmic streaming.

1. Perfused Cells
The internodal cell of Characeae has played a central role in studies on
the mechanism of cytoplasmic streaming. The cell has a large cylindrical
shape and a large central vacuole. Kamiya and Kuroda (1 955) first applied
a perfusion technique to internodal cells and this technique was later
improved (Tazawa, 1964). The perfused cells are still alive because only
the vacuole is perfused. Although such vacuole-perfused cells have con-
tributed to the study of vacuole functions (see Shimmen et al., 1994), it
was not thought to be an ideal cell model for studying the mechanism of
cytoplasmic streaming because exogenously applied substances exerted
their effects through the tonoplast. Subsequently, a perfused cell from
which the tonoplast was removed was made by perfusing the cell with a
solution that contained ethyleneglycol-bis(P-amhoethylether)-N,N,N',N'-
tetraacetic acid (EGTA) (Tazawa et al., 1976; Williamson, 1975). In ATP-
depleted tonoplast-free cells, movements of organelles that are tightly as-
sociated with actin bundles were reactivated upon addition of Mg-ATP
(Williamson, 1975) and the movements were regulated by Ca2+(Shimmen
et al., 1984). Tonoplast-free cells have also been used for detecting myosin
activities by introducing organelles derived from heterogenous cells (Adams
and Pollard, 1986;Kohno and Shimmen, 1988; Shimmen and Tazawa, 1982a;
Sheetz and Spudich, 1983b), skeletal muscle myosin (Shimmen and Yano,
1984), and myosin-coated beads (Sheetz and Spudich, 1983a). Cytoplasmic
streaming was also reactivated on an internodal cell that had been cut open
longitudinally (Kuroda, 1983).

2. Permeabilized Cells
Another type of cell model of Characeae cells was a plasma membrane-
permeabilized cell model. Shimmen and Tazawa (1983a) succeeded in per-
CELL MODEL SYSTEMS IN CYTOSKELETON STUDIES 5
meabilizing Churaceae cells by inducing rapid plasmolysis. When the cell
is transferred from a solution with low osmolarity to one with high osmolar-
ity in the presence of EGTA, the plasma membrane rapidly becomes de-
tached from the cell wall and this causes breakage of the plasma membrane.
To make the plasma membrane labile, EGTA must be added to both
solutions and all treatments must be performed under cold conditions. As
a result, ATP in the cytoplasm flows out and the cytoplasmic streaming
stops. It can be reactivated by adding ATP. This method is considered to
be useful for permeabilizing the plasma membrane of other plant species
because it does not seem to require cells as large as Characean cells. In
fact, successful permeabilization of the plasma membrane has been reported
for potato tubular slices (Ponstein et al., 1990), isolated and in situ mesophyl
cells of Barley leaves. suspension-cultured cells of Cutharuntus roseiis (Mi-
mura and Shimmen, 1992), and protoplasts of suspension-cultured soybean
cells (Saleem and Cutler, 1986), although these studies did not focus on
the cytoskeleton. If the cell wall is partially digested and then permeabilized
by osmotic change, it is also expected that large molecules such as proteins
will be introduced. Permeabilization of Characene cells using detergent
(Shimmen and Tazawa, 1982a) and electroporation (Shimmen and Tazawa,
1983b) has also been reported.

3. Cytoplasmic Droplets
When one end of an internodal cell is cut, endoplasm flows out to form
droplets that are surrounded by membrane. The cytoplasmic droplets dis-
play various patterns of movements; that is, moving motile fibrils and rota-
tion of chloroplasts and nuclei (Jarosh, 1956; Kuroda and Kamiya, 1975),
which are not seen in vivo. Also, the chloroplasts continue to rotate even
after removal of the surface membrane of the droplets in the presence of
Mg-ATP. Moreover, the chloroplasts are reactivated by heavy meromyosin
prepared from skeletal muscle after suppression of the function of the
putative Nitella myosin (Kuroda and Kamiya, 1975), indicating that their
rotation was induced by interaction between actin and myosin. Movements
of fibrils that were identified as actin filaments were also observed in
squeezed cytoplasm (Higashi-Fujime, 1980).

