Cas utilizes Nck2 to activate Cdc42 and regulate cell
polarization during cell migration in response to wound
healing
Kohei Funasaka
1
, Satoko Ito
2
, Hitoki Hasegawa
2
, Gary S.Goldberg
3
, Yoshiki Hirooka
1
,
Hidemi Goto
1
, Michinari Hamaguchi
2
and Takeshi Senga
2
1 Department of Gastroenterology, Nagoya University Graduate School of Medicine, Japan
2 Division of Cancer Biology, Nagoya University Graduate School of Medicine, Japan
3 Molecular Biology Department, University of Medicine and Dentistry of New Jersey, Stratford, NJ, USA
Introduction
The establishment of cell polarity is essential for a
variety of cellular functions, such as cell division, dif-
ferentiation and migration; however, the molecular
mechanisms underlying cell polarization have not been
elucidated thoroughly. Genetic and cell biological stud-
ies have identified several molecules that are important
for cell polarity. Among these proteins, Cdc42, a Rho
family GTPase conserved in a wide range of organ-
isms, has been found to play a pivotal role for the
establishment of cell polarity [1–3]. In yeast, Cdc42 is
required for polarized bud formation during cell divi-
sion and morphological changes in response to phero-
mone signaling [4]. In multicellular organisms, cell
polarity is determined by extracellular stimuli, such as
chemoattractant gradients and cell–cell contact. Locali-
zation and activation of Cdc42 in response to these
environmental changes are key events leading to cell
polarization [5,6].
Cas is a multiadaptor protein that regulates various
signaling pathways in response to extracellular stimuli,
including growth factors and integrin-mediated cell
adhesion [7–9]. Cas was originally identified as a
highly phosphorylated protein in cells transformed by
v-Src and v-Crk [10,11]. Cas contains an N-terminal
SH3 domain, proline-rich regions and a substrate
domain with multiple tyrosine phosphorylation sites
Keywords
Cas; Cdc42; Crk; Nck; polarity
Correspondence
T. Senga, Division of Cancer Biology,
Nagoya University Graduate School of
Medicine, 65 Tsurumai-cho, Showa-ku,
Nagoya 466-8550, Japan
Fax: +81 52 744 2464
Tel: +81 52 744 2463
E-mail: tsenga@med.nagoya-u.ac.jp
(Received 14 April 2010, revised 1 June
2010, accepted 28 June 2010)
doi:10.1111/j.1742-4658.2010.07752.x
Integrin-mediated activation of Cdc42 is essential for cell polarization,
whereas the integrin adaptor protein Cas is required for cell migration dur-
ing wound healing. After phosphorylation on tyrosine residues, Cas recruits
the adaptor proteins Crk and Nck to execute integrin-mediated signals.
However, the mechanisms leading to Cdc42 activation and its relationship
with Cas, Crk and Nck have not been elucidated clearly. In the present
study, we demonstrate that Cas utilizes Nck2 to activate Cdc42 and induce
cell polarization in response to wounding. By contrast, Cas recruits CrkII
to activate Rac1 and promote the extension of cell protrusions needed for
cell motility. These results indicate that Cas utilizes Nck2 and CrkII in a
coordinated set of distinct pathways leading to cell migration.
Structured digital abstract
lMINT-7909509:Cas (uniprotkb:Q61140) and Nck2 (uniprotkb:Q8BQ28)colocalize (MI:0403)
by fluorescence microscopy (MI:0416)
Abbreviations
CasKo, homozygous null Cas knockout; CasWt, CasKo transfected with wild-type Cas; DAPI, 4¢,6¢-diamino-2-phenylindole dihydrochloride;
GST, glutathione S-transferase; PAK, p21-activated kinase; PBD, p21 binding domain; PIX, PAK-interacting guanine nucleotide exchange
factor; PP2, 4-amino-5-(4-chlorophenyl)-7-(t-butyl)pyrazolo[3,4-D]pyramidine; siRNA, small interfering siRNA.
