The inhibitory region of troponin-I alters the ability of F-actin
to interact with different segments of myosin
Valerie B. Patchell
1
, Clare E. Gallon
2
, Matthew A. Hodgkin
3
, Abdellatif Fattoum
4
, S. Victor Perry
1
and Barry A. Levine
1,2
1
Department of Physiology, School of Medicine and
2
School of Biosciences, University of Birmingham, Birmingham, UK;
3
School
of Biological Sciences, University of Warwick, Warwick, UK;
4
CRBM, CNRS, INSERM U249, F-34090 Montpellier, France
Peptides corresponding to the N-terminus of skeletal
myosin light chain 1 (rsMLC1 1–37) and the short loop of
human cardiac b-myosin (hcM398–414) have been shown
to interact with skeletal F-actin by NMR and fluorescence
measurements. Skeletal tropomyosin strengthens the
binding of the myosin peptides to actin but does not
interact with the peptides. The binding of peptides cor-
responding to the inhibitory region of cardiac troponin I

[1,2]. This implies that TnI prevents the interaction of actin
with the myosin head that leads to the activation of the
MgATPase. In the presence of tropomyosin, the inhibitory
influence of TnI is much increased and the maximum effect
is obtained when the stoichiometry approaches one mole-
cule of TnI to seven actin monomers [1–5]. When troponin
C and troponin T are absent this inhibition is calcium
insensitive [6] but otherwise corresponds to the ÔoffÕ state in
intact muscle.
The region of rabbit fast skeletal TnI represented by
residues 96–116, known as the inhibitory peptide (IP),
possesses properties that are very similar to the intact
molecule in that it binds to troponin C, and in the
presence of tropomyosin the inhibition of the MgATPase
of actomyosin by the peptide is markedly increased [7].
The inhibitory peptide in the presence of tropomyosin is
about 50% as effective as the intact TnI molecule when
assayed under similar conditions. Only about half of the
residues of IP, as originally isolated, appear to be essential
for this property because a synthetic duodecapeptide
comprising residues 104–115 (short IP) has equivalent
inhibitory activity [8]. Recent evidence suggests that
additional regions of TnI, C-terminal to the IP, may be
required to obtain inhibitory activity equal to the intact
molecule [9,10].
The mechanism of action of TnI on the regulation of
the contractile process is not as yet understood (see [11]
for a review). Despite the inhibitory properties of TnI the
current view is that tropomyosin regulates the actomyosin
ATPase in situ by a steric mechanism [12–14] and it has

surface of actin flanked on three sides by additional contacts
involving myosin surface loops [16]. One of these loops,
Pro402–Lys415, is modelled as interacting with actin near
residues 331–332 [16] at the junction of subdomains 1 and 3
of actin and appears to be important for normal muscle
activity.Deletionofthisloopregionresultedinthelossof
strong binding of myosin to actin [17] while a single amino
acid residue change, ArgfiGln, in this loop region of the
b-chain of human cardiac myosin is associated with familial
hypertrophic cardiomyopathy [18,19] and has been reported
to result in altered kinetic properties of the myosin
subfragment 1 ATPase [20].
Although there is no doubt that tropomyosin moves on
contraction it is difficult, in view of the somewhat limited
knowledge of the nature of the actin–myosin interaction, to
decide on the role of actin in the activation process. X-ray
analysis provides some evidence for movement of the actin
domains during contraction [21] and it is likely that in model
systems using mutant proteins the movement of tropo-
myosin observed in the presence of myosin and troponin is a
consequence of conformational changes in actin [22,23]. The
binding of ligands at discrete and specific binding sites on
actin during the contractile cycle would be expected to
induce conformational changes that influence its interaction
with myosin. Cross linking studies with the zero length
carbodiimidate reagent specific for lysine–carboxylate con-
tacts suggest that one such ligand, TnI, binds close to the
region represented by residues 1–12 of actin [24]. Proton
MR studies have indicated that IP interacts with residues
1–7 and 24–25 of the N-terminal region of actin [25]. These

