Structural Basis of Integrin
Activation by Talin
Kate L. Wegener,1,5Anthony W. Partridge,2,5Jaewon Han,2Andrew R. Pickford,4Robert C. Liddington,3
Mark H. Ginsberg,2,* and Iain D. Campbell1,*
1Department of Biochemistry, University of Oxford, South Parks Road, Oxford OX1 3QU, England, UK
2Department of Medicine, University of California San Diego, La Jolla, CA 92093, USA
3Program on Cell Adhesion, Cancer Center, Burnham Institute for Medical Research, La Jolla, CA 92037, USA
4Biological Sciences, University of Portsmouth, King Henry I Street, Portsmouth PO1 2DY, England, UK
5These authors contributed equally to this work.
*Correspondence: firstname.lastname@example.org (I.D.C.), email@example.com (M.H.G.)
Regulation of integrin affinity (activation) is es-
sential for metazoan development and for
lin phosphotyrosine-binding (PTB) domain to
integrin b subunit cytoplasmic domains (tails)
causes activation, whereas numerous other
PTB-domain-containing proteins bind integrins
without activating them. Here we define the
structure of a complex between talin and the
membrane-proximal integrin b3 cytoplasmic
domain and identify specific contacts between
talin and the integrin tail required for activation.
We used structure-based mutagenesis to engi-
neer talin and b3 variants that interact with
comparable affinity to the wild-type proteins
but inhibit integrin activation by competing
with endogenous talin. These results reveal
the structural basis of talin’s unique ability to
activate integrins, identify an interaction that
could aid in the design of therapeutics to block
integrin activation, and enable engineering of
cells with defects in the activation of multiple
classes of integrins.
Integrins are found throughout the animal kingdom, where
they play important roles in cell adhesion, migration, pro-
liferation, and survival. They are membrane-spanning het-
erodimers of a and b subunits, both of which typically
comprise a short cytoplasmic tail (?20 to 50 residues),
main (?700 to 1000 residues). In mammals, there are 18
identified a subunits and 8 b subunits that combine to
form 24 distinct heterodimers. In humans, integrins play
critical roles in development and participate in the patho-
genesis of heart disease, chronic inflammation, and can-
cer (Ginsberg et al., 2005; Hynes, 2002).
mains in adefault low-affinity ligand-binding state(the ‘‘off
state’’); however, cells can change the conformation and
affinity of these receptors in response to cellular stimula-
tion in a process often termed ‘‘integrin activation.’’ This
conformational change results in increased adhesion
and subsequent signaling, which mediates events such
as cell migration, platelet aggregation, leukocyte exit
from the vasculature, and assembly of the extracellular
matrix. The binding of a cytoskeletal protein called talin
to the b subunit cytoplasmic tail is a common final step
in the activation process (Tadokoro et al., 2003; Tanent-
zapf and Brown, 2006).
tains bundles of a helices and an N-terminal FERM (band
4.1, ezrin, radixin, and moesin) domain with three subdo-
mains: F1, F2, and F3 (Chishti et al., 1998; Garcia-Alvarez
et al., 2003; Papagrigoriou et al., 2004; Rees et al., 1990).
Talin colocalizes with integrins (Horwitz et al., 1986) and
binds to F-actin and actin-binding proteins (reviewed in
Critchley, 2005), thus linking the actin-cytoskeletal net-
work to the extracellular matrix. The F3 subdomain of
the FERM domain contains the highest affinity integrin-
binding site for integrin b tails and is sufficient to activate
integrins(Calderwood etal.,2002).Apartial viewofthisin-
teraction was obtained in a crystal structure of the F2 and
F3domains of thetalinFERMdomain incomplexwitha12
residue fragment (739WDTANNPLYDEA750) comprising
?25% of the b3-integrin tail (Garcia-Alvarez et al., 2003).
osine-binding (PTB) domain fold and that its interaction
with integrins strongly resembles the interaction of other
PTB domains with peptide ligands (Garcia-Alvarez et al.,
2003). Several other PTB domains bind to b3 in a similar
fashion to talin (Calderwood et al., 2003; Garcia-Alvarez
tivate integrins. Thus, we reasoned that additional unique
features of the integrin/talin interaction enable talin to
cause activation. Indeed, nuclear magnetic resonance
Cell 128, 171–182, January 12, 2007 ª2007 Elsevier Inc. 171
(NMR) spectroscopic evidence has suggested an interac-
tion between talin F3 and the membrane-proximal (MP) b-
tail region (Ulmer et al., 2003; Vinogradova et al., 2002)
that is important for activation (Hughes et al., 1995; Ulmer
et al., 2003). However, such an interaction has not been
Here, we provide the first atomic-level description of
the interaction between talin and the MP region of the
b3-integrin cytoplasmic domain. Structure-based muta-
genesis shows that disruption of the newly identified inter-
actions blocks integrin activation in cells while retaining
binding to the membrane-distal (MD) region of b3. These
results define the talin/integrin interactions that lead to in-
tegrin activation and explain why talin’s activating ability is
unique among PTB-containing proteins. They also identify
aninteraction thatisatarget forthe rational designof ther-
apeutics that will block integrin activation while allowing
other MD-interacting proteins to maintain their functions.
or animals with selective defects in integrin activation.
Talin Interacts with Both the MP and MD Regions
of the b3 Tail
We set out to obtain a structural explanation for the un-
usual ability of the talin PTB domain to activate integrins.
An initial exploration of the interactions was carried out
by adding various tail-derived b3-integrin peptides to the
F3 subdomain of talin while monitoring the chemical shift
perturbations in1H-15N-HSQC NMR spectra. The peptide
sequences and the nomenclature used are set out in
Figure 1A. The MP peptide corresponds to the N-terminal
region of the cytoplasmic b tail. A peptide comprised of
the b-tail region previously visualized in the crystal struc-
ture (Garcia-Alvarez et al., 2003) is denoted MD; this
Figure 1. NMR Titration Studies of Integ-
rin and PIPKIg Peptide
(A) Design of peptides.
(B and C) Weighted shift maps (seeExperimen-
talProcedures)of induced chemical shifts seen
in1H-15N-HSQC spectra when the (B) MP pep-
tide and (C) MD peptides are added to the talin
(D–H) Weighted shift data for each peptide
titration are mapped onto the structure of the
F3 subdomain with the largest shifts shown in
blue (the full data set is given in Figure S1).
(D) MP peptide.
(E) MD peptide.
(F) Full-length b3-integrin tail. Because of
the unfavorable exchange rates, the data for
the full-length b3 peptide were obtained from
(G) PIPKIg peptide.
(H) b3/PIPKIg-chimeric peptide.
172 Cell 128, 171–182, January 12, 2007 ª2007 Elsevier Inc.
peptide primarily causes chemical shift perturbations in
strand S5 (de Pereda et al., 2005).
acts with a site on F3 that is distinct from the MD-binding
site (Figure 1). The new perturbations found in the MP-
binding site mainly arise in residues that form a loop be-
tween the first and second b strands (S1 and S2) of the
F3 structure. Residues 358 and 359 of F3 are also per-
turbed, presumably due to the b-tail residue W739 that
is included in both the MP and MD peptides and binds
tightly in a pocket on the surface of F3.
