Binding of GEF-H1 to the tight junction-associated adaptor cingulin results in inhibition of Rho signaling and G1/S phase transition.
ABSTRACT The activity of Rho GTPases is carefully timed to control epithelial proliferation and differentiation. RhoA is downregulated when epithelial cells reach confluence, resulting in inhibition of signaling pathways that stimulate proliferation. Here we show that GEF-H1/Lfc, a guanine nucleotide exchange factor for RhoA, directly interacts with cingulin, a junctional adaptor. Cingulin binding inhibits RhoA activation and signaling, suggesting that the increase in cingulin expression in confluent cells causes downregulation of RhoA by inhibiting GEF-H1/Lfc. In agreement, RNA interference of GEF-H1 or transfection of GEF-H1 binding cingulin mutants inhibit G1/S phase transition of MDCK cells, and depletion of cingulin by regulated RNA interference results in irregular monolayers and RhoA activation. These results indicate that forming epithelial tight junctions contribute to the downregulation of RhoA in epithelia by inactivating GEF-H1 in a cingulin-dependent manner, providing a molecular mechanism whereby tight junction formation is linked to inhibition of RhoA signaling.
- SourceAvailable from: Sandra Citi[show abstract] [hide abstract]
ABSTRACT: Cingulin, a component of vertebrate tight junctions, contains a head domain that controls its junctional recruitment and protein interactions. To determine whether lack of junctional cingulin affects tight-junction organization and function, we examined the phenotype of embryoid bodies derived from embryonic stem cells carrying one or two alleles of cingulin with a targeted deletion of the exon coding for most of the predicted head domain. In homozygous (-/-) embryoid bodies, no full-length cingulin was detected by immunoblotting and no junctional labeling was detected by immunofluorescence. In hetero- and homozygous (+/- and -/-) embryoid bodies, immunoblotting revealed a Triton-soluble, truncated form of cingulin, increased levels of the tight junction proteins ZO-2, occludin, claudin-6 and Lfc, and decreased levels of ZO-1. The +/- and -/- embryoid bodies contained epithelial cells with normal tight junctions, as determined by freeze-fracture and transmission electron microscopy, and a biotin permeability assay. The localization of ZO-1, occludin and claudin-6 appeared normal in mutant epithelial cells, indicating that cingulin is not required for their junctional recruitment. Real-time quantitative reverse-transcription PCR (real-time qRT-PCR) showed that differentiation of embryonic stem cells into embryoid bodies was associated with up-regulation of mRNAs for several tight junction proteins. Microarray analysis and real-time qRT-PCR showed that cingulin mutation caused a further increase in the transcript levels of occludin, claudin-2, claudin-6 and claudin-7, which were probably due to an increase in expression of GATA-6, GATA-4 and HNF-4alpha, transcription factors implicated in endodermal differentiation. Thus, lack of junctional cingulin does not prevent tight-junction formation, but gene expression and tight junction protein levels are altered by the cingulin mutation.Journal of Cell Science 11/2004; 117(Pt 22):5245-56. · 5.88 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Across the animal kingdom the apical junction complex of epithelial cells creates both a permeability barrier and cell polarity. Although based on overlapping and evolutionarily conserved proteins, the cell-cell contacts of nematodes, flies and mammals appear to differ in morphology and functional organization. Emerging evidence shows that the selective pore-like properties of vertebrate and invertebrate barriers are created by the claudin family. Similarly, assembly of the barriers requires a conserved set of polarity-generating protein complexes, particularly the PAR protein complexes.Current Opinion in Cell Biology 05/2004; 16(2):140-5. · 11.41 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Members of the Dbl-homology (DH) family of proteins promote guanine nucleotide exchange on Rho GTPases. Lbc, which specifically acts on Rho but not Rac or Cdc42, was isolated as a transforming oncogene and is composed of a DH and a Pleckstrin-homology (PH) domain. We show here that the Lbc DH domain alone is capable of stimulating new DNA synthesis in quiescent fibroblasts and Rho-dependent actin stress fiber assembly. However, the PH domain is required for subcellular localization of Lbc along actin stress fibers and for efficient transformation of NIH3T3 cells. The results show that, in contrast to other Dbl-homology proteins, the PH domain of Lbc is dispensable for activation of Rho in vivo. The PH domain-dependent subcellular localization of Lbc may, however, be important for growth factor activation of endogenous Lbc and for oncogenic transformation.Oncogene 01/1998; 15(23):2827-31. · 7.36 Impact Factor
Developmental Cell, Vol. 8, 777–786, May, 2005, Copyright ©2005 by Elsevier Inc. DOI 10.1016/j.devcel.2005.03.003
Binding of GEF-H1 to the Tight Junction-Associated
Adaptor Cingulin Results in Inhibition
of Rho Signaling and G1/S Phase Transition
Saima Aijaz,1Fabio D’Atri,2Sandra Citi,2,3
Maria S. Balda,1,* and Karl Matter1,*
1Division of Cell Biology
Institute of Ophthalmology
University College London
2Department of Molecular Biology
University of Geneva
3Dipartimento di Biologia
Università di Padova
components and the cytoskeleton (Anderson et al.,
2004; Gonzalez-Mariscal et al., 2003).
Rho GTPases are molecular switches that are impor-
tant components of many subcellular signaling pro-
cesses that govern cell proliferation and differentiation
(Etienne-Manneville and Hall, 2002; Ridley, 2004). In
their GTP-bound state, they can bind effector mole-
cules that activate downstream components; in their
GDP-bound state, they are inactive. Activation is cata-
lyzed by guanine nucleotide exchange factors (GEFs)
that stimulate the exchange of GDP by GTP and inacti-
vation by GTPase-activating proteins that promote GTP
hydrolysis. The spatial and temporal control of signal-
ing by Rho GTPases is thought to be determined by
regulating the localization and activation of these regu-
lators at specific subcellular sites, but our knowledge
about these processes is still very limited (Etienne-
Manneville and Hall, 2002).
In epithelial cells, confluence is paralleled by a reduc-
tion of active RhoA levels and activation of Rac1 and
Cdc42 (Braga, 2002; Fukata and Kaibuchi, 2001; Noren
et al., 2001). Activation of Rac1 and Cdc42 is triggered
by E-cadherin engagement and promotes formation of
the junctional complex. Downregulation of RhoA at cell
confluence is observed in different cell types, resulting
in inhibition of G1/S phase progression (Coleman et al.,
2004). One way of inhibiting RhoA signaling is activa-
tion of p190RhoGAP, which has been observed upon
cadherin engagement by immobilized ligands in trans-
fected CHO cells, which do not form a junctional com-
plex (Noren et al., 2003). The relevance of GEFs for the
confluence-dependent regulation of RhoA and the con-
tribution of mechanisms not associated with cadherins
Here we focus on GEF-H1/Lfc, a GEF for Rho that
associates with tight junctions in epithelial cells and
regulates paracellular permeability (Benais-Pont et al.,
2003). GEF-H1, originally cloned in mice and called Lfc,
is an oncoprotein of the Dbl family that activates RhoA
but not Rac1 or Cdc42 (Glaven et al., 1996; Krendel et
al., 2002; Ren et al., 1998). GEF-H1 can associate with
different cytoskeletal structures, microtubules as well
as the actin cytoskeleton, and has been proposed to
mediate crosstalk between the two types of filaments
(Benais-Pont et al., 2003; Glaven et al., 1999; Krendel
et al., 2002; Ren et al., 1998). The molecular basis for
the association of GEF-H1 with different cytoskeletal
structures is unknown. We now show that GEF-H1
binds to the F-actin binding junctional adaptor cingulin.
