Content uploaded by David A Bobak
Author content
All content in this area was uploaded by David A Bobak on Jul 14, 2016
Content may be subject to copyright.
Inactivation of the small GTPase Rho disrupts cellular attachment and
induces adhesion-dependent and adhesion-independent apoptosis
David Bobak
1,2,3
, Jonathan Moorman
1
, Angelo Guanzon
3
, Linda Gilmer
1
and Chang Hahn
2,3
Departments of
1
Medicine,
2
Microbiology and
3
Beirne Carter Center for Immunology Research, University of Virginia School of
Medicine, Charlottesville, Virginia 22908, USA
Rho small GTPases regulate a variety of cellular
signaling pathways involved in cell growth and transfor-
mation. In this study, we examined potential roles for
Rho in adhesion-dependent and -independent pathways
regulating apoptosis. Rho GTPases are speci®cally
inactivated by exoenzyme C3 (C3) of Clostridium
botulinum. Using a novel Sindbis virus-based gene
expression system, we created a double subgenomic
recombinant (dsSIN:C3) capable of expressing active C3
in intact cells. Infection of L929 ®broblasts with
dsSIN:C3 caused essentially complete ADP-ribosylation
of intracellular Rho within 1 h. dsSIN:C3-infected cells
also became rounded within 1 ± 2 h and detached by 5 h
post-infection. Infection of L929 in suspension with
dsSIN:C3 disrupted the ability for normal cellular
attachment and spreading. Infection of primary cell
explants of chicken embryo ®broblasts (CEF) and rat
aortic smooth muscle cells (RSM) with dsSIN:C3 caused
cytoskeletal eects similar to those seen in L929. We
also observed that C3 markedly decreased the basal
phosphorylation state of focal adhesion kinase (FAK).
Most intriguingly, we found that dsSIN-based expression
of C3 or loss of function mutants of Rho could each
induce apoptosis and, in RSM, this eect was observed
to be adhesion-independent. Rho GTPases, therefore,
appear to regulate signal pathways that are required for
cell survival and growth that are separate from, but
likely overlap with, Rho-dependent pathways involved in
cellular adhesion.
Keywords: Rho; apoptosis; exoenzyme C3; adhesion
Introduction
Adhesion-dependent intracellular signals play critical
roles in cell growth, proliferation, and transformation.
Common to each of these cellular functions is the
requirement for stimulus-dependent re-organization of
the cytoskeleton. Moreover, severe disruption of
cytoskeletal organization, as can occur with the loss
of adhesion, can cause certain types of cells to undergo
programmed cell death, or apoptosis (Frisch and
Francis, 1994; Meredith et al., 1993; Re et al., 1994).
Many dierent signal transduction pathways are
involved in adhesion-dependent signaling and include
the generation of intracellular calcium transients,
formation of phospholipid metabolites, stimulation of
serine/threonine kinases, and activation of tyrosine
kinases (Clark and Brugge, 1995). Many of these
adhesion-dependent signaling pathways are now
known to be regulated by members of the Rho family
of small GTPases. Rho family GTPases have been
found to modulate many actin micro®lament-depen-
dent cytoskeletal functions, including formation of
focal adhesions, stress ®bers, and membrane ruing
(Tapon and Hall, 1997; Takai et al., 1995).
Rho proteins are members of the ras proto-oncogene
superfamily of small GTPases and are unique in that
they serve as speci®c intracellular substrates for certain
bacterial toxins and exoenzymes (Aktories et al., 1992;
Just et al., 1992a,b; Sugai et al., 1992a,b). Exoenzyme
C3 (C3) of Clostridium botulinum is the best described
of these bacterial products (Aktories, 1994). C3 ADP-
ribosylates Rho at Asn
14
, an amino acid residue located
within the putative eector domain of Rho (Sekine et
al., 1989). This covalent modi®cation functionally
inactivates Rho, likely by blocking the interaction of
Rho with its downstream eector molecules and/or
Rho-associated kinases (Lim et al., 1996; Tapon and
Hall, 1997). C3, therefore, has become a valuable
reagent for the study of Rho-dependent cellular
functions. C3-treatment of cells, for example, alters a
variety of cytoskeleton-dependent functions including
assembly of focal adhesions and stress ®ber networks
(Chardin et al., 1989; Ridley and Hall, 1992; Rubin et
al., 1988), lymphocyte aggregation (Tominaga et al.,
1993) and cytotoxicity (Lang et al., 1992), cell motility
(Ridley, 1996; Takaishi et al., 1995), neurite outgrowth
(Jalink et al., 1994), and formation of actin cleavage
furrows (Kishi et al., 1993). C3 has a serious limitation,
however, in that it is relatively cell-impermeant.
Investigators have traditionally used techniques such
as microinjection, permeabilization, or extremely
prolonged incubation times to achieve functionally
signi®cant intracellular levels of C3 (Aktories, 1994;
Narumiya and Morii, 1993). In order to investigate the
potential roles of Rho in apoptosis, we sought to
develop a system that could rapidly and eciently
produce high intracellular levels of C3 in a large
number of intact cells. To circumvent the problems
inherent with previous C3 delivery techniques, we
adapted a newly described Sindbis virus (SIN) based
transient expression system for our purpose (Hahn et
al., 1992a; Xiong et al., 1989). We generated a
recombinant double subgenomic Sindbis virus able to
express C3 (dsSIN:C3) in infected cells. Virtually 100%
of targeted cells were infected with dsSIN:C3, resulting
in rapid and synchronized intracellular expression of
C3 and essentially complete ADP-ribosylation of
endogenous Rho.
We found that infection of monolayer L929
®broblasts with dsSIN:C3 sequentially induced cellular
Correspondence: DA Bobak
Received 15 January 1997; revised 5 July 1997; accepted 6 July 1997
Oncogene (1997) 15, 2179 ± 2189
1997 Stockton Press All rights reserved 0950 ± 9232/97 $12.00
retraction, rounding, and detachment. Further, infec-
tion of suspension preparations of L929 cells with
dsSIN:C3 blocked the subsequent ability of these cells
to attach and spread normally. These eects of
dsSIN:C3 were not limited to transformed cell lines.
Infection with dsSIN:C3 also disrupted adhesion in
primary explants of chicken embryo ®broblasts (CEF)
and rat aortic smooth muscle cells (RSM). Addition-
ally, we observed that the disruption of Rho by C3
caused a marked decrease in the basal phosphorylation
state of focal adhesion kinase (FAK). Finally, we
determined that the dsSIN:C3-mediated inactivation of
Rho was a potent activator of apoptosis in CEF and
RSM, but not in the L929 ®broblast cell line. Most
intriguingly, we found that the Rho-dependent
activation of apoptosis also occurred in suspension
preparations of RSM, thereby identifying adhesion-
dependent and adhesion-independent roles for Rho in
signal pathways leading to apoptosis in adherent cells.
Results
We found that the infection of L929 with dsSIN:C3
produced rapid and ecient intracellular expression of
C3, occurring as early as 1 h post-infection (Figure 1b).