NMR studies of the interaction with full-length b3 cyto-
plasmic tails were more challenging. The off rates of the
complex correspond to an ‘‘intermediate exchange’’ re-
gime, with peaks broadening and disappearing rather
than changing position. It was possible, however, to
define perturbed residues by plotting the percentage
decrease in peak height upon addition of the peptide (Fig-
ures 1D and S1). The perturbed residues are essentially
the sum of those affected by the MD peptide and the
MPpeptide (Figures 1andS1). Theb3-tail/talin F3interac-
tions can thus be considered in two parts, with one corre-
sponding to the previously defined interaction at the MD
site and the other corresponding to the MP site. The lim-
ited solubility of the full-length b3 tail meant that it was
not possible to produce a 1:1 complex at the concentra-
tions used. This, together with exchange broadening of
lines, made the b3 tail an unsuitable ligand for studying
the structure of the talin/integrin complex by NMR at
We have, however, previously shown that a peptide
segment from phosphatidylinositol phosphate kinase
peptide in the MD site (de Pereda et al., 2005). These data
are shown again here for comparison (Figures 1D and S1).
by a reverse turn, but the PIPKIg peptide binds with much
higher affinity in a slow-exchange NMR regime, and a 1:1
complex is readily achieved. Critically, W739 from the b3
tail and W642 from the PIPKIg peptide occupy essentially
identical positions and adopt identical main-chain and
side-chain torsion angles, each binding in a deep pocket
on the talin protein surface (de Pereda et al., 2005; Gar-
cia-Alvarez et al., 2003). This strong similarity led us to be-
lieve that a synthetic chimeric peptide comprised of b3-in-
tegrin tail residues on the N-terminal side of the critical
tryptophan and PIPKIg residues on the C-terminal side
would yield a peptide ligand with higher affinity and better
solubility that would besuitable forhigh-resolution studies
of the F3-MP interaction. The sequence of the designed
chimeric peptide is shown in Figure 1A. NMR experiments
showed that this chimeric peptide does indeed bind to
talin and forms a tight, highly soluble 1:1 complex. More-
over, comparison of the protein-backbone chemical shifts
in the bound and unbound states indicates that the resi-
dues affected by the chimeric peptide are essentially the
sum of residues affected by the MP-integrin peptide and
those affected by the PIPKIg peptide (Figures 1 and S1),
supporting the idea that the chimera is suitable for struc-
tural studies of the F3-MP interaction.
The Structure of the Talin F3/Peptide Complex
Reveals a Binding Interface with the MP Site
The structure of the complex was calculated from NMR
data using a total of 2187 experimental restraints, 138 of
which were unambiguous intermolecular nuclear Over-
hauser effects (NOEs; Table S1 and Figure 2). As ex-
pected, the PIPKIg-derived portion of the peptide binds
(de Pereda et al., 2005); the SPLH sequence (residues
742–745) forms a reverse turn, and residues 739–741 cre-
ate a b strand that augments the b sheet formed by
strands S5–S7 from talin (Figure 2). Importantly, the
pivotal tryptophan residue (W739) that connects the two
idue in the crystal structure of the b3-tail complex (Garcia-
residues 716–739 of the b3 tail are free to adopt their
native orientation in the F3 complex. The extended nature
of the complex and the relative rigidity of the F3 domain
are consistent with the notion that the MD site and the
MP site behave relatively independently (Figure 2B).
Thus, any differences between the native integrin peptide
and the chimeric peptide are restricted to the MD site. The
structural and functional implications of differences be-
tween the binding of the PIPKIg peptide and the b-tail
MD site have been discussed previously (de Pereda
et al., 2005).
The new structure reveals important novel contacts be-
tween the MP region of b3 and talin. Specifically, the b3-
derived portion of the peptide forms an a helix between
H722 and R736. This helix lies across strands S1–S2 and
lecular NOEs used in the structure calculation, 57 define
the two phenylalanine residues F727 (23 NOEs) and F730
(18 NOEs). The hydrophobic talin residue L325 appears
important for this interaction, and 7 NOEs were identified
between this residue and the phenylalanine side chains
(see Figure S2). The calculated structure is consistent
during NMR titration experiments (Figure S1).
Thereis a propensity for residues 722–732of the free b3
peptide to form an a-helical structure since the a-proton
chemical shifts of these residues are consistently down-
field from random-coil values. Binding to the talin F3 do-
main further increases the downfield shifts of these b-tail
resonances, which indicates that the helix is stabilized in
the complex (data not shown). Heteronuclear
NOEs were recorded for the protein alone as well as in
complex with various peptides (Figure S3). The results
indicatethattheloopbetween S1andS2oftalin F3isstiff-
supports the calculated structure.
Cell 128, 171–182, January 12, 2007 ª2007 Elsevier Inc. 173
Mutation of b-MP Residues that Contact F3
Diminishes the Ability of Talin to Activate Integrin
Contacts between talin and the MP region of the b3 cyto-
plasmic tail may constitute an interaction that is a key to
the activation process. To test this concept, the effects
of a number of designed mutations in both the b3 tail
and the talin F3 subdomain were investigated. Integrin ac-
tivation was assessed using the antibody PAC1 and flow
cytometry as previously described (Partridge et al., 2005).
The effects of mutating F727 and F730, which both
make intimate contact with F3, were investigated using a
mutant integrin aIIb subunit, R995A. When paired with a
wild-type (WT) b3 subunit, the assembled integrin is in
an activated state in transfected cells (Figure 3A), which
is consistent with previous reports (Hughes et al., 1996).
The activated state is dependent on endogenous talin,
as shown by the decrease or abolition of ligand binding
whentheinfluence oftalin isreducedeitherbytalinknock-
down or by a mutation in the b tail (Y747A) that disrupts
talin binding (Tadokoro et al., 2003; Figure 3A). When
aIIb(R995A) was paired with b3(F727A) or b3(F730A), b3
mutants designed to disrupt the F3-MP interaction, the
activating effect of the aIIb(R995A) mutation was dramat-
ically reduced (Figure 3B).
Mutation of F3 Residues that Contact the b Tail
Diminish the Ability of Talin to Activate Integrin
Mutations to the contact residues in talin F3 (L325R,
S365D, S379R, or Q381V) were also made to see if they
disrupted the F3-MP interaction. Mutants were correctly
folded, as judged by their dispersed NMR spectra (data
not shown), and chemical shifts induced by addition of
chimeric peptide indicated that binding to the MD site
was unchanged. In contrast, perturbation of contact resi-
dues in the MP region was either markedly reduced, as
with F3(Q381V) and b3(F727A), or undetectable, as with
F3(L325R) and b3(F730A) (Figure 4). Transfection of cells
with cDNA encoding the F2 and F3 subdomains (F23) of
talin (residues 206–405) is known to activate several
integrins (Calderwood et al., 2002), but each of the four
mutations diminished the ability of F23 to activate integrin
aIIbb3 (Figures 3C and S4). These results establish the
Figure 2. Structure of the Talin F3-Chi-
meric Peptide Complex
(A) Stereo view of 20 calculated structures
superimposed along the well-ordered residues
of the talin F3 domain (15N-1H-NOEs > 0.65).
