| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departments of Anatomy and Cell Biology and Pathology, Columbia University, New York, New York 10032, USA
1Correspondence: Department of Anatomy and Cell Biology, Columbia University, 630 West 168th Street, New York, NY 10032, USA. E-mail: ggg1{at}columbia.edu
 |
INTRODUCTION |
|---|
 TOP INTRODUCTION REFERENCES |
|---|
, endoplasmic reticulum (ER)
(2)
, lysosomes and endosomes (3)
, Golgi
(4)
, and vimentin intermediate filaments (IFs)
(5)
. MTs have been shown to be dynamic structures turning
over with a half-life of 5–10 min in proliferating cells in culture
(6
, 7)
. This raises the question of how such dynamic
structures as MTs can contribute to cellular organization, which would
appear to require at least some structural stability. Kirschner and
Mitchison (8)
hypothesized that MT dynamics may allow MTs
to efficiently search the 3-dimensional space of the cell and, in
response to environmental cues, dynamic MTs may become locally
stabilized. Indeed, in polarized and differentiating cells, a subset of
MTs are much more stable than the bulk, dynamic MTs, and, consistent
with a role for MTs in organizing cellular organelles, the stable MTs
are usually oriented in the axis of polarization or differentiation
(reviewed in ref 9
).
One of the questions we have been addressing in our laboratory is how
cells distinguish stabilized MTs from dynamic MTs. One clue to this
question is that when MTs are stabilized in vivo, the
tubulin comprising the MTs is modified by one or more
post-translational modifications (reviewed in ref 9
). One
of the best characterized of these modifications is detyrosination,
which involves the removal of the carboxyl-terminal tyrosine residue
from 
-tubulin by tubulin carboxypeptidase (10)
, which
preferentially utilizes polymeric tubulin as a substrate
(11)
. Thus, only stabilized MTs that have persisted long
enough have a high level of detyrosinated tubulin, and detyrosination
thus serves as a marker for stable MTs. Detyrosination generates
-tubulin with a glutamate residue at the carboxyl terminus. (The two
forms of tubulin are called Tyr and Glu tubulin after their respective
carboxyl termini.) Detyrosination is reversible and Glu tubulin can be
retyrosinated by tubulin tyrosine ligase (12)
. This enzyme
is active only on soluble tubulin and is responsible for keeping the
monomer pool of tubulin completely tyrosinated in vivo
(13)
. As a result of the properties of tubulin
carboxypeptidase and tubulin tyrosine ligase, and the dynamics of
cellular MTs, cells have a mixture of dynamic MTs enriched in Tyr
tubulin (Tyr MTs), stable MTs enriched in Glu tubulin (Glu MTs), and
subunit tubulin comprised of only Tyr tubulin. We can distinguish the
two forms of MTs by using antibodies that specifically react with Glu
and Tyr tubulin (14)
.
The biological system we use in most of our studies is an in
vitro wound healing model. Fibroblasts are grown to confluency,
and narrow strips of cells are then scraped off the substratum. The
cells at the wound edge are initially unpolarized, but, in response to
wounding, they become polarized with their long axis perpendicular to
the wound and with an active leading edge at the portion of the cell
facing the cell-free area. As a result of this polarization, the cells
are able to move directionally into the wound. One response of
wound-edge cells to wounding is the stabilization of a subset of MTs
that are oriented toward the wound (15
, 16)
. The
stabilization of the MTs oriented toward the wound edge requires a
serum factor (17)
, which we subsequently identified as the
mitogenic lipid lysophosphatidic acid (LPA) (18)
. LPA or
serum is capable of generating stabilized MTs in as little as 5 min
after the addition to serum-starved cells (18)
. This
suggests that MTs are poised to respond rapidly to extracellular cues
and can function on a time scale that is consistent with many dynamic
cell behaviors. In our initial effort to identify intracellular factors
that mediate the LPA-induced stabilization of MTs, we found that
activated Rho (a Ras-related, small GTPase) is both necessary and
sufficient to mediate the selective stabilization of MTs at the wound
edge (18)
. Activated Rho did not alter the parameters of
the dynamic MTs but instead affected only whether MTs were in a dynamic
state. We propose that Rho regulates MT stability and is part of a
signaling cascade that locally recruits or activates MT stabilizing
factors in the leading edge of the cell. We are currently working on
identifying other factors in this signaling pathway, as well as the
factors mediating the selective stabilization of MTs.
