ABBS 2005,38(02): Roles of cortactin, an actin polymerization mediator, in cell endocytosis

*Corresponding author: Tel, 86-513-5629631; Fax, 86-513-85519820; E-mail,
[email protected]

Abstract        Cortactin, an actin-binding protein and
a substrate of Src, is encoded by the EMS1 oncogene. Cortactin is known to
activate Arp2/3 complex-mediated actin polymerization and interact with
dynamin, a large GTPase and proline rich domain-containing protein. Transferrin
endocytosis was significantly reduced in cells by knock-down of cortactin
expression as well as in vivo introduction of cortactin immunoreagents.
Cortactin-dynamin interaction displayed morphologically dynamic co-distribution
with a change in the endocytosis level in cells treated with an actin
depolymerization reagent, cytochalasin D. In an in vitro beads assay, a
branched actin network was recruited onto dynamin-coated beads in a cortactin
Src homology domain 3 (SH3)-dependent manner. In addition, cortactin was found
to function in the late stage of clathrin coated vesicle formation. Taken
together, cortactin is required for optimal clathrin mediated endocytosis in a
dynamin directed manner.

Key words        cortactin; dynamin; actin polymerization;
endocytosis

The uptake of
extracellular materials through ­endocytosis requires an orchestrated
interaction of the endocytic ­components [1–3]. In this process, coated vesicle ­formation is an
essential step, known as a dynamic ­membrane activity mediated by the large
GTPase dynamin [4–11]. A number
of Src-homology domain containing ­proteins are implicated in coated vesicle
formation through interaction with the carboxy-terminal proline-rich domain of
dynamin. These accessory proteins, including Grb2 [12], intersectin [13],
endophilin [14], syndaptin I [15], amphiphysin I [16], amphiphysin II [17],
profilin [18] and mammalian Abp1 [19], either recruit dynamin or are ­recruited
by dynamin to activate dynamin GTPase activity and function in clathrin coated
pit invagination and scission.

Actin cytoskeleton
implicated in endocytosis is a ­well-grounded concept. Recent studies indicated
that dynamin, rather than being a mechano-enzyme, may instead ­activate a
downstream mediator of fission in clathrin coated vesicle (CCV) formation [20].
This highlights the importance of the actin skeleton as a potential mediator in
power motion [20]. In fact, multiple proteins binding to dynamin, such as
syndapin I, profilin and mAbp1, interact with actin or ­regulate actin
assembly. Recently, cortactin, an ­­actin-binding partner and direct substrate
of Src tyrosine kinase, was found to be a dynamin-binding protein [21]. ­Morphological
findings in the ultrastructural level indicated that cortactin was distributed
over the surface or base of clathrin lattices, as well as actin filaments
associated with the pits [22]. Cortactin contains an N-terminal acidic ­domain
that binds to actin-related protein complex (Arp2/3) that ­mediates actin
polymerization [23,24]. In addition, cortactin has a unique structure that
features 6.5 tandem repeats and a carboxy-terminal Src homology ­domain 3 (SH3)
domain that form the F actin binding ­domain and the dynamin binding region,
respectively [21,23].

Here, we demonstrate
that normal cortactin level in cells is required for optimal clathrin mediated
endocytosis. ­Unlike lamellipodia and cortical ruffles, dynamin recruits
cortactin to endocytic sites during the process in an actin-dependent fashion.
To address further the role of cortactin in CCV formation, we used a functional
assay for CCV ­formation to determine the hierarchy by which cortactin is
targeted to and acts in coated pit invagination and scission. We found that
cortactin plays an important role in CCV fission from plasma membranes. This
study ­provides in vitro and in vivo evidence of the
dynamin-directed function of cortactin in endocytosis, suggesting that
cortactin may act as an adaptor in downstream CCV scission.

Materials and Methods

DNA constructs and
recombinant proteins

The cDNAs encoding
cortactin and its mutants [full-length cortactin, cortactin D (1–80), cortactin D (1–193) and cortactin D (1–352)] were generated by polymerase chain reaction (PCR)
using the murine cortactin ­complementary ( pXZ112) [23,25] as a
template, and were subcloned into mammalian expression vector pCMV-tagcDNA encoding proline
rich domain (PRD) of dynamin 2 was PCR-amplified using full-length dynamin 2 ­plasmid
(Dyn-GFP) (a gift from Dr. McNiven,
Mayo Clinic, To prepare recombinant
GST-cortactin, GST-cortactin D(1–80), GST-Cort-SH3 and
GST-CortDSH3, cDNAs encoding corresponding fragments were
inserted into pGEX-2T vector at EcoRI/BamHI sites. The detailed
methods for preparation of GST-cortactin derivates and GST-free proteins have
been described ­previously [26].