Cingulin binding inhibits GEF-H1 and, hence, results in
downregulation of RhoA and inhibition of G1/S phase
transition, providing a molecular mechanism whereby
tight junction formation is linked to RhoA inactivation.
The activity of Rho GTPases is carefully timed to con-
trol epithelial proliferation and differentiation. RhoA
is downregulated when epithelial cells reach conflu-
ence, resulting in inhibition of signaling pathways
that stimulate proliferation. Here we show that GEF-
H1/Lfc, a guanine nucleotide exchange factor for
RhoA, directly interacts with cingulin, a junctional
adaptor. Cingulin binding inhibits RhoA activation
and signaling, suggesting that the increase in cingulin
expression in confluent cells causes downregulation
of RhoA by inhibiting GEF-H1/Lfc. In agreement, RNA
interference of GEF-H1 or transfection of GEF-H1
binding cingulin mutants inhibit G1/S phase transi-
tion of MDCK cells, and depletion of cingulin by regu-
lated RNA interference results in irregular mono-
layers and RhoA activation. These results indicate
that forming epithelial tight junctions contribute to
the downregulation of RhoA in epithelia by inactiva-
ting GEF-H1 in a cingulin-dependent manner, provid-
ing a molecular mechanism whereby tight junction
formation is linked to inhibition of RhoA signaling.
The epithelial junctional complex mediates adhesion
and regulates cell proliferation and differentiation (Balda
and Matter, 2003). Tight junctions are the most apical
component of the junctional complex and separate the
apical from the basolateral membrane (Cereijido et al.,
2000; Schneeberger and Lynch, 2004; Tsukita et al.,
2001). They regulate paracellular permeability and re-
strict apical/basolateral intramembrane diffusion of lip-
ids. Multiple signaling components have been localized
to epithelial tight junctions, some of which function in
the regulation of epithelial polarization, differentiation,
and growth control (Matter and Balda, 2003). These sig-
naling components interact with different types of
adaptor proteins that also bind junctional membrane
GEF-H1 Interacts with the F-Actin Binding
GEF-H1 can associate with two different types of actin-
based structures in different cell types: tight junctions
*Correspondence: firstname.lastname@example.org (M.S.B.); email@example.com
in epithelial cells and stress fibers in fibroblasts (Be-
nais-Pont et al., 2003). We therefore tested whether
tight junction-associated actin binding proteins copre-
cipitate with endogenous GEF-H1. Figure 1A shows
that GEF-H1 was precipitated by monoclonal antibody
(mAb) B4/7. We then blotted the same samples with
antibodies specific for cingulin, a tight junction-associ-
ated F-actin binding protein (Citi et al., 1988; D’Atri and
Citi, 2001). Anti-cingulin antibodies detected a band
of approximately 150 kDa in B4/7 immunoprecipi-
tates, suggesting that cingulin exists in a complex with
We next determined the domain of GEF-H1 required
for complex formation with cingulin using GEF-H1/GST
fusion proteins (Figure 1B). Glutathione beads loaded
with equal amounts of GEF-H1/GST fusion proteins
were incubated with MDCK cell extracts. Specific pre-
cipitation of full-length cingulin was only observed with
constructs containing the PH domain (Figure 1C). The
anti-cingulin antibody also recognized a band of ap-
proximately 70 kDa that was only present in pull-downs
with fusion proteins containing the PH domain. Since
this antibody had been generated against a recombi-
nant protein containing the rod and tail domains (Figure
1D), this suggests that the PH domain of GEF-H1 binds
to either one of these two cingulin domains. This was
confirmed with experiments with recombinant GST fu-
sion proteins containing different regions of cingulin
that were tested for pull-down of recombinant His6-
tagged PH domain. The PH domain of GEF-H1 was effi-
ciently precipitated by a cingulin fusion protein contain-
ing residues 782 to 1025, suggesting that the GEF-H1/
cingulin interaction is due to direct binding of GEF-H1’s
PH domain to the cingulin rod domain (Figures 1E and
1F). Because GEF-H1 and cingulin colocalize at inter-
cellular junctions (Figure 1G), GEF-H1/cingulin com-
plexes are likely to be primarily associated with tight
We next tested whether cingulin can influence the
distribution of GEF-H1. When MDCK cells were trans-
fected with myc-tagged wild-type cingulin, the trans-
fected protein was found to colocalize with GEF-H1 at
cell-cell junctions (Figure 2A). In cells that stained more
brightly, myc-cingulin was not only closely associated
with cell-cell junctions but could also be seen in the
cytosol. This behavior was enhanced by removal of the
head domain (myc-cingulinR+T), a mutation known to
inhibit junctional recruitment of cingulin and to induce
cytosolic aggregates (D’Atri et al., 2002). Endogenous
GEF-H1 colocalized with both cingulin constructs. Ex-
pression of a construct containing cingulin residues
782-1025, which include the GEF-H1 binding site,
yielded a diffuse GEF-H1 distribution. Transfection of
the cingulin head domain did not affect the distribution
of GEF-H1. Staining for GEF-H1 generally appeared to
be brighter in cells expressing a GEF-H1 binding cin-
gulin fragment. This was not due to a crossreaction of
the secondary antibodies since it was not observed
when the anti-GEF-H1 antibody was omitted (not
shown). Immunoblotting did not reveal upregulation of
GEF-H1 expression. Although this might be due to a
low transfection efficiency, it is likely that cytosolic
GEF-H1 was more efficiently labeled. ZO-1, another
tight junction protein known to associate with cingulin,
Figure 1. GEF-H1 Directly Interacts with Cingulin
(A) Association of endogenous GEF-H1 and cingulin. MDCK cell
extracts were immunoprecipitated with the anti-GEF-H1 mAb B4/7
(IP B4/7), a control antibody (mAb), or empty beads. The precipi-
tates and total cell extract were analyzed by immunoblotting using
the anti-GEF-H1 antibody B4/7 or an anti-cingulin polyclonal anti-
body. The positions of GEF-H1, cingulin, and the antibody heavy
chain (γ) are indicated.
(B) Domain map of GEF-H1. The C1, the Dbl homology (DH), the
pleckstrin homology (PH), and the C-terminal (CTD) domain are in-
(C) Mapping of the interacting domain in GEF-H1. MDCK cell ex-
tracts were loaded on beads conjugated with either GST or GST
fusion proteins containing different domains of GEF-H1. The pull-
downs were analyzed by immunoblotting for cingulin using an anti-
body generated against the C-terminal half of cingulin. The asterisk
marks the position of a cingulin degradation product.
(D) Domain map of cingulin. Indicated are the head, the central rod,
and the C-terminal tail domain along with the number of the amino
acid residues forming the domain borders.
(E and F) Identification of the GEF-H1 binding site in cingulin. Pull-
down assays with cingulin/GST fusion proteins of recombinant
His6-PH domain of GEF-H1 were analyzed by immunoblotting with
an anti-His6or anti-GST antibodies. (F) Increasing concentrations
of His6-PH domain were added to test saturation of binding.
(G) Colocalization of GEF-H1 and cingulin at the junctional com-
plex. Semiconfluent MDCK cells were stained with a mAb against
GEF-H1 (FITC) and a rabbit polyclonal antibody specific for cingulin
(Cy3). The sample was analyzed by optical sectioning with a confo-
cal microscope. Shown is an xy-section taken at the level of tight
junctions and an xz-section to show the distribution along the
z-axis. The colored panels represent the corresponding overlays.
