Directory UMM :Data Elmu:jurnal:P:PlantScience:Plant Science_BioMedNet:081-100:
58
Division decisions and the spatial regulation of cytokinesis
Anne W Sylvester
Cytokinesis in plant cells is accomplished when a membranous
cell plate is guided to a pre-established division site. The
orientation of the new wall establishes the starting position of a
cell in a growing tissue, but the impact of this position on
future development varies. Recently, proteins have been
identified that participate in forming, stabilizing and guiding the
cell plate to the correct division site. Mutations that affect
cytokinesis with varying impacts on plant development are
providing information about the mechanics of cytokinesis and
also about how the division site is selected.
Abbreviations
dcd
discordia
KCBP
kinesin-like calmodulin binding protein
KatAp
kinesin from Arabidopsis thaliana A peptide
MAPK
mitogen-activated protein kinase
MMK3
Medicago mitogen-activated protein kinase 3
MT
microtubule
NPK1
Nicotiana protein kinase 1
PP
phragmoplast
PPB
preprophase band
TKRP125 tobacco kinesin-related polypeptide of 125 kDa
wty1
warty1
equatorial zone between recently formed daughter nuclei
[2–5]. The growing cell plate then spreads out radially, as
an expanding interlacing membranous network, until it
joins with the parental walls at the periphery. A
cytoskeletal structure called a phragmoplast (PP) not only
guides vesicles to the equatorial zone but also directs the
growing cell plate toward a specific site at or near the
membrane of the parental cell wall — the cortical division
site. This ill-defined site is first evident prior to mitosis
when the preprophase band (PPB) appears transiently as
a continuous band of cytoskeletal elements in the cell
cortex. Remarkably, the position of the PPB forecasts the
future site of attachment between the cell plate and the
parental cell walls. The integrated behavior of these
cytoskeletal structures thus accomplishes the feat of
depositing a new wall at a specific position. What is the
mechanism of cell plate formation? What determines
where this new wall will form? How does cell position
affect future development? This review presents recent
progress in answering these questions and also evaluates
potential versatility in the process, considering that some
divisions have more impact on future development than
others. Recent advances in understanding cell cycle regulation [6], mechanisms of mitosis [7] or cell expansion
[8,9 (DJ Cosgrove, pp 73–78)] are relevant topics, but are
not covered here. Excellent diagrams showing the
detailed structure of the growing cell plate and PP are
available elsewhere [2,4,5].
Introduction
Forming the cell plate
Plant cell division serves a dual function of increasing cell
number and also positioning new cross walls during cytokinesis. Once the new wall is formed, cells acquire a final
shape by differential cell expansion. Several important
decisions are thus made during division and expansion,
including when and how often division should occur,
where a new cell wall should be located, and in what direction to expand. The outcome of any one of these decisions
could be critical to the successful development of an organ
because new daughter cells are joined from inception by
cell walls. In reality, however, the position of new walls has
a variable influence on the shape of an organ or function of
a tissue. This is because the spatial regulation of cytokinesis, which determines where new walls will be positioned,
affects development to different degrees [1]. Regardless of
the impact of division pattern on development, successful
cytokinesis is needed to maintain the integrity of the plant
body. Understanding how and when cytokinesis occurs will
help to evaluate the effect of different stages of development on its spatial regulation.
Completion of cytokinesis requires the interaction of fusing secretory vesicles, dynamic microtubules (MTs), actin
and numerous associated proteins (Figure 1), as confirmed
by electron and immunofluorescence microscopy and by
experimental manipulation of cytokinesis. Disruption of
the Golgi complex with the inhibitor brefeldin A, for
example, prevents completion of the cell plate, suggesting
that new vesicles are Golgi-derived [10]. Vesicles will also
not accumulate in the equatorial zone of cells treated with
inhibitors that depolymerize MTs [11], as a result of either
transport inhibition or altered organization of the Golgi
complex [12]. On the other hand, taxol, which hyperstabilizes MTs by preventing their depolymerization, increases
the number of vesicles in the equatorial zone initially,
although cytokinesis is eventually arrested [13]. These
results demonstrate that MTs must be present but not necessarily dynamic for successful vesicle transport,
supporting the idea that vesicles migrate along MTs with
the aid of motor proteins [14].
Addresses
Department of Botany, PO Box 3165, University of Wyoming, Laramie,
Wyoming 82071-3165, USA; e-mail: annesyl@uwyo.edu
Current Opinion in Plant Biology 2000, 3:58–66
1369-5266/00/$ — see front matter © 2000 Elsevier Science Ltd.
All rights reserved.
Vesicle transport
Cytokinesis is accomplished by progressive deposition of
membranes and associated wall synthesizing compounds
into a cell plate, which is first located centrally in the
The exact mechanism of vesicle transport is still an open
question. The bipolar MTs in the PP are arranged with
their plus ends overlapping in the equatorial zone and
Division decisions and the spatial regulation of cytokinesis Sylvester
59
Figure 1
Diagrammatic view of a plant cell in early
cytokinesis. Proteins recently identified in
each compartment of the dividing cell are
shown. Kinesin-related proteins (TKRP125,
KCBP, KatAp) show distribution patterns that
suggest an involvement in organizing the MT
array, rather than transporting vesicles. KCBP
localizes to the nuclear membrane of
endosperm cells during late telophase only.
Several other proteins probably involved in
formation of the tubulo-vesicular membrane
network of the cell plate (barely visible as
wavy lines in the equatorial zone) include
KNOLLE and phragmoplastin. CDC48 may
function with KNOLLE in docking of vesicles
to their target but localization details are not
sufficient to suggest this. Kinases (such as
p43Ntf6) locate to the same region, but the
phosphorylation substrates are not known.
Vinculin-like antigens have recently been
observed in the forming cell plate of
cellularizing endosperm, possibly serving in
association with the actin filaments found
there. Actin appears to link the phragmoplast
to the cortex. The actin-depleted zone might
carry residual actin patches but this
connection is not yet clear.
Actin
Daughter nuclear
membrane
*KCBP
Phragmoplast
MTs
*TKRP125
*KCBP
*KatAp
Actin-depleted
zone
Cell Plate in
equatorial zone
*KCBP
*Phragmoplastin
(& ADL1)
*KNOLLE
*MAPKs (p43 Ntf6 )
*Vinculin antigen
minus ends directed toward the daughter nuclei
(Figure 1), suggesting vesicles might be transported
using a plus-end directed motor protein [5]. A candidate
kinesin-related protein isolated from tobacco PPs
(TKRP125) has the expected domains for motor activity
[15]. The protein localizes, however, along the entire
length of MTs in the PP, rather than in the punctate patterns expected for vesicle association. TKRP125 could
still interact with other as yet unidentified proteins associated with vesicles. In permeablized cells, however,
antibodies against TKRP125 actually inhibit plus-end
MT translocation [15]. Together, these results suggest
that TKRP125 might contribute to maintaining the structure of the MTs in the PP, rather than transporting
vesicles to the cell plate.
Minus-end directed kinesin-like proteins have been found
associated with the PP, including kinesin-like calmodulin
binding protein (KCBP) and a kinesin from Arabidopsis
thaliana A peptide (KatAp). KCBP mediates MT translocation in vitro [16] and its motor domain binds MTs [17•].
KCBP localizes to the MTs in the PP and also to the
nuclear membrane early in telophase and within the equatorial zone [18•]. KatAp is initially present in the spindle
mid-zone before the PP is organized and then in association with the PP itself [19]. The localization patterns for
both proteins are consistent with a function in organizing
the MTs in the PP, rather than in directing vesicle transport
[18•,19], particularly as similar kinesins in other systems
help to assemble and organize the mitotic spindle [20].
The translocation of MTs toward the minus end could propel vesicles toward the plus ends. Most evidence, however,
Vesicles
Current Opinion in Plant Biology
points to an organizational function for these proteins, perhaps mediated by these minus-end directed motors.
F-actin could be a candidate for vesicle motility because it
co-localizes with MTs in the PP, and is oriented as expected
for a motility system [21]. Myosin may also be present in the
PP [22]. Although disruption of actin filaments with cytochalasin does not affect vesicle accumulation, it does disorient
the cell plate [11]. F-actin could facilitate vesicle transport
through stabilizing the MTs, as discussed below but whether
it plays a direct role in vesicle transport is not clear.
Organizing the phragmoplast
MT translocation might help to establish and then maintain the bipolar MT array of the PP [23], possibly mediated
by the kinesin-related proteins discussed above.
Differential MT depolymerization is also important in PP
function [10]. Centrifugal growth of the cell plate is prevented when MTs are stabilized, mirroring an effect of
preventing depolymerization on spindle function [7].
These studies point to a required balance between polymerization at the leading edge and depolymerization at the
trailing edge of the PP. It is likely that similar organizational mechanisms are found for the PP as found for other
MT structures, such as the mitotic spindle.
Vesicle accretion and fusion in the cell plate
Newly arriving vesicles coalesce into a membranous network
as the cell plate extends out radially [4]. Clathrin-coated vesicles are observed in conjunction with the growing
tubulo-vesicular network of the cell plate, providing evidence for fusion and membrane recycling as part of cell plate
60
Growth and development
Figure 2
Figure 3
Recent
cross walls
Cell
size
Cell
differentiation
Pre-differentiation
zone
Current Opinion in Plant Biology
Epidermal cells in a fully-grown maize leaf of (a) a wild-type plant
compared with (b) a wty1 mutant. (a) The cells in a normal leaf are
linear and aligned with the long axis of the leaf blade, as emphasized in
the cell outlines in the right of the panel. (b) The mutant leaf has overly
enlarged cells in the wart, which result from early events of overexpansion and improper cytokinesis (see also Figure 4).
formation [4]. A protein that probably facilitates this process
is phragmoplastin, a dynamin-related protein originally isolated from soybean [24] and also from Arabidopsis (called
Arabidopsis dynamin-like protein 1 [ADL1]) [25]. In animal
systems, dynamin is a GTPase that functions in endocytosis,
particularly in membrane separation [26]. Both phragmoplastin and ADL1 localize to the growing cell plate [24,27].
Phragmoplastin is not dependent on MT organization, as it
remains at the cell plate margins after the PP has dissociated
near the end of cytokinesis. Taxol-stabilization of MTs does
not eliminate phragmoplastin from its earliest position in the
equatorial zone but does prevent its redistribution to the
leading edge [24]. In vivo observations using green-fluorescent protein (GFP)-fusion proteins show that
phragmoplastin associates with sites of high membrane activity [28], consistent with a dynamin-like role for the protein.
Many essential protein functions are regulated by phosphorylation, including dynamin assembly and kinesin
activity [29,30]. Recently, mitogen-activated protein kinases (MAPKs) were found in the cell plate. MAPKs are
cell-cycle regulated proteins involved in organizing the
cytoskeleton and are also abundant in diverse signaling
pathways [31,32]. Two proteins similar to MAPKs, (p43Ntf6
from tobacco and Medicago mitogen-activated protein
kinase 3 [MMK3] from alfalfa) are activated during late
anaphase if functional MTs are present [33•,34•]. MTs are
not required to maintain distribution of MMK3 in the cell
plate once the PP is established [34•]. Another protein,
Nicotiana protein kinase 1 (NPK1) is similar to upstream
kinases in a MAPK chain, referred to as MAPK kinase
kinases [35]. NPK1 is expressed in dividing cells [36•] and
Simplified view of the pattern of division and expansion in the
developing maize leaf primordium. In general, cells in the maize leaf
adhere to an inverse pattern of dividing and expanding, particularly
below the zone where epidermal cells begin to differentiate (predifferentiation zone). Here the gradual increase in cell size is due to the
rate of expansion exceeding that of division. Orientation of new crosswalls can be observed and quantified in leaf surface impressions made
from this region (see Figure 4). Figure reprinted with permission [54•].
complements yeast mutations in homologs to NPK1 [35],
but its localization patterns relative to the cell plate have
not yet been reported. Pending these results, NPK1 in
conjunction with p43Ntf6 and MMK3 could be candidate
transducers in a phosphorylation cascade in the cell plate.