(B) Ribbon representation of the complex
showing the approximate location of the MD-
and MP-binding sites.
(C) The talin F3 domain is shown in white as an
accessible surface representation with resi-
dues that were mutated for subsequent exper-
iments indicated in color: L325 (green), W359
(pink), S365 (purple), S379 (blue), and Q381
(yellow). The chimeric peptide backbone (resi-
dues 723–746), as well as the side chains of
the two phenylalanine residues (F727 and
F730), are shown in red. Overlaid on the com-
plex structure in cyan is the position of the
b3-integrin tail fragment found in the crystal
structure 1MK7 (Garcia-Alvarez et al., 2003).
The critical tryptophan (W739) side chains
from the crystal structure and the NMR struc-
ture are also shown.
174 Cell 128, 171–182, January 12, 2007 ª2007 Elsevier Inc.
biological relevance of the interactions seen between the
talin F3 protein and the MP region and show that these
interactions are critical for integrin activation.
Talin Mutants that Inhibit Integrin Activation
The observation that talin mutants with impaired contacts
to the MP region still bind to the chimeric peptide via the
MD site suggested that they might compete with and dis-
place endogenous talin, thus inhibiting integrin activation.
To investigate this possibility, we used abPy cells from a
chinese hamster ovary (CHO) cell line that expressed a
talin-dependent, constitutively active chimeric integrin
that was composed of the extracellular and transmem-
brane domains of aIIbb3 as well as the cytoplasmic tails of
a6b1 (Baker et al., 1997). We transiently transfected abPy
cells with cDNAs encoding the mutant F23 constructs
Figure 3. Talin-Mediated Integrin Activa-
tion Requires Specific Contacts between
the b3-MP Region and Talin F3
(A and B) CHO cells were transiently trans-
fected with plasmids encoding aIIb and b3 or
mutants thereof. In some samples (indicated
above) a plasmid encoding a shRNA for talin-
1 (Tadokoro et al., 2003) or a control shRNA
was cotransfected with the integrin plasmids.
Cells were double-stained for integrin-aIIbb3
expression (B-D57) and activated aIIbb3-
(PAC1). Flow cytometry was used to measure
the geometric mean fluorescence intensity (F)
of PAC1 binding. Nonspecific PAC1 binding
itive antagonist of aIIbb3, Ro43-5054, and
maximal binding (Fmax) was measured in the
presence of the activating antibody anti-
LIBS6. Activation index was defined as 100 3
(F ? F0) / (Fmax? F0) (see Experimental Proce-
dures). (A) Activation of aIIb(R995A) b3 integrin
was diminished in the cells cotransfected with
b3(Y747A) mutation blocks talin binding to the
MD site, and the aIIb(R995A)b3(Y747A) integrin
is not activated. (B) Talin-mediated activation
requires specific contacts between talin F3
and the b3-MP region. Depicted is the activa-
tion index (gray bars) or the aIIbb3 expression,
indicated as fluorescence intensity of D57
staining (black bars). The activating effect of
aIIb(R995A) is diminished when it is paired
(C and D) CHO cells stably expressing WT hu-
man aIIbb3 (A5 cells) or a high-affinity chimeric
integrin (abPy cells) were transiently cotrans-
fected with vectors for WT or mutant HA-
tagged talin F23 and eGFP protein as a trans-
fection marker. PAC1 binding was measured
by flow cytometry in the subset of cells
expressing GFP. Insets show representative
western blots of cell lysates probed with anti-
HA antibody and indicate that WT and mutant
HA-talin F23 expressed to similar levels. (C)
Specific PAC1 binding in aIIbb3-expressing
A5 cells transfected with talin F23 mutants rel-
ative to that in cells transfected with WT talin
F23 (set as 100). Mutants to MP-contacting
residues (L325R, S365D, S379R, and Q381V)
and MD-contacting residues (W359A) prevent
activation. (D) PAC1 binding in aIIba6b3b1-
expressing abPy cells transfected with talin
F23 or its mutants relative to that in cells transfected with empty vector (set as 100). The MP-contacting talin F23 mutants (L325R, S365D,
S379R, and Q381V) inhibited activation, whereas the MD-contacting mutant talin F23(W359A) had no significant effect on activation levels. All
data are an average of three of more measurements; error bars are standard deviations.
Cell 128, 171–182, January 12, 2007 ª2007 Elsevier Inc. 175
(L325R, S365D, S379R, or Q381V). Each of the four mu-
tants blocked integrin activation (Figure 3D), a result con-
sistent with competition between endogenous full-length
talin and the activation-defective F23 mutants for binding
to the b-MD site. In contrast, transfection of these cells
with cDNA encoding WT F23 modestly increased activa-
tion, whereas a mutant that strongly inhibits the MD-bind-
ing site, F23(W359A), had no effect on activation, which is
consistent with its markedly reduced affinity for the b tail
To assess whether other PTB domains could recapitu-
late the dominant-negative effect of the talin mutants,
abPy cells were transfected with the PTB-containing
protein DOK1. As predicted, this protein inhibited integrin
activation (Figure 5A), which is consistent with the idea
that it competes with endogenous talin for the MD site
but does not induce activation. Comparison of the struc-
tures of the DOK1 and talin PTB domains shows that they
differ primarily in the region between strands S1 and S2
(Figure 5B), which is consistent with the idea that this
region is vital for activation. Titration of the DOK1 PTB
domain with the b3/PIPKIg peptide revealed interactions
with S5, S6, and H2, which is indicative of binding to the
MD region, whereas no interactions with the S1 to S2
regions were observed (Figure 5C). These results explain
the unique ability of talin to activate integrins and provide
a structural explanation for the capacity of certain other
PTB-domain-containing proteins (Huang et al., 2004) to
inhibit integrin function.
Thermodynamic parameters for the binding of WT and
mutant talin F3 to the b3-chimeric peptides were mea-
sured using isothermal titration calorimetry (ITC; Table S2
and Figure S5). In general, mutations that disrupt the MP
site, either in the b tail or talin F3, have relatively little effect
on the Kdvalues (Table S2). This probably arises because
of a balance between enthalpy gain and entropy loss,
which is consistent with stabilization of the MP helix on ta-
lin binding in the WT but not in mutant structures. In sharp
contrast, the talin (W359A) mutant, which inhibits binding
of the MD site of the b3 tail, reduced the affinity ?1000-
fold, similar to the effect of mutations in the NPxY region
of the integrin tail that abrogate MD binding (Ulmer et al.,
2003). These results suggest that the MD site provides a
substantial fraction of the binding energy and explain why
compete with talin.
