The Dynamic Cell
GRAF1-dependent endocytosis
Gary J. Doherty* and Richard Lundmark†1
*MRC Laboratory of Molecular Biology, Hills Road, Cambridge CB2 OQH, U.K., and †Department of Medical Biochemistry and Biophysics, Umeå University,
901 87 Umeå, Sweden
Abstract
The role of endocytosis in controlling a multitude of cell biological events is well established. Molecular
and mechanistic characterization of endocytosis has predominantly focused on CME (clathrin-mediated
endocytosis), although many other endocytic pathways have been described. It was recently shown that the
BAR (Bin/amphiphysin/Rvs) and Rho GAP (GTPase-activating protein) domain-containing protein GRAF1
(GTPase regulator associated with focal adhesion kinase-1) is found on prevalent, pleiomorphic endocytic
membranes, and is essential for the major, clathrin-independent endocytic pathway that these membranes
mediate. This pathway is characterized by its ability to internalize GPI (glycosylphosphatidylinositol)anchored proteins, bacterial toxins and large amounts of extracellular fluid. These membrane carriers are
highly dynamic and associated with the activity of the small G-protein Cdc42 (cell division cycle 42). In the
present paper, we review the role of GRAF1 in this CLIC (clathrin-independent carrier)/GEEC (GPI-anchored
protein-enriched early endocytic compartment) endocytic pathway and discuss the current understanding
regarding how this multidomain protein functions at the interface between membrane sculpting, small
G-protein signalling and endocytosis.
Introduction
Endocytosis describes the process by which transmembrane
and membrane-associated proteins and their ligands, extracellular fluid with nutrients, and plasma membrane lipids, are
internalized by a cell. In so doing, endocytosis regulates the
concentration of lipids and proteins on the cell surface (and
therefore cell–environment interactions), as well as creating
novel internal vesicles (endosomes) from which signalling
events can occur and which are subsequently trafficked to
other regions of the cell to which they deliver their contents
by fusing with other membrane-bound compartments. Cells
are specialized for different functions, and endocytosis
plays distinct roles in each cell type, being coupled
with countless processes including neurotransmission, cell–
cell and cell–matrix adhesion, cell morphology changes,
growth factor signalling and mitosis. As well as regulating
these essential cell physiological phenomena, endocytic
pathways can also be hijacked by pathogens and toxins.
Endocytosis necessarily requires the production of highly
curved membranes from the flat plasma membrane, and
cellular proteins mediate this membrane sculpting by a
variety of mechanisms (see [1] for a review). The best-studied
Key words: Bin/amphiphysin/Rvs domain (BAR domain), clathrin-independent endocytosis,
glycosylphosphatidylinositol-linked protein (GPI-linked protein), GTPase regulator associated with
focal adhesion kinase-1 (GRAF1), oligophrenin, small G-protein.
Abbreviations used: Arf, ADP-ribosylation factor; BAR, Bin/amphiphysin/Rvs; CCV, clathrincoated vesicle; Cdc42, cell division cycle 42; CLIC, clathrin-independent carrier; CME, clathrinmediated endocytosis; CTxB, cholera toxin B subunit; FAK, focal adhesion kinase; GAP, GTPaseactivating protein; GFP, green fluorescent protein; GPI, glycosylphosphatidylinositol; GEEC, GPIanchored protein-enriched early endosomal compartment; GEF, guanine-nucleotide-exchange
factor; GRAF1, GTPase regulator associated with FAK-1; N-WASP, neuronal Wiskott–Aldrich
syndrome protein; PH, pleckstrin homology; SH3, Src homology 3; SNX9, sorting nexin 9.
1
To whom correspondence should be addressed (email richard.lundmark@medchem.
umu.se).
