1676rr Page 994 Wednesday, December 15, 2004 11:35 AM
Published December 13, 2004
Research Roundup
D y n a m i n i n fi s s i o n a n d f u s i o n
he same GTPase that tears membranes apart is needed to put them back
together, based on results from Christopher Peters, Andreas Mayer, and
colleagues (Université de Lausanne, Switzerland). The dual regulation may
ensure that only one of the two competing processes goes on at any one time.
The Swiss group finds that yeast cells lacking the dynamin homologue
Vps1p resemble both fission and fusion mutants. The effects were seen in the
morphology of the vacuoles, where Vps1p was localized. Some cells had a
single enlarged vacuole, whereas others had many small vacuolar fragments.
Without dynamin, yeast vacuoles have both fission
and fusion failures.
The two phenotypes were seen because Vps1p functions in both
pathways. As expected given dynamin’s known fission activity, vacuole
fragmentation in response to salt stress was disrupted in the mutant. But vacuole fusion reactions
also required Vps1p, which was found to interact with the Vam3p t-SNARE.
Vps1p recruited the t-SNARE into large membrane complexes containing multiple copies of both
proteins. This organization favors fragmentation, since dynamin polymers are the fission-active form. At
the same time, fusion was inhibited until Vps1p was released from the vacuole membrane, suggesting
that the Vps1p-bound t-SNARE is inactive. The release of Vps1p was controlled by the t-SNARE and its
ATPase chaperone, NSF. How NSF is regulated to support or limit fission remains to be addressed.
The linking of dynamin with the fusion machinery may prevent repeated futile cycles of fission
and fusion. “Intuitively, [this counter-regulation] makes sense,” says Mayer. “When a vesicle is
pinched off, SNAREs must be incorporated to make it fusogenic for future reactions. So why are they
not active while pinching is going on?” Silencing the t-SNAREs at the site of fission by dynamin’s
intervention is one way to solve this problem.
As vps1 mutants were deficient in vacuole fusion, dynamin must also somehow promote fusion,
perhaps by organizing cooperative t-SNARE complexes. Dynamin release from the vacuole might also
induce necessary conformational changes in the t-SNARE.
MAYER/ELSEVIER
T
Downloaded from on June 17, 2017
Reference: Peters, C., et al. 2004. Cell. 119:667–678.
Opening closed pores
A
994
JCB • VOLUME 167 • NUMBER 6 • 2004
be completely open to allow diffusion.”
But diffusion is just what his group
noted during Aspergillus mitosis. The
nuclear pores of this fungus were opened
by the dispersal of FG-repeat containing
nucleoporins (FG-Nups), which normally
form the diffusive barrier to the nucleus.
Core structural proteins, however, were
retained at the pore.
The dispersal of FG-Nups was accompanied by their phosphorylation and
required the mitotic kinases NIMA and
Cdk1. NIMA moved to the nuclear periphery at mitosis and may alter FG–Nup
interactions via direct phosphorylation.
Many proteins that are restricted to
the nucleus or cytoplasm during interphase were found in both compartments
during mitosis, including RanGAP. The
likely resulting loss of the RanGTP gradient is expected to impair regulated
transport further. Since the gradient must
be rebuilt to reestablish active transport
after mitosis, Osmani speculates that the
nuclear envelope may break down transiently and reform around the daughter
masses of DNA, where the RanGEF is
concentrated. No one has seen such an
event as yet, as the envelope is difficult to
visualize by this stage of division.
Reference: De Souza, C.P.C., et al. 2004.
Curr. Biol. 14:1973–1984.
OSMANI/ELSEVIER
closed mitosis may be less
closed than was thought, based
on results from Colin De Souza,
Stephen Osmani, and colleagues (The
Ohio State University, Columbus, OH).
The authors find that a fungus opens
nuclear pores during mitosis to permit
diffusion into and out of the nucleus.
Simple organisms may thus be viable
model systems for the study of nuclear
changes during mammalian mitosis.
As the nuclear envelope is not broken
down during a closed mitosis, the cell
must regulate the nuclear entry of mitotic
kinases and tubulin. This was assumed
to occur through cell cycle regulated
alterations of specific transport pathways. “I, and probably most everyone,”
says Osmani, “almost took for granted
that subtle changes in the transport properties of the pore alter import pathways
slightly. Probably in [budding yeast] this
is the case. But no one thought it could
An FG–Nup (green) is at the nucleus during
interphase (i) but is dispersed at prophase (p)
and telophase (t).
1676rr Page 995 Wednesday, December 15, 2004 11:35 AM
Published December 13, 2004
Text by Nicole LeBrasseur
[email protected]
I
Reference: Wessells, R.J., et al. 2004. Nat. Genet.