using Fmoc polyamide chemistry and purified as described
previously [28]. The peptide comprising residues 398–414 of
human cardiac b-myosin was synthesized by Syntem
(Montpellier) and purified as reported previously [29]. The
composition and purity of all peptides was confirmed by
NMR and mass spectral analysis.
Muscle proteins
Freeze dried actin prepared by the method of Spudich and
Watt [30] was dissolved in 5 m
M
triethanolamine/HCl,
pH 8.0, 0.2 m
M
CaCl
2
,0.2 m
M
ATP, 0.2 m
M
dithiothreitol,
and dialysed for 3 h against the same buffer until fully
depolymerized. It was then centrifuged at 30 000 g for
30minandtheconcentrationoftheG-actininthe
supernatant determined by measuring absorbance at
290 nm using an absorption coefficient of 0.63 mgÆ
mL
)1
Æcm
)1
. The G-actin was polymerized by making the

rsMLC1 1–37 (MLC1) APKKDVKKPAAAAAAPAPAPAPAPAPAKPKEEKIDL
rsMLC1 1–13 (MLC1) APKKDVKKPAAAA
HA306–318 (HA peptide) PKYVKQATLKLAT
Ó FEBS 2002 F-actin interactions inhibited by Troponin-I (Eur. J. Biochem. 269) 5089
a stock solution of 1 mgÆmL
)1
rabbit skeletal tropomyosin
in 50 m
M
Tris/HCl, pH 7.6, 100 m
M
KCl, to give a final
concentration of 2.5 mgÆmL
)1
actin, 0.5 mgÆmL
)1
tropo-
myosin, i.e. a molar ratio of actin : tropomyosin of
approximately 7 : 1. The complex was dialysed into several
changes of 5 m
M
phosphate buffer, pH 7.0, in H
2
Oor
[
2
H]
2
O. Complex formation and the absence of free protein
was confirmed by comparison of the electrophoretic

M
. The actin was then centrifuged at 100 000 g for 1 h
and the pellet resuspended in Buffer A. This was dialysed
extensively against Buffer A to remove excess IAEDANS.
The concentration of the resulting G-actin was determined
using an absorption coefficient of 0.63 mgÆmL
)1
Æcm
)1
at
290 nm. A correction for the IAEDANS contribution at
290 nm was made using absorbance at 290 nm ¼
0.21 · absorbance at 336 nm, for bound IAEDANS.
The concentration of IAEDANS was determined using
the absorption coefficient of 6100
M
)1
Æcm
)1
at 336 nm. The
extent of labelling was normally 0.8–0.9 molÆmole
)1
G-actin. The labelled G-actin was polymerized by making
the solution 50 m
M
with respect to KCl, 2 m
M
with respect
to MgCl
2

fraction of the complexes formed. The accuracy of the K
d
values was gauged from curve fit obtained, the associated R
2
value (> 0.95) and the requirement that iterative fit of the
linear representation of the experimental data extrapolated
to an intercept value of 1.
Surface plasmon resonance
Direct binding of the TnI inhibitory peptide to actin was
investigated using surface plasmon resonance (SPR)
analysis to evaluate the association and dissociation rate
constants, K
a
and K
d
respectively, for the binding of the
peptide to immobilized F-actin using a BIAcore 3000
system. F-actin or BSA was covalently linked to carboxy-
methyldextran surfaces using standard amine coupling.
One surface was derivatized in the absence of protein.
Following immobilization the chip surfaces were capped
with ethanolamine and subject to surface equilibration
(BIApplications Handbook, 1993). Non-specific binding
was monitored using the control BSA and underivatized
flow cells. Sensorgrams were obtained using different
immobilization densities and the binding of the TnI
inhibitory peptide was assessed at various flow rates
(5–30 lLÆmin
)1
) and over a range of concentrations

dures. Spectra were obtained at 500 MHz on a Bruker
spectrometer at a sample temperature of 285K. Titration
of the peptides with F-actin was carried out by addition of
aliquots of F-actin (10 mgÆmL
)1
) or F-actin–tropomyosin
(5 mgÆmL
)1
F-actin). Titration of the inhibitory peptide
with F-actin or F-actin–tropomyosin was also carried out
by the addition of small aliquots (1–5 lL) of a stock
solution of the peptide to 0.5 mL of solution containing
F-actin at a concentration of 2.5–4.0 mgÆmL
)1
. The broad
signals of the spectrum of F-actin obtained at these
concentrations contributed relatively little to the spectra of
the peptides in the presence of actin. Two-pulse spin-echo
spectra (1024 transients) were obtained using a (180-t-90-t)
sequence with a delay time, t ¼ 60 ms, and an overall
interpulse delay of 3 s to enable complete magnetization
recovery. Signal amplitude in these experiments is modu-
lated by the corresponding coupling constant and relax-
ation time of each resonance and is a very sensitive
indicator of the effect of binding. As observed in previous
studies of actin binding [25,28] interaction results in
marked reduction of the bound peptide ligand resonances
consistent with the high molecular weight and slow
tumbling of the complex. Both direct signal linewidth
and the signal intensity in the two-pulse spin-echo spectra