Proteins that participate in vesicle targeting are also found
in the cell plate. One example is the syntaxin-related protein KNOLLE [27,37]. Syntaxins (also called t-SNARES)
are located in target membranes and specifically interact
with proteins in arriving vesicles, called v-SNARES [2,38].
Syntaxins in nonplant systems participate in vesicle fusion
during cytokinesis [39]. Localization of KNOLLE to vesicles and to the equatorial zone suggests it might serve
similar recognition and docking functions in the growing
cell plate [27,37]. Mutations in KNOLLE cause abnormal
cytokinesis consistent with this role: the cell plate is often
incomplete, malformed or even absent in mutant cells [37].
In nonplant systems, interactions between target membrane and vesicle proteins are mediated by other soluble
proteins, including Vaccinia virus complement control protein (VCP) and its homolog in yeast, Cdc48p [40,41]. An
Arabidopsis homolog to Cdc48p (AtCDC48) is present in
dividing plant cells, localizes to the PP area and is able to
complement yeast cdc48 mutations [42].
A multi-step process of membrane targeting must be present in plant cytokinesis: target membranes must be first
established before the arriving vesicles can be docked.
Next, the cell plate membrane becomes a functional plasma membrane, which may occur after cytokinesis is
complete. It is interesting that a multi-step process is
described ultrastructurally [4], is demonstrated in inhibitor
studies [28,43] and is also postulated on the basis of the distribution of typical plasma membrane proteins such as
Division decisions and the spatial regulation of cytokinesis Sylvester
ATPase, which is absent in the forming cell plate [27]. It
will be interesting to find out precisely where and when
KNOLLE, for example, is localized in relation to other proteins potentially involved in targeting, such as AtCDC48.
61
Figure 4
Guiding the cell plate to the division site
The nature of the ‘correct site’ at the parental cell wall
To accomplish cytokinesis, the PP has to lead the forming
cell plate to the right place at about the right time. This
intriguing process involves interaction among dynamic
MTs of the PP, actin filaments and the cortical division site
near the parental cell. The nature of this site is still a mystery. Both MTs and actin filaments are present and are
required to mark the correct site prior to mitosis when the
transient PPB forms [2]. Disassembly of the PPB can be
accelerated by injection of CDC2, suggesting cell cycle progression may participate in marking the site [44]. Several
hypotheses could explain how the cortical division site is
marked. One hypothesis is that marker proteins are phosphorylated by CDC2 [45]. Alternatively, a localization of
ion pumps at the membrane site has been proposed [46].
Another intriguing idea is that the site could be specified by
the absence, rather than the presence, of a marker protein.
It is interesting that actin remains in the cortex throughout
division except at the division site, a position referred to as
the actin-depleted zone (Figure 1; [47]). This hypothesis
suggests that the absence of scaffolding in the division site
balances the PP against the extensive actin scaffolding elsewhere in the cell. Other than the actin-depleted zone, no
persistent structural or biochemical changes have been
detected in the cortical division site.
Selecting the division site
An explanation is still needed for how the site is chosen,
even if specific markers are eventually identified. The
division site could indeed be established earlier than just
prior to mitosis. For example, the site for any single new
cross wall could be influenced by the size and shape of the
parent cell and neighboring cells [48,49]. Change in cell
wall characteristics could be induced by local strains,
caused by differential expansion among interconnected
cells [48]. The new division site would then be selected in
response to the changed physical parameters of the tissue.
Another related view is that the division site is selected on
the basis of geometrically correct positions with the cell
[50]. Such physical explanations could signal a cell to
divide in a particular location, but the biochemical
responses elicited by the signal remain unknown.
Spatial signals of cytokinesis were studied recently in onion
root cells [51•]. When division is temporarily disrupted by
caffeine, the resulting binucleate cells will eventually adjust
and begin dividing again [51•]. Enlarged binucleate cells
generally produced a single PPB at a site predicted from the
size and shape of the cell prior to the treatment. These cells
would then continue to divide with altered spindle and PP
alignments, frequently not related to the most recent PPB
position. The results contradict other cases, however, in
Leaf surface impressions showing patterns of division and expansion at
the base of a developing primordium, in wild-type maize (a,c) and in a
wty1 mutant (b,d). (c) The base of the primordium depicted in Figure 3
was examined under the light microscope and shows the typical
squarish cells of a rapidly dividing epidermis. Recent cross walls are
apparent because they are light and less indented than older parental
walls. In this basal zone, cell divisions are similar in (d) the mutant
compared with (c) the non-mutant. (b) Further up the primordium, mutant
cells begin to over-expand (not shown) and subsequently incomplete
cytokinesis is observed in the enlarged cells (arrowhead) compared with
(a) non-mutant. Cells such as these are endoreduplicated but appear to
have normally arrayed microtubules (data not shown). Quantitative
analysis of the cytokinesis defects in relation to the enlarged cells
suggests that incomplete cytokinesis may result from loss of correct
spatial information in the growing cell. Leaf shape is unaffected, in part
due to the fact that cells adjoining the wart adjust their growth pattern
accordingly. The scale bar represents 60 µm.
which the PPB is absolutely required for correct wall placement, such as when a mutant lacks the PPB [52] or when
PPs adjust to PPB position after normal spindle rotations
[53]. In the experimentally induced binucleate cells, PP
position may be signaled by the nucleus or alternatively, by
older signals remaining from a prior division. These observations help define the normal parameters of site selection
but we clearly do not know all the molecules involved in
selecting the division site.
A genetic approach may help explain how the division site is
selected. The mutation warty 1 (wty1) is providing information about cell sizes that constrain normal cytokinesis [54•].
The wty1 mutants have cytokinesis defects restricted to
groups of cells, producing warts in the leaf blade (Figure 2).
Cell sizes can be measured in developing warts near the base
of the maize leaf primordium, because of the graduated balance between division and expansion (Figure 3). Epidermal
cells in a developing wart initially appear normal with respect
to size (Figure 3) and in distribution of cytoskeletal arrays
(AW Sylvester, unpublished data). Mutant cells that begin to
62
Growth and development
exceed a constant aspect ratio (length to width ratio) at a
given position in the primordium will eventually be unable
to complete cytokinesis (Figure 3). One explanation for
these observations is that over-enlargement prevents normal
cytokinesis, possibly by destroying spatial signals within the
cell. In these cells, expansion may be prolonged due to an
inability to enter mitosis, as wart cells also have endoreduplicated nuclei with condensed chromosomes (AW Sylvester,
unpublished data). Apparently, PP are able to form in warts
but appear to lack directional information. Ongoing analysis
of the gene and its product will help clarify the role of WTY1
in site selection.
Stabilizing and guiding the phragmoplast
The same proteins may be involved in selecting and
also guiding the PP to the division site. For example,
actin patches were observed in the actin-depleted zone
in dividing cells [47]. Actin filaments emanating from
the PP could attach to these patches left behind when
the PPB disappears [47,55]. The PP could thus be tethered to the division site by the same proteins that
originally established the site. Injected profilin, which
effectively binds monomeric actin, disrupts cytokinesis,
potentially by abolishing an actin-based mechanical
support to the forming cell plate [55]. Similarly, caffeine, which is thought to inhibit cytokinesis by
disrupting necessary Ca 2+ gradients, degrades actin filaments at the leading edge of the cell plate [43].
Consequently, the cell plate arrives at the wrong location, if at all, and is often incomplete. It is reasonable to
conclude that one role of actin is in the guidance of the
PP to the division site [55].
The PP and cell plate may also be stabilized internally by
actin. In Clivia endosperm cells, actin filaments are initially interspersed among spindle MTs but soon begin to
shorten and polymerize in the equatorial zone, eventually
interdigitating among the MTs of the PP (Figure 1; [56•]).
These shorter actin filaments are also associated with a
vinculin-like protein [56•]. Vinculin is an actin-associated
protein known to moderate actin–membrane connections
[57]. It is possible, therefore, that actin and associated proteins, such as vinculin, could be binding the vesicular
network into a stable structure, while also guiding it to the
division site.
Analysis of the discordia (dcd) mutations dcd1 and dcd2 may
lead us to the molecules that participate in PP guidance
[58••]. The dcd mutations cause abnormal growth of cells
that undergo asymmetric cell division in the leaf epidermis, such as the subsidiary cells in the stomatal complex.
Disruption of cell shape is preceded by improperly positioned cell walls, caused by misguided PPs. Actin
disruption by cytochalasin phenocopies the mutant defect
and the mutant phenotype is also exacerbated by cytochalasin treatment. These results suggest that DCD may
function in either residual marking of the division site
and/or in successful guidance of the PP to the site.
Even if actin proves to be a structural guide, other proteins may participate in the process. Clues are coming
from another maize gene, tangled1 (tan1) [59,60••]. Mutant
tan1 leaves have disorganized cells compared with the
well-oriented cells in the normal maize leaf. Cell disorganization is attributed partly to the inability of mutant cells
to divide in a particular longitudinal orientation (parallel
to the leaf axis). Mutant tan1 cells also lack the normal distrubution of longitudinally oriented PPB, demonstrating
the protein is directly or indirectly required for specifying
the longitudinal division plane. Furthermore, TAN1
appears to be necessary for proper PP guidance: all PPs in
the mutant are slightly disoriented compared with those
in normal cells, including those that are presumed to give
rise to transversely oriented cell walls. These results are
particularly intriguing because they point to the potential
for post-cytokinesis adjustments in wall position: cells that
have slightly oblique PPs may expand differentially and
very slightly, thereby adjusting the new cross wall to a
more typical transverse orientation. Preliminary reports
suggest the gene may encode a protein with hydrophobicity characteristics of a cell-wall protein [61]. Completed
analysis of the gene and its protein will be very interesting, as will an evaluation of the localization of TAN1
relative to PP and PPBs.
Impact of division ‘decisions’ on the plant body
Division orientations of individual cells vary in their
impact on the plant body. For example, morphogenesis of
some leaves does not require carefully orchestrated cell
divisions [1,59]. On the other hand, some roots have
stereotyped division patterns that result in proper disposition of cell layers, required for the acquisition of cell
identities [62]. The impact of cytokinesis defects depends
on the organ, the developmental stage and the nature of
the gene involved. Currently, the phenotypes of mutants
that affect cytokinesis reflect this variability. Some of the
genes identified are probably redundant, others are directly involved in the mechanics of cytokinesis and others may
play a more indirect role in division and expansion. The
remainder of this review will evaluate these mutants based
on the severity of their phenotype.
Fatal divisions
Many mutants with defects in cytokinesis were identified
in genetic screens of seedling-lethal embryo-defective
mutants [63]. The cyt1 mutant of Arabidopsis has defective
cross walls and excessive accumulation of callose, evident
early in embryogenesis [64••]. Interfering with cellulose
synthesis can phenocopy the mutant, suggesting the gene
plays a role in wall production during cytokinesis [64••].
The well-characterized knolle mutation affects body organization of the Arabidopsis embryo and specifically alters
the forming cell plate. As discussed earlier, KNOLLE
must be one of the key players in successful cytokinesis,
probably through a syntaxin-like role in vesicle targeting.