3502 FEBS Journal 277 (2010) 3502–3513 ª2010 The Authors Journal compilation ª2010 FEBS
that associate with SH2 domains to direct protein
interactions mediating signaling events leading to cell
migration [12,13].
Cas is ubiquitously expressed and its deletion in
mice is embryonic lethal [14]. Fibroblasts derived from
Cas-deficient mice showed cytoskeletal abnormalities
and defects in cell migration and spreading, indicating
an essential role of Cas for integrin-mediated signals
[15]. Tyrosine phosphorylation of Cas is mostly medi-
ated by the Src family kinases, and its phosphorylation
is required for Cas-mediated cell migration and trans-
formation [16–19]. Phosphorylated Cas recruits adap-
tor proteins such as Crk and Nck [20–22]. Association
of Crk with Cas enhances cell migration and spreading
by activating Rac1 [23]. Nck is important for regulat-
ing signals from cell surface receptors to the actin
cystoskeleton, as well as for cell movement. A number
of signaling molecules have been found to associate
with Nck; however, the physiological importance of
these interactions remains uncertain [24].
A wound-healing assay comprises a simple in vitro
experiment used to examine cell migration that is
enabled as a result of the release of physical con-
straints. A scratch in the confluent monolayer initiates
cell migration in the direction perpendicular to the
scratch until the gap is filled with cells [3]. Several
hours after the wound is made, cells on the edge of the
wound develop a polarized morphology [1]. Polarized
cells on the wound edge extend membrane protrusions
and reorient the Golgi in the direction of migration
[25]. Integrin-mediated activation of Cdc42 has been
shown to be critical for this polarization during cell
migration [1]; however, the signaling molecules
involved in the integrin-mediated activation of Cdc42
remain unknown. In the present study, we show that
Cas utilizes Nck2 to regulate cell polarization and
Cdc42 activity during cell migration in response to
wound healing.
Results
Cas is required for the polarization of migrating
cells
To examine the role of Cas in the establishment of cell
polarity during cell migration, we performed a wound-
healing assay using Cas deficient CasKo cells (homo-
zygous null Cas knockout cells) and CasWt cells
(generated by transfecting CasKo cells with wild-type
Cas). Cas expression in CasWt cells was similar to that
in Balb3T3 cells, and Cas was absent in CasKo cells
(Fig. 1A). As shown in Fig. 1B, CasWt cells migrated
faster than CasKo cells in this assay. In addition to
the wound healing assay, CasWt cells also migrated
approximately 40% better than CasKo cells through a
modified Boyden chamber (Fig. 1C).
Because cell polarization is an important prelude to
migration [26], we examined the effects of Cas on cell
polarization in response to wound healing. As shown
in Fig. 2A, CasWt cells at the wound edge started to
extend protrusions toward the free space within 4 h,
and over 90% of the cells at the edge were polarized,
with one side pointed toward the wound within 6 h.
By contrast, < 10% of the CasKo cells at the wound
edge displayed a polarized morphology 6 h after the
wound was made.
Measurement of protrusion length also indicated
that Cas was required for the formation of cell protru-
sions. As shown in Fig. 2B, CasWt cells exhibited cell
protrusions with a length of 61 ± 23 lm (mean ± SD)
by 3 h after wounding. This was almost twice the aver-
age protrusion length exhibited by CasKo cells, which
measured 35 ± 15 lm.
Microtubule elongation forms toward the leading
edge of cells during wound healing [1]. Tubulin stain-
ing indicates that Cas promoted this directional forma-
tion of microtubules within 3 h after wounding. As
shown in Fig. 2C, elongation of microtubules between
the nucleus and wound was observed in over 80%
of the CasWt cells on the wound edge. By contrast,
< 10% of the CasKo cells displayed this directional
organization of microtubules.