have been reported elsewhere [25,33]. Resonance line width
changes can also originate, however, from any increase in
solution viscosity that significantly alters the rotational
diffusion of the ligand. As F-actin solutions have significant
viscosity we therefore first studied the effect of an increase in
solution viscosity on the linewidth characteristics of the IP.
Minimal spectral effects were observed for the hcTnI128–
153 peptide over a concentration range of 0–500 l
M
in 10%
(v/v) deuterated glycerol (MSD Isotopes). These observa-
tions indicated that viscosity effects on resonance and
linewidth in the peptide spectrum were not significant.
Evidence for the absence of viscosity effects on peptide
resonance and linewidth as a result the presence of actin
were obtained by comparing the spectrum of a control
peptide, the HA peptide (Table 1), in the presence and
absence of F-actin (Fig. 1). The absence of detectable
alterations in the spectrum of the peptide indicated that any
changes in viscosity due to the presence of F-actin have
negligible effects on the rotational diffusion in solution and
hence linewidth of the peptide resonances. These results
(Fig. 1) also served as control data indicating that there was
no nonspecific HA peptide interaction with F-actin.
Inspection of the spectrum of the HA peptide in the
presence of F-actin also indicates that although interaction
did not occur, there is a detectable contribution to signal
intensity deriving from F-actin at the relatively high
concentrations of the protein used in this control experiment
(Fig. 1). The broad signals of the spectrum of F-actin did

with F-actin
To investigate the interaction between actin and the
inhibitory region of TnI, we monitored the NMR spectral
Fig. 1. Proton magnetic resonance spectra demonstrating that the
presence of F-actin does not result in broadening of signals of peptide in
the absence of complex formation whilst interaction with F-actin results
in specific spectral changes. Spectra determined in 5 m
M
sodium
phosphate buffer, pH 7.4, T ¼ 285K. (A) HA306–318 peptide,
200 l
M
, (B) HA306–318 peptide, 200 l
M
, in the presence of F-actin,
200 l
M
. The spectral region between 1.2 and 1.4 p.p.m. under these
conditions is shown on an expanded scale as inset. The fine structure
for the HA306–318 peptide resonances is retained indicating lack of
interaction with F-actin and the absence of broadening due to non-
specific viscosity effects over the actin concentration range studied
(0–8 mgÆmL
)1
). Peak at 1.34 p.p.m. in inset B is due to actin. (C)
rcTnI161–181 peptide, 200 l
M
. (D) rcTnI161–181 peptide, 200 l
M
,in

slow tumbling of the complex formed. Similar results were
reported earlier [25] for the binding to F-actin by the
inhibitory peptide from rabbit fast skeletal muscle TnI
(residues 96–116) that differs from the homologous human
cardiac peptide by four conservative replacements. Since
almost all the resonances of the peptide hcTnI128–153 were
affected in the presence of F-actin (Fig. 3) the extent of the
spectral changes suggests that the entire length of the
peptide is constrained by attachment to the actin filament.
The kinetics of the interaction of the TnI inhibitory
region with actin were characterized using surface plasmon
resonance to monitor binding to immobilized F-actin. The
sensorgrams obtained recorded the association and disso-
ciation phases of the interaction (Fig. 4A). Analysis of the
dissociation phase for hcTnI128–153 peptide concentrations
in the range 1–10 l
M
gave an off rate constant of 10
3
Æs
)1
consistent with the NMR observation of fast exchange on
the relaxation time scale. The equilibrium constant for the
interaction was obtained by fitting the sensorgram data to a
model employing 1 : 1 complex formation. The value of the
dissociation constant derived, 3 l
M
(Table 2) was consistent
with an analysis of the dependence of the equilibrium
plateau signal on the concentration of the TnI inhibitory