4. Other Plants
There have been few reports employing cell models of other plant cells.
Takata (1961) succeeded in reactivating the glycerinated model of Acetahu-
laria stalk by Mg-ATP and Ca”. Cytoplasmic streaming in Acetubulariu
was also studied by employing perfused cell (Nagai and Fukui, 1985) and
isolated cytoplasm in vitro (Menzel and Elsner-Menzel, 1989). La Claire
6 SElJl SONOBE

(1984) reported that a Triton-permeabilized cell model of the coenocytic

green alga Ernodesrnis showed remarkable cytoplasmic contraction, which
was thought to reflect a motile reaction during wound healing of this alga
in vivo, upon addition of ATP in the presence of M Ca2+.

6.Cortical Microtubules

Since “microtubules” were first found ultrastructurally by Ledbetter and

Porter (1963) in the cortical region of pea epicotyl, the orientation of MTs
in the cortical region, so-called cortical microtubules (CMTs), has been
examined by electron and immunofluorescence microscopy (Gunning and
Hardham, 1982; Williamson, 1991; Lloyd, 1987).
CMTs are thought to regulate cellular morphogenesis by controlling the
direction of cellulose microfibril deposition on the outer surface of the
plasma membrane (Delmer, 1987; Giddings and Staehelin, 1991), and their
orientation and organization are regulated by many internal and external
factors, such as plant hormones (Shibaoka, 1994), light irradiation (Iwata
and Hogetsu, 1989; Laskowski, 1990; Zandomeni and Schopfer, 1993), pro-
gression of cell cycle (Ledbetter, 1967), aging (Hogetsu and Ohshima,
1986), and mechanical forces (Williamson, 1990). However, the molecular
mechanism of CMT organization remains unsolved.

1. Perfused Cells
Wasteneys and Williamson (1989a) observed MT assembly in perfused
Nitella internodal cells after introduction of biotinylated brain tubulin.
Exogenous tubulin polymerized in the endoplasm to form MT bundles and
some of them were associated with nuclei. In the cortical region, a limited
number of MTs was observed at the site where a normal array of chloro-
plasts was retained, whereas a mass of MTs was observed at the edge of a
region where chloroplasts had been removed, suggesting the presence of
a factor that nucleated MTs at the cortical region. These results suggested
the presence of sites responsible for nucleation of MTs in the cortical region.
Chloroplasts that associated with the plasma membrane were thought to
interfere with exogenous tubulin accessing the sites.

2. Membrane Ghosts
When protoplasts burst after attachment to the surface of polylysine-coated
coverslips in a hypotonic solution, they leave fragments of the plasma
CELL MODEL SYSTEMS IN CYTOSKELETON STUDIES 7
membrane on the coverslip (Figs. l a and lb). This technique was introduced
by Marchant (1 978) in studies of CMTs of the green alga Mougeotia and
was based on a technique used for ameba of slime mold Dictyostelium
(Clark et al., 1975). Thus, the membrane ghosts have been thought to be a
kind of model of CMTs because the CMTs on the ghosts can be manipulated
directly. Kakimoto and Shibaoka (1986) have suggested a significant role
of the plasma membrane in stabilizing CMTs. CMTs of membrane ghosts
of Mougeotia were resistant against the mM level of Ca” but were depoly-
merized by Ca2+after removal of the plasma membrane by treatment with
Triton X-100. These results suggested the presence of a factor that conferred
stability to CMTs on the plasma membrane. Kakimoto and Shibaoka also
observed projections on MTs and cross-bridge structures between neighbor-
ing MTs, suggesting a significant role of microtubule-associated proteins
(MAPS) in the organization of CMTs. Involvement of a transmembrane
protein(s) in stabilization of CMTs was demonstrated by Akashi and Shi-
baoka (1991). The effect of Ca2+and calmodulin on CMT stability was also
reported in the membrane ghosts (Cyr, 1991).
CMTs on the membrane ghosts disappeared upon treatment with ATP
(Sonobe and Shibaoka, 1990; Katsuta and Shibaoka, 1992). These results
and those showing that the arrangement and stability of CMTs were affected
by kinase inhibitors (Mayumi and Shibaoka, 1996; Mizuno, 1992) suggested
the participation of protein phosphorylation in regulation of CMT organiza-
tion. Phosphorylation of a 65-kDa plant MAP (Jiang and Sonobe, 1993)
that was shown to be colocalized with CMTs by endogenous kinase was
found in isolated CMTs (Yamamoto et al., 1994, see below), but its function
in CMT organization is not known.
An experimental system in which CMTs were reconstructed has been
developed using both membrane ghosts and a cytoplasmic extract of mini-
protoplasts of tobacco BY-2 cells (Sonobe and Takahashi, 1994). When
CMTs on the ghosts (Figs. l a and l b ) were removed by incubation with a
high concentration (mM order) of Ca” at low temperature (Figs. l c and
Id) and then CMT-free ghosts were incubated with a cytoplasmic extract
of miniprotoplasts, MTs reappeared on the ghosts (Figs. l e and lf), but
none did when the ghosts were preincubated with trypsin (Figs. l g and
lh). These results indicated that on the ghosts there was the presence of
a factor that facilitates association of MTs with the plasma membrane. The
factor was thought to be a membrane-associated protein because a high
concentration of KCl prior to incubation with the extract inhibited the
reappearance of MTs on the ghosts. In comparison with the system employ-
ing the introduction of purified tubulin to a cell model, such as perfused
Characean internodal cell (Wasteneys and Williamson, 1989a), the cyto-
plasmic extract is thought to be more complicated, but all components
responsible for CMT organization are present. Wasteneys and Williamson
a SElJl SONOEE