Biochem. Soc. Trans. (2009) 37, 1061–1065; doi:10.1042/BST0371061
endocytic pathway is CME (clathrin-mediated endocytosis),
where transmembrane receptors have cytosolic motifs (or
bind proteins that function analogously) that recruit a large
network of proteins that stimulate the endocytic process,
resulting in the formation of a densely coated, highly
curved CCV (clathrin-coated vesicle). However, it is also
evident that many other internalized proteins, including
GPI (glycosylphosphatidylinositol)-anchored proteins, use
distinct, clathrin-independent endocytic pathways, whereas
some exogenous proteins used as endocytic pathway probes,
such as CTxB (cholera toxin B subunit), enter cells via various
pathways. Many clathrin-independent endocytic pathways
have been described (see [2,3] for extensive reviews of
these pathways and their cell biological functions). These
include a highly prevalent pathway involving pleiomorphic
CLICs (clathrin-independent carriers) that deliver material
to GEECs (GPI-anchored protein-enriched early endosomal
compartments), the so-called CLIC/GEEC pathway that is
the focus of the present review. It appears that clathrinindependent pathways have similar lipid requirements and
function independently of rigid protein networks such as
those found around CCVs. Instead, these pathways have reliance on particular protein kinases, small G-proteins and associated effector proteins, implying the involvement of a more
dynamic protein assembly and endosomal generation process.
Small G-proteins and endocytic membrane
carrier formation
Small G-proteins of the Rho, Arf (ADP-ribosylation factor)
and Rab families are associated with membrane compartments via post-translational modifications and loading of
GTP. They signal via a variety of effector molecules to
C The
C 2009 Biochemical Society
Authors Journal compilation 1061
1062
Biochemical Society Transactions (2009) Volume 37, part 5
direct and co-ordinate events such as actin polymerization
and membrane remodelling, thus acting as master regulators
to facilitate processes such as membrane protrusion and
invagination, cell movement and adhesion. Cycles of GTP
loading (turning them ‘on’) and hydrolysis (turning them
‘off’) by these proteins are in turn regulated by GEFs
(guanine-nucleotide-exchange factors) and GAPs (GTPaseactivating proteins) (see [4,5] for reviews). Cycles of GTP
hydrolysis seem to be important for the progression and
maturation of individual cellular processes and function as
checkpoints. It is striking that particular small G-proteins
have been characterized as regulators of clathrin-independent
endocytic pathways [2]. This implies that these endocytic
pathways are likely to be closely coupled with other
cellular processes controlled by small G-proteins and that
endocytosis and these processes require close co-ordination.
Membrane nanodomains enriched in certain lipids (including
cholesterol) are not only known to be essential for clathrinindependent endocytic pathways but also for receptor
clustering and Rho family small G-protein localization.
Formation of such domains would therefore be an ideal way
to sense the cellular milieu and to concomitantly transmit
signals to stimulate an endocytic apparatus and one or more
small G-protein (and their effector molecules). Both RhoA
and Rac1 have been suggested to play a role in the uptake of
the interleukin-2 receptor [6,7], whereas Cdc42 (cell division
cycle 42) is necessary for the uptake of GPI-linked proteins,
CTxB and the Helicobacter pylori VacA (vacuolating
cytotoxin A) toxin via the CLIC/GEEC pathway [8–10] (see
below). Proteins of the Arf family have the ability to induce
and sense high membrane curvature in their GTP-bound
state in addition to recruiting effector molecules [11,12].
Not surprisingly therefore, these proteins are found to play
important roles in several membrane trafficking events in
the cell, including endocytosis. Arf6 has been shown to be
required for uptake of the MHC class I receptor via a tubular,
clathrin-independent pathway [13], whereas Arf1 is involved
in the distinct CLIC/GEEC pathway [14].
The multidomain GRAF [GTPase regulator
associated with FAK-1 (focal adhesion
kinase-1)] protein family
GRAF1 belongs to a subfamily of GAPs whose members
share a high degree of sequence identity and similar
domain composition and organization [with BAR
(Bin/amphiphysin/Rvs), PH (pleckstrin homology), GAP
and SH3 (Src homology 3) domains] (Figure 1A). Proteins
of this family include GRAF1–GRAF3 and oligophrenin
(which lacks the SH3 domain present in other family
members). It is interesting to note that normal GRAF1 and
oligophrenin function is required for health. Oligophrenin
mutations are frequently found in patients suffering from
X-linked mental retardation [15], whereas GRAF1 appears
to act as a tumour suppressor in leucocytes, where deletions,
truncations and mutations in both alleles have been found
associated with acute myeloid leukaemia and myelodysplastic
C The
C 2009 Biochemical Society
Authors Journal compilation syndrome [16,17]. The G:C-rich promoter of GRAF1 is
ordinarily unmethylated, but approx. 38 % of biopsies of
bone marrow from patients with these conditions exhibit
GRAF1 promoter methylation, which is associated with
reduced protein expression [18].
