Annotated Video Collection
|
Annotated up to and including
Vol 168 No 1
|
As videos
have become more popular in biology, they have found many applications. They have
been used to focus through
a thick sample, or provide a three-dimensional
view. But, for the most part, they are a way of seeing what happens in a cell
over time. As can be seen in this collection, the results are often spectacular.
Not all videos published in The
JCB are shown below (for the complete list of available videos, see the
Supplemental Material Archive). Videos have been
selected because they illustrate the central result of a paper, and do so in
a way that is more instructive than the use of still images. Each selected paper
is represented by a one-line heading, which can be used to jump down to the
short description of and links to the selected videos. The description includes
direct links to the abstract of the paper and to the individual videos that
illustrate a particular result. For many papers there are more videos than there
are links in the description; to access the full list of videos for a particular
paper please see the Supplemental Material Archive.
Papers have been placed into topic
categories and subcategories. Within each subcategory, more recent papers are
shown at the top of the list. The categorization of selected papers and videos
is approximate, so it is worth scanning several sections for a desired topic.
The same paper on microtubule function in neurons could, for example, easily
find its place under either the microtubule or neurite-outgrowth heading.
Annotations of Videos
Cell
cycle proteins
Fzy and
Fzr cooperate to destroy cyclin B in flies
Raff
et al. examine two regulators of fly cyclin
B destruction: Fizzy (Fzy)/Cdc20 and Fzy-related (Fzr)/Cdh1. Fzy/Cdc20
is concentrated at kinetochores and centrosomes early in mitosis, whereas Fzr/Cdh1
is concentrated at centrosomes throughout the cell cycle. In syncytial embryos,
only Fzy/Cdc20 is present, and only the spindle-associated cyclin B is degraded
at the end of mitosis. A mutant form of cyclin B that cannot be targeted for
destruction by Fzy/Cdc20 is no longer degraded on spindles of syncytial
embryos, but still targeted by Fzr/Cdh1 in cellularized
embryos, albeit more slowly than normal. This suggest that Fzy/Cdc20
is responsible for catalyzing the first phase of cyclin B destruction that
occurs on the mitotic spindle, whereas Fzr/Cdh1 is responsible for catalyzing
the second phase of cyclin B destruction that occurs throughout the cell.
Chromosome
dynamics
A motor helps holocentrics pull away from
poles
Holocentric chromosomes in worms have kinetochores distributed along the length
of the chromosome. Powers
et al. show that the plus-end microtubule motor KLP-19 helps pull
chromosomes away from a monopole. This may put tension on a mono-attached
chromosome, swinging the attached kinetochore towards the pole that it is attached
to and away from the opposite pole. Without this system a single kinetochore
can attach to both poles leading to lagging
chromosomes and errors relative to a normal
mitosis.
Kinetochores switch to depolymerization at anaphase
Maddox et al. find
that during
anaphase both poleward flux of microtubules and net depolymerization of
microtubules at the kinetochores contribute to the movement of chromosomes. During metaphase,
however, flux is compensated for by net polymerization of microtubules at the
kinetochores.
The
polar wind in action
The distribution of chromokinesins
along chromosome arms makes these motors excellent candidates for mediating
the polar ejection force. This force, which projects chromosome arms away from
the poles, has been proposed to aid in chromosome congression to the spindle
midzone. Levesque and Compton
test this model by injecting antibodies specific for the chromokinesin Kid.
They start by looking in cells that
have monopolar spindles because of the injection of Eg5 antibodies. Polar
ejection, which results in chromosomes spreading outwards from the spindle
pole, can be seen in these cells. But in cells injected with antibodies to both
Eg5 and Kid the chromosomes
cluster around the single spindle pole.
In normal spindles the difference
is less severe. Normal chromosome
congression is perturbed only slightly by the projection
of chromosome arms towards spindle poles, and only a minority of mitoses
are delayed.
Merotelic
attachments are a major source of segregation errors
A chromosome can be left behind
during anaphase if it has a merotelic orientation, i.e., a single chromatid
has its kinetochore attached to kinetochore microtubule bundles extending towards
both spindle poles. Cimini et al.detect
such merotelic orientations by light
microscopy and confirm them by electron
microscopy. They represent a major mechanism of aneuploidy not detected
by the mitotic spindle checkpoint.
CENP-meta
is a motor for maintaining chromosome alignment
Yucel
et al. find that CENP-meta, a fly kinesin-like motor most similar to vertebrate
CENP-E, is required for the maintenance
of chromosomes at the metaphase plate (compare this video to that of a normal
mitosis). Thus there are distinguishable requirements for either establishing
or maintaining chromosome congression.
Spindle
checkpoint
Dynein
shuts down the spindle checkpoint
Injection of a prometaphase cell
with dynamitin, to inhibit dynein, blocks a cell in metaphase. But this block is relieved
by a second injection of anti-Mad2 antibodies, which inhibit the spindle checkpoint.
Howell et al. propose that
the mitotic block occurs because dynein's normal function is to transport several
kinetochore proteins, including checkpoint proteins, away from the kinetochore
and towards the spindle poles. This action of dynein helps shut down the checkpoint
once kinetochores are correctly attached to the spindle.
Kinetochores
may act as sites for catalytically sequestering checkpoint proteins
Howell
et al. find that the spindle checkpoint protein Mad2 (red) is a transient
component of unattached kinetochores, as predicted by the catalytic model.
In this model, unattached kinetochores provide sites for assembling and releasing
Mad2-Cdc20 complexes, which sequester Cdc20 and prevent it from activating the
events necessary for mitotic exit.
Spindle
construction
Microtubule ejection thins out spindles
Rusan and Wadsworth
show that microtubules are ejected
from centrosomes later in mitosis, thus thinning out the astral arrays in favor
of the central spindle. This may help the central spindle to provide a unique
signal specifying the position of the cytokinetic furrow.