FIG. 1 Fluorescence micrographs of membrane ghosts of tobacco BY-2 cells. Freshly prepared
membrane ghosts (a and b) were incubated with cold Ca*'-containing solution to remove
preexisting CMTs for 60 rnin (c and d) and then the ghosts werc incubated with the cytoplasmic
extract of miniproloplasts for 10 min (e and f). MTs reappeared on the ghosts but not o n the
ghosts pretreated with trypsin prior to incubation with the extract (g and h). Fluorescent (a,
c, e , and g) and phase-contrast images are shown. Bar = 20 p m .
CELL MODEL SYSTEMS IN CYTOSKELETON STUDIES 9
(1989a) pointed out the possibility that reassembly of exogenous tubulin
at the cortical region of the perfused NitellLr cells without any branching.
which had been observed in intact cells during recovery from disassembly
of CMT with oryzalin (Wasteneys and Williamson, 1989b). should be due
to the absence from purified tubulin of a factor that was essential for CMT
rearrangement.
As has been repeatedly demonstrated by electron microscopy (Hardham
and Gunning, 1978; Seagull and Heath, 1980; Lancelle el ul., 1986;Giddings
and Staehelin, 1991), cross-bridge structures exist between CMT and the
plasma membrane and between neighboring MTs. Although these struc-
tures are thought to play an important role in CMT organization, they have
not been biochemically charactcrized. The system mentioned previously
should be a powerful tool for biochemically identifying cross-bridge proteins
by detecting their ability to reconstruct MTs on the ghosts.

3. Isolated CMTs
As described previously, in order to understand the mechanism of CMT
organization, cross-bridges between CMTs and the plasma membrane and
between neighboring MTs must be characterized biochemically. We first
attempted to isolate cross-bridge proteins from membrane ghosts because
the finding that KCI extraction of membrane ghosts inhibited attachment
of MTs seemed to indicate that a high concentration of KC1 could release
cross-bridge proteins from the plasma membrane. However, there was
difficulty in obtaining a sufficient amount of proteins from membrane
ghosts. Therefore, we attempted mass isolation of CMTs (Sonobe ef al.,
1994; Yamamoto ef al., 1994).
Protoplasts of tobacco BY-2 cells were gently homogenized in the pres-
ence of taxol and separated by centrifugation through a density gradient
of Percoll. The uppermost layer in the centrifuge tube contained vesicles
to whose membrane CMTs were attached (Fig. 2a). CMTs were isolated
from these vesicles by solubilization of the membrane with Triton X-100.
Cross-bridge structures remained associated with the isolated CMTs (Fig.
2b). After cycling of depolymerization and polymerization of the isolated
CMTs, we found that four kinds of polypeptides (i.e., 200-, 120-, SO-. 65-
kDa polypeptides) remained associated with MTs. The 65-kDa polypeptide
was previously isolated by Jiang and Sonobe ( I 993).