In addition to immunoblotting, we were also able to
document C3 expression by immunoprecipitation of C3
from extracts of
35
S-methionine-labeled L929 cells
infected with dsSIN:C3 (data not shown). Similar
levels of expression of functional C3 were observed in
many other cell types, including BHK-21, CEF, RSM,
P815 (murine mastocytoma) and EL-4 (murine T
lymphoma) (Moorman et al., 1996 and data not
shown).
Initially, we sought to determine the eects of
dsSIN:C3 infection on the morphology of monolayer
®broblasts (L929). Infection of L929 with dsSIN:C3,
but not parental virus or dsSIN:CAT, produced
dramatic morphological changes. Within 2 h of
dsSIN:C3 infection, we observed cellular retraction,
formation of neurite-like extensions and cell rounding
in essentially 100% of infected cells (Figure 2a ± c).
Other investigators have reported that microinjection
of C3 causes a generally similar type of cellular
retraction within 15 ± 60 min (Hall, 1994; Takai et al.,
1995). When the time required for virus internalization,
RNA transcription, and protein translation is taken
into account, the time course of morphologic changes
caused by dsSIN:C3 infection (within 1 h) is similar to
that reported previously by investigators using micro-
injected C3.
Intriguingly, longer incubation times with dsSIN:C3
caused additional changes in cellular morphology. At
6 h post-infection, dsSIN:C3-treated cells lost the
previously observed neurite-like extensions and be-
came completely rounded (Figure 2d), an eect not
previously described for C3 treatment of monolayer
cells. Using a quantitative assay of cellular attachment,
we next found that mock- or dsSIN:CAT-infected L929
monolayers exhibited minimal loss of cellular adher-
ence (Figure 2e). Somewhat similarly, dsSIN:C3
infection also caused little loss of adherence for up to
2 h post-infection (Figure 2e). However, by 3 h post
infection with dsSIN:C3, more than 30% of the cells
became detached and by 5 h, essentially all C3-treated
cells were detached (Figure 2e). Neither dsSIN:C3- or
dsSIN:CAT-infected L929 showed signi®cant loss of
viability when examined up to 8 h post-infection (data
not shown). dsSIN:C3 infection of a L929 cell variant
(which displays about a two log decreased eciency for
dsSIN viral replication) caused cellular retraction, but
complete cell rounding and detachment did not occur
(data not shown).
We next examined how the time course of
morphological changes induced by dsSIN:C3-infected
L929 correlated with biochemical modi®cation of Rho
by C3. C3 ADP-ribosylates Rho via the covalent
transfer of the ADP-ribose moiety of nicotine adenine
dinucleotide (NAD) to the asparagine residue at
position 41 of Rho (Sekine et al., 1989). The mobility
of ADP-ribosylated Rho on SDS ± PAGE is retarded
by about 1 kDa as compared with unmodi®ed Rho
(Dillon et al., 1995; Just et al., 1994; Ridley and Hall,
1992). Using this assay, we determined that infection of
L929 with dsSIN:C3 caused rapid and ecient ADP-
ribosylation of Rho (Figure 2f). Similar results were
obtained when we used a dierential ADP-ribosylation
assay (Tominaga et al., 1993). These results con®rmed
our initial interpretation that infection with dsSIN:C3
a
C 3 —
1 2 3 4
b
Figure 1 Expression of exoenzyme C3 by dsSIN:C3. (a) The
recombinant vectors contain a second subgenomic promoter
introduced into the 3'untranslated region of the RNA genome.
In this `double subgenomic' expression vector (dsSIN), sequences
encoding a foreign protein can be expressed via the second
subgenomic promoter. The cDNA fragment encoding the C3
polypeptide was subcloned into the phagemid shuttle vector. The
phagemid cassette (containing the second subgenomic promoter,
3'-UTR and cDNA for C3 polypeptide) was then subcloned into
the full-length recombinant dsSIN cDNA clone, which was
linearized, and full length infectious 5'Capped run-o RNA
transcripts were synthesized. RNA transfection of BHK-21 cells
yielded recombinant dsSIN:C3 virus stocks for use in target cells.
Except as speci®cally noted in the legend to b, all dsSIN
infections used a MOI of 20. (b) L929 cells were infected with
media alone (mock infection, lane 1), a dsSIN recombinant
expressing CAT (dsSIN:CAT, lane 2), or a dsSIN recombinant
expressing C3 (dsSIN:C3 @ MOI of 5, lane 3, and MOI of 20,
lane 4). After 1 h of incubation, cellular extracts were made,
subjected to SDS ± PAGE, and immunoblotted with anti-C3 as
outlined in Materials and methods. The immunoreactive bands
are of expected molecular weight (*28 kDa) for C3. Only the
relevant portion of the blot is displayed because no other
signi®cant immunoreactive bands were observed
Inactivation of Rho disrupts cellular attachment and induces apoptosis
DA Bobak et al
2180
produced a rapid and essentially complete ADP-
ribosylation of the pool of intracellular Rho. A
similarly ecient and rapid modi®cation of Rho was
noted in several other cell types including BHK-21,
CEF, RSM, P815, and EL-4 (Moorman et al., 1996
and data not shown). Although the Rac and Cdc42
sub-families of Rho-like proteins contain an Asn
residue in a homologous position to Asn
41
of Rho,
C3 does not appear to ADP-ribosylate Rac or Cdc42
to any appreciable extent (Tapon and Hall, 1997;
Ridley and Hall, 1992; Symons, 1996; Takai et al.,
1995). In the present study, we also found no
detectable ADP-ribosylation of either Rac or Cdc42
in lysates of dsSIN:C3-treated cells (data not shown).
We believe, therefore, that the cellular eects produced
by dsSIN:C3 in our study likely result from the speci®c
inactivation of Rho proteins.
Most of the previous studies on Rho and adhesion
examined stimulus-dependent induction of stress ®ber
formation in serum-starved monolayer cell lines. In
contrast, the roles of Rho in the initial steps of cellular
adhesion and spreading are largely unknown. To
address this question, we produced a suspension of
L929 cells and followed the course of these cells during
attachment and spreading. Within 3 h after being
placed in tissue culture wells, essentially all mock- and
ab
cd
e
f
1 2 3 4 5 6 7
Figure 2 Infection of adherent L929 cells with dsSIN:C3 ADP-ribosylates induces cellular retraction, rounding, and detachment
(a±d) L929 cells were infected with media alone (mock infection (a), dsSIN:CAT (b), or dsSIN:C3 (cand d). After 2 h (a±c)or6h
(d) of incubation, cells were visualized and photographed using inverted phase contrast microscopy. No signi®cant morphologic
changes from baseline were observed for mock- or dsSIN:CAT-infected cells at 2 or 6 h (6 h photos not shown). Representative
®elds are displayed. (e) L929 cells were cultured onto ¯at-bottom 96-well microtiter plates (2.5610
4
cells/well). Cells were infected
with media alone (mock), dsSIN:CAT, or dsSIN:C3. At the indicated times, supernatants were collected from the wells and the
number of cells counted. The percentage of detached cells (nonadherent cells) was measured and expressed as a function of time.
