J. Mol. Biol. (1996) 260, 756–766
Cooperativity in F-Actin: Binding of Gelsolin at the
Barbed End Affects Structure and Dynamics of the
Whole Filament
Ewa Prochniewicz1, Qingnan Zhang1, Paul A. Janmey2 and
David D. Thomas1*
1
Department of Biochemistry
University of Minnesota
Medical School, Minneapolis
MN 55455, USA
2
Division of Experimental
Medicine, Harvard Medical
School, Boston, MA 02115
USA
We have studied the effect of gelsolin, a Ca-dependent actin-binding
protein, on the microsecond rotational dynamics of actin filaments, using
time-resolved phosphorescence (TPA) and absorption anisotropy (TAA) of
erythrosin iodoacetamide attached to Cys374 on actin. Polymerization of
actin in the presence of gelsolin resulted in substantial increases in the rate
and amplitude of anisotropy decay, indicating increased rotational motion.
Analysis indicates that the effect of gelsolin cannot be explained by
increased rates of overall (rigid-body) rotations of shortened filaments, but
reflects changes in intra-filament structure and dynamics. We conclude
that gelsolin induces (1) a 10° change in the orientation of the absorption
dipole of the probe relative to the actin filament, indicating a conformational change in actin, and (2) a threefold decrease in torsional
rigidity of the filament. This result, which is consistent with complementary electron microscopic observations on the same preparations, directly
demonstrates long-range cooperativity in F-actin, where a conformational
change induced by the binding of a single gelsolin molecule to the barbed
end is propagated along inter-monomer bonds throughout the actin
filament.
7 1996 Academic Press Limited
*Corresponding author
Keywords: actin; gelsolin; dynamics; cooperativity; phosphorescence
Introduction
The organization of actin in living cells is
regulated by interaction with a multitude of actinbinding proteins. Gelsolin, a Ca-dependent actinbinding protein, can interact with either monomeric
(G) or filamentous (F) actin. Interaction of gelsolin
with monomeric actin results in formation of a tight
complex with two actin monomers, which acts as a
nucleus of polymerization and produces short
filaments with one gelsolin molecule remaining
bound to the barbed end of each filament (Bryan &
Kurth, 1984; Yin et al., 1981; Janmey et al., 1986). A
similar equilibrium distribution of short filaments
is produced when gelsolin is added to filamentous
actin (Yin et al., 1980; Yin & Stossel, 1980); in this
case gelsolin also remains bound as a stable ‘‘cap’’
at the barbed end of the severed filament. The
regulation of the length of actin filaments by
Abbreviations used: TPA, transient phosphorescence
anisotropy; TAA, transient absorption anisotropy;
ErIA, erythrosin iodoacetamide.
0022–2836/96/300756–11 $18.00/0
gelsolin is thought to be part of the mechanism of
Ca-dependent solation of actin gels in cells (Yin
et al., 1980, 1981).
Gelsolin is composed of six structurally similar
segments (Way et al., 1989; Way & Matsudaira,
1993). The crystal structure of a complex between
gelsolin segment 1 and G-actin shows that segment
1 binds to the cleft between subdomains 1 and 3 of
the actin monomer (McLaughlin et al., 1993).
However, segment 1 itself does not have severing
activity and is not required for nucleation and
polymerization. These functions require the presence of the remaining segments, particularly
segments 2 to 3 for severing and segments 2 to 6 for
nucleating (Way et al., 1989; Way & Matsudaira,
1993). The relationship between the structure of
gelsolin and its actin-binding activities (severing,
capping and nucleating), remains a subject of
intensive study (Sun et al., 1994; Yin et al., 1981;
Ditsch & Wegner, 1994).
The postulated role of gelsolin in regulating the
dynamics of the actin-based cytoskeleton and cell
motility prompted us to investigate the effect of
7 1996 Academic Press Limited
757
Cooperativity in Actin Filaments
gelsolin on structure and internal dynamics of pure
actin filaments. In the present work we have
applied transient phosphorescence anisotropy
(TPA) and transient absorption anisotropy (TAA),
using an actin-bound phosphorescent probe,
erythrosin iodoacetamide (ErIA), to study the
effect of gelsolin on the microsecond time scale rotational dynamics of actin filaments. The theoretical
basis of the present work has been established
(Prochniewicz et al., 1996). We have shown that:
(1) TPA of F-actin cannot be explained by rigidbody rotations (Sawyer et al., 1988), but reflects
primarily intra-filament torsion (Yoshimura et al.,
1984); (2) the analysis of TPA in terms of the theory
developed by Schurr (Allison & Schurr, 1979;
Schurr, 1984) allows determination of the torsional
rigidity of actin filaments; and (3) the orientation of
the absorption dipole of the actin-bound probe with
respect to the actin filament axis can be determined
using the TAA method.
Electron microscopic examination of the samples
used in the present work has already shown that a
single bound gelsolin affects the helical structure of
the whole actin filament (Orlova et al., 1995). We
now present spectroscopic evidence that binding
of gelsolin results in long-range changes in the
structure and dynamics of F-actin.
Results
The interaction of gelsolin with
ErIA-labeled actin
The average filament length of control ErIAlabeled F-actin was Ln = 4.2 mm (Figure 1(a))
(Prochniewicz et al., 1996), while the length of
filaments polymerized in the presence of gelsolin
Figure 2. TPA of actin polymerized in the presence
(+GS) and absence (−GS) of gelsolin. Actin (0.06 mg/ml)
in F-buffer, 25°C. The two decays are superimposed on
the best fits (in the time window from 2.75 to 895 ms) to
the sum of three exponential terms plus a constant
(equation (3)). Residuals (data minus fit) are plotted for
both fits. The fitted correlation times Fi and amplitudes
Ai (i = 1, 2, 3) are F1 = 5.15 ms, F2 = 33 ms, F3 = 240 ms,
A1 = 0.182, A2 = 0.317, A3 = 0.320, Aa = 0.179 (−GS),
and F1 = 4.65 ms, F2 = 20.8 ms, F3 = 94.2 ms, A1 = 0.201,
A2 = 0.394, A3 = 0.295, Aa = 0.108 (+GS).
(referred to as actin–gelsolin) decreased with increasing gelsolin:actin ratio. For example, at a
gelsolin:actin molar ratio of 1:250, Ln decreased to
0.68 mm (Figure 1(b)), very near the length predicted if each gelsolin produces one filament and
all filaments are capped. The length distributions
in the presence and absence of gelsolin (Figure 1)
were not significantly affected by the ErIA label.
Gelsolin had little or no effect on the fraction
of actin polymerized. Sedimentation experiments
(one hour at 230,000 g) showed that 93% and
96% of actin sedimented at gelsolin:actin = 1:180
(Ln = 0.5 mm) and 1:290 (Ln = 0.8 mm), respectively,
as compared with 99% of sedimenting control actin
(Ln = 4.2 mm). The small fraction of unpolymerized
actin was not enough to affect the spectroscopic
data.
The effect of gelsolin on actin TPA
Figure 1. Length distribution of filaments of actin
polymerized in the absence (a) and presence (b) of
gelsolin, measured from electron micrographs.
Gelsolin increased both the rate and amplitude of
time-resolved phosphorescence anisotropy (TPA)
decay of ErIA-labeled F-actin (Figure 2). These
effects did not depend on whether gelsolin was
added before or after actin polymerization. Fitting
each decay to a sum of three exponential terms
plus a constant (equation (3)) confirmed a
significant increase in the total amplitude of
anisotropy decay, as indicated by a decrease in
the normalized residual anisotropy Aa = ra /r0 .
At gelsolin:actin ratios from 1:230 to 1:366, Aa
(0.111(20.005), SEM, n = 21) was less than half of
that obtained for control actin (Aa = 0.236(20.008),
SEM, n = 18). On the other hand, the initial
anisotropy of actin-gelsolin, r0 = 0.126(20.002)
758
Figure 3. TAA of actin–gelsolin at 0.5 mg/ml (other
conditions as in Figure 2), superimposed on the best fit
(in the time window from 250 ns to 891 ms) to the sum of
three exponential terms and a constant (equation (3)),
resulting in F1 = 2.44 ms, F2 = 12.9 ms, F3 = 59.9 ms,
a
A1 = 0.280, A2 = 0.352, A3 = 0.302, Aa = 0.065. Aa = r0a /ra
was used to calculate the angle ua (equation (15)).
(SEM, n = 22), remained the same as the control,
r0 = 0.131(20.006) (SEM, n = 17), indicating that
gelsolin did not affect the angular amplitude of
submicrosecond wobbling motions of the dye
and/or actin (equations (4), (5); Prochniewicz et al.,
1996). Further model-based analysis of the effect
of gelsolin was performed in the range of
gelsolin:actin ratios from 1:230 to 1:366, corresponding to the range of filament lengths from
Ln = 0.63 mm to Ln = 1 mm.
The effect of gelsolin on actin
filament structure
The transient absorption anisotropy (TAA) decay
for actin–gelsolin (Figure 3) was best fit to a sum
of three exponentials plus a constant, equation
(3), and the fitted values of initial (r0a ) and final
a
(ra
) anisotropies of TAA were used to calculate
the orientation of the absorption dipole (ua in Figure 8) of actin-bound erythrosin, using equation
(15), as described previously for control actin
(Prochniewicz et al., 1996). Polymerization of actin
in the presence of gelsolin significantly changed ua :
in the range of gelsolin:actin ratio from 1:245 to
1:400, ua = 41.7(20.5)° (mean 2 SEM, n = 29),
which is significantly higher than ua = 30.2(20.7)°
(mean 2 SEM, n = 14) obtained in parallel experiments on actin polymerized in the absence of
gelsolin (Prochniewicz et al., 1996).
Rigid-body rotation is insufficient to explain
dynamics of actin–gelsolin
When we attempted to fit the TPA decay for
actin–gelsolin (Figure 2, +GS) to the rigid-body
rotation model (equation (12)), the fit (not shown)
was very poor (x2 = 49), indicating that much more
Cooperativity in Actin Filaments
Figure 4. TAA of actin (−GS) and actin–gelsolin (+GS)
(conditions as in Figure 3) in the time window from
0.016 ms to 60 ms, and best fits of these decays to the
rigid-body rotation model (equation (12)), using the
length distributions from Figure 1; ua was allowed to
vary within the range 30.2(22.5)° (−GS) and 41.7(22.2)°
(+GS) and d = 34.5°.
rotational motion is occurring than would be
expected for rigid actin filaments. The inadequacy
of the rigid-body rotation model was further
demonstrated by transient absorption anisotropy
(TAA). In this method the time window of the
experiment is shifted from 3 to 890 ms (phosphorescence) to a shorter range, 0.016 to 60 ms, allowing
us to focus on the initial period of the microsecond
time-scale decay; furthermore, the number of
adjustable parameters in the fit decreases from two
(ue , and k) to one (k), (equation (16), Figure 8). The
results of the fit (Figure 4) show that the dynamics
of actin–gelsolin in the microsecond time range has
components that cannot be explained by rigid body
rotation, just as it could not explain the microsecond dynamics of control F-actin (Prochniewicz
et al., 1996). Therefore, we must analyze the TPA
and TAA decays in terms of rotational dynamics
within the actin filament.
The effect of gelsolin on internal dynamics of
the actin filament
The effect of gelsolin on the torsion constant a of
actin filaments was analyzed by fitting our TPA
data to the theory of Schurr (Allison & Schurr, 1979;
Schurr, 1984; Prochniewicz et al., 1996; equation
(10)). This model includes both rigid-body rotations
and torsional flexing about the actin filament axis.
The value of ua was constrained within the range of
values consistent with the TAA results. The high
quality of the fit, with residuals no higher than 6%
of the experimental values (Figure 5), shows that
this model adequately describes the microsecond
time scale rotational dynamics of actin–gelsolin
filaments, as found in the absence of gelsolin
(Prochniewicz et al., 1996). The results (Table 1)
show that gelsolin not only changes the structure of
the actin filament, as indicated by the 10° change in
dye orientation shown above by TAA, but also
759
Cooperativity in Actin Filaments
Figure 5. Fit of actin–gelsolin TPA (data from Figure 2,
+GS) to the theory of Schurr (equation (10), which
includes torsional flexing as well as rigid-body rotation)
superimposed on the experimental decay (r(t)), using the
length distribution from Figure 1(b). Residuals = r(t) − fit.
d was fixed at 34.5°, and ua was allowed to vary within
the range 41.7(22.4)° (determined from TAA, Figure 3).
The two other parameters, k and ue were allowed to vary,
and the angle Cea was calculated from equation (16). The
result of this particular fit is: a = 0.17E-11 dyn cm,
ua = 39.4°, ue = 49.5°, and k = 0.640. Average fit results
from eight experiments are given in Table 1.
increases the internal flexibility of the filament, as
indicated by a threefold decrease in the torsion
constant a.
To confirm the uniqueness of this interpretation,
we attempted to fit the actin–gelsolin TPA data to
equation (10), assuming that gelsolin’s only effect is
on filament length, with no other effect on the
structure and dynamics of actin. The filament
length distribution was fixed at the values
determined for actin–gelsolin from electron micrographs (Figure 1(b)), and the parameters ua , ue , and
a were fixed at the values for control actin in
Table 1. This produced a very bad fit to the TPA
data for actin–gelsolin: residuals (Figure 6(a)) were
as large as 66% of the experimental values of
anisotropy. The fit was even worse (not shown) if
we considered rigid-body motions, neglecting
internal torsional dynamics. Thus, although the
shorter filament lengths do predict an increase in
both the rigid-body and internal rotations of the
actin filaments, this was much less than the
observed increase in the rate and amplitude of
Figure 6. Residuals (r(t) − fit) from fitting the TPA of
actin–gelsolin to equation (10), using the measured
filament length distribution from Figure 1(b)
(Ln = 0.63 mm). (a) All parameters other than filament
length were fixed within the ranges of the values
(mean 2 SEM) determined for control actin (Figure 2;
Table 1), corresponding to the assumption that gelsolin
has no effect on the internal structure or dynamics of the
actin filament. (b) Parameters were fixed as in (a), except
that the torsion constant a was allowed to vary,
corresponding to the assumption that gelsolin affects
actin’s flexibility but not its internal structure.
(c) Parameters were fixed as in (a), except that transition
dipole angles ua and ue were allowed to vary,
corresponding to the assumption that gelsolin affects
actin’s internal structure but not its flexibility.
rotational motions induced by gelsolin. Large
residuals remained when the torsion constant a was
an adjustable parameter in the fit, while transition
dipole angles ua and ue were still fixed at the values
for control actin (Figure 6(b)). When ua and ue were
allowed to vary, but a remained fixed at the value
for control actin, residuals were still large, up to
20% of the value of anisotropy (Figure 6(c)). We
conclude that the change in the microsecond
dynamics of actin induced by gelsolin must reflect
long-range changes in both structure (ua ) and
internal dynamics (a) of the actin filament. The
results of fitting the TPA of actin–gelsolin to
equation (10), assuming variable height h and radii
a of the ‘‘elementary rod’’, show that, although the
changes in h affected a, the product a*h (defined as
the torsional rigidity C) of the filament remained
Table 1. The effect of gelsolin on the structure and dynamics of actin, from TPA
Sample
Actin–gelsolin
Control actin
ua (deg)
ue (deg)
a × 1012 (dyn cm)
k
39.4 2 0.0
29.1 2 0.4
47.5 2 0.5
47.3 2 0.4
1.3 2 0.2
3.5 2 0.4
0.65 2 0.01
0.60 2 0.02
The results of fitting TPA to equation (10), presented as mean 2 SEM. For actin–gelsolin (example
of data and fit shown in Figure 5), the actin:gelsolin ratio ranged from 230 to 340, corresponding to
Ln = 0.63 to 0.95 mm, and the length distribution was determined by electron microscopy separately
for each sample (n = 8) as in the example shown in Figure 1(b). For control actin (data shown in Figure 2, −GS), the average length distribution of Figure 1(a) was used, corresponding to Ln = 4.2 mm
(n = 7) (Prochniewicz et al., 1996). ua and ue are the orientations of the transition dipoles of actin-bound
erythrosin (Figure 8), a is the torsion constant, and k is the amplitude-reduction factor corresponding
to submicrosecond motions (equations (4) and (5)).
760
Cooperativity in Actin Filaments
Discussion
Figure 7. TPA of actin–gelsolin, before and after
crosslinking with 0.4% glutaraldehyde. Conditions of
TPA measurement as in Figure 2. Filament length:
actin–gelsolin Ln = 1.01 mm, crosslinked actin–gelsolin
Ln = 1.05 mm.
constant (fits not shown). Thus, the effect of
gelsolin on F-actin can be interpreted in terms of
changes in the continuum flexibility of the filaments
(Prochniewicz et al., 1996).
The effect of crosslinking with glutaraldehyde
Further evidence that gelsolin affects the internal
dynamics of actin filaments was obtained by using
glutaradehyde crosslinking (Lehrer, 1972; Prochniewicz & Yanagida, 1990) to restrict the internal
motions, without changing the length distribution
(Figure 7). Crosslinking actin–gelsolin with glutaraldehyde greatly decreased the rate and extent of
TPA decay, indicating decreased rotational motion,
without any change in the filament length (Figure 7). The rotational dynamics of crosslinked
actin–gelsolin are similar to those of control actin
(Figure 2). This decreased motion could not be due
to inter-filament cross-links, since the diffusive
motions of crosslinked and control actin–gelsolin
filaments measured by dynamic light scattering (Lo
et al., 1994) were essentially the same (data not
shown). Thus, the crosslinking results support our
conclusion that gelsolin increases the internal
rotational dynamics of actin filaments, independently of the effect on filament length.
The effect of inter-filament interaction on TPA
The TPA decay of actin–gelsolin was independent of protein concentration (data not shown) over
a range from 0.06 mg/ml, where filaments are
essentially non-interacting, to 2 mg/ml, where
there is sufficient filament interaction to constrain
translational diffusion (Janmey et al., 1986). The
same result was obtained in the absence of gelsolin,
where filaments are sufficiently long to interact
already at a concentration as low as 0.06 mg/ml.
These results strongly support our conclusion that
the observed TPA decay is dominated by rotational
dynamics within the actin filament.
The effect of gelsolin on TPA and TAA of F-actin
cannot be explained simply by the shortening of
actin filaments (Figures 4 and 7). There must also
be changes in the internal structure and dynamics
of actin (Figures 5 and 6; Table 1). The 10° change
in the orientation of the absorption dipole (ua ) of
actin-bound erythrosin and the threefold decrease
in the torsion constant a induced by binding of
one gelsolin molecule at the barbed end of actin
filament strongly indicate long-range cooperativity
of gelsolin-induced structural changes in F-actin.
The change in ua was first concluded from TAA
measurement (Figure 3) and then confirmed by
fitting TPA to the theory of Schurr (equation (10))
(Figures 5 and 6; Table 1). A critical assumption in
the Schurr model, which permits the determination
of ua and the torsion constant a, is that the only
detectable motions in the time window of interest
(1 ms to 1 ms) are rigid-body and torsional flexing
motions about the actin filament axis. This
assumption is justified by the uniaxial character of
motions of actin–gelsolin as in our previous work
on control actin (Prochniewicz et al., 1996): (1)
simulations showed that end-over-end tumbling,
perpendicular to the actin filament long axis, is out
of the time scale of the phosphorescence measurements; and (2) gelsolin did not affect r0 , so it did not
affect the amplitude of nanosecond wobbling
motions (equations (4) and (5)), which remained
out of the time scale of the TAA and TPA
experiments.
The effect of gelsolin on the process of
polymerization of actin
The structural effect of gelsolin could be due to
growth of the filaments on structurally altered
nuclei: two monomers directly bound to gelsolin
seem to be oriented in a different way than the
monomers in spontaneously nucleated F-actin (Doi,
1992; Hesterkamp et al., 1993). The effect of nuclei
on the structure of actin polymer has been
previously considered as one of the explanations
of the difference between the structure of actin
polymerized in the presence of Ca2+ and Mg2+
(Orlova & Egelman, 1995; Orlova et al., 1995). On
the other hand, proposed fragmentation of the
polymerizing actin by free gelsolin (Ditsch &
Wegner, 1994) suggests that the effect of gelsolin
could be due to propagation of structural changes
induced by binding to and capping the barbed end.
The present results could be explained by either of
these hypotheses, nucleation or capping, since
gelsolin-induced changes in TPA (Figure 2) were
independent of whether gelsolin was added to
monomeric or polymerized actin.
The effect of gelsolin on the orientation of the
C terminus
Since erythrosin is rigidly bound to Cys374
(Prochniewicz et al., 1996), the change in the
761
Cooperativity in Actin Filaments
orientation of its absorption dipole (ua ) probably
reflects the change in the orientation of Cys374 itself
and/or a larger region in the C terminus. Such a
possibility is supported by the previously reported
effect of gelsolin on the distance between Cys374
of two neighboring monomers (Hesterkamp et al.,
1993) and by the result of electron microscopic
reconstruction (Orlova et al., 1995). A homogeneous
increase of density in the bridge between two
strands of the actin–gelsolin helix has been
correlated with a change in the orientation of the C
terminus, since (1) this bridge is involved in the
connection between the C terminus and subdomain
4 of the protomers in the opposite strands (Milligan
et al., 1990; Holmes et al., 1990), and (2) the changes
in its density accompany the changes in the
resistance of two C-terminal residues to trypsin
cleavage (Orlova & Egelman, 1995). The orientation
of the actin-bound erythrosin could be directly
affected by the binding of gelsolin segment 1 to the
lower part of subdomain 1 of the actin monomer
(Boyer et al., 1987; McLaughlin et al., 1993), and/or
by the propagation of structural changes induced
by gelsolin in other regions of actin protomer: the
nucleotide-binding cleft (Tellam, 1986; Harris, 1985,
1988; Bryan & Kurth, 1984) and the phalloidin
binding site (Allen & Janmey, 1994). The possibility
of an allosteric effect is supported by the recently
demonstrated structural connectivity between the
C terminus and the nucleotide binding cleft
(Crosbie et al., 1994).
Gelsolin-induced local structural changes
affect flexibility of the whole filament
The change in the torsion constant a, which
characterizes continuum flexibility of F-actin
(Prochniewicz et al., 1996), indicates that the effect
of gelsolin should not be regarded as an effect on
motions of one structural element in actin, but
rather as an effect on a range of intra-filament
motions. Normal mode analysis of the currently
available models of the actin filament led to the
conclusion that the torsional flexibility of the whole
filament could be significantly affected by reorientation of only a few residues, such as in the region
of the hydrophobic plug and subdomain 2 (Holmes
et al., 1990; Lorenz et al., 1993; Tirion et al., 1995;
ben-Avraham & Tirion, 1995). The possibility of a
relationship between intra-monomer structural
changes and filament flexibility is further supported by thermodynamical considerations, which
located the source of actin flexibility in intramonomer ‘‘hinges’’ (Erickson, 1989); such hinges
were later predicted by the normal mode analysis
of G-actin (Tirion & ben-Avraham, 1993) and
considered as one of the explanations of motions
revealed by time-resolved intra-monomer energy
transfer (Miki & Kouyama, 1994). On the other
hand, electron microscopic reconstructions of
filaments suggested an explanation of actin
flexibility not only in terms of mobility of the
intra-monomer hinges but also in terms of lateral
slipping of two strands of the helix and/or random
dislocations of the whole monomers, which implies
that there are mobile hinges on the intermonomer
interface (Egelman et al., 1982; Bremer et al., 1991;
Egelman & Orlova, 1995). Both approaches to the
molecular origin of actin flexibility could explain
the effect of gelsolin, since recent biochemical
studies revealed significant interconnectivity of
structural changes in actin. The C terminus, which
is probably affected by gelsolin, participates in
formation of the intermonomer bonds and its
modifications can affect the kinetics of polymerization and the shape of the filaments (O’Donoghue
et al., 1992; Drewes & Faulstich, 1993; Mossakowska
et al., 1993). Furthermore, there is a possibility of an
allosteric effect of the changes in the C terminus on
subdomain 2, which plays an important role in
intermonomer interactions and dynamics of actin
(Holmes et al., 1990; Khaitlina et al., 1993; Crosbie
et al., 1994; Orlova & Egelman, 1995). More direct
indication of gelsolin-induced changes in the
inter-monomer interactions is provided by the
capping effects, which seem to strengthen the
bonds between the capped monomers in the barbed
end (Weber et al., 1991) and also cause substantial
decrease in the rate of dissociation of monomers
from the pointed end (Janmey & Stossel, 1986).
Structural transitions in actin
The gelsolin-induced changes in F-actin reported
here represent a new case of long-range cooperative
transitions in the filament, with one site of
structural perturbation, the barbed end. Functional
similarities among gelsolin, the cytochalasins,
b-actinin, villin, fragmin/severin, scinderin,
Acanthamoeba capping protein, capZ and capG, as
well as structural similarities among many of these
proteins (Yin et al., 1981; Weeds & Maciver, 1993)
suggest that long-range cooperative structural
transitions in actin structure may be involved in the
regulation of the dynamics of actin-based cytoskeleton by many actin-binding proteins (Way et al.,
1989). A possibility of such transitions has already
been suggested as an explanation of the significantly reduced rate of elongation at the barbed end
by binding of b-actinin to the pointed end, and of
the reduced rate of elongation at the pointed end by
blocking the barbed end with cytochalasin D
(Maruyama et al., 1984).
Cooperativity of actin structure was originally
predicted from the theoretical analysis of the
mechanism of polymerization (Oosawa & Kasai,
1971; Wegner & Engel, 1975). Experimental
evidence for cooperativity has been provided by
non-linearity of changes in stability, spectroscopic
signal or functional properties of actin induced by
factors randomly bound along actin: phalloidin or
BeFx (Drewes & Faulstich, 1993; Muhlrad et al.,
1994), myosin heads (Thomas et al., 1979; Miki et al.,
1982; Mossakowska et al., 1988; Schwyter et al.,
1990; Prochniewicz et al., 1993), the regulatory
proteins tropomyosin and troponin (Butters et al.,
762
1993; Geeves & Lehrer, 1994) or specific antibodies
(DasGupta & Reisler, 1992). Determination of the
crystal structure of actin (Holmes et al., 1990;
Lorenz et al., 1993) opened up the possibility of
studying the molecular basis of structural transitions in the actin filament. The first such approach
was the normal modes analysis of the currently
available models of F-actin filaments assuming that
the monomers are rigid spheres (ben-Avraham &
Tirion, 1995): while the models differed in the
numbers of longitudinal and diagonal bonds, their
combined Gibbs contact energy of bonds was found
to be essentially the same. More complex future
calculations, including internal flexibility of
monomers, will further promote understanding of
the mechanisms of structural transitions in actin.
Conclusion
Analysis of the effect of gelsolin on the
microsecond time-scale phosphorescence anisotropy decay of actin shows that binding of one
gelsolin to the barbed end of the actin filament
affects the orientation of the Cys374-bound dye on
all protomers and increases about threefold the
torsional flexibility of the whole filament. The most
plausible mechanism is that binding of gelsolin
induces structural changes in the directly bound
protomers, and these changes are subsequently
propagated along the whole actin filament.
Materials and Methods
Preparation of actin and labeling with erythrosin
iodoacetamide (ErIA)
Purification and labeling of actin with the phosphorescent dye erythrosin iodoacetamide (ErIA) was performed
as described (Prochniewicz et al., 1996). Briefly, actin was
purified by one cycle of polymerization–depolymerization, using 30 mM KCl for polymerization. Freshly
prepared actin was polymerized at 47 mM in 100 mM
KCl, 20 mM bicarbonate (pH 8.0), 0.3 mM ATP and
0.2 mM CaCl2 and labeled with 420 mM ErIA for three
hours at 25°C. Excess dye was removed by dialysis,
ultracentrifugation (one hour at 200,000 g) and chromatography on Sephadex G-25. Specificity of labeling at
Cys374 was confirmed using N-ethylmaleimide-blocked
actin (Prochniewicz et al., 1996). The extent of labeling
was 0.83(20.09; mean 2 SD, n = 14). Labeled actin was
polymerized with 50 mM KCl, ultracentrifuged and
depolymerized in G-buffer (5 mM Tris (pH 8.0), 0.2 mM
ATP and 0.1 mM CaCl2 ). Freshly prepared actin was used
for gelsolin-binding studies. Before any physical or
functional measurement, each actin sample was stabilized by a 1:1 molar ratio of phalloidin to actin monomers,
to ensure that that fraction of monomeric actin was
negligible even at very low protein concentrations.
Human plasma gelsolin was purified by elution from
a DE-52 ion exchange matrix in 30 mM NaCl, 3 mM
CaCl2 , 25 mM Tris-HCl (pH 7.4) as described by
Kurokawa et al. (1990), rapidly frozen in liquid nitrogen
and stored at −80°C. Some preliminary work used
recombinant human plasma gelsolin, a generous gift
from Blake Pepinsky, Biogen, Inc, Cambridge MA. The
gelsolin concentration was determined from the ab-
Cooperativity in Actin Filaments
sorbance at 280 nm using the extinction coefficient
1.8 ml mg−1 cm−1.
Polymerization of actin in the presence of gelsolin
Monomeric ErIA actin was diluted to 0.5 mg/ml in
G-buffer, and gelsolin was added at molar ratios of 1:25
to 1:200. The samples were polymerized by addition of
KCl to 50 mM, followed by 2.53 to 3 hours of incubation
at 25°C (Yin et al., 1981), stabilized by the addition of a
1:1 molar ratio of phalloidin, and buffer conditions were
adjusted to F-buffer (50 mM KCl, 20 mM Tris (pH 8.0),
0.2 mM ATP, 0.1 mM CaCl2 ). Polymerized actin was
stored on ice overnight and used for experiments within
two days. The ratio of bound gelsolin to actin was
estimated from the average length of the filaments,
assuming that binding of one gelsolin molecule to the
barbed end of a filament n mm long corresponds to one
molecule of gelsolin per 366*n monomers (Yin et al., 1981;
Janmey et al., 1986). ErIA actin polymerized in the
presence as well as in the absence of gelsolin was motile
in the in vitro motility assay.
Glutaraldehyde crosslinking
Crosslinking was performed as described (Prochniewicz & Yanagida, 1990). Polymerized actin and
actin–gelsolin in F'-buffer (50 mM KCl, 20 mM Hepes
(pH 7.0), 0.2 mM ATP, 0.1 mM CaCl2 ) (replacement of
Tris with Hepes prevented side reactions between the
amino groups of Tris and glutaraldehyde) was diluted to
0.4 mg/ml in the same F'-buffer, glutaraldehyde was
added to 0.4%, and samples were incubated for
30 minutes at 25°C. The reaction was stopped by addition
of 20 mM Tris (pH 8.0), and the sample was dialyzed
overnight against F-buffer and clarified by ten minutes of
centrifugation at 21,000 g.
Electron microscopy
Electron microscopy was carried out as described
(Prochniewicz et al., 1996). Samples of phalloidin-stabilized, phosphorescent actin-gelsolin were diluted to
0.02 mg/ml in F-buffer, applied to glow-discharged,
collodion- and carbon-film-coated copper grids, stained
with 1% uranyl-acetate, and observed with a JEOL 100
CX electron microscope at an accelerating voltage of
80 kV. Photographs were taken at 15,000 × magnification,
and negatives were printed on 8 inch × 10 inch Ilford
photographic paper. The distribution of filament lengths
was determined by tracing the prints with a digitizing
tablet.
Spectroscopic experiments
Phalloidin-stabilized actin samples in F-buffer were
diluted in the same buffer to 0.06 mg/ml for transient
phosphorescence anisotropy (TPA) and to 0.5 mg/ml
for transient absorption anisotropy (TAA), as described
(Prochniewicz et al., 1996). All spectroscopic measurements were performed in F-buffer at 23 to 25°C in the
presence of oxygen-removing enzymes (Eads & Thomas,
1984), glucose oxidase (220 mg/ml) and catalase (36 mg/
ml), and glucose (45 mg/ml).
Microsecond rotational motions of phosphorescentlabeled actin were detected essentially as described
for actin by both TPA and TAA (Ludescher & Thomas,
1988; Ludescher et al., 1987; Prochniewicz et al., 1995). In
763
Cooperativity in Actin Filaments
TPA experiments, the time-resolved phosphorescence
anisotropy was calculated from:
r(t) =
Ivv − GIvh
Ivv + 2GIvh
(1)
where Ivv (t) and Ivh (t) are the vertically and horizontally
polarized components of the emission signal. G is an
instrumental correction factor, determined by performing
the experiment with a solution of free dye under
experimental conditions and adjusting G to give an
anisotropy value of zero, the theoretical value for a
rapidly and freely tumbling chromophore. In TAA
experiments, the change in absorbance D A(t) is defined
as log(I0 /I(t)), where I(t) and I0 are the intensities of
the transmitted light with and without an excitation
pulse. The absorption anisotropy rA (t) was calculated
from:
rA (t) =
DAvv − DAvh
DAvv + 2DAvh
(2)
where DAvv and DAvh correspond to absorption with the
polarizer oriented vertically and horizontally.
Figure 8. The orientation of the transition dipoles of the
probe relative to the local coordinate system fixed on
an ‘‘elementary rod’’. ma and me are the absorption and
emission transition dipoles, respectively. ua and ue define
the orientation of these two dipoles relative to the axis of
rotation Z (the actin filament axis); d is the angle between
the two dipoles.
Anisotropy data analysis
Model-independent fit: sum of exponentials
Model-independent parameters of anisotropy decay:
initial anisotropy r0 = r(t = 0), rotational correlation times
fi , amplitudes Ai = ri /r0 and the final (or residual)
anisotropy ra were determined by fitting of anisotropy to
a sum of exponential terms and a constant:
n
r(t) = s ri e−t/Fi + ra
(3)
i=1
Optimal fits, based on x2 values and plots of the residuals,
were consistently obtained for three exponential terms
(n = 3), as reported for ErIA-labeled actin (Prochniewicz
et al., 1996).
The observed initial anisotropy, as fitted to equation
(3), is lower than the theoretical maximum value r00 by an
amplitude reduction factor kE1, due to motions on the
time-scale too fast to be detected (Prochniewicz et al.,
1996):
r(t = 0) = r0 = kr00 = 0.4kP2 (cos d)
k = r0 /r00 =
$
1
cos uc (1 + cos uc )
2
%
(4)
2
(5)
where d is the angle between the excited (absorption) and
observed (emission) transition dipoles of the probe,
P2 (cos d) = (3 cos2 d − 1)/2, and uc is the radius of a
cone describing the submicrosecond wobble of the probe.
The angle d = 34.8° was determined directly from the
measured anisotropy of ErIA immobilized by incorporation into a solid block of polymethylmethacrylate, as
described by Prochniewicz et al. (1996).
flexible filament with mean local cylindrical symmetry
(Allison & Schurr, 1979; Schurr, 1984) as applied to TPA
of ErIA–F-actin (Prochniewicz et al., 1996). According
to this model, the filament is regarded as a randomly
labeled linear array of cylindrical elementary rods
with a local coordinate system embedded in each rod;
the orientation of absorption (ma ) and emission (me )
dipoles of the rod-bound probe in the local coordinate
system is shown in Figure 8. The anisotropy r(t) describes
the mean-squared displacements of rods (and the
bound dye) about body-fixed x, y and z axes due to
combined intra-filament twisting motions and motions
of the whole filament (rigid body rotations) and its
general form is (as adapted from Allison & Schurr,
1979; Schurr, 1984):
2
r(t) = k s An Fn (t)Cn (t)
(6)
n=0
where k is the amplitude reduction factor (equations (4)
and (5)), An are amplitudes, Cn (t) is the torsional
correlation function, and Fn (t) is the tumbling correlation
function. The amplitudes An are defined by:
A0 = 101 (3 cos2 ua − 1)(3 cos2 ue − 1)
(7)
A1 = 65 sin ua cos ua cos cea sin ue cos ue
(8)
A2 = 103 sin2 ua cos2 cea sin2 ue
(9)
If rotational motions are uniaxial, i.e. filaments do not
undergo tumbling or bending within the time window of
detection, Fn (t) = 1, and the filaments have a broad length
distribution, the anisotropy r(t) becomes sum of
contributions from filaments within each particular
length group li :
Rotational diffusion model: segmented flexible cylinder
n=2 i=p
Anisotropy decays were analyzed in terms of the
theory of Schurr, describing the rotational diffusion of a
r(t) = k s s An Cni (t)
n=0 i=1
(10)
764
Cooperativity in Actin Filaments
where p is the number of length groups in the histogram
(Figure 1). The twisting correlation function for filaments
in the length group li , Cni (t), is defined as:
exp
Cni (t) =
$
−n 2kB Tt
(Ni + 1)g
Ni + 1
Ni + 1
%
$
Ni + 1
× s exp − n 2 s d Q (1 − e−t/tsi )
m=1
2
msi
s=2
tsi =
ga
(s − 1)p
sin2
4
2(Ni + 1)
dsi2 =
kB Ttsi
g
Qmsi =
2
si
X$ X %
a
ua = arccos 2 12 raa
3 2
r0
%
+
cos d = cos ue cos ua + sin ue cos cea sin ua
1
d
(Ni + 1) s1
m = 1, 2, . . . , Ni + 1
(11)
where Ni is the number of rods in the filament of length
li , a is the torsion constant and g is the friction factor for
rotation of a rod about its z axis. kB is Boltzmann constant,
T is temperature. For a cylindrical rod g = 4pha2h where
h and a are its height and radius, respectively, and h is
the solvent viscosity.
For completely rigid filaments of length li , a = a and
anisotropy due to rigid-body rotation is expressed as:
n
ki li
r(t) = s ri (t)bi , bi =
i=1
n
(15)
a
are the initial and final values of the TAA,
where r0a and ra
obtained by fitting the decay to the sum of three
exponential terms, equation (3). The TAA decay depends
on the orientation of only one transition dipole
(absorption), whereas the TPA decay depends on the
orientation of both the absorption and emission dipoles.
Since the angles determining the orientation of the two
transition dipoles of the probe (Figure 8) are constrained
by:
(m − 12 )(s − 1)p
2
cos
(1 − ds1 )
(Ni + 1)
(Ni + 1)
X
X
independently determining the length distribution of
actin filaments, using electron microscopy, and the
orientation of the absorption dipole of actin-bound ErIA
(ua , Figure 8), using transient absorption anisotropy
(TAA). The angle ua was calculated using the formula of
Kinosita et al. (1977):
(12)
s ki li
(16)
the fitting of TPA data was performed using two
adjustable parameters (k and ue ) in the rigid body
rotation model (equation (12)), and three adjustable
parameters (k, ue , and a) in the theory of Schurr (equation
(10)). In the case of the TAA fit to the rigid-body rotation
model, the number of adjustable parameters decreased to
one (k).
Reagents
The phosphorescent dye ErIA was purchased from
Molecular Probes (Eugene, OR) and stored at −20°C.
ATP, EM grade 25% glutaraldehyde solution in ampules,
and phalloidin, were obtained from Sigma (St Louis,
MO). All other chemicals were of reagent grade.
i=1
where ki is the number of filaments with length li ,
ri (t) = k[A0 exp(−t6Di_ ) + A1 exp(−t(Di> + 5Di_ ))
+ A2 exp(−t(4Di> + 2Di_ ))]
0
l
3kB T ln i − si
a
Di_ =
hpl3i
1
$
l
, si = 1.57 − 7 1/ln i − 0.28
a
Di> =
kB T
4pha2li
(13)
%
2
(14)
where a is the radius of the rod, h is the solvent viscosity,
and Di> and Di_ are the diffusion coefficients around long
z and short x axes, respectively.
Methods of simulation and fitting
The anisotropy decay was fitted to the indicated
expressions (equations (3), (10) and (12)) as described
(Prochniewicz et al., 1996). In most cases, this was done
using a microcomputer, but in the case of equation (10),
this was a lengthy calculation and was carried out with
a Cray-YMP C90 supercomputer (Cray Research, Eagan,
MN). A bounded modified Levenberg–Marquardt algorithm using a finite-difference Jacobian (IMSL function
BCLSF, IMSL, Inc.) was used to solve the non-linear least
squares problem.
The number of adjustable parameters in all specific
forms of equation (6) used here was reduced by
Acknowledgements
We thank Drs E. Egelman and A. Orlova for technical
advice and use of a digitizing tablet, M. Witzman for
writing the Cray program implementing the Schurr
theory, and R. L. H. Bennet and N. Cornea for excellent
technical assistance.
This work was supported by grant to D.D.T. from the
National Institutes of Health (AR 32961) and the
Minnesota Supercomputer Institute, to E.P. from the
American Heart Association, Minnesota Affiliate, and to
P.J. from the National Institutes of Health (AR 38910).
References
Allen, P. G. & Janmey, P. A. (1994). Gelsolin displaces
phalloidin from actin filaments. A new fluorescence
method shows that both Ca and Mg affect the rate
at which gelsolin severs F-actin. J. Biol. Chem. 269,
32916–32923.
Allison, S. A. & Schurr, M. (1979). Torsion dynamics and
depolarization of fluorescence of linear macromolecules. I. Theory and application to DNA. Chem.
Phys. 41, 35–59.
ben-Avraham, D. & Tirion, M. M. (1995). Dynamic and
elastic properties of F-actin: a normal-modes
analysis. Biophys. J. 68, 1231–1245.
Boyer, M., Feinberg, J., Hue, H. K., Capony, J. P.,
Benyamin, Y. & Roustan, C. (1987). Antigenic probes
Cooperativity in Actin Filaments
locate a serum–gelsolin-interaction site on the
C-terminal part of actin. Biochem. J. 248, 359–364.
Bremer, A., Millonig, R. C., Sutterlin, R., Engel, A.,
Pollard, T. D. & Aebi, U. (1991). The structural
basis for the intrinsic disorder of the actin filament:
the ‘‘lateral slipping’’ model. J. Cell. Biol. 115,
689–703.
Bryan, J. & Kurth, M. C. (1984). Actin–gelsolin
interactions. Evidence for two actin-binding sites.
J. Biol. Chem. 259, 7480–7487.
Butters, C. A., Willadsen, 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.
Crosbie, R. H., Miller, C., Cheung, P., Goodnight, T.,
Muhlrad, A. & Reisler, E. (1994). Structural
connectivity in actin: effect of C-terminal modifications on the properties of actin. Biophys. J. 67,
1957–1964.
DasGupta, G. & Reisler, E. (1992). Actomyosin interactions in the presence of ATP and the N-terminal
segment of actin. Biochemistry, 31, 1836–1841.
Ditsch, A. & Wegner, A. (1994). Nucleation of actin
polymerization by gelsolin. Eur. J. Biochem. 224,
223–227.
Doi, Y. (1992). Interaction of gelsolin with covalently
cross-linked actin dimer. Biochemistry, 31, 10061–
10069.
Drewes, G. & Faulstich, H. (1993). Cooperative effect on
filament stability in actin modified at the C-terminus
by substitution or truncation. Eur. J. Biochem. 212,
247–253.
Eads, T. & Thomas, D. D. (1984). Microsecond rotational
motions of eosin-labeled myosin measured by
time-resolved anisotropy of absorption and phosphorescence. J. Mol. Biol. 179, 55–81.
Egelman, E. H. & Orlova, A. (1995). New insight into
actin filament dynamics. Curr. Opin. Cell Biol. 5,
172–180.
Egelman, E. H., Francis, N. & DeRosier, D. J. (1982).
F-actin is a helix with a random variable twist.
Nature, 298, 131–135.
Erickson, H. P. (1989). Co-operativity in protein–protein
association. The structure and stability of the actin
filament. J. Mol. Biol. 206, 465–474.
Geeves, M. A. & Lehrer, S. S. (1994). Dynamics of the
muscle thin filament regulatory switch: the size of
cooperative unit. Biophys. J. 67, 273–282.
Harris, H. E. (1985). Lack of nucleotide cleavage on the
binding of G-actin-ATP to plasma gelsolin. FEBS
Letters, 90, 81–83.
Harris, H. E. (1988). The binary complex of pig plasma
gelsolin with Mg2+-G-actin in ATP and ADP. FEBS
Letters, 233, 359–362.
Hesterkamp, T., Weeds, A. G. & Mannhertz, H. G. (1993).
The actin monomers in the ternary gelsolin: 2 actin
complex are in an antiparallel orientation. Eur. J.
Biochem. 218, 507–513.
Holmes, K. C., Popp, D., Gebhard, W. & Kabsch, W.
(1990). Atomic model of the actin filament. Nature,
347, 44–49.
Janmey, P. A. & Stossel, T. P. (1986). Kinetics of actin
monomer exchange at the slow growing ends of
actin filaments and their relation to the elongation of
filaments shortened by gelsolin. J. Muscle Res. Cell
Motil. 7, 446–454.
Janmey, P. A., Peetermans, J., Zaner, K. S., Stossel, T. P.
& Tanaka, T. (1986). Structure and mobility of actin
filaments as measured by quasielastic light scatter-
765
ing, viscometry and electron microscopy. J. Biol.
Chem. 261, 8357–8362.
Khaitlina, S. Y., Moraczewska, J. & StrzeleckaGolaszewska, H. (1993). The actin/actin interactions
involving the N-terminus of the DNase-I-binding
loop are crucial for stabilization of the actin filament.
Eur. J. Biochem. 218, 911–920.
Kinosita, K., Kawato, S. & Ikegami, A. (1977). A theory
of fluorescence polarization decay in membranes.
Biophys. J. 20, 289–305.
Kurokawa, H., Fujii, W., Ohmi, K., Sakurai, T. &
Nonomura, Y. (1990). Simple and rapid purification
of brevin. Biochem. Biophys. Res. Commun. 168,
451–457.
Lehrer, S. (1972). The crosslinking of actin and of
tropomyosin by glutaraldehyde. Biochem. Biophys.
Res. Commun. 48, 967–975.
Lo, S. H., Janmey, P. A., Hartwig, J. H. & Chen, L. B.
(1994). Interactions of tensin with actin and
identification of its three distinct actin-binding
domains. J. Cell Biol. 125, 1067–1075.
Lorenz, M., Popp, D. & Holmes, K. C. (1993). Refinement
of the F-actin structure against X-ray fiber diffraction
data by the use of a direct mutation algorithm. J. Mol.
Biol. 234, 826–836.
Ludescher, R. D. & Thomas, D. D. (1988). Microsecond
rotational dynamics of phosphorescent-labeled
muscle crossbridges. Biochemistry, 27, 3343–3351.
Ludescher, R. D., Eads, T. M. & Thomas, D. D. (1987).
Triplet anisotropy studies of restricted rotational
motion in myosin monomers and filaments. In
Optical Studies of Muscle Crossbridges (Baskin, R. J. &
Yeh, Y., eds), pp. 3–65, CRC Press Inc., Boca Raton,
Florida.
Maruyama, K., Yamada, N. & Mabuchi, I. (1984).
Capping of one end of an actin filament affects
elongation at the other end. J. Biochem. 96, 613–620.
McLaughlin, P. J., Gooch, J. T., Mannhertz, H. G. &
Weeds, A. G. (1993). Structure of gelsolin segment
1–actin complex and the mechanism of filament
severing. Nature, 364, 685–692.
Miki, M. & Kouyama, T. (1994). Domain motion in actin
observed by fluorescence energy transfer. Biochemistry, 33, 10171–10177.
Miki, M., Wahl, P. & Auchet, J. C. (1982). Fluorescence
anisotropy of labeled F-actin: influence of divalent
cations on the interaction between F-actin and
myosin heads. Biochemistry, 20, 3661–3665.
Milligan, R. A., Whittaker, M. & Safer, D. (1990).
Molecular structure of F-actin and location of surface
binding sites. Nature, 348, 217–221.
Mossakowska, M., Belagyi, J. & Strzelecka-Golaszewska,
H. (1988). An EPR study of the rotational dynamics
of actins from skeletal and smooth muscle and their
complexes with heavy meromyosin. Eur. J. Biochem.
175, 557–564.
Mossakowska, M., Moraczewska, J., Khaitlina, S. &
Strzelecka-Golaszewska, H. (1993). Proteolytic removal of three C-terminal residues of actin alters the
monomer–monomer interactions. Biochem. J. 289,
897–902.
Muhlrad, A., Cheung, P., Phan, B. C., Miller, C. & Reisler,
E. (1994). Dynamic properties of actin. Structural
changes induced by beryllium fluoride. J. Biol. Chem.
269, 11852–11858.
O’Donoghue, S. L., Miki, M. & dosRemedios, C. G. (1992).
Removing of the two C-terminal residues of actin
affects the filament structure. Arch. Biochem. Biophys.
293, 110–116.
766
Cooperativity in Actin Filaments
Oosawa, F. & Kasai, M. (1971). Actin. In Subunits in
Biological Systems (Timasheff, M., ed.), part A,
pp. 261–322. Dekker, New York.
Orlova, A. & Egelman, E. H. (1995). Structural dynamics
of F-actin: I. Changes in the C terminus. J. Mol. Biol.
245, 582–597.
Orlova, A., Prochniewicz, E. & Egelman, E. H.
(1995). Structural dynamics of F-actin: II. Cooperativity in structural transitions. J. Mol. Biol. 245,
598–607.
Prochniewicz, E. & Yanagida, T. (1990). Inhibition of
sliding movement of F-actin by cross-linking
emphasize the role of F-actin structure in the
mechanism of motility. J. Mol. Biol. 216, 761–772.
Prochniewicz, E., Katayama, E., Yanagida, T. & Thomas,
D. D. (1993). Cooperativity in F-actin: chemical
modifications of actin monomers affect the functional interactions of myosin with unmodified
monomers in the same actin filament. Biophys. J. 65,
113–123.
Prochniewicz, E, Zhang, Q., Howard, E. & Thomas, D. D.
(1996). Microsecond rotational dynamics of actin:
spectroscopic detection and theoretical simulation.
J. Mol. Biol. 255, 446–457.
Sawyer, W. H., Woodhouse, A. G., Czarnecki, J. J. & Blatt,
E. (1988). Rotational dynamics of actin. Biochemistry,
27, 7733–7740.
Schurr, J. M. (1984). Rotational diffusion of deformable
macromolecules with mean cylindrical symmetry.
Chem. Phys. 84, 71–96.
Schwyter, D. H., Kron, S. J., Toyoshima, Y. Y., Spudich,
J. A. & Reisler, E. (1990). Subtilisin cleavage of actin
inhibits in vitro sliding movement of actin filaments
over myosin. J. Cell Biol. 111, 465–470.
Sun, H.-Q., Wooten, D. C., Janmey, P. A. & Yin, H. L.
(1994). The actin-binding domain of gelsolin
also caps actin filaments. J. Biol. Chem. 269,
9473–9479.
Tellam, R. L. (1986). Gelsolin inhibits nucleotide exchange
from actin. Biochemistry, 25, 5799–5804.
Thomas, D. D., Seidel, J. C. & Gergely, J. (1979).
Rotational dynamics of spin-labeled F-actin in the
sub-millisecond time range. J. Mol. Biol. 132, 257–273.
Tirion, M. M. & ben-Avraham, D. (1993). Normal mode
analysis of G-actin. J. Mol. Biol. 230, 186–195.
Tirion, M. M., ben-Avraham, D., Lorentz, M. & Holmes,
K. C. (1995). Normal modes refinement parameters
for the F-actin model. Biophys. J. 68, 5–12.
Way, M. & Matsudaira, P. (1993). The secrets of
severings? Curr. Biol. 3, 887–890.
Way, M., Gooch, J., Pope, B. & Weeds, A. G. (1989).
Expression of human plasma gelsolin in Escherichia
coli and dissection of actin binding sites by segmental
deletion mutagenesis. J. Cell Biol. 109, 593–605.
Weber, A., Pring, M., Lin, S. L. & Bryan, J. (1991). Role
of the N- and C-terminal actin binding domains of
gelsolin in barbed filament end capping. Biochemistry, 30, 9327–9334.
Weeds, A. & Maciver, S. (1993). F-actin capping proteins.
Curr. Opin. Cell Biol. 5, 63–69.
Wegner, A. & Engel, J. (1975). Kinetics of cooperative
association of actin to actin filaments. Biophys. Chem.
3, 215–225.
Yin, H. L. & Stossel, T. P. (1980). Purification and
structural properties of gelsolin, a Ca2+-activated
regulatory protein of macrophages. J. Biol. Chem. 255,
9490–9493.
Yin, H. L., Zaner, K. S. & Stossel, T. P. (1980). Ca2+ control
of actin gelation. J. Biol. Chem. 255, 9494–9500.
Yin, H. L., Hartwig, J. H., Maruyama, K. & Stossel, T. P.
(1981). Ca2+ control of actin filament length. J. Biol.
Chem. 256, 9693–9697.
Yoshimura, H., Nishio, T. & Mihashi, K. (1984). Torsional
motion of eosin-labeled F-actin as detected in the
time-resolved anisotropy decay of the probe on the
sub-millisecond time range. J. Mol. Biol. 179, 453–467.
Edited by P. E. Wright
(Received 8 January 1996; received in revised form 6 May 1996; accepted 10 May 1996)