The keule mutant of Arabidopsis also lacks normal cytokinesis [65] as does the cyd mutant of pea [66]. In these cases,
Division decisions and the spatial regulation of cytokinesis Sylvester
cross walls are initiated but not completed, mimicking the
effect of caffeine on cytokinesis.
The gnom/emb30 mutation appears to affect all aspects of
cell division and expansion. Most gnom mutant cells are
able to complete cytokinesis normally, but are disoriented
from the first zygotic division [67]. The EMB30/GNOM
protein shares a domain with a secretory protein in yeast
called Sec7p [68]. Other regions of the protein are similar
to a non-essential yeast protein Yec2p [67]. Although the
precise role of the protein is not known, the pleiotropic
effects of the mutation may be due to generalized defects
in secretion or vesicle function, rather than just secretion
during cell plate formation.
63
organ specificity of mutant defects suggests either redundant functions of the genes in other organs or that the
specific controls of division may vary in different organs.
Spatially adjustable divisions
Differentiation of specific cell types sometimes requires a
highly regulated pattern of division [1,69]. Mistakes frequently cannot be tolerated in these types of
differentiation divisions. One example is in the required
asymmetric division that yields polarized cells. Pollen
grains will differentiate appropriately only if the first
mitotic cross wall is deposited asymmetrically. Incomplete
cytokinesis, as seen in gem1 mutants, disrupts differentiation of the vegetative and generative cells [70•]. Similarly,
asymmetric cell division precedes differentiation of the
layered Arabidopsis root, which is prevented in scr mutants
[71]. In this case, the division alteration is more likely due
to a response to a change in positional information rather
than a direct effect on cytokinesis. Even so, mutations
such as scr may provide information about how these signals are read by cells, an important aspect of the spatial
regulation of division [69].
Several mutants have problems with cytokinesis but show
quite normal development. Loss of spatial information by
altered cytokinesis may be corrected in a growing organ by
adjusting the direction or rate of cell expansion or increasing the rate of cell division. For example, tan1 mutants
have normally shaped leaves and organs, despite incorrectly placed cross walls. Cells in tan1 mutants that are
experimentally forced to expand more, by growing in dark
or low-light conditions, show corrected orientations of cell
walls (M Mitkovski, AW Sylvester, unpublished data). Cell
adjustments are also seen in wty1 mutants: improper
growth of warts is balanced by neighboring cells, which
divide more as the leaf is growing [54•]. This neighborly
behavior of cells is consistent with the idea that the emerging organ shape prevails as a signal for the spatial
regulation of cell division during morphogenesis of leaves.
The idea requires that cells respond together and communicate locally, via regulated plasmodesmatal trafficking
[75] or other means of intercellular communication. MTs
and cellulose microfibrils, which together direct oriented
cell expansion, are observed to be co-aligned in cell groups
[49,60••,76], consistent with other observations that neighboring cells grow in coordinated patterns [49]. Mutants,
such as tan1 and wty1, provide information about the
capacity of cells to adjust within a given framework. Such
adjustments suggest that the selection of the division site
itself may be flexibly regulated.
Near fatal divisions
Conclusions
Drastic alterations in cell division may have only partial
effects on development. For example, the fass mutants in
Arabidopsis have defects in cytokinesis that result in abnormally shaped cells [72]. Organ morphogenesis is also
severely altered, with plants acquiring a short stubby
appearance, but tissues are distributed normally. These
mutants also lack the normal PPB, suggesting the gene
may be involved in selecting and/or marking the division
site [52]. The ton/fass mutants, among others, show that
tight spatial regulation of cytokinesis is not necessary for
basic tissue differentiation.
The steps of cytokinesis are integrated so that a growing
cell plate is guided to a pre-established division site.
Several proteins are known that probably contribute to
the membranous construction of the cell plate, including
phragmoplastin and KNOLLE. Potential phosphorylation requirements in the cell plate will be understood as
more kinases are identified. Also, other proteins that
might be involved specifically in vesicle targeting, such
as CDC48, will provide information about the interactions between soluble and membrane-bound proteins.
There are several phases in the construction and growth
of the cell plate. In the future, it will be interesting to
understand how these vesicle proteins may behave
uniquely in the cell plate.
Division effects may vary based on whether a gene is
required throughout development or in all cells. The tso1
mutants, for example, show cytokinesis defects only in floral organs and ovules [73,74]. In floral meristems and sepals,
cross walls are incomplete and nuclei endoreduplicated,
similar to the cytokinetic defects in plants with knolle, keule,
cyd and wty1 mutations. Developmental arrest of ovules in
tso1 mutants appears to be due to abnormal and uncoordinated cell expansion, a good example of the same gene
being required for division or expansion, depending on the
organ [73]. Similarly, wty1 mutants show altered cytokinesis,
but only in leaf blade cells [54•]. If these are all null alleles,
Two major cytoskeletal components are required for successful cytokinesis — MTs and actin. The relationship of
the arrays to the forming cell plate has been clarified experimentally and by cytological observations. Kinesin-related
proteins that participate in maintaining the structure and
function of the PP are also being identified. Motorized
transport along MTs is still the best hypothesis for how
vesicles get to the cell plate but, as yet, proteins that mediate this process have not been definitively identified. Actin
64
Growth and development
is also likely to be responsible for unifying the cell plate and
for providing internal as well as lateral support for the PP.
Actin may well serve as a scaffolding upon which the PP is
poised as it extends radially to the division site. The mechanism of guiding the PP to its division site is just beginning
to unfold, as new mutants are found that influence this
process. Mutations such dcd1 and dcd2 are likely to provide
key information about PP guidance.
One outcome of cytokinesis is that two daughter cells are
initially positioned relative to neighbors and to the original parent cell. On the basis of mutant phenotypes, the
impact of this division ‘decision’ on subsequent development depends on the developmental stage, the type of
protein required, or on the extent of genetic redundancy.
Some mutants, such as tan1 and wty1 can adjust for cytokinesis defects, presumably by expanding or dividing
differently to compensate for the loss of spatially correct
information. In other cases, a correct cell division orientation may be required prior to a critical differentiation step,
so that a spatially defective cytokinesis cannot be tolerated, as in the gem1 mutant. Finally, seedling-lethal
mutations, such as cyt1, knolle, keule, cyd may cause the loss
of some essential function in cytokinesis itself. While the
mechanics of cytokinesis is probably shared for different
cells in different organs, it is possible that the means of
selecting the division site may prove to be variable. In the
long term, the question of how a division site is selected
should be considered in the context of the entire growing
tissue. The interconnections of plant cells suggest that
any single division ‘decision’ influences neighboring cells
biochemically and ultimately physically. Continued characterization of mutants, and molecular analysis of the
genes involved, will help to clarify both the mechanics
and the determinants of cytokinesis.
Acknowledgements
I wish to thank Jim Reynolds for contributions at all levels to this work. I
thank Miso Mitkovski for micrographs and the US Department of
Agriculture and the National Science Foundation—Experimental
Program for Stimulating Competitive Research for financial support. This
review is dedicated to the memory of Professor Paul Green, a source of
inspiration and insight.
References and recommended reading
Papers of particular interest, published within the annual period of review,
have been highlighted as:
• of special interest
•• of outstanding interest
6.
Mironov V, De Veylder L, Van Montagu M, Inze D: Cyclin-dependent
kinases and cell division in plants-the nexus. Plant Cell 1999,
11:509-522.
7.
Franklin AE, Cande WZ: Nuclear organization and chromosome
segregation. Plant Cell 1999, 11:523-534.
8.
Wymer C, Lloyd C: Dynamic microtubules: implications for cell
wall patterns. Trends Plant Sci 1996, 1:222-228.
9.
Cosgrove DJ: New genes and new biological roles for expansins.
Curr Opin Plant Biol 2000, 3:73-78.
10. Yasuhara H, Sonobe S, Shibaoka H: Effects of brefeldin A on the
formation of the cell plate in tobacco BY- 2 cells. Eur J Cell Biol
1995, 66:274-281.
11. Hardham AR, Gunning BES: Some effects of colchicine on
microtubules and cell division in roots of Azolla pinnata.
Protoplasma 1980, 102:31-51.
12. Thyberg J, Moskalewski S: Role of microtubules in the organization
of the Golgi complex. Exp Cell Res 1999, 246:263-279.
13. Yasuhara H, Sonobe S, Shibaoka H: Effects of taxol on the
development of the cell plate and of the phragmoplast in tobacco
BY-2 cells. Plant Cell Physiol 1993, 34:21-29.
14. Asada T, Collings D: Molecular motors in higher plants. Trends
Plant Science 1997, 2:29-37.
15. Asada T, Kuriyama R, Shibaoka H: TKRP125, a kinesin-related
protein involved in the centrosome-independent organization of
the cytokinetic apparatus in tobacco BY-2 cells. J Cell Sci 1997,
110:179-189.
16. Song H, Golovkin M, Reddy AS, Endow SA: In vitro motility of
AtKCBP, a calmodulin-binding kinesin protein of Arabidopsis.
Proc Natl Acad Sci USA 1997, 94:322-327.
17.
•
Narasimhulu SB, Reddy AS: Characterization of microtubule
binding domains in the Arabidopsis kinesin-like calmodulin
binding protein. Plant Cell 1998, 10:957-965.
The motor domain of the kinesin-related calmodulin binding protein, KCBP,
is shown to bind microtubules in cosedimentation assays with the purified
protein. Two MT-binding domains are identified in KCBP, one near the carboxyl terminus and one near the amino terminus. The protein could thus be
involved in bundling MTs. Evidence for the function of this protein in organizing MTs is thus growing.
18. Smirnova EA, Reddy AS, Bowser J, Bajer AS: Minus end-directed
•
kinesin-like motor protein, Kcbp, localizes to anaphase spindle
poles in Haemanthus endosperm. Cell Motil Cytoskeleton 1998,
41:271-280.
KCBP distribution is observed in endosperm cells during mitosis and cytokinesis. Interesting shifts in localization patterns are seen during a transition
through telophase. KCBP shifts from an association with the spindle poles
at the end of mitosis to the equatorial zone. KCBP is eventually localized
along the phragmoplast MTs, but also to the nuclear membrane early in
telophase and to the cell plate in late telophase. Slight differences in KCBP
distribution with BY-2 cells may be due to unique properties of the cellularizing endosperm.
19. Liu B, Cyr RL, Palevitz BA: A kinesin-like protein, KatAp, in the cells
of Arabidopsis and other plants. Plant Cell 1996, 8:119-132.
20. Compton DA: Focusing on spindle poles. J Cell Sci 1998,
111:1477-1481.
21. Kakimoto T, Shibaoka H: Cytoskeletal ultrastructure of
phragmoplast-nuclei complexes isolated from cultured tobacco
cells. Protoplasma 1988, 2:95-103.
22. Parke J, Miller C, Anderton BH: Higher plant myosin heavy-chain
identified using a monoclonal antibody. Eur J Cell Biol 1986, 41:9-13.
1.
Sylvester AW, Smith L, Freeling M: Acquisition of identity in the
developing leaf. Annu Rev Cell Dev Biol 1996, 12:257-304.
2.
Assaad FF, Mayer U, Lukowitz W, Juergens G: Cytokinesis in
somatic plant cells. Plant Physiol Biochem 1997, 35:177-184.
3.
Fowler JE, Quatrano RS. Plant cell morphogenesis: plasma
membrane interactions with the cytoskeleton and cell wall. Annu
Rev Cell Dev Biol 1997, 13:697-743.