Once MTs become stabilized, their tubulin subunits become
post-translationally modified. What is the consequence of this
post-translational modification? To answer this question, we have
examined cellular organelles to see if any were colocalized with Glu
MTs. We found that vimentin IFs are preferentially localized on stable
Glu MTs (19)
. This coalignment was determined by the
post-translational detyrosination of the tubulin comprising the Glu MTs
rather than the increased stability of Glu MTs, as shown by further
studies. First, the distribution of IFs was disrupted (‘collapsing’
from an extended array to a perinuclear aggregate) by microinjecting
cells with affinity-purified antibodies to Glu tubulin but not with
affinity-purified antibodies to Tyr tubulin (19)
. Second,
when monomeric Glu tubulin was generated in the cytoplasmic pool by
microinjecting cells with nonpolymerizable and nonretyrosinatable Glu
tubulin, it acted as a dominant negative inhibitor of the IF–Glu MT
interaction (i.e., vimentin IFs collapsed to the perinuclear area).
Microinjection of the same amount of nonpolymerizable Tyr tubulin into
wound-edge cells did not affect the distribution of vimentin IFs
(20)
. These results demonstrate directly that the
post-translational detyrosination of tubulin itself is critical for
maintaining an extended distribution of IFs on MTs and suggests that
Glu tubulin is the preferred site for IF–MT interaction. We have also
learned that vimentin IFs did not stabilize Glu MTs, because collapse
of IFs to a perinuclear location by microinjection of antibodies to
vimentin IFs had no noticeable effect on the array of Glu MTs
(19)
. It is not clear what the functional consequence of
the IF–Glu MT association is. Nonetheless, it is clear that
stabilization and subsequent post-translational detyrosination of MTs
can polarize the intracellular distribution of IFs, and this may serve
as a paradigm to understanding other organelle–MT interactions.
What is the molecular mechanism that accounts for the preferential
interaction between IFs and Glu MTs? To identify candidate proteins
that selectively interact with Glu tubulin and mediate the IF–Glu MT
interaction, we tested whether antibodies to Glu or Tyr tubulin would
selectively block the binding of proteins to MTs. We incubated
taxol-stabilized MTs prepared from brain tubulin, an ~50:50 mixture
of Glu and Tyr tubulin, with MT-interacting proteins in the presence of
saturating levels of antibodies to Glu or Tyr tubulin. No significant
difference in the binding of brain microtubule associated proteins
(MAPs) to MTs was detected in the presence of either antibody.
However, the binding of fibroblast or brain conventional kinesin to MTs
was prevented more effectively by antibodies to Glu tubulin compared
with antibodies to Tyr tubulin, suggesting that kinesin preferentially
binds to Glu tubulin compared with Tyr tubulin (21)
.
Direct measurement of the binding affinity of conventional kinesin
heads to pure Glu and Tyr MTs confirmed that kinesin binds to Glu MTs
with an approximately threefold higher affinity than to Tyr MTs
(21)
. These results indicate that detyrosination of
tubulin can regulate the interaction of kinesin with MTs and that
conventional kinesin may be responsible for mediating the preferential
interaction between vimentin IFs and Glu MTs. Indeed, when we tested if
kinesin can associate with vimentin IFs, we found that conventional
kinesin specifically cosedimented with vimentin IFs without the need
for any accessory molecules (21)
. We also have preliminary
results suggesting that the kinesin light chain involved in interacting
with vimentin IFs is a specific one, because the kinesin light chain
that sedimented with vimentin IFs reacted with a kinesin light chain
antibody but not two other antibodies known to react with known
isoforms of kinesin light chain (21)
.
Microinjecting cells with nonpolymerizable Glu tubulin interfered with
the IF–MT interaction. To map the epitope on Glu tubulin that is
responsible for interfering with the IF–MT interaction, we have
microinjected fragments of tubulin into cells and tested if they are
capable of disrupting the distribution of IFs in vivo.
Microinjected 14 kDa carboxyl-terminal trypsin fragment of -Glu
tubulin (-C Glu) induced IF collapse whereas the
NH2-terminal trypsin fragment of -tubulin did
not alter the IF array. The epitope required more than the
detyrosination site at the carboxyl terminus, because a short peptide
(a 7-mer) mimicking the carboxyl terminus of Glu tubulin did not
disrupt the IF distribution (20)
. Remarkably, the same
reagents that disrupted the IF–Glu MT interaction in vivo
also inhibited the binding of kinesin to MTs in vitro
(20)
. These results further support the notion that
kinesin is responsible for mediating the IF–Glu MT interaction. That
kinesin is involved in MT–IF interaction was initially implied by the
study of Gyoeva and Gelfand (22)
, which showed that
microinjected kinesin antibody caused collapse of vimentin IFs.