Dynein takes depolymerization to the poles
Spindle microtubules constantly move towards spindle poles
in a process called flux,
but the length of the pre-anaphase spindle is maintained by microtubule polymerization
at plus ends and microtubule depolymerization at minus ends near the poles. In
theory the microtubules could be pulled towards the poles by depolymerization,
but Gaetz and
Kapoor separate the movement and depolymerization by inhibiting dynein
action. The result is cessation of pole-localized depolymerization and thus elongating
spindles;
it appears that dynein normally transports the microtubule-depolymerizing motor
Kif2a to the poles. But flux is not affected, suggesting that other translocating
motors, not pole-localized depolymerization, drive flux.
Kinetochore-nucleated
microtubules help build spindles
A new contributor to spindle assembly is described by Khodjakov
et al. They find that microtubules can be nucleated from kinetochores
of monooriented chromosomes, and those microtubules are then incorporated into
spindles that are either recovering
from a monopolar state or bipolar but recruiting
a monooriented chromosome.
Interphase
microtubules are dragged into the spindle region during prophase
The rapid
disassembly of the interphase microtubule network during prophase was observed
by Rusan et al. The microtubules
move inward along other microtubules,
sometimes even doing a U-turn
before moving back toward the centrosome.
EB1-driven
dynamics helps during spindle construction
Rogers
et al. find that depletion of EB1, a protein that localizes to the plus ends
of growing microtubules, suppresses the normal dynamic
growth and shrinkage of microtubules, leaving only passive,
lateral movements. In fly embryos, this results in the failure
of spindle elongation seen during normal
mitotic divisions.
Aurora-A
is required for spindle assembly
Normal spindle
assembly in worm embryos requires the activity of Aurora-A kinase, according
to Hannak et al. They find that
embryos lacking Aurora-A fail in the microtubule-independent accumulation
of γ-tubulin and two other pericentriolar components, which results
in a failure
of spindle assembly in which asters collapse onto each other. In a later
paper, the same authors showed that centrosomal arrays can still
form in the absence of γ-tubulin,
but
γ-tubulin is necessary for formation of
a full mitotic spindle, and is the kinetically dominant centrosomal microtubule
nucleator.
Evidence
for a possible spindle matrix
Kapoor
and Mitchison investigate the possibility that a static spindle matrix
helps tether the spindle. They find that the mitotic kinesin Eg5 remains relatively
stationary even as the
microtubule spindle fluxes towards the spindle poles. This could be explained if Eg5 is walking along
microtubules at the rate of flux but in the opposite direction, but the authors
find that Eg5 distribution remains stationary even when Eg5 motor activity is
inhibited. This leads to the proposition that a spindle matrix may exist and
have Eg5 as one component.
Skeletor
as a candidate component of a spindle matrix
Walker
et al. propose that Skeletor, a protein they identify from flies, could
form the basis of a microtubule-independent spindle matrix. Such a matrix could
stabilize force production by the microtubule-based spindle. The Skeletor-based
structure can form independently of microtubules, and normally arises during prophase before the microtubule-based spindle has been completed (see also the enlarged image
of a single nucleus, with
microtubules in green and Skeletor in red). The two structures coalign during metaphase.
Daughter centrioles move and are
then anchored
Piel
et al. show that mother centrioles stay immobile during G1, but daughter
centrioles move extensively in a process dependent on the presence of either actin or microtubules.
The gradual decrease in these movements as the cell cycle progresses suggests
that there is a maturation-dependent process of centriole anchoring.
Centriolar satellites are transported towards centrosomes
Kubo
et al. find that particles called centriolar satellites (green) are transported to their final resting place near centrosomes.
Organelle
partitioning
Mitochondrial replication factories
DNA replication
factories in mitochondria divide even
in the absence of mitochondrial DNA, according to Meeusen
and Nunnari.
Myosin V may help partition the
ER
Wöllert
et al. find that, during mitosis, myosin V–driven movement of small
globular vesicles along F-actin is strongly inhibited, but the movement
of ER and ER network
formation on F-actin is up-regulated. Thus F-actin may help partition the ER
during cell division.
Peroxisomes
segregate using Myo2p
Hoepfner
et al. analyzed the movement
of peroxisomes in budding yeast. In the absence of the dynamin-like protein
Vps1p only one or two giant peroxisomes remained, but segregation still
occurred. Peroxisome movement was abolished by
latrunculin A treatment, and movement was also found to be dependent on the
myosin motor Myo2p.
Golgi clusters partition
during mitosis
Golgi partitioning during mitosis
has been suggested to occur via fusion of the Golgi with the endoplasmic reticulum.
In contrast, Jokitalo et al. find
that the Golgi membranes split into small clusters that persist through mitosis
and partition early in mitosis
to two sides of the nucleus. These two sets of clusters are pushed
apart during anaphase, and form the basis for the re-formation
of the Golgi in the two daughter cells.
No tetraploidy checkpoint
Concentrated actin polymerization inhibitors can both induce cytokinesis failure
and halt the cell cycle, leading some to propose the existence of a tetraploidy
checkpoint that detects the aberrant binucleate state. But Uetake
and Sluder show that lower levels
of the inhibitors still result in cytokinesis failure but now allow
continued division of the binucleate cells,
suggesting that such a checkpoint does not exist in these cell types.
Minimal requirements for inducing cytokinesis
Cytokinesis is known to be induced by DNA-containing mitotic spindles, but Alsop
and Zhang
find that micromanipulated cells can still go through cytokinesis even if they
have only an aster
or bundle of
microtubules.
Rho activation during cytokinesis
Rho-family GTPases are thought to stimulate contraction of the actomyosin
ring during cytokinesis. Yoshizaki
et al. find, based on FRET activity that rises when the GTPases are activated,
that the dynamics of activation during cytokinesis are not quite such a simple
story. Although RhoA
is active at the cleavage furrow, the patterns for Rac1 and Cdc42
activity are not as clear.
Rings that promote cytokinesis
Tasto et al. characterize
Mid2, a fission yeast homologue of the cell division protein anillin. Mid2
forms a ring around
the center of dividing cells, colocalizing and splitting along with the septin
ring. Both rings
help promote late events in cytokinesis.