4.
Samuels AL, Giddings TH, Staehelin LA: Cytokinesis in tobacco BY2 and root tip cells: a new model of cell plate formation in higher
plants. J Cell Biol 1995, 130:1345-1357.
25. Park JM, Kang SG, Pih KT, Jang HJ, Piao HL, Yoon HW, Cho MJ,
Hwang I: A dynamin-like protein, ADL1, is present in membranes
as a high-molecular-mass complex in Arabidopsis thaliana. Plant
Physiol 1997, 115:763-771.
5.
Staehelin LA, Hepler PK: Cytokinesis in higher plants. Cell 1996,
84:821-824.
26. Kozlov MM: Dynamin: possible mechanism of ‘Pinchase’ action.
Biophys J 1999, 77:604-616.
23. Asada T, Shibaoka H: Isolation of polypeptides with microtubuletranslocating activity from phragmoplasts of tobacco BY-2 cells.
J Cell Sci 1994, 107:2249-2257.
24. Gu XJ, Verma DPS: Phragmoplastin, a dynamin-like protein
associated with cell plate formation in plants. EMBO J 1996,
15:695-704.
Division decisions and the spatial regulation of cytokinesis Sylvester
27.
Lauber MH, Waizenegger I, Steinmann T, Schwarz H, Mayer U, Hwang
I, Lukowitz W, Jurgens G: The Arabidopsis KNOLLE protein is a
cytokinesis-specific syntaxin. J Cell Biol 1997, 139:1485-1493.
28. Gu X, Verma DP: Dynamics of phragmoplastin in living cells during
cell plate formation and uncoupling of cell elongation from the
plane of cell division. Plant Cell 1997, 9:157-169.
29. Slepnev VI, Ochoa GC, Butler MH, Grabs D, Camilli PD: Role of
phosphorylation in regulation of the assembly of endocytic coat
complexes. Science 1998, 281:821-824.
30. Sharp DJ, McDonald KL, Brown HM, Matthies HJ, Walczak C, Vale RD,
Mitchison TJ, Scholey JM: The bipolar kinesin, KLP61F, cross-links
microtubules within interpolar microtubule bundles of Drosophila
embryonic mitotic spindles. J Cell Biol 1999, 144:125-138.
31. Hirt H: Multiple roles of MAP kinases in plant signal transduction.
Trends Plant Sci 1997, 2:11-15.
32. Reszka AA, Seger R, Diltz CD, Krebs EG, Fischer EH: Association of
mitogen-activated protein kinase with the microtubule
cytoskeleton. Proc Natl Acad Sci USA 1995, 92:8881-8885.
33. Calderini O, Bogre L, Vicente O, Binarova P, Heberle-Bors E,
•
Wilson C: A cell cycle regulated MAP kinase with a possible role
in cytokinesis in tobacco cells. J Cell Sci 1998, 111:3091-3100.
This MAPK from tobacco is activated during late anaphase and the protein
localizes to the cell plate. Disruption of MTs with inhibitors prevents activation of the kinase.
34. Bogre L, Calderini O, Binarova P, Mattauch M, Till S, Kiegerl S,
•
Jonak C, Pollaschek C, Barker P, Huskisson NS et al.: A MAP kinase
is activated late in plant mitosis and becomes localized to the
plane of cell division. Plant Cell 1999, 11:101-114.
A MAPK is characterized from alfalfa. The protein appears to be distributed
in the region of the cell plate and remains there even as the phragmoplast
extends radially.
35. Banno H, Hirano K, Nakamura T, Irie K, Nomoto S, Matsumoto K,
Machida Y: NPK1, a tobacco gene that encodes a protein with a
domain homologous to yeast BCK1, STE11, and Byr2 protein
kinases. Mol Cell Biol 1993, 13:4745-4752.
36. Nakashima M, Hirano K, Nakashima S, Banno H, Nishihama R,
•
Machida Y: The expression pattern of the gene for NPK1 protein
kinase related to mitogen-activated protein kinase kinase kinase
(MAPKKK) in a tobacco plant: correlation with cell proliferation.
Plant Cell Physiol 1998, 39:690-700.
An upstream MAPK (identified as a MAPK kinase kinase) was found in
tobacco and is expressed in dividing cells. Subcellular localization is still
needed to clarify whether or not this protein is part of a phosphorylation cascade in the cell plate or phragmoplast.
37.
Lukowitz W, Mayer U, Jurgens G: Cytokinesis in the Arabidopsis
embryo involves the syntaxin-related KNOLLE gene product. Cell
1996, 84:61-71.
38. Robinson DG, Hinz G, Holstein SE: The molecular characterization
of transport vesicles. Plant Mol Biol 1998, 38:49-76.
39. Jantsch-Plunger V, Glotzer M: Depletion of syntaxins in the early
Caenorhabditis elegans embryo reveals a role for membrane
fusion events in cytokinesis. Curr Biol 1999, 9:738-745.
40. Mellman I: Enigma variations: protein mediators of membrane
fusion. Cell 1995, 82:869-872.
41. Frohlich KU, Fries HW, Rudiger M, Erdmann R, Botstein D, Mecke D:
Yeast cell cycle protein CDC48p shows full-length homology to
the mammalian protein VCP and is a member of a protein family
involved in secretion, peroxisome formation, and gene
expression. J Cell Biol 1991, 114:443-453.
42. Feiler HS, Desprez T, Santoni V, Kronenberger J, Caboche M, Traas J:
The higher plant Arabidopsis thaliana encodes a functional
CDC48 homologue which is highly expressed in dividing and
expanding cells. EMBO J 1995, 14:5626-5637.
43. Valster AH, Hepler PK: Caffeine inhibition of cytokinesis: effect on
the phragmoplast cytoskeleton in living Tradescantia stamen hair
cells. Protoplasma 1997, 196:155-166.
44. Hush J, Wu L, John PC, Hepler LH, Hepler PK: Plant mitosis
promoting factor disassembles the microtubule preprophase
band and accelerates prophase progression in Tradescantia. Cell
Biol Int Rep 1996, 20:275-287.
45. Cleary AL: F-actin redistributions at the division site in living
Tradescantia stomatal complexes as revealed by microinjection of
rhodamine–phalloidin. Protoplasma 1995, 185:152-165.
65
46. Cyr R: Calcium/calmodulin affects microtubule stability in lysed
protoplasts. J Cell Sci 1991, 100:311-317.
47.
Cleary AL, Gunning BES, Wasteneys GO, Hepler PK: Microtubule
and F-actin dynamics at the division site in living Tradescantia
stamen hair cells. J Cell Sci 1992, 103:977-988.
48. Green PB: Transductions to generate plant form and pattern: an
essay on cause and effect. Ann Bot 1996, 78:269-281.
49. Sylvester AW, Williams MH, Green PB: Orientation of cortical
microtubules correlates with cell shape and division direction:
immunofluorescence of intact epidermis during development of
Graptopetalum paraguayensis. Protoplasma 1989, 153:91-103.
50. Flanders DJ, Rawlins DJ, Shaw PJ, Lloyd CW: Nucleus-associated
microtubules help determine the division plane of plant epidermal
cells: avoidance of four way junctions and the role of cell
geometry. J Cell Biol 1990, 110:1111-1122.
51. Gimenez-Abian MI, Utrilla L, Canovas JL, Gimenez-Martin G,
•
Navarrete MH, De la Torre C: The positional control of mitosis and
cytokinesis in higher-plant cells. Planta 1998, 204:37-43.
Binucleate wheat cells are produced by caffeine treatment and subsequent
patterns of phragmoplast arrangement are recorded in cells of different
sizes. The authors conclude that PP could form without respect to the PPB
in some cases, suggesting pre-existing spatial cues may dictate cell wall
position under certain circumstances.
52. Traas J, Bellini C, Nacry P, Kronenberger J, Bouchez D, Caboche M:
Normal differentiation patterns in plants lacking microtubular
preprophase bands. Nature 1995, 375:676-677.
53. Palevitz BA: Division plane determination in guard mother cells of
Allium: video time-lapse analysis of nuclear movements and
phragmoplast rotation in the cortex. Dev Biol 1986, 117:644-654.
54. Reynolds JO, Eisses JF, Sylvester AW: Balancing division and
•
expansion during maize leaf morphogenesis: analysis of the
mutant, warty-1. Development 1998, 125:259-268.
Cells with wty1 mutations undergo faulty cytokinesis once the cells have
begun to over-expand. The results suggest that one spatial cue for the division site may be appropriate length to width or volume ratio in the cell. In
addition, neighboring cells in the wty1 mutant, which can look normal,
appear to adjust to the defect by dividing more.
55. Valtzer AH, Pierson ES, Valenta R, Hepler PK, Emons AMC: Probing
the plant actin cytoskeleton during cytokinesis and interphase by
profilin microinjection. Plant Cell 1997, 9:1815-1824.
56. Endle MC, Stoppin V, Lambert AM, Schmit AC: The growing cell
•
plate of higher plants is a site of both actin assembly and
vinculin-like antigen recruitment. Eur J Cell Biol 1998, 77:10-18.
The distribution of actin and an associated vinculin-like protein is described
in endosperm cells of Clivia. Short actin filaments are seen in the center of
the PP and co-localize with vinculin. The authors suggest the proteins are
important in maintaining and perhaps guiding the PP during cytokinesis.
57.
Johnson RP, Craig SW: F-actin binding site masked by the
intramolecular association of vinculin head and tail domains.
Nature 1995, 373:261-264.
58. Gallagher K, Smith LG: discordia mutations specifically misorient
•• asymmetric cell divisions during development of the maize leaf
epidermis. Development 1999, 126:4623-4633.
These mutants show an interesting and informative phenotype. Asymmetric
divisions that occur during differentiation of the epidermis are affected so
that cytokinesis is disrupted and cells expand in abnormal orientations.
Normally positioned PPB are present, predicting the correct site for the
impending new wall. PP lose direction relative to the site, however, suggesting that they are not guided properly or that residual information marking the site is lost. Particularly intriguing is the fact that actin filament
disruption can phenocopy the mutants. The results implicate the genes in a
PP guidance process that may require actin or associated proteins.
59. Smith LG, Hake S, Sylvester AW: The tangled-1 mutation alters cell
division orientations throughout maize leaf development without
altering leaf shape. Development 1996, 122:481-489.
60. Cleary AL, Smith LG: The Tangled1 gene is required for spatial
•• control of cytoskeletal arrays associated with cell division during
maize leaf development. Plant Cell 1998, 10:1875-1888.
Stunning views of the distribution of MTs and actin filaments in tan1 and nonmutant leaf primordia are presented. Detailed analysis of cytoskeletal arrays
in mutants compared with non-mutants confirm that longitudinal divisions are
specifically disrupted in the mutant. PPBs do not occupy the expected longitudinally oriented position. PP formations are also slightly disoriented compared with normal.
66
Growth and development
61. Miller D, Hable W, Gottwald J, Ellard-Ivey M, Demura T, Lomax T,
Carpita N: Connections: the hard wiring of the plant cell for
perception, signaling, and response. Plant Cell 1997, 9:2105-2117.
69. Scheres B, Heidstra R: Digging out roots: pattern formation, cell
division and morphogenesis in plants. Curr Top Dev Biol 1999,
45:207-247.
62. Berger F, Hung CY, Dolan L, Schiefelbein J: Control of cell division in
the root epidermis of Arabidopsis thaliana. Dev Biol 1998,
194:235-245.