Each value represents the mean+s.d. of quadruplicate assays. (f) L929 cells were infected with media alone (mock, lane 1),
dsSIN:CAT (lanes 2, 4, 6), or dsSIN:C3 (lanes 3, 5, 7). After 1 h (lanes 2 and 3), 2 h (lanes 4 and 5), and 3 h (lanes 6 and 7), cells
were harvested, cellular extracts made and subjected to SDS ± PAGE followed by immunoblotting with anti-Rho as indicated in the
Materials and methods section. Lane 1 represents extracts from uninfected cells at time 0 h. Only the relevant portion of the blot is
shown. The lower arrow identi®es the mobility of unmodi®ed Rho and the upper arrow the mobility of ADP-ribosylated Rho
Inactivation of Rho disrupts cellular attachment and induces apoptosis
DA Bobak et al
2181
dsSIN:CAT-infected cells showed evidence of spread-
ing (de®ned by loss of cellular rounding and the
formation of new cellular extensions) (Figure 3a, b and
d). In contrast, infection of suspension L929 cells with
dsSIN:C3 prevented subsequent cell spreading for as
long as 6 h (Figure 3c and d). Although the C3-treated
cells remained completely rounded and refractile, we
could not conclusively determine whether these cells
had actually attached. When we placed the culture
dishes at an angle or gently tapped them, no movement
of the cells occurred. Aspiration of the media with a
micropippetor was sucient to dislodge the cells,
however, indicating that attachment was labile at
best. Nonetheless, the C3-treated cells clearly did not
spread, a ®nding that was in dramatic contrast to the
normal course of attachment and spreading exhibited
by mock- or control dsSIN:CAT-infected cells. Similar
to results observed with monolayer L929 (Figure 2d),
the morphological eects of dsSIN:C3 infection in
suspension L929 correlated with the intracellular
expression of functional C3 (Figure 3e).
We next sought to determine whether the dsSIN:C3
eects we had observed in L929 were relevant to
primary, non-transformed monolayer cells as well. We
found that infection of CEF with dsSIN:C3 also
produced marked cellular retraction and rounding
(Figure 4). At later time points, we further observed
that dsSIN:C3-infected CEF became rounded, lost
neurite-like structures and detached (data not shown).
In fact, the time course of C3-mediated morphological
ab
c
d
e
1 2 3
Figure 3 Infection of suspension L929 cells with dsSIN:C3 ADP-ribosylates Rho and disrupts cellular spreading. (a±c) L929 cells
were detached with trypsin/EDTA and the suspension cells were infected with media alone (mock), dsSIN:CAT or dsSIN:C3, and
then added to multiwell plates. After 6 h of incubation, wells were photographed. (d). In experiments identical to those performed in
a±cabove, L929 cells were detached with trypsin/EDTA and the suspension cells were infected with media alone (mock),
dsSIN:CAT, or dsSIN:C3, and then added to multiwell plates. At the indicated times, the percentage of cells remaining rounded
(i.e., unattached or attached but unspread) was measured and expressed as a function of time. Each value represents the mean of
triplicate assays, the s.d. for each value was 410%. (e) Suspension L929 cells were infected with media alone (mock, lane 1),
dsSIN:CAT (lanes 2), or dsSIN:C3 (lanes 3). After 1 h of incubation, cellular extracts wer collected, analysed, and displayed as
indicated in the legend to Figure 2f
Inactivation of Rho disrupts cellular attachment and induces apoptosis
DA Bobak et al
2182
changes was somewhat faster for CEF compared with
L929, likely re¯ecting the ecacy of infection of dsSIN
(Hahn, unpublished data). The eects of dsSIN:C3 on
the cellular morphology of primary cell types were
observed not only in ®broblasts, but in explants of rat
aortic smooth muscle (RSM) cells as well (Figure 4).
Cellular adhesion requires the assembly of a variety
of intracellular proteins into focal adhesion complexes,
a process which appears to require the participation of
protein tyrosine kinases (Parsons, 1996). In particular,
the phosphorylation-dependent activation of FAK is
necessary for focal adhesion formation by integrins.
Treatment of monolayer cells with tyrosine kinase
inhibitors decreases the level of phosphorylation of
FAK and also disrupts adhesion (Parsons, 1996).
Correlating this information with our own results, we
next decided to examine whether C3 altered the
phosphorylation state of FAK. Because the activation
and regulation of FAK are especially well-described for
CEF (Schaller et al., 1992), we infected CEF
monolayers with dsSIN:C3 and observed a signi®cant
decrease in the basal state of tyrosine phosphorylation
on FAK in dsSIN:C3 infected cells (as compared with
the eects of mock or dsSIN:CAT infection) (Figure
5).
In some types of primary cell monolayers, cellular
detachment leads to apoptosis (Frisch and Francis,
1994; Meredith et al., 1993; Re et al., 1994). In this
situation, the cell likely activates its cell death pathway
secondary to the withdrawal of adhesion-dependent
signals required for cell survival (Bates et al., 1995).
Because we had observed that the dsSIN:C3-mediated
inactivation of Rho led to cellular detachment, it
seemed possible that C3-treatment of monolayer cells
could lead to apoptosis as well. Even though infection
of L929 monolayers with dsSIN:C3 disrupted cellular
adhesion (Figure 2), such treatment did not induce
appreciable levels of apoptosis (data not shown). Other
investigators have also shown that many types of
transformed monolayer cells lack adhesion-dependent
regulation of apoptosis (Dehar, 1995; Frisch and
Francis, 1994; Meredith et al., 1993; Re et al., 1994).
Intriguingly, we obtained dierent results when we
examined the C3 eects on primary monolayer cells.
Mock- or dsSIN:CAT-infected CEF or RSM exhibited
no discernible apoptosis (Figure 6, lanes 2, 3, 5 and 6).
In contrast, dsSIN:C3 was a potent inducer of
apoptosis in both cell types (Figure 6, lanes 4 and 7).
In addition to using DNA gel electrophoresis, we also
measured apoptosis by a ¯ow cytometry-based
technique (Moorman et al., 1996). DNA histograms
of dsSIN:C3-treated cells displayed the pattern of sub-
diploid low molecular weight DNA fragmentation
characteristic of apoptosis: e.g., at 6 h post infection,
mock- and dsSIN:CAT-infected CEF exhibited less
than 1% apoptosis, whereas 22% of dsSIN:C3-infected
CEF were apoptotic (mean values of duplicate assays
with s.e.m.510%).
Figure 4 dsSIN:C3 infection causes cellular retraction and rounding of primary explants of ®broblasts and smooth muscle cells.