When cells are polarized for migration, the Golgi
becomes oriented between the nucleus and the direc-
tion of migration [3]. To examine the effects of Cas on
Golgi orientation, the localization of the Golgi matrix
protein, GM130 [27], was examined in CasKo and
CasWt cells on the wound edge after wounding. As
shown in Fig. 2D, polarized localization of the Golgi
in CasKo cells was clearly delayed compared to that of
CasWt cells. Approximately one-third of the Golgi was
localized within a 120arc between the nucleus and
the wound edge upon the wounding, which was the
result of chance because cells were sectioned into three
120arcs. Three hours after wounding, approximately
two-thirds of CasWt showed polarized localization of
the Golgi, whereas < 40% of CasKo cells showed
polarized localization of the Golgi.
Cas promotes Cdc42 activation and trafficking
during wound healing
Cdc42 is a Rho GTPase that traffics to the leading
edge of cell protrusions and regulates cell polarity dur-
ing wound healing [1]. The effects of Cas on Cdc42
localization during wound healing were evaluated by
K. Funasaka et al. Cas ⁄Nck2 regulates cell polarity
FEBS Journal 277 (2010) 3502–3513 ª2010 The Authors Journal compilation ª2010 FEBS 3503
immunofluorescence microscopy. As shown in Fig. 3A,
whereas more than 50% of the CasWt cells at the
wound edge contained Cdc42 localized on the leading
edge, < 10% of the CasKo cells at the wound edge
showed localization of Cdc42 on the leading edge.
Thus, Cas is required for trafficking of Cdc42 to the
leading edge of migrating cells.
In addition to intracellular location, the effects of
Cas on Cdc42 activation were also examined. A previ-
ous study demonstrated the activation of Cdc42 during
wound healing [1]. Cdc42 activity was assessed by
affinity precipitation of Cdc42-GTP with a glutathione
S-transferase–p21-activated kinase–p21 binding domain
(GST-PAK-PBD) fusion protein. As shown in Fig. 3B,
wound-induced activation of Cdc42 was reduced in
CasKo cells compared to CasWt cells. To further con-
firm the reduced activation of Cdc42 in CasKo cells,
we examined the activity of Cdc42 in both cell lines
3 h after wounding. Three independent experiments
demonstrated that the Cdc42 activity 3 h after wounding
in CasWt cells was almost twice that of CasKo cells
(Fig. 3C).
Silencing of Cas in Balb3T3 cells inhibits cell
polarization
To further evaluate the requirement of Cas for cell
polarization, we used small interfering RNA (siRNA) to
knockdown Cas expression in Balb3T3 cells. As shown
in Fig. 4A, transfection with Cas siRNA effectively sup-
pressed Cas expression. Three days after the transfection
of either control or Cas siRNA, orientation of the Golgi
during wound healing was examined by immunostain-
ing. As shown in Fig. 4B, an average of 29 ± 3.8% of
the cells transfected with Cas siRNA contained polar-
ized Golgi by 3 h after wounding compared to an aver-
age of 66 ± 3.2% seen in control transfectants.
In addition to reducing cell polarization, Cas siRNA
transfection also reduced Cdc42 activation and traf-
ficking during wound healing. As shown in Fig. 4C,
0
100
200
300
400
500
600
CasKo CasWt
Distance of migration
CasKo CasWt
0
24
Wound healing assay
(h)
(µm) (24 h)
*
200 µm
Migration assay (3 h)
CasKo CasWt
0
50
100
150
200
250
300
350
400
450
CasKo CasWt
No. of migrated cells per field
*
CasKo CasWt
Cas
Actin
Balb3T3
A
B
C
Fig. 1. Cas is essential for cell migration.