70. Park SK, Howden R, Twell D: The Arabidopsis thaliana
•
gametophytic mutation gemini pollen1 disrupts micros
Division decisions and the spatial regulation of cytokinesis
Anne W Sylvester
Cytokinesis in plant cells is accomplished when a membranous
cell plate is guided to a pre-established division site. The
orientation of the new wall establishes the starting position of a
cell in a growing tissue, but the impact of this position on
future development varies. Recently, proteins have been
identified that participate in forming, stabilizing and guiding the
cell plate to the correct division site. Mutations that affect
cytokinesis with varying impacts on plant development are
providing information about the mechanics of cytokinesis and
also about how the division site is selected.
Abbreviations
dcd
discordia
KCBP
kinesin-like calmodulin binding protein
KatAp
kinesin from Arabidopsis thaliana A peptide
MAPK
mitogen-activated protein kinase
MMK3
Medicago mitogen-activated protein kinase 3
MT
microtubule
NPK1
Nicotiana protein kinase 1
PP
phragmoplast
PPB
preprophase band
TKRP125 tobacco kinesin-related polypeptide of 125 kDa
wty1
warty1
equatorial zone between recently formed daughter nuclei
[2–5]. The growing cell plate then spreads out radially, as
an expanding interlacing membranous network, until it
joins with the parental walls at the periphery. A
cytoskeletal structure called a phragmoplast (PP) not only
guides vesicles to the equatorial zone but also directs the
growing cell plate toward a specific site at or near the
membrane of the parental cell wall — the cortical division
site. This ill-defined site is first evident prior to mitosis
when the preprophase band (PPB) appears transiently as
a continuous band of cytoskeletal elements in the cell
cortex. Remarkably, the position of the PPB forecasts the
future site of attachment between the cell plate and the
parental cell walls. The integrated behavior of these
cytoskeletal structures thus accomplishes the feat of
depositing a new wall at a specific position. What is the
mechanism of cell plate formation? What determines
where this new wall will form? How does cell position
affect future development? This review presents recent
progress in answering these questions and also evaluates
potential versatility in the process, considering that some
divisions have more impact on future development than
others. Recent advances in understanding cell cycle regulation [6], mechanisms of mitosis [7] or cell expansion
[8,9 (DJ Cosgrove, pp 73–78)] are relevant topics, but are
not covered here. Excellent diagrams showing the
detailed structure of the growing cell plate and PP are
available elsewhere [2,4,5].
Introduction
Forming the cell plate
Plant cell division serves a dual function of increasing cell
number and also positioning new cross walls during cytokinesis. Once the new wall is formed, cells acquire a final
shape by differential cell expansion. Several important
decisions are thus made during division and expansion,
including when and how often division should occur,
where a new cell wall should be located, and in what direction to expand. The outcome of any one of these decisions
could be critical to the successful development of an organ
because new daughter cells are joined from inception by
cell walls. In reality, however, the position of new walls has
a variable influence on the shape of an organ or function of
a tissue. This is because the spatial regulation of cytokinesis, which determines where new walls will be positioned,
affects development to different degrees [1]. Regardless of
the impact of division pattern on development, successful
cytokinesis is needed to maintain the integrity of the plant
body. Understanding how and when cytokinesis occurs will
help to evaluate the effect of different stages of development on its spatial regulation.
Completion of cytokinesis requires the interaction of fusing secretory vesicles, dynamic microtubules (MTs), actin
and numerous associated proteins (Figure 1), as confirmed
by electron and immunofluorescence microscopy and by
experimental manipulation of cytokinesis. Disruption of
the Golgi complex with the inhibitor brefeldin A, for
example, prevents completion of the cell plate, suggesting
that new vesicles are Golgi-derived [10]. Vesicles will also
not accumulate in the equatorial zone of cells treated with
inhibitors that depolymerize MTs [11], as a result of either
transport inhibition or altered organization of the Golgi
complex [12]. On the other hand, taxol, which hyperstabilizes MTs by preventing their depolymerization, increases
the number of vesicles in the equatorial zone initially,
although cytokinesis is eventually arrested [13]. These
results demonstrate that MTs must be present but not necessarily dynamic for successful vesicle transport,
supporting the idea that vesicles migrate along MTs with
the aid of motor proteins [14].
Addresses
Department of Botany, PO Box 3165, University of Wyoming, Laramie,
Wyoming 82071-3165, USA; e-mail: annesyl@uwyo.edu
Current Opinion in Plant Biology 2000, 3:58–66
1369-5266/00/$ — see front matter © 2000 Elsevier Science Ltd.
All rights reserved.
Vesicle transport
Cytokinesis is accomplished by progressive deposition of
membranes and associated wall synthesizing compounds
into a cell plate, which is first located centrally in the
The exact mechanism of vesicle transport is still an open
question. The bipolar MTs in the PP are arranged with
their plus ends overlapping in the equatorial zone and
Division decisions and the spatial regulation of cytokinesis Sylvester
59
Figure 1
Diagrammatic view of a plant cell in early
cytokinesis. Proteins recently identified in
each compartment of the dividing cell are
shown. Kinesin-related proteins (TKRP125,
KCBP, KatAp) show distribution patterns that
suggest an involvement in organizing the MT
array, rather than transporting vesicles. KCBP
localizes to the nuclear membrane of
endosperm cells during late telophase only.
Several other proteins probably involved in
formation of the tubulo-vesicular membrane
network of the cell plate (barely visible as
wavy lines in the equatorial zone) include
KNOLLE and phragmoplastin. CDC48 may
function with KNOLLE in docking of vesicles
to their target but localization details are not
sufficient to suggest this. Kinases (such as
p43Ntf6) locate to the same region, but the
phosphorylation substrates are not known.
Vinculin-like antigens have recently been
observed in the forming cell plate of
cellularizing endosperm, possibly serving in
association with the actin filaments found
there. Actin appears to link the phragmoplast
to the cortex. The actin-depleted zone might
carry residual actin patches but this
connection is not yet clear.
Actin
Daughter nuclear
membrane
*KCBP
Phragmoplast
MTs
*TKRP125
*KCBP
*KatAp
Actin-depleted
zone
Cell Plate in
equatorial zone
*KCBP
*Phragmoplastin
(& ADL1)
*KNOLLE
*MAPKs (p43 Ntf6 )
*Vinculin antigen
minus ends directed toward the daughter nuclei
(Figure 1), suggesting vesicles might be transported
using a plus-end directed motor protein [5]. A candidate
kinesin-related protein isolated from tobacco PPs
(TKRP125) has the expected domains for motor activity
[15]. The protein localizes, however, along the entire
length of MTs in the PP, rather than in the punctate patterns expected for vesicle association. TKRP125 could
still interact with other as yet unidentified proteins associated with vesicles. In permeablized cells, however,
antibodies against TKRP125 actually inhibit plus-end
MT translocation [15]. Together, these results suggest
that TKRP125 might contribute to maintaining the structure of the MTs in the PP, rather than transporting
vesicles to the cell plate.
Minus-end directed kinesin-like proteins have been found
associated with the PP, including kinesin-like calmodulin
binding protein (KCBP) and a kinesin from Arabidopsis
thaliana A peptide (KatAp). KCBP mediates MT translocation in vitro [16] and its motor domain binds MTs [17•].
KCBP localizes to the MTs in the PP and also to the
nuclear membrane early in telophase and within the equatorial zone [18•]. KatAp is initially present in the spindle
mid-zone before the PP is organized and then in association with the PP itself [19]. The localization patterns for
both proteins are consistent with a function in organizing
the MTs in the PP, rather than in directing vesicle transport
[18•,19], particularly as similar kinesins in other systems
help to assemble and organize the mitotic spindle [20].
The translocation of MTs toward the minus end could propel vesicles toward the plus ends. Most evidence, however,
Vesicles
Current Opinion in Plant Biology
points to an organizational function for these proteins, perhaps mediated by these minus-end directed motors.
F-actin could be a candidate for vesicle motility because it
co-localizes with MTs in the PP, and is oriented as expected
for a motility system [21]. Myosin may also be present in the
PP [22]. Although disruption of actin filaments with cytochalasin does not affect vesicle accumulation, it does disorient
the cell plate [11]. F-actin could facilitate vesicle transport
through stabilizing the MTs, as discussed below but whether
it plays a direct role in vesicle transport is not clear.
Organizing the phragmoplast
MT translocation might help to establish and then maintain the bipolar MT array of the PP [23], possibly mediated
by the kinesin-related proteins discussed above.
Differential MT depolymerization is also important in PP
function [10]. Centrifugal growth of the cell plate is prevented when MTs are stabilized, mirroring an effect of
preventing depolymerization on spindle function [7].
These studies point to a required balance between polymerization at the leading edge and depolymerization at the
trailing edge of the PP. It is likely that similar organizational mechanisms are found for the PP as found for other
MT structures, such as the mitotic spindle.
Vesicle accretion and fusion in the cell plate
Newly arriving vesicles coalesce into a membranous network
as the cell plate extends out radially [4]. Clathrin-coated vesicles are observed in conjunction with the growing
tubulo-vesicular network of the cell plate, providing evidence for fusion and membrane recycling as part of cell plate
60
Growth and development
Figure 2
Figure 3
Recent
cross walls
Cell
size
Cell
differentiation
Pre-differentiation
zone
Current Opinion in Plant Biology
Epidermal cells in a fully-grown maize leaf of (a) a wild-type plant
compared with (b) a wty1 mutant. (a) The cells in a normal leaf are
linear and aligned with the long axis of the leaf blade, as emphasized in
the cell outlines in the right of the panel. (b) The mutant leaf has overly
enlarged cells in the wart, which result from early events of overexpansion and improper cytokinesis (see also Figure 4).
formation [4]. A protein that probably facilitates this process
is phragmoplastin, a dynamin-related protein originally isolated from soybean [24] and also from Arabidopsis (called
Arabidopsis dynamin-like protein 1 [ADL1]) [25]. In animal
systems, dynamin is a GTPase that functions in endocytosis,
particularly in membrane separation [26]. Both phragmoplastin and ADL1 localize to the growing cell plate [24,27].
Phragmoplastin is not dependent on MT organization, as it
remains at the cell plate margins after the PP has dissociated
near the end of cytokinesis. Taxol-stabilization of MTs does
not eliminate phragmoplastin from its earliest position in the
equatorial zone but does prevent its redistribution to the
leading edge [24]. In vivo observations using green-fluorescent protein (GFP)-fusion proteins show that
phragmoplastin associates with sites of high membrane activity [28], consistent with a dynamin-like role for the protein.
Many essential protein functions are regulated by phosphorylation, including dynamin assembly and kinesin
activity [29,30]. Recently, mitogen-activated protein kinases (MAPKs) were found in the cell plate. MAPKs are
cell-cycle regulated proteins involved in organizing the
cytoskeleton and are also abundant in diverse signaling
pathways [31,32]. Two proteins similar to MAPKs, (p43Ntf6
from tobacco and Medicago mitogen-activated protein
kinase 3 [MMK3] from alfalfa) are activated during late
anaphase if functional MTs are present [33•,34•]. MTs are
not required to maintain distribution of MMK3 in the cell
plate once the PP is established [34•]. Another protein,
Nicotiana protein kinase 1 (NPK1) is similar to upstream
kinases in a MAPK chain, referred to as MAPK kinase
kinases [35]. NPK1 is expressed in dividing cells [36•] and
Simplified view of the pattern of division and expansion in the
developing maize leaf primordium. In general, cells in the maize leaf
adhere to an inverse pattern of dividing and expanding, particularly
below the zone where epidermal cells begin to differentiate (predifferentiation zone). Here the gradual increase in cell size is due to the
rate of expansion exceeding that of division. Orientation of new crosswalls can be observed and quantified in leaf surface impressions made
from this region (see Figure 4). Figure reprinted with permission [54•].
complements yeast mutations in homologs to NPK1 [35],
but its localization patterns relative to the cell plate have
not yet been reported. Pending these results, NPK1 in
conjunction with p43Ntf6 and MMK3 could be candidate
transducers in a phosphorylation cascade in the cell plate.