Cortactin small RNA
interference

Cortactin siRNA was
transcribed in vitro using T7 RNA polymerase according to the manufacturer’s
protocol for the Silencer siRNA construction kit (Ambion,

DNA transfection and
immunofluorescence microscopy

MDA-MB-231,
3T3-L1 cells and NIH/3T3 cells were obtained from ATCC (For
immunofluorescence microscopy, cells were either permeabilized with P-buffer
(0.02% Triton X-100,

Visual inspection and
biochemical assay of transferrin uptake

Cells were incubated
with 10 mg/ml

Analysis of the
interaction between cortactin and PRD of dynamin

Cortactin was incubated
with GST-Dyn-PRD ­immobilized on glutathione-Sepharose at different
concentrations. After incubation, samples were centrifuged and the supernatants
were fractionated by sodium dodecyl ­sulfate-polyacrylamide gel electrophoresis
(SDS-PAGE) and analyzed by immunoblot using cortactin antibody (

Actin branching bead
assay

Fluorescent carboxylated
polystyrene beads were coated with GST-Dyn-PRD by incubating beads in protein ­solution
(5 mg/ml) in room temperature for 2 h. The extent of protein coating was
examined by immunoblot assay using GST antibody. Beads were washed twice with
the 1´actin
polymerization buffer (

Cell-free assay for CCV
formation

Rat brain cytosol was
prepared exactly as previously ­described [29]. The cell-free internalization
assay using ­perforated 3T3-L1 cells was performed based on ­previously
described methods [29,30]. 3T3-L1 cells were first washed in KSHM (

Results

Cortactin is required
for clathrin-mediated ­endocytosis

We applied cortactin RNA
and interference ­technique to MDA-MB-231 cells, a breast cancer cell line in
which cortactin is usually overexpressed, to directly examine the effect of
cortactin on transferrin uptake. As shown in Fig. 1(A), after cortactin siRNA treatment, the
expression of cortactin in MDA-MB-231 cells was ­dramatically reduced to trace
level. While GFP ­­­knock-down as a mock treatment had no effect on transferrin
uptake, knock-down of cortactin substantially attenuated the ligand uptake [Fig.
1(B,C)]. However, cortactin RNA interference had no effect on fluid-phase
endocytosis of dextran (data not shown).

As an additional
approach to interfere with cortactin function in vivo, we introduced
anti-cortactin immunoreagents into MDA-MB-231 cells using the BioPorter system. While incubation with
BioPorter alone did not lead to
impairment in endocytosis, ­introduction of either polyclonal or monoclonal
antibody of cortactin caused an apparent attenuation of transferrin uptake. As
depicted in Fig. 2, based
on ­capture-ELISA-based internalization assay, introduction of cortactin
immunoreagents resulted in a 30%–60% reduction
in ­transferrin uptake. The control performed with rabbit IgG or labeled IgG
failed to show any effects on endocytosis.

Association of cortactin
with dynamin at endocytosis sites in an actin-dependent manner

We applied cytochalasin
D, known to cause perturbation on actin polymerization, to live NIH/3T3 cells,
and observed the variation of the association between cortactin and dynamin as
well as the alteration of transferrin endocytosis. Dynamin antibody was used to
label the dynamin in endocytosis sites [31,32]. As shown in Fig. 3,
after exposure to 10 mM ­cytochalasin
D for 1 h, transferrin endocytosis in NIH/3T3 cells was apparently impaired in
accompaniment with submerging of the cortactin fluorescent signal which was
associated with dynamin, while the dynamin signal at the sites seemed
unaffected. Interestingly, we ­noticed the return of the colocalization of
cortactin with dynamin after drug washing-out. In the meantime, the endocytosis
was restored as well.