Proteins of the GRAF family are able to stimulate the
GTPase activity of proteins of the Rho family of small
G-proteins. This activity has been most extensively studied in
GRAF1 where the crystal structure of the GAP domain has
been solved [19,20] (see Figure 1B for a structural model of
GRAF1). In vitro studies demonstrated that the GAP domain
of GRAF1 is more active against Cdc42 and RhoA than Rac1,
and in vivo the overexpression of GRAF1 resulted in cell morphology changes that mimic those seen when RhoA activity
is inhibited [20–22]. BAR modules, which form through BAR
domain dimerization to produce a highly curved membraneinteracting module, are often found in conjunction with other
lipid-binding regions. The N-terminal BAR and PH domains
of proteins in the GRAF family localize these proteins to
lipid membranes and work together to generate and sense
membrane curvature in vitro [23], suggesting that they act
similarly in this respect to other BAR domain-containing
proteins (see Figure 1A, inset). The BAR and PH domains
of GRAF family proteins appear to work in concert since
the individual domains, or proteins harbouring specific point
mutants in either domain, are severely impaired in their
abilities to bind membranes and to localize appropriately in
cells. This type of co-operation has been shown for similar
proteins with adjacent membrane-binding domains [24], and
bestows on GRAF family members the ability to detect the
coincidence in cells of a highly curved membrane and the
presence of the plasma membrane-enriched phosphoinositide
PtdIns(4,5)P2 . Interestingly, it has been shown that the GAP
domains of GRAF1 and oligophrenin are autoinhibited by
their BAR and PH domains. This inhibition is due to a direct
interaction between these protein regions and is independent
of membrane binding of the BAR and PH domains [25].
GRAF1 was originally identified as an interactor of FAK,
which binds GRAF1’s SH3 domain [21]. Later, it was shown
that this domain also, and more strongly, interacts with the
membrane scission protein dynamin with which GRAF1
forms a tight complex, suggesting that these proteins may
function together in membrane sculpting processes in vivo
[23]. In summary, by virtue of their domain structures,
GRAF family members are perfectly designed to work at
the interface between small G-protein regulation and plasma
membrane remodelling.
Highly dynamic GRAF1-positive
tubulovesicular membranes are required
for a high-volume constitutive endocytic
pathway
The localization of endogenous GRAF1 in fibroblasts is striking. GRAF1 is found on pleiomorphic tubular and punctate
membrane structures that are easily obliterated by cold fixation (probably hampering their discovery until recently) [23].
The Dynamic Cell
Figure 1 Role of the multidomain protein GRAF1
(A) Schematic representation of the domain structure in GRAF1 with the function of each domain indicated in text. ‘PIP2’ and
‘Curvature’ indicate PtdIns(4,5)P2 -containing and curved membranes respectively. The inset shows the extensive membrane
tubular network generated in a cell overexpressing the membrane-remodelling region of GRAF1 as a GFP–BAR+PH protein.
(B) Structural model of dimeric GRAF1 with the crystallized domains depicted in ribbon representation with their corresponding
molecular modelling database (mmdb) numbers. Domains whose structures remain unsolved (BAR and PH) are depicted
as cross-hatched replicas patterned from the crystal structures of SNX9 (BAR) and ARAP (Arf GAP with Rho GAP, ankyrin
repeats and PH domains) (PH). (C) Model of the endocytic uptake of fluid, GPI-linked proteins and CTxB via a GRAF1-mediated
pathway. Receptors and toxins are clustered in nanodomains of the plasma membrane that are recognized by the small
G-proteins Arf1 and Cdc42 inside the cell. Subsequently, GRAF1 is recruited to the membrane bud, which results in the
formation of a membrane tubule that is eventually released from the plasma membrane via a scission reaction.