36:1275–1281.
Ti m e k e e p e r s a b o u n d
very cell in the body has its own little clock, according to Emi
Nagoshi, Ueli Schibler (University of Geneva, Switzerland), and
colleagues. These peripheral clocks may be closely related to the
body’s central clock—the suprachiasmatic nucleus (SCN).
Neurons of the SCN maintain circadian gene expression
for long periods of time.
Other cells, however, were
thought to rely on the SCN to
maintain their oscillations, as
the amplitude of gene expression oscillations in fibroblast
cultures lessens rapidly with
time in culture. But Schibler’s
Oscillations can be seen in gene expression
group shows that the exampatterns of individual fibroblasts.
ination of mass cultures masks
the continued ability of each cell to maintain strong oscillations.
By looking at individual fibroblasts, the authors were able to see
persistent oscillations. Even after mitosis, daughter cells continued the rhythm
of the mother, although with a slight phase shift, probably due to the
transient suspension of transcription. Mitosis itself is also gated by the
oscillations, although the cellular advantage to this gating is unclear.
Mathematical modeling suggested that global culture oscillations
are diluted over time by slight differences in the periodicities of each
cell’s clock. Resynchronization was achieved in cultured cells by activation of a wide variety of signaling pathways, which reset all cells to
a common point. In animals, peripheral clocks are synchronized by
feeding cycles, which depend on sleep–wake cycles. As the SCN controls
sleep cycles, it automatically synchronizes the oscillations in peripheral
cell types.
E
Downloaded from on June 17, 2017
nsulin signaling, which is required to keep
blood sugar levels steady, may be deleterious
for heart cells as they age, according to
Robert Wessells, Rolf Bodmer (The Burnham
Institute, La Jolla, CA), and colleagues. When
insulin signaling is compromised, however, the
old flies remain young at heart.
Animals with less insulin live longer, but
the relevant effects on individual organs had not
been assessed. The new results show that insulin
is deleterious to fly heart function. As flies age,
their hearts beat more slowly and are more prone
to failure under stress. But this deterioration
was delayed in mutant flies that did not perceive
as much insulin.
Insulin’s effects on the heart are not mediated by systemic actions downstream of insulin.
Heart-specific reduction in insulin signaling
also improved heart performance in elderly
flies. Activation of cardiac insulin signaling, in
contrast, made young flies more prone to stressinduced heart failure.
Heart-specific improvements did not extend
the fly’s lifespan, but heart activity is much
more critical to mammalian survival. If the
effects of insulin and insulin-like growth factors are similar in humans, Bodmer dreams of
“organ-specific interventions to improve the
quality of old age.”
SCHIBLER/ELSEVIER
Insulin breaks hearts
Reference: Nagoshi, E., et al. 2004. Cell. 119:693–705.
Checking in on spindle size
nly replicated chromosomes should journey down
mitotic spindles. The premature segregation of unreplicated DNA is prevented by the replication checkpoint,
which is widely believed to block entry into mitosis when
replication forks are stalled. But Vaidehi Krishnan, Uttam
Surana, and colleagues (Institute of Molecular and Cell Biology,
Singapore) show that the checkpoint has a more direct target—
it prevents spindle elongation.
During replication stalls, budding yeast cells build short
mitotic spindles. These spindles elongate in replication
checkpoint-defective rad53 or mec1 mutants, thus causing
untimely and uneven distribution of the DNA. Surana’s group
shows that this elongation occurs in the absence of most of
the hallmarks of mitosis, including APC activation, cohesin
cleavage, and biorientation.
“At least in early S phase, when cohesion and biorientation have not been established, cells are not looking ahead
[to mitosis],” says Surana. “They are just solving a local
problem of preserving nuclear integrity by preventing spindle
elongation.” Rad53 and Mec1 kinases accomplish this by
SURANA/ELSEVIER
O
Cin8 expression causes DNA segregation (middle) and spindle
elongation (right) during replication stalls.
down-regulating the microtubule-binding proteins Cin8 and
Stu2, which elongate spindles. The microtubule destabilizing
motor Kip3 also helped by restricting spindle elongation during synthesis delays.
Cin8 overexpression forced premature nuclear division
in checkpoint-competent cells. Conversely, overexpression of
Rad53 in cells whose DNA replicated normally resulted in
abnormal and fragmented spindles.
Reference: Krishnan, V., et al. 2004. Mol. Cell. 16:687–700.
RESEARCH ROUNDUP • THE JOURNAL OF CELL BIOLOGY
995