Centrosomes
are needed for spindle orientation but not cytokinesis
Khodjakov
and Rieder find that centrosomes are not necessary for a normal
cytokinesis, but cells lacking centrosomes do suffer from a lack of astral
microtubules and therefore sometimes misorient their spindles leading to failures in cytokinesis. Cells lacking centrosomes also fail to enter S phase in the
following cell cycle.
Chloroplast FtsZ forms a ring at the division site
Bacterial FtsZ mediates cell division
by forming a ring at the cell division site. Vitha
et al. show that chloroplast FtsZ also forms a ring
and thus probably acts in a similar manner.
Chromosomes
TopoII α
is mobile
The DNA-decatenating protein topoisomerase
II has been proposed as a stable scaffold for mitotic chromosomes. Although
Tavormina et al. do not rule
out this idea, they show that a fluorescent version of DNA topoisomerase II α
(topoII α) protein turns
over rapidly. This may allow topoIIα,
which is concentrated toward
the axes of mitotic chromosome arms, to quickly reach and relieve areas
of chromosomal strain that develop during mitosis. In another study examining
the localization of
topoII, Christensen et al.
come to similar conclusions.
The nuclear envelope may help retard origin firing
Time-lapse microscopy of GFP-marked
origins allows Heun et al. to show that late-firing origins are enriched in a zone immediately adjacent
to the nuclear envelope during G1, at which time a modified chromatin structure
may be established to retard origin firing.
Nuclear
pores
Nuclear pore complexes are fixed in place
Daigle
et al. report that nuclear pore complexes (NPCs) undergo limited
movements that match the deformations of the nuclear envelope as tracked
using a grid of bleached
nuclear lamins. NPCs are therefore remarkably stable complexes, and are probably
anchored to a protein network in the nuclear envelope.
Nucleoporins reassemble around post-mitotic chromatin
A conserved nuclear pore subcomplex
was characterized and tracked by Belgareh
et al., who found that the proteins were recruited during telophase in a rim pattern surrounding the chromosomes. A low level
of staining was also apparent on the kinetochores throughout mitosis.
Nucleoli
Nucleolar re-formation
after mitosis
Savino
et al. follow the re-formation of nucleoli after mitosis. Prenucleolar bodies (PNB) form on the chromosome
surface and nucleolar material flows along links
between PNBs and towards a developing nucleolar organizer
region (NOR). Eventually this leads to the fusion of nucleoli to form a single entity.
Processing complexes may help reassemble nucleoli
Nucleolar
reassembly during telophase is shown by Dundr
et al. to require mitotically preserved processing complexes.
Speckles
A
splicing factor has limited mobility
Based on the limited
mobility of a splicing factor, Kruhlak
et al. determine that the factor undergoes frequent but transient interactions
with relatively immobile nuclear binding sites, both when associated with speckles
and when dispersed in the nucleoplasm. This a 3-D video that should be viewed
using red/green 3-D glasses.
ER
to and from Golgi
Sar1 makes tubular ER
export sites
Sar1-GTP induces the formation of
elongated tubules that
are also seen during export from
the ER in living cells. Aridor
et al. therefore suggest that Sar1 links cargo selection with ER morphogenesis
through the generation of these transitional tubular ER export sites.
Two
distinct pathways from Golgi to ER: Rab6 and KDEL
Recycling of proteins from the Golgi
to the ER via the KDEL receptor (red) is distinct
from retrograde Golgi to ER transport dependent on Rab6 (green), as determined
by White et al. Only the KDEL
pathway is dependent on COP-I.
Endosomes
Endosome to trans-Golgi
traffic is vesicular
Coexpression of fluorescent Rab9
and Rab7 by Barbero et al.
revealed that these two late endosome Rabs occupy distinct
domains within late endosome membranes. Rab9 is present on endosomes
that display bidirectional
microtubule-dependent motility, and Rab9-positive transport vesicles
(rather than tubules) can be seen moving
along microtubule tracks and fusing
with the trans-Golgi network.
Recycling
of endosomes by Arf6
Activation of nucleotide exchange
on Arf6 causes an increase in both membrane internalization and the return of the resultant structures
to the plasma membrane. However, when Brown
et al. block Arf6 in its GTP-loaded ("on") state, this results in accumulation of endosome-derived vacuoles.
Actin tails form behind motile endosomes and lysosomes
Taunton
et al. document the formation of actin comet tails (red) behind motile
endosomes and lysosomes. The tails may help the organelles move around
the cell, and influence the construction of the cytoskeleton.
Lysosomes
Lysosomes repair the plasma
membrane
When cells are injured, calcium rushes
in and prompts exocytosis of membranes to repair the gap. Jaiswal
et al. confirm that the repairing membranes are from lysosomes.
Vam6p
clusters and fuses lysosomes
Caplan
et al. find that human Vam6p promotes clustering
and then fusion of lysosomes.
Plasma
membrane
A marker for sequential exocytosis
The SNARE protein SNAP25, say Takahashi
et al.,
marks the plasma membrane after an initial exocytic event to allow rapid sequential
exocytic events.
A myosin V moves yeast secretory vesicles
Secretory vesicles actively
move to the site of exocytosis in yeast. Schott
et al. find that multiple secretory vesicles often follow
the same linear track and frequently enter and cross the bud. This movement
requires the activity
of the myosin-V heavy chain encoded by the MYO2 gene. When the predicted
lever arm of this motor is progressively shortened (with the most extreme example
being the 0IQ mutant), the vesicle movements are progressively slowed.
Rapid
cycling of lipid rafts to and from the Golgi
Nichols
et al. detect rapid cycling of lipid raft markers between the plasma
membrane and the Golgi. Through selective photobleaching, they are able to study
transport either out from the
Golgi to the plasma membrane, or in
from the plasma membrane to the Golgi.
Membrane
docking at the immunological synapse requires Rab27a
Stinchcombe
et al. find that normal
membrane docking of lytic granules at the immunological synapse is defective
in cells lacking Rab27a. In cells lacking other Rab proteins, polarization of
the secretory granules is incomplete.