L929 (a±c), chicken embryo ®broblasts (CEF) (d±f), and rat aortic smooth muscle cells (RSM) (g±i) were infected with media
alone (a,d,g), dsSIN:CAT (b,e,h), or dsSIN:C3 (c,e,i). After 1 h of incubation, cells were photographed using phase contrast
microscopy
Inactivation of Rho disrupts cellular attachment and induces apoptosis
DA Bobak et al
2183
Having found that infection of primary cell types
with dsSIN:C3 disrupted adherence, it at ®rst seemed
plausible to us that the induction of apoptosis by C3
could have been entirely accounted for by the loss of
adhesion-mediated signals resulting from the C3-
mediated inactivation of Rho. We knew that this could
not be the sole explanation, however, because in other
related work we had recently found that dsSIN:C3
infection activates apoptosis in the suspension cell line,
EL4 (Moorman et al., 1996). We reasoned, therefore,
that the inactivation of Rho must also be able to lead to
apoptosis in a manner independent of Rho's roles in
adhesion-mediated signaling.
Because we were able to eciently infect suspen-
sions of L929 with dsSIN:C3 (Figure 3), we used this
approach to test whether dsSIN:C3 could induce
apoptosis in suspensions of CEF and RSM. In CEF
cells detached by treatment with trypsin-EDTA alone,
apoptosis rapidly ensued (Figure 7, lane 1),
demonstrating that CEF are exquisitely sensitive to
loss of adhesion-dependent cellular survival signals
and that we could not determine adhesion-indepen-
dent eects of C3 on apoptosis in CEF (Figure 7
lanes 2 and 3). On the other hand, RSM cells
detached by trypsin-EDTA showed little baseline
apoptosis during the same 6 h incubation time
(Figure 7, lane 4). Suspension RSM, therefore,
represented a model that could enable us to examine
the eects of C3 on apoptosis independent of
adhesion. Infection of suspension RSM by
dsSIN:CAT showed no appreciable eect on apopto-
sis (Figure 7, lane 5). However, dsSIN:C3 infection
of RSM in suspension was also a potent activator of
apoptosis (Figure 7, lane 6). These results indicate
that, in RSM, C3 can induce apoptosis independent
of the adhesion status of the cells.
Using the ¯ow cytometric assay outlined above, we
con®rmed the eects of C3 on apoptosis in RSM and
additionally found that the overexpression of `loss of
function' mutants of Rho also led to apoptosis (Figure
8). Sindbis-based expression of a dominant negative
form of Rho (Rho
19N
) or an inactive form of Rho
(Rho
37A
) eciently induced apoptosis in adherent as
well as suspension RSM. These results con®rmed that
the C3-induced eects on apoptosis were due to
inactivation of Rho and not any theoretical Rho-
independent eects.
220 —
97 —
66 —
anti-pTyr anti-FAK
1 2 3 4 5 6
Figure 5 Infection of CEF cells with dsSIN:C3 decreases
phosphotyrosine levels of FAK. Monolayers of CEF were
infected with media alone (lanes 1 and 4), dsSIN:CAT (lanes 2
and 5), or dsSIN:C3 (lanes 3 and 6). After 3 h of incubation,
cellular extracts were made, normalized by protein content, and
subjected to immunoprecipitation with anti-FAK followed by
SDS ± PAGE and immunoblotting with anti-pTyr (lanes 1 ± 3) or
anti-FAK (lanes 4 ± 6) as indicated in Materials and methods
1 2 3 4 5 6 7
Figure 6 dsSIN:C3 infection induces adhesion-dependent apop-
tosis in primary explants of ®broblasts and smooth muscle cells.
Monolayers of CEF (lanes 2 ± 4) and RSM (lanes 5 ± 7) were
infected with media alone (lanes 2 and 5), dsSIN:CAT (lanes 3
and 6), or dsSIN:C3 (lanes 4 and 7). After 6 h of incubation, cells
were harvested, the genomic DNA puri®ed and fractionated by
agarose electrophoresis as outlined in Materials and methods
section. Lane 1 is the DNA standard (HindIII-lambda phage
digest)
1 2 3 4 5 6
Figure 7 dsSIN:C3 infection induces adhesion-independent
apoptosis in smooth muscle cells. CEF (lanes 1 ± 3) and RSM
(lanes 4 ± 6) were treated with trypsin-EDTA to generate single
cell suspensions. After washing with serum-containing media to
inactivate trypsin, the suspension cells were infected with media
alone (lanes 1 and 4), dsSIN:CAT (lanes 2 and 5), or dsSIN:C3
(lanes 3 and 6). After 6 h of incubation in which cells were
maintained in suspension, the cellular genomic DNA was
extracted, puri®ed and fractionated by agarose electrophoresis
exactly as in the experiment shown in Figure 6. The DNA
standard (HindIII-lambda phage digest) is to the left
Inactivation of Rho disrupts cellular attachment and induces apoptosis
DA Bobak et al
2184
Discussion
In this study, we observed that dsSIN:C3 caused
morphologic changes of cellular retraction similar to
those changes reported in studies using microinjection
of C3 (reviewed in Takai et al., 1995; Zigmond, 1996).
In addition, though, we also observed that, over time,
C3 expression caused complete cell rounding, loss of
neurite-like extensions, and actual cellular detachment
(Figure 2), results that have not previously been
reported with C3. These more dramatic changes may
well result from the ecacy of the C3 expression
system we used here and/or the prolonged inactivation
of Rho due to continual C3 expression. With dsSIN:C3
treatment, the time between ADP-ribosylation of Rho
and onset of cell retraction (51 h) was short, but there
was a noticeable lag between the ADP-ribosylation of
Rho and the time to complete cell rounding/detach-
ment (*6 h) (Figure 2). Cellular retraction is likely
caused by the rapid disruption of actin stress ®ber
organization. Actual cellular detachment, as occurs
during cell division, tissue dierentiation or metastasis,
appears to be an even more complexly regulated event
(Zigmond, 1996). In this context, then, we are tempted
to speculate that our results here indicate that the
regulation of stress ®ber organization by Rho is
relatively labile and easily disrupted. In contrast, the
continuous and prolonged inactivation of Rho-
dependent signaling pathways may be needed to
trigger disassembly of the network of subcellular
complexes required for cellular detachment.
We also demonstrated here that Rho is required for
the spreading of cells following contact with substra-
tum (Figure 3). Cellular attachment signals alone,
therefore, appear to be insucient to trigger spreading
in the absence of functional Rho, a result in general
agreement with the ®ndings of Hotchin and Hall using
a dierent approach (Hotchin and Hall, 1995). Rho
would therefore appear to be one of the most upstream
regulators involved in cellular attachment and spread-
ing. Rho has been hypothesized to be required for
growth factor-dependent integrin activation and
clustering (Hotchin and Hall, 1995; Parsons, 1996).
However, in this study we observed that continued
disruption of Rho in cells that were already ®rmly
adherent led to cellular detachment (Figure 2). Clearly,
then, some type of Rho-sensitive signal or feedback
continues to exist even after the initial steps of integrin-
matrix interaction and focal adhesion complex
assembly are complete, possibly at the level of
biomechanical cell shape regulation.