peptide
hcTnI128–147, in the presence of 120 l
M
BSA. This spectrum is
indistinguishable from the algebraic sum of the individual spectra
(A + B) indicating lack of nonspecific interaction with BSA. Signals
deriving from the hcTnI128–147 are labelled.
Fig. 3. Interaction of the human cardiac TnI inhibitory peptide with
F-actin illustrated by proton magnetic resonance spectroscopy to show
the residues involved in complex formation. Spectra determined in 5 m
M
sodium phosphate buffer, pH 7.2, T ¼ 293K. (A) peptide hcTnI128–
153, 200 l
M
, (B) hcTnI128–153, 200 l
M
, in the presence of F-actin,
18 l
M
. (C) hcTnI128–153, 200 l
M
, in the presence of F-actin, 50 l
M
.
(D) difference spectrum, A–C, highlighting the residues whose side-
chain signals are perturbed by binding to F-actin. Signals of the
hcTnI128–153 are labelled. Complex formation characterized by rel-
atively fast exchange between the free and actin-bound states of the
peptide population is indicated by the resonance broadening that
occurs during titration with increasing concentrations of F-actin. Note

reversed the enhancement seen in the presence of F-actin or
F-actin–tropomyosin in a manner consistent with the
derived affinity of the IP (Table 2). These data confirmed
that the IP formed a 1 : 1 complex with F-actin the affinity
of which is enhanced by tropomyosin.
Interaction of the myosin light chain N-terminal peptide
with actin and reversal by the TnI inhibitory peptide
The effect of the IP on the interaction with F-actin of the
myosin light chain peptides, MLC1 1–37 and MLC1 1–13
was studied in view of the evidence that this region, localized
to the head of the myosin molecule, can bind to the
C-terminal of actin [15,37,38]. The MLC1 peptides were
found to bind to F-actin both in the absence and presence of
tropomyosin with the interaction resulting in the reduction
of the NMR resonance intensity for the majority of the
peptide sidechain signals (Fig. 6). Tropomyosin alone did
not affect the MLC1 peptide spectra nor did it result in the
dissociation of the MLC1 peptides from F-actin. On the
contrary it increased their affinity. The progressive changes
observed with increasing concentrations of F-actin reflected
complex formation in fast exchange and indicated the
Fig. 4. Interaction of the TnI inhibitory peptide with F-actin determined by surface plasmon resonance (25 mm Hepes, pH 7.4, 150 m
M
NaCl) and by
proton magnetic resonance spectroscopy (5 mm sodium phosphate buffer pH 7.2). (A) Sensorgrams showing the kinetics of binding of the human
cardiac TnI128–153 inhibitory peptide to immobilized F-actin at the peptide concentrations indicated. The fit of these data to 1 : 1 complex
formation yielded a dissociation constant of 3 ± 2 l
M
for the F-actin complex (Table 2). (B) The cardiac TnI inhibitory peptide forms a complex
with F-actin whose affinity is enhanced by tropomyosin as shown by the influence of tropomyosin on the change in resonance line width of the

)6
) F-actin–tropomyosin (
M
)6
)
Dansylated inhibitory peptide, hcTnI128-153 28 ± 5 13 ± 3
Inhibitory peptide, hcTnI128–153 3 ± 2
Myosin loop peptide, hcM398–414 32 ± 5 18 ± 5
Ó FEBS 2002 F-actin interactions inhibited by Troponin-I (Eur. J. Biochem. 269) 5093
involvement of the N-terminal residues of MLC1 in actin
binding both in the absence and presence of tropomyosin.
(Fig. 6i).
Addition of hcTnI128–153 at a much lower relative con-
centration than either MLC1 peptide brought about disso-
ciation of the latter from F-actin and F-actin–tropomyosin.
Fig. 5. The TnI inhibitory peptide forms a 1 : 1 complex with F-actin whose affinity is enhanced by tropomyosin as indicated by fluorescence emission
spectra. The experimental conditions were 5 m
M
phosphate buffer, pH 7.2, T ¼ 293K. The relative fluorescence intensity is shown in arbitrary
units. Excitation was at 340 nm and the spectra were recorded from 420 to 600 nm. (A) Fluorescence emission spectra of dansylated TnI inhibitory
peptide complexed with F-actin-tropomyosin. Titration of the dansylated TnI inhibitory peptide with F-actin or F-actin–tropomyosin (molar ratio
of actin : tropomyosin of  7 : 1) resulted in enhancement of the fluorescence emission intensity of the dansyl group. Shown are fluorescence
emission spectra of 10 l
M
dansylated hcTnI128–153 in the presence of increasing concentrations of F-actin–tropomyosin (2, 5, 20 and 60 l
M
actin,
in traces 2–5, respectively). Competition by 2 l
M
unlabelled hcTnI128–153 in the presence of 60 l