Figure 4. Weighted Shift Maps Obtained
Talin F3 Constructs on Addition of b3/
Talin F3 L325R and Q381V mutants were ti-
trated with the original b3/PIPKIg chimera,
while the effect of peptide mutations (F727A
and F730A) was assessed using WT talin F3.
The shift maps (see Experimental Procedures)
indicate that binding to the MP region (via the
talin loop between S1 and S2) is markedly re-
duced for the Q381V and F727A mutants or
completely abolished for the L325R and
F730A mutants, while interactions with the
MD region are retained.
1H-15N-HSQC Spectra of
176 Cell 128, 171–182, January 12, 2007 ª2007 Elsevier Inc.
An Activating Mutant in the Transmembrane Domain
Is Talin Independent
We previously proposed that talin binding to the b subunit
cytoplasmic domain results in a mismatch in the packing
of the transmembrane regions of a and b subunits (Par-
tridge et al., 2005). Modeling and mutational approaches
indicate that interactions between the transmembrane re-
gions stabilize the integrin low-affinity state (Bennett,
2005; Gottschalk, 2005; Luo et al., 2005; Partridge et al.,
2005). Another interaction that helps maintain the inactive
conformation is the aIIb(R995)/b3(D723) MP salt bridge
(Hughes et al., 1996). We note, however, that while break-
ing this link shifts the equilibrium toward the active state, it
does not by itself promote full activation, which requires
endogenous talin (Tadokoro et al., 2003). To test if trans-
membrane mutationsactivate in atalin-independent man-
ner, the apolar-to-charged mutation b3(L712R) was ana-
lyzed. A combination of this mutant with b3(Y747A),
a mutation that prevents binding to the talin F3-MD site
confirming that disruption of transmembrane helix pack-
ing activates integrin aIIbb3 independent of talin binding.
A combination of NMR studies, together with cell-based
functional assays, has revealed how the talin F3 domain
is uniquely designed to activate integrins. The talin F3 do-
main forms a well-defined complex with the helix-forming
MP region of the b-integrin tail, and this interaction holds
the key to the molecular recognition required for activa-
tion. Mutations in integrin or talin that inhibit this interac-
tion in vitro also prevent integrin activation in cells, and
mutants with intermediate functional effects in cells retain
Figure 5. b-Tail Binding to the DOK1 PTB Domain Inhibits Activation of a High-Affinity Integrin
(A) CHO cells stably expressing a high-affinity chimeric integrin (abPy cells) were transiently cotransfected with empty vector (pCDNA3.1), WT talin
F23, or full-length DOK1; eGFP protein was used as a transfection marker. The level of integrin activation was measured by flow cytometry of har-
vested cells using PAC1 antibody as described in the legend to Figure 3. Data are an average of three or more measurements; error bars are standard
(B) Overlay of the DOK1 (cyan) and talin (yellow) PTB domains. The critical differences between the two domains are the residues between strands S1
PTB domain (10:1 molar ratio). Gray columns indicate residues whose peaks experienced severe line broadening in the bound state and could not be
Cell 128, 171–182, January 12, 2007 ª2007 Elsevier Inc. 177
a partial ability to form the F3-MP interaction (compare
Figures 3 and 4). These findings are consistent with previ-
ous studies that identified a variety of activating mutations
within the MP region of integrins, thus establishing that
this region is critical for stabilizing the low-affinity confor-
mation (Ginsberg et al., 2005; Partridge et al., 2005).
The unique feature of talin F3 that promotes interaction
with the MP region of b-integrin appears to be the flexible
loop betweenstrandsS1and S2thatforms ahydrophobic
pocket that accepts the side chains of F727 and F730 in
the complex. The talin mutation L325R within this pocket
abolishes binding to the MP region (Figure 4) and thus hin-
ders activation (Figure 3C). Other PTB-domain-containing
proteins (DOK [‘‘IRS-like’’], Shc [’’Shc-like’’], and NUMB
[’’Dab-like’’]; Uhlik et al., 2005) lack such a mobile loop
and so are unlikely to interact with the MP region
(Figure 5B). Indeed, we showed that DOK binds and in-
hibits integrin activation and that binding occurs to the
MD site but not the MP site (Figure 5). Several point mu-
tants in talin also transform the F3 domain from an activa-
tor into an inhibitor (Figure 3D). This is consistent with
competition between endogenous talin and various PTB
domains, including mutated talin F3, for the MD site. The
concept of inhibition of talin activation by competitive
binding to the b tails by a variety of proteins, especially
those with PTB domains, may be an important feature in
the regulation of integrin activity. These conclusions also
apply to PTB (F3)subdomains in other FERM-domain pro-
teins. The FERM-domain structures of band 4.1, ezrin,
radixin, and moesin all have very short loops between
strands 1 and 2 with no hydrophobic residues that could
bury the two integrin phenylalanines (Edwards and
Keep, 2001; Hamada et al., 2000; Han et al., 2000; Smith
et al., 2003).
Integrin activation is critical for a variety of pathological
events such as thrombosis, inflammation, and tumor me-
tastasis (Campbell and Ginsberg, 2004). The interface be-
tween the MP region of b3 and talin suggests that it might
be readily accessible to pharmacological inhibition. The
inhibitory talin mutants provide interesting tools to study
talin, as they should maintain other talin functions, includ-
ing the formation of links to the cytoskeleton and other
focal adhesion proteins.
How does the talin/MP interaction lead to activation?
One possibility is that the configuration of the complex
disrupts the putative salt bridge between aIIb(R995) and
b3(D723). Several distinct models for the a-b tail complex
have been published, and the F3-MP complex would
Figure 6. Effect of a Transmembrane
Integrin Mutation and Talin S1-S2 Loop
Lysine Mutations on Integrin Activation
(A) CHO cells were transiently transfected with
plasmids encoding aIIb and b3 or mutants
thereof, and the cells were double-stained
with PAC1 and D57 to measure integrin activa-
tion and aIIbb3 expression, respectively. Integ-
legend to Figure 3. Both the putative trans-
and the salt-bridge-breaking aIIb(R995A) mu-
tation result in an activated integrin. However,
when paired with a mutation that disrupts talin
binding to form aIIb(R995A)b3(Y747A), the
R995A mutation loses the ability to induce the
high-affinity state, whereas the b3(L712R) mu-
tation’s activation level is unchanged in the
aIIbb3(L712R, Y747A) integrin.
(B) CHO cells stably expressing aIIbb3 (A5
cells) were transiently cotransfected with vec-
tors for WT or mutant HA-tagged talin F23
and eGFP protein as a transfection marker.