The regulation of cellular adhesion is exceedingly
complex and the nature of the adhesion-induced
signaling varies according to the type of cell and
nature of the extracellular matrix. For example, many
types of primary cells exhibit anchorage-dependent
apoptosis, a phenomenon whereby adhesion-dependent
signals normally suppress apoptosis but cellular
detachment usually leads to apoptosis. In particular,
certain types of cell-matrix interactions generate
patterns of cell signaling that serve to promote cell
survival and cell growth (Meredith et al., 1993; Zhang
et al., 1995; Ruoslahti, 1997). Data presented here, as
well as that of other investigators, implicate Rho
proteins as critical regulators of these adhesion-
dependent survival signals. It is not yet known exactly
which adhesion-mediated signal pathways are required
to suppress apoptosis in normal cells, although
evidence suggests that integrin-activated FAKs and
other tyrosine kinases are important (Clark and
Brugge, 1995; Meredith et al., 1993; Frisch et al.,
1996). Recent work suggests that biomechanical signals
generated by cell shape and tension may also be critical
in suppressing apoptosis in adherent cells (Chen et al.,
1997). Many transformed cell lines do not exhibit the
phenomenon of anchorage-dependent apoptosis, an
attribute that is essential for cellular metastasis
(Meredith et al., 1993; Zhang et al., 1995; Malik and
Parsons, 1996; Varner and Cheresh, 1996). The results
of our present study are consistent with these
observations: we found that C3-mediated cellular
detachment induced apoptosis only in primary fibro-
blasts (CEF) and smooth muscle cells (RASM), but not
in a transformed ®broblast line (L929).
In addition to the roles of Rho in adhesion-
dependent signaling, evidence also exists to support
an adhesion-independent function for Rho in suppres-
Figure 8 Expression of loss of function Rho mutants induces
adhesion-independent apoptosis in smooth muscle cells. Mono-
layer (a) or suspension RSM (b) were infected with media alone
(Mock, dsSIN:CAT (CAT), dsSIN:C3 (C3), dsSIN:RhoA (wt),
dsSIN:RhoA
19N
(19N), or dsSIN:RhoA
37A
(37A). After 6 h of
incubation, the cells were harvested and ®xed for ¯ow cytometry.
Apoptosis was determined according to the methods outlined in
the Materials and methods section. The expression levels of
exogenous Rho proteins were similar (not shown). Each value
represents the mean+s.e.m. of duplicate experiments. In both
panels, the C3, 19N, and 37A values are each signi®cantly
increased compared to the CAT and wt values (at least P50.05
for each comparison)
Inactivation of Rho disrupts cellular attachment and induces apoptosis
DA Bobak et al
2185
sing and/or regulating apoptosis. In other work, we
have found that the C3-mediated inactivation of Rho
in two forms of leukocytes leads to apoptosis in these
suspension cells. Infection of EL4 lymphocytes with
dsSIN:C3 eciently inactivates endogenous Rho and
induces apoptosis (Moorman et al., 1996). In related
work, we found that another Rho-inactivating
exoenzyme known as toxin A of Clostridium dicile
(TxA) activates apoptosis in neutrophils, HL-60 and
EL4 (Thorne and Bobak, 1997). Aktories and
colleagues have determined that the enzymatic action
of TxA is to mono-glucosylate Rho family GTPases on
Thr
37
, an amino acid residue within the Rho eector
region that is located very near the site of C3-mediated
ADP-ribosylation at Asn
41
(Just et al., 1995a,b).
Support for a role for Rho in regulating cellular
survival signals comes from the work of other
investigators as well. Avraham and Weinberg origin-
ally demonstrated that overexpression of Rho elim-
inates the serum requirement in proliferating NIH3T3
cells (Avraham and Weinberg, 1989). Other investiga-
tors have con®rmed those results and additionally
found that the overexpression of a constitutively active
form of Rho also reduces anchorage-dependence for
cell growth (Perona et al., 1993; Prendergast et al.,
1995). Giry et al. demonstrated that overexpression of
Rho proteins potentiates cellular resistance to the
cytotoxic eects of C. dicle toxins A and B (Giry et
al., 1995). Furthermore, Prendergast and colleagues
recently found that certain farnesyltransferase inhibi-
tors apparently induce apoptosis in Ras-transformed
cells through a Rho-dependent pathways (i.e., apopto-
sis was inhibited by expression of a farnesyl-
independent form of Rho (Lebowitz et al., 1997)).
Finally, Henning et al. achieved thymus-speci®c
expression of a C3 transgene in a murine system and
demonstrated that functional Rho is required for the
survival and proliferation of normal T-cells (Henning
et al., 1997). In apparent contrast to these studies are
reports by Lacal and colleagues in which the over-
expression of constitutively active forms of Rho were
observed to induce apoptosis (Esteve et al., 1995;
Jimenez et al., 1995). In general, though, the design of
these particular studies diered from those outlined
above, particularly in that Rho-induced apoptosis was
observed in the setting of serum starvation.
If Rho GTPases do indeed act downstream of
growth- and survival-promoting pathways, then it is
conceivable that expression of constitutively active Rho
in the setting of serum starvation creates antagonistic
signals within the cell, resulting in a signalling con¯ict
that leads to the activation of the programmed cell
death pathway. Therefore, we speculate that either the
activation or inactivation or Rho GTPases can lead to
apoptosis, depending on the state of the cell (e.g.,
normal or stressed). This hypothesis mirrors that of
Lacal's group in a recent report in which they also
propose that `Rho-mediated signals may lead to a
protective response and others to cell suicide,
dependent of cell type and the speci®c extracellular
stimulus (Perona et al., 1997)'. In that same study, the
investigators demonstrated that activation of nuclear
factor-kB (NF-kB) is regulated by Rho family GTPases
(Perona et al., 1997). In many systems, synthesis of
survival and anti-apoptotic factors requires activation
of NF-kB (Baichwal and Baeuerle, 1997). It is tempting
to speculate then, that disrupting Rho function in
certain types of non-transformed and primary cell
types disturbs NF-kB-sensitive growth- and survival-
promoting factors, leading to activation of apoptosis.
Thus, we now believe that there is compelling evidence
to place Rho GTPases within signal pathways required
for cell survival and growth that are separate from, but
may well overlap with, Rho-dependent pathways
regulating cellular adhesion. In some systems, for
example, Rho regulates key growth-related signals but
apparently does not regulate actin micro®lament
organization (Thorburn et al., 1997). Future work in
this area will likely be directed at identifying the
speci®c functions of Rho in the related, but distinct,
intracellular signaling pathways activated by integrins
and growth factors that exist to control cell survival
and death.
Materials and methods
Materials
Tissue culture media, saline solutions and heat inactivated
fetal bovine serum were purchased from commercial
sources including Gibco-BRL (Grand Island, NY). Bio-
Whittaker (Walkersville, MD). Hyclone (Logan, UT), Cell-
Gro and Sigma (St Louis, MO). Restriction enzymes and
modifying enzymes were purchased from commercial
sources and used essentially according to manufacturers
speci®cations. Biochemicals and chemicals were from
Sigma (St Louis, MO) and Calbiochem (San Diego, CA).