C. Phragmoplast

Cytokinesis of higher plant cells occurs by the formation of a cell plate

that centrifugally develops after mitosis. The cell plate is thought to be
10 SElJl SONOBE

FIG. 2 Electron micrographs of isolated plasma membrane vesicles and CMTs. Plasma mem-
brane vesicles were isolated from protoplasts of tobacco BY-2cells under conditions in which
CMTs were preserved (a). CMTs were isolated from the vesicles by solubilization with Triton
X-100 (b). In this case, complexes of CMTs and the plasma membrane were isolated at a low
concentration of Triton X-100 (0.2%).Bar = 0.2 pm.
CELL MODEL SYSTEMS IN CMOSKELETON STUDIES 11
formed by fusion of Golgi apparatus-derived vesicles (Hepler, 1982;Whaley
and Mollenhauser, 1963) that are thought to be transported in an MT-
dependent manner (Bajer, 1968; Gunning, 1982; Hepler and Jackson, 1968;
Yasuhara et al., 1993). Two sets of MTs at the edge of the cell plate
are organized perpendicular to it with opposite polarities (Euteneuer and
McIntosh, 1980), namely, each set of MTs places the plus end at the cell
plate. This organization seems to be reasonable considering the fact that
Golgi-derived vesicles are transported to the cell plate from both directions.
Observations by Inou6 (1964) demonstrated the important role of phrag-
moplast in constructing MT organization at the place where the plus end
of MTs are interdigitated.

1. Permeabilized Cells
Microtubule organization in the phragmoplast was studied with cell model
systems. Vantard et al. (1990) introduced exogenous Paramecium tubulin
into saponin-permeabilized endosperm cells of Haemanthus, which have
no cell walls. Cells were treated with a lysis buffer containing saponin and
tubulin that was purified from Paramecium axonemes. The incorporated
tubulin was detected with a specific antibody against Paramecium tubulin
to distinguish it from preexisting MTs of Haemanthus. The Paramecium
tubulin was incorporated at the equatorial plane of phragmoplast and nu-
clear surfaces, suggesting its function in the MT organization of these
regions in plant cells.
Asada et al. (1991) succeeded in reactivating translocation of MTs of
phragmoplasts in a glycerinated cell model of tobacco BY-2 cells. Tobacco
suspension-cultured cells, BY-2 cells, whose cell cycle was synchronized
at anaphase or telophase (see below for the synchronization method)
were treated with a wall-digesting solution and subsequently with a so-
lution containing 60% glycerol solution on ice to permeabilize the plasma
membrane. Bovine brain tubulin labeled with dichlorotriazonil amino-
fluorescein (DTAF) was introduced into glycerinated BY-2 cells in the
presence of nucleotides. DTAF tubulin was initially incorporated into the
equatorial plane of the phragmoplast. The incorporated tubulin was de-
tected as a single bright line with a width that increased with time. When
nonlabeled tubulin was introduced after incubation with DTAF tubulin,
the bright band at the equatorial plane separated into two parts. These
observations could be explained as the incorporation of exogenous tubulin
at the plus end of preexisting phragmoplast MTs and simultaneous translo-
cation of MTs toward their minus ends. GTP was preferred over ATP for
inducing MT translocation. The nonhydrolyzable analog of GTP or ATP,
GMPPMP or AMPPMP, could induce only polymerization of tubulin at
the plus ends of MTs, and subsequent application of GTP or ATP induced
12 SElJl SONOBE

MT translocation even in the absence of exogenous tubulin. These findings

predicted the presence of GTP or ATP at the equatorial plane of an MT-
dependent motor protein. In fact, Asada and Shibaoka (1994) isolated
kinesin-related motor protein from phragmoplasts of tobacco BY-2 cells,
which translocates MTs toward their minus end in vitro, as observed in
glycerinated models.