presence of 25 l
M
F-actin. (B) As for A and upon addition of 10 l
M
cardiac inhibitory peptide, hcTnI128–153
.
(C) Difference spectrum, B-A,
showing the sidechain groups of MLC1 1–37 whose resonances displayed actin-dependent broadening that is reversed by the presence of the
inhibitory peptide. The increase in signal intensity of the proton NMR spectra of the MLC1 peptide indicates that its interaction with F-actin is
abolished in the presence of the inhibitory peptide.
5094 V. B. Patchell et al.(Eur. J. Biochem. 269) Ó FEBS 2002
This was clearly indicated by the reversal of the actin-
associated spectral changes for resonances unique to the
MLC1 peptide, e.g. the trimethylalanine signal (Fig. 6ii).
Taken together these results suggested that, while tropo-
myosin on its own did not hinder the binding of MLC1 to
actin, the dissociation of the MLC1 1–37 by the IP binding
to F-actin or F-actin–tropomyosin may have resulted from
a conformational change in subdomain 1 of actin rather
than as a consequence of competition for binding at
identical or overlapping sites. The possibility that the IP
produced its effect by inducing a conformational change in
actin was explored further by studying its influence on the
binding of the loop peptide hcM398–414. This region of the
myosin molecule is believed to dock at the junction between
subdomain 1 and 3 of actin [16,39] whereas the LC1 peptide
binds close to the C terminus of actin.
Interaction of the myosin loop peptide, residues
398–414, with F-actin occurs at a region
that does not overlap with the binding site

(Fig. 7B). Comparable quenching effects were observed in
the presence of tropomyosin (1 : 7, tropomyosin : actin)
while the titration data were consistent with 1 : 1 complex
formation as judged by the goodness of fit of the data to a
1 : 1 binding curve. The derived K
d
values were similar to
those obtained by monitoring the actin–tryptophan fluor-
escence changes on the addition of the hcM398–414 peptide
(Table 2).
Titration of IAEDANS-labelled F-actin with the inhi-
bitory peptide, hcTnI128–153 was also carried out in the
absence and presence of tropomyosin. Under both condi-
tions the inhibitory peptide led to enhancement of the
IAEDANS emission ( 16% enhancement at saturation,
Fig. 7B) with a shift of the fluorescence emission maximum
from 475–470 nm. These titration data were consistent with
1 : 1 complex formation and yielded K
d
values similar to
those obtained using unlabelled F-actin (Table 2). The
observations that binding of the TnI inhibitory region led to
fluorescence enhancement and a blue-shift of the emission
maximum are consistent with the IAEDANS label on
Cys374 experiencing a less polar environment upon complex
formation. This contrasts with the change in environment of
the label upon interaction of actin with the hcM398–414
myosin loop peptide. The markedly different response of the
IAEDANS label to the binding of the two peptides provides
direct experimental evidence that the myosin loop and the

of His401, Arg403, Asn408, Tyr410 and Thr412 that
indicated complex formation with F-actin (Fig. 8). Less
notably perturbed is the sidechain signal of Val411. The
nature of the residues affected was unchanged upon interac-
tion with F-actin–tropomyosin while the increased broaden-
ing effects observed at low peptide : actin in the presence of
tropomyosin are consistent with an enhanced affinity result-
ing from a decrease in peptide dissociation kinetics.
The binding of the TnI inhibitory region simultaneously
displaces peptides bound at nonoverlapping sites
on actin
Competition experiments were carried out to monitor the
ability of different peptides to simultaneously bind to
F-actin. The peptide derived from TnI, rcTnI161–181,
bound to F-actin in the presence or absence of tropomyosin
without displacing the myosin loop peptide (Fig. 9i). This
TnI peptide represents a region C-terminal to the IP of TnI
and has been proposed as an additional actin-binding site
[9,10]. Binding of the rcTnI161–181 peptide was judged
from the spectral broadening of its clearly distinguishable
sidechain signals (e.g. His170, c.f. Fig. 1) that occurred
without any concurrent changes in the resonances unique to
the hcM398–414 myosin loop peptide (His401 and Tyr410,
c.f. Fig. 8). Competition from the myosin loop peptide with
peptide rcTnI161–181 for interaction with actin would have
resulted in its displacement and the consequent appearance
of signals broadened as a consequence of interaction with
F-actin. These results indicated that the myosin loop peptide
and rcTnI161–181 are bound simultaneously at different
sites on actin as might be expected from the differences in