Speci®c monoclonal and polyclonal antibodies recognizing
RhoA were from Santa Cruz Biotechnology (Santa Cruz,
CA). Rabbit polyclonal anti-serum recognizing exoenzyme
C3 of Clostridium botulinum was a gift from Dr M Popo
(Institut Pasteur, Paris, France) (Popo et al., 1990, 1991).
Rabbit polyclonal and murine monoclonal antibodies
recognizing focal adhesion kinase (FAK) were a gift from
Dr J Thomas Parsons, University of Virginia. Murine
monoclonal antibody recognizing phosphotyrosine (pTyr)
(clone G10) was from Upstate Biotechnology Incorporated
(Rochester, NY). Horseradish peroxidase-conjugated sec-
ondary antibodies were from BioRad (Hercules, CA). All
other reagents and chemicals were purchased from
commercial sources of the highest grade available. Purified
exoenzyme C3 of C. botulinum was a gift from Dr J Moss
(National Institutes of Health, Bethesda, MD) and also
from Upstate Biotechnology Incorporated (Rochester,
NY); biochemical activities were similar with either form
of C3. cDNA encoding the mature C3 exoenzyme was a
gift from Dr M Popo (Institut Pasteur, Paris, France)
(Popo et al., 1990, 1991). cDNA encoding wild type
RhoA was a gift from Dr C Der (University of North
Carolina-Chapel Hill, NC).
Cells
Murine L929 cells were obtained from ATCC (Rockville,
MD), maintained in Dulbecco's Modi®ed Eagle's medium
(DME) (supplemented with 10% heat-inactivated fetal
bovine serum, 10 m/ml penicillin, 10 mg/ml streptomycin
and 2 mML-glutamine), split every 3.5 days, and used
between passages 3 and 13 of ATCC. Baby Chinese
hamster kidney cells (BHK21 clone 13) were obtained
from American Type Culture Collection (Rockville, MD),
maintained in DME (supplemented with 10% heat-
inactivated fetal bovine serum, 10 m/ml penicillin, 10 mg/
ml streptomycin and 2 mMglutamine) and used between
passages 6 and 14 after acquisition from ATCC. Chicken
Inactivation of Rho disrupts cellular attachment and induces apoptosis
DA Bobak et al
2186
embryo ®broblast (CEF) cells were isolated from 10 day
old chicken embryo and split once to generate secondary
CEF in MEM supplemented with 5% heat-inactivated fetal
bovine serum as described (Pierce et al., 1973). Explants of
rat aortic smooth muscle cells were a generous gift from Dr
G Owen (University of Virginia) and were maintained in
DMEM/F12 medium (supplemented with 10% heat-
inactivated fetal bovine serum, 10 m/ml penicillin and
10 mg/ml streptomycin), and were used between passages
8 and 13 (Geisterfer et al., 1988).
Generation of dsSIN recombinants
dsSIN recombinants were generated as described (Hahn et
al, 1992a). Brie¯y, the coding sequence of C. botulinum
exoenzyme C3 was ampli®ed by polymerase chain reaction
using the appropriate primers which added ¯anking
sequences representing restriction sites for NcoI(5'-
CCGGTGCCATGGCACCCTGCATGCTGC-3')andBam-
HI (5'-CCCGGGGATCCTCACGCTTTACAATCTGGG-
3'). The ampli®ed fragment was digested with NcoIand
BamHI and inserted into the phagemid shuttle vector
pH3'2J1. The recombinant plasmid with the appropriate
insert was isolated, ampli®ed, and shuttled into plasmid
pTE3'2JC1, which contains the full-length double sub-
genomic Sindbis virus cDNA. Two independent recombi-
nant plasmids containing the appropriate insert, referred to
as pTE3'2J:C3, were ampli®ed and puri®ed for use in vitro
transcription/transfection. pTE3'2J:C3 plasmids were linear-
ized by digesting with the restriction enzyme XhoI and a run
o in vitro transcription reaction was performed (Hahn et al.,
1992a). The integrity and quantity of the RNA transcripts
were examined by agarose gel electrophoresis and 5 mgof
full-length RNA transcripts were transfected into subcon-
¯uent monolayers of BHK cells by either lipofection or
electroporation. Sixteen to 20 h post transfection, the media
were collected and titered by either plaque assay or lytic unit
assay in L929 to determine the plaque forming unit (p.f.u.).
Usually 0.5 to 1.0610
10
p.f.u. of recombinant viruses were
generated from 5610
6
BHK cells. The overall strategy for
construction of dsSIN:C3 is displayed in Figure 1a.
Using the RhoA cDNA (see Materials and methods),
cDNAs corresponding to the 19N (dominant negative) and
37A (constitutively inactive) mutant forms of RhoA were
constructed by site-directed mutagenesis (Ausubel et al., 1997;
Hahn et al., 1992b). All cDNAs were con®rmed by DNA
sequencing and used to create Sindbis recombinants capable
of expressing wild type RhoA (dsSIN:RhoA), dominant
negative RhoA (dsSIN:RhoA
19N
) and inactive RhoA
(dsSIN:Rho
37A
). Acting in an analogous manner to that of
dominant negative Ras (Ras
17N
), Rho
19N
remains stably
complexed with GDP and functions as a dominant negative
form of Rho (Feig and Cooper, 1988; Prendergast et al.,
1995; Ridley and Hall 1992). It is believed that RhoA
37A
cannot coordinate Mg
2+
binding and thus is `inactive' due to
its inability to bind GDP or GTP (Just et al., 1995b;
Hirshberg et al., 1997).
Although exoenzyme C3 is not considered a toxin and SIN
is not a human pathogen, dsSIN:C3 recombinants could have
potential biohazard concerns. Therefore, we used BL-2
laboratories and precautions for the construction and
experimental use of the dsSIN:C3 recombinants in accor-
dance with protocols approved by the Institutional Biosafety
and Recombinant DNA Committee of the University of
Virginia for the use of recombinant DNA and materials with
biohazard potential.
Infections with dsSIN recombinants
Monolayers of cells at approximately 80% con¯uency were
infected with various Sindbis virus recombinants. Cells
were seeded to form monolayers and at 80% con¯uency,
media were removed and viruses were added to a
designated multiplicity of infection (MOI) in the presence
of PBS (with 2 mMCaCl
2
/2 mMMgCl
2
). Unless otherwise
noted in the ®gure legend, all dsSIN infections used a MOI
of 20. After cells were incubated in the presence of viruses
for an hour, the inocula were removed and appropriate
media were added. At the times indicated, media were
removed and cells were processed as described. For each
experiment two controls were used: (1) addition of media
alone (mock infection); and (2) infection with a dsSIN
recombinant capable of expressing the irrelevant marker
enzyme chloramphenicol acetyltransferase (dsSIN:CAT).
Measurement of cell detachment and cell attachment/spreading
Cell detachment L929 were plated into ¯at-bottom multi-
well plates to achieve a density of 2.5610
4
cells/well.