PAC1 binding was assessed in the GFP-
expressing subset of cells by flow cytometry
and analyzed as described in the legend to
Figure 3. Mutation of lysine K320 (K320D), a
residue that is predicted to point away from
the membrane, had no effect on activation,
while mutation of K322 (K322D), a residue
that is predicted to be directed toward the
membrane, prevented activation. The inset to
(B) shows representative western blots of cell
lysates probed with anti-HA antibody and
indicate that WT and mutant HA-talin F23
expressed to similar levels. The data are an
average of three of more measurements; error
bars are standard errors of the mean.
178 Cell 128, 171–182, January 12, 2007 ª2007 Elsevier Inc.
sterically prevent the formation of the salt bridge in some
of these (Gottschalk, 2005; Vinogradova et al., 2002; Wel-
jie et al., 2002; Table S3). We note, however, that while
breaking this salt bridge seems to be necessary, it is not
sufficient for full activation (Figure 3A; Tadokoro et al.,
2003). Prevention of b3 binding to the MD site by the
b3(Y747A) mutation also inhibits activation of the aII-
b(R995A)b3 (Figure 3) and b3(D723R) forms of integrin
(data not shown). Talin binding to the MD site might con-
tribute to activation by displacing this part of the integrin
tail from a membrane-anchored position (Vinogradova
et al., 2004); however, it is hard to reconcile this idea
with the inactive state of the aIIb(R995A)b32(F727A) or
aIIb(R995A)b3(F730A) integrins (Figure 3B) in spite of de-
stabilization of the salt bridge and talin still binding via
the MD site. In contrast, our data suggest that the primary
function of the talin/MD interaction is to provide an initial
strong linkage between talin and integrin and that activa-
tion arises from the subsequent talin/MP interaction.
We have proposed that activation involves disruption of
transmembrane helix interactions (Partridge et al., 2005).
A related model with the additional possibility of homo-
and heterodimeric interactions has been proposed by Li
et al. (2005), and interactions between membrane-span-
ning regions have also been modeled (Gottschalk, 2005)
and implied byleucine mutagenesis (Luoet al.,2005). Gly-
cosylation mapping was used to explore whether the
membrane-spanning regions changed positions in the
membrane (Stefansson et al., 2004). Here we made a mu-
tation in the membrane-spanning region, b3(L712R),
which is expected to reposition the transmembrane helix,
thus allowing the arginine guanidinium group to snorkel
out of the bilayer into a more hydrophilic environment.
The activating yettalin-independent
b3(L712R) (Figure 6) is consistent with integrin activation
that involves repositioning of the b-transmembrane helix.
librium between the inactive and active states, one possi-
bility is that the b-integrin subunit transmembrane domain
is dynamic, bobbing in and out of the membrane, with
its transmembrane helices sampling various degrees of
burial within the bilayer. Formation of a stable MP-b3 helix
in intimate contact with talin F3 would promote a confor-
mation in which transmembrane domain residues are fur-
ther out of the membrane than in the ’’off state,’’ pushing
the equilibrium toward the active conformation.
The structure-function analysis reported here provides
a cogent structural model to explain talin-dependent in-
tegrin activation (Figure 7). When the F3 domain engages
the b-MP region, additional favorable electrostatic con-
tacts between F3 and the lipid head groups can be
made. For example, two lysine residues (316 and 322)
withinthe largeS1-S2 loop, whichcontacts theN-terminal
end of the b-MP helix, would point toward the acidic head
groups of the membrane bilayer. K343 and K345 from the
flanking loop also contribute to making the MP surface of
Figure 7. Model of Talin-Induced Integrin Activation
(A) The talin F3 domain (surface representation; colored by charge), freed from its autoinhibitory interactions in the full-length protein, becomes avail-
able for binding to the integrin.
(B) F3 engages the MD part of the b3-integrin tail (in red), which becomes ordered, but the a-b integrin interactions that hold the integrin in the
low-affinity conformation remain intact.
(C) In a subsequent step, F3 engages the MP portion of the b3 tail while maintaining its MD interactions. Consequences of this additional interaction
are (1) destabilization of the putative integrin salt bridge; (2) stabilization of the helical structure of the MP region; and (3) electrostatic interactions
between F3 and the acidic lipid head groups. The net result is a change in the position of the transmembrane helix, which is continuous with the
MP-b-tail helix. This position change causes a packing mismatch with the aIIb-transmembrane helix, separation or reorientation of the integrin tails,
and activation. Mutants of F3 that have compromised interactions with the MP region and other PTB domains that lack an MP-binding site stall at
point B, consistent with their dominant-negative behavior.
Cell 128, 171–182, January 12, 2007 ª2007 Elsevier Inc. 179
talin F3 highly basic. When the b-MP and transmembrane
regions are modeled as a continuous a helix and the F3 ly-
sines are brought into apposition with a model membrane
is H722, in agreement with glycosylation-mapping studies
(Stefansson et al., 2004). To test this model, we mutated
two basic residues in the S1-S2 loop: mutation of K322
(K322D), which points toward the bilayer in our model,
prevented talin activation, while mutation of K320
(K320D), which points away from the bilayer, had no effect
(Figure 6B). We speculate, therefore, that F3-membrane
interactions, together with the formation of a b-tail helix
in intimate contact with F3, make significant contributions
to the energy required to stabilize the integrin-activated
state. In the context of a complete FERM domain, the ori-
entation and location of F3, in which the MP helix points
vertically through the membrane, also bring the surfaces
of the F1 and F2 domains into close apposition with the
membrane. We note that in the F2 domain, four lysines
(263, 268, 272, and 274) also appear appropriately posi-
tioned to interact with the membrane. This orientation of
the FERM domain on the membrane surface is similar to
that proposed previously for a radixin-ICAM-2 complex
(Hamada et al., 2003).
In summary, we have demonstrated the structural basis
for the unique ability of talin to activate integrins. The spe-
cific interface identified here completes the molecular pic-
ture of the talin/b3 interaction and provides a target for the
design of therapeutics aimed at disrupting integrin activa-
tion while leaving talin’s other activities intact. Finally, we
provide a structural template for mutational studies to de-
fine the biological role of talin’s ability to activate integrins
in cells and, ultimately, in whole animals.
Peptides were synthesized commercially by Alta Bioscience (Birming-
ham, England, UK) or EZBiolab (Westfield, IN, USA) or produced as
a GST-b3 recombinant fusion protein with the peptide liberated by
thrombin digestion. All peptides were further purified using reverse-
phase high-performance liquid chromatography and validated by
electrospray-ionization mass spectrometry.
U-15N-labeled talin F3 domain was expressed and purified as de-
scribed previously (de Pereda et al., 2005). U-15N-labeled DOK1 PTB
domain (residues 154–256; Swiss-Prot Q99704) was subcloned into
the pGEX-6P2 vector and produced using the same methods.
U-15N,13C and U-15N,2H doubly labeled talin was produced in the
same way, using13C-glucose or D6-glucose and D2O.