Visualizing
the location and dynamics of exocytosis
Schmoranzer
et al. use total internal reflection (TIR) fluorescence microscopy to
visualize exocytosis in
mammalian cells (e.g., see event on left side of video). The analysis reveals
that there are no preferred sites for constitutive exocytosis in this system.
Toomre
et al. use a combination of TIR microscopy (green, labeling molecules
close to or at the membrane) and standard fluorescence microscopy (red, for
molecules further from the membrane) to visualize trafficking
to and fusion with the plasma membrane during exocytosis. Red dots turn
yellow then green as they approach the membrane, and then explode in a burst
of light as they fuse with the plasma membrane during exocytosis. The transport
containers appear to be partially anchored at the membrane before fusion, and
can undergo either partial or complete fusion events.
Caveolae
Caveolin
helps traffic lipids
Pol
et al. observe the formation of lipid
droplets in cells expressing a dominant negative caveolin protein, and
suggest that caveolin helps traffic lipids to and from lipid droplets.
Axonal
Special microtubules for getting into axons
Nakata and Hirokawa
find that preferential
transport of cargoes into axons is directed by a special population of microtubules,
which have a high turnover rate and increased binding of the tip-binding protein
EB1.
Retrograde
transport in axons
Lalli
and Schiavo measure retrograde
transport in axons. They show that the transport vesicles carry both
tetanus toxin and NGF, but do
not colocalize with lysosomes.
Wound
healing
Using acid to close a wound
The sodium-proton exchanger NHE1 is needed for normal
migration during wound closing, according
to Denker and
Barber.
In cells with a mutant NHE1 that cannot translocate ions, directed migration fails.
Normal ion exchange by NHE1 may change the pH at the front of the migrating
cells, but the direct consequence of such a putative change is not known.
ARNO
induces migration
Healing of large wounds normally
progresses with the leading edge of cells migrating
smoothly as a unit. Santy
and Casanova find that cells expressing ARNO pull
away from the wound edge and exhibit a distinct fan-shaped leading edge
and a trailing edge with a tail that often remains attached to the body of the
monolayer. ARNO acts as an exchange factor for the small GTPase ARF6, leading
to increased activation of both Rac1 and phospholipase D. These two independent
pathways function together to increase cell migration.
Actomyosin
flow, accumulation, and contraction heal a wound
Mandato
and Bement find that wounds heal using two distinct components: a highly dynamic assembly zone, in which myosin
2 and actin preferentially accumulate, and a stable contractile actomyosin ring.
The contractile nature of the apparatus is revealed in several experiments:
corners of a wound round
up; contacting fingers of actin pull
in the edges of a wound; and breakage
of the contractile ring prevents proper closure. Other videos reveal the flow
of actin, whereas myosin accumulates with little obvious flow.
Dorsal closure resembles wound healing
During fly
development, the process of dorsal
closure (see also another
video) brings together epidermal cells to form a sealed tube. Kiehart
et al. demonstrate that dorsal closure can be largely, but not fully,
explained by a purse-string model in which the leading-edge epidermal cells
contract to seal the epidermal layer. This suggests that fly dorsal closure
has many similarities with mammalian wound healing.
Actin
Microtubule and actin movements are coordinated
Fluorescent speckle microscopy (FSM)
allowed Salmon et al. to examine
both microtubules (MTs) and filamentous actin (f-actin) in migrating newt cells.
F-actin exhibited four zones of dynamic behavior: rapid retrograde flow in the lamellipodium, slow retrograde
flow in the lamellum, anterograde flow in the cell body, and no movement in
the convergence zone between the lamellum and cell body. MTs
moved at the same trajectory and velocity as f-actin in the cell body
and lamellum, but not
in the lamellipodium
or convergence zone. MTs grew
along f-actin bundles, and quiescent MT ends moved in association with f-actin bundles. Thus f-actin movements have a profound
effect on MTs in migrating cells, and MTs and f-actin may bind to one another
in vivo.
Actin-dependent
picket fences slow diffusion in the plasma membrane
Fujiwara
et al. track the movement
of single phospholipid molecules in the plasma membrane. Over the short-term,
these molecules appear to stay within defined compartments, but after an average
of 11ms they hop to
an adjacent compartment. Over even longer time periods (an average of 0.33 s),
the lipids hop between
larger compartments. Similar compartments are not
apparent on vesicles, as trajectories are less closely apposed. Trajectories
within a single compartment are not significantly slowed relative to diffusion
rates in vesicles, so the delays in hopping between compartments must explain
the lower overall diffusion rate for lipids in cellular membranes.
The compartmentalization depends
on the actin-based membrane
skeleton, but not on the extracellular matrix, extracellular domains
of membrane proteins, or cholesterol-enriched rafts. The authors propose that
various transmembrane proteins anchored to the actin-based membrane skeleton
meshwork act as rows of pickets that temporarily confine phospholipids.
An
actin fragment that relaxes myofibroblasts
Myofibroblasts are specialized fibroblasts
that can contract to aid in wound healing, possibly by using stress fibers containing
α-smooth muscle actin (α-SMA).
Hinz et al. join the N-terminal
sequence of α-SMA to a fusion peptide.
Application of the resulting protein to myofibroblasts relaxes the cells reversibly. Such an agent may be useful in treating fibrocontractive
diseases.
A Rop GTPase controls
pollen tube tip growth
Fu
et al. show that Rop1At, a Rop GTPase belonging to the Rho family, controls
actin dynamics and thus
pollen tube tip growth in tobacco plants.
Myosin
Myosin
recruitment drives the distribution of nuclei in fly embryos
As nuclei divide in the early fly
embryo, they are actively distributed along the long axis of the embryo. Royou
et al. show that the cortical contractions that drive this are accompanied
by periodic accumulation of myosin to the cortex. Recruitment occurs at the end
of telophase, correlated with the drop
in cdc2/cyclin B activity, and results in a cortical contraction during
interphase.