2. Isolated Phragmoplasts
Isolation of phragmoplasts has long been a target for enabling the study
of their structure and function. Because phragmoplasts appear for only a
short period throughout the cell cycle, the cell cycle needs to be synchro-
nized for their mass isolation.
Tobacco BY-2 cells have a characteristic feature with respect to their
exceptionally high rate of growth compared with that of other cell lines of
plant cells (Nagata et al., 1992). Nagata et al. (1981) first introduced a
technique of synchronization of the cell cycle of BY-2 cells using aphidicolin,
an inhibitor of DNA polymerase-a of eukaryotes. Kakimoto and Shibaoka
(1988) developed the technique by introducing metaphase arrest using
propyzamide, a specific inhibitor of polymerization of plant tubulin (Akashi
et al., 1988). They obtained a mass culture of BY-2 cells with the cell cycle
synchronized at the mitotic phase for a mitotic index of more than 90% and
isolated phragmoplasts after wall digestion and subsequent solubilization of
the plasma membrane by Triton X-1 00. The isolated phragmoplasts retained
MTs accompanied by daughter nuclei but not actin filaments, which are
known as another cytoskeletal component of the phragmoplast. Microtu-
bules of isolated phragmoplasts became depolymerized after treatment
with a cold Ca2+-containing solution, whereas daughter nuclei remained
associated with the isolated phragmoplasts, suggesting the presence of a
connection structure(s) other than MTs and actin filaments between phrag-
moplasts and daughter nuclei. Actin filaments were preserved in the isolated
phragmoplasts by adding tropomyosin and myosin heavy meromyosin
(HMM) purified from skeletal muscle to the isolation buffer. Using this
technique, the polarity of actin filaments of phragmoplasts was determined;
arrows formed by HMM pointed toward the daughter nuclei. Small vesicles
that were thought to be derived from the Golgi apparatus associated with
MTs and cross-bridge structures between these vesicles and MTs were
observed (Kakimoto and Shibaoka, 1988; Yasuhara et al., 1993), suggesting
the presence of a protein(s) responsible for MT-dependent vesicle transport
in the phragmoplast.
Yasuhara et al. (1992) identified a 100-kDa polypeptide in the ATP
extract of isolated phragmoplasts and spindles. This polypeptide was bound
to MTs in vitro and dissociated upon addition of ATP as well as AMPPNP.
CELL MODEL SYSTEMS IN CYTOSKELETON STUDIES 13
Because of the lack of ATPase activity in the 100-kDa polypeptide, it was
not thought to be a motor protein. Its function in phragrnoplasts and spin-
dles is still unknown.
During development of the cell plate, polysaccharides, which are major
components of the cell wall, are synthesized. Kakimoto and Shibaoka (1992)
succeeded in synthesizing polysaccharides in isolated phragrnoplasts. They
isolated phragmoplasts from protoplasts of synchronized BY-2 cells by
pushing through a nylon mesh to remove the plasma membrane instead of
using solubilization with detergents, which might disturb the polysaccharide
synthesis thought to be closely dependent on the plasma membrane. Poly-
saccharide synthesis occurred at the equatorial plane of isolated phrag-
rnoplasts when they were incubated with a buffer containing UDP-glucose.
The synthesized polysaccharide at the equatorial plane was thought to be
/3-1.3-glucan because /3-1,3-glucanase solubilized them. On the other hand,
when isolated phragmoplasts were incubated with a buffer containing UDP-
xylose, polysaccharide synthesis occurred in Golgi-like structures that had
been associated with isolated phragmoplasts. These results suggested that
P-1.3-glucan is synthesized in the cell plate, whereas xyloglucan is synthe-
sized in the Golgi apparatus and transported to the cell plate.

D. Microtubule-Organizing Center

In animal cells, MTs are usually organized by characteristic structures, such

as centrosomes, spindle poles, and basal bodies. These structures are called
microtubule-organizing centers (MTOCs) and their biochemical composi-
tions have been studied extensively (Brinkley, 1985; Stearns and Kirschner,
1994). In contrast, we have little information on MTOCs in higher plant
cells because of the absence of discrete structures responsible for MT
organization in plant cells, such as the centrosomes with centrioles found
in animal cells.
Attention has been drawn to the nucleus of higher plant cells as a candi-
date for MTOCs in higher plant cells from observations that MTs were
closely associated with the nuclear surface in endosperm cells of Heaman-
thus (De Mey et aL, 1982; Lambert, 1980, 1993), in root tip cells of Allium
(Falconer et al., 1988), and in tobacco BY-2 cells (Katsuta et al., 1990).
Observations that MT reappearance occurred on the nuclear surface during
restoration from MT disruption (Wasteneys and Williamson, 1989a,b) and
that MTs were nucleated on the nuclear surface in living cells rnicroinjected
with fluorescently labeled tubulin (Wasteneys et al., 1993; Zang et al., 1990)
and in cell models to which exogenous tubulin had been introduced (Vant-
ard et al., 1990; Wasteneys and Williamson, 1989a) also indicated an irnpor-
tant role of the nucleus in MT organization. Microtubules associated with