Monolayers were infected with media alone (mock),
dsSIN:CAT, or dsSIN:C3 and incubated at 378Cin5%
CO
2
. At the times indicated the supernatant was removed
and wells were washed once with PBS. The original
supernatant and the wash ¯uid were pooled, the number
of cells counted, and this value was considered to represent
the number of cells detached. The percentage of detached
cells was de®ned as [(number of cells detached/number of
cells originally attached)6100].
Cell attachment/spreading L929 were detached with
trypsin/EDTA, washed and the suspension cells were
infected with media alone (mock), dsSIN:CAT, or
dsSIN:C3 as indicated. Infected cells were added to ¯at-
bottom multiwell plates, incubated at 378Cin5%CO
2
,and
the plates were removed at the indicated times for
evaluation by inverted phase microscopy. The percentage
of cells remaining completely rounded was measured; a
minimum of 200 cells (in at least four random ®elds) were
evaluated for each condition. Criteria for rounded and
spread cells were essentially as described (Jalink et al.,
1994); i.e., only completely spherical, refractile cells with
no visible extensions were considered rounded.
Cellular extracts, immunoblotting, and immunoprecipitation
After speci®ed hours of incubations, cells were washed with
cold PBS, resuspended at a concentration of 1 ± 2610
6
cells/500 ml in RIPA buer (50 mMTris pH 7.5, 150 mM
NaCl, 1 mMEGTA, 1.0% NP-40, 0.5% deoxycholate,
1m
MAEBSF, 0.3 mMaprotinin, 1 mMleupeptin), incu-
bated for 15 min at 48C, and centrifuged at 10 000 g for
10 min (Harlow and Lane, 1989; Schaller et al., 1992,
1994). The supernatant lysates were collected, and the
protein concentration determined using the Pierce BCA
protein assay kit (Rockford, IL) according to the
manufacturer's instructions. The individual sample vo-
lumes were normalized by protein concentration and
separated by 12% SDS ± PAGE (acrylamide gels). Gels
were electroblotted onto polyvinylidene ¯uoride (PVDF)
membranes (Millipore Corp., Bedford, MA), which were
immunoblotted using the speci®c primary antibodies
described above. Immunoreactive bands were detected
using horseradish peroxidase-conjugated secondary anti-
bodies in conjunction with an enhanced chemiluminescent
system (ECL, Amersham Corporation, Arlington Heights,
IL).
For immunoprecipitation experiments, cells were harvested
and lysed as outlined above with the exception that the RIPA
buer additionally contained 1 mMsodium vanadate.
Following lysis, cellular extracts were equalized on the basis
of protein concentration and subjected to immunoprecipita-
tion with rabbit polyclonal anti-FAK, essentially as
previously described (Schaller et al., 1992, 1994). The
resulting samples were separated by 8% SDS ± PAGE,
immunoblotted using either monoclonal anti-FAK or
Inactivation of Rho disrupts cellular attachment and induces apoptosis
DA Bobak et al
2187
monoclonal anti-pTyr, and immunoreactive bands were
visualized as outlined above.
Determination of apoptosis
Cellular apoptosis was determined by the observation of
nucleosomal DNA fragmentation using agarose gel
electrophoresis of genomic DNA (i.e., observation of the
characteristic low molecular weight DNA ladders). At the
indicated times post-infection, cells were harvested and
lysed in digestion buer (100 mMTris-Cl-pH 8.0, 100 mM
NaCl, 25 mMEDTA, 0.5% sodium dodecyl sulfate and
0.1 mg/ml proteinase K) for 1 h at 508C (Ausubel et al.,
1997). Genomic DNA was isolated by sequential extraction
with equal volumes of phenol and chloroform, followed by
ethanol precipitation. Puri®ed DNA was treated with
100 mg/ml DNAse-free RNAse A for 1 h at 378C. DNA
content was normalized and separated by 2% agarose gel
electrophoresis. The gel was stained with 10 mg/ml
ethidium bromide and the DNA bands were visualized by
ultraviolet transillumination. Cellular apoptosis was also
determined by ¯ow cytometry, essentially as described by
Moorman et al., 1996.
Experimental results
Each result displayed as a ®gure, stated in the text or cited
as `data not shown' is representative of two or more similar
and separate experiments. In the ®gures displaying
quantitative data, each value represents the mean of
triplicate or quadruplicate assays, with standard deviation
shown or stated as indicated in the ®gure legend.
Acknowledgements
This work was supported by research grants from the
DHHS/NIH (GM54572-DB, CH), National Leukemia
Association (DB), and Council for Tobacco Research
(DB, CH). JM was supported by a post-doctoral fellow-
ship training grant from the DHHS/NIH.
References
Aktories K. (1994). Mol. Cell. Biochem.,138, 167 ± 176.
Aktories K, Mohr C and Koch G. (1992). Curr. Top.
Microbiol. Immunol.,175, 115 ± 131.
AusubelFM,BrentR,KingstonRE,MooreDD,Seidman
JG, Smith JA and Struhl K. (eds) (1997). Current
Protocols in Molecular Biology. Wiley-Interscience, New
York.
Avraham H and Weinberg RA. (1989). Mol. Cell. Biol.,9,
2058 ± 2066.
Bates RC, Lincz LF and Burns GF. (1995). Cancer Metast.
Rev.,14, 191 ± 203.
Baichwal VR and Baeuerle PA. (1997). Curr. Biol.,7, R94 ±
R96.
Chardin P, Boquet P, Madaule P, Popo MR, Rubin EJ and
Gill DM. (1989). EMBO J.,8, 1087 ± 1092.
Chen CS, Mrksich M, Huang S, Whitesides GM and Ingber
DE. (1997). Science,276, 1425 ± 1428.
Clark EA and Brugge JS. (1995). Science,268, 233 ± 239 .
Dehar S. (1995). Cancer Metast. Rev.,14, 165 ± 172.
Dillon ST, Rubin EJ, Yakubovich M, Pothoulakis C,
LaMont JT, Feig LA and Gilbert RJ. (1995). Infect.
Immunol.,63, 1421 ± 1426.
Esteve P, del Peso L and Lacal JC. (1995). Oncogene,11,
2657 ± 2665.
Feig LA and Cooper GM. (1988). Mol. Cell Biol.,8, 3235 ±
3241.
Frisch S and Francis H. (1994). J. Cell Biol.,124, 619 ± 626.
Frisch SM, Vuori K, Ruoslahti E and Chan-Hui PY. (1996).
J. Cell Biol.,134, 793 ± 799.
Geisterfer AAT, Peach MJ and Owens GK. (1988). Circ.
Res.,62, 749 ± 756.
Giry M, Popo MR, von Eichel-Strieber C and Boquet P.
(1995). Infect. Immun.,63, 4063 ± 4071.
Hahn CS, Hahn YS, Braciale TJ and Rice CM. (1992a).
Proc. Natl. Acad. Sci. USA,89, 2679 ± 2683.