Proteins that participate in vesicle targeting are also found
in the cell plate. One example is the syntaxin-related protein KNOLLE [27,37]. Syntaxins (also called t-SNARES)
are located in target membranes and specifically interact
with proteins in arriving vesicles, called v-SNARES [2,38].
Syntaxins in nonplant systems participate in vesicle fusion
during cytokinesis [39]. Localization of KNOLLE to vesicles and to the equatorial zone suggests it might serve
similar recognition and docking functions in the growing
cell plate [27,37]. Mutations in KNOLLE cause abnormal
cytokinesis consistent with this role: the cell plate is often
incomplete, malformed or even absent in mutant cells [37].
In nonplant systems, interactions between target membrane and vesicle proteins are mediated by other soluble
proteins, including Vaccinia virus complement control protein (VCP) and its homolog in yeast, Cdc48p [40,41]. An
Arabidopsis homolog to Cdc48p (AtCDC48) is present in
dividing plant cells, localizes to the PP area and is able to
complement yeast cdc48 mutations [42].
A multi-step process of membrane targeting must be present in plant cytokinesis: target membranes must be first
established before the arriving vesicles can be docked.
Next, the cell plate membrane becomes a functional plasma membrane, which may occur after cytokinesis is
complete. It is interesting that a multi-step process is
described ultrastructurally [4], is demonstrated in inhibitor
studies [28,43] and is also postulated on the basis of the distribution of typical plasma membrane proteins such as
Division decisions and the spatial regulation of cytokinesis Sylvester
ATPase, which is absent in the forming cell plate [27]. It
will be interesting to find out precisely where and when
KNOLLE, for example, is localized in relation to other proteins potentially involved in targeting, such as AtCDC48.
61
Figure 4
Guiding the cell plate to the division site
The nature of the ‘correct site’ at the parental cell wall
To accomplish cytokinesis, the PP has to lead the forming
cell plate to the right place at about the right time. This
intriguing process involves interaction among dynamic
MTs of the PP, actin filaments and the cortical division site
near the parental cell. The nature of this site is still a mystery. Both MTs and actin filaments are present and are
required to mark the correct site prior to mitosis when the
transient PPB forms [2]. Disassembly of the PPB can be
accelerated by injection of CDC2, suggesting cell cycle progression may participate in marking the site [44]. Several
hypotheses could explain how the cortical division site is
marked. One hypothesis is that marker proteins are phosphorylated by CDC2 [45]. Alternatively, a localization of
ion pumps at the membrane site has been proposed [46].
Another intriguing idea is that the site could be specified by
the absence, rather than the presence, of a marker protein.
It is interesting that actin remains in the cortex throughout
division except at the division site, a position referred to as
the actin-depleted zone (Figure 1; [47]). This hypothesis
suggests that the absence of scaffolding in the division site
balances the PP against the extensive actin scaffolding elsewhere in the cell. Other than the actin-depleted zone, no
persistent structural or biochemical changes have been
detected in the cortical division site.
Selecting the division site
An explanation is still needed for how the site is chosen,
even if specific markers are eventually identified. The
division site could indeed be established earlier than just
prior to mitosis. For example, the site for any single new
cross wall could be influenced by the size and shape of the
parent cell and neighboring cells [48,49]. Change in cell
wall characteristics could be induced by local strains,
caused by differential expansion among interconnected
cells [48]. The new division site would then be selected in
response to the changed physical parameters of the tissue.
Another related view is that the division site is selected on
the basis of geometrically correct positions with the cell
[50]. Such physical explanations could signal a cell to
divide in a particular location, but the biochemical
responses elicited by the signal remain unknown.
Spatial signals of cytokinesis were studied recently in onion
root cells [51•]. When division is temporarily disrupted by
caffeine, the resulting binucleate cells will eventually adjust
and begin dividing again [51•]. Enlarged binucleate cells
generally produced a single PPB at a site predicted from the
size and shape of the cell prior to the treatment. These cells
would then continue to divide with altered spindle and PP
alignments, frequently not related to the most recent PPB
position. The results contradict other cases, however, in
Leaf surface impressions showing patterns of division and expansion at
the base of a developing primordium, in wild-type maize (a,c) and in a
wty1 mutant (b,d). (c) The base of the primordium depicted in Figure 3
was examined under the light microscope and shows the typical
squarish cells of a rapidly dividing epidermis. Recent cross walls are
apparent because they are light and less indented than older parental
walls. In this basal zone, cell divisions are similar in (d) the mutant
compared with (c) the non-mutant. (b) Further up the primordium, mutant
cells begin to over-expand (not shown) and subsequently incomplete
cytokinesis is observed in the enlarged cells (arrowhead) compared with
(a) non-mutant. Cells such as these are endoreduplicated but appear to
have normally arrayed microtubules (data not shown). Quantitative
analysis of the cytokinesis defects in relation to the enlarged cells
suggests that incomplete cytokinesis may result from loss of correct
spatial information in the growing cell. Leaf shape is unaffected, in part
due to the fact that cells adjoining the wart adjust their growth pattern
accordingly. The scale bar represents 60 µm.
which the PPB is absolutely required for correct wall placement, such as when a mutant lacks the PPB [52] or when
PPs adjust to PPB position after normal spindle rotations
[53]. In the experimentally induced binucleate cells, PP
position may be signaled by the nucleus or alternatively, by
older signals remaining from a prior division. These observations help define the normal parameters of site selection
but we clearly do not know all the molecules involved in
selecting the division site.
A genetic approach may help explain how the division site is
selected. The mutation warty 1 (wty1) is providing information about cell sizes that constrain normal cytokinesis [54•].
The wty1 mutants have cytokinesis defects restricted to
groups of cells, producing warts in the leaf blade (Figure 2).
Cell sizes can be measured in developing warts near the base
of the maize leaf primordium, because of the graduated balance between division and expansion (Figure 3). Epidermal
cells in a developing wart initially appear normal with respect
to size (Figure 3) and in distribution of cytoskeletal arrays
(AW Sylvester, unpublished data). Mutant cells that begin to
62
Growth and development
exceed a constant aspect ratio (length to width ratio) at a
given position in the primordium will eventually be unable
to complete cytokinesis (Figure 3). One explanation for
these observations is that over-enlargement prevents normal
cytokinesis, possibly by destroying spatial signals within the
cell. In these cells, expansion may be prolonged due to an
inability to enter mitosis, as wart cells also have endoreduplicated nuclei with condensed chromosomes (AW Sylvester,
unpublished data). Apparently, PP are able to form in warts
but appear to lack directional information. Ongoing analysis
of the gene and its product will help clarify the role of WTY1
in site selection.
Stabilizing and guiding the phragmoplast
The same proteins may be involved in selecting and
also guiding the PP to the division site. For example,
actin patches were observed in the actin-depleted zone
in dividing cells [47]. Actin filaments emanating from
the PP could attach to these patches left behind when
the PPB disappears [47,55]. The PP could thus be tethered to the division site by the same proteins that
originally established the site. Injected profilin, which
effectively binds monomeric actin, disrupts cytokinesis,
potentially by abolishing an actin-based mechanical
support to the forming cell plate [55]. Similarly, caffeine, which is thought to inhibit cytokinesis by
disrupting necessary Ca 2+ gradients, degrades actin filaments at the leading edge of the cell plate [43].
Consequently, the cell plate arrives at the wrong location, if at all, and is often incomplete. It is reasonable to
conclude that one role of actin is in the guidance of the
PP to the division site [55].
The PP and cell plate may also be stabilized internally by
actin. In Clivia endosperm cells, actin filaments are initially interspersed among spindle MTs but soon begin to
shorten and polymerize in the equatorial zone, eventually
interdigitating among the MTs of the PP (Figure 1; [56•]).
These shorter actin filaments are also associated with a
vinculin-like protein [56•]. Vinculin is an actin-associated
protein known to moderate actin–membrane connections
[57]. It is possible, therefore, that actin and associated proteins, such as vinculin, could be binding the vesicular
network into a stable structure, while also guiding it to the
division site.
Analysis of the discordia (dcd) mutations dcd1 and dcd2 may
lead us to the molecules that participate in PP guidance
[58••]. The dcd mutations cause abnormal growth of cells
that undergo asymmetric cell division in the leaf epidermis, such as the subsidiary cells in the stomatal complex.
Disruption of cell shape is preceded by improperly positioned cell walls, caused by misguided PPs. Actin
disruption by cytochalasin phenocopies the mutant defect
and the mutant phenotype is also exacerbated by cytochalasin treatment. These results suggest that DCD may
function in either residual marking of the division site
and/or in successful guidance of the PP to the site.
Even if actin proves to be a structural guide, other proteins may participate in the process. Clues are coming
from another maize gene, tangled1 (tan1) [59,60••]. Mutant
tan1 leaves have disorganized cells compared with the
well-oriented cells in the normal maize leaf. Cell disorganization is attributed partly to the inability of mutant cells
to divide in a particular longitudinal orientation (parallel
to the leaf axis). Mutant tan1 cells also lack the normal distrubution of longitudinally oriented PPB, demonstrating
the protein is directly or indirectly required for specifying
the longitudinal division plane. Furthermore, TAN1
appears to be necessary for proper PP guidance: all PPs in
the mutant are slightly disoriented compared with those
in normal cells, including those that are presumed to give
rise to transversely oriented cell walls. These results are
particularly intriguing because they point to the potential
for post-cytokinesis adjustments in wall position: cells that
have slightly oblique PPs may expand differentially and
very slightly, thereby adjusting the new cross wall to a
more typical transverse orientation. Preliminary reports
suggest the gene may encode a protein with hydrophobicity characteristics of a cell-wall protein [61]. Completed
analysis of the gene and its protein will be very interesting, as will an evaluation of the localization of TAN1
relative to PP and PPBs.
Impact of division ‘decisions’ on the plant body
Division orientations of individual cells vary in their
impact on the plant body. For example, morphogenesis of
some leaves does not require carefully orchestrated cell
divisions [1,59]. On the other hand, some roots have
stereotyped division patterns that result in proper disposition of cell layers, required for the acquisition of cell
identities [62]. The impact of cytokinesis defects depends
on the organ, the developmental stage and the nature of
the gene involved. Currently, the phenotypes of mutants
that affect cytokinesis reflect this variability. Some of the
genes identified are probably redundant, others are directly involved in the mechanics of cytokinesis and others may
play a more indirect role in division and expansion. The
remainder of this review will evaluate these mutants based
on the severity of their phenotype.
Fatal divisions
Many mutants with defects in cytokinesis were identified
in genetic screens of seedling-lethal embryo-defective
mutants [63]. The cyt1 mutant of Arabidopsis has defective
cross walls and excessive accumulation of callose, evident
early in embryogenesis [64••]. Interfering with cellulose
synthesis can phenocopy the mutant, suggesting the gene
plays a role in wall production during cytokinesis [64••].
The well-characterized knolle mutation affects body organization of the Arabidopsis embryo and specifically alters
the forming cell plate. As discussed earlier, KNOLLE
must be one of the key players in successful cytokinesis,
probably through a syntaxin-like role in vesicle targeting.