These structures have been shown to co-localize with a large
amount of endocytosed dextran and yet be devoid of known
markers for CME events, suggesting that they are involved
in clathrin-independent endocytosis. Indeed, depletion of
GRAF1 resulted in a major reduction in fluid-phase uptake in
unstimulated HeLa cells to an extent similar to that produced
by blocking CME. GRAF1-positive tubules contained
endocytosed material after short incubation times, indicating
that they are of an early endocytic nature. Overexpressed
wild-type GRAF1 showed a similar localization pattern and
allowed the visualization of these structures in living cells,
where they were shown to be highly dynamic membranes and
also capable of internalizing CTxB and a model GPI-linked
protein [GFP (green fluorescent protein)–GPI]. In contrast,
a truncated version of GRAF1 (comprising the lipid-binding
BAR and PH domains but lacking the C-terminal GAP and
SH3 domains) localized to much more static tubules and
puncta and inhibited the uptake of both CTxB and GFP–GPI.
The surface label of these markers co-localized with truncated
GRAF1 for long periods, suggesting that this protein
functions as a dominant-negative mutant by stabilizing static,
non-functional carrier intermediates at the cell surface.
Role of GRAF1 in the CLIC/GEEC endocytic
pathway
While these studies of GRAF1 highlighted the nature and
extent of a highly prevalent, pleiomorphic, clathrinindependent endocytic pathway using fluorescence
C The
C 2009 Biochemical Society
Authors Journal compilation 1063
1064
Biochemical Society Transactions (2009) Volume 37, part 5
microscopy approaches, electron microscopy approaches had
previously identified the CLIC/GEEC endocytic pathway
to be a major source of fluid-phase uptake [8]. These CLICs
are morphologically distinct from both CME intermediates
and caveolae in that they have a tubular and ring-like
appearance. Molecules such as GPI-linked receptors and
CTxB are internalized into these membrane structures, whose
production is dependent on activity of the small G-proteins
Cdc42 and Arf1 [8,10,14]. However, the lack of specific
proteins known to be involved in the generation of CLICs
had hampered the further characterization of this pathway.
Since the morphological features and cargo specificity
described for CLICs also apply to GRAF1-dependent
endocytosis, we propose that GRAF1 is the first non-cargo
marker and membrane remodelling protein described for the
CLIC pathway. As mentioned above, GRAF1 has the ability
to localize to membrane tubular structures via its BAR and
PH domains, and to regulate the activity of both Cdc42
and RhoA. Using electron microscopy, GRAF1-positive
carriers in cells were confirmed to contain endocytosed cargo
and shown to be approx. 40 nm in diameter [23]. Interestingly,
the BAR and PH domains of GRAF1 work together to
sculpt spherical liposomes to 40 nm diameter tubules in
vitro, suggesting that its membrane curvature modulating
ability exists similarly in vivo. Furthermore, GRAF1 was
found to co-localize with dominant-active Cdc42 at discrete
structures at the cell surface, suggesting that it functions
together with this small G-protein in endocytic membrane
manufacture.
The close interactions between cell membranes and
cytoskeletal elements have been extensively reviewed
elsewhere [26]. In addition to having likely roles in
promoting CLIC membrane manufacture, the actin- and
microtubule-based cytoskeletons are expected to be central
highways upon which CLIC/GEEC membranes traffic. It
has been shown that both N-WASP (neuronal Wiskott–
Aldrich syndrome protein; which is activated by Cdc42 and
promotes actin polymerization) and actin are involved in the
uptake of GPI-linked proteins [27]. Further, the membrane
remodelling protein SNX9 (sorting nexin 9), which interacts
with N-WASP, has been suggested to participate in clathrinindependent endocytosis [28] (alongside its canonical role
in CME). While the GRAF family member oligophrenin
interacts with actin directly (via its C-terminus) [29], GRAF1
does not bind directly to actin but participates indirectly in
this process via regulation of RhoA/Cdc42 activity. Other
proteins that play roles in the CLIC/GEEC pathway include
Arf1, and its effector ARHGAP10, which is a GAP for
Cdc42 [14]. These observations highlight the importance of
GEFs and GAPs in the regulation of membrane transport
and it is likely that many similar types of proteins are
intimately involved in carrier formation and processing, and
that they might be used differentially in distinct cell types.