Recently it was shown that the formation of vimentin IFs network in
spreading cells was inhibited by kinesin antibodies (23)
,
further supporting the role of kinesin in mediating IF–MT interaction.
Direct binding of kinesin to vimentin IFs in vitro (see ref
21
) and the observation that individual vimentin IF
fragments move toward the cell periphery at rates consistent with
kinesin-driven motility (24)
, suggest that vimentin IFs
may be a simple, nonmembraneous protein cargo for kinesin.
In addition to IFs, the distribution of other peripherally localized
organelles, such as the ER (2)
, mitochondria
(1)
, and lysosomes (3)
, are also dependent on
MTs. As MT-based, plus-end directed motors, kinesins have been
implicated in the transport and maintenance of the organization of the
ER (25
26
27)
, mitochondria and lysosomes (28
, 29)
, and membrane-bound vesicles (30
, 31)
. Based on
our results demonstrating the preferential interaction of conventional
kinesin with Glu MTs (ref 21
and unpublished observations)
and the competitive inhibition of kinesin binding by nonpolymerizable
Glu tubulin, but not nonpolymerizable Tyr tubulin (20)
, it
seems possible that Glu MTs may also be the preferred MT substrates in
other MT–organelle interactions. We are currently investigating
whether other members of the kinesin superfamily also react
differentially with Glu and Tyr MTs. It is possible that specific
kinesin light chains are involved in interaction with different
cellular organelles, because the kinesin light chain involved in
interacting with vimentin is likely specific (21)
, and the
association of a specific kinesin light chain with mitochondria was
recently reported (32)
.
Our data support a model in which post-translational modification of
tubulin plays a central role in generating cellular polarity
(Fig. 1
). This model hypothesizes that the rapid dynamics of MTs allows them to
explore the 3-dimensional space of the cell in search of signals from
the extracellular environment (‘MT exploration’). In response to
external signals, MTs may become selectively stabilized (‘MT
stabilization’). In our studies, we found that an extracellular
factor, LPA, presumably in conjunction with a wounding stimulus, works
by activating a Rho GTPase-dependent pathway to convert dynamic MTs to
stabilized MTs in portions of the cell adjacent to the extracellular
wounding stimulus. After stabilization, the tubulin comprising the MTs
is modified by post-translational modification (e.g., detyrosination
[‘MT modification’]). This step is notable in that it does not
alter the spatial information resident in the array of stabilized MTs
but translates the change in MT dynamics into a biochemical signal that
can be interpreted by other cellular constituents. Modified MTs may be
preferentially utilized in the interaction of other organelles with MTs
(‘MT interaction’). We have shown that one of the cellular organelles
capable of responding to tubulin detyrosination is vimentin IFs and
that the biochemical signal is interpreted by kinesin, which binds
better to detyrosinated tubulin compared with tyrosinated tubulin. In
Fig. 1
, a cell shape change is depicted as accompanying the
polarization of IFs on stable, modified MTs, although we have not shown
this and the cell shape polarization may occur earlier.
|
There may be important advantages in generating the various biochemical forms of tubulin that result from post-translational modification. By relying on detyrosination of tubulin to mediate the interaction of MTs with IFs (and possibly other organelles), the cell may prevent organelles from interfering with MT dynamics by reducing their interaction with Tyr tubulin monomer or Tyr MTs. Alternatively, by enhancing kinesin interaction with a form of tubulin (Glu tubulin) that is only generated in MTs, the cell may limit the possible interference of monomeric tubulin with motor activity. There are at least six other tubulin post-translational modifications, and it is possible that they may also regulate the interaction of MTs with organelles. Perhaps the diversity of modifications allows for specific regulation of different organelles with MTs—a ‘one modification, one organelle’ model. Alternatively, separate modifications may specify regulation of different cellular pathways or localizations.
|
REFERENCES |
|---|
|
TOPINTRODUCTIONREFERENCES |
|---|
-tubulin: a mechanism for subcellular differentiation of microtubules. J. Cell Biol. 105,251-264This article has been cited by other articles:
|
|
 |
|
  M. Al-Maghrebi, H. Brule, M. Padkina, C. Allen, W. M. Holmes, and Z. E. Zehner The 3' untranslated region of human vimentin mRNA interacts with protein complexes containing eEF-1{gamma} and HAX-1 Nucleic Acids Res., December 1, 2002; 30(23): 5017 - 5028. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Infante, M. Stein, Y Zhai, G. Borisy, and G. Gundersen Detyrosinated (Glu) microtubules are stabilized by an ATP-sensitive plus-end cap J. Cell Sci., January 11, 2000; 113(22): 3907 - 3919. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