concentration of 50 l
M
. The presence of the TnI inhibitory
peptide (0.55 l
M
) led to a decrease in the amount of myosin
peptide bound to F-actin–tropomyosin (with hcM398–414
at  100-fold excess over TnI peptide). This is seen from the
Fig. 8. Proton MR spectral changes upon titration of hcM398–414 with F-actin identifying the residues involved in complex formation. Spectra
determined in 5 m
M
sodium phosphate buffer, pH 7.2, T ¼ 293K. (i) (A) Peptide hcM398–414 (200 l
M
). (B) In the presence F-actin, 28 l
M
.(C)
Difference spectrum, A-B, highlighting the residues whose sidechain signals are perturbed by binding to F-actin. (ii) as in (i) but spectra acquired by
the use of a two-pulse spin-echo sequence. Spectral accumulation in this way distinguishes signals on the basis of their J-coupling patterns and
highlights even small changes in signal linewidth resulting in readily detectable changes in intensity. Signals of hcM398–414 are labelled. Complex
formation characterized by relatively fast exchange between the free and actin-bound states of the peptide population is indicated by the resonance
broadening that occurs during titration with increasing concentrations of F-actin. The unique sidechain resonances of His401, Arg403, Tyr410,
Val411 and Thr412 display marked perturbation upon complex formation.
5096 V. B. Patchell et al.(Eur. J. Biochem. 269) Ó FEBS 2002
change in the myosin peptide signals, for example, Arg403,
Val411 and Thr412 that, as highlighted by difference
spectroscopy (Fig. 9ii), revert towards those of the free
peptide in the presence of hcTnI128–147 at an
actin : Tm : TnI peptide ratio of 7 : 1 : 1. These observa-
tions reinforce the suggestion that conformational changes
which occur when one molecule of troponin I interacts with

C-terminally labelled actin [42] it can be concluded that the
N-terminus of MLC1 binds close to the C-terminus of actin.
The N-terminal region APKK (residues 1–4) of MLC1
appears to be particularly important since modification of
these residues by recombinant DNA technology results in
changes in the kinetics of the actomyosin MgATPase [41].
Other residues at the N-terminus of MLC1 are involved in
binding and have indeed been shown to be important for the
activity of cardiac myosin. A peptide corresponding to
residues 5–14 of human ventricular MLC1 increased the
contractility of intact and skinned human heart fibres [43]
and a similar peptide added to rat cardiac myofibrils
induced a supramaximal increase in the MgATPase activity
at submaximal calcium levels [44].
The NMR and fluorescence studies both indicate inter-
action of F-actin with another region of myosin, the loop
peptide, hcM398–414. The interaction appears to occur at a
region on F-actin that is different from that involved in
binding the TnI inhibitory peptide as shown by the
distinctive response of the IAEDANS probe to each
peptide. This is consistent with the earlier observations that
Fig. 9. (i) The TnI inhibitory region displaces both peptide hcM398–414 and peptide rcTnI161–181 that interact concurrently with F-actin at distinct
binding locations. The aromatic sidechain NMR resonances are shown since these provide unique reporter signals for each of the peptides. Spectra
determined in 5 m
M
phosphate buffer, pH 7.2, T ¼ 293K. (A) Myosin loop peptide (hcM398–414), 108 l
M
. (B) hcM398–414, 108 l
M
in the

Under the conditions of low ionic strength at which these
studies were carried out there was no evidence that
tropomyosin inhibited the binding of either myosin peptide,
indeed the evidence was that their affinity for actin was
increased. This implies that the binding site(s) occupied by
these peptides are different from those involved in binding
tropomyosin.
The results of the competition experiments and the
observations that the affinities of the TnI inhibitory peptide
and the myosin peptides for actin are all very similar suggest
that displacement does not explain our results. A more
likely explanation is that all the peptides have specific
binding sites and that the binding of the IP in a 1 : 1
complex induces a conformational change that is wide-
spread in the actin molecule leading to the dissociation of
the MLC1–37, hcM398–414 and rcTnI161–181 peptides
(Fig. 10). On the other hand any conformational change
induced in actin by binding of the myosin loop peptide or
rcTnI161–181 is more restricted since any allosteric effect
does not extend from the regions of actin where these
peptides are bound.
Both NMR and fluorescence studies indicate that tropo-
myosin enhances the binding of the inhibitory peptide
region to actin presumably by enabling slower dissociation
of the peptide. From the results of early ultracentrifugation
studies [3,45] in which tropomyosin–actin binding was
assayed by co-centrifugation it has been assumed that these
two proteins did not interact at low Mg concentration and
low ionic strength. The fact that the K
d