A sensor for the activity
and abundance of MLCK
Chew
et al. construct a sensor that can read out both the abundance (shown as
peak height) and activity (red is inactive and blue is active) of myosin light
chain kinase (MLCK), a protein that activates myosin during nonmuscle cell contraction.
The sensor has a Ca2+/calmodulin binding site placed between two
added fluorescent domains. When MLCK is activated by the binding of Ca2+/calmodulin,
FRET between the two fluorescent domains is disrupted.
In contracting cells, MLCK is recruited
to and activated along contracting stress fibers (also visible in a second
cell). MLCK is also activated in the lamella of motile and stationary cells,
and at the cleavage furrow during cytokinesis.
MLCK
stimulates rapid contraction; Rho kinase stimulates sustained contraction
Katoh
et al. report that the calcium-dependent myosin light chain kinase (MLCK)
triggers rapid stress fiber contraction, whereas Rho-kinase elicits sustained
contraction, which is necessary for maintaining stress
fibers, focal adhesions, and cytoplasmic tension. The authors separate
the effects of these two contractile systems by preparing contractile fibers
either in glycerol (which
maintains both contractile systems) or Triton
X-100 (which removes the Rho-kinase system).
Microtubules
Tea1p rides on microtubule ends
Feierbach et al.
confirm that Tea1p moves
to the ends of fission yeast cells on microtubules.
They find that Tea1p does so by attaching to the ends
of microtubules, which deposit
it directly
at the cell surface. Tea1p is probably held on microtubule ends by Tip1p, as
in cells lacking Tip1p the Tea1p wanders along
microtubules.
Microtubule
catastrophe under pressure
Janson et al. report
that force stalls MT growth and induces rapid catastrophes visible as single and repeated
events. Detailed measurements suggest the barrier acts simply by slowing tubulin
addition, thus giving more time for structural changes leading to catastrophe.
This behavior may make microtubules a more
adaptable positioning device.
Dynactin and microtubules search out their organelle targets
Vaughan
et al. visualize the p150Glued subunit of dynactin (a binding
partner for cytoplasmic dynein). They find that it associates with growing
microtubule plus-ends to form "comet tails." When these tails encounter
Golgi-derived membranes, the membranes are seen to initiate rapid
movement (see the same phenomenon in close-up). This supports the search-and-capture model in which microtubules probe the
cytoplasm for organelles in need of transport. A similar mechanism may underlie
microtubule-kinetochore interactions.
Tea1p
travels to cell ends and keeps polarity factors anchored there
Tea1p is needed to keep fission
yeast growing linearly; in its absence cells become bent and branched. Behrens
and Nurse demonstrate that tea1p is transported on the plus ends of microtubules from the vicinity of the nucleus to the cell
ends. Tea1p prevents the curling of microtubules around the cell ends, and helps
retain polarity factors at the cell ends.
Cytoplasmic
microtubules position the nucleus
Cytoplasmic
microtubules in fission yeast run from one end of the cell to another.
Tran et al. suggest that these
microtubules position the nucleus by attaching to the nucleus and then pushing
on the ends of the cell. They first show that the nucleus undergoes microtubule-dependent
deformations, and then
that these deformations correlate with growing microtubules pushing (see also a second video) against the ends of the cell.
Cortex-microtubule
interactions position the nucleus and spindle
Adames
and Cooper find that the budding yeast nucleus moves
to the mother-bud neck via capture of microtubule ends at one cortical
region at the incipient bud site or bud tip, followed by microtubule depolymerization.
Subsequent spindle movement into
the neck is mediated by microtubule sliding along the bud cortex, which
can sometimes be seen to occur with
free microtubules. In a later paper, Heil-Chapdelaine
et al. found that the cortical protein Num1p provides an essential attachment
point for the sliding machinery (see the color version, or grayscale
version with full legend).
Kinesin
KIFC3 and dynein
cooperate in Golgi positioning
Golgi membranes disperse in cells
treated with nocodazole, but then recover their perinuclear distribution as microtubules regrow. In cells lacking the
kinesin KIFC3, however, Xu et al. find that Golgi membranes remain scattered. Similar results are evident when the Golgi is initially dispersed by BFA
treatment of either wild-type
or kifC3-/- cells. The effect on kifC3-/- cells is only present when
dynein is inhibited by cholesterol depletion or dynamitin expression, suggesting
that dynein and KIFC3 cooperate in Golgi positioning.
Kinesin's neck linker
drives processive movement
Conventional kinesin is a dimeric
motor that moves along microtubules. When Tomishige
and Vale restrained motion of kinesin's neck linker via an oxidative crosslink,
the movement of kinesin was
restricted to brief one-dimensional diffusion events. This indicates that conformational
changes in the neck linker, not in the neck coiled-coil, drive processive movement
by the kinesin motor.
Determining direction of movement
Rafts to the front and back
Gómez-Moutón
et al.
find that lipid rafts that include PI3 kinase localize
to the front of moving cells; other rafts move to the back.
Disruption of raft structure by extraction of cholesterol prevents
both PI3 kinase localization and cell movement.
Cdc42 focuses the
direction of movement
After inhibition of Cdc42 function by Srinivasan
et al., cells fail to orient their movement correctly and
instead form pseudopods in an irrelevant
orientation.
Cells can move by either destroying or squeezing through the matrix
Wolf et al. find
that cells can use either of two methods to move: they can create
a path by proteolytically destroying extracellular matrix, or they can resort
to amoeboid
movement using propulsive
squeezing through gaps in the extracellular matrix. Whereas proteolytic movement remodels the
surrounding matrix via traction and bundling, during amoeboid movement the matrix
is not affected.
During amoeboid movement the cells are forced to squeeze
through small holes in the matrix.
Electrotaxis
uses a unique signal transduction pathway
Wound healing may be enhanced by the
presence of electrical fields, which are induced by spatial and temporal variations
in epithelial transport or electrical resistance. Similar fields have been shown
to affect the migration of many cell types. As a model for such migration events, Zhao
et al. examine the ability of Dictyostelium amoebae to undergo electrotaxis.