All NMR experiments were carried out at 25?C on spectrometers
equipped with Oxford Instruments superconducting magnets (500,
600, and 750 MHz
1H-15N-HSQC titrations wereperformedas previously described (de
Pereda et al., 2005). Weighted shifts (Dd [HN,N]) were calculated using
the equation: Dd [HN,N] = ([DdHNWHN]2+ [DdNWN]2)1/2, where WHNand
1H operating frequencies) and GE Omega
WN= 0.154; Ayed et al., 2001), and Dd = dbound– dfree.Talin F3 domains
in complex with slow-exchanging peptides as well as mutant F3 do-
mains were assigned by correlation to the WT talin F3 assignments
(de Pereda et al., 2005) and confirmed using three-dimensional (3D)
gradient-enhanced [1H-15N]-NOESY-HSQC (tm = 100 ms) spectra
and/or [1H-15N]-TOCSY-HSQC (tm = 70 ms) spectra.
and13Cb shifts were obtained using a series of 2D1H-15N-13C exper-
iments (Bersch et al., 2003). Bound peptide shifts were obtained using
U-15N,2H F3 and unlabeled chimeric peptide (1.3:1). Both bound and
(2D) NOESY, TOCSY, and COSY spectra. DOK PTB domain assign-
ments were obtained previously (unpublished data).
NOE data for structure calculations were obtained from a [1H-15N]-
NOESY-HSQC spectrum on U-15N F3 and unlabeled chimeric peptide
(1:1.2), a 2D NOESY spectrum on unlabeled F3 and chimeric peptide
(1:1) in 99.96% D2O, and 2D NOESY spectra from a sample of
U-15N,2H F3 and unlabeled chimeric peptide (1.3:1) in both 90%
H2O/D2O and 99.96% D2O (tm= 100 ms in each case). Intermolecular
NOEs were identified in each NOESY spectrum, apart from that of
U-15N,2H F3 and unlabeled chimeric peptide in D2O. All samples con-
tained 50 mM phosphate buffer (pH 6.1) and 100 mM NaCl.
3JHNHA-coupling constants were determined from peak-fitting anal-
15N-1H-NOEs were determined from a pair of HSQC experiments re-
corded with and without1H saturation during the recycle delay (Farrow
et al., 1994). NMR experiments were processed using NMRPipe (Dela-
glio et al., 1995) and visualized with Sparky (www.cgl.ucsf.edu/home/
15N-HMQCJ spectra (Redfield et al., 1991). Heteronuclear
NOE distance restraints were calibrated on the basis of NOEs from
regular secondary-structure elements and were grouped into four
classes (1.8–3.0 A˚, 1.8–3.8 A˚, 1.8–4.6 A˚, and 1.8–6.0 A˚) corresponding
to strong, medium, weak, and very weak NOEs. Dihedral restraints
(F/c) were obtained from the TALOS database (Cornilescu et al.,
1999). Additional F angle restraints were derived from the3JHNHA-
coupling constants and a modified Karplus equation (Pardi et al.,
1984). Hydrogen-bond restraints were incorporated based on hydro-
gen-exchange data and secondary-structure elements identified
from initial rounds of structure calculation.
Structure calculations were performed using a simulated annealing
protocol within the program CNS, Version 1.1 (Brunger et al., 1998).
The initial conformation of the talin F3 domain was that from the crystal
structure of talin in complex with the PIPKIg peptide (de Pereda et al.,
2005), while the chimera peptide started from randomized coordi-
nates. One hundred structures were calculated, and the top 20 of
these were further refined. Floating assignment of prochiral groups
was achieved using the SOPHIE procedure (Pickford et al., 2001). Dur-
ing final minimization each group was eased into the pro-R or pro-S
position by enforcing the correct bond angles at the prochiral center.
The stereochemical quality of the structures was assessed using the
program PROCHECK-NMR (Laskowski et al., 1996). Molecular
models were generated using the program MOLMOL (Koradi et al.,
Cell Culture, Cell Lines, and Reagents
CHO cells were obtained from American Type Culture Collection
(ATCC) and cultured in Dulbecco’s Modified Eagle’s Medium with
10% fetal bovine serum (FBS), 1% nonessential amino acids (Sigma),
penicillin (50 units/ml), and streptomycin sulfate (50 mg/ml) in a 37?C
tissue-culture incubator. The anti-aIIbb3 antibodies D57, PAC1, and
anti-LIBS6, as well as Ro43-5054, an aIIbb3-specific peptide-mimetic
competitive inhibitor, have been described (Tadokoro et al.,2003).The
D57 antibody was biotinylated with biotin-N-hydroxy-succinimide
(Sigma; B-D57) according to the manufacturer’s instructions.
180 Cell 128, 171–182, January 12, 2007 ª2007 Elsevier Inc.
Site-directed mutations in both the aIIbb3 subunits and the talin F3
construct were generated using the QuikChange mutagenesis kit
(Stratagene). Mutants were confirmed by DNA sequencing.
For experiments to assess the functional effect of point mutants in the
b3tail, CHO cells werecotransfected with pCDM8 plasmids coding for
WT or mutant sequences of aIIb and b3. Transfection was done using
the Plus Reagent and Lipofectamine (Invitrogen Life Technologies); 24
hr later cells were stained with B-D57 and PAC1 in the presence and
absence of Ro43-5054 or anti-LIBS6. The biotinylated monoclonal an-
tibody B-D57 was used to detect expression of aIIbb3, while PAC1, an
activation-specific, monoclonal IgM antibody, was used to assess the
activation state of the aIIbb3 integrin. This antibody is an authentic li-
gandfor integrin aIIbb3(Abrams etal.,1994),and itsbinding correlates
with the binding of natural ligands such as fibrinogen (Shattil et al.,
1985). Cells were analyzed on a FACScan using both B-D57 and
PAC1 antibodies as described previously (Partridge et al., 2005). The
geometric mean fluorescence intensity (MFI) of PAC1 staining in the
presence of Ro43-5054 (2 mM) was used to estimate nonspecific
PAC1 binding (F0). The fluorescence intensity in the presence of anti-
LIBS6 was used to estimate maximal PAC1 binding (Fmax) since anti-
LIBS6 directly induces aIIbb3 binding to PAC1 regardless of the status
of cellular activation mechanism (Baker et al., 1997). The activation in-
dex was calculated using the formula: 100 3 (F ? F0) / (Fmax? F0),
where F = MFI under the test condition.
For experiments to assess the functional effect of F3-domain muta-
tions, WT or mutant versions of pCDNA3.1 plasmids coding for N-ter-
minally HA-tagged proteins consisting of the F2 and F3 domains of
mouse talin (residues 206–405; Swiss-Prot P26039) were transiently
transfected into cells stably expressing WT aIIbb3, an aIIb(D723A)b3
mutant or an aIIba6Ab3b1A-chimeric integrin (abPy cells; Baker
et al., 1997). In these experiments eGFP vector was cotransfected
as a transfection marker at a HA-F23/eGFP ratio of 20:1.