(A) Western blot analysis of Cas in CasWt,
CasKo and Balb3T3 cells. (B) Confluent
monolayers of CasWt and CasKo cells were
wounded with a pipette tip and incubated
for 24 h. Data are the mean ± SD of the
distance that leading edge of the monolayer
traveled into the wound area in five
randomly selected fields from three
independent experiments (*P< 0.01); scale
bar = 200 lm. (C) 5 ·10
4
CasKo and CasWt
cells were loaded onto the upper surface
of Boyden chambers, incubated for 3 h,
fixed, and examined by microscopy. Cells
that migrated to the lower surface of the
chamber are shown as the mean ± SD from
five randomly selected fields in three
independent experiments (*P< 0.01).
Cas ⁄Nck2 regulates cell polarity K. Funasaka et al.
3504 FEBS Journal 277 (2010) 3502–3513 ª2010 The Authors Journal compilation ª2010 FEBS
cells transfected with Cas siRNA exhibited approxi-
mately half of the Cdc42 activity found in control
transfectants 3 h after wounding. Cdc42 was also evi-
dent at the ends of cell protrusions on the wound edge
in control transfectants, although it was not detected
in cells transfected with Cas siRNA (Fig. 4D). Taken
together with the results obtained from Cas knockout
cells, these data indicate that Cas is an important com-
ponent of the signaling cascade that directs cell polari-
zation, Cdc42 activity and cell migration in response
to wound healing.
Src kinase inhibition disrupts polarization of
migrating cells
The Src tyrosine kinase phosphorylates Cas to pro-
mote cell migration [18]. We employed a Src kinase
inhibitor [4-amino-5-(4-chlorophenyl)-7-(t-butyl)pyra-
zolo[3,4-d]pyramidine; PP2] to determine whether Cas
phosphorylation was needed for the establishment of
polarity during wound healing. As shown in Fig. 5A,
tyrosine phosphorylation of Cas was induced by
wounding, which was effectively suppressed by PP2
treatment. This inhibition of Cas phosphorylation by
PP2 caused a decrease in cell elongation during wound
healing. As shown in Fig. 5B, cells treated with PP2 did
not extend protrusions into the wound area within 6 h
after wounding. In addition, PP2 treatment reduced
Golgi mobilization between the nucleus and wound
edge to levels seen in CasKo cells (Fig. 5C). These data
suggest that Src phosphorylates Cas to induce cell
polarization and migration during wound healing.
Nck2 is crucial for cell polarization and Cdc42
activation during wound healing
Crk and Nck are adaptor proteins that can associate
with phosphorylated tyrosine residues of Cas [8]. Two
024
6 (h)
CasKo
CasWt
Percentage of cells with
polarized golgi
0
10
20
30
40
50
60
70
80
90
CasKo CasWt
Distance of cell protrusion
(µm)
*
CasKo CasWt
100 µm
20 µm
CasKo CasWt
20 µm
0
10
20
30
40
50
60
70
80
0 h 1 h 3 h 6 h
CasKo CasWt
**
A
B
C
D
Fig. 2. Cas promotes wound-induced cell
polarization. (A) Confluent monolayers of
CasKo and CasWt cells were wounded and
cells were incubated at 37 C with 5% CO
2.
Photographs were taken at the indicated
time points (scale bar = 100 lm). (B) Three
hours after wounding, the cells were fixed,
immunostained with anti-a-tubulin serum
and DAPI, and the length of the protrusions
of wound edge cells was measured. Thirty
cells in randomly selected fields were mea-
sured in each of three independent experi-
ments. Data are the distance (mean ± SD)
between the leading edge and the nucleus
(*P< 0.01). (C) Three hours after wounding,
the cells were fixed and immunostained
with anti-a-tubulin serum and DAPI (scale
bar = 20 lm). (D) CasWt and CasKo cells
were wounded, fixed and immunostained
with anti-GM130 serum and DAPI at the
indicated time points to evaluate the per-
centage of cells with Golgi located in the
120arc facing the wound. One hundred
cells were evaluated for Golgi localization in
each of two independent experiments. Data
are the mean ± SEM (*P< 0.01). Images
on the right panel are representative images
of immunostained cells 3 h after wounding.