Further work, using live cell microscopy to visualize the
recruitment and activation of small G-proteins (including
Cdc42 and Arf1) and their modulators (including GRAF1 and
ARHGAP10), will be necessary to elucidate how these
C The
C 2009 Biochemical Society
Authors Journal compilation proteins function spatiotemporally in the CLIC/GEEC
endocytic pathway.
Interestingly, CLIC formation and Cdc42 activation have
been shown to depend on the presence of cholesterol in the
plasma membrane [27], suggesting further that membrane
nanodomains are somehow integral to the coupling of small
G-protein activity and endocytosis. Flotillins are membrane
proteins that are proposed to form a hairpin in the plasma
membrane similarly to caveolins, are found highly enriched
in cholesterol-rich membranes and are proposed to function
in CTxB endocytosis [30]. Although tubules decorated with
wild-type GRAF1 did not accumulate flotillin1, expression
of dominant-negative GRAF1 stabilized early endocytic
tubules that were able to mix with flotillin1-positive
membrane regions [23]. This suggests that flotillin1 may act
upstream of GRAF1 in specifying domains to be internalized
into CLICs, or simply that similar types of membrane regions
are involved in these trafficking processes but that they are
not interdependent. These questions are currently being
answered using siRNA (small interfering RNA) approaches.
Little is known about the driving force for membrane
deformation and fission during CLIC generation. GRAF1
has the ability to generate membrane curvature via its BAR
domain, but compared with other BAR-containing proteins it
is relatively inefficient in this respect. It might be that GRAF1
functions more as a sensor of curvature and thereby localizes
to membrane tubules (produced by other proteins) and
functions to stabilize their high curvature. It will be necessary
to identify the lipids and proteins that collaborate in this
process in order to understand more about the mechanism
of how carriers form, and perhaps Arf1 acts upstream
in the CLIC/GEEC pathway, generating high membrane
curvature, which is then stabilized by GRAF1 (Figure 1C).
CLIC/GEEC pathway carriers form even in the presence of
a mutated dynamin protein that blocks CME. However, it is
highly likely that dynamin participates in CLIC/GEEC endocytosis but may not be absolutely required for membrane
fission as it is with CME. GRAF1 forms a complex together
with dynamin, which suggests that these proteins function
together at some stage of membrane carrier processing. The
identification of the specific dynamin isoform involved in the
CLIC/GEEC pathway will allow further characterization of
its role. CLICs mature into GEECs that can eventually fuse
with early endosomes in a Rab5-dependent manner, although
it is likely that (perhaps most of) the CLIC/GEEC membranes can also traffic directly to perinuclear compartments
such as the Golgi apparatus. The sorting and processing of
lipids and proteins in these steps will also require an additional
arsenal of regulatory proteins that remain to be identified.
Concluding remarks
While the molecular and mechanistic characterization of
clathrin-independent endocytic pathways is only beginning,
recent progress has been made on defining one such pathway,
namely the CLIC/GEEC pathway. This pathway is a major
mediator of fluid-phase and GPI-linked protein uptake and
The Dynamic Cell
can be hijacked by cholera toxin. This pathway requires the
function of the membrane sculpting, and small G-protein
modulator, GRAF1, which lines its highly curved membrane
intermediates, as well as the small G-proteins Cdc42, Arf1
and the ArfGAP ARHGAP10. Further work will identify
how these proteins function together in endocytic membrane
production, identify other endogenous and exogenous
cargoes of the CLIC/GEEC pathway, and determine the cell
biological functions of this exciting endocytic portal.
Acknowledgements
We thank Harvey McMahon (MRC Laboratory of Molecular Biology,
Cambridge, U.K.) in whose laboratory we had the privilege of
performing experiments that led to this review.
Funding
We thank the Swedish Research Council, Swedish Cancer Society
and Umeå University for funding.