widespread conformational changes in the actin monomer
occur on interaction with TnI. Inspection of the actin
structure indeed reveals intramonomer contacts within actin
subdomain 1 that may underlie the ability of the inhibitory
region of TnI to influence the surface activity of actin
towards other actin binding proteins. The site of binding of
the inhibitory region is the N-terminal of actin that is
structurally linked to the C-terminal and residues 99–101.
Displacement of myosin contacts would therefore be
facilitated by small conformational changes distributed
through the residue network making intramolecular con-
tacts between the N- and C-terminal regions on subdomain
1ofactin.
To maintain the integrity of the F-actin filament
structure after interaction with TnI with a monomer it
would be expected that the conformational changes would
be transmitted to neighbouring actin monomers. These
could take place through the contact points between actin
monomers in the F-actin filament, of which it has been
postulated that there are at least four per actin monomer
[46]. Biochemical and physiological data suggest that the
response of the actin protein assembly involves longitu-
dinal cooperativity along the thin filament. The regulatory
interactions would be expected to be the same as the
nearest neighbour interactions that govern the actin
monomer contacts upon which thin filament assembly is
based. Several studies have suggested that structural
changes in the actin monomer result from polymerization
to form F-actin [47,48]. These changes may be fundamen-
tal for the longitudinal cooperativity observed in func-

described here provide direct of evidence of sites of
interaction and of conformational changes occurring in
actin that are an important aspect of the regulatory process
(Fig. 10). They also imply that the role of tropomyosin in
filament function may be to stabilize the actin filament and
facilitate its cooperative function rather than directly
blocking the interaction of actin with myosin as postulated
by the steric hypothesis.
ACKNOWLEDGEMENTS
The work described has been supported by grants from the British
Heart Foundation and the Wellcome Trust. The work is part of the
Bioinformatics Initiative at the University of Birmingham supported by
the Medical Research Council.
REFERENCES
1. Schaub, M.C. & Perry, S.V. (1971) The regulatory proteins of the
myofibril. Characterisation and properties of the inhibitory factor
(Troponin B). Biochem. J. 123, 367–377.
2. Perry, S.V., Cole, H.A., Head, J.F. & Wilson, F.J. (1972) Loca-
lisation and mode of action of the inhibitory component of the
troponin complex. Cold Spring Harbor Symp Quant Biol. 37, 251–
262.
3. Eaton, B.L., Kominz, D.R. & Eisenberg, E. (1975) Correlation
between the inhibition of the acto-heavy meromyosin ATPase and
the binding of tropomyosin to F-actin: effects of Mg
2+
,KCl,
troponin I and troponin C. Biochem. 14, 2718–2725.
4. Lehrer, S.S., Chai, M. & Geeves, M.A. (1997) Effect of troponin I
(TnI) on actin S1 ATPase and S1 binding in the absence and
presence of rabbit skeletal tropomyosin. Biophys. J. 72,MP198.

muscle. Cold Spring Harbor Symp Quant Biol. 37, 341–352.
13. Huxley, H.E. (1972) Structural changes in the actin- and myosin-
containing filaments during contraction. Cold Spring Harbor
Symp Quant Biol. 37, 361–378.
14. Parry, D.A.D. & Squire, J.M. (1973) Structural role of tropo-
myosin in muscle regulation: analysis of the x-ray diffraction
patterns from relaxed and contracting muscles. J. Mol. Biol. 75,
35–55.
15. Milligan, R.A. (1996) Protein–protein interactions in the rigor
actomyosin complex. Proc.NatlAcad.Sci.USA93, 21–26.
16. Rayment,I.,Holden,H.M.,Whittaker,M.,Yohn,C.B.,Lorenz,
M., Holmes, K.C. & Milligan, R.A. (1993) Structure of the actin–
myosin complex and its implications for muscle contraction.
Science 261, 58–65.
17. Sasaki, N., Asukagawa, H., Yasuda, R., Hiratsuka, T. & Sutoh,
K. (1999) Deletion of the myopathy loop of Dictyostelium myosin
II and its impact on motor functions. J. Biol. Chem. 274, 37840–
37844.
18. Cuda, G., Fananapazir, L., Zhu, W.S., Sellers, J.R. & Epstein,
N.D. (1993) Skeletal-muscle expression and abnormal function of
beta-myosin in hypertrophic cardiomyopathy. J. Clin. Invest. 91,
2861–2865.
19.Geisterfer-Lowrance,A.A.T.,Kass,S.,Tanigawa,G.,Vosberg,
H.P.,McKenna,W.,Seidman,C.E.&Seidman,J.G.(1990)A
molecular basis for familial hypertrophic cardiomyopathy – a
beta-cardiac myosin heavy-chain gene missense mutation. Cell 62,
999–1006.
20. Cuda, G., Fananapazir, L., Epstein, N.D. & Sellers, J.R. (1997)
The in vitro motility activity of beta-cardiac myosin depends on
the nature of the beta-myosin heavy chain gene mutation in