Active recruitment of
cortactin to endocytosis sites by dynamin through its proline rich domain

Transfected NIH/3T3
cells were subjected to pulse ­exposure to transferrin for 10 min. Results from
­­­­multiple-color staining showed that dynamin and cortactin ­colocalized
exactly with APIn the in vitro
experiment, we observed that the affinity of Dyn-PRD to cortactin represented by
an apparent K_d value was
less than 1 mM, which
supported the notion that Dyn-PRD could substantially pull down cortactin from
quiescent NIH/3T3 total lysate (Fig. 5). Notably, in the actin branching
experiment, cortactin-mediated actin branching could hatch around Dyn-PRD
coated beads. This phenomenon was abolished once an SH3-depleted cortactin was
in place of the wild-type cortactin. Based on all of these results, we believe
that dynamin would be first localized to endocytosis sites followed by recruitment
of cortactin through the interaction of Dyn-PRD with the Cort-SH3 domain, and
this event was naturally ­accompanied by actin assembly. In fact, according to
a recent study, imaging of actin and dynamin recruitment in the early stage of
endocytosis showed that a burst of dynamin recruitment was an earlier event
than actin recruitment to the endocytosis area [20].

Cortactin functions at
the late stage of CCV ­formation

We continuously
investigated the role of cortactin in CCV formation, an essential course in
clathrin-mediated endocytosis, by virtue of a cell free functional assay
[30,33] using 3T3-L1 cells [29], which normally express cortactin. In this
assay, transferrin biotinylated via a cleavable disulphide bond (B-SS-Tfn)
binds to its receptor in perforated cells and is further sequestered into
deeply invaginated coated pits and internalized into coated vesicles in a
cytosol- and ATP-dependent manner [27,30,33,34]. In the present work, we
prepared rat brain cytosol in which cortactin protein was almost completely
depleted with cortactin antibody [Fig. 6(A)]. While cortactin alone has
no effect on CCV formation in the absence of cytosol, cortactin-depleted
cytosol displayed inhibitory activity on the process. Based on MesNa resistance
assay, B-SS-Tfn internalization, which reflected the amount of CCV formation,
was reduced in the presence of cortactin-depleted cytosol. Addition of
recombined cortactin protein in cytosol dramatically restored B-SS-Tfn
internalization. However, preincubating of the membrane with N-terminal-deleted
cortactin or Cort-SH3 domain exhibited distinct effects on rescuing the
reduction of B-SS-Tfn internalization by cortactin depletion. Addition of
cortactin SH3 domain resulted in a further inhibition of CCV formation. Notably,
when N-terminal-deleted cortactin was added, the restoration of CCV formation
amounted to 70% of that produced by full-length cortactin (Fig. 6),
indicating N-terminal-deleted cortactin was less efficient than the wild-type
cortactin.

Discussion

In the present study we
further examined the function of cortactin, an actin binding partner encoded by
the EMS1 oncogene, in clathrin-mediated endocytosis and its ­relationship with
dynamin in the process. We initiated this study based on the understanding of
the relationship ­between cytoskeleton and endocytosis [35–40] and the morphological findings that cortactin
associated with clathrin coated pits and endosomal vesicles [22,41], and that
cortactin is known as a potent mediator of actin ­nucleation [23]. We first
combined the application of cortactin siRNA and immunoreagents ­introduction
into live cells to perturb the function of cortactin in cells. We demonstrated
that interference of cortactin function in cells could result in the impairment
of transferrin endocytosis (Figs. 1 and 2). This result is ­consistent with the findings that
microinjection of ­anti-cortactin antibodies into living cells inhibited
endocytosis [22]. Based on these results we believe that cortactin plays a
substantial role in constitutive endocytosis, however, so far the detailed
function of this molecule in clathrin-mediated endocytosis has not been fully
elucidated.