References
1 McMahon, H.T. and Gallop, J.L. (2005) Membrane curvature and
mechanisms of dynamic cell membrane remodelling. Nature 438,
590–596
2 Doherty, G.J. and McMahon, H.T. (2009) Mechanisms of endocytosis.
Annu. Rev. Biochem. 78, 857–902
3 Mayor, S. and Pagano, R.E. (2007) Pathways of clathrin-independent
endocytosis. Nat. Rev. Mol. Cell Biol. 8, 603–612
4 Gillingham, A.K. and Munro, S. (2007) The small G proteins of the Arf
family and their regulators. Annu. Rev. Cell Dev. Biol. 23, 579–611
5 Jaffe, A.B. and Hall, A. (2005) Rho GTPases: biochemistry and biology.
Annu. Rev. Cell Dev. Biol. 21, 247–269
6 Grassart, A., Dujeancourt, A., Lazarow, P.B., Dautry-Varsat, A. and
Sauvonnet, N. (2008) Clathrin-independent endocytosis used by
the IL-2 receptor is regulated by Rac1, Pak1 and Pak2. EMBO Rep. 9,
356–362
7 Lamaze, C., Dujeancourt, A., Baba, T., Lo, C.G., Benmerah, A. and
Dautry-Varsat, A. (2001) Interleukin 2 receptors and detergent-resistant
membrane domains define a clathrin-independent endocytic pathway.
Mol. Cell 7, 661–671
8 Kirkham, M., Fujita, A., Chadda, R., Nixon, S.J., Kurzchalia, T.V., Sharma,
D.K., Pagano, R.E., Hancock, J.F., Mayor, S. and Parton, R.G. (2005)
Ultrastructural identification of uncoated caveolin-independent early
endocytic vehicles. J. Cell Biol. 168, 465–476
9 Gauthier, N.C., Monzo, P., Kaddai, V., Doye, A., Ricci, V. and Boquet, P.
(2005) Helicobacter pylori VacA cytotoxin: a probe for a
clathrin-independent and Cdc42-dependent pinocytic pathway routed to
late endosomes. Mol. Biol. Cell 16, 4852–4866
10 Sabharanjak, S., Sharma, P., Parton, R.G. and Mayor, S. (2002)
GPI-anchored proteins are delivered to recycling endosomes via a
distinct cdc42-regulated, clathrin-independent pinocytic pathway.
Dev. Cell 2, 411–423
11 Lee, M.C., Orci, L., Hamamoto, S., Futai, E., Ravazzola, M. and Schekman,
R. (2005) Sar1p N-terminal helix initiates membrane curvature and
completes the fission of a COPII vesicle. Cell 122, 605–617
12 Lundmark, R., Doherty, G.J., Vallis, Y., Peter, B.J. and McMahon, H.T.
(2008) Arf family GTP loading is activated by, and generates, positive
membrane curvature. Biochem. J. 414, 189–194
13 Naslavsky, N., Weigert, R. and Donaldson, J.G. (2004) Characterization of
a nonclathrin endocytic pathway: membrane cargo and lipid
requirements. Mol. Biol. Cell 15, 3542–3552
14 Kumari, S. and Mayor, S. (2008) ARF1 is directly involved in
dynamin-independent endocytosis. Nat. Cell Biol. 10, 30–41
15 Zanni, G., Saillour, Y., Nagara, M., Billuart, P., Castelnau, L., Moraine, C.,
Faivre, L., Bertini, E., Durr, A., Guichet, A. et al. (2005) Oligophrenin 1
mutations frequently cause X-linked mental retardation with cerebellar
hypoplasia. Neurology 65, 1364–1369
16 Borkhardt, A., Bojesen, S., Haas, O.A., Fuchs, U., Bartelheimer, D.,
Loncarevic, I.F., Bohle, R.M., Harbott, J., Repp, R., Jaeger, U. et al. (2000)
The human GRAF gene is fused to MLL in a unique t(5;11)(q31;q23) and
both alleles are disrupted in three cases of myelodysplastic
syndrome/acute myeloid leukemia with a deletion 5q. Proc. Natl. Acad.