Tight Junctions and Regulation of Rho
cingulin can thus interact in vivo and this interaction is
sufficiently strong to affect the distribution of GEF-H1.
Because the PH domain of GEF-H1 interacts with
cingulin, it might be sufficient for junctional recruitment.
Although a VSV-tagged PH domain was partially tar-
geted to intercellular junctions, the transfected protein
also localized to cytoplasmic aggregates (Figure 2D).
Mutation of tryptophane-563 (PHW563A-VSV), a resi-
due that is conserved in PH domains, or deletion of the
PH domain in full-length GEF-H1 resulted in proteins
that accumulated in the cytosol (Figure 2E), suggesting
that the PH domain is important for junctional re-
The N-Terminal Domains of GEF-H1 Regulate
Since overexpressed GEF-H1 is not targeted efficiently
to junctions, we expressed N- and C-terminally trun-
cated GEF-H1-VSV to determine whether a specific do-
main inhibits junctional recruitment. Removal of the
C-terminal domain (GEF-?CTD-VSV) did not affect the
distribution of the protein: This mutant protein was still
associated with filaments. Cells expressing GEF-?CTD-
VSV appeared more spread out and flatter; this appear-
ance was paralleled by increased RhoA activation (not
shown) as in other cell lines (Glaven et al., 1996; Kren-
del et al., 2002). In contrast, removal of the N-terminal
domain (GEF-?NTD-VSV), which contains the C1 do-
main, eliminated the filamentous staining, and a signifi-
cant fraction of the protein colocalized with cingulin at
junctions. Additionally, a nonfilamentous cytoplasmic
pool was observed. These data indicate that the N ter-
minus regulates junctional recruitment of GEF-H1,
which is in agreement with the observation that muta-
tion of the Zinc-fingers in the C1 domain impedes
microtubule binding (Krendel et al., 2002).
We next tested whether the N terminus is sufficient
for microtubule binding. A construct containing the C1
and the intervening domain (C1/ID-VSV) was recruited
to microtubules and induced microtubule bundling; the
latter was not observed with the full-length construct
(Figure 3B). Constructs containing either one of these
domains alone remained cytosolic and often aggre-
gated (not shown).
Since the N-terminal domain is sufficient and re-
quired for microtubule binding but a construct lacking
the PH domain does not bind microtubules (Figure 2E),
it is possible that the two domains interact with each
other and thereby modulate targeting to microtubules
and junctions. A recombinant His6-tagged PH domain
construct was indeed pulled down by a GST fusion pro-
tein containing the C1 domain but not by GST alone
(Figure 3C). Binding of the PH domain to a GST fusion
protein containing the C-terminal domain was not ob-
served (not shown). Thus, the C1 and PH domains of
GEF-H1 can interact, providing a possible explanation
for the observed effects on each other’s targeting activ-
Figure 2. Redistribution of GEF-H1 by Transfection of Cingulin
(A and B) Wild-type MDCK cells or a cell line overexpressing VSV-
tagged GEF-H1 (C) were transfected with myc-cingulin or truncated
mutants containing the indicated domains. The cells were pro-
cessed for double immunofluorescence with a rabbit anti-myc anti-
body to detect transfected cingulin and either monoclonal antibod-
ies against GEF-H1 (A), ZO-1 (B), or the VSV-epitope to stain
transfected GEF-H1 (C). In (C), myc-cingulinR+T transfection, a
GEF-H1-VSV-expressing (arrow) and a nonexpressing (asterisk) cell
are shown. (D) VSV-tagged PH domain of GEF-H1 was transiently
expressed in MDCK cells, and the cells were stained with anti-VSV
and anti-cingulin. Shown are confocal XY, upper two panels, and
XZ sections, lower two panels. (E) cDNAs coding for VSV-tagged
PH domain with a W563 to A substitution or a GEF-H1 mutant with
an internal deletion of the PH domain were transiently transfected
and localized by immunofluorescence.
did not coaggregate with truncated cingulin, indicating
that the observed redistribution of GEF-H1 was not due
to a general effect on junctional proteins (Figure 2B).
To further test antibody specificity, we transfected
full-length and truncated myc-cingulin into a cell line
overexpressing VSV-tagged GEF-H1. A considerable
fraction of overexpressed GEF-H1 was associated with
microtubules as observed previously (Figure 2C) (Be-
nais-Pont et al., 2003; Krendel et al., 2002). In cells
expressing myc-cingulin, GEF-H1-VSV was recruited
more efficiently to intercellular junctions (middle pan-
els). The colocalization of GEF-H1-VSV and myc-cin-
gulin was not due to antibody crossreaction since it
was only observed in cells expressing GEF-H1-VSV
(Figure 2C, myc-cinR+T: GEF-H1-VSV expressing (ar-
row) and nonexpressing (asterisk) cells are shown that
both contain large cingulin aggregates). GEF-H1 and
Cingulin Inhibits RhoA Activation by GEF-H1
RhoA is downregulated when cells reach confluence
and cease to proliferate (Coleman et al., 2004). Since
cingulin expression increases with cell density whereas
loss of FRET and, hence, reduced YFP emission. Coex-
pression of full-length cingulin with the FRET probe re-
sulted in increased YFP emission, indicating reduced
Rho-GTP levels (Figure 4D). Only mutants containing
the GEF-H1 binding site (myc-cingulin R+T, myc-cin-
gulin 782-1025) increased YFP emission. Although Rho
inhibition was significant, the increase in YFP emission
was more pronounced when Rho was inhibited by add-
ing TAT-C3, a membrane permeable version of C3
transferase (Coleman et al., 2001). In contrast, transfec-
tion of GEF-H1 reduced YFP emission indicative of Rho
activation. Cotransfection of full-length, myc-cingulin
R+T and myc-cingulin 782-1025 again counteracted
Rho activation by GEF-H1, supporting the conclusion
that cingulin inhibits GEF-H1.
To test whether inhibition of Rho activation depends
on GEF-H1, we made use of a cell line permitting the
tetracycline-regulated depletion of the exchange factor
by RNA interference (Figure 4E). In GEF-H1-depleted
cells, transfection of cingulin no longer affected the
YFP/CFP ratio, suggesting that inhibition required the
presence of GEF-H1 (Figure 4F).
Since the above-used FRET probe can respond to
other Rho GTPases than RhoA, we used a similar probe
consisting of YFP separated from CFP by RhoA fused
to a Rho binding domain, resulting in FRET when the
GTPase is in the GTP-bound form (Yoshizaki et al.,
2003). Analogous probes specific for Rac1 and Cdc42
were used for comparison. RhoA inactivation was ob-
served when cingulin constructs containing the GEF-
H1 binding site were cotransfected (Figure 4G). No ef-
fects on the Rac1 probe were observed, and the Cdc42
probe was only affected by the head domain, which
stimulated Cdc42 activation, suggesting that the stimu-
lation of SRE-driven transcription (Figure 4C) might
have been mediated by Cdc42.
Figure 3. The C1 Domain of GEF-H1 Modulates Junctional
(A) MDCK cell lines expressing GEF-H1-VSV or truncated proteins
lacking either the C- (GEF-?CTD-VSV) or the N-terminal domain
(GEF-?NTD-VSV) under the control of a tetracycline-regulated pro-
moter were induced for 24 hr. The cells were processed for double
immunofluorescence using anti-VSV and anti-cingulin antibodies.