Hall A. (1994). Annu. Rev. Cell Biol.,10, 31 ± 54.
Harlow E and Lane D. (1989). Antibodies: a Laboratory
Manual. Cold Spring Harbor Laboratory, Cold Spring
Harbor.
Henning SW, Galandrini R, Hall R and Cantrell DA. (1997).
EMBO J.,16, 2397 ± 2407.
Hirshberg M, Stockley RW, Dodson G and Webb GR.
(1997). Nature Struct. Biol.,4, 147 ± 152.
Hotchin NA and Hall A. (1995). J. Cell Biol.,131, 1857 ±
1865.
Jalink K, van Corven EJ, Hengeveld T, Morii N, Narumiya S
and Moolenaar WH. (1994). J. Cell. Biol.,126, 801 ± 810.
Jimenez B, Arends M, Esteve P, Perona R, Sanchez R,
Ramon y Cajal S, Wyllie A and Lacal JC. (1995).
Oncogene,10, 811 ± 816.
Just I, Fritz G, Aktories K, Giry M, Popo MR, Boquet P,
Hegenbarth S and von Eichel-Streiber C. (1994). J. Biol.
Chem.,269, 10706 ± 10712.
Just I, Mohr C, Schallehn G, Menard L, Didsbury JR,
Vandekerckhove J, van Damme J and Aktories K.
(1992a). J. Biol. Chem.,267, 10274 ± 10280.
Just I, Schallehn G and Aktories K. (1992b). Biochem.
Biophys. Res. Commun.,183, 931 ± 936.
Just I, Selzer J, Wilm M, von Eichel-Streiber C, Mann M and
Aktories K. (1995a). Nature,375, 500 ± 503.
Just I, Wilm M, Selzer J, Rex G, von Eichel-Streiber C,
Mann M and Aktories K. (1995b). J. Biol. Chem.,270,
13932 ± 13936.
Kishi K, Sasaki T, Kuroda S, Itoh T and Takai Y. (1993). J.
Cell. Biol.,120, 1187 ± 1195.
Lang P, Guizani L, Vitte-Mony I, Stancou R, Dorseuil O,
Gacon G and Bertoglio J. (1992). J. Biol. Chem.,267,
11677 ± 11680.
Lebowitz PF, Sakamuro D and Prendergast GC. (1997).
Cancer Res.,57, 708 ± 713.
Lim L, Manser E, Leung T and Hall C. (1996). Eur. J.
Biochem.,242, 171 ± 185.
Malik RK and Parsons JR. (1996). Biochem. Biophys. Acta,
1287, 73 ± 76.
Meredith JE, Fazeli B and Schwartz MA. (1993). Mol. Biol.
Cell,4, 953 ± 961.
Moorman JP, Bobak DA and Hahn CS. (1996). J. Immunol.,
156, 4146 ± 4153.
Narumiya S and Morii N. (1993). Cell Signal,5, 9±19.
Parsons J. (1996). Curr. Opin. Cell Biol.,8, 146 ± 152.
Perona R, Esteve P, Jimenez B, Ballestero RP, Ramon y
Cajal S and Lacal JC. (1993). Oncogene,11, 1285 ± 1292.
Perona R, Montaner S, Saniger L, Sa
Ânchez-Pe
Ârez I, Bravo R
and Lacal JC. (1997). Genes Develop.,11, 463 ± 475.
Pierce JS, Strauss EG and Strauss JH. (1973). J. Virol.,13,
1030 ± 1036.
Popo M, Boquet P, Gill DM and Eklund MW. (1990). Nucl.
Acid. Res.,18, 1291.
Popo MR, Hauser D, Boquet P, Eklund MW and Gill DM.
(1991). Infec. Immunity,59, 3673 ± 3679.
Prendergast GC, Khosravi-Far R, Solski PA, Kurzawa H,
Lebowitz PF and Der CJ. (1995). Oncogene,10, 2289 ±
2296.
Inactivation of Rho disrupts cellular attachment and induces apoptosis
DA Bobak et al
2188
Re F, Zanetti A, Sironi M, Polentarutti N, Lanfrancone L,
Dejana E and Colotta F. (1994). J. Cell. Biol.,127, 537 ±
546.
Ruoslahti E. (1997). Science,276, 1345 ± 1346.
Ridley AJ. (1996). Curr. Biol.,6, 1256 ± 1264.
Ridley AJ and Hall A. (1992). Cell,70, 389 ± 399.
Rubin EJ, Gill M, Boquet P and Popo MR. (1988). Mol.
Cell. Biol.,8, 418 ± 426.
Schaller MD, Borgman CA, Cobb BS, Vines RR, Reynolds
AB and Parsons JT. (1992). Proc. Natl. Acad. Sci. USA,
89, 5192 ± 5196.
Schaller MD, Hildebrand JD, Shannon JD, Fox JW, Vines
RR and Parsons JT. (1994). Mol. Cell Biol.,14, 1680 ±
1688.
Sekine A, Fujiwara M and Narumiya S. (1989). J. Biol.
Chem.,264, 8602 ± 8605.
Sugai M, Chen CH and Wu HC. (1992a). Proc. Natl. Acad.
Sci. USA,89, 8903 ± 8907.
Sugai M, Hashimoto K, Kikuchi A, Inoue S, Okumura H,
Matsumoto K, Goto Y, Ohgai H, Moriishi K, Syuto B and
et al. (1992b). J. Biol. Chem.,267, 2600 ± 2604.
Symons M. (1996). Trends Biochem. Sci.,21, 178 ± 181.
Takai Y, Sasaki T, Tanaka K and Nakanishi H. (1995).
Trends Biochem. Sci.,20, 227 ± 231.
TakaishiK,KikuchiA,KurodaS,KotaniK,SasakiTand
Takai Y. (1993). Mol. Cell. Biol.,13, 72 ± 79.
Takaishi K, Sasaki T, Kameyama T, Tsukita S, and Takai Y.
(1995). Oncogene,11, 39 ± 48.
Tapon N and Hall A. (1997). Curr. Opin. Cell Biol.,9, 86 ±
92.
Thorburn J, Xu SC and Thorburn A. (1997). EMBO J.,16,
1888 ± 1900.
Thorne J and Bobak D. (1997). Submitted for publication.
Tominaga T, Sugie K, Hirata M, Morii N, Fukata J, Uchida
A, Imura H and Narumiya S. (1993). J. Cell Biol.,120,
1529 ± 1537.
Varner JA and Cheresh DA. (1996). Curr. Opin. Cell Biol.,8,
724 ± 730.
Xiong C, Levis R, Shen P, Schlesinger S, Rice C and Huang
HV. (1989). Science,243, 1188 ± 1191.
Zhang Z, Vuori K, Reed JC and Ruoslahti E. (1995). Proc.
Natl. Acad. Sci. USA,92, 6161 ± 6165.
Zigmond SH. (1996). Curr. Opin. Cell Biol.,8, 66 ± 73.
Inactivation of Rho disrupts cellular attachment and induces apoptosis
DA Bobak et al
2189