The keule mutant of Arabidopsis also lacks normal cytokinesis [65] as does the cyd mutant of pea [66]. In these cases,
Division decisions and the spatial regulation of cytokinesis Sylvester
cross walls are initiated but not completed, mimicking the
effect of caffeine on cytokinesis.
The gnom/emb30 mutation appears to affect all aspects of
cell division and expansion. Most gnom mutant cells are
able to complete cytokinesis normally, but are disoriented
from the first zygotic division [67]. The EMB30/GNOM
protein shares a domain with a secretory protein in yeast
called Sec7p [68]. Other regions of the protein are similar
to a non-essential yeast protein Yec2p [67]. Although the
precise role of the protein is not known, the pleiotropic
effects of the mutation may be due to generalized defects
in secretion or vesicle function, rather than just secretion
during cell plate formation.
63
organ specificity of mutant defects suggests either redundant functions of the genes in other organs or that the
specific controls of division may vary in different organs.
Spatially adjustable divisions
Differentiation of specific cell types sometimes requires a
highly regulated pattern of division [1,69]. Mistakes frequently cannot be tolerated in these types of
differentiation divisions. One example is in the required
asymmetric division that yields polarized cells. Pollen
grains will differentiate appropriately only if the first
mitotic cross wall is deposited asymmetrically. Incomplete
cytokinesis, as seen in gem1 mutants, disrupts differentiation of the vegetative and generative cells [70•]. Similarly,
asymmetric cell division precedes differentiation of the
layered Arabidopsis root, which is prevented in scr mutants
[71]. In this case, the division alteration is more likely due
to a response to a change in positional information rather
than a direct effect on cytokinesis. Even so, mutations
such as scr may provide information about how these signals are read by cells, an important aspect of the spatial
regulation of division [69].
Several mutants have problems with cytokinesis but show
quite normal development. Loss of spatial information by
altered cytokinesis may be corrected in a growing organ by
adjusting the direction or rate of cell expansion or increasing the rate of cell division. For example, tan1 mutants
have normally shaped leaves and organs, despite incorrectly placed cross walls. Cells in tan1 mutants that are
experimentally forced to expand more, by growing in dark
or low-light conditions, show corrected orientations of cell
walls (M Mitkovski, AW Sylvester, unpublished data). Cell
adjustments are also seen in wty1 mutants: improper
growth of warts is balanced by neighboring cells, which
divide more as the leaf is growing [54•]. This neighborly
behavior of cells is consistent with the idea that the emerging organ shape prevails as a signal for the spatial
regulation of cell division during morphogenesis of leaves.
The idea requires that cells respond together and communicate locally, via regulated plasmodesmatal trafficking
[75] or other means of intercellular communication. MTs
and cellulose microfibrils, which together direct oriented
cell expansion, are observed to be co-aligned in cell groups
[49,60••,76], consistent with other observations that neighboring cells grow in coordinated patterns [49]. Mutants,
such as tan1 and wty1, provide information about the
capacity of cells to adjust within a given framework. Such
adjustments suggest that the selection of the division site
itself may be flexibly regulated.
Near fatal divisions
Conclusions
Drastic alterations in cell division may have only partial
effects on development. For example, the fass mutants in
Arabidopsis have defects in cytokinesis that result in abnormally shaped cells [72]. Organ morphogenesis is also
severely altered, with plants acquiring a short stubby
appearance, but tissues are distributed normally. These
mutants also lack the normal PPB, suggesting the gene
may be involved in selecting and/or marking the division
site [52]. The ton/fass mutants, among others, show that
tight spatial regulation of cytokinesis is not necessary for
basic tissue differentiation.
The steps of cytokinesis are integrated so that a growing
cell plate is guided to a pre-established division site.
Several proteins are known that probably contribute to
the membranous construction of the cell plate, including
phragmoplastin and KNOLLE. Potential phosphorylation requirements in the cell plate will be understood as
more kinases are identified. Also, other proteins that
might be involved specifically in vesicle targeting, such
as CDC48, will provide information about the interactions between soluble and membrane-bound proteins.
There are several phases in the construction and growth
of the cell plate. In the future, it will be interesting to
understand how these vesicle proteins may behave
uniquely in the cell plate.
Division effects may vary based on whether a gene is
required throughout development or in all cells. The tso1
mutants, for example, show cytokinesis defects only in floral organs and ovules [73,74]. In floral meristems and sepals,
cross walls are incomplete and nuclei endoreduplicated,
similar to the cytokinetic defects in plants with knolle, keule,
cyd and wty1 mutations. Developmental arrest of ovules in
tso1 mutants appears to be due to abnormal and uncoordinated cell expansion, a good example of the same gene
being required for division or expansion, depending on the
organ [73]. Similarly, wty1 mutants show altered cytokinesis,
but only in leaf blade cells [54•]. If these are all null alleles,
Two major cytoskeletal components are required for successful cytokinesis — MTs and actin. The relationship of
the arrays to the forming cell plate has been clarified experimentally and by cytological observations. Kinesin-related
proteins that participate in maintaining the structure and
function of the PP are also being identified. Motorized
transport along MTs is still the best hypothesis for how
vesicles get to the cell plate but, as yet, proteins that mediate this process have not been definitively identified. Actin
64
Growth and development
is also likely to be responsible for unifying the cell plate and
for providing internal as well as lateral support for the PP.
Actin may well serve as a scaffolding upon which the PP is
poised as it extends radially to the division site. The mechanism of guiding the PP to its division site is just beginning
to unfold, as new mutants are found that influence this
process. Mutations such dcd1 and dcd2 are likely to provide
key information about PP guidance.
One outcome of cytokinesis is that two daughter cells are
initially positioned relative to neighbors and to the original parent cell. On the basis of mutant phenotypes, the
impact of this division ‘decision’ on subsequent development depends on the developmental stage, the type of
protein required, or on the extent of genetic redundancy.
Some mutants, such as tan1 and wty1 can adjust for cytokinesis defects, presumably by expanding or dividing
differently to compensate for the loss of spatially correct
information. In other cases, a correct cell division orientation may be required prior to a critical differentiation step,
so that a spatially defective cytokinesis cannot be tolerated, as in the gem1 mutant. Finally, seedling-lethal
mutations, such as cyt1, knolle, keule, cyd may cause the loss
of some essential function in cytokinesis itself. While the
mechanics of cytokinesis is probably shared for different
cells in different organs, it is possible that the means of
selecting the division site may prove to be variable. In the
long term, the question of how a division site is selected
should be considered in the context of the entire growing
tissue. The interconnections of plant cells suggest that
any single division ‘decision’ influences neighboring cells
biochemically and ultimately physically. Continued characterization of mutants, and molecular analysis of the
genes involved, will help to clarify both the mechanics
and the determinants of cytokinesis.
Acknowledgements
I wish to thank Jim Reynolds for contributions at all levels to this work. I
thank Miso Mitkovski for micrographs and the US Department of
Agriculture and the National Science Foundation—Experimental
Program for Stimulating Competitive Research for financial support. This
review is dedicated to the memory of Professor Paul Green, a source of
inspiration and insight.
References and recommended reading
Papers of particular interest, published within the annual period of review,
have been highlighted as:
• of special interest
•• of outstanding interest
6.
Mironov V, De Veylder L, Van Montagu M, Inze D: Cyclin-dependent
kinases and cell division in plants-the nexus. Plant Cell 1999,
11:509-522.
7.
Franklin AE, Cande WZ: Nuclear organization and chromosome
segregation. Plant Cell 1999, 11:523-534.
8.
Wymer C, Lloyd C: Dynamic microtubules: implications for cell
wall patterns. Trends Plant Sci 1996, 1:222-228.
9.
Cosgrove DJ: New genes and new biological roles for expansins.
Curr Opin Plant Biol 2000, 3:73-78.
10. Yasuhara H, Sonobe S, Shibaoka H: Effects of brefeldin A on the
formation of the cell plate in tobacco BY- 2 cells. Eur J Cell Biol
1995, 66:274-281.
11. Hardham AR, Gunning BES: Some effects of colchicine on
microtubules and cell division in roots of Azolla pinnata.
Protoplasma 1980, 102:31-51.
12. Thyberg J, Moskalewski S: Role of microtubules in the organization
of the Golgi complex. Exp Cell Res 1999, 246:263-279.
13. Yasuhara H, Sonobe S, Shibaoka H: Effects of taxol on the
development of the cell plate and of the phragmoplast in tobacco
BY-2 cells. Plant Cell Physiol 1993, 34:21-29.
14. Asada T, Collings D: Molecular motors in higher plants. Trends
Plant Science 1997, 2:29-37.
15. Asada T, Kuriyama R, Shibaoka H: TKRP125, a kinesin-related
protein involved in the centrosome-independent organization of
the cytokinetic apparatus in tobacco BY-2 cells. J Cell Sci 1997,
110:179-189.
16. Song H, Golovkin M, Reddy AS, Endow SA: In vitro motility of
AtKCBP, a calmodulin-binding kinesin protein of Arabidopsis.
Proc Natl Acad Sci USA 1997, 94:322-327.
17.
•
Narasimhulu SB, Reddy AS: Characterization of microtubule
binding domains in the Arabidopsis kinesin-like calmodulin
binding protein. Plant Cell 1998, 10:957-965.
The motor domain of the kinesin-related calmodulin binding protein, KCBP,
is shown to bind microtubules in cosedimentation assays with the purified
protein. Two MT-binding domains are identified in KCBP, one near the carboxyl terminus and one near the amino terminus. The protein could thus be
involved in bundling MTs. Evidence for the function of this protein in organizing MTs is thus growing.
18. Smirnova EA, Reddy AS, Bowser J, Bajer AS: Minus end-directed
•
kinesin-like motor protein, Kcbp, localizes to anaphase spindle
poles in Haemanthus endosperm. Cell Motil Cytoskeleton 1998,
41:271-280.
KCBP distribution is observed in endosperm cells during mitosis and cytokinesis. Interesting shifts in localization patterns are seen during a transition
through telophase. KCBP shifts from an association with the spindle poles
at the end of mitosis to the equatorial zone. KCBP is eventually localized
along the phragmoplast MTs, but also to the nuclear membrane early in
telophase and to the cell plate in late telophase. Slight differences in KCBP
distribution with BY-2 cells may be due to unique properties of the cellularizing endosperm.
19. Liu B, Cyr RL, Palevitz BA: A kinesin-like protein, KatAp, in the cells
of Arabidopsis and other plants. Plant Cell 1996, 8:119-132.
20. Compton DA: Focusing on spindle poles. J Cell Sci 1998,
111:1477-1481.
21. Kakimoto T, Shibaoka H: Cytoskeletal ultrastructure of
phragmoplast-nuclei complexes isolated from cultured tobacco
cells. Protoplasma 1988, 2:95-103.
22. Parke J, Miller C, Anderton BH: Higher plant myosin heavy-chain
identified using a monoclonal antibody. Eur J Cell Biol 1986, 41:9-13.
1.
Sylvester AW, Smith L, Freeling M: Acquisition of identity in the
developing leaf. Annu Rev Cell Dev Biol 1996, 12:257-304.
2.
Assaad FF, Mayer U, Lukowitz W, Juergens G: Cytokinesis in
somatic plant cells. Plant Physiol Biochem 1997, 35:177-184.
3.
Fowler JE, Quatrano RS. Plant cell morphogenesis: plasma
membrane interactions with the cytoskeleton and cell wall. Annu
Rev Cell Dev Biol 1997, 13:697-743.
4.