(B) MDCK cells were transiently transfected with cDNAs coding for
either GEF-H1-VSV or a construct containing the N-terminal do-
main of GEF-H1, which contains the C1 domain as well as the se-
quences between the C1 and DH domains (C1/ID-VSV). After 24 hr,
the cells were stained with antibodies against the VSV-epitope and
(C) In vitro binding of the C1 to the PH domain. GST with or without
the C1 domain were bound to glutathione beads and used to pre-
cipitate recombinant His6-PH domain of GEF-H1. The pull-downs
were analyzed by immunoblotting with an anti-His6monoclonal an-
expression of GEF-H1 remains comparatively constant
(Figure 4A), the interaction between cingulin and GEF-
H1 might be part of a mechanism that inactivates RhoA
when epithelial cells reach confluence. Indeed, trans-
fection of a GEF-H1 binding cingulin fragment resulted
in a reduction of stress fibers, suggesting inhibition of
RhoA (Figure 4B). To test whether increased expression
of cingulin affects GEF-H1 activity, we measured tran-
scriptional activity of SRF (Hill et al., 1995). Because
MDCK cells were found not to stimulate SRE-driven
transcription in response to RhoA activation, we used
the retinal pigment epithelium cell line ARPE-19 as a
reporter system. Transfection of myc-cingulin and of
the GEF-H1 binding mutant myc-cingulinR+T inhibited
SRE-driven transcription (Figure 4C). Transfection of
the cingulin head domain stimulated the response
whereas expression of the GEF-H1 binding myc-cin-
gulin 782-1025 fragment was sufficient to inhibit tran-
scription. Importantly, expression of GEF-H1 stimulated
luciferase expression, and cotransfection of full-length
cingulin or mutants that contained the GEF-H1 binding
site inhibited, indicating that cingulin can inhibit GEF-
To monitor Rho activation directly, we used a FRET-
based assay that makes use of a fusion protein in which
YFP is separated from CFP by a Rho binding domain
(Yoshizaki et al., 2003). Binding of Rho-GTP results in a
GEF-H1 Regulates G1/S Phase Transition
In fibroblasts as well as epithelial cells, RhoA signaling
regulates G1/S phase transition (Auer et al., 1998; Cole-
man et al., 2004; Liberto et al., 2002; Olson et al., 1998).
Since tight junctions regulate this step of the cell cycle
(Balda et al., 2003), we tested whether GEF-H1 is im-
portant for G1/S phase transition in MDCK cells using
tetracycline-regulated depletion of GEF-H1 by RNA in-
terference. Cells were plated at low density and syn-
chronized in G1 phase in medium with low serum, a
treatment that results in efficient accumulation of
MDCK cells in G1 phase (Balda et al., 2003). G1/S phase
transition was then stimulated by the addition of serum
and cells synthesizing DNA were visualized by bromo-
deoxyuridine labeling. More than 70% of cells entered
S phase in wild-type MDCK cells (Figure 5A). In con-
trast, GEF-H1 depletion inhibited G1/S transition as
only about 30% of the cells incorporated bromodeoxy-
uridine. GEF-H1 is thus required for efficient serum-
induced G1/S phase transition in MDCK cells.
We next tested whether inhibition of GEF-H1 by cin-
gulin is sufficient to inhibit G1/S phase transition. Trans-
fected MDCK cells were synchronized and labeled with
bromodeoxyuridine. The cells were then double labeled
for transfected myc-tagged cingulin and bromodeoxy-
uridine, and the fractions of cells stained with either
Tight Junctions and Regulation of Rho
Figure 4. Inhibition of GEF-H1 and Rho Activation by Cingulin
(A) MDCK cells grown to 20%, 50%, or 100% confluence were lyzed and equal amounts of protein were analyzed by immunoblotting for the
expression of GEF-H1, cingulin, and, as a loading control, α-tubulin. Note the pronounced upregulation of cingulin with cell confluence.
(B) MDCK cells were transiently transfected with a truncated cingulin construct lacking the head domain (myc-cingulinR+T). Cells were
stained with antibodies against myc, GEF-H1, and fluorescent phalloidin. Note the reduced appearance of stress fibers in transfected cells.
(C) Inhibition of SRE-driven transcription by cingulin. ARPE-19 cells were cotransfected with a plasmid containing a SRE driving firefly
luciferase expression, one with a control promoter regulating renilla luciferase expression, and the indicated expression vectors. After 30 hr,
the luciferases were assayed and the ratios of the values for firefly divided by those from renilla luciferase calculated. Shown are the
means ± 1 SD of a typical experiment performed in triplicates. Both panels were normalized to plasmid controls without GEF-H1 cotransfec-
tion. Asterisks indicate p values smaller than 0.05 that were calculated with two-tailed t tests comparing the single transfections to plasmid
controls and the double transfections to GEF-H1 transfections.
(D) Inhibitin of RhoA activation by cingulin. MDCK cells were transfected with pRaichu-RBD, a Rho-specific FRET probe, and the indicated
expression vectors. TAT-C3 labels cells that were incubated with TAT-modified C3 between the transfection and cell lysis. After cell lysis, the
emission for YFP (530 nm) and CFP (475 nm) was measured with an excitation wavelength of 430 nm, and the ratios were calculated. Shown
are the means ± 1 SD of a typical experiment (n = 4). Asterisks label p values smaller than 0.05 that refer to comparisons of single transfections
with plasmid controls and double transfections with GEF-H1-transfected samples. Note, increased YFP emission indicates Rho inactivation.
(E) Downregulation of GEF-H1 in MDCK cells by tetracycline-regulated RNA interference. Confluent MDCK cells expressing control RNA
duplexes or GEF-H1-directed RNA duplexes were treated with tetracycline for 3 days. Expression of GEF-H1 and α-tubulin was then analyzed
by immunoblotting with monoclonal antibodies B4/7 and 1A2, respectively.
(F) The Rho-specific FRET probe pRaichu-RBD was transfected into tetracycline treated control or GEF-H1 RNAi cells together with empty
expression vector or full-length or truncated cingulin. The extracts were then analyzed as in (D) (asterisks, p < 0.05).
(G). FRET probes specific for RhoA, Rac1, or Cdc42 were transfected into MDCK cells together with the indicated cingulin constructs
(asterisks, p < 0.05). Note, decreased YFP emission indicates inactivation.
one or both antibodies were determined. Myc-cingulin
efficiently inhibited G1/S phase transition and so did
the expression of the construct containing the rod and
the tail domain (Figure 5B). Whereas expression of the
head domain alone did not have an effect, transfection
of myc-cingulin 782-1025, which binds and inhibits
GEF-H1, inhibited bromodeoxyuridine incorporation.
These observations suggest that increased expression
of cingulin inhibits G1/S phase transition and that this
inhibitory function maps to the GEF-H1 binding site.
Cingulin Depletion Activates RhoA Signaling
To test whether endogenous cingulin is indeed impor-
tant for the regulation of RhoA signaling, we depleted
depletion thus results in activation of RhoA and stimu-
lation of RhoA signaling.
We present evidence that the RhoA exchange factor
GEF-H1 interacts with cingulin, resulting in inhibition of
the GEF. Because cingulin expression increases with
increasing cell confluence (Figure 4), the cingulin/GEF-
H1 interaction provides a mechanism that links regula-
tion of RhoA signaling to cell confluence.