They find that the early stages of signal reception and transduction are not
shared between electrotaxis and the well-characterized, G-protein-dependent process
of chemotaxis. Inositol phospholipids, which are concentrated at the leading
edge of cells during chemotaxis, did
not show an intracellular gradient during electrotaxis, and G-protein-coupled
receptors were not redistributed. But the respective signaling strategies
must converge somewhere upstream of directed actin polymerization, as an actin
binding protein shows polarized localization during electrotaxis.
RhoA
helps pick up tails of moving cells
Worthylake
et al. look at monocyte movement and
find that monocytes lacking RhoA activity have trouble
picking up their tails.
Making and remodeling
adhesions
Splitting podosomes to make new adhesive contacts
Podosomes are actin-rich adhesions in macrophages that have many similarities
to the focal complexes used for leading-edge adhesion in other cell types. Evans
et al. track
the formation
and turnover of podosomes
at the leading edge of a moving macrophage. Fission and fusion events
can both be seen; the fission may be explained by branching
polymerization by actin. Both de novo assembly and fission events contribute
to new podosome assembly at the advancing front of the cell.
Focal
complexes behave differently at the front and back of the cell
Focal contacts at the front and rear
of a migrating cell must behave differently so that cells can grab on at the
front and let go at the back. Ballestrem
et al. examine this behavior using a labeled integrin subunit. At the rear
of the cell, RhoA induces high-density focal complexes that slide
inwards. Integrin turnover is fast, possibly allowing polarized renewal and
thus movement of the focal contacts. In contrast, the low-density focal complexes
in lamellipodia induced by Rac1 are stationary
and transient.
HA
and CD44 work together to form adhesion-filled protrusions
Localized application of hyaluronic
acid (HA), a major carbohydrate component of the extracellular matrix found mostly
in skin, joints, and eyes, can promote the formation of
local adhesion-filled protrusions. As documented by Oliferenko
et al., this is dependent on the CD44 adhesion receptor and the small GTP-binding
protein Rac1. The resulting motility is used in processes from metastasis to
wound healing.
Actions of microtubules
Arg connects microtubules and actin to help protrusion
Cells with Arg ruffle
and protrude actively, with Arg concentrated
in these protrusions. Miller
et al.
show evidence that Arg connects microtubules to actin bundles, thus allowing
the microtubules to deliver protrusion-promoting proteins to the cell edges.
Microtubules target adhesions precisely
Krylshkina et al.
demonstrate that microtubules probe close
to the cell–substrate interface (microtubules close
to the substrate appear darker) and specifically target
cell–substrate
adhesions. A suggestion
of such targeting was seen by the same group in an earlier paper by Kaverina
et al., and more examples
of targeting can be seen here
and here and here.
The microtubules may be guided to the adhesion sites by actin bundles, and may
bring molecules that regulate turnover of the adhesions.
Microtubules make a persistent push to the front
Wittmann et al. find
that the few microtubules that make it into protrusions at the front of the
cell grow
more persistently than do other microtubules. Many more such “pioneer” microtubules
are evident in cells
with constitutively active Rac,
even though Rac is normally thought of as a regulator of actin not microtubule
dynamics. Pioneer microtubules are also swept backwards by actin
retrograde flow.
Released microtubules dictate the direction of cell movement
Abal et al. find
that short microtubules labeled at their growing ends with EB1-GFP track
outwards towards the periphery.
When the microtubules are instead anchored at the centrosome by overexpressed
ninein, cells can still form polarized ruffles but they no
longer move,
suggesting that released microtubules help cells to move in a single direction.
Kinesin
delivers a signal inhibiting adhesion sites
Microtubules target
substrate adhesions and thus promote their disassembly. Krylyshkina
et al. investigate which motor might deliver this signal. They first inhibit
dynein, as evidenced by the dispersion
of lysosomes, but this has no
effect on adhesion site dynamics. Inhibition of kinesin, however, induced
a dramatic increase in the
size and reduction in number of substrate adhesions, mimicking the effect observed
after microtubule disruption by nocodazole. Microtubules still
target substrate adhesions. So conventional kinesin must be required only
for the focal delivery of a component(s) that retards the growth or promotes
the disassembly of adhesion sites.
APC
is deposited from microtubules at the cell surface to promote outgrowth
Mimori-Kiyosue
et al. report that the APC protein moves
along and concentrates at
the ends of microtubules. Shrinking microtubules deposit their
load of APC near the cell
surface, where the APC (shown in green) may promote cell outgrowth or migration.
Microtubules
deliver relaxing signals to contact sites
Kaverina
et al. suggest that microtubules deliver localized doses of relaxing signals
to contact sites to retard or reverse their development. Further, they propose
that it is via this route that microtubules exert their well-established control
on cell polarity. In spreading cells, for example, contacts turn over readily
(arrow in left panel of video
1 but persist and enlarge when microtubules are depolymerized (arrow in right
panel of video
1). Decreases in contact size occur thanks to direct
targeting of microtubules to these sites, a phenomenon that can also be seen
when microtubules regrow after
nocodazole treatment. The microtubules may deliver factors that inhibit contractility.
Applying inhibitors of
contractility causes shrinkage of the contacts, loss of microtubules, and retraction
of the cell.
Leukocyte transmigration
A docking structure
for transmigrating leukocytes
During inflammation, leukocytes
must exit the blood and enter the tissues via a process called transmigration.
Barreiro et al. examine this
process and the importance of ezrin, radixin, and moesin (ERM), which connect
membrane adhesion receptors to the actin-based cytoskeleton. Moesin is clustered around a lymphoblast during both adhesion to the endothelium
and transmigration across it, whereas VCAM-1 is clustered only during apical adherence, and I-CAM-1 is clustered during transmigration. These proteins provide the basis for an
endothelial docking structure that plays a key role in the firm adhesion of
leukocytes to the endothelium during inflammation.