29. Bartegi, A., Roustan, C., Chavanieu, A., Kassab, R. & Fattoum,
A. (1997) Interaction of F-actin with synthetic peptides spanning
the loop region of human cardiac beta-myosin heavy chain con-
taining Arg403. Eur. J. Biochem. 250, 484–491.
30. Spudich, J.A. & Watt, S. (1971) The regulation of rabbit skeletal
muscle contraction. Biochemical studies of the interaction of the
tropomyosin-troponin complex with actin and the proteolytic
fragments of myosin. J. Biol. Chem. 246, 4866–4487.
31. Miki, M., Dos Remedios, C.G. & Barden, J.A. (1987) Spatial
relationship between the nucleotide-binding site, Lys-61 and Cys-
374 in actin and a conformational change induced by myosin
subfragment-1 binding. Eur. J. Biochem. 168, 339–345.
32. Jardetzky, O. & Roberts, G.C.K. (1981) NMR in Molecular
Biology p.59. Academic Press, New York/London.
33. Huber, P.A.J., Gao, Y., Fraser, I.D.C., Copeland, O., El-Mez-
gueldi,M.,Slatter,D.A.,Keane.N.E.,Marston,S.B.&Levine,
B.A. (1998) Structure-activity studies of the regulatory interaction
of the 10 kilodalton C-terminal fragment of caldesmon with
actin and the effect of mutation of caldesmon residues 691–696.
Biochemistry 37, 2314–2326.
34. Lian, L Y. & Roberts, G.C.K. (1993) Effects of chemical
exchange on NMR spectra. In NMR of Macromolecules (Roberts,
G.C.K., ed.), pp 151–182. IRL Press, Oxford.
35. Jardetzky, O. & Jardetzky, C.D. (1962) Introduction to magnetic
resonance spectroscopy methods and biological applications.
Methods Biochem. Anal. 9, 225.
36. McConnell, H.M. (1958) Reaction rates by nuclear magnetic
resonance. J. Chem. Phys. 28, 430–431.
37. Sutoh, K. (1982) Identification of myosin-binding sites on the
actin sequence. Biochemistry 21, 3654–3661.

46. Holmes, K.C., Popp, D., Gebhard, W. & Kabsch, W. (1990)
Atomic model of the actin filament. Nature 347, 44–49.
47. Steinmetz, M.O., Stoffler, D. & Hoenger, A. (1997) Actin: From
cell biology to atomic detail. J. Struct. Biol. 119, 295–320.
48. Moraczewska, J., Wawro, B., Seguro, K. & Strzelecka-
Golaszewska, H. (1999) Divalent cation, nucleotide and poly-
merization-dependent changes in the conformation of subdomain
2ofactin.Biophys. J. 77, 373–385.
49. Peptides as probes in Muscle Research. (1991) (J.C. Ruegg, ed.),
pp 1–174. Springer-Verlag, Berlin.
50. Bremel, R.D. & Murray, J.M. (1972) and. Weber. Manifestations
of cooperative behaviour in the regulated actin filament during
actin-activated ATP hydrolysis in the presence of calcium. Cold
Spring Harbor Symp Quant. Biol. 37, 267–275.
51. Butters, C.A., Willasden, K.A. & Tobacman, L.S. (1993)
Cooperative interactions between adjacent troponin-tropomyosin
complexes may be transmitted through the actin filament. J. Biol.
Chem. 268, 15565–15570.
52. Levine, B.A., Patchell, V.B. & Perry, S.V. (1999) Troponin I and
conformational changes in actin. J. Muscle Res. Cell Motil. 20,
828–829.
53.Kabsch,W.,Mannherz,H.G.,Suck,D.,Pai,E.F.&Holmes,
K.C. (1990) Atomic structure of the actin – DNase-I complex.
Nature 347, 37–44.
5100 V. B. Patchell et al.(Eur. J. Biochem. 269) Ó FEBS 2002