ITC was performed on a VP-ITC calorimeter (Microcal, Northampton,
MA, USA). Aliquots (8 ml) of WT (KKLLITIHDRKEFAKFEEERARAKWV-
pYSPLHYSAR) or mutant chimeric peptide were injected into the cell
containing WT or mutant talin F3 domain. Peptide concentrations
were typically 150 to 300 mM, while protein concentrations ranged
from 25 to 50 mM. For the W359A mutant, the protein concentration
was 50 mM, and the peptide concentration was 1 mM. Prior to ITC
titrations, the peptides and proteins were in 200 mM Tris, 300 mM
NaCl, pH 7.5. In each experiment 37 injections were made. The exper-
iments were performed at 23?C. All titrations were performed at
least twice and with different protein preparations. Experimental data
were analyzed using MicroCal Origin software.
Supplemental Data include five figures and three tables and can be
found with this article online at http://www.cell.com/cgi/content/full/
I.D.C., K.W., M.H.G., and R.C.L. acknowledge support from the NIH
Cell Migration Consortium. I.D.C. is also supported by The Wellcome
Trust. A.W.P. acknowledges support from the Tobacco-Related
Disease Research Program. M.H.G.’s laboratory is supported by
grants from the NIH. We thank Camilla Oxley for contributions to the
DOK data; Roman Melnyk for careful reading of the manuscript; and
Wilma Puzon-McLaughlin and Nima Yousefi for valuable technical
Received: July 6, 2006
Revised: September 5, 2006
Accepted: October 20, 2006
Published: January 11, 2007
Determinants of specificity of a baculovirus-expressed antibody Fab
fragment that binds selectively to the activated form of integrin alpha
IIb beta 3. J. Biol. Chem. 269, 18781–18788.
Ayed, A., Mulder, F.A., Yi, G.S., Lu, Y., Kay, L.E., and Arrowsmith, C.H.
(2001). Latent and active p53are identical inconformation.Nat. Struct.
Biol. 8, 756–760.
Baker, E.K., Tozer, E.C., Pfaff, M., Shattil, S.J., Loftus, J.C., and
Ginsberg, M.H. (1997). A genetic analysis of integrin function:
Glanzmann thrombasthenia in vitro. Proc. Natl. Acad. Sci. USA 94,
Bennett, J.S. (2005). Structure and function of the platelet integrin
alphaIIbbeta3. J. Clin. Invest. 115, 3363–3369.
Bersch, B., Rossy, E., Coves, J., and Brutscher, B. (2003). Optimized
set of two-dimensional experiments for fast sequential assignment,
secondary structure determination, and backbone fold validation of
13C/15N-labelled proteins. J. Biomol. NMR 27, 57–67.
Brunger, A.T., Adams, P.D., Clore, G.M., DeLano, W.L., Gros, P.,
Grosse-Kunstleve, R.W., Jiang, J.S., Kuszewski, J., Nilges, M., Pannu,
N.S., et al. (1998). Crystallography & NMR system: A new software
suite for macromolecular structure determination. Acta Crystallogr.
D Biol. Crystallogr. 54, 905–921.
Calderwood, D.A., Yan, B., de Pereda, J.M., Alvarez, B.G., Fujioka, Y.,
Liddington, R.C., and Ginsberg, M.H. (2002). The phosphotyrosine
binding-like domain of talin activates integrins. J. Biol. Chem. 277,
Calderwood, D.A., Fujioka, Y., de Pereda, J.M., Garcia-Alvarez, B.,
Nakamoto, T., Margolis, B., McGlade, C.J., Liddington, R.C., and
Ginsberg, M.H. (2003). Integrin beta cytoplasmic domain interactions
with phosphotyrosine-binding domains: a structural prototype for
diversity in integrin signaling. Proc. Natl. Acad. Sci. USA 100, 2272–
Campbell, I.D., and Ginsberg, M.H. (2004). The talin-tail interaction
places integrin activation on FERM ground. Trends Biochem. Sci.
Chishti, A.H., Kim, A.C., Marfatia, S.M., Lutchman, M., Hanspal, M.,
Jindal, H., Liu, S.C., Low, P.S., Rouleau, G.A., Mohandas, N., et al.
(1998). The FERM domain: a unique module involved in the linkage
of cytoplasmic proteins to the membrane. Trends Biochem. Sci. 23,
Cornilescu, G., Delaglio, F., and Bax, A. (1999). Protein backbone
angle restraints from searching a database for chemical shift and
sequence homology. J. Biomol. NMR 13, 289–302.
to talin function. Biochem. Soc. Trans. 33, 1308–1312.
de Pereda, J.M., Wegener, K.L., Santelli, E., Bate, N., Ginsberg, M.H.,
Critchley, D.R., Campbell, I.D., and Liddington, R.C. (2005). Structural
basis for phosphatidylinositol phosphate kinase type Igamma binding
to talin at focal adhesions. J. Biol. Chem. 280, 8381–8386.
Delaglio, F.,Grzesiek, S.,Vuister, G.W., Zhu,G.,Pfeifer, J., and Bax,A.
(1995). NMRPipe: a multidimensional spectral processing system
based on UNIX pipes. J. Biomol. NMR 6, 277–293.
activated FERM domain of moesin: an analysis of structural changes
on activation. Biochemistry 40, 7061–7068.
Farrow, N.A., Muhandiram, R., Singer, A.U., Pascal, S.M., Kay, C.M.,
Gish, G., Shoelson, S.E., Pawson, T., Forman-Kay, J.D., and Kay,
Cell 128, 171–182, January 12, 2007 ª2007 Elsevier Inc. 181
L.E. (1994). Backbone dynamics of a free and phosphopeptide- Download full-text
complexed Src homology 2 domain studied by 15N NMR relaxation.
Biochemistry 33, 5984–6003.
Garcia-Alvarez, B., de Pereda, J.M., Calderwood, D.A., Ulmer, T.S.,
Critchley, D., Campbell, I.D., Ginsberg, M.H., and Liddington, R.C.
(2003). Structural determinants of integrin recognition by talin. Mol.
Cell 11, 49–58.
Ginsberg, M.H., Partridge, A., and Shattil, S.J. (2005). Integrin regula-
tion. Curr. Opin. Cell Biol. 17, 509–516.
Gottschalk, K.E. (2005). A coiled-coil structure of the alphaIIbbeta3
integrin transmembrane and cytoplasmic domains in its resting state.
Structure 13, 703–712.
Hamada, K., Shimizu, T., Matsui, T., Tsukita, S., and Hakoshima, T.
(2000). Structural basis of the membrane-targeting and unmasking
mechanisms of the radixin FERM domain. EMBO J. 19, 4449–4462.
Hamada, K., Shimizu, T., Yonemura, S., Tsukita, S., Tsukita, S., and
Hakoshima, T. (2003). Structural basis of adhesion-molecule recogni-
tion by ERM proteins revealed by the crystal structure of the radixin-
ICAM-2 complex. EMBO J. 22, 502–514.
Han, B.G., Nunomura, W., Takakuwa, Y., Mohandas, N., and Jap, B.K.