The large GTPase dynamin
is a crucial component of endocytosis machinery which was known as a
mechano-enzyme functioning in the initial stage of endocytosis [4–10,35,37,42]. Dynamin directly associates with
cortactin through the proline rich domain of dynamin and the SH3 domain of
cortactin [21]. This interaction was expansively reflected by their
co-distribution in lamellipodia and ruffles, induced membrane waves, podosomes
and actin tails of vesicles [21,32,43], and was always accompanied by ­actin
rearrangement. Thus, we further examined the effect of actin dynamics on the
association of dynamin and cortactin in endocytosis sites which was probed by
the anti-dynamin antibody Hudy 1 [31,32]. Unexpectedly, when NIH/3T3 cells were
treated with an actin ­polymerization inhibitor cytochalasin D for 1 h, the
association of cortactin with dynamin became invisible, but was restored in
less than 1 h after cytochalasin D was washed out (Fig. 3). It was
demonstrated that the association of cortactin with dynamin in endocytosis
sites could not be dependent on actin dynamics, in which cortactin was found to
play a modulator role [23,24]. This result also implied that the localization
of cortactin to endocytic required interaction with F-actin via dynamin. In
order to fully understand the function of cortactin-dynamin association in
endocytosis sites, it is necessary to know the dynamics of this interaction. We
thus ­performed a further experiment in vivo and ­established a system in
vitro to demonstrate the ­recruitment between cortactin and dynamin. Based
on the results from these two approaches, cortactin appears to be passively
recruited by dynamin to endocytosis-related sites (Figs. 4 and 5). Interestingly,
branched actin filaments could be ­partially recruited to dynamin PRD coated
beads, which was ­abolished as mutant cortactin with no SH3 domain and was in
place of the wild-type one (Fig. 5). This is ­consistent with the cell
staining results.

To investigate if
cortactin is implicated at an early stage of endocytosis, we used a functional
assay based on a cell-free system that analyzes initial events immediately
after membrane invagination. As a result, we prepared rat brain cytosol extract
treated with an anti-cortactin antibody, which depleted cortactin proteins
completely but had no effect on actin in the extract. The cortactin-depleted ­extract
was incubated with 3T3-L1 cells. The extract without cortactin induced CCV
formation poorly with an efficiency approximately 60% less than that of the
mock-treated extract. Furthermore, adding a recombinant wild-type cortactin
resulted in a 90% restoration of CCV formation (Fig. 6). These data
demonstrate that cortactin is essential for optimal cell endocytosis, which
involves a dynamin-directed interaction in the early stage of CCV formation.

References

 1    Marsh
M, McMahon HT. The structural era of endocytosis. Science 1999, 285: 215–220

 2    McPherson
PS, Kay BK, Hussain NK. Signaling on the endocytic pathway. Traffic 2001, 2:
375–384

 3    Munn
AL. Molecular requirements for the internalisation step of endocytosis: insights from yeast. Biochim Biophys
Acta 2001, 1535: 236–257

 4    Damke
H. Dynamin and receptor-mediated endocytosis. FEBS Lett 1996, 389: 48–51

 5    Hinshaw
JE. Dynamin and its role in membrane fission. Annu Rev Cell Dev Biol 2000, 16:
483–519

 6    Liu
JP, Robinson PJ. Dynamin and endocytosis. Endocr Rev 1995, 16: 590–607

 7    McClure
SJ, Robinson PJ. Dynamin, endocytosis and intracellular signalling. Mol Membr
Biol 1996, 13: 189–215

 8    McNiven
MA,  9    Sever
S. Dynamin and endocytosis. Curr Opin Cell Biol 2002, 14: 463–467

10   Vallee RB,
Herskovits JS, Aghajanian JG, Burgess CC, Shpetner HS. Dynamin, a GTPase
involved in the initial stages of endocytosis. Ciba Found Symp 1993, 176: 185–193

11   Warnock DE,
Schmid SL. Dynamin GTPase, a force-generating molecular switch. Bioessays 1996,
18: 885–893

12   Gout I,
Dhand R, Hiles ID, Fry MJ, Panayotou G, Das P, Truong O et al. The
GTPase dynamin binds to and is activated by a subset of SH3 domains. Cell 1993,
75: 25–26

13   Roos J,
Kelly RB. Dap160, a neural-specific Eps15 homology and multiple SH3
domain-containing protein that interacts with Drosophila dynamin. J Biol Chem
1998, 273: 19108–19119

14   Ringstad N,
Nemoto Y, De Camilli P. The SH3p4/Sh3p8/SH3p13 protein family: binding partners for synaptojanin and
dynamin via a Grb2-like Src homology 3 domain. Proc Natl Acad Sci 15   Qualmann B,
Roos J, DiGregorio PJ, Kelly RB. Syndapin I, a synaptic dynamin-binding protein
that associates with the neural Wiskott-Aldrich syndrome protein. Mol Biol Cell
1999, 10: 501–513