Sci. U.S.A. 97, 9168–9173
17 Panagopoulos, I., Kitagawa, A., Isaksson, M., Morse, H., Mitelman, F. and
Johansson, B. (2004) MLL/GRAF fusion in an infant acute monocytic
leukemia (AML M5b) with a cytogenetically cryptic
ins(5;11)(q31;q23q23). Genes Chromosomes Cancer 41, 400–404
18 Bojesen, S.E., Ammerpohl, O., Weinhausl, A., Haas, O.A., Mettal, H.,
Bohle, R.M., Borkhardt, A. and Fuchs, U. (2006) Characterisation of the
GRAF gene promoter and its methylation in patients with acute myeloid
leukaemia and myelodysplastic syndrome. Br. J. Cancer 94,
323–332
19 Longenecker, K.L., Zhang, B., Derewenda, U., Sheffield, P.J., Dauter, Z.,
Parsons, J.T., Zheng, Y. and Derewenda, Z.S. (2000) Structure of the BH
domain from GRAF and its implications for Rho GTPase recognition.
J. Biol. Chem. 275, 38605–38610
20 Jelen, F., Lachowicz, P., Apostoluk, W., Mateja, A., Derewenda, Z.S. and
Otlewski, J. (2009) Dissecting the thermodynamics of GAP–RhoA
interactions. J. Struct. Biol. 165, 10–18
21 Hildebrand, J.D., Taylor, J.M. and Parsons, J.T. (1996) An SH3
domain-containing GTPase-activating protein for Rho and Cdc42
associates with focal adhesion kinase. Mol. Cell. Biol. 16, 3169–3178
22 Taylor, J.M., Macklem, M.M. and Parsons, J.T. (1999) Cytoskeletal
changes induced by GRAF, the GTPase regulator associated with focal
adhesion kinase, are mediated by Rho. J. Cell Sci. 112, 231–242
23 Lundmark, R., Doherty, G.J., Howes, M.T., Cortese, K., Vallis, Y.,
Parton, R.G. and McMahon, H.T. (2008) The GTPase-activating protein
GRAF1 regulates the CLIC/GEEC endocytic pathway. Curr. Biol. 18,
1802–1808
24 Pylypenko, O., Lundmark, R., Rasmuson, E., Carlsson, S.R. and Rak, A.
(2007) The PX–BAR membrane-remodeling unit of sorting nexin 9.
EMBO J. 26, 4788–4800
25 Eberth, A., Lundmark, R., Gremer, L., Dvorsky, R., Koessmeier, K.T.,
McMahon, H.T. and Ahmadian, M.R. (2009) A BAR domain-mediated
autoinhibitory mechanism for RhoGAPs of the GRAF family. Biochem. J.
417, 371–377
26 Doherty, G.J. and McMahon, H.T. (2008) Mediation, modulation, and
consequences of membrane-cytoskeleton interactions. Annu. Rev.
Biophys. 37, 65–95
27 Chadda, R., Howes, M.T., Plowman, S.J., Hancock, J.F., Parton, R.G. and
Mayor, S. (2007) Cholesterol-sensitive Cdc42 activation regulates actin
polymerization for endocytosis via the GEEC pathway. Traffic 8,
702–717
28 Yarar, D., Waterman-Storer, C.M. and Schmid, S.L. (2007) SNX9 couples
actin assembly to phosphoinositide signals and is required for
membrane remodeling during endocytosis. Dev. Cell 13, 43–56
29 Fauchereau, F., Herbrand, U., Chafey, P., Eberth, A., Koulakoff, A.,
Vinet, M.C., Ahmadian, M.R., Chelly, J. and Billuart, P. (2003) The RhoGAP
activity of OPHN1, a new F-actin-binding protein, is negatively controlled
by its amino-terminal domain. Mol. Cell. Neurosci. 23, 574–586
30 Glebov, O.O., Bright, N.A. and Nichols, B.J. (2006) Flotillin-1 defines a
clathrin-independent endocytic pathway in mammalian cells.
Nat. Cell Biol. 8, 46–54
Received 16 April 2009
doi:10.1042/BST0371061
C The
C 2009 Biochemical Society
Authors Journal compilation 1065