You are on page 1of 42

C. G. Dos Remedios, D. Chhabra, M. Kekic, I. V. Dedova, M. Tsubakihara, D. A.

Berry and N. J. Nosworthy


Physiol Rev 83:433-473, 2003. doi:10.1152/physrev.00026.2002

You might find this additional information useful...

This article cites 301 articles, 99 of which you can access free at:
http://physrev.physiology.org/cgi/content/full/83/2/433#BIBL

This article has been cited by 53 other HighWire hosted articles, the first 5 are:
Helix Straightening as an Activation Mechanism in the Gelsolin Superfamily of Actin
Regulatory Proteins
H. Wang, S. Chumnarnsilpa, A. Loonchanta, Q. Li, Y.-M. Kuan, S. Robine, M. Larsson, I.
Mihalek, L. D. Burtnick and R. C. Robinson
J. Biol. Chem., August 7, 2009; 284 (32): 21265-21269.
[Abstract] [Full Text] [PDF]

Essential Role of Cofilin-1 in Regulating Thrombin-induced RelA/p65 Nuclear


Translocation and Intercellular Adhesion Molecule 1 (ICAM-1) Expression in Endothelial
Cells
F. Fazal, K. M. Bijli, M. Minhajuddin, T. Rein, J. N. Finkelstein and A. Rahman
J. Biol. Chem., July 31, 2009; 284 (31): 21047-21056.
[Abstract] [Full Text] [PDF]

SUMOylation of nuclear actin


W. A. Hofmann, A. Arduini, S. M. Nicol, C. J. Camacho, J. L. Lessard, F. V. Fuller-Pace and P.
de Lanerolle
J. Cell Biol., July 27, 2009; 186 (2): 193-200.
[Abstract] [Full Text] [PDF]

Manipulating Actin Dynamics Affects Human In Vitro Decidualization


I. Ihnatovych, M. Livak, J. Reed, P. de Lanerolle and Z. Strakova
Biol Reprod, July 1, 2009; 81 (1): 222-230.
[Abstract] [Full Text] [PDF]

Dopamine D4 Receptors Regulate GABAA Receptor Trafficking via an


Actin/Cofilin/Myosin-dependent Mechanism
N. M. Graziane, E. Y. Yuen and Z. Yan
J. Biol. Chem., March 27, 2009; 284 (13): 8329-8336.
[Abstract] [Full Text] [PDF]

Medline items on this article's topics can be found at http://highwire.stanford.edu/lists/artbytopic.dtl


on the following topics:
Physiology .. Actin
Biochemistry .. Actin-Binding Protein
Endocrinology .. Gelsolin
Biochemistry .. Motor Protein
Physiology .. Myofibrils
Oncology .. Actin Filaments

Updated information and services including high-resolution figures, can be found at:
http://physrev.physiology.org/cgi/content/full/83/2/433

Additional material and information about Physiological Reviews can be found at:
http://www.the-aps.org/publications/prv

This information is current as of October 2, 2009 .

Physiological Reviews provides state of the art coverage of timely issues in the physiological and biomedical sciences. It is
published quarterly in January, April, July, and October by the American Physiological Society, 9650 Rockville Pike, Bethesda MD
20814-3991. Copyright © 2005 by the American Physiological Society. ISSN: 0031-9333, ESSN: 1522-1210. Visit our website at
http://www.the-aps.org/.
Physiol Rev
83: 433– 473, 2003; 10.1152/physrev.00026.2002.

Actin Binding Proteins:


Regulation of Cytoskeletal Microfilaments
C. G. DOS REMEDIOS, D. CHHABRA, M. KEKIC, I. V. DEDOVA, M. TSUBAKIHARA,
D. A. BERRY, AND N. J. NOSWORTHY

Institute for Biomedical Research, Muscle Research Unit, Department of Anatomy and Histology,
University of Sydney, Sydney, Australia

I. Introduction 434
A. The cytoskeleton 434
B. What do we exclude from this review? 434
II. Structure and Dynamics of Monomeric Actin 434
A. The first actin crystals 434
B. Actin microcrystals and tubes 434
C. Actin monomer structure 435
D. MreB, an ancestral actin 436
E. Models of F-actin 437
F. Assembly of actin filaments 439
G. Elongation and annealing of F-actin 439
III. Actin Binding Proteins 440
A. ADF/cofilin family 441
B. Profilin family 449
C. Gelsolin superfamily 450
D. Thymosins 453
E. DNase I 455
F. Capping proteins 456
G. The Arp2/3 complex 458
IV. Role of Ternary Complexes in Regulating Actin Cytoskeleton Assembly 460
A. Cofilin, actin, and DNase I 461
B. Ternary complexes of actin with two or more ABPs 461
V. The Cytoskeleton and Pathology 462
A. Actin 462
B. Gelsolin 462
C. Tropomodulin 463
D. Connections between the extracellular matrix and the nucleus 463
E. Heart failure 463
F. Gene arrays 464
G. Single nucleotide polymorphisms, diseases, and the cytoskeleton 464
H. Cell signaling and actin microfilaments 464
VI. Unresolved Issues 464
A. Atomic structure of F-actin 464
B. Functional differences between actin isoforms 465
C. Ternary complexes of monomeric actin with ABPs 465
D. Undiscovered ABPs 465
E. Cooperative binding of ABPs along filaments 465
F. Cytoskeletal proteins in the nucleus 465
G. Disease and the cytoskeleton 465
H. Prokaryotic cytoskeletal elements 466

dos Remedios, C. G., D. Chhabra, M. Kekic, I. V. Dedova, M. Tsubakihara, D. A. Berry, and N. J. Nosworthy.
Actin Binding Proteins: Regulation of Cytoskeletal Microfilaments. Physiol Rev 83: 433– 473, 2003; 10.1152/phys-
rev.00026.2002.—The actin cytoskeleton is a complex structure that performs a wide range of cellular functions. In
2001, significant advances were made to our understanding of the structure and function of actin monomers. Many

www.prv.org 0031-9333/03 $15.00 Copyright © 2003 the American Physiological Society 433
434 DOS REMEDIOS ET AL.

of these are likely to help us understand and distinguish between the structural models of actin microfilaments. In
particular, 1) the structure of actin was resolved from crystals in the absence of cocrystallized actin binding proteins
(ABPs), 2) the prokaryotic ancestral gene of actin was crystallized and its function as a bacterial cytoskeleton was
revealed, and 3) the structure of the Arp2/3 complex was described for the first time. In this review we selected
several ABPs (ADF/cofilin, profilin, gelsolin, thymosin !4, DNase I, CapZ, tropomodulin, and Arp2/3) that regulate
actin-driven assembly, i.e., movement that is independent of motor proteins. They were chosen because 1) they
represent a family of related proteins, 2) they are widely distributed in nature, 3) an atomic structure (or at least a
plausible model) is available for each of them, and 4) each is expressed in significant quantities in cells. These ABPs
perform the following cellular functions: 1) they maintain the population of unassembled but assembly-ready actin
monomers (profilin), 2) they regulate the state of polymerization of filaments (ADF/cofilin, profilin), 3) they bind to
and block the growing ends of actin filaments (gelsolin), 4) they nucleate actin assembly (gelsolin, Arp2/3, cofilin),
5) they sever actin filaments (gelsolin, ADF/cofilin), 6) they bind to the sides of actin filaments (gelsolin, Arp2/3), and
7) they cross-link actin filaments (Arp2/3). Some of these ABPs are essential, whereas others may form regulatory
ternary complexes. Some play crucial roles in human disorders, and for all of them, there are good reasons why
investigations into their structures and functions should continue.

I. INTRODUCTION II. STRUCTURE AND DYNAMICS


OF MONOMERIC ACTIN

A. The Cytoskeleton Actin is only found in eukaryotes. It comprises a


highly conserved family of proteins that fall into three
The internal cytoskeleton of eukaryotic cells is com- broad classes: "-, !-, and #-isoforms. It is mainly located
posed of actin microfilaments, microtubules, and interme- in the cytoplasm, but it is also present in the nucleus
diate filaments. The cytoskeleton is dynamic and strong, where it may or may not have motor-associated functions.
The highest concentrations (!20% of the total protein) of
ever ready to adapt to demands on the cell. An important
actin are in striated muscles; however, significant quanti-
property of actin is its ability to produce movement in the
ties of actin are present in nonmuscle cells where it plays
absence of motor proteins. At the cell membrane micro-
a variety of roles including myosin-independent changes
filament assembly protrudes the membrane forward pro-
of cells shape, motor-based organelle transport, regula-
ducing the ruffling membranes in actively moving cells.
tion of ion transport, and receptor-mediated responses of
Microfilaments can also play a passive structural role by the cell to external signals.
providing the internal stiffening rods in microvilli, main-
taining cell shape, and anchoring cytoskeletal proteins.
A. The First Actin Crystals
The major focus of this review is to examine how actin
binding proteins (ABPs) control these processes. Finally, Few people in the field of actin remember that Chris-
we raise some interesting and challenging questions for tine Oriol-Audit was the first to report the crystallization
future research. of actin in the absence of an actin-binding protein (212).1
These crystals were achieved in the presence of polyeth-
ylene glycol but were too small to be useful for X-ray
B. What Do We Exclude From This Review? diffraction at that time. If today’s high-powered X-ray
beam facilities were available then, they may well have
Actin microfilaments provide the “rails” along which yielded the atomic structure of actin in the absence of an
myosin “motors” perform work in a variety of cellular ABP. Perhaps it was a cruel twist of fate that she sadly
functions. A major review of myosin motors (263) has died early in 2001 without knowing that the first structure
been published, and they are excluded from this review. of uncomplexed actin was soon to be published (216).
Microfilaments cooperate with microtubules via microtu- The principal obstacle to crystallizing actin was its pro-
bule-associated proteins (MAPs) during the transport of pensity to spontaneously self-assemble under solvent
vesicles and organelles, and this interesting aspect of conditions conventionally used to grow protein crystals.
microfilament function was recently reviewed (34). Actin
filaments also interact with intermediate filaments, a func- B. Actin Microcrystals and Tubes
tion that may play an important role in enabling extracel-
In their pioneering book on the biophysics of protein
lular stimuli to be transmitted to key targets like ribo-
polymerization, Fumio Oosawa and Sho Asakura (210)
somes and chromosomes deep within the cell. This field is
an emerging one. It will be covered by Quinlan et al. (235)
and is beyond the scope of this review. 1
Christine Oriol-Audit died in January 2001.

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


ACTIN AND ACTIN BINDING PROTEINS 435

predicted that actin, like tubulin, would form helically


tubular crystalline assemblies. We subsequently showed
that these structures could be formed in the presence of
the trivalent lanthanide ions, particularly Gd3" (3, 87).
These lanthanide-induced actin tubes (composed of a
single layer of monomers) and microcrystals (bilayered
sheets) provided the first views of the shape of the actin
monomer (3, 73). The computed average image of the
monomer was “pear” shaped, having a “large” and a
“small” domain. Electron diffraction studies revealed
structure out to at least 14.5-Å resolution, but this was
simply not good enough to provide reliable molecular
details (74). In solution, actin filaments can form su-
pramolecular assemblies called paracrystals (99) as well
as forming liquid crystalline arrays (67).

C. Actin Monomer Structure

Although the cocrystallization of actin with bovine


pancreatic DNase I was first reported in the late 1970s
(176), 13 years passed before the first atomic resolution
(2.4 Å) structure of actin was reported (137). The actin
monomers used in these crystals were lightly digested
with trypsin to remove the COOH-terminal three residues
before crystallization. Because the crystal structure of
DNase I had already been determined at 2.5-Å resolution
(272), it was a relatively simple task to subtract its struc-
ture from the actin-DNase I complex to obtain a clear
view of the actin monomer. The resulting structure (illus-
trated in ribbon form in Fig. 1A) produced a number of
surprises. For example, the large and small domains de-
scribed at low resolution contained nearly equal numbers
of residues and were nearly equal in size. Both the NH2
and COOH termini were located in the same subdomain.
DNase I makes crystal contacts across the “top” of the
FIG. 1. A: actin monomer ribbon structure of G-actin from DNase I
nucleotide cleft, thus explaining how it inhibits nucleotide cocrystals at 2.8-Å resolution (137). The numbered amino acid residues
exchange. However, since these were cocrystals, did this were chosen because of their relevance to this review. B: actin monomer
structure represent a monomer conformation when it was ribbon structure of uncomplexed rabbit skeletal muscle actin at 1.54-Å
resolution (217).
not complexed to DNase I?
Monomeric (G-)actin has dimensions of !67 # 40 #
37 Å and a molecular mass of nearly 43,000 Da. The affinity to Mg-G-actin and !200-fold weaker binding with
overall features of the structure include four quasi-subdo- Ca-G-actin (147). ATP binds as a complex with either
mains (here they are numbered 1– 4 but an alternative Mg2" [dissociation constant (Kd) 1.2 nM] or Ca2" (Kd 0.12
nomenclature of IA, IB, IIA, and IIB has an historical nM) (79, 270). Mg2", the dominant cation in vivo, deter-
basis, Ref. 266), each having a repeating motif comprising mines how tightly or weakly the nucleotide binds. Six or
a multi-stranded !-sheet, a !-meander, and a right-handed seven relatively low-affinity divalent cation-binding sites
!"!-unit. About 40% of the structure is "-helical. In Figure have been identified that appear to be important for actin
1A, the monomer is oriented so that the top has about the paracrystal (94, 271), and actin microcrystal formation
same orientation it would have in the filament with its (73), but their physiological importance is not well under-
“pointed” end directed at the top of the page and the stood. A notable feature of the structure is a two-stranded
“barbed” end toward the bottom. “hinge” at residues 140 and 338 joining the large and small
A tightly bound nucleotide lies in a deep cleft in the domains. Tirion et al. (283) describe a “propeller” motion
center of G-actin. This site is usually occupied by ATP or between the two domains that produces an opening and
ADP-Pi rather than ADP, which binds with !10-fold lower closing of the nucleotide cleft. This putative slewing mo-

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


436 DOS REMEDIOS ET AL.

tion is permitted because of this hinge and is viewed as a A significant structural difference actin crystallized
rigid body movement between the subdomains. alone and complexed with an ABP protein is relatively
There are several other atomic structures of actin. In large coil-to-"-helix conversion and a 10° rotation at the
1993, McLaughlin’s group at the Medical Research Coun- top of subdomain 2. Although it is not clear if these
cil (MRC) in Cambridge (186) reported the structure of differences represent native cellular G-actin, Dominguez
actin complexed with gelsolin segment 1. This segment et al. are currently completing the analysis of an uncom-
binds to the bottom of the monomer, making contacts plexed ATP-G-actin structure that should satisfy their
with subdomains 1 and 3 (see discussion below). Unlike critics.
the DNase I-actin structure, the COOH-terminal three res- The substantial conformational change reported by
idues were not cleaved before crystallization. However, Otterbein et al. (217) is well supported by other laborato-
the DNase I binding loop was not well resolved, presum- ries including our own. Dedova et al. (81) recently dem-
ably because it was not sufficiently stabilized in the ab- onstrated, using fluorescence spectroscopy, that signifi-
sence of DNase I. In the same year Schutt et al. (259) cant (1– 4 Å) changes occur in subdomains 1 and 2 when
reported the structure of spleen actin (a !-isoform) com- cofilin and DNase I bind either separately or as a ternary
plexed to profilin at a resolution of 2.55 ÅA. This group complex. Several other authors have suggested that actin
had been refining their structure for many years and can undergo allosteric conformational changes, and this
demonstrated that, although there was broad similarity to is discussed in more detail in section IV. The major chal-
the structure of McLaughlin et al. (186), there were sig- lenge for structural biologists now lies in relating the
nificant differences that they attributed to the sequence structural changes to functional changes.
differences between the actin isoforms.
The Schutt/Lindberg consortium that captured actin
in a different conformation subsequently reported a D. MreB, an Ancestral Actin
fourth actin-profilin structure. They again used bovine
!-actin cocrystallized with bovine pancreatic profilin (60). Actin is an essential and ubiquitous cytoskeletal com-
They named the two structures the open and closed states ponent of all eukaryotic cells and is absent from prokary-
and used them to argue that the transition between the otic cells. However, although it has long been suspected
two states represented a structural change that could that the evolutionary origins of actin lie in their prokary-
convert their “ribbon” polymer into conventional F-actin. otic ancestors, until 2001 no one had reported a credible
In 2001, the inevitable happened. Dominguez et al. at candidate. The first suggestion that MreB may be an an-
the Boston Biomedical Research Institute succeeded in cestral actin gene was published 10 years ago (30), and
producing small, high-quality crystals of actin that were recently (136), it was localized in Bacillus to form distinct
not complexed to ABPs (Fig. 1B). This finding was re- filaments located close to the cell surface. The atomic
ported at the Boston meeting of the Biophysical Society structure of these filaments was more recently reported
colleagues (216) and was subsequently published in Sci- by Fusinita van den Ent and colleagues (289) at the MRC,
ence (217). They achieved the result that had eluded so Cambridge, who provided convincing evidence that MreB
many others by modifying the COOH-terminal cysteinyl is indeed an ancestral actin gene. This actin look-alike is
(Cys-374) with a rhodamine label that suppressed the present in all nonspherical bacteria.
tendency of actin to polymerize. Amino acid sequence homology to actin is limited to
Dominguez and colleagues (217) were also the first to 15% (289), and although its overall size and shape strongly
crystallize actin with ADP present in the nucleotide-bind- resemble actin, there are differences. The sequence cor-
ing cleft. This was important because we know that the responding to the actin DNase I-binding loop in subdo-
critical concentration of ADP-G-actin is about an order of main 2 is larger, and the COOH terminus, located in
magnitude higher than for ATP or ADP-Pi-G-actin (241). subdomain 1, is 20 residues longer. A 2.1-Å resolution
Kabsch et al. (137) had also produced crystals of ADP-G- crystal structure reveals that, like actin, MreB has two
actin and showed that the ADP structure was not remark- major domains separated by a nucleotide-binding cleft.
ably different from the ATP form. However, the Kabsch Also, like actin, each domain has two subdomains with
crystals were formed in the presence of ATP, and the essentially the same topology. The structure of MreB is
nucleotide was cleaved to ADP subsequent to crystalliza- shown in Figure 2.
tion. MreB can self-assemble into 51-Å-wide filaments that
The structure of Dominguez et al. significantly differs are about half the diameter of actin microfilaments, but
from the structure of Kabsch et al., particularly in subdo- unlike F-actin, they are linear (nonhelical) polymers re-
main 2 where DNase I binds. However, it could be argued sembling the linear polymers seen in actin crystalline
that this Boston/Heidelberg difference is because ADP sheets first reported by ourselves (87) and later by others
was present before the crystals were formed, i.e., before (3). The axial repeat for MreB microfilaments is 51.1 Å,
the crystal contact points were established. somewhat less than the axial repeat of F-actin (55 Å) (Fig.

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


ACTIN AND ACTIN BINDING PROTEINS 437

analogs and control the shape of bacteria. However, we


know nothing of how assembly of this structure is regu-
lated, or whether it binds to motor proteins. Figure 3
illustrates the striking similarities between the crystal
structure of protofilaments of actin (right) and MreB
(left). Nearly identical molecular orientations and con-
tacts are seen between monomers for the two proteins.
Conversion of a pair of actin protofilaments into F-actin is
achieved by simply twisting a pair of actin protofilaments
to conform to the actin filament symmetry (see below).

E. Models of F-actin

The field has made substantial progress in under-


standing the structure of F-actin. Nearly 40 years ago Jean
Hanson and Jack Lowy were the first to describe the
helical nature of actin filaments (116). In this model,
F-actin can be viewed either as a single-start, left-handed
tightly wound helix of monomers (the so-called genetic
helix) or as a two-start, right-handed long-pitch helix. The
conventional view of F-actin is the two-start helix because
FIG. 2. Crystal structure of MreB showing the four-domain motif
it is probable that the monomer-monomer affinity is stron-
seen in the actin monomer structure (136). The NH2 and COOH termini
are located in subdomain 1A (equivalent to subdomain 1 of actin) (F. van ger along the two long-pitch strands than between those
den Ent, personal communication). strands (128).
No atomic structure has yet been determined for
3). These polymers form spirals beneath the bacterial cell F-actin, although highly plausible models have been pro-
wall, and although their precise function remains elusive, posed based on the model of Holmes et al. (128). Figure 4
it is likely that MreB filaments behave like microfilament illustrates five monomers in a filament. Viewed along the

FIG. 3. A comparison of crystal structure of MreB


(left) with a model of actin protofilaments (right) de-
rived from linear polymers along a single strand of
F-actin. Five monomers are illustrated in both proto-
filaments. Although the axial repeats between mono-
mers differ, the molecular contacts between mono-
mers in protofilaments of MreB are essentially the
same as the contact between actin monomers along the
two strands of F-actin (van den Ent, personal
communication).

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


438 DOS REMEDIOS ET AL.

FIG. 4. The Holmes model of F-actin showing six


monomers, three in each of the two-start long-pitch helix
originally described by Holmes et al. (128).

genetic helix, monomers rotate by $166° and have an fragments, but it can also be seen in good images of
axial translation of 27.5 Å. There are 13 monomers in 6 F-actin.
turns with a pitch of 59 Å yielding a filament diameter of Thus, despite the passage of nearly 40 years and the
!90 –100 Å (128). Contacts between the monomers along best efforts of laboratories in Germany, the United King-
this helix occur across the diameter of the filament. The dom, the United States, Japan, and elsewhere, the atomic
two-start long-pitch helix is right-handed with a some- structure of F-actin remains elusive. A major part of the
what variable half pitch of 360 –390 Å comprising 12–14 problem is the inherent disorder in linear aggregations of
monomers per half turn. In this model each monomer is filaments. Even if they could be aligned with the same
still surrounded by four others. The center-to-center dis- polarity, it is difficult to pack F-actin complexed to phal-
tance between monomers along this helix is !55 Å. These loidin (a 7-residue peptide that dramatically stabilizes
filaments have a distinct structural polarity that was first actin filaments) so there are precise contacts between
noticed when the filaments were “decorated” with myosin filaments. Devices like the innovative trumpet-shaped

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


ACTIN AND ACTIN BINDING PROTEINS 439

quartz tubes used by David Popp et al. (233) achieved Thus it is now more common to view F-actin as a dy-
greatly improved filament alignment that yielded the first namic, responsive structure than a passive structural el-
low-angle X-ray diffraction data. More recently, Yuichiro ement.
Maeda (who also worked in Heidelberg and has since
returned to the RIKEN SPring8 Institute synchrotron fa-
F. Assembly of Actin Filaments
cility in Japan) and colleagues achieved even better align-
ment of F-actin without the help of phalloidin using very
The most authoritative dissertation on the biophysics
high (13.5 Tesla) magnetic fields. They already have struc-
of F-actin is Oosawa and Asakura’s slim volume, Thermo-
tural data out to 5-Å resolution (207), and we can expect
dynamics of the Polymerization of Protein (210). Despite
them to report further progress in the near future.
its age, this book continues to provide a comprehensive
In the meantime, we must content ourselves with the
theoretical background for understanding the assembly of
available models of F-actin based on data from low-reso-
actin into filaments and higher order assemblies. The
lution electron microscope images. The original model
conditions under which actin monomers self-assemble in
(128) was created by fitting the atomic structure of the
vitro are well documented (211), although it is fair to say
actin monomer (137) to models based on the limited-
they are less well understood for in vivo assembly.
resolution X-ray diffraction data from aligned filaments.
Polymerization is essentially a condensation reac-
This model is illustrated in Figure 4. Subsequently, Mi-
tion. The main features of this process are 1) a slow initial
chael Lorenz et al. (167) refined the Holmes (128) model
association to a dimer that is more likely to rapidly dis-
using data from actin mutations. In the same year Tirion
sociate to monomers than to assemble; 2) the formation
et al. (283) described the propeller motion of the two
of a stable trimer that represents the nucleus of polymer-
major domains about the hinge and then refined the F-
ization, a state where actin assembly is more likely than is
actin model to take into account domain motion. Thus it
disassembly; and 3) the elongation phase during which
emerged that the actin monomer was not the rigid, static
actin monomers are rapidly assembled. Solvent condi-
(double strand of pearls) structure so commonly de-
tions that promote polymerization include high ionic
picted. Monomers in this model are positioned slightly
strength (KCl concentrations %50 mM), neutral or slightly
tangential to the filament axis, with subdomain 1 (the
acidic pH, high Mg2" (rather than high Ca2"), and ele-
major myosin S-1 binding site) being located at the high-
vated temperature (9, 111, 294), in other words, the con-
est radius.
ditions found in cells.
F-actin displays an arrowhead-like appearance when
In addition to these three processes, actin filaments
decorated with myosin subfragment 1 (S-1) (131, 188). For
are in a continuous state of assembly/disassembly. As a
this reason the opposite ends of the filament have been
consequence, in the steady state, a small but finite con-
named the “barbed” and “pointed” ends. These ends cor-
centration of actin monomers will be present in any fila-
respond to exposed subdomains 1 and 3 and subdomains
ment population (140, 211). This concentration of free
2 and 4, respectively.
monomers in equilibrium with a population of actin fila-
Quite a different model was proposed by Schutt,
ments is referred to as the critical concentration and is
Lindberg, and colleagues (259) based on crystallographic
dependent on solvent conditions, particularly on the pres-
principles and known examples from other structures in
ence of certain ABPs and/or actin ligands such as phal-
biology. In profilin-actin crystals, a “ribbon” structure can
loidin (80) or latrunculin A (193).
be seen in which monomer subdomains 1 and 2 lie close
to the ribbon axis. Alternating monomers displayed front
and back views. They proposed that monomers were G. Elongation and Annealing of F-actin
reoriented to produce F-actin. In essence, this group
turned the model of Holmes et al. (128) inside out by Elongation involves association and dissociation of
rotating the actin monomer about 180° in the plane of the monomers from the filament. These processes can occur
filament axis. The effect of this was to locate subdomains at either end of the filament, but association predomi-
1 and 2 closer to the filament axis, leaving subdomains 3 nantly occurs at the barbed end and dissociation at the
and 4 at higher radii. These authors further suggested that pointed end. The nucleotide binding site of monomeric
the subdomains were able to undergo significant move- actin is almost exclusively associated with ATP in vivo.
ments relative to each other and that this could feasibly Conversely, the majority of F-actin subunits contain
be built into a mechanism of contraction (258). bound ADP. While the G-actin bound ATP is readily ex-
Low-resolution (25–30 Å) electron microscope stud- changeable with solvent nucleotides, the ADP of F-actin is
ies have revealed that the actin filament can exist in essentially nonexchangeable (121). Hydrolysis of bound
multiple conformations depending on the type of bound ATP was originally thought to be tightly coupled to the
cation and nucleotide, the isoform of actin (213, 214), and polymerization process (210, 300); however, subsequent
the presence of other proteins bound to actin (182, 218). investigations (44, 46, 77, 221) revealed that a time lag

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


440 DOS REMEDIOS ET AL.

exists between the incorporation of ATP-G-actin onto the filament stabilizers (e.g., tropomyosin), capping proteins
filament end and hydrolysis of the bound nucleotide. The (e.g., CapZ, tropomodulin), ABPs that promote branching
kinetics of ATP hydrolysis and release of its product, Pi (e.g., Arp2/3), and ABPs that sequester G-actin and thus
(43), suggests that the bound ATP of newly added actin maintain a pool of monomers in solution (e.g., thymosin
subunits is hydrolyzed through two distinct sequential !4, profilin).
steps as illustrated in the following reaction scheme Finally, filament lengths are affected by fragmenta-
tion and annealing. Fragmentation can arise from thermal
motion in vitro, but in vivo much of this will be con-
ATP-G-actin 7 ATP-F-actin 3 strained by the other contents of the cytoplasm. Anneal-
Fast
ing occurs when an existing filament binds to the appro-
priate end of a second filament. This has been observed
directly using fluorescent-phalloidin-labeled single fila-
ADP-Pi-F-actin 7 ADP-F-actin"Pi ments (Ishiwata, personal communication) and has been
Slow quantified (2.2 $M$1!s$1) under similar conditions (145).
Arp2/3 (discussed below) is believed to create branch
The rate of ATP hydrolysis was calculated to be 10 times points in actin microfilaments by “capturing” existing fil-
higher than the rate of Pi release, suggesting that ADP-Pi- aments.
F-actin is the major intermediate in the nucleotide hydro-
lysis process. Conversion of ATP to ADP-Pi and then to
ADP " Pi is not random but primarily occurs at the III. ACTIN BINDING PROTEINS
growing (barbed) ends of filaments (47).
Depolymerization of actin is not simply the opposite Actin binds a substantial number of proteins collec-
of polymerization, principally because actin cannot regen- tively called ABPs. Some years ago, Pollard and Cooper
erate ATP from ADP and Pi (45). Instead, dissociated (231) identified a large number of ABPs, and recently we
ADP-actin subunits rapidly exchange their bound ADP for counted 162 distinct and separate proteins without in-
ATP in solution (200), a process that is accelerated by cluding their many synonyms or isoforms. No doubt more
profilin (see below). Polymerization of actin is not depen- will be identified. Many of the known ABPs bind to the
dent on nucleotide hydrolysis. It is possible under special same loci on the surface of actin and therefore can be
circumstances to form F-actin from G-actin containing expected to compete. A few bind with positive coopera-
either no bound nucleotide (141, 227) or a nonhydrolyz- tivity and tend to form ternary complexes (see below) but
able analog of ATP (63). However, nucleotide hydrolysis rather more bind with negative cooperativity. In myofi-
is required for the normal function of F-actin. brils, at least eight sarcomeric proteins bind to the thin
The structural polarity of the actin filament and the filaments. At least 12 ABPs are membrane-associated pro-
irreversible nature of ATP hydrolysis during actin assem- teins, and another nine are membrane receptors or ion
bly have implications for the rate and direction of filament transporters. Thirteen ABPs cross-link actin filaments,
growth at opposite ends of F-actin. The critical concen- whereas others enable filaments to interact with other
tration for the pointed end is 12- to 15-fold higher than for elements of the cytoskeleton. Microfilaments probably do
the barbed end under physiological conditions (301). This not interact directly with microtubules and/or intermedi-
difference may result in the unidirectional growth of the ate filaments but do so via linker proteins.
actin filament due to a continual flux of actin subunits Actin also binds !30 other ligands including drugs
from the pointed to the barbed end of the filament. This and toxins. Thus the sheer number of ligands that have a
reaction is called “treadmilling” (300). significant affinity for actin strongly suggests there is
Under in vivo conditions one might expect that all or probably a large number of binding sites that cover much
most of the actin exists in an assembled (filamentous) of the exposed surface of the molecule. A comprehensive
form. Slow addition at the barbed end and even slower list is available from Actin (266), the Guidebook to the
dissociation at the pointed end of the filaments produces Cytoskeletal and Motor Proteins (151), and most recently
a rate of treadmilling (300) of monomers (!2 $m/h) that from the two-volume Molecular Interactions of Actin
is !200-fold slower than observed in vivo. Monomers (88). Attempts to classify these ABPs leave many “or-
move progressively along the filament from the barbed phans” that do not fit into families, so any attempt to
end (the end associated with Z-disks of the sarcomere, or group them is bound to be somewhat arbitrary. Classifi-
the cell membrane where ruffling occurs in motile cells) cations according to the function of ABPs or their con-
toward the free pointed end of the filament (located in the sensus sequences can also be problematic.
middle of the sarcomere or oriented away from the cell Several types of ABPs facilitate disassembly and as-
membrane). ABPs present in vivo regulate different as- sembly. Below is a review of the most common ABPs.
pects of the assembly/disassembly process. These include Classification of these ABPs can be reduced to seven

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


ACTIN AND ACTIN BINDING PROTEINS 441

groups. 1) Monomer-binding proteins sequester G-actin virtually all eukaryotic cells. They are relatively small
and prevent its polymerization (e.g., thymosin !4, DNase (15–19 kDa) proteins that exist in multiple isoforms. Their
I). (Note: C. E. Schutt and Uno Lindberg are completing main functions include the rapid recycling of actin mono-
an extensive review of Actin Monomer Binding Proteins mers associated with membrane ruffling and with cytoki-
to be published in the Protein Profile series of Oxford nesis. Different genes encode for ADF and cofilin, but
University Press.) 2) Filament-depolymerizing proteins although it is common to regard them as synonymous,
induce the conversion of F- to G-actin (e.g., CapZ and they are distinctly different.
cofilin). 3) Filament end-binding proteins cap the ends of
the actin filament preventing the exchange of monomers
at the pointed end (e.g., tropomodulin) and at the barbed 1. Members of the family
end (e.g., CapZ). 4) Filament severing proteins shorten
the average length of filaments by binding to the side of ADF/cofilin was first discovered and purified in 1980
F-actin and cutting it into two pieces (e.g., gelsolin). from embryonic chick brain extracts by Bamburg et al.
(Note: H. Hinssen is completing an extensive review on (20). Since then, the family has grown to include a number
the Gelsolin Family to be published in the Protein Profile of related proteins, including invertebrate depactin
series of Oxford University Press.) 5) Cross-linking pro- (named because it depolymerizes actin) (173); porcine
teins contain at least two binding sites for F-actin, thus ADF or destrin (destroys F-actin) (191); cofilin (cosedi-
facilitating the formation of filament bundles, branching ments with filamentous actin) (1), Acanthamoeba acto-
filaments, and three-dimensional networks (e.g., Arp2/3). phorin (236); Dictyostelium coactosin (82), Drosophila
6) Stabilizing proteins bind to the sides of actin filaments twinstar or D-61 (91); unc-60A and unc-60B from Caeno-
and prevent depolymerization (e.g., tropomyosin). 7) Mo- rhabditis (185); Xenopus Acs (or XAC1 and XAC2) (2);
tor proteins that use F-actin as a track upon which to and finally Toxoplasma ADF (4). All of these share con-
move (e.g., the myosin family of motors). Here we con- siderable (30 – 40%) amino acid sequence identity. Fur-
sider only groups 1– 4. thermore, two other major protein families are related to
ABPs are not limited to one class, for example, gel- ADF through the presence of an ADF homology domain.
solin is capable of severing and capping the barbed end of One protein has a duplication of this domain and is con-
actin filaments, and the Arp2/3 complex can nucleate sequently called twinfilin; the other contains a single ADF
filament formation, elongate filaments, and establish homology domain linked to another motif and encodes
branch points in actin networks (69). In this review we the drebin family of proteins (16).
have selected eight ABPs that are biologically relevant Despite this somewhat confusing array of homologs,
and have atomic structures or good working models. vertebrates have genes for only two forms, ADF and
The rapid assembly of actin filaments is the principal cofilin. The names alone suggest that one depolymerizes
driving force behind many forms of cell locomotion. Cells F-actin (ADF) while cofilin cosediments with F-actin, but
can migrate at rates up to !0.5 $m/s (264). This means actually both can elevate the levels of monomeric actin
that filaments must have a net rate of elongation that is and both can bind to F-actin. Only one isoform of ADF is
slightly less than 200 monomers/s. Principally this occurs known in mammals (and birds), whereas two are known
at their barbed ends where MgATP-actin assembles about for cofilin. ADF and cofilin are clearly different but related
five times faster than MgADP-actin (149). The exchange of proteins.
ATP for ADP in filaments is so slow (of the order of days, Cofilin is diffusely distributed in the cytoplasm of
Ref. 201) it can be ignored. At steady state, filament length quiescent cells. However, in active cells, it translocates to
is constant, but there is a slow turnover (!2 $m/h) (264) cortical regions where the actin cytoskeleton is highly
of monomers due to a treadmilling process that involves dynamic and drives the ruffling of membranes of motile
the hydrolysis of ATP to ADP-Pi-actin and then a slower cells (18, 309), the cleavage furrow of dividing cells (197),
dissociation of Pi leaving most of the monomers in the the advancing of neuronal growth cones (18, 198), and
filament with bound MgADP. Thus, if pure actin filaments myofibrillogenesis (206). ADF and cofilin are not neces-
treadmill very slowly and if actin assembly is the driver of sarily expressed at the same level in all tissues. In adults,
cell motility, then the intrinsic rate of treadmilling must ADF is relatively highly expressed in nerves, intestine,
increase in vivo. This is the task of the ABPs. kidney, and testes (18), whereas cofilin levels are higher
in hematopoietic tissues, bone osteoclasts, and fibro-
blasts (309). In baby hamster kidney cells, ADF consti-
A. ADF/Cofilin Family tutes 0.4% of the soluble protein, while cofilin accounts
for 1.3% (148). The total ADF and cofilin amounts to 20
Unlike so many of the ABPs, the actin depolymeriz- $M in cells where the total actin concentration is more
ing factor (ADF)/cofilin family of proteins is expressed in than three times higher (67 $m) (148) (see Table 1).

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


442

TABLE 1. Actin binding protein characteristics

Actin Binding Arp2/3 (to


Protein Actin ADF/Cofilin Profilin Gelsolin Thymosin !4 DNase I CapZ Filaments)

Relevant figures Figs. 1, 4, 6 Figs. 5B, 6 (row 1), Figs. 6 (row 2), Figs. 6 (row 3), Figs. 6 (row 2), Figs. 6 (row 1), Figs. 6 (row 4), Fig. 17
(row 1) 7B, 8, 10 9, 10 7C, 11, 12A, 14 15 16
12B, 13
Amino acids; 375 amino acids; 165–168 amino 125–153 amino 720–730 amino 42 amino acids, 260 amino acids, " & 286, ! & 384 (Arp2) "
molecular 43 kDa (288) acids; 18–19 kDa acids; 138 acids in 6 4,963 Da (248) 29 kDa (222) 277 amino 391 (Arp3) 7
mass (261) amino acids subunits, 80 acids; 59 kDa subunit
14.9 kDa (138) kDa (86) (38) protein 220
kDa (172)
Residues on 2–5, 40–50, 61– 1–12, 25, 147, 167, 113, 166, 167, 1–18, 112–117, 1–4, 41, 154, 155, 39–46, 50, 53, 338–348 (156), N/A (actin
actin that bind 64, 89–100, 288, 292, 328, 334, 169, 171–173, 144, 146, 148, 167 (247, 251) 60–64, 68, 69, 360–372 (131) subdomains 2
ABP (Fig. 6) 110–112, 130– 341, 344, 345, 348, 284, 286–288, 167, 169, 285– 202–204 and and 4)
131, 166–173, 357–375 (304) 290, 354, 355, 375, 334, 350, 207 (137)

Physiol Rev • VOL


195–197, 202– 361, 364, 369, 351, 354, 356–
205, 243–247, 371–373 and 375, 374 and
266–269, 285– 375 (259) 375 (186)
289, 322–325,
357–360, and
375 (128)

83 • APRIL 2003 •
Concentration in 65–70 $M (19); 20 $M (0.44% of 20–100 $M (230) !5 $M Up to 500 $M (Not known) 0.6 $M (1:500 2 $M Arp2, 5
cells 300 $M (299) total protein) (18) (249) actin $M Arp3 (41,
DOS REMEDIOS ET AL.

20% of total monomers) 143)


muscle (305)
protein (266)
Binding to Weak Segment 1 to 0.05 nM (177) Only the !- Binds weakly to

www.prv.org
pH 8, ATP-GA Kd 8 Mg-ATP Kd 0.5 Mg-ATP-GA Kd
G-actin $M; ADP-GA Kd $M; Mg-ADP barbed end Kd 1.7 $M; Mg- subunit can actin dimers
0.1 $M (107, 240) Kd 5 $M 1 $M; ADP-GA & 80 bind GA (53) (242)
(261); Mg-ATP segments bind $M (42)
Kd 0.1 (6) to 2 GAs 0.07
$M (257)
Binding to actin (ATP-GA) Kd 0.1 pH %7, ADP-FA Delivers ATP-GA 5–60 nM at Weak (41) Kd 0.1 mM (177) 1:1, Kd 0.5–1.5 Kd 10 $M (242,
filaments $M at the (0.3 $M); ATP-FA to the barbed barbed end nM (53) 262)
barbed end; (10 $M) (42, 202) end of FA (146); 50 nM
Kd 0.6 $M at (262)
the pointed
end (228)
Pointed end Critical Accelerates off rate Very weak Yes ? Yes Yes No (Tmod binds Kd 10 nM (242,
effect concentration at pointed end, 9 (disassembles pointed ends) 262)
is 0.6 $M s$1 (240) FA)
(228). ATP on
rate 0.5 $M/s;
off rate 1 $M/
s; ADP on
rate 0.1 $M/s;
off rate 0.2
$M/s
Barbed end Critical Inserts monomers With cofilin, No ? Yes No Blocks No
effect concentration 0.15 $M (262). disassembly is exchange (Kd
is 0.1 $M No change in off fast: 125 s$1 1 nM) (255)
(228); ATP on rate (85)
rate 5 $M/s;
off rate 1 $M/

Physiol Rev • VOL


s; ADP on
rate 1 $M/s;
off rate 6 $M/
s
Caps FA N/A No Yes Yes No No Yes Yes

83 • APRIL 2003 •
Nucleates Nucleus is an No Inhibits Yes No No Yes Yes (302) near
assembly actin trimer nucleation membrane
(143)
Cross-links No No No No No No No Yes at 70°
filaments
ACTIN AND ACTIN BINDING PROTEINS

www.prv.org
Severs FA No Yes No Yes No No No No
Side binding to Stronger along a No No Yes No No No Yes Kd 0.5 $M
F-actin strand
Important Mg-ATP (Kd 1.2 PIP2 inhibits Promotes PIP2 inhibits Inhibits PIP2 inhibits PIP/PIP2 inhibit Arp2 binds
functional nM) (79) depolymerization nucleotide severing (275) nucleotide nucleotide binding of profilin (143);
ligands (283); inhibits exchange exchange exchange CapZ to actin activated by
nucleotide (203); PIP2 (107) (127); inhibits filaments WASP, Scar 1
exchange (27) inhibits binding to (123) (103)
binding (110) actin (130)

ABP, actin binding protein; Kd, dissociation constant; ADF, actin depolymerizing factor.
443
444 DOS REMEDIOS ET AL.

Recently, Vartiainen et al. (291) clarified the ADF/ 2. Structure of ADF/cofilin


cofilin nomenclatures for vertebrates. Using mice, they
proposed that cofilin 1 is expressed in most tissues of Porcine destrin was the first structure in this family
embryonic and adult cells, cofilin 2 is expressed in muscle of proteins to be solved by NMR spectroscopy (118).
cells, and ADF is limited to epithelial and endothelial Subsequently, atomic structures of three more members
cells. More importantly, they defined the functional simi- of the ADF/cofilin family have been determined. A ribbon
larities and differences between them. representation of yeast cofilin is shown in Figure 5A

FIG. 5. A: the atomic structure of yeast cofilin solved


at 1.8 Å showing the NH2 and COOH termini. Putative sites
for binding F-actin and G-actin (purple helix) are located
in the COOH-terminal region of the molecule. The solid
spheres identify residues that are essential for actin bind-
ing. [Modified from Fedorov et al. (95).] B: the crystal
structure of yeast cofilin (red) was fitted using molecular
dynamics to the actin monomer structure (grey) using
energy minimization. This approximate orientation is pre-
served in views of subsequent actin-ABP complexes.
(Model kindly provided by Willy Wriggers.)

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


ACTIN AND ACTIN BINDING PROTEINS 445

solved at 1.8 Å (95). Crystal structures have also been the presence of muscle and nonmuscle tropomyosins
determined for both Acanthamoeba actophorin (161) and (17). Tropomyosin is located deeper into the cortex from
a plant ADF from Arabidopsis thaliana (32). No atomic the site of active actin assembly and its presence probably
structure has been published for any vertebrate cofilin or inhibits cofilin binding (28). Cofilin and phalloidin also
for any muscle ADF/cofilin, but the atomic coordinates compete for binding to F-actin (120), even though their
for a high-resolution NMR structure for chick cofilin from binding sites are separated by 15–20 Å. McGough et al.
embryonic skeletal muscle were recently deposited in the (182) proposed that an allosteric change in the twist of
BioMagRes Bank (http://bmrb.wisc.edu) under BMRB ac- F-actin induced by cofilin may inhibit phalloidin from
cession number 5177 (11). binding between monomers close to the filament axis.
Indirect methods have been used to examine the sites The blue residues in Figure 6 represent the probable
on cofilin that bind to actin. Despite the best efforts of minimal molecular contacts for ADF/cofilin on actin. The
several groups, these two proteins have not yet been principal motif shown in Figure 5, A and B (a 4-stranded
cocrystallized, and consequently, unlike DNase I, gelsolin !-sheet surrounded by 4 "-helices), is also found in an
segment 1, and profilin, the precise contact sites between unrelated sequence of gelsolin segment 1. It is therefore
cofilin and actin are not known. Structural homologies not surprising that cofilin competes with gelsolin segment
between cofilin, gelsolin, and profilin strongly suggest 1 and profilin (85) for binding to G-actin. Both profilin and
they bind to the same region of G-actin (i.e., to subdo- gelsolin segment 1 bind to the barbed end of the actin
mains 1 and 3). However, the functional products of sys- monomer (subdomains 1 and 3), even though only Glu-
tematic mutagenesis of yeast cofilin indicate that the co- 167 and Tyr-169 (see Table 1) are common binding sites.
filin-actin interaction is distinctly different. Because cofilin competes with both proteins, it is possible
Competitive binding of a synthetic dodecapeptide that it too interacts with residues in the 162–176 loop of
(residues 104 –115 of vertebrate cofilin) has implicated actin subdomain 3 (see Fig. 1A).
this region in the interaction of cofilin with monomeric The structural sites to which cofilin binds along the
but not with filamentous actin (308). This peptide is outside of the filament are not known in atomic detail, so
largely conserved in other members of the ADF/cofilin the blue residues in Figure 6 are based on chemical
family including destrin and depactin, suggesting that it cross-linking and other data. At low resolution (25 Å),
may be a consensus sequence essential for actin binding ADF/cofilin bridges subdomains 1 and 3 of an upper actin
and depolymerizing activities. to subdomains 1 and 2 of a lower actin in a filament. The
Mutagenesis of surface residues for both yeast (158) major axis of cofilin makes a 30° angle with the plane
and chick (154) cofilin have implicated the NH2-terminal normal to the helical axis of the filament. It is centered
region in G-actin binding. Zero-length cross-linking of (axially) at about the position of subdomain 2 of the lower
G-actin to vertebrate cofilin by 1-ethyl-3-(3-dimethylamino- actin subunit and radially at the cleft between subdo-
propyl)carbodiimide (308) revealed that either Lys-112 mains 1 and 3 of the upper actin subunit. Electron cryo-
and/or Lys-114 of cofilin is in direct contact with actin. microscopy and helical reconstructions of F-actin deco-
Mutations of these cross-linked residues (192), or the rated with cofilin revealed its ability to reduce the angle of
corresponding Arg-96 and Lys-98 of yeast cofilin, totally rotation between subunits along the short-pitched left-
abolished actin binding. Thus a significant fraction of the handed genetic helix by 3–5° (182). There is no effect on
surface residues has been implicated in the binding of the axial rise per actin subunit, and consequently, the
cofilin to G- and F-actin (Fig. 5A, circles) and within these long-pitch helical repeat of F-actin is decreased from
areas, eight or nine residues (solid spheres) are consid- !360 to 270 Å, and the diameter increases slightly. The
ered essential. Despite the absence of cocrystals, the change of helical pitch can even be seen in negatively
cofilin structure has been “docked” onto the Kabsch et al. stained filaments in the absence of cofilin (Fig. 7a) and in
(137) actin structure (Fig. 5B) using energy minimization the presence of stoichiometric cofilin (Fig. 7b). We are
techniques (304). This figure provides a clear view of the grateful to Roger Craig for providing these micrographs.
structural relationship between these two proteins and is
presented so it can be compared with the binding sites for 3. Cellular functions
profilin and DNase I (see below).
Bamburg (16) recently stressed the importance of A) TREADMILLING. The ADF/cofilin family is almost sin-
switching from an ADF/cofilin system to a tropomyosin- gle-handedly credited for the high rate of treadmilling of
based regulation of F-actin function. Tropomyosin binds monomers in actin filaments in vivo. This is achieved in
along the grooves of the long pitch helix of F-actin. In part by increasing (by 30-fold) the off rate (9 s$1, Ref. 240)
myofibrils, where tropomyosin is constitutively ex- at the pointed ends of filaments without changing the off
pressed, the binding of tropomyosin and cofilin to actin rate at the barbed ends (42). However, in the presence of
filaments is mutually exclusive and competitive (311). profilin (see sect. IIIB), the rate of pointed-end disassem-
Thus the depolymerizing activity of cofilin is inhibited by bly is even faster (125 s$1, Ref. 85). The resulting elevated

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


446 DOS REMEDIOS ET AL.

concentration of cofilin-G-actin in the cytoplasm can be


rapidly recycled at the barbed end of filaments, provided
that ATP can replace the actin-bound nucleotide. This
concept is illustrated in Figure 8. In vivo, the concentra-
tion of ADF/cofilin is low relative to other ABPs such as
profilin and thymosin !4 (see discussion below). ADF/
cofilin “decoration” of actin filaments is probably re-
stricted to their “aged” ends where ATP has been con-
verted to ADP, i.e., some distance from the membrane
surface where the filaments are actively growing (see
below).
B) DEPOLYMERIZATION. At steady state, increasing the
rate of dissociation at the pointed end would not depoly-
merize actin filaments if the released monomers were able
to reassemble at the barbed end. However, if a barbed-end
capping protein (e.g., CapZ) blocks reassembly, then co-
filin will depolymerize actin filaments.
C) NUCLEATION OF POLYMERIZATION. ADF/cofilin can nucle-
ate the assembly of actin, and this function is likely to be
especially important in the presence of barbed-end block-
ers like CapZ. The ability to nucleate assembly is probably
pH dependent and may vary between different isoforms of
ADF/cofilin.
D) ADP-ACTIN. Binding of ADF/cofilin to actin is regu-
lated, at least in part, by the state of the nucleotide bound
to actin. It binds to ADP-actin with about two orders of
magnitude greater affinity than to ATP-actin (Table 1) or
ADP-Pi-actin, and this is true for both the G- and F-actin
(42) at pH 7.8. Thus cofilin not only promotes the disas-
sembly of ADP-actin monomers from filaments, but it also
binds to released ADP-actin monomers and inhibits the
exchange of their bound nucleotide (202). The off rate at
the pointed ends of filaments is probably the rate-limiting
step in the recycling of disassociated actin subunits (281).
The extent of depolymerization by ADF/cofilin depends
on several factors; for example, different isoforms may
have differing pH sensitivities (240, 183), and other ABPs
can influence the binding of cofilin to actin because they
share binding sites.
E) SEVERING. Marie-France Carlier et al. (42) argue that
the kinetic effects of ADF/cofilin are specific for the ends
of filaments and therefore it cannot sever. However, the
rate of filament fragmentation would not have to be very
great to increase the disassembly rate from !125 s$1 in
vitro to the rate required to match the speed of motility
(!200 s$1). Evidence based on light microscopy (119)
FIG. 6. The molecular contact sites between actin (shown here as
Van der Waals representations) and ABP sites. “Front” (A) and “back” suggests that filaments are rapidly depolymerized by ADF
(B) views of the monomer are shown where the binding sites are at slightly alkaline pH. Also, there is a difference in sev-
colored: row 1: yellow, actin monomer-monomer interaction sites; light ering activities of recombinant and native cofilin where
green, DNase I; and blue, ADF/cofilin; row 2: dark green, thymosin !4;
and red, profilin; row 3: pink, gelsolin segment 1; and black, vitamin the native form is more active than the recombinant co-
D-binding protein; row 4: red, CapZ sites. There is significant overlap filin. Recombinant cofilins can vary in their severing ac-
between the actin binding sites for ADF/cofilin, profilin, gelsolin seg- tivity and can give the impression that they have only
ment 1, and CapZ and are therefore shown on separate figures. These are
the minimal contacts and are based on chemical cross-linking and weak severing activity (132). Thus it remains unclear
mutagenesis data. whether ADF/cofilin can sever actin filaments like true
severing proteins or whether the highly cooperative bind-

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


ACTIN AND ACTIN BINDING PROTEINS 447

FIG. 7. Electron micrographs of negatively stained


(uranyl acetate) rabbit skeletal muscle actin. a: F-actin
only. b: F-actin " chick cofilin (molar ratio 1:1). c: F-actin
" human gelsolin (1:0.2). (Micrographs provided by Roger
Craig.)

ing of cofilin at substoichiometric ratios makes the fila- presence of two cofilins (yeast) per actin subunit. These
ments “brittle” at the points where the decorated and were seen by electron microscopy as two distinct sites in
undecorated regions meet (19). Recently, Ed Egelman image reconstructions. This raises the question of
and colleagues (104) reported filament severing in the whether one or two cofilins bind to actin monomers, a
feature which may depend on the isoform of cofilin. Thus
the question is whether cofilin accelerates treadmilling by
severing or depolymerizing F-actin. Current opinion
seems to favor the severing ability of cofilin.

4. Cellular concentrations

The cellular concentrations of ADF/cofilin are sub-


stantially lower than the concentration of actin, so it is
unlikely there is sufficient ADF/cofilin present in cells (20
$M, Ref. 148) to bind to all polymerized and unpolymer-
ized actins (65 $M; see Table 1), even taking into account
its low affinity for ATP- or ADP-Pi-actins. Furthermore,
the binding of ADF/cofilin to filaments is cooperative
(182). If there is not enough ADF/cofilin to act as a mono-
mer-sequestering protein, this role probably belongs to
profilin and thymosin !4 (see below).
Relatively few ABPs bind to the pointed compared
with the barbed end of a filament. Cofilin can compete
with 1) spectrin that stabilizes short-actin oligomers (see
below) (254); 2) tropomodulin that caps the pointed end
of tropomyosin-coated actin filaments in muscle and non-
muscle cells (97); 3) DNase I, which binds very tightly to
the pointed end of actin monomers but only weakly to actin
filaments (see sect. IIIE); and 4) new pointed-end ABPs that
are likely to be identified from proteins that currently are
either orphans or have not yet been identified in gene
arrays and two-dimensional gel electrophoresis.
A number of isoforms of cofilin have been described,
and it is possible that not all members of the family
perform the same functions under the same conditions
FIG. 8. A cartoon representation of F-actin in the absence (left) and (e.g., divalent cations and pH) (128). Some cofilins bind
presence (right) of cofilin. Cofilins (red crescents) preferentially bind to tightly to actin monomers while others do not (Noswor-
ADP-containing monomers in the filament and induce a small change in
the angular rotation of monomers in the filament with a consequent
thy, unpublished observations). Some cosediment with
slight increase in filament diameter and decrease in filament length. F-actin while others have only a low affinity (120). Some
Cofilin accelerates the off rate at the pointed ends of the filaments. Actin promote actin assembly while others rapidly depolymer-
monomers are represented as having a small and a large domain and are
color coded according to their bound nucleotide (brown for ATP, orange ize it (128). Human cofilin seems to be a better nucleator
for ADP-Pi, and yellow for ADP). of assembly whereas ADF is a better polymerizing agent,

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


448 DOS REMEDIOS ET AL.

and this is reflected in their respective Kd values (42, 240) expressed in cultured cells, LIMK-1 induces actin reorga-
(Table 1). nization and reverses cofilin-induced depolymerization
If different isoforms of cofilin have different effects (307). Expression of inactive LIMK-1 results in the accu-
on microfilament assembly, are they differentially ex- mulation of F-actin filaments (7). Phosphorylation of co-
pressed within a cell? Mouse myocyte cells lines express filin appears to increase with pH-induced activation of
significant amounts of ADF. Elevated expression of actin cofilin, consistent with a compensating homeostatic
causes a downregulation of ADF and decreases spreading mechanism (27).
of these cells without changing the levels of cofilin (256). Although NH2-terminal Ser residues (Ser-1, -3, -4,
Regulating ADF/cofilin expression is likely to be slow (on and/or -6 in different species) are recognized as the site
a time scale of tens of minutes/hours) and, apart from for phosphorylation, it is not clear whether the resulting
being energetically wasteful, it is unlikely to be an effec- inhibition of actin binding involves a steric or conforma-
tive way of enabling microfilaments to respond quickly to tional change. Recently, Blanchoin et al. (29) examined
environmental stimuli. However, because most ABPs bind this question by crystallizing Acanthamoeba actophorin
with moderate to low affinity (in the micromolar range), (ADF/cofilin) in the phosphorylated state. Their structure
elevated levels of ADF/cofilin could, over time, subtly was essentially identical to unphosphorylated ADF/cofi-
shift the balance of ABPs and, by mass action, alter the lin, and they concluded that inhibition of activity by LIM-
state of assembly of cytoskeletal microfilaments. kinase was not due to a conformational change at the
ADF/cofilin-actin interface. Their reported steric blocking
5. Phosphorylation of ADF/cofilin of actin binding was consistent with molecular dynamics
predictions (Fig. 5B) and with site-directed mutant stud-
The ability of cofilin to bind G-actin is also regulated ies. Frustratingly, Ser-1 residue in these crystals could not
by phosphorylation of Ser-3, which is conserved in most be resolved.
members of the ADF/cofilin family. When ADF/cofilin is Until recently, the mechanism by which cofilin is
phosphorylated, there is a sharp fall in its affinity for actin dephosphorylated was not well understood. Slingshot, a
(190). ADP exchange in G-actin is strongly inhibited by protein phosphatase with F-actin binding ability, was
cofilin, but once cofilin is phosphorylated, the released shown by Niwa et al. (204) to dephosphorylate cofilin in
ADP-actin monomer can exchange with cytoplasmic ATP, cultured cells and in cell-free assays. In Drosophila, loss
and it is now ready for reincorporation at the barbed end of this enzyme results in a dramatic increase in cellular
of a growing filament. Typically, this occurs at the inter- levels of both F-actin and phosphorylated cofilin.
face of the microfilaments and the membrane at the lead-
ing edge of the moving cell. Thus cellular microfilament 6. Inhibition by phosphatidylinositol 4,5-bisphosphate
turnover can potentially be regulated by cycles of phos-
The activity of ADF/cofilin can be regulated by the
phorylation and dephosphorylation. As we will see below,
membrane lipids phosphatidylinositol 4-phosphate (PIP)
profilin plays an important role here.
and phosphatidylinositol 4,5-bisphosphate (PIP2) (309).
We know that in yeast, phosphorylation is not a
These bind to and inhibit the actin-binding domain of
regulatory factor whereas it is in most vertebrates. An
ADF/cofilin at residues 104 –115 and the NH2-terminal
important question is, How much of total cellular ADF/
region (154) and suggest that transmembrane signaling by
cofilin is phosphorylated? For most cells, we simply do
PIP2 can regulate the function of ADF/cofilin. Little is
not have this information, but in amoeba !30% of acto-
known of the molecular nature of this binding site.
phorin is phosphorylated (29). On balance, it seems likely
that, since not all of ADF/cofilin is constitutively active,
7. pH effect
there may be mechanisms other than phosphorylation
that control actin assembly-disassembly in vertebrates. The ability of ADF/cofilin to assemble or disassemble
Phosphorylation of vertebrate cofilin is achieved by F-actin is pH dependent in vitro (311). Acidic conditions
LIM-kinase proteins 1 and 2 (LIMK-1 and LIMK-2). LIMK-1 (less than pH 6.8) enhance the ability of ADF/cofilin to
is predominantly neuronal, whereas LIMK-2 is more wide- stabilize F-actin while at more alkaline pH (%7.3), cofilin
spread (26, 187). Although the signaling pathway for ac- can rapidly depolymerize F-actin. Furthermore, the criti-
tivation of the LIM-kinases is not yet fully understood, it is cal concentration of the cofilin-actin complex is known to
known that LIMK-1 is under the control of the small be pH dependent, being low (!2 $M) at pH 6.5 and
GTPase Rac (307), whereas LIMK-2 is regulated by the significantly higher (7 $M) at pH 8.2 (240). Conversely, it
GTPase cdc42 and rho (21) (see discussion below). LIMKs has been suggested that the severing activity of cofilin is
are themselves regulated by phosphorylation of a Thr pH independent (89). The physiological significance of
residue (90). In Acanthamoeba, activation of LIM-kinase this pH dependency is that actin filaments may be stabi-
by Pak1 (91) results in phosphorylation of Ser-1 (the lized in the vicinity of membrane Na"-H" antiports (16). A
Acanthamoeba version of Ser-3) of actophorin (29). When more recent report (27) focused on the role of pH in vivo.

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


ACTIN AND ACTIN BINDING PROTEINS 449

Using pH-sensitive fluorescent probes introduced into


mouse fibroblasts (Swiss 3T3 cells), they (27) showed that
by lowering intracellular pH, ADF colocalized with actin
filaments and that when the pH was increased, ADF par-
titioned more with monomeric actin. In contrast, cofilin is
much less responsive to altered pH, and there are no
reports of structural changes induced in either ADF or
cofilin, despite the fact that its structure has been deter-
mined by both NMR spectroscopy (11) and crystallogra-
phy (158).

8. Nuclear translocation sequence

All vertebrate ADF/cofilin (destrin) sequences con-


tain a putative nuclear localization sequence (NLS) that
enables it to migrate from its normal location in the
cytoplasm to the nucleus under conditions of cellular
stress (203). Heat shock and other forms of stress prob-
ably affect a few residues immediately preceding the NLS,
which expose it for subsequent binding to a nuclear trans-
port factor and thus passage through the nuclear pores
(32). Transport is an active process, and actin accompa-
nies the cofilin into the nucleus, although the significance
of this is not clear. FIG. 9. A ribbon diagram of the atomic (2.55-Å resolution) structure
of profilin (red) cocrystallized with monomeric actin where the orien-
tation of actin is about the same as seen in Figure 5B (259).
B. Profilin Family

355, 361, 364, 369, 371, 373, 375; subdomain 3, 166, 167,
The profilin family of ABPs is found in eukaryotic
169, 171–173, 284, 286 –288, 290 (see Fig. 6).
cells from Acanthamoeba through to human and in many
but not all species there are two or more profilin genes
2. Cellular functions
(230). Profilins are small proteins with an approximate
molecular mass of 19 kDa (6). They are among the most Profilin has several cellular functions. It is essentially
highly expressed (20 –100 $M, Ref. 37) of the cytoplasmic a high-affinity (Ka & 107 M$1) monomer-binding protein
proteins and are distributed throughout the cytoplasm. (225). It is best known for its ability to promote the
Like cofilin and DNase I, profilin can be denatured in 8 M exchange of nucleotide in actin monomers released from
urea and then renatured by dialysis (138), a property used filaments (109) and, because the high (millimolar) con-
during its purification. centrations of Mg-ATP in the cytoplasm, it catalyzes the
exchange of ADP for ATP. It achieves this by binding to
1. Atomic structure subdomains 1 and 3 near the hinge between the two major
domains (see Figs. 6 and 9) and modulates the opening of
X-ray diffraction and NMR atomic resolution struc- the nucleotide cleft. Profilin enhances filament turnover
tures are available for no less than six isoforms of profilin. in the presence of cofilin (85) because the two proteins
All are very similar. Profilin has also been cocrystallized act at opposite ends of a filament, with profilin adding
with actin in two different conformations, the so-called ATP-actin to the barbed end and cofilin dissociating ADP-
“closed” and “open” states (60). It was on the basis of actin from the pointed end (see Fig. 10). Profilin also
these two structures that Schutt et al. (259) developed inhibits the hydrolysis of ATP bound to actin, thus main-
their model of F-actin (discussed above). Within the pro- taining actin monomers in a state where they retain a high
filins, the NH2-terminal sequence is conserved whereas affinity for the growing barbed end of filaments (6).
the COOH-terminal region of certain isoforms resembles While profilin acts as an effective buffer for high
gelsolin (see below). The atomic structure of the profilin- monomer concentrations in cells, it can promote polymer-
#-actin complex is illustrated in Figure 9. Profilin binds to ization by transporting monomers to the fast-growing
subdomains 1 and 3 at loci that substantially overlap the barbed ends of filaments. Profilin binds to a site on actin
binding sites of gelsolin segment-1 (259). Its main crystal (see row 2 in Fig. 6) that is inaccessible in the Holmes
contacts with actin are as follows: subdomain 1, 113, 354, model of the filament (128) (see Fig. 4). Thus, after deliv-

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


450 DOS REMEDIOS ET AL.

terminal half (180), whereas CapG (aka gCap39) is a


smaller, three-domain gelsolin analog that cannot sever
F-actin but responds to Ca2" and PIP2 under conditions
where gelsolin is ineffective (312).
Unlike ADF/coflin and profilin, an isoform of intra-
cellular gelsolin can circulate in plasma where it severs
and caps actin filaments released into the circulation (e.g.,
following cell death). The resulting actin monomers and
oligomers are then strongly sequestered by vitamin D-
binding protein (DBP) and ultimately removed from the
circulation in the liver (124). Plasma gelsolin is slightly
larger (83 kDa) than cytosolic gelsolin and is derived by
alternative splicing of a single gene resulting in a short
peptide appended to its NH2 terminus (155). Otherwise,
the sequences of plasma and intracellular gelsolins are
identical.

1. Structure of gelsolin and gelsolin/F-actin

FIG. 10. A cartoon representating the cooperative roles of ADF/ Gelsolin is an 80-kDa protein consisting of two tan-
cofilin (red crescents) and profilin (blue rhomboids) in regulating the dem homologous halves (segments 1–3 and 4 – 6), each
turnover of actin monomers (yellow) in a microfilament containing ATP
and ADP. Cofilin accelerates the dissociation of monomers from the
containing threefold repeats (Fig. 11). Its structure in the
pointed ends of filaments. Phosphorylation of cofilin (dark red) disso- absence of Ca2" has been determined at 2.5-Å resolution
ciates it from ADP-actin and profilin promotes the exchange of ADP for (35). The isolated NH2-terminal half can bind to (cap) two
ATP that facilitates the addition of profilin-ATP-actin at the barbed end.
actin monomers and can sever F-actin without the need
for free Ca2". In contrast, the COOH-terminal half binds
a single actin and is Ca2" dependent. The molecule is
ering a monomer to the growing end of the filament, it
compact and globular having dimensions of !85 # 36
must either move to a new site that does not block
# 55 Å. Segment 1 cocrystallized with monomeric actin is
assembly or completely dissociate. This can occur even
shown in Figure 12A. It has two bound Ca2" (186). Gel-
when actin is complexed to thymosin (see below) be-
solin segments 4 – 6 also bind to monomeric actin, and
cause of the high exchange rates for both thymosin and
their relationship to the actin monomer is illustrated in
profilin (219). It is not clear why some cells employ thy-
Figure 12B.
mosin to buffer actin monomers while others use profilin.
McGough et al. (181) had earlier suggested a mech-
Profilin binds only very weakly to the pointed ends of
anism in which segments 2 and 3 bind to the actin filament
filaments. Dissociation of profilin from actin is stimulated
while segment 1 wedges itself between two adjacent actin
by PIP and PIP2 (108); thus profilin may play a role in
monomers along the longitudinal axis. This step requires
transmitting signals between the cell membrane and the
a rearrangement of the interface between segments 1 and
actin cytoskeleton. Profilin binds to Arp2 in the Arp2/3
2 to lengthen the linker between these segments. Seg-
complex (see sect. IIIG).
ments 4 – 6 reach across the filament to bind a monomer
These functions are schematically illustrated in Fig-
in the other strand. This step would also require extension
ure 10. Here profilin is shown as blue rhomboids that bind
of the convoluted linker between segments 3 and 4.
to the barbed end of F-actin. The blue arrows indicate the
The following year, on the basis of crystallographic
cycle of profilin-actin interactions.
data, Robinson et al. (243) concluded that Ca2" binding
induces a conformational rearrangement in which seg-
C. Gelsolin Superfamily ment 6 is flipped over and translated by !40 Å relative to
segments 4 and 5. The structural reorganization tears
Gelsolin belongs to a superfamily of ABPs expressed apart the continuous !-sheet core of segments 4 and 6.
in all eukaryotes. The grouping includes (but is not lim- This exposes the actin-binding site on segment 4, enabling
ited to) gelsolin, villin, adseverin (also known as scind- severing and capping of actin filaments to proceed (Fig.
erin), CapG, flil, and severin (aka fragmin). All contain at 13). Severing of F-actin occurs only when sufficient actin-
least one of the 120-amino acid structural repeats and actin bonds are broken (36). The severing of F-actin by
many have three or six gelsolin repeats (157). For exam- gelsolin is illustrated in Figure 7C where only short seg-
ple, villin (92.5 kDa) is a six-domain ABP that regulates ments of actin filaments are observed in the presence of
actin assembly in microvilli. It has a Ca2"-dependent NH2- recombinant gelsolin (electron micrograph courtesy of

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


ACTIN AND ACTIN BINDING PROTEINS 451

FIG. 11. A ribbon diagram of the


atomic structure of inactive gelsolin
showing segment 1 (red) (containing the
NH2 terminus), segments 2 and 3 (green,
yellow), and segments 4 – 6 (magenta,
green, gold). (Figure kindly provided by
Bob Robinson.)

Roger Craig, University of Massachusetts Medical Center, 2. Functional properties in vitro


Worcester, MA). Here the molar ratio of gelsolin (seg-
ments 1–3) to actin is 0.05:1. A) ACTIN SEVERING. Gelsolin certainly lives up to its

The crystal structure of the F-actin binding domain of name. It rapidly dissolves actin gels in vitro by severing,
severin, the gelsolin homolog from Dictyostelium discoi- capping, and nucleating growth. Like cofilin, it can disso-
deum, has recently been described at high (1.75 Å) reso- ciate phalloidin from F-actin (262). These functions are
lution (243). The structure reveals an "-helix-!-sheet achieved by attacking filaments from their barbed ends.
sandwich that contains two cation-binding sites and F- Gelsolin is an important ABP because of its high affinity
actin binding residues in the equivalent region of gelsolin for actin filaments (Kd 50 nM) (229).
segment 2 (234). This segment also mediates the binding Segments 1–3 and 4 – 6 of intact gelsolin both play an
of gelsolin to tropomyosin (172). active role in the severing of F-actin. Binding of gelsolin

FIG. 12. A: crystal ribbon structure of


gelsolin segment 1 (red) and actin. B:
gelsolin segments 4, 5, and 6 and actin.
These figures are shown in the same ori-
entation as Figs. 5B and 9A. Colors are
those used in Fig. 11. (Figures kindly pro-
vided by Bob Robinson.)

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


452 DOS REMEDIOS ET AL.

trigger for events that involve the cytoskeleton, and gel-


solin is also the only known Ca2"-dependent severing
protein with a Kd for Ca2" in the micromolar range. The
six motifs contain two sites that bind to G-actin and one
that binds to filaments. The Ca2"-dependent monomer-
binding site in the COOH-terminal half plays a critical role
both in the cooperative binding to actin by gelsolin and in
its nucleating activity. Actin binding was localized to seg-
ment 4 by expressing segments 4, 4 –5, 5, and 5– 6 in
Escherichia coli (232). The nonbinding segments play an
important role in the Ca2" regulation of actin binding and
other activities of gelsolin (232). The segmental arrange-
ments are shown in Figure 10.
An analysis of the gelsolin construct lacking the
FIG. 13. A cartoon representation of three monomers in a filament
showing a possible rearrangement of gelsolin segments involved in COOH-terminal helix indicated the importance of this
cleavage of the actin filament. part of molecule for Ca2" regulation and supported the
so-called “tail helix latch” mechanism of activation by
Ca2" (170). This latch is primarily responsible for trans-
initiates the severing, but while binding to filaments is mitting Ca2" binding from the COOH-terminal half to the
rapid, its severing action is slow. Segments 4 – 6 confer a NH2-terminal half of gelsolin. Occupancy of this site
Ca2" sensitivity in the micromolar range, and binding primes gelsolin for severing, but it is unable to sever
involves a cooperative intramolecular structural rear- because the tail latch is still engaged. Unlatching the tail
rangement within gelsolin (229). This can be visualized by helix by micromolar Ca2" releases the final constraint to
comparing the arrangement of segments 4 – 6 of inactive initiate the severing cascade (163).
gelsolin (Fig. 11) with the active form shown in Figure
12B. A cartoon of active gelsolin segments on F-actin is 3. Comparison with ADF/cofilin
shown in Figure 13, which was redrawn from a figure
ADF/cofilin competes with gelsolin segments 2 and 3
kindly provided by Bob Robinson (University of Uppsala,
for binding to actin filaments. This suggests they share a
Uppsala, Sweden). Severing is effective even in the pres-
common structural topology for binding to actin (290),
ence of tropomyosin and has been used by several inves-
and this was subsequently confirmed by molecular dy-
tigators to solubilize actin filaments from skeletal muscle
namics simulation (304). Both proteins alter the confor-
sarcomeres.
mation of the actin filament, dissociate phalloidin, sever,
B) CAPPING. Not all authors agree that gelsolin is prin-
and cap F-actin. They both bind to ADP-G-actin with
cipally a severing protein. Gremm and Wegner (112) ar-
higher affinity than to ATP-G-actin, and the binding of
gued that at micromolar and submicromolar free Ca2"
both to actin is inhibited by PIP2 (275). Recent expression
concentrations, gelsolin acts more as a Ca2"-regulated
studies suggest both proteins play an important role in
capping protein. They observed that depolymerization
regulating actin dynamics in vivo (268). However, despite
was mainly caused by dissociation of actin subunits from
the more powerful severing action of gelsolin, ADF/cofilin
the pointed ends of actin filaments and that inhibition of
is currently favored as the major regulatory mechanism of
polymerization was achieved by gelsolin at barbed ends.
actin cytoskeleton reorganization (16).
The rate of capping at the barbed end was directly pro-
portional to the free Ca2" concentration (112).
4. Knock-out and overexpression
After severing, gelsolin remains attached to the
barbed end of the filament as a cap, thereby preventing Transgenic gelsolin-null mice have normal embry-
the reannealing of actin fragments. Capping prevents onic development and longevity but suffer subtle changes
rapid growth at the barbed ends while disassembly pro- suggesting that it is needed for rapid changes in cell
ceeds unchecked at the pointed ends, resulting in the motility in hemostasis (reduced platelet function), inflam-
rapid disassembly until an equilibrium is established be- mation (delayed neutrophil migration), and wound heal-
tween capped filaments and free gelsolin (239). Dissocia- ing (fibroblasts move more slowly) (303). In these null
tion (uncapping) of gelsolin from these filaments gener- mice, gelsolin is an essential effector of rac-mediated
ates numerous polymerization-competent ends that seed actin dynamics, acting downstream of rac recruitment to
the rapid assembly of new filaments. Thus the cytoskele- the membrane (10). Gelsolin also appears to play a role in
ton can be rapidly rebuilt in a new direction and cell the initiation of filopodial retraction. An absence of gel-
movement can be redirected (66). solin delays its retraction and produces “studded” filopo-
2"
C) CALCIUM DEPENDENCE. Ca is an important cellular dia along growing neurites (169).

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


ACTIN AND ACTIN BINDING PROTEINS 453

Overexpression of gelsolin in cultured mouse fibro-


blasts enhances their migration rate by !125% via gelso-
lin-promoted actin assembly and disassembly (72). Con-
trol of the rate of cell migration is closely linked to the
rates of actin filament severing. In fibroblasts that over-
express gelsolin, there is a direct correlation between
movement velocity and actin turnover rates. This demon-
strates that gelsolin is an important control mechanism
for actin cycling in cells (184). Overexpression of gelsolin
in cells also inhibits phospholipase C-# activity by com-
peting with it for available PIP2. A rise in free Ca2"
increases the affinity of gelsolin for PIP2, and as the level
and availability of PIP2 changes during signaling, cross-
talk between PIP2-regulated proteins (e.g., gelsolin and
phospholipase C-#) provides a mechanism for positive or
negative regulation of the signal transduction cascade
(274).

D. Thymosins

Thymosin !4 (T!4) is a small (molecular mass !5 FIG. 14. A modeled complex of thymosin !4 and an actin monomer
(78) using an orientation of the actin monomer comparable to Figs. 5A,
kDa) protein that belongs to a family of highly conserved 9A, and 12. No crystal structure is available for this complex.
small proteins originally isolated from the thymus gland
but now known to be more widely distributed (93). It
contains 43 residues, !50% of which are charged. The bind to the NH2-terminal region of actin and have char-
amino acid composition of T!4 is remarkably deficient in acteristic temperature-dependent conformations (75).
hydrophobic residues (168), which suggests that T!4 These data indicate that many parts of T!4 are not in tight
probably has very little defined structure in solution. contact with actin. Similarly, circular dichroism spectra
indicate that free T!4 is predominantly unstructured but
1. Structure increases its "-helical content when it binds to actin. At
least one research group has made progress with a crystal
Until very recently, the structure of T!4 was based
structure of T!4. De La Cruz et al. (78) recently concluded
on a model constructed by Dan Safer et al. (251) who
that T!4 changes the conformation and structural dynam-
docked it to the crystal structure of monomeric actin
ics (the so-called “breathing”) of actin monomers. The
solved in the presence of DNase I (Fig. 14). These authors
conformational change may reflect the unique ability of
regard this protein as having an essentially unfolded con-
T!4 to sequester actin monomers and inhibit nucleotide
formation in solution and, although there may be an in-
exchange (compared with profilin).
crease in its "-helical content, it remains unfolded once it
binds to actin (251). Based on a range of data (250), it has
2. Cellular distribution and functions
been suggested that T!4 drapes itself around nearly half
the perimeter of the actin. The molecular contact points !-Thymosins are widely distributed in nature. Trans-
between T!4 and the actin monomer investigated using lational levels can vary from tissue to tissue (262). In
zero-length cross-linkers show that Lys-3 of T!4 is located chick brain, concentrations are an order of magnitude
close to Glu-167 (in subdomain 3 near the hinge), and that higher than profilin and cofilin. It is abundant in neural
Lys-18 of T!4 cross-links to one of the four NH2-terminal tissue as well as in circulating cells such as platelets,
acidic residues of actin (Asp-1 through Glu-4). These two leukocytes, and macrophages. T!4 binds stoichiometri-
contacts flank the "-helical region of T!4 at the barbed cally to monomeric actin and inhibits polymerization. In
end of the monomer. The COOH-terminal region of T!4 the developing chick brain, the ratio of ADF to cofilin to
(Lys-38) was cross-linked to Gln-41 at the top of actin T!4 increases between embryonic days 13 and 17. Be-
subdomain 2, a known DNase I binding site. cause no other thymosin isoforms were detected in brain
The NMR structure of free T!4 in solution (75) ex- extracts, it is probable that it is the major actin-seques-
hibits a range of conformational preferences. In water it tering protein in this tissue (84). T!4 mRNA is mainly
appears to lack a folded conformation, but a preferential expressed in neurons of the hippocampus, neocortex, and
"-helical conformation can be observed in solution at the amygdala, as well as in oligodendrocytes and probably
different temperatures and pH. Residues 5–16 putatively has a function in the production and remodeling of neu-

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


454 DOS REMEDIOS ET AL.

ronal processes (48). T!4 is upregulated in the pyramidal trostatic contact site with actin. A nonconserved NH2-
neurons of the hippocampus where its function may be terminal segment preceding the motif exerts an inhibitory
related to restoration of neurite circuits after focal isch- activity on actin polymerization probably by steric hin-
emic damage (292). drance. Substitution of a His or Lys for a Glu at the third
T!4 is present in some cells at concentrations as high position in the motif converts the peptide from an inhib-
as 0.5 mM (249). High expression levels have been ob- itor of polymerization into a stimulator. Thus subtle se-
served in young embryonic muscles but not in adult skel- quence changes in ABPs can have remarkable effects on
etal muscles. Consequently, it seems that the G-actin pool their functions.
in young embryonic skeletal muscle is mainly regulated
by T!4 and profilin (206). However, T!4 may become less 3. Molecular interaction with actin
important as muscle develops (199). In contrast, there is
evidence that it plays a role in angiogenesis in coronary Binding of T!4 to actin is coupled to the dissociation
smooth muscle, acting as a chemoattractant by stimulat- of bound water molecules, which is greater for CaATP-
ing the migration of endothelial cells in vivo (174). actin than MgATP-actin monomers. The COOH terminus
In quiescent NIH 3T3 cells, T!4 was undetectable, of T!4 makes contact with actin at His-40 near the mouth
but after restoring serum to the culture medium where of the nucleotide cleft and binding slows the 3H exchange
cells were induced to proliferate, there was a pronounced rate, suggesting there are changes in conformation of
increase in T!4 mRNA (316). Overexpression in these actin monomers when it binds. This may reflect the ability
cells increases F-actin levels while maintaining a constant of T!4 to both sequester actin monomers and inhibit
level of total actin (273). In contrast, others reported that nucleotide exchange (78).
a twofold increase in T!4 in these same cells doubled the Gelsolin and T!4 compete for binding to actin, indi-
total actin expression but maintained the ratio of G- to cating that T!4 binds to a site close or identical to the
F-actin (110). Unexpectedly, other cytoskeletal proteins gelsolin segment 1-binding site (13). Figure 6 shows that
such myosin IIA, "-actinin, and tropomyosin also in- the binding loci for gelsolin and T!4 do not precisely
creased nearly twofold (273). The T!4 cell lines spread coincide but that there must be sufficient overlap to ex-
more fully and adhered to the dish more strongly than plain the competition. If T!4 is to be a major cellular
vector controls, suggesting a coordinated regulation of actin-sequestering factor, it should bind, among other
the cytoskeleton by T!4. In addition, transcriptional lev- things, with a slightly higher affinity (0.7 $M) than gelsolin
els of several related cytoskeletal proteins also increased. (!1 $M) (262), thus subtly shifting the assembly/disas-
These observations suggest there is a coordinated cy- sembly equilibrium of actin toward the monomeric state.
toskeletal expression of this protein (110). The interactions of these two proteins with actin are
Like several other ABPs, thymosins bind to and co- complex and require the participation of several comple-
ordinately control the availability of actin monomers for mentary peptide sequences. A common motif (I,V)EKFD
incorporation into filaments. T!4 is a contender with in these two proteins exerts a major inhibitory effect on
profilin for the title of the universal actin-monomer bind- salt-induced actin polymerization (96).
ing protein. It has a putative actin-binding motif To make matters more complex, T!4 and profilin
(LKHAET) (287) found in other ABPs including acto- (and presumably gelsolin) compete with WASP (an acti-
bindin and myosin. Nachmias (196), and more recently vator of Arp2/3) for binding to actin monomers and pre-
Chen et al. (54), reviewed the discovery and primary sumably have overlapping binding sites (see sect. IIIG).
structure of T!4. DNase I and T!4 bind to sites on actin that partially
Acting alone, T!4 inhibits polymerization, and like overlap. The apparent Kd of the actin-T!4 complex gen-
ADF/cofilin it also inhibits the exchange of the actin- erated from a preformed actin-DNase I complex is 160
bound nucleotide (107). This contrasts with the action of $M. A fivefold excess of DNase I over the preformed
profilin which, acting alone, facilitates the exchange of actin-T!4 complex is necessary to observe a comparable
ADP for ATP and promotes actin assembly (109). Thus, Kd (129). Thus DNase I can displace T!4 from the mono-
while it acts as a passive buffer of actin monomers, pre- mer (and probably vice versa).
venting their polymerization, T!4 can cooperate with pro- Chemical cross-linking experiments also indicate
filin to regulate actin assembly (273). However, T!4 is not that profilin competes with T!4 for actin binding. In the
a simple G-actin-sequestering protein because, at least at presence of Mg2", profilin promotes assembly of T!4-
high concentrations, it also interacts with F-actin (40). actin but in the presence of Ca2", profilin was unable to
The actin-binding sites of T!4 resemble that of an- initiate this polymerization (12). Phalloidin, rabbit skele-
other small ABP, actobindin, which also inhibits polymer- tal muscle myosin S1, and chicken intestinal myosin I all
ization. Their binding loci on monomeric actin have been polymerize actin from the T!4-actin complex (15). Thus,
localized using peptide mimetics (286). Both have an despite the very high affinity of DNase I for actin, its
NH2-terminal hexapeptide motif forming the major elec- binding can be inhibited by a lower affinity T!4. The

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


ACTIN AND ACTIN BINDING PROTEINS 455

explanation for this may be an allosteric conformational diated by G-actin sequestering and involve interaction
change in the actin monomer as reported by us recently with DNase I (114).
(81) for DNase I and cofilin. NIH 3T3 cells that overexpress (x3) T!10 have been
Competitive binding data suggest that T!4 also shown to have 23–33% more polymerized actin than con-
makes contact with the myosin binding site as well as trol cells, and their microfilaments appear thicker. There
with the bottom of subdomain 3 where it overlaps the is no change in total actin, profilin, or cofilin/ADF content.
binding sites of several ABPs including DBP (Fig. 6, row These cells are more motile, spread faster, and have
3) at the barbed end of the monomer. Therefore, the higher chemotactic and wound healing activity (273). Im-
distance between the pointed and barbed ends requires munofluorescent labeling of peritoneal macrophages
the COOH-terminal half of T!4 to be in a predominantly showed that both T!4 and T!10 were uniformly distrib-
extended conformation (247). Point mutations revealed uted within the cytoplasm. Overexpression in fibroblasts
that the middle section of T!4 (residues 17–22) also in- induces extensive loss of stress fibers, indicating that
teracts with actin (267). T!4 and T!10 are functionally similar and that both are
Although T!4 is widely recognized as a monomer effective regulators of actin filament dynamics in living
binding protein, at least three reports have demonstrated cells (313).
its ability to bind to F-actin (14, 40, 273). Addition of 200 Both T!4 and T!10 are elevated in a human prostate
$M T!4 during actin assembly decreases the critical con- cancer cell line and correlate positively with the grade of
centration of free actin. This is consistent with the in vivo tumor. Transfection with antisense T!15 constructs into
observation that overexpression of T!4 increases stress rat prostatic carcinoma cells demonstrates that the T!4
fibers and decreases the amount of monomeric actin peptide positively regulates cell motility. Increased motil-
(273). A chemically cross-linked binary complex of T!4 ity is a critical component of the metastatic pathway (22).
actin is incorporated into F-actin only when both phalloi- T!15 is also upregulated in nonmetastatic breast cancer
din and gelsolin are present (14). T!4 increases the long- and even in ductal carcinoma (compared with benign
pitch helical repeat (from 35 to 40.5 nm determined from breast tissue). Consequently, it might represent a poten-
3-dimensional helical reconstructions) of F-actin. Like its tial early marker for breast malignancy (106).
interaction with actin monomers, T!4 does not “dock” T!9 and T!met9 (the minor variants of T!4) inhibit
neatly onto the F-actin model, particularly at the NH2- polymerization of ATP-actin with identical Kd values (0.7–
terminal region of actin subdomain 3. This difference may 0.8 $M). Like T!4, these thymosins bind to ADP-G-actin
well be due to conformational changes in both proteins as with a 100-fold lower affinity than to ATP-G-actin (134).
suggested recently by at least two authors (14, 78).
E. DNase I
4. Other thymosins
DNase I is a secretory glycoprotein with an endonu-
Thymosin isoforms are distributed throughout the
clease activity that cleaves double-stranded DNA to yield
vertebrate phyla (247). While T!4 is the predominant
5'-phosphorylated polynucleotides. It has a requirement
expressed isoform in mammalian cells, many species pro-
for both Ca2" and Mg2". Bovine pancreatic DNase I is the
duce at least two T! isoforms, in some cases in the same
best characterized of a group of DNases. It has a molec-
cell (49). Differences in the distributions of the mRNAs
ular mass of 31 kDa and an optimum pH of 7.8 (152). It
encoding T!10 and T!4 suggest there are different roles
was first purified from pancreas and parotid gland that
for these peptides during mouse early embryogenesis
contained four DNases (A, B, C, and D), differing accord-
(49). The levels of T!10 mRNA strongly increase during
ing to their carbohydrate side chain and polypeptide com-
early postimplantation mouse embryogenesis, suggesting
ponents. DNase I is glycosylated at Asn-18 (189). These
that T!10 plays an important role in early development.
factors, together with the presence of sialic acid, result in
Changes in T!10 and T!4 mRNA expression are also
a charge heterogeneity that can be detected by isoelectric
important in the control of actin dynamics in developed
focusing (162).
neurons and glia since T!10 mRNA is present in different
regions of the rat forebrain, including hippocampus, neo-
1. Atomic structure
cortex, and several brain nuclei (48). T!10 is also down-
regulated in failing human hearts with dilated cardiomy- The structure of DNase I was solved crystallographi-
opathy (Tsubakihara, unpublished data). cally (272) at 2-Å resolution and was found to have two
The underlying mechanism for these cellular changes Ca2" binding sites. DNase I has a high affinity for actin
remains to be elucidated. However, plasmid-driven over- (Kd & 5 # 108 M$1) (177). The crystal structure of the
expression of the T!10 gene is known to increase the actin-DNase I complex (with actin in the ATP form) has
susceptibility of apoptosis in transfected fibroblasts (115). been determined at 2.8-Å resolution (137). Most ABPs
These proapoptotic properties of thymosins may be me- interact with actin subdomain I, whereas DNase I binds to

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


456 DOS REMEDIOS ET AL.

levels in cells (71), although its function in cytoskeletal


dynamics is yet to be elucidated. However, the tight bind-
ing and specificity of DNase I for actin and its wide
distribution suggests it has a biological role. It binds very
tightly to G-actin (Kd in the nM range) (266) and essen-
tially removes all free actin monomers from solution. It
binds much more weakly to F-actin (100 $M) where it is
a capping protein that increases the rate of dissociation
from the pointed ends of filaments (298).
We recently examined the possible cellular codistri-
bution of DNase I and cofilin in cultured mammalian cells.
Using fluorescent-labeled antibodies to cofilin and DNase
I, we found the cytoplasmic distributions of these two
proteins overlap but are not identical (56). Like T!4,
DNase I may act as a stabilizer of actin monomers by
effectively removing them from the available pool of
monomers available for assembly. In section IV we raise
the possibility that it may also be a natural modulator of
the action of cofilin on the cytoskeleton.

F. Capping Proteins

Monomers in actin filaments “age” as they progress


FIG.15. This ribbon structure of DNase I (red) and actin is derived from the assembling (barbed) end to the disassembling
from the 1990 2.8-Å crystal structure of the complex (137). The orien-
tation is essentially the same as Figs. 5A, 9A, 12, and 14. Note the
(pointed) end of a filament. This aging process is driven
contacts for DNase I are located in subdomain 2. by the chemical energy of Mg-ATP. Free monomers pre-
dominantly contain bound ATP that is hydrolyzed to
ADP-Pi only after incorporation at the barbed end, with
the part corresponding to the pointed end of the actin the subsequent release of Pi. At the barbed end, the on
(298). The atomic structure of rabbit skeletal muscle actin rate for ATP monomers (5 $M/s) exceeds the off rate (1
complexed with bovine DNase I is shown in Figure 15. $M/s) while for ADP monomers the corresponding on
rate (1 $M/s) is slower than the off rate (6 $M/s). Thus,
2. Cellular functions while monomers at the barbed end nearly all contain ATP,
monomers at the pointed end mostly contain ADP. At the
DNase I is an enzyme secreted by exocrine glands pointed end, the on rate for ATP monomers (0.5 $M/s) is
such as the pancreas and the parotid into the alimentary slower than its off rate (1 $M/s) while the off rate for ADP
tract. However, it has a much wider tissue distribution. (0.2 $M/s) is faster than its on rate (0.1 $M/s) (149).
This suggests it may play roles other than to hydrolyze Because the rates of monomer exchange are 20 times
DNA. DNase I, as well as other DNases, has been impli- faster at the barbed than at the pointed end, there is a high
cated in apoptosis (223) where it hydrolyzes DNA into probability that an ATP monomer that adds to this end
short fragments (DNA laddering). will be hydrolyzed before another exchange occurs.
Lazarides and Lindberg (159) discovered that actin Clearly, capping proteins at these ends will alter mono-
was a naturally occurring inhibitor of DNase I, and they mer kinetics depending on their binding affinity and rate
were the first to suggest that DNase I may be a cytoskel- of exchange.
etal protein. They proposed three possible biological roles Many, perhaps most, cells control the length and
for this interaction: 1) actin controls the nucleolytic ac- number distribution of thin filaments under conditions
tivity of this enzyme during the cell cycle, 2) the complex where total actin concentration remains approximately
has a specific function in DNA metabolism, and 3) the constant. This assembly and disassembly of actin requires
primary function of DNase I is related to the formation a dynamic equilibrium between actin monomers and the
and function of actin filaments rather than to the degra- two, nonequivalent ends of actin filaments. Proteins that
dation of DNA. The fact that phosphoinositides can dis- cap or prevent this exchange may variously nucleate,
sociate the binary actin-DNase I complex (309) also sug- promote, stabilize, or inhibit assembly. They are therefore
gests that this interaction may have biological relevance. in a strong position to regulate actin polymerization. This
DNase I has been useful tool in measuring G-actin simple concept is illustrated in Figure 16. Here we focus

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


ACTIN AND ACTIN BINDING PROTEINS 457

assembly at the barbed ends of filaments, 4) elimination


of annealing at the barbed ends of filaments (265), and 5)
correct assembly of filaments at the Z-disk (254). CapZ
does not affect the rate of fragmentation of actin filaments
and does not bind to their pointed ends (53). Figure 16
shows CapZ at the barbed end of a filament.
CapZ is a heterodimer comprising an "-subunit
(there are at least 2 isoforms in eukaryotic cells) and a
!-subunit (2 or 3 isoforms), and both subunits are re-
quired for effective barbed-end capping of filaments (53).
The "1- and "2-isoforms result from alternative gene ex-
pression (117). Capping proteins containing the "1-iso-
form display fourfold higher affinity for actin than capping
protein containing the "2-isoform, although the physiolog-
ical significance of this remains unclear. The !1- and
!2-isoforms result from alternative splicing of a single
mRNA transcript (255), and although both isoforms dem-
onstrate similar affinity for actin, they are differentially
expressed and distributed. The !1-isoform is predomi-
nantly expressed in muscle cells and is localized to the
Z-line in striated muscle, whereas the !2 is located under
FIG. 16. A cartoon representation of the interaction of a barbed-end the sarcolemma (51). Consequently, capping protein con-
(CapZ, purple) and a pointed-end capping protein, tropomodulin taining the !1-isoform is often referred to as CapZ and is
(brown) (Tmod). One CapZ binds to two monomers at the barbed end of analogous to !-actinin (179). The !2-isoform is predomi-
an actin (yellow) filament. The NH2 terminus of Tmod binds tightly to
NH2 terminus of tropomyosin (green) at the pointed end of filaments and nantly expressed in nonmuscle cells where it is localized
also to one (or two) monomers. The on rate and off rates at the filament to the cell periphery (255).
ends are indicated by lengths of the arrows. Two views of the Tmod B) STRUCTURE OF CAPZ. CapZ was recently crystallized
molecule [originally published in Krieger et al. (153)], modeled from
X-ray scattering data, are shown at the top of this figure. by Yuchiro Maeda and colleagues, and a preliminary re-
port was presented at the SPring8 actin conference held
in Japan in November 2001. Details of this structure are
on the most abundant barbed-end capping protein CapZ not yet available but are likely to be published later in
and the predominant pointed-end capping protein tropo- 2003. Despite the fact there are no cocrystals of CapZ-
modulin. actin, considerable detail is available on the molecular
contacts between these proteins. The contact sites on
1. CapZ, a barbed-end capping protein actin are illustrated in Figure 6, row 4. Deletion of the
COOH terminus of the !-subunit of CapZ eliminates its
Several different proteins can bind to the barbed end capping activity as does the binding of a monoclonal
of filaments. Some of them like gelsolin (and its close antibody directed against the COOH-terminal 12 residues
relatives adseverin and fragmin) (254) and profilin have (130). Later, this same group showed that 31– 40%, respec-
already been discussed above in a different context. Villin, tively, of the "- and !-subunits are not required for strong
CapG, tensin, and fragmin have been reviewed else- binding to actin, suggesting that these regions may be
where (254). important for other functions (53).
CapZ is present in all eukaryotic cells (65). In skeletal As its name suggests, capping protein binds tightly to
muscle it is localized at the Z-disk where it is probably the barbed end of actin filaments. In vitro, this capping
held onto the Z-filaments by a unique protein (see review prevents the exchange of actin subunits from the barbed
in Ref. 165). It is usually prepared from chicken or rabbit end. Furthermore, capping protein facilitates actin poly-
skeletal muscle but, unless G-actin is twice purified by gel merization by binding to and stabilizing monomers or
filtration, traces of CapZ remain in the preparation and oligomers of actin forming nuclei for elongation of fila-
affect actin assembly characteristics (52). The difficulty in ments (51).
preparing actin free of CapZ can be appreciated when one C) REGULATION OF CAPZ ACTIVITY. In vivo, the capping of
realizes that one heterodimer of CapZ can cap a filament actin filaments is regulated by second messengers, PIP
containing a thousand or more actin monomers (265). and PIP2. This is also true for other ABPs including cofilin,
A) FUNCTIONS OF CAPZ. The reported biological roles of destrin and DNase I (309), profilin (108), and gelsolin
CapZ are as follows: 1) nucleation of actin assembly, 2) (285) (see Table 1). Upon signal transduction, PIP and
capture of preexisting filaments, 3) regulation of actin PIP2 promote removal of capping protein from actin fila-

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


458 DOS REMEDIOS ET AL.

ments, resulting in localized pockets containing a dra- i.e., at their respective ends. Blocking barbed ends with
matic increase in the number of free barbed ends. This cytochalasin D had no effect on filament length, whereas
allows actin polymerization to proceed in localized re- overexpression of Tmod reduced actin filament length,
gions of the cell. Capping protein is involved in a variety possibly because Tmod promotes the conversion of ATP
of biological roles including the activation of platelets and or ADP-Pi monomers to ADP with a consequent raising of
maintaining the proper assembly of actin filaments at the the critical concentration from 0.1 to 1.0 $M (241).
Z-line of sarcomere (254).
G. The Arp2/3 Complex
2. Tropomodulin, a pointed-end capping protein
Arp2 and Arp3 belong to a recently discovered family
Unlike barbed-end capping proteins, pointed-end
of proteins that are related, at least in a limited way, to the
capping proteins are something of a rarity. The name
sequence and structure of actin. Their cellular concentra-
tropomodulin (Tmod) is apt because it binds strongly to
tions are 2 and 5 $M, respectively (171). These two actin-
actin in the presence of tropomyosin. With the exception
like proteins are assembled into the Arp2/3 complex,
of Arp2/3 (see sect. IIIG), no other protein binds to the
which is known to occur in a wide range of organisms
pointed end of F-actin, and no other protein inhibits fila-
from yeast to mammals. The seven-subunit protein com-
ment elongation at this end. Although Tmod binds to
plex comprises Arp2 and Arp3 and five smaller (Arc)
F-actin (it does not bind monomeric actin). Its affinity is
proteins that do not appear to be related to any proteins
weak (Kd !0.1 $M) compared with its affinity for tropo-
currently in the databases. Both Arp2 and Arp3 sequences
myosin-decorated actin filaments that is about three or-
have nucleotide-binding and divalent cation-binding re-
ders of magnitude stronger (Kd is submicromolar) (297).
gions, but otherwise they are divergent from each other
For this reason, some authors originally referred to it as a
and from actin (143). Arp2/3 is associated with dynamic
tropomyosin binding protein (295).
cortical regions of mammalian cells that are rich in cy-
Human Tmod has been cloned and is expressed in a
toskeletal actin (69). It colocalizes with the actin “comet
wide range of tissues with high levels in skeletal and
tails” of Listeria and Shigella where actin polymerization
cardiac muscle (276). It contains 359 amino acids and has
drives filament assembly and propels these intracellular
a molecular mass of 40,476 Da determined by mass spec-
bacterial pathogens through the cytoplasm of the host cell
trometry. Overexpression of Tmod in cardiomyocytes
at speeds of 10 –90 $m/min (302).
produces short, disorganized, and destabilized thin fila-
ments (277). A Tmod overexpressing transgenic mouse is
1. Structure
now available (278). A related, somewhat large (64 kDa)
protein, leiomodin, is highly expressed in smooth as well The structure of the Arp2/3 complex at 2.0-Å resolu-
as cardiac and skeletal muscle (62). tion was only recently described (244). It is remarkable
Yuichiro Maeda’s group (at the RIKEN SPring8 Insti- that a complex this big crystallized at all. Proteins were
tute, Japan) has reported crystals of the actin-binding extracted from bovine thymus gland, and the isolated
COOH-terminal half of Tmod that diffract to 1.45 Å (153). complex was in an inactive state when it was crystallized.
This structure is now in press in the Biophysical Journal. Conversion to an active form is a matter of speculation
The overall shape of a slightly truncated Tmod has been but will probably involve a rearrangement of several sub-
determined from low-angle X-ray scattering by Fujisawa units to allow the Arp2 and Arp3 to be repositioned so
et al. (101), also in Maeda’s group. Tmod is 115 Å long and that they can act as nucleation site for the assembly of
consists of two domains, one 65 Å long, the other slightly new filaments. This rearrangement will also be needed if
smaller, where the long axes of the two are tilted by !40° the Arp2/3 complex is to “capture” preexisting filaments
relative to each other and is illustrated in Figure 16. The and incorporate them with the correct orientation toward
authors proposed a model for Tmod in which the rod- the advancing cell membrane. The structure is now de-
shaped NH2-terminal half binds to the NH2 terminus of posited in the Protein Data Bank (PDB accession number
tropomyosin, and the COOH-terminal region protrudes 1k8k).
from the pointed end of the actin filament and is slightly Briefly, the structure has the following features. The
bent toward the actin subunit at the end, capping the overall shape of the complex is a flat ellipsoid with di-
pointed end (101). There appears to be 1–2 Tmod mole- mensions of 150 # 140 # 70–100 Å. The structures of Arp2
cules/actin filament (98). and Arp3 strongly resemble the actin monomer. Arp2 has
Both CapZ and Tmod can regulate filament length a strong resemblance to subdomains 1 and 2 of actin
and assembly in vitro and in vivo. By studying the incor- except that the interdomain cleft is more open (by !18°)
poration of rhodamine-labeled actin monomers into na- and contains no bound ATP. It also has a profilin binding
tive thin filaments of cardiomyocytes, Littlefield et al. domain (143). Arp3 contains large inserts into subdo-
(164) concluded that both proteins can rapidly exchange, mains 2, 3, and 4 and has 42 residues more than actin. The

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


ACTIN AND ACTIN BINDING PROTEINS 459

cleft is less open (!12°), a feature that lowers its affinity Wiskott-Aldrich syndrome protein (WASP), the neuronal
for ATP. Subdomains 1 and 2 were difficult to resolve in homolog of WASP (N-WASP), or suppressor of cAMP
the crystals, and the authors replaced them with a poly- receptor 1 (Scar1). PIP2 may be also involved in the
alanine backbone. These and other features give the Arps activation of WASP (242). Activated WASP has sites that
distinctive surface features. In mammalian cells, the five bind both an actin monomer and the Arp2/3 complex.
associated proteins are p16, p20, p21, p34, and p40 (pro- Arp2/3 then changes its conformation so that it can nu-
teins are also called ARPC-5, -4, -3, -2, and -1, respectively) cleate the assembly of new actin filaments and assemble
of which the last is a seven-blade propeller structure with a “daughter” filament at an angle of !70° to the “mother”
a large insert that may be responsible, along with p20 and filament (Fig. 17). As a consequence, Arp2/3 is localized at
p21, for anchoring the complex to the side of the actin branch points in the cortical filament network, binding
filament. p20 and p34 are located in the center of the the pointed end of a new (uncapped) daughter filament to
structure and may move as a block relative to Arp2 (244). the mother filament thus leaving the barbed (growing)
end available for filament elongation. It is not clear at this
2. Cellular functions stage whether WASP dissociates from activated Arp2/3 at
The principal function of Arp2/3 is to create branch the branch point.
points by nucleating the assembly of filaments near ruf- Zalevsky et al. (314) recently examined how N-WASP
fling membranes, a process that needs micromolar con- and Scar1 differentially activate the Arp2/3 complex to
centrations of ATP (195). Arp2/3 is also considered to be produce actin networks with different three-dimensional
a cross-linking protein, and some authors believe it can architectures. They found that nucleation was slowest for
cap the pointed (depolymerizing) ends of actin filaments, Scar1, 16-fold higher for WASP, and 70-fold higher still for
although this is still controversial. the neuronal analog N-WASP. Their data fit a mathemat-
Nucleation of filament assembly by Arp2/3, first ob- ical model in which one activated Arp2/3 complex, one
served by Welch et al. (302), is unlike any other ABP. actin monomer, and an actin filament combine into a
Extracellular signals such as growth factors are transmit- preactivation complex. This then undergoes a first-order
ted through G protein-coupled receptors (e.g., Cdc42) to activation step to become a nucleus. They went on to

FIG. 17. A cartoon of the molecular interactions Arp2/3. Components are as follows: activated Arp2/3 (red), inactive
Arp2/3 (pink), N-WASP (green), Cdc42 (blue), phosphatidylinositol 4,5-bisphosphate (PIP2) (yellow), and actin (orange).
Inactive Arp2/3 is activated by N-WASP under the control of Cdc42 and PIP2, and filament formation is nucleated. Actin
filaments assemble by addition of monomers at the barbed end. The N-WASP-Arp2/3 complex can bind to the side of the
“mother” filament to create a branch point that also polymerizes. This branching assembly creates an “out-pushing” of
the cell membrane. The branching of “daughter” filaments occurs at an angle of 70° to the mother filament. The solid
black arrows indicate the directions of the polymerization forces.

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


460 DOS REMEDIOS ET AL.

show that WASP and Scar1 bind actin and Arp2/3 with “comet tails” (the review by Pantaloni et al. is available at
nearly identical affinities but stimulate rates of actin nu- www.sciencemag.org/cgi/content/full/292/5521/1502/DC1).
cleation that vary by almost 100-fold. Differences in nu- The role of WASP/Scar proteins is critical for cyto-
cleation rates were attributed to differences in the num- plasmic organization in a range of organs during develop-
ber of COOH-terminal acidic amino acids that are respon- ment in Drosophila (315) and presumably Arp2/3 and the
sible for different rates of actin assembly. The authors WASP family of proteins play essential roles in the devel-
therefore concluded that Arp2/3 is not regulated by a opment of mammalian cells.
simple on-off switch. Several reviews have recently been published on the
In the vicinity of the ruffling membrane, new mono- role of Arp2/3 in actin polymerization and the factors that
mers are added to the barbed-end of the growing filament control the functions of Arp2/3. These include the role of
creating a polymerization-based “out-pushing” of the cell WASP family of proteins (50, 125, 242), and a review of
membrane (Fig. 17). The Y-shaped branch points produce the structure of Arp2/3 itself (31).
a cross-linked meshwork of filaments that act as a plat- Thus Arp2/3 is a complex protein with an even more
form against which the actin polymerization can push. complicated method of activation. Single-headed myosin I
The entire Arp2/3 complex along with WASP have been motor proteins interact with Arp2/3 via their Scar-homol-
observed in electron microscopic images of branching ogy 3 (SH3) domain (160), but what are motor protein
filaments (293). Mullins et al. (195) refer to this process of doing in a process that is usually regarded as an actin-only
branching filaments as “dendritic nucleation.” Assembly process? Certainly there is no evidence to suggest that
of monomers at the barbed ends of these filaments gen- Arp2/3 is involved in muscle contraction, although it
erates the force needed to push the ruffling membrane clearly plays a role in actin-based motility.
forward. Branching may occur at the barbed ends or Several other questions remain. What is the nature of
along the side of the mother filament (or both). Ishiwata the conformational changes needed to activate Arp2/3?
and co-workers (103) recently addressed this question by Can Arp2/3 promote filament growth at the pointed ends
the directly visualizing actin assembly using total internal of filaments? What is the role of profilin? What is the role
reflection fluorescence microscopy. In the presence of of ATP hydrolysis, and is this related to changes in the
Arp2/3 and N-WASP, branching occurred predominantly putative nucleotide cleft in Arp2 or Arp3? Does activation
along the sides of growing filaments. of filament nucleation produce filament growth at the
Nucleation and branching are carried out close to the barbed or pointed end? Why do branch points occur
leading edge of motile cells. Arp2/3 is not generally asso- where they do? How do Cdc42 and PIP2 cooperate to
ciated with filaments deeper into the cortex of the cell release the inhibitory effect of WASP on Arp2/3?
where the actin filaments are known to bind tropomyosin.
A recent report (28) demonstrated that tropomyosin lo-
cated deeper into the cytoplasmic cortex inhibits branch- IV. ROLE OF TERNARY COMPLEXES IN
ing and nucleation by WASP-activated Arp2/3, thus re- REGULATING ACTIN CYTOSKELETON
stricting its distribution to the cytoplasmic perimeter. ASSEMBLY
In Listeria, nucleation of new filaments occurs in
response to the expression of ActA that mimics the func- The foregoing discussion has probably created the
tion of WASP in recruiting Arp2/3. ActA contains 612 impression that regulation of assembly and disassembly
amino acids and is anchored into the bacterial membrane. of the actin cytoskeleton is complex. In fact, the complex-
In a fascinating series of experiments, Cameron et al. (39) ity may be comparable to the system that controls gene
showed that small microspheres coated with purified expression in eukaryotic genes where a promoter element
ActA could undergo movement generated by actin poly- contains the promoter-proximal upstream control ele-
merization. Thus ActA alone can nucleate actin assembly. ments as well as enhancers and silencer elements. A
The minimum set of proteins and conditions needed for further level of regulation is then provided by distal reg-
actin-based motility was recently defined for the propul- ulatory sequences. DNA-bound proteins that can alter the
sion of Listeria by Marie-France Carlier and colleagues state of condensation of the chromatin DNA also influ-
(166). ATP, Arp2/3, actin filaments with bound ADP, ADF/ ence these DNA sequences. The control of actin cytoskel-
cofilin, and a capping protein (e.g., CapZ) were all that eton assembly/disassembly may be even more complex.
was required. ADF/cofilin accelerates the treadmilling of Just as we do not have a precise understanding of the
monomers through the filaments (see above) and drives control of gene expression, we are also only beginning to
the Listeria through the actin-rich cytoplasm of its host understand the control of location and mechanism of
cell. A review from this group (220) highlighted the com- assembly of the actin cytoskeleton. Not only are there
plexity of the mechanism of actin assembly in cell biol- potentially multiple ABPs, but they are also capable of
ogy. Clearly other intracellular factors such as profilin and simultaneous expression within a single cell. The pres-
"-actinin regulate the rates of actin assembly in these ence of actin ligands such as ATP or ADP, Ca2" or Mg2"

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


ACTIN AND ACTIN BINDING PROTEINS 461

can also be important, as well as pH. Under these condi- that DNase I may act as a regulator of actin-cofilin inter-
tions, it would be surprising if control of the state of action and that the ternary complex may therefore be an
assembly was left to single ABPs as suggested in the important element in the regulation of microfilament as-
preceding sections. The next level of complexity to con- sembly (142), effectively removing cofilin from the cellu-
sider is whether ternary complexes play a role and if so lar pool and thereby preventing its normal ability to po-
how is this achieved. Of course, quaternary and higher lymerize. The ternary complex may be destabilized by
orders of complexity may also exist. There can be no other ABPs such as profilin or gelsolin, which alter the
argument that complexes of more than one ABP control bound nucleotide presumably via an allosteric conforma-
actin filament functions (see sect. IIIG). Activation by tional change. Such a regulatory mechanism would occur
Ca2" via the tropomyosin-troponin C/I/T complex is now on a slow time scale equivalent to the rate of appearance
well established, but there are also proteins that regulate and disappearance of DNase I and would depend on the
filament length, anchoring into the Z-disk of sarcomeres, balance of free ABPs, pH, and divalent cations present in
cap thin filaments, and connect actin to microtubules and the cell.
the intermediate filament system.

B. Ternary Complexes of Actin With Two


A. Cofilin, Actin, and DNase I or More ABPs

It is widely accepted that regulation of the rate of The actin-DNase I binary complex is unusual but not
assembly/disassembly (treadmilling) of the actin cy- unique in its ability to form a ternary complex with cofilin.
toskeleton is achieved by a single ABP, probably ADF/ Profilin can be chemically cross-linked to actin-DNase I;
cofilin (21). However, the widespread use of DNase I however, in the absence of cross-linking, a negative co-
affinity chromatography to capture complexes of actin operativity was observed (12), i.e., profilin dissociates
with other ABPs (cofilin, destrin, actobindin, adseverin, DNase I from actin, and vice versa. DNase I and gelsolin
fimbrin, and radixin) suggests that ternary complexes also bind to actin in a negatively cooperative way. Chem-
exist, at least in vitro (133). We recently reported the ical cross-linking of Cys residues between T!4 and actin-
formation of ternary complexes of actin, cofilin, and DNase also revealed the formation of a ternary complex
DNase I (57) using nondenaturing polyacrylamide gel (239). These authors found that the Kd of T!4 for actin-
electrophoresis. The affinity of DNase I for G-actin is high DNase I was the same as for actin alone (see Table 1), and
(nM Kd), significantly higher than for actin-cofilin ($M) they concluded that the binding loci for these two ABPs
(266). However, the affinity of cofilin for the binary DNase do not overlap. Others (13) reported that T!4 competes
I-actin complex is higher than for actin alone, and the with gelsolin segment 1, but both are dissociated by the
reverse is also true, namely, the affinity of DNase I is binding of DNase I to actin. A similar negative coopera-
higher for the actin-cofilin complex than for actin alone. tivity between spatially separated actin-binding sites was
Thus when DNase I is present in a mixture of actin observed between DNase I and gelsolin segment 1 (12).
monomers and the cofilin-actin binary complex, it binds Thus, although the binding site for DNase I is physically
to the binary complex forming a ternary (cofilin-actin- separated from the sites for gelsolin segment 1, T!4, and
DNase I) complex before it binds to the remaining free profilin, the negatively cooperative nature of their inter-
actin (205). This ternary complex was recently suggested actions suggests that an allosteric conformational change
based on colocalization in monkey kidney epithelial may be transmitted along the “length” of the small domain
cells (56). of actin.
In a report only just completed, we (81) have docu- Capping of the pointed ends of actin filaments by
mented the magnitude of actin conformational changes as tropomodulin (Tmod, discussed above) appears to in-
a consequence of separately binding cofilin and DNase I volve a ternary complex. Tmod binds only weakly to
and in the ternary cofilin-actin-DNase I complex. We used F-actin, but the affinity of Tmod for actin increases nearly
fluorescence and fluorescence resonance energy transfer three orders of magnitude in the presence of tropomyo-
spectroscopy to demonstrate significant conformational sin. Rapid exchange of Tmod at this end permits assembly
changes in the small domain (subdomains 1 and 2) of and disassembly of these filaments. The same is probably
monomeric actin. DNase I binding to actin increases by true for CapZ and other proteins that cap the barbed end
!1 Å the distance between probes attached to Gln-41 and of filaments.
Cys-374. When cofilin alone binds to actin, this distance In a two-hybrid system, actin interacting protein 1
decreases by !3 Å, but when both bind (ternary complex) (Aip1; a 65-kDa protein) was shown to form a ternary
the distance increases by !4 Å. This effect is completely complex with actin and ADF/cofilin (245). In yeast, Aip1
reversible and is independent of the order of addition. interacts with actin via subdomains 3 and 4 (5), and some
Thus the effects are not simply additive. We speculated of these residues do not overlap those for ADF/cofilin and

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


462 DOS REMEDIOS ET AL.

other ABPs. AIP1 alone has little or no filament severing Only one report (100) suggested that bacterial expression
activity, but it exhibits a cooperative filament severing could yield native actin, and the findings of this report
activity in the presence of Xenopus ADF/cofilin (245). have not been repeated. Yeast can be used to express
This ability of Aip1 to recognize ADF/cofilin bound to mutant actins (246), although the yield of the inserted
filaments may be important in the modulation of ADF/ actin gene product is impractically low.
cofilin activity. There are mouse fibroblast cell lines that express
Drugs like cytochalasin D act directly on the actin almost pure nonmuscle !-actin (252), but these are spe-
cytoskeleton by altering the rate of assembly/disassembly cial cases. Most cells express more than one actin isoform
at both the pointed and barbed ends of the filaments. A and, although several methods have been described for
recent structural analysis of latrunculin-A (193) revealed separating them, for example, hydroxyapatite column
that it binds only to actin monomers and prevents poly- chromatography (260), they are technically challenging.
merization by inducing specific structural changes. The Some actin mutations (for example, removal of the NH2-
binding site was not previously known to be an ABP terminal acidic residues) are not essential for the survival
binding locus. Crystals of actin-gelsolin segment 1 were of yeast but are needed to activate myosin ATPase activ-
soaked in latrunculin-A, and crystallographic refinement ity (64).
of the resulting ternary complex revealed that it tightly More than 150 papers have reported !300 actin mu-
links subdomains 2 and 4 of actin. Latrunculin-A produces tations (266). Only a handful deal with human actin mu-
local rather than global structural changes in actin, allo- tations, and most of these are deletion mutants near the
sterically altering the surfaces involved in the assembly of COOH terminus. Three at residues 336, 346, and 356 re-
F-actin. This effect is achieved by restricting the rotations sulted in the production of nonfunctional actin. The re-
of subdomains 2 and 4. Thus ternary complexes involving maining deletions at residues 366, 373, and 374 had sig-
actin appear to be important for the action of some drugs. nificantly reduced functionality. In a recent review, Mar-
ston and Hodgkinson (178) observed that 13 actin
mutations were known to cause hypertrophic cardiomy-
V. THE CYTOSKELETON AND PATHOLOGY
opathy (E99K, A331P, P164A, A295S), dilated cardiomy-
opathy (R312H, E361G), severe nemaline myopathy
Many of the proteins discussed above are affected by
(R183C, R256H, N280K, V370F, L94P), and actin myopathy
mutations resulting in the development of significant pa-
(G15R, V163L). Clearly, caution is needed in drawing
thology. Given the central role played by actin and the
sweeping conclusions about the functional effects of
ABPs, this outcome is not surprising (see Ref. 21a).
these structural changes. However, when these mutant
actins can be expressed in substantial quantities, we can
A. Actin begin to understand the functional consequences of mu-
tant actins in organs where more than one actin gene is
Actin has a remarkably conserved amino acid se- coexpressed (25).
quence, probably more than any other protein. This alone
suggests that mutations have been eliminated rather than
tolerated by the genome. Because actins are essential for B. Gelsolin
motility, force generation, and other important cellular
functions, it is remarkable that so few diseases are asso- If the actin cytoskeleton plays a significant role in
ciated with primary defects in microfilaments and their disease, and in particular tumors (215), then we would
regulators. No doubt, part of the reason is that there are also expect its associated ABPs to be involved in the
usually multiple actin genes, i.e., the problem is circum- disorder or disease. The functions of gelsolin have been
vented by redundancy. described above, but it is also a substrate for the apopto-
Perhaps the same evolutionary pressures that main- tic enzyme caspase-3 (150) that cleaves the actin-binding
tained the highly conserved sequences of actins also pro- domain of gelsolin from its calcium-binding domain. Phys-
tected them from being a major player in the disease ical separation of these two domains coincides with
process. Because of its pivotal role, defects or alterations plasma membrane blebbing, one of the hallmarks of apo-
in actin probably have dramatic effects on survival when ptosis. Overexpression of the Ca2"-independent NH2-ter-
no alternate actin gene is available (e.g., yeast has a single minal half of gelsolin induces apoptosis, whereas gelsolin
actin gene which is lethal when disrupted). null neutrophils have a delayed onset of apoptosis (150).
A major obstacle to studying the altered properties of Cancer cells characteristically lack apoptotic mecha-
mutant actins is the considerable difficulty cell biologists nisms and express less gelsolin (157), whereas expression
have had in expressing actin using standard protein ex- of mutated (His-321) gelsolin is associated with the loss of
pression systems. Bacteria have no actin of their own and tumorigenicity of transformed cells (102).
are therefore a logical choice as an expression system. Normally, plasma gelsolin functions as a part of the

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


ACTIN AND ACTIN BINDING PROTEINS 463

actin-scavenging system to assemble and disassemble ac- integrin receptors, the nucleus was elongated in the di-
tin filaments. The D187N and D187Y gelsolin mutations rection of the pull. Furthermore, they showed that this
generate a 71-residue fragment that forms amyloid fibrils effect on the nuclear membrane was mediated by the
in humans. These mutations were also associated with actin cytoskeleton. The actin cytoskeleton also provides a
familial amyloidosis of the Finnish type (FAF) (139). The physical connection between ion channels such as ClC3
symptoms of FAF patients include the accumulation in chloride ion channels in the cell membrane and the inte-
their tissues of amyloid derived from plasma gelsolin rior of the cell (33).
(285). Neither the recombinant wild-type nor the D187N Defects, deficiencies, or deletions of any component
FAF-associated gelsolin fragments self-associate into of the linkage between the extracellular matrix (lami-
amyloid at pH 7.5, but incubation of either fragment at nin-2, "-dystroglycan), trans-sarcolemmal proteins ("-,
low pH values (6.0–4.0) produces well-defined fibrils (237). !-, and #-sarcoglycans, !-dystroglycan), the cytoskeleton
(dystrophin, dystrobrevin, syntrophin, actin), and the nu-
cleus (emerin, lamin A/C) would dramatically alter this
C. Tropomodulin
putative mechanical signaling (175) and thus affect the
mechanical stability of cardiomyocytes.
Changes in gene expression of this thin filament cap-
Altered expression of cytoskeletal, linkage, and ex-
ping protein produce a surprisingly extensive range of
tracellular proteins has been reported in hypertrophied
phenotypes. Increases in Tmod expression levels are as-
cardiomyocytes (280, 282) and also in failing myocytes
sociated with the dilation (rather than the hypertrophic)
(61, 83, 122). Changes in expression levels of fibronectin,
component in human dilated cardiomyopathy (306). In a
laminin, fibrillin, fibulin SC1, and decorin, all components
transgenic mouse model, inhibition of Tmod expression
of the extracellular matrix, are upregulated in rat exper-
prevents dilation (279) but not hypertrophy of the myo-
imental myocardial infarctions (269). A number of mono-
cardium. Tmod is a promising development in our under-
genetic cardiomyopathies have recently been detailed in a
standing of this baffling disorder (278).
selected number of families, which include a range of
Graves’ disease is a complex autoimmune disease
different genes involved in the surface-nucleus pathway
characterized by multiple susceptible loci involving cyto-
(55). Thus cytoskeletal proteins may be involved in the
toxic T cells, HLA antigens, and thyroid-stimulating hor-
progressive defects leading to heart failure. Hypertrophic
mone receptors (299). Human Tmod has !40% identity at
cardiomyopathy appears to be a final common outcome
the NH2 and COOH termini of the 64-kDa autoantigen
for gene defects involving an astonishing variety of cy-
protein identified in Graves’ disease. The insertion of
toskeletal and sarcomeric proteins.
several homologous repeats in the midsection of the
Graves’ protein together with an extension of the COOH
terminus of Tmod accounts for the differences in length E. Heart Failure
between the Graves’ protein (572 residues) and tropo-
Familial hypertrophic cardiomyopathy (FHC) is an
modulin (359 residues) (299). It is therefore likely the two
autosomal dominant disorder resulting in increased myo-
genes evolved from a common ancestral gene (276). The
cardial mass and myofibrillar disarray. It is the most
gene for Tmod was assigned to human chromosome 9 in
common cause of sudden death in the young. Linkage
a region also known to contain several other genes and
studies and candidate-gene approaches have revealed dis-
disease loci, for example, an unrelated cytoskeletal gene
ease-causing mutations in more than eight different genes
for gelsolin.
all encoding sarcomeric proteins. More than 100 muta-
Finally, two genes encoding for Tmod that map next
tions have already been identified, and they involve prac-
to another on human chromosome 15 make them candi-
tically every protein component in the sarcomere. Muta-
dates for amyotrophic lateral sclerosis and a dyslexia
tions are observed in !-myosin heavy chain (105), tropo-
gene as well as a candidate gene for limb girdle muscular
nin I (144, 92), troponin T (238), "-tropomyosin (282),
dystrophy (70). At this stage it is not clear how these
myosin-binding protein C (296), regulatory and essential
syndromes are related to Tmod expression disorders, but
myosin light chains (226), cardiac "-actin (209), titin
the clear link suggests that Tmod is a key gene.
(253), and desmin (122, 253). However, these mutations
are present in only half the patients (296), and conse-
D. Connections Between the Extracellular Matrix quently, there must be other gene mutations that cause
and the Nucleus FHC, some of which will probably involve the cytoskele-
ton. The common phenotype resulting from a wide range
Maniotis et al. (175) showed that a mechanical tug on of cytoskeletal/sarcomeric proteins may well prove to be
cell surface receptors can immediately change the orga- a reflection of the functional linkage of these proteins to
nization of proteins in the cytoplasm and the nucleus. both the extracellular matrix and the nucleus in cardio-
They showed that by specifically pulling on cell surface myocytes.

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


464 DOS REMEDIOS ET AL.

Studies of human dilated cardiomyopathy (DCM) (SNPs) from human populations. The idea here is that
(306) and of mouse models (279) of DCM also point to a SNPs (unlike most other analytical methods) are poten-
critical role of cytoskeletal elements at the Z-disks in tially capable of discovering the cause of disease.
chamber dilation. Mice lacking the muscle LIM-domain A number of companies have developed technologies
protein (MLP), a Z-disk protein that interacts directly with for analyzing SNPs, but the most challenging aspect of the
"-actinin, displayed many of the features of human DCM technology is its ability to accurately detect very small
(8). Mutations (usually more than one per protein) in differences between DNA restriction fragments in which
other proteins associated with "-actinin, cardiac "-actin, only one base pair has been replaced (e.g., T instead of A
desmin, and titin were subsequently found in familial or G instead of a C base pair, or vice versa). Sequenom
forms of DCM (76, 209, 253). Also, mutations of other (www.sequenom.com) has developed a method that very
Z-disk components in the mouse result in DCM (55, 59). accurately “weighs” the DNA fragments using mass spec-
Mechanical stress on the extracellular matrix leads to trometry which achieves precisions of %99.9%. It is de-
defects in Z-disk formation, sarcomeric stabilization, and batable whether this accuracy is really needed, although
altered muscle-cell signaling, thereby driving the onset of their approach was both unusual and effective. They have
chamber dilation and heart failure (58). In a dog model of identified at least three cytoskeletal proteins that are
heart failure, cytoskeletal actin was upregulated while statistically associated with specific diseases in specific
desmin (a protein in the intercalated disks) was down- human populations. These diseases have a late onset and
regulated (122). are not associated with mortality in infants, juveniles, or
young adults.
F. Gene Arrays
H. Cell Signaling and Actin Microfilaments
Yang et al. (306) used gene array technology to ex-
amine altered genes in failing human hearts. They ob- In mammals, there is a well-defined connection be-
served an increased expression of gelsolin and a signifi- tween extracellular signals (like insulin), their membrane
cant upregulation of "-B-crystallin, a desmin-associated receptors, intracellular signals (Ras and Rac), and the
protein that protects or chaperones actin and desmin regulators of actin microfilament assembly (Arp2/3, gel-
filaments from stress-induced damage. "-B-crystallin is a solin, ADF/cofilin). Because the binding of ligands to
member of 27-kDa heat shock protein family. These data some surface receptors is known to activate genes, the
(using Affymetrix gene arrays) were confirmed by us cytoskeletal links may extend from the membrane all the
(using NEN arrays) (135). They support the view that the way to the nuclei and perhaps even to the chromosomes.
actin cytoskeleton is damaged in DCM. Samples of failing The control of these connections in yeast, Drosophila,
left ventricles have been studied by two-dimensional gel and mammalian cells was recently summarized by Hall
electrophoresis of human DCM patients (68) and of an and colleagues (24). The relationships between compo-
animal model of DCM (122). In another form (familial nents of the signaling pathways are complex and are
hypertrophic) of human heart failure, the expression of beyond the scope of this review.
"-actin, plasma gelsolin, thymosin !4, and several other
cytoskeletal proteins (23) changed more than twofold.
Finally, gene arrays have been used to examine an exper- VI. UNRESOLVED ISSUES
imental myocardial infarction using failing rat hearts
(269). Cytoskeletal actin and several ABPs ("-actin, thy- A. Atomic Structure of F-actin
mosin !4 and !19, Arp2/3 and a number of other ABPs not
reviewed here) were upregulated. The data clearly impli-
It is common to assume that because we know all or
cate the involvement of cytoskeletal and myofibrillar
most of the structural atomic details of the actin mono-
genes in human heart failure. Thus there is growing evi-
mer, the available models of F-actin are equally valid.
dence for altered expression of cytoskeletal proteins in
How good is this assumption?
heart failure. It is likely that similar results will be re-
It is true that the known atomic structures of mono-
ported for the failure of other organs.
meric actins are very similar. However, significant differ-
ences are beginning to emerge. Nearly all of the available
G. Single Nucleotide Polymorphisms, Diseases, and atomic structures were achieved using actin cocrystal-
the Cytoskeleton lized with an ABP. Recently, a new structure of actin was
reported (217) that is not complexed to an ABP. These
A relatively new method of determining the connec- authors observed “a dramatic conformational change in
tions between human diseases and gene mutations in- subdomain 2” compared with previous structures. It will
volves analysis of single nucleotide polymorphisms therefore be of considerable interest to see the effect the

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


ACTIN AND ACTIN BINDING PROTEINS 465

new structure from Dominguez et al. (217) has on a model complex involving monomeric actin in vitro, is there a
of F-actin with its bound ADP. Bound nucleotide can have role for ternary complexes in vivo? The more details we
dramatic effects on the structure of actin and on the learn about the ABPs, the more they seem to fit into
affinities of ABPs for actin. Three-dimensional recon- families with common cellular functions. Not all DNases
structions of electron micrographs of actin filaments are bind actin, so it is still not clear if they can play a universal
based on low-resolution structural data. Thus the avail- role.
able models make assumptions about the arrangement
and orientation of the monomers in the filament. To date,
D. Undiscovered ABPs
most filament models (128, 167, 283) have been con-
structed using the structures from actin cocrystals with
Despite the proliferation of ABPs over recent years
ABPs, and there is an assumption that this is unaltered in
(65 new proteins have been discovered over the past 5
the filament. Arguably, filaments are the most biologically
years, Ref. 151), there are still significant areas of the
relevant form of actin in cells, so there is a real need to
actin monomer surface that do not appear to be the site
resolve their structure at high resolution. Several groups
for ABPs (see Fig. 6). Some of this surface may come to
are currently working on this most difficult, perhaps in-
be occupied by known but not yet fully characterized
tractable, structural problem.
ABPs. It is unlikely that we have discovered all of the
ABPs, and it seems a safe bet to suggest that as the human
B. Functional Differences Between Actin Isoforms genome and other genomes become refined, at least some
of the “orphan” genes will turn out to be new ABPs.
The great majority of the several thousand references
on this topic have used rabbit skeletal muscle actin. Only
E. Cooperative Binding of ABPs Along Filaments
very small sequence differences exist between different
actin isoforms, and these probably have functional con-
There is a spatial separation and organization of fil-
sequences. For example, the myocardium of small mam-
ament nucleation, elongation, branching, and other pro-
mals like mice and rats expresses almost pure cardiac
cesses involving actin assembly and disassembly, but lit-
"-actin, whereas cardiomyocytes in larger animals (in-
tle or nothing is known about the controls of these pro-
cluding humans) express about equal amounts of cardiac
cesses. For example, we know that Arp2/3 produces a
and skeletal "-actin (25, 113). Why do larger hearts ex-
regular branching of actin microfilaments and that this
press skeletal muscle actin? #-Actin isoforms are located
regular branching is important for the protrusion of cyto-
just beneath the sarcolemma, whereas only "-actins are
plasmic blebs and fingers (195), but what controls the
incorporated into sarcomeres (113). We do not know if
distribution of Arp2/3? In muscle thin filaments, binding
isoforms can coassemble or whether such filaments have
of a myosin head at one point alters the probability of a
different effects on activities such as myosin ATPase.
second head binding up to several monomers along the
Very little is known about the functional differences be-
filament. What conformational changes are involved here,
tween actin isoforms. Even less is known about whether
and how are they propagated along the filaments? ADF/
certain ABPs bind to particular actin isoforms and
cofilin binds cooperatively to F-actin, but does this coop-
whether differences in actin isoform structure determine
erativity depend on the actin isoform?
which ABPs bind.
The situation is no clearer for ADF/cofilin. We know
there are functional differences (nucleation, elongation, F. Cytoskeletal Proteins in the Nucleus
fragmenting, binding ratios to actin) between ADFs and
cofilins, but comparisons based on existing literature are A number of cytoskeletal proteins are found in the