16   David C,
McPherson PS, Mundigl O, De Camilli P. A role of amphiphysin in synaptic
vesicle endocytosis suggested by its binding to dynamin in nerve terminals.
Proc Natl Acad Sci 17   Ramjaun AR,
Micheva KD, Bouchelet I, McPherson PS. Identification and characterization of a
nerve terminal-enriched amphiphysin isoform. J Biol Chem 1997, 272: 16700–16706

18   Witke W,
Podtelejnikov AV, Di Nardo A, Sutherland JD, Gurniak CB, Dotti C, Mann M. In
mouse brain profilin I and profilin II associate with regulators of the
endocytic pathway and actin assembly. EMBO J 1998, 17: 967–976

19   Kessels MM,
Engqvist-Goldstein AE, Drubin DG, Qualmann B. ­Mammalian Abp1, a
signal-responsive F-actin-binding protein, links the actin ­cytoskeleton to
endocytosis via the GTPase dynamin. J Cell Biol 2001, 153: 351–366

20   Merrifield
CJ, Feldman ME, Wan L, Almers W. Imaging actin and dynamin recruitment during
invagination of single clathrin-coated pits. Nat Cell Biol 2002, 4: 691–698

21   McNiven MA,
Kim L, Krueger EW, Orth JD, 22   23   Uruno T,
Liu J, Zhang P, Fan YX, Egile C, Li R, Mueller SC et al. Activation of
Arp2/3 complex-mediated actin polymerization by cortactin. Nat Cell Biol 2001,
3: 259–266

24   Weaver AM,
Karginov AV, Kinley AW, Weed SA, Li Y, Parsons JT, Cooper JA. Cortactin
promotes and stabilizes Arp2/3-induced actin filament network formation. Curr
Biol 2001, 11: 370–374

25   Zhan X, Hu
X, Hampton B, Burgess WH, Frirsel R, Maciag T. Murine cortactin is
phosphorylated in response to fibroblast growth factor-1 on ­tyrosine residues
late in the G1 phase of the BALB/c 3T3 cell cycle. J Biol Chem 1993, 268: 24427–24431

26   Huang C,
Zhan X. Proteolysis of cortactin by calpain in platelets and in vitro.
Methods Mol Biol 2000, 144: 289–295

27   Smythe E,
Carter LL, Schmid SL. Cytosol- and clathrin-dependent ­stimulation of
endocytosis in vitro by purified adaptors. J Cell Biol 1992, 119: ­1163–1171

28   Blanchoin
L, Amann KJ, Higgs HN, Marchand JB, Kaiser DA, Pollard TD. Direct observation
of dendritic actin filament networks nucleated by Arp2/3 complex and WASP/Scar
proteins. Nature 2000, 404: 1007–1011

29   Simpson F,
Hussain NK, Qualmann B, Kelly RB, Kay BK, McPherson PS, Schmid SL.
SH3-domain-containing proteins function at distinct steps in clathrin-coated
vesicle formation. Nat Cell Biol 1999, 1: 119–124

30   Carter LL,
Redelmeier TE, Woollenweber LA, Schmid SL. Multiple ­GTP-binding proteins
participate in clathrin-coated vesicle-mediated endocytosis. J Cell Biol 1993,
120: 37–45

31   Kessels MM,
Engqvist-Goldstein AE, Drubin DG, Qualmann B. Mammalian Abp1, a
signal-responsive F-actin-binding protein, links the actin ­cytoskeleton to
endocytosis via the GTPase dynamin. J Cell Biol 2001, 153: 351–366

32   Ochoa GC,
Slepnev VI, Neff L, Ringstad N, Takei K, Daniell L, Kim W et al. A
functional link between dynamin and the actin cytoskeleton at podosomes. J Cell
Biol 2000, 150: 377–389

33   Schmid SL,
Smythe E. Stage-specific assays for coated pit formation and coated vesicle
budding in vitro. J Cell Biol 1991, 114: 869–880

34   Hill E, van
Der Kaay J, Downes CP, Smythe E. The role of dynamin and its binding partners
in coated pit invagination and scission. J Cell Biol 2001, 52: 309–323

35   da Costa SR,