(2000). Protein 4.1R core domain structure and insights into regulation
of cytoskeletal organization. Nat. Struct. Biol. 7, 871–875.
Horwitz, A., Duggan, K., Buck, C., Beckerle, M.C., and Burridge, K.
(1986). Interaction of plasma membrane fibronectin receptor with
talin–a transmembrane linkage. Nature 320, 531–533.
Huang, C.L., Cheng, J.C., Liao, C.H., Stern, A., Hsieh, J.T., Wang,
C.H., Hsu, H.L., and Tseng, C.P. (2004). Disabled-2 is a negative
regulator of integrin alpha(IIb)beta(3)-mediated fibrinogen adhesion
and cell signaling. J. Biol. Chem. 279, 42279–42289.
Hughes, P.E., O’Toole, T.E., Ylanne, J., Shattil, S.J., and Ginsberg,
M.H. (1995). The conserved membrane-proximal region of an integrin
cytoplasmic domain specifies ligand binding affinity. J. Biol. Chem.
Hughes, P.E., Diaz-Gonzalez, F., Leong, L., Wu, C., McDonald, J.A.,
Shattil, S.J., and Ginsberg, M.H. (1996). Breaking the integrin hinge.
A defined structural constraint regulates integrin signaling. J. Biol.
Chem. 271, 6571–6574.
Hynes, R.O. (2002). Integrins: bidirectional, allosteric signaling
machines. Cell 110, 673–687.
Koradi, R., Billeter, M., and Wuthrich, K. (1996). MOLMOL: a program
for display and analysis of macromolecular structures. J Mol Graph.
14, 51–55, 29–32.
Laskowski, R.A., Rullmannn, J.A., MacArthur, M.W., Kaptein, R., and
Thornton, J.M. (1996). AQUA and PROCHECK-NMR: programs for
checking the quality of protein structures solved by NMR. J. Biomol.
NMR 8, 477–486.
Li, W., Metcalf, D.G., Gorelik, R., Li, R., Mitra, N., Nanda, V., Law, P.B.,
Lear, J.D., Degrado, W.F., and Bennett, J.S. (2005). A push-pull mech-
anism for regulating integrin function. Proc. Natl. Acad. Sci. USA 102,
Luo, B.H., Carman, C.V., Takagi, J., and Springer, T.A. (2005). Disrupt-
ing integrin transmembrane domain heterodimerization increases
ligand binding affinity, not valency or clustering. Proc. Natl. Acad.
Sci. USA 102, 3679–3684.
Papagrigoriou, E., Gingras, A.R., Barsukov, I.L., Bate, N., Fillingham,
I.J., Patel, B., Frank, R., Ziegler, W.H., Roberts, G.C., Critchley, D.R.,
and Emsley, J. (2004). Activation of a vinculin-binding site in the talin
rod involves rearrangement of a five-helix bundle. EMBO J. 23,
Pardi, A., Billeter, M., and Wuthrich, K. (1984). Calibration of the
angular dependence of the amide proton-C alpha proton coupling
constants, 3JHN alpha, in a globular protein. Use of 3JHN alpha for
Partridge, A.W., Liu, S., Kim, S., Bowie, J.U., and Ginsberg, M.H.
(2005). Transmembrane domain helix packing stabilizes integrin
alphaIIbbeta3 in the low affinity state. J. Biol. Chem. 280, 7294–7300.
Pickford, A.R., Smith, S.P., Staunton, D., Boyd, J., and Campbell, I.D.
(2001). The hairpin structure of the (6)F1(1)F2(2)F2 fragment from
human fibronectin enhances gelatin binding. EMBO J. 20, 1519–1529.
Redfield, C., Smith, L.J., Boyd, J., Lawrence, G.M., Edwards, R.G.,
Smith, R.A., and Dobson, C.M. (1991). Secondary structure and topol-
ogy of human interleukin 4 in solution. Biochemistry 30, 11029–11035.
Rees, D.J., Ades, S.E., Singer, S.J., and Hynes, R.O. (1990). Sequence
and domain structure of talin. Nature 347, 685–689.
Shattil, S.J., Hoxie, J.A., Cunningham, M., and Brass, L.F. (1985).
Changesin the platelet membrane glycoprotein IIb.IIIa complex during
platelet activation. J. Biol. Chem. 260, 11107–11114.
Smith, W.J.,Nassar,N.,Bretscher,A.,Cerione,R.A.,and Karplus, P.A.
(2003). Structure of the active N-terminal domain of Ezrin. Conforma-
tional and mobility changes identify keystone interactions. J. Biol.
Chem. 278, 4949–4956.
Stefansson, A., Armulik, A., Nilsson, I., von Heijne, G., and Johansson,
S. (2004). Determination of N- and C-terminal borders of the trans-
membrane domain of integrin subunits. J. Biol. Chem. 279, 21200–
Tadokoro, S., Shattil, S.J., Eto, K., Tai, V., Liddington, R.C., de Pereda,
J.M., Ginsberg, M.H., and Calderwood, D.A. (2003). Talin binding to
integrin beta tails: a final common step in integrin activation. Science
Tanentzapf, G., and Brown, N.H. (2006). An interaction between
integrin and the talin FERM domain mediates integrin activation but
not linkage to the cytoskeleton. Nat Cell Biol. 8, 601–606.
Uhlik, M.T., Temple, B., Bencharit, S., Kimple, A.J., Siderovski, D.P.,
and Johnson, G.L. (2005). Structural and evolutionary division of phos-
photyrosine binding (PTB) domains. J. Mol. Biol. 345, 1–20.
Ulmer, T.S., Calderwood, D.A., Ginsberg, M.H., and Campbell, I.D.
(2003). Domain-specific interactions of talin with the membrane-prox-
imal region of the integrin beta3 subunit. Biochemistry 42, 8307–8312.
Vinogradova, O., Velyvis, A., Velyviene, A., Hu, B., Haas, T., Plow, E.,
and Qin, J. (2002). A structural mechanism of integrin alpha(IIb)beta(3)
‘‘inside-out’’ activation as regulated by its cytoplasmic face. Cell 110,
Vinogradova, O., Vaynberg, J., Kong, X., Haas, T.A., Plow, E.F., and
Qin, J. (2004). Membrane-mediated structural transitions at the cyto-
plasmic face during integrin activation. Proc. Natl. Acad. Sci. USA
Weljie, A.M., Hwang, P.M., and Vogel, H.J. (2002). Solution structures
of the cytoplasmic tail complex from platelet integrin alpha IIb- and
beta 3-subunits. Proc. Natl. Acad. Sci. USA 99, 5878–5883.
Chemical shift assignments for the talin/chimeric peptide complex
have been deposited in the BioMagResBank with the accession num-
ber 7150. The coordinates of the calculated structure ensemble and
the minimized average structure have been deposited in the Protein
Data Bank with the ID codes 2H7D and 2H7E, respectively.
182 Cell 128, 171–182, January 12, 2007 ª2007 Elsevier Inc.