Tether
durations needed for leukocyte attachment and rolling
When a tissue is inflamed, leukocytes
use L-selectin to attach to endothelial cells lining the blood vessels. The
initial capture is followed by leukocyte rolling along the endothelium. Dwir
et al. find that individual L-selectin tethers must persist for at least
20 ms to support continued rolling adhesion. The probability of successive tether
formation apparently collapses when tether duration drops below this critical
period. Failure in tethering can be seen,
for example, when the tail domain of L-selectin is truncated by the 358stop
mutation, which abolishes L-selectin's attachment to α-actinin
and possibly other cytoskeletal proteins.
NCAM traps machinery at new
synapses
For a contact between axon and dendrite
to turn into a synapse, a great deal of machinery must be recruited. Sytnyk
et al. find evidence that the neural cell adhesion molecule (NCAM) does
this recruiting by anchoring intracellular organelles in nascent synapses. Clusters
of NCAM, linked via spectrin to TGN organelles, translocate along neurites until
the complex is immobilized
via contact with NCAM in another neurite.
Actin
and microtubule movements are coupled in growth cones
Actin and microtubules interact in growth
cones, according to Schaefer et al.
They make growth cone movies of both microtubules
(MTs), and actin (with the first
actin movie and second actin
movie showing entire growth cones, and the third
actin movie showing a close-up of the peripheral domain and transition zone).
MTs enter the periphery by
polymerization and are cleared by a combination of catastrophe and coupling to
actin retrograde flow. The combination of MT polymerization and retrograde flow
can also lead to treadmilling.
Single MTs in the peripheral domain align
predominantly along, or very near, filopodial F-actin bundles (also visible
in close-up, and in another
example that includes both tracking on actin and a catastrophe event). Retrograde
flow can also lead to frequent MT bending,
buckling, and breakage in the transition zone. Finally, actin
arcs help bundle MTs
into the central domain (see also a dual
color movie). Schaefer et al. propose that the steady-state movement of F-actin
and microtubules in the filopodia allows the system to adapt quickly: a slight
decrease in retrograde F-actin flow, for example, would drive rapid microtubule
advance along the filopodia.
Growth cones grab sites that can withstand tension
Like a rock climber, a neuronal
growth cone senses its substrate to identify the route that will provide the
best grip. Suter and Forscher use beads coated with the Aplysia growth cone adhesion molecule apCAM
to show that the growth cone steers across the surface of a bead, but only if the bead is physically restrained.
The tension from apCAM binding to the restrained bead leads to Src family tyrosine
kinase activation, which then promotes the strengthening of apCAM-actin linkages.
The stronger linkages further increase tension, until the apCAM-actin linkage
is strong enough to guide growth cone extension. This process should drive growth
cone migration along the path providing the best molecular grip.
PKC increases neurite growth by increasing microtubule growth
Normal
microtubule dynamics at a neuronal growth cone are changed
after activation of protein kinase C (PKC), as demonstrated by Kabir
et al. PKC activation increases neurite outgrowth by increasing the lifetime
of microtubule growth episodes two-fold, increasing rescue frequencies 1.7-fold,
and decreasing catastrophe frequencies
two-fold.
TI-VAMP is needed for neurite outgrowth
Martinez-Arca
et al. are able to visualize
vesicles containing tetanus neurotoxin-insensitive vesicle-associated membrane
protein (TI-VAMP) and show that this protein, probably via its fusion activity,
is needed for neurite outgrowth.
Adhesion is necessary for elongation
The elongation
of the worm body during development involves a concerted contraction of the epidermis,
so when Petitt et
al. mutated
the cadherin-catenin system that helps hold cells together the result was a failure
in elongation.
An adhesion protein for slime molds
Fey et al. identify SadA, the first adhesion protein known in the social slime mold Dictyostelium.
Cells that lack the protein have problems – they
move more erratically and sometimes their division is delayed or aberrant compared
to wildtype
cells.
But they still stream
normally
after being starved (wildtype streaming is shown here).
Intermediate filaments help cells to hold together
Cells lacking the attachment between cytoplasmic intermediate filaments and
the plasma membrane can be pulled apart more
readily
than can control
cells. Huen et al.
therefore conclude that intermediate filaments contribute to the strength of
intercellular attachments.
Myo-endothelial progenitors isolated from muscle
Tamaki
et al. examine cells from the interstitial spaces of murine skeletal muscle,
and isolate a population of these cells that can form both endothelial and myogenic
cells. The latter fate is clearly evident in vitro, as the progenitors give
rise to cells that can undergo spontaneous contractions. This is seen in both
round cells and myotubes. The myo-endothelial progenitors may contribute to postnatal skeletal muscle
growth.
Skeletal muscle cells can couple with heart muscle cells
Heinecke
et al. find that skeletal muscle cells (elongated myotubes) can become
electrically coupled to heart muscle cells, which results in synchronous
beating directed by the heart muscle cells. This suggests that skeletal
muscle cells might become productively synchronized with heart muscle after
being used to repair tissue damaged by a heart attack.
Release of platelets from megakaryocytes
Megakaryocytes release platelets from long tube-like extensions called pro-platelets. Italiano
et al. show that platelets are liberated from the ends of proplatelets,
but that the number of ends is also increased by bending
and branching.
Frog Dishevelled is transported dorsally to specify dorsal fate
Frog embryos specify their dorsal
axis during the first cell cycle by transporting a determinant along microtubules
from the vegetal pole to the prospective dorsal side. Miller
et al. demonstrate that Dishevelled (Dsh) fulfills the criteria for such
a determinant. The directional transport of vesicle-like organelles coated with
GFP-marked Dsh can be visualized
but is disrupted
by treatments that disrupt dorsal specification. Once at the dorsal side, Dsh
activates the Wnt signaling pathway to specify dorsal cell fates.
Execution
proteases
Caspase cleavage of GRASP65 helps fragment the Golgi during apoptosis
Lane
et al. find that during apoptosis the Golgi ribbon is fragmented into dispersed clusters of tubulo-vesicular membranes. Fragmentation is caspase
dependent and GRASP65 (Golgi reassembly and stacking protein of 65 kD) is a
caspase-3 substrate. Expression of a caspase-resistant form of GRASP65 partially
preserved cisternal stacking and inhibited breakdown of the Golgi ribbon
in apoptotic cells.