White lines indicate wound direction (green,
GM130; blue, DAPI; scale bar = 20 lm).
K. Funasaka et al. Cas ⁄Nck2 regulates cell polarity
FEBS Journal 277 (2010) 3502–3513 ª2010 The Authors Journal compilation ª2010 FEBS 3505
Crk family members, CrkII and CrkL, can associate
with phosphorylated Cas to regulate the actin cytoskel-
eton, cell migration, invasion and survival [28,29]. The
Nck family has two known members, Nck1 and Nck2,
and both proteins can associate with phosphorylated
Cas [20,22,24].
As shown in Fig. 6A, Crk and Nck proteins were
expressed to similar levels in CasKo and CasWt cells.
We performed siRNA knockdown experiments to
determine whether these proteins were involved in the
Cas-mediated polarization of cells. As shown in
Fig. 6B, transfection of specific siRNA to CasWt cells
effectively suppressed the expression of target Crk or
Nck proteins, but not other proteins. Cells transfected
with Nck2 siRNA displayed significantly less polarized
Golgi than other transfectants during wound healing,
indicating that Nck2 played a critical role in the polar-
ization of CasWt cells.
In addition to inhibiting orientation of the Golgi,
cell protrusions were more randomly oriented in
Nck2 knockdown cells compared to either control or
CrkII siRNA-transfected cells (Fig. 6D, E). Interest-
ingly, the elongation of protrusions was reduced in
CrkII knockdown cells but not in Nck2 knockdown
cells (Fig. 6D, F). CrkII siRNA reduced CrkII expres-
sion by approximately 50%, leading to a significant
reduction in cell protrusion distance of approximately
30% compared to control cells (t-test: P< 0.01).
These results indicate that Cas ⁄CrkII association was
required for the formation of protrusions, whereas
Cas ⁄Nck2 association was essential for the polariza-
tion of cells.
To further confirm the role of Nck2 for cell polari-
zation in cells expressing Cas, its localization and
effects on Cdc42 activity during wound healing were
examined. As shown in Fig. 7A, Nck2 co-localized
with Cas on the leading edge of cells. By contrast,
localization of Nck2 on the leading edge was not
observed in CasKo cells (Fig. 7B), indicating that Cas
was required for the polarized localization of Nck2.
CasKo
CasWt
0
10
20
30
40
50
60
70
CasKo CasWt
0
0.5
1
1.5
CasKo CasWt
Relative ratio of
active Cdc42
Percentage of cells with
Cdc42 localized on the leading edge
*
20 µm
Active Cdc42
Total Cdc42
CasWt CasKo
Time (h) 01 3 013
CasKo CasWt
Total Cdc42
Active Cdc42
A
B
C
Fig. 3. Cas promotes Cdc42 activation and
intracellular trafficking during wound healing.
(A) Three hours after wounding, cells were
fixed and immunostained for Cdc42. DAPI
was used to stain nuclei. Arrows indicate
Cdc42 localized on the leading edge. Fifty
cells on the wound edge in each of three
independent experiments were evaluated
for the localization of Cdc42. The percent-
age of these cells with Cdc42 localized on
the leading edge is presented as the
mean ± SEM (*P< 0.01). (B) Forty
scratches were made on the confluent
monolayers of cells, and cells were lysed at
the indicated time points to detect total
Cdc42 and active, GTP bound, Cdc42. (C)
Forty scratches were made and, 3 h later,
cells were lysed to detect total and active
Cdc42. Three independent experiments
were performed and relative ratios of Cdc42
activity are shown as the mean ± SD.
A representative result from the western
blotting is shown.
Cas ⁄Nck2 regulates cell polarity K. Funasaka et al.
3506 FEBS Journal 277 (2010) 3502–3513 ª2010 The Authors Journal compilation ª2010 FEBS