Samuels AL, Giddings TH, Staehelin LA: Cytokinesis in tobacco BY2 and root tip cells: a new model of cell plate formation in higher
plants. J Cell Biol 1995, 130:1345-1357.
25. Park JM, Kang SG, Pih KT, Jang HJ, Piao HL, Yoon HW, Cho MJ,
Hwang I: A dynamin-like protein, ADL1, is present in membranes
as a high-molecular-mass complex in Arabidopsis thaliana. Plant
Physiol 1997, 115:763-771.
5.
Staehelin LA, Hepler PK: Cytokinesis in higher plants. Cell 1996,
84:821-824.
26. Kozlov MM: Dynamin: possible mechanism of ‘Pinchase’ action.
Biophys J 1999, 77:604-616.
23. Asada T, Shibaoka H: Isolation of polypeptides with microtubuletranslocating activity from phragmoplasts of tobacco BY-2 cells.
J Cell Sci 1994, 107:2249-2257.
24. Gu XJ, Verma DPS: Phragmoplastin, a dynamin-like protein
associated with cell plate formation in plants. EMBO J 1996,
15:695-704.
Division decisions and the spatial regulation of cytokinesis Sylvester
27.
Lauber MH, Waizenegger I, Steinmann T, Schwarz H, Mayer U, Hwang
I, Lukowitz W, Jurgens G: The Arabidopsis KNOLLE protein is a
cytokinesis-specific syntaxin. J Cell Biol 1997, 139:1485-1493.
28. Gu X, Verma DP: Dynamics of phragmoplastin in living cells during
cell plate formation and uncoupling of cell elongation from the
plane of cell division. Plant Cell 1997, 9:157-169.
29. Slepnev VI, Ochoa GC, Butler MH, Grabs D, Camilli PD: Role of
phosphorylation in regulation of the assembly of endocytic coat
complexes. Science 1998, 281:821-824.
30. Sharp DJ, McDonald KL, Brown HM, Matthies HJ, Walczak C, Vale RD,
Mitchison TJ, Scholey JM: The bipolar kinesin, KLP61F, cross-links
microtubules within interpolar microtubule bundles of Drosophila
embryonic mitotic spindles. J Cell Biol 1999, 144:125-138.
31. Hirt H: Multiple roles of MAP kinases in plant signal transduction.
Trends Plant Sci 1997, 2:11-15.
32. Reszka AA, Seger R, Diltz CD, Krebs EG, Fischer EH: Association of
mitogen-activated protein kinase with the microtubule
cytoskeleton. Proc Natl Acad Sci USA 1995, 92:8881-8885.
33. Calderini O, Bogre L, Vicente O, Binarova P, Heberle-Bors E,
•
Wilson C: A cell cycle regulated MAP kinase with a possible role
in cytokinesis in tobacco cells. J Cell Sci 1998, 111:3091-3100.
This MAPK from tobacco is activated during late anaphase and the protein
localizes to the cell plate. Disruption of MTs with inhibitors prevents activation of the kinase.
34. Bogre L, Calderini O, Binarova P, Mattauch M, Till S, Kiegerl S,
•
Jonak C, Pollaschek C, Barker P, Huskisson NS et al.: A MAP kinase
is activated late in plant mitosis and becomes localized to the
plane of cell division. Plant Cell 1999, 11:101-114.
A MAPK is characterized from alfalfa. The protein appears to be distributed
in the region of the cell plate and remains there even as the phragmoplast
extends radially.
35. Banno H, Hirano K, Nakamura T, Irie K, Nomoto S, Matsumoto K,
Machida Y: NPK1, a tobacco gene that encodes a protein with a
domain homologous to yeast BCK1, STE11, and Byr2 protein
kinases. Mol Cell Biol 1993, 13:4745-4752.
36. Nakashima M, Hirano K, Nakashima S, Banno H, Nishihama R,
•
Machida Y: The expression pattern of the gene for NPK1 protein
kinase related to mitogen-activated protein kinase kinase kinase
(MAPKKK) in a tobacco plant: correlation with cell proliferation.
Plant Cell Physiol 1998, 39:690-700.
An upstream MAPK (identified as a MAPK kinase kinase) was found in
tobacco and is expressed in dividing cells. Subcellular localization is still
needed to clarify whether or not this protein is part of a phosphorylation cascade in the cell plate or phragmoplast.
37.
Lukowitz W, Mayer U, Jurgens G: Cytokinesis in the Arabidopsis
embryo involves the syntaxin-related KNOLLE gene product. Cell
1996, 84:61-71.
38. Robinson DG, Hinz G, Holstein SE: The molecular characterization
of transport vesicles. Plant Mol Biol 1998, 38:49-76.
39. Jantsch-Plunger V, Glotzer M: Depletion of syntaxins in the early
Caenorhabditis elegans embryo reveals a role for membrane
fusion events in cytokinesis. Curr Biol 1999, 9:738-745.
40. Mellman I: Enigma variations: protein mediators of membrane
fusion. Cell 1995, 82:869-872.
41. Frohlich KU, Fries HW, Rudiger M, Erdmann R, Botstein D, Mecke D:
Yeast cell cycle protein CDC48p shows full-length homology to
the mammalian protein VCP and is a member of a protein family
involved in secretion, peroxisome formation, and gene
expression. J Cell Biol 1991, 114:443-453.
42. Feiler HS, Desprez T, Santoni V, Kronenberger J, Caboche M, Traas J:
The higher plant Arabidopsis thaliana encodes a functional
CDC48 homologue which is highly expressed in dividing and
expanding cells. EMBO J 1995, 14:5626-5637.
43. Valster AH, Hepler PK: Caffeine inhibition of cytokinesis: effect on
the phragmoplast cytoskeleton in living Tradescantia stamen hair
cells. Protoplasma 1997, 196:155-166.
44. Hush J, Wu L, John PC, Hepler LH, Hepler PK: Plant mitosis
promoting factor disassembles the microtubule preprophase
band and accelerates prophase progression in Tradescantia. Cell
Biol Int Rep 1996, 20:275-287.
45. Cleary AL: F-actin redistributions at the division site in living
Tradescantia stomatal complexes as revealed by microinjection of
rhodamine–phalloidin. Protoplasma 1995, 185:152-165.
65
46. Cyr R: Calcium/calmodulin affects microtubule stability in lysed
protoplasts. J Cell Sci 1991, 100:311-317.
47.
Cleary AL, Gunning BES, Wasteneys GO, Hepler PK: Microtubule
and F-actin dynamics at the division site in living Tradescantia
stamen hair cells. J Cell Sci 1992, 103:977-988.
48. Green PB: Transductions to generate plant form and pattern: an
essay on cause and effect. Ann Bot 1996, 78:269-281.
49. Sylvester AW, Williams MH, Green PB: Orientation of cortical
microtubules correlates with cell shape and division direction:
immunofluorescence of intact epidermis during development of
Graptopetalum paraguayensis. Protoplasma 1989, 153:91-103.
50. Flanders DJ, Rawlins DJ, Shaw PJ, Lloyd CW: Nucleus-associated
microtubules help determine the division plane of plant epidermal
cells: avoidance of four way junctions and the role of cell
geometry. J Cell Biol 1990, 110:1111-1122.
51. Gimenez-Abian MI, Utrilla L, Canovas JL, Gimenez-Martin G,
•
Navarrete MH, De la Torre C: The positional control of mitosis and
cytokinesis in higher-plant cells. Planta 1998, 204:37-43.
Binucleate wheat cells are produced by caffeine treatment and subsequent
patterns of phragmoplast arrangement are recorded in cells of different
sizes. The authors conclude that PP could form without respect to the PPB
in some cases, suggesting pre-existing spatial cues may dictate cell wall
position under certain circumstances.
52. Traas J, Bellini C, Nacry P, Kronenberger J, Bouchez D, Caboche M:
Normal differentiation patterns in plants lacking microtubular
preprophase bands. Nature 1995, 375:676-677.
53. Palevitz BA: Division plane determination in guard mother cells of
Allium: video time-lapse analysis of nuclear movements and
phragmoplast rotation in the cortex. Dev Biol 1986, 117:644-654.
54. Reynolds JO, Eisses JF, Sylvester AW: Balancing division and
•
expansion during maize leaf morphogenesis: analysis of the
mutant, warty-1. Development 1998, 125:259-268.
Cells with wty1 mutations undergo faulty cytokinesis once the cells have
begun to over-expand. The results suggest that one spatial cue for the division site may be appropriate length to width or volume ratio in the cell. In
addition, neighboring cells in the wty1 mutant, which can look normal,
appear to adjust to the defect by dividing more.
55. Valtzer AH, Pierson ES, Valenta R, Hepler PK, Emons AMC: Probing
the plant actin cytoskeleton during cytokinesis and interphase by
profilin microinjection. Plant Cell 1997, 9:1815-1824.
56. Endle MC, Stoppin V, Lambert AM, Schmit AC: The growing cell
•
plate of higher plants is a site of both actin assembly and
vinculin-like antigen recruitment. Eur J Cell Biol 1998, 77:10-18.
The distribution of actin and an associated vinculin-like protein is described
in endosperm cells of Clivia. Short actin filaments are seen in the center of
the PP and co-localize with vinculin. The authors suggest the proteins are
important in maintaining and perhaps guiding the PP during cytokinesis.
57.
Johnson RP, Craig SW: F-actin binding site masked by the
intramolecular association of vinculin head and tail domains.
Nature 1995, 373:261-264.
58. Gallagher K, Smith LG: discordia mutations specifically misorient
•• asymmetric cell divisions during development of the maize leaf
epidermis. Development 1999, 126:4623-4633.
These mutants show an interesting and informative phenotype. Asymmetric
divisions that occur during differentiation of the epidermis are affected so
that cytokinesis is disrupted and cells expand in abnormal orientations.
Normally positioned PPB are present, predicting the correct site for the
impending new wall. PP lose direction relative to the site, however, suggesting that they are not guided properly or that residual information marking the site is lost. Particularly intriguing is the fact that actin filament
disruption can phenocopy the mutants. The results implicate the genes in a
PP guidance process that may require actin or associated proteins.
59. Smith LG, Hake S, Sylvester AW: The tangled-1 mutation alters cell
division orientations throughout maize leaf development without
altering leaf shape. Development 1996, 122:481-489.
60. Cleary AL, Smith LG: The Tangled1 gene is required for spatial
•• control of cytoskeletal arrays associated with cell division during
maize leaf development. Plant Cell 1998, 10:1875-1888.
Stunning views of the distribution of MTs and actin filaments in tan1 and nonmutant leaf primordia are presented. Detailed analysis of cytoskeletal arrays
in mutants compared with non-mutants confirm that longitudinal divisions are
specifically disrupted in the mutant. PPBs do not occupy the expected longitudinally oriented position. PP formations are also slightly disoriented compared with normal.
66
Growth and development
61. Miller D, Hable W, Gottwald J, Ellard-Ivey M, Demura T, Lomax T,
Carpita N: Connections: the hard wiring of the plant cell for
perception, signaling, and response. Plant Cell 1997, 9:2105-2117.
69. Scheres B, Heidstra R: Digging out roots: pattern formation, cell
division and morphogenesis in plants. Curr Top Dev Biol 1999,
45:207-247.
62. Berger F, Hung CY, Dolan L, Schiefelbein J: Control of cell division in
the root epidermis of Arabidopsis thaliana. Dev Biol 1998,
194:235-245.
70. Park SK, Howden R, Twell D: The Arabidopsis thaliana
•
gametophytic mutation gemini pollen1 disrupts micros