Cathepsin B can execute tumor cells
Apoptosis in a cancer cell line is shown by Foghsgaard
et al. to be minimally dependent on proteases called caspases -- the
usual executioners in apoptosis. Instead, a lysosomal protease called cathepsin
B takes on this execution function, perhaps after being translocated to the
cytoplasm. Many tumor cells upregulate cathepsin B because of its invasion-promoting
properties, but this may come at a cost for the tumor cells, as the excess cathepsin
B apparently makes the cells more sensitive to suicidal cell death.
Mitochondria
Apoptotic
cells show a transient loss in mitochondrial membrane potential
Waterhouse
et al. suggest that the transient
loss of mitochondrial membrane potential that can be seen in individual apoptotic
cells may be masked by cell-to-cell asynchrony when looking at a population
of cells.
Protein
degradation and stress responses |
|
Ubiquitination
Misfolded proteins are actively transported into aggresomes
Misfolded proteins are gathered
into structures called aggresomes, and García-Mata
et al. find that the accumulation occurs via directed
transport to the aggresome along microtubules.
Autophagy
Apg5 helps form autophagosomes
Autophagosomes form
as cup-shaped organelles that engulf large parts of the cytoplasm. As shown
by Mizushima et al., Apg5, part
of a ubiquitin-like conjugation system, localizes to the forming autophagosomes,
and is essential for their formation.
Stress
granules
Stress granule assembly is dynamic
Stressed cells inhibit translation
initiation, and the resultant free mRNAs accumulate
in stress granules (SGs). These can disassemble
to yield polysomes once the stressor (in this case, arsenite) is removed. Kedersha
et al. find that labeled SG-associated proteins rapidly and continuously
shuttle in and out of SGs, indicating that the assembly of SGs is a highly dynamic
process. Thus mammalian SGs may be sites at which untranslated mRNAs are sorted
and processed for either reinitiation, degradation, or packaging into stable
nonpolysomal mRNP complexes.
Mitochondria slow down for calcium
Mitochondria move
along microtubules,
but Yi et al.
find that they slow
down
when calcium levels rise, probably so that they can help buffer the ion back
to normal levels.
Activated Ras defines the front of chemotaxing cells
Sasaki et al.
show that localized
Ras activation at the front of slime mold cells allows successful chemotaxis
to a point source.
TCR signaling without a synapse
Bunnell
et al. dispute the idea
that signaling through the T cell antigen receptor is dependent on assembly
of a large, complicated structure called the immune synapse. They find that
the signaling component ZAP-70 is rapidly
recruited into small contacts that can almost immediately initiate
calcium influxes. The ZAP-70 is initially
but not eventually subject
to rapid exchange.
Stress induces oscillations
Msn2 and Msn4 are two yeast proteins that move into the nucleus in response
to severe stress; once there they help the cell to respond to stress. Jacquet
et al. find that under
intermediate stress conditions the two proteins instead oscillate
into and out of the nucleus. Perhaps this oscillation prevents unnecessarily
intense activation but keeps the cell primed for a full response should conditions
worsen.
Dynein delivers HIV to the nucleus
McDonald et al. find that HIV
uses dynein-dependent movement
on microtubules, possibly to deliver viral genomes to the nucleus.
Listeria
use cellular engulfment machinery to invade neighboring cells
Listeria monocytogenes bacteria
move
around and between cells by co-opting the cell's actin polymerization machinery.
Robbins et al. suggest that the
bacteria exploit the cells' normal mechanism for engulfing neighboring cell
surface fragments to achieve spread between cells. When the bacterial cells
encounter a junction between two mammalian cells, they can be seen
either protruding (red P) into the neighboring cell or ricocheting (yellow R)
off of it.
A
dynamic Golgi is built from the ER
Brefeldin A is known to inhibit
ER to Golgi transport, and thus to lead to Golgi disappearance through the continuation
of Golgi to ER transport. Ward et
al. use this treatment to observe the displacement of a variety of Golgi
proteins, including those thought to be stable Golgi components. This furthers
their view of the Golgi as a dynamic, rather than stable structure. The abrupt
appearance of some of the proteins in punctate structures suggests that
they do not track out from the Golgi, but are reassembled at ER exit sites.
Ward et al. suggest that these exit sites are the building blocks for the Golgi
apparatus.
Parallel fibropositors make for strong tendons
Canty et al.
map out long, thin plasma-membrane extensions that they call fibropositors.
These structures extend between adjacent cells and deposit collagen fibers in
a parallel arrangement that confers strength to tendons.
Calcium
inhibition of AC6 allows formation of gaps between endothelial cells
The barrier between blood and tissues,
formed by endothelial cells, can be breached by inflammatory mediators. They
elevate cytosolic calcium ([Ca2+]i) in the endothelial cells, thus
increasing actomyosin tension and decreasing adhesions. This results in the
generation of focal intercellular
gaps that form a paracellular pathway to promote fluid, solute, and protein
permeability. These changes are only possible, say Cioffi
et al., because the [Ca2+]i also inhibits type 6 adenyl cyclase
(AC6) in endothelial cells. When the authors boost cAMP levels by introducing
the calcium-stimulated AC8 this prevents
gap formation.
Matrix meshes can be strapped together with small-scale forces
Fibroblast explants in collagen
gels, like cells in tissues, can mold the surrounding matrix. They exert mechanical
forces that lead to the formation of ligament-like
straps between the explants. Somehow, the micrometer-scale
cellular traction forces produced by the explants end up generating millimeter-scale
structural changes in the matrix. Sawhney
and Howard studied this process with fiduciary beads and a computer algorithm.
They found that the collagen forms a mesh of interconnected fibers. Small movements
along one axis of this collagen mesh (the axis running between explants) generate
a large movement perpendicular to the axis, which draws collagen into the strap.