Regulation of GEF-H1 and RhoA
Signaling by Cingulin
Cingulin binds to the PH domain of GEF-H1, a domain
important for GEF-H1’s transforming activity (White-
head et al., 1995). PH domain-mediated interactions are
important for the targeting of several Dbl family mem-
bers to specific subcellular sites, including the actin cy-
toskeleton, and mediate interactions with proteins and
lipids (Bellanger et al., 2000; Hoffman and Cerione,
2002; Olson et al., 1997; Zheng, 2001). PH domains of
some GEFs bind phosphoinositides; however, binding
affinities are relatively low and phosphoinositides may
not be required for membrane targeting (Baumeister et
al., 2003; Snyder et al., 2001). The PH domain of GEF-
H1 does not bind to the main plasma membrane phos-
phoinositides in vitro (not shown).
Junctional targeting of GEF-H1 is regulated by the
N-terminal domain, which binds microtubules as well
as the PH domain. In the full-length protein, the PH do-
main is required for microtubule binding, suggesting
that the interaction between the PH domain and the C1
domain is part of a mechanism that regulates recruit-
ment to different cytoskeletal structures. The C1 do-
main might also have other interaction partners such
as diacylglycerol. However, we have not been able to
detect binding of this lipid (not shown).
Binding to cingulin as well as microtubules inhibits
GEF-H1 function (Krendel et al., 2002). It will thus be
important to understand how the C1/PH domain in-
teraction is modulated to determine the sizes of these
pools as well as the amount of free GEF. Our data indi-
cate that the level of cingulin expression is an important
determinant of GEF-H1-mediated Rho regulation. It is
possible that other parameters affect GEF-H1 activity
as well. GEF-H1 can be phosphorylated by Pak1, re-
sulting in binding to 14-3-3 (Zenke et al., 2004). How-
ever, Pak1 phosphorylation and 14-3-3 binding do not
appear to affect the activity or localization of GEF-H1
(Zenke et al., 2004).
Cingulin binds to several TJ proteins and to F-actin
(Cordenonsi et al., 1999; D’Atri and Citi, 2001). Cingulin
expression in several cell types increases upon inhibi-
tion of histone deacetylase, a treatment that can induce
differentiation (Bordin et al., 2004). This is supported by
the observation that cingulin expression increases with
cell density (Figure 4). Targeted deletion of the head
domain of cingulin results in expression of a truncated
form of cingulin and altered expression of genes regu-
lating endodermal differentiation (Guillemot et al.,
2004). Although this was paralleled by altered expres-
sion of GEF-H1, it is not clear whether this was caused
Figure 5. GEF-H1 Regulates G1/S Phase Transition
(A) Wild-type, GEF-H1 RNAi, and control RNAi cells were plated on
coverslips synchronized by incubation in 0.1% serum in the pres-
ence or absence of tetracycline for 2 days. Serum was then added
back together with bromodeoxyuridine to trigger cell cycle pro-
gression and label cells in S phase. After 6 hr, the cells were fixed
and stained with antibodies against bromodeoxyuridine and
Hoechst 33258. The percentages of cells were then determined by
counting all cells and bromodeoxyuridine-positive cells in at least
12 fields per experiment and sample. Shown are means ± 1 SD of
three experiments (asterisks, p < 0.05). White bars, bromodeoxyuri-
dine positive; gray bars, bromodeoxyuridine negative.
(B) MDCK cells were transiently transfected with the indicated cin-
gulin constructs and then plated on coverslips, synchronized, and
labeled with bromodeoxyuridine as in (A). Shown are means ± 1 SD
of three experiments (asterisks, p < 0.05).
cingulin expression by RNA interference using a tetra-
cycline regulated promoter. Incubation with tetracycline
resulted in a strong reduction of cingulin expression in
cells expressing cingulin-directed RNA duplexes but
not in cells expressing control RNA duplexes (Figure
6A). The expression level of GEF-H1 was not signifi-
cantly affected by the tetracycline treatment. Cingulin-
depleted cells had an irregular morphology (Figure 6B),
reminiscent of cells expressing constitutively active
RhoA (Jou and Nelson, 1998). The remaining cingulin
still associated with cell-cell junctions, and the GEF-H1
staining became more cytoplasmic (Figure 6C). The
tight junction proteins ZO-1 and occludin as well as the
adherens junction protein β-catenin remained junction
associated (Figure 6D).
To test whether depletion of cingulin stimulates Rho
activation, we used the FRET probe. Depletion of cin-
gulin decreased YFP emission, suggesting stimulation
of Rho (Figure 7A). If activation of RhoA was monitored
with a pull-down experiment using a fusion protein con-
taining the Rho binding domain of rhotekin, increased
levels of RhoA were detected in precipitates derived
from cingulin-depleted cells (Figure 7B), indicating in-
creased activation of RhoA. RhoA signaling inhibits
myosin light chain phosphatase by phosphorylation.
Cingulin depletion indeed stimulated enhanced phos-
phorylation of the phosphatase (Figure 7C). Cingulin
Tight Junctions and Regulation of Rho
Figure 6. Regulated Depletion of Cingulin Af-
fects Cell Morphology and GEF-H1 Local-
(A) Stable MDCK cells expressing control or
cingulin-directed RNA duplexes under the
control of a tetracycline-inducible promoter
were induced for 40 hr. Expression of cin-
gulin, α-tubulin, and GEF-H1 was then ana-
lyzed by immunoblotting.
(B) Depletion of cingulin and cell morphol-
ogy. Control- and Cingulin-RNAi cells were
cultured at 40% confluence with or without
tetracycline. After 2 days, the cell morphol-
ogy was analyzed by phase contrast micro-
(C and D) Cingulin-RNAi cells were cultured
without or with tetracycline for 40 hr before
processing for immunofluorescence with
anti-GEF-H1 and anti-cingulin antibodies, or
antibodies specific for ZO-1, occludin, or
β-catenin. Note, cingulin depletion results in
increased cytoplasmic staining of GEF-H1
but not ZO-1, occludin, and β-catenin.
by the expression of the truncated form of cingulin,
which contains the GEF-H1 binding site, or an adaptive
response that occurred during the selection. However,
the present results suggest that the effects of cingulin
mutation on gene expression may in part be due to ef-
fects on Rho signaling.
kuyama et al., 2004; Irie et al., 2004; Kawakatsu et al.,
2005; Sander et al., 1998). Our observations now link
tight-junction formation with inactivation of GEF-H1
and inhibition of RhoA signaling, suggesting that the
changes in Rho GTPase signaling that occur at conflu-
ence are orchestrated by the recruitment of different
GEFs to the forming junctional complex.
Inactivation of RhoA signaling is thought to be impor-
tant for the inhibition of signaling pathways that pro-
mote proliferation in various cell types including epithe-
lial cells (Coleman et al., 2004). Our results indicate that
GEF-H1 promotes G1/S phase transition and that its
inhibition by cingulin binding regulates cell-cycle pro-
gression. Tight junctions recruit other signaling compo-
nents that have been linked to the downregulation of
signaling pathways that promote proliferation and, in
particular, G1/S phase transition. These include the tu-
mor suppressor PTEN and a complex formed by the
transcription factor ZONAB and CDK4 (Balda et al.,
2003; Matter and Balda, 2003; Wu et al., 2000). It is not
known whether these signaling systems are regulated
by RhoA. Nevertheless, the data described here link
formation of tight junctions to inactivation of RhoA sig-
naling and inhibition of cell-cycle progression, support-
ing a model according to which tight-junction assembly
serves as an indicator of epithelial cell density that pro-
gressively inhibits different proliferation promoting sig-
naling pathways with increasing cell density.
Junction Formation and RhoA Inactivation
Increased expression of cingulin results in lower levels
of active RhoA, whereas reduced cingulin expression
causes higher levels of active RhoA. These observa-
tions suggest a molecular mechanism for the inhibition
of RhoA signaling with cell confluence that is linked to
the formation of tight junctions: in low confluent cells
expression of cingulin is low and GEF-H1 is primarily
cytoplasmic; with increasing cell density, cingulin accu-
mulates at forming tight junctions resulting in seque-
stration of free GEF-H1 at tight junctions and inhibition
of RhoA signaling.
Engaging cadherins by plating transfected CHO cells
on immobilized ligands stimulates the activity of
p190RhoGAP (Noren et al., 2003). Although CHO cells
do not form a junctional complex, this suggests that
different types of intercellular junctions contribute to
the regulation of RhoA signaling: adherens junctions by
activating a Rho GAP and tight junctions by inhibiting
a Rho GEF. Our results indicate that the regulation of
GEF-H1 by cingulin makes a significant contribution to
the regulation of RhoA in epithelial cells because total
cellular levels of active RhoA were found to change sig-
nificantly in response to changes in cingulin expression.
When epithelial cells reach confluence, RhoA is inac-
tivated, whereas Rac1 and Cdc42 are activated (Braga,
2002). Strikingly, both processes appear to be medi-
ated by GEFs that are recruited to the forming junc-
tional complex. Adherens junctions recruit GEFs for
Cdc42 and Rac1, resulting in stimulation of Rac1 and
Cdc42, and stabilization of the junctional complex (Fu-
cDNA Constructs, RNA Interference, and Cell Lines
The canine GEF-H1 sequence was used for all cDNA constructs
(Benais-Pont et al., 2003). A C-terminally VSV-tagged full-length
protein was cloned into pcDNA4/TO (Invitrogen). All mutants were
generated by PCR and confirmed by sequencing. GEF-?CTD-VSV
was generated by inserting the VSV-tag after the codon for amino
acid residue 600 and GEF-?NTD-VSV by deleting the sequence
coding for residues 2 to 216. Amino acids 447 to 575 were removed
for the GEF-H1-?PH construct. For regulated expression, pcDNA6/
residues 217 to 600; PH, residues 447 to 600; CTD, residues 577
to 986. For cingulin fusion proteins, the indicated sequences were
cloned into pGEX-4T-1 and pcDNA3.1myc-His to generate C-ter-
minally tagged molecules. For regulated RNAi, the 3#-end of the
mouse U6 promoter in mU6pro was changed into a tetracycline
operator, and annealed oligonucleotides coding for RNA duplexes
were inserted (van de Wetering et al., 2003; Yu et al., 2002). The
sequence 5#-AAGGCCACCATCTATGGCATC-3# of cingulin was
targeted. For GEF-H1, the sequences were as previously described
(Benais-Pont et al., 2003). The RNAi vectors were cotransfected
with pcDNA6/TR and pCB6, and clones were selected with blastici-
din and G418.
Immunoprecipitations, Immunoblotting, and Pull-Down Assays
For immunoprecipitations, MDCK cells were extracted with 10 mM
Hepes (pH 7.4), containing 150 mM NaCl, 1% Triton X-100, 0.5%
sodium deoxycholate, 0.2% sodium dodecylsulfate, 1% Empigen
BB, and a cocktail of protease inhibitors. GEF-H1 was immuno-
precipitated with antibody B4/7 conjugated to Protein G-sepharose
(Benais-Pont et al., 2003). GST and His6-fusion proteins were gen-
erated as described (Balda and Matter, 2000; Balda et al., 1996).
For GST pull-down assays with GEF-H1 fusion proteins, cells were
extracted with 1% Triton X-100 and extracts were preadsorbed
with inactive beads for 15 min prior to incubation with glutathione-
sepharose beads coated with 15 ?g of fusion proteins. To monitor
interactions between recombinant proteins, 0.5 ?M GST-cingulin
proteins were bound to beads, washed three times with buffer S
(PBS, 1% Triton X-100, 1 mM DTT, and protease inhibitors), incu-
bated with 0.5 ?M His6-PH domain for 3 hr at 4°C in the presence
of 3% BSA, and then washed twice with buffer S and once with
PBS. Saturation of binding was tested using 0.5, 1, and 1.5 ?M
His6-PH domain. For immunoblotting, samples were separated on
6 to 15% gradient gels and then transferred to nitrocellulose mem-
branes (Balda et al., 1996). GEF-H1 was detected with mAb B4/7,
α-tubulin with mAb 1A2, cingulin with a rabbit anti-cingulin anti-
body, the VSV-epitope with mAb P5D4, the myc-epitope with mAb
9E10 (Benais-Pont et al., 2003; Cordenonsi et al., 1999; Kreis,
1987). RhoA pull-down assays were performed as described (Be-
nais-Pont et al., 2003). Antibodies specific for myosin light chain
phosphatase and phosphorylated (T696) MYPT-1 were obtained
form Upstate Biotechnology (Lake Placid, NY).
Immunofluorescence and Bromodeoxyuridine Incorporation
Cells were fixed with methanol at −20°C or 3% paraformaldehyde
and then processed for double immunofluorescence using FITC-
and Cy3-conjugated donkey secondary antibodies (Balda et al.,
1996). For triple labeling with phalloidin, FITC-phalloidin was used
together with Cy3- and Cy-5-conjugated secondary antibodies.
The following antibodies were used for immunofluorescence: GEF-
H1, mAb B4/7; cingulin, rabbit polyclonal antibody; α-tubulin, mAb
1A2; the VSV-epitope, mAb P5D4 or a rabbit anti-peptide antibody;
the myc-epitope, a rabbit polyclonal antibody (MBL Laboratories).
For bromodeoxyuridine incorporation, cells were plated on cover-
slips and incubated in 0.1% serum, resulting in accumulation of
about 75% of the cells in Go/G1 phase (Balda et al., 2003). If cells
were double labeled for transfected cingulin, anti-myc antibody in-
cubations were added after the initial methanol fixation. After
washing, the samples were fixed with 3% paraformaldehyde for 30
min and then denatured with hydrochloric acid followed by staining
with anti-bromodeoxyuridine antibodies (Balda et al., 2003). Epiflu-
orescence images were taken with a Leica DM1 RB and confocal
images with a Zeiss LSM 510 or a Leica LCS SP2 using 63× oil
Figure 7. Depletion of Cingulin and RhoA Activation
(A) MDCK cells were transfected with the Rho-specific FRET probe
and the indicated vectors. After 30 hr, emission for YFP and CFP
was measured (excitation, 430 nm) and the ratios were calculated.
Shown are the means ± 1 SD of a typical experiment (n = 6).
(B) Wild-type, two different clones of cingulin RNAi cells (CinRiTc1/
c2), as well as control RNAi cells (ConRiT) were cultured without or
with tetracycline for 40 hr. Cell extracts were then incubated with
GST with or without the Rho binding domain of rhotekin. Samples
of total cell extracts and pellets were analyzed by immunoblotting
using anti-RhoA antibodies.
(C) GEF-H1-VSV expressing cells, control RNAi cells (ConRiT), and
two different clones of cingulin RNAi cells (CinRiTc1/c2) were cul-
tured without or with tetracycline for 40 hr, lysed, and analyzed by
immunoblotting with antibodies against phosphorylated (p-MYP) or
total myosin light chain phosphatase (MYP). The graph shows
quantification of the immunoblots by densitometric scanning.
Shown are the means ± 1 SD (n = 4).
SRE Reporter and FRET Assays
Reporter assays were done by cotransfecting ARPE-19 cells with
a plasmid containing a SRE-containing promoter driving firefly lu-
ciferase expression (Clonetech), a control plasmid for renilla lucifer-
ase expression, and the indicated expression vectors (Balda and
Matter, 2000). After 30 hr, firefly luciferase was measured and stan-
dardized using the renilla luciferase values. For FRET assays,
pRaichu-RBD, pRaichu-RhoA, pRaichu-Rac1, and pRaichu-Cdc42
were cotransfected with the indicated expression and RNAi plas-
TR was cotransfected and cells were selected with blasticidin and
zeocin. For GEF-H1 domain constructs, the sequences encoding
the following residues were inserted into pcDNA4/TO, pGEX-4T-3,
or pRSET-C: C1, residues 1 to 96; C1/ID, residues 1 to 231; DH/PH,
Tight Junctions and Regulation of Rho
mids into 50% confluent MDCK cells cultured in 12-well plates (Itoh
et al., 2002; Yoshizaki et al., 2003). 30 hr after the transfection, the
cells were washed with cold PBS and lyzed with 300 ?l/well of 20
mM Tris (pH 7.5), 100 mM NaCl, 5 mM MgCl2, 0.5% Triton X-100,
and 20 ?g/ml PMSF. The lysates were transferred to 96-well plates
for centrifugation at 2000 g for 5 min. Fluorescence was measured
with a Perkin Elmer LS 50B fluorometer (excitation, 430 nm; emis-
sion, 475 and 530 nm). The ratios of emission at 530 nm/475 nm
were calculated. Significance of the results in both types of assay
was determined with two-tailed t tests.
Cereijido, M., Shoshani, L., and Contreras, R.G. (2000). Molecular
physiology and pathophysiology of tight junctions. I. Biogenesis of
tight junctions and epithelial polarity. Am. J. Physiol. 279, G477–
Citi, S., Sabanay, H., Jakes, R., Geiger, B., and Kendrick Jones, J.
(1988). Cingulin, a new peripheral component of tight junctions. Na-
ture 333, 272–276.
Coleman, M.L., Marshall, C.J., and Olson, M.F. (2004). RAS and
RHO GTPases in G1-phase cell-cycle regulation. Nat. Rev. Mol. Cell
Biol. 5, 355–366.
Coleman, M.L., Sahai, E.A., Yeo, M., Bosch, M., Dewar, A., and
Olson, M.F. (2001). Membrane blebbing during apoptosis results
from caspase-mediated activation of ROCK I. Nat. Cell Biol. 3,
Cordenonsi, M., D’Atri, F., Hammar, E., Parry, D.A., Kendrick-Jones,
J., Shore, D., and Citi, S. (1999). Cingulin contains globular and
coiled-coil domains and interacts with ZO-1, ZO-2, ZO-3, and myo-
sin. J. Cell Biol. 147, 1569–1582.
D’Atri, F., and Citi, S. (2001). Cingulin interacts with F-actin in vitro.
FEBS Lett. 507, 21–24.
D’Atri, F., Nadalutti, F., and Citi, S. (2002). Evidence for a functional
interaction between cingulin and ZO-1 in cultured cells. J. Biol.
Chem. 277, 27757–27764.
Etienne-Manneville, S., and Hall, A. (2002). Rho GTPases in cell
biology. Nature 420, 629–635.
Fukata, M., and Kaibuchi, K. (2001). Rho-family GTPases in cad-
herin-mediated cell-cell adhesion. Nat. Rev. Mol. Cell Biol. 2, 887–
Fukuyama, T., Ogita, H., Kawakatsu, T., Fukuhara, T., Yamada, T.,
Sato, T., Shimizu, K., Nakamura, T., Matsuda, M., and Takai, Y.
(2004). Involvement of the c-Src-Crk-C3G-Rap1 signaling in the
nectin-induced activation of Cdc42 and formation of adherens
junctions. J. Biol. Chem. 280, 815–825.
Glaven, J.A., Whitehead, I.P., Nomanbhoy, T., Kay, R., and Cerione,
R.A. (1996). Lfc and Lsc oncoproteins represent two new guanine
nucleotide exchange factors for the Rho GTP-binding protein. J.
Biol. Chem. 271, 27374–27381.
Glaven, J.A., Whitehead, I., Bagrodia, S., Kay, R., and Cerione, R.A.
(1999). The Dbl-related protein, Lfc, localizes to microtubules and
mediates the activation of Rac signaling pathways in cells. J. Biol.
Chem. 274, 2279–2285.
Gonzalez-Mariscal, L., Betanzos, A., Nava, P., and Jaramillo, B.E.
(2003). Tight junction proteins. Prog. Biophys. Mol. Biol. 81, 1–44.
Guillemot, L., Hammar, E., Kaister, C., Ritz, J., Caille, D., Jond, L.,
Bauer, C., Meda, P., and Citi, S. (2004). Disruption of the cingulin
gene does not prevent tight junction formation but alters gene ex-
pression. J. Cell Sci. 117, 5245–5256.
Hill, C.S., Wynne, J., and Treisman, R. (1995). The Rho family
GTPases RhoA, Rac1, and CDC42Hs regulate transcriptional acti-
vation by SRF. Cell 81, 1159–1170.
Hoffman, G.R., and Cerione, R.A. (2002). Signaling to the Rho
GTPases: networking with the DH domain. FEBS Lett. 513, 85–91.
Irie, K., Shimizu, K., Sakisaka, T., Ikeda, W., and Takai, Y. (2004).
Roles and modes of action of nectins in cell-cell adhesion. Semin.
Cell Dev. Biol. 15, 643–656.
Itoh, R.E., Kurokawa, K., Ohba, Y., Yoshizaki, H., Mochizuki, N., and
Matsuda, M. (2002). Activation of rac and cdc42 video imaged by
probes in the membrane of living cells. Mol. Cell. Biol. 22, 6582–
Jou, T.S., and Nelson, W.J. (1998). Effects of regulated expression
of mutant RhoA and Rac1 small GTPases on the development of
epithelial (MDCK) cell polarity. J. Cell Biol. 142, 85–100.
Kawakatsu, T., Ogita, H., Fukuhara, T., Fukuyama, T., Minami, Y.,
Shimizu, K., and Takai, Y. (2005). Vav2 as a Rac-GEF responsible
for the nectin-induced, c-Src- and Cdc42-mediated activation of
Rac. J. Biol. Chem. 280, 4940–4947.
Kreis, T.E. (1987). Microtubules containing detyrosinated tubulin are
less dynamic. EMBO J. 6, 2597–2606.
We thank Dr. Matsuda for the FRET probes and the members of
our laboratories for critical reading of the manuscript. This research
was supported by Cancer Research UK (CUK) grant number
C1483/A2632, The Wellcome Trust (063661 and 066100), and the
Medical Research Council. The Citi team was supported by the
Swiss National Science Foundation and the Swiss Cancer League.
Received: June 15, 2004
Revised: January 5, 2005
Accepted: March 3, 2005
Published: April 3, 2005
Anderson, J.M., Van Itallie, C.M., and Fanning, A.S. (2004). Setting
up a selective barrier at the apical junction complex. Curr. Opin.
Cell Biol. 16, 140–145.
Auer, K.L., Park, J.S., Seth, P., Coffey, R.J., Darlington, G., Abo,
A., McMahon, M., Depinho, R.A., Fisher, P.B., and Dent, P. (1998).
Prolonged activation of the mitogen-activated protein kinase path-
way promotes DNA synthesis in primary hepatocytes from p21Cip-
1/WAF1-null mice, but not in hepatocytes from p16INK4a-null mice.
Biochem. J. 336, 551–560.
Balda, M.S., and Matter, K. (2000). The tight junction protein ZO-1
and an interacting transcription factor regulate ErbB-2 expression.
EMBO J. 19, 2024–2033.
Balda, M.S., and Matter, K. (2003). Epithelial cell adhesion and the
regulation of gene expression. Trends Cell Biol. 13, 310–318.
Balda, M.S., Garrett, M.D., and Matter, K. (2003). The ZO-1 associ-
ated Y-box factor ZONAB regulates epithelial cell proliferation and
cell density. J. Cell Biol. 160, 423–432.
Balda, M.S., Whitney, J.A., Flores, C., González, S., Cereijido, M.,
and Matter, K. (1996). Functional dissociation of paracellular per-
meability and transepithelial electrical resistance and disruption of
the apical-basolateral intramembrane diffusion barrier by expres-
sion of a mutant tight junction membrane protein. J. Cell Biol. 134,
Baumeister, M.A., Martinu, L., Rossman, K.L., Sondek, J., Lemmon,
M.A., and Chou, M.M. (2003). Loss of phosphatidylinositol 3-phos-
phate binding by the C-terminal Tiam-1 pleckstrin homology do-
main prevents in vivo Rac1 activation without affecting membrane
targeting. J. Biol. Chem. 278, 11457–11464.
Bellanger, J.M., Astier, C., Sardet, C., Ohta, Y., Stossel, T.P., and
Debant, A. (2000). The Rac1- and RhoG-specific GEF domain of
Trio targets filamin to remodel cytoskeletal actin. Nat. Cell Biol. 2,
Benais-Pont, G., Punn, A., Flores-Maldonado, C., Eckert, J., Ra-
poso, G., Fleming, T.P., Cereijido, M., Balda, M.S., and Matter, K.
(2003). Identification of a tight junction-associated guanine nucleo-
tide exchange factor that activates Rho and regulates paracellular
permeability. J. Cell Biol. 160, 729–740.
Bordin, M., D’Atri, F., Guillemot, L., and Citi, S. (2004). Histone de-
acetylase inhibitors up-regulate the expression of tight junction
proteins. Mol. Cancer Res. 2, 692–701.
Braga, V.M. (2002). Cell-cell adhesion and signalling. Curr. Opin.
Cell Biol. 14, 546–556.
Krendel, M., Zenke, F.T., and Bokoch, G.M. (2002). Nucleotide ex-
change factor GEF-H1 mediates cross-talk between microtubules
and the actin cytoskeleton. Nat. Cell Biol. 4, 294–301.
Liberto, M., Cobrinik, D., and Minden, A. (2002). Rho regulates
p21(CIP1), cyclin D1, and checkpoint control in mammary epithelial
cells. Oncogene 21, 1590–1599.
Matter, K., and Balda, M.S. (2003). Signalling to and from tight junc-
tions. Nat. Rev. Mol. Cell Biol. 4, 225–236.
Noren, N.K., Niessen, C.M., Gumbiner, B.M., and Burridge, K.
(2001). Cadherin engagement regulates Rho family GTPases. J.
Biol. Chem. 276, 33305–33308.
Noren, N.K., Arthur, W.T., and Burridge, K. (2003). Cadherin engage-
ment inhibits RhoA via p190RhoGAP. J. Biol. Chem. 278, 13615–
Olson, M.F., Sterpetti, P., Nagata, K., Toksoz, D., and Hall, A. (1997).
Distinct roles for DH and PH domains in the Lbc oncogene. Onco-
gene 15, 2827–2831.
Olson, M.F., Paterson, H.F., and Marshall, C.J. (1998). Signals from
Ras and Rho GTPases interact to regulate expression of p21Waf1/
Cip1. Nature 394, 295–299.
Ren, Y., Li, R., Zheng, Y., and Busch, H. (1998). Cloning and charac-
terization of GEF-H1, a microtubule-associated guanine nucleotide
exchange factor for Rac and Rho GTPases. J. Biol. Chem. 273,
Ridley, A.J. (2004). Rho proteins and cancer. Breast Cancer Res.
Treat. 84, 13–19.
Sander, E.E., van Delft, S., ten Klooster, J.P., Reid, T., van der Kam-
men, R.A., Michiels, F., and Collard, J.G. (1998). Matrix-dependent
Tiam1/Rac signaling in epithelial cells promotes either cell-cell ad-
hesion or cell migration and is regulated by phosphatidylinositol
3-kinase. J. Cell Biol. 143, 1385–1398.
Schneeberger, E.E., and Lynch, R.D. (2004). The tight junction: a
multifunctional complex. Am. J. Physiol. 286, C1213–C1228.
Snyder, J.T., Rossman, K.L., Baumeister, M.A., Pruitt, W.M., Sider-
ovski, D.P., Der, C.J., Lemmon, M.A., and Sondek, J. (2001). Quanti-
tative analysis of the effect of phosphoinositide interactions on the
function of Dbl family proteins. J. Biol. Chem. 276, 45868–45875.
Tsukita, S., Furuse, M., and Itoh, M. (2001). Multifunctional strands
in tight junctions. Nat. Rev. Mol. Cell Biol. 2, 286–293.
van de Wetering, M., Oving, I., Muncan, V., Pon Fong, M.T.,
Brantjes, H., van Leenen, D., Holstege, F.C., Brummelkamp, T.R.,
Agami, R., and Clevers, H. (2003). Specific inhibition of gene ex-
pression using a stably integrated, inducible small-interfering-RNA
vector. EMBO Rep. 4, 609–615.
Whitehead, I., Kirk, H., Tognon, C., Trigo-Gonzalez, G., and Kay, R.
(1995). Expression cloning of lfc, a novel oncogene with structural
similarities to guanine nucleotide exchange factors and to the regu-
latory region of protein kinase C. J. Biol. Chem. 270, 18388–18395.
Wu, X., Hepner, K., Castelino-Prabhu, S., Do, D., Kaye, M.B., Yuan,
X.J., Wood, J., Ross, C., Sawyers, C.L., and Whang, Y.E. (2000).
Evidence for regulation of the PTEN tumor suppressor by a mem-
brane-localized multi-PDZ domain containing scaffold protein
MAGI-2. Proc. Natl. Acad. Sci. USA 97, 4233–4238.
Yoshizaki, H., Ohba, Y., Kurokawa, K., Itoh, R.E., Nakamura, T.,
Mochizuki, N., Nagashima, K., and Matsuda, M. (2003). Activity of
Rho-family GTPases during cell division as visualized with FRET-
based probes. J. Cell Biol. 162, 223–232.
Yu, J.Y., DeRuiter, S.L., and Turner, D.L. (2002). RNA interference by
expression of short-interfering RNAs and hairpin RNAs in mamma-
lian cells. Proc. Natl. Acad. Sci. USA 99, 6047–6052.
Zenke, F.T., Krendel, M., DerMardirossian, C., King, C.C., Bohl, B.P.,
and Bokoch, G.M. (2004). p21-activated kinase 1 phosphorylates
and regulates 14–3-3 binding to GEF-H1, a microtubule-localized
Rho exchange factor. J. Biol. Chem. 279, 18392–18400.
Zheng, Y. (2001). Dbl family guanine nucleotide exchange factors.
Trends Biochem. Sci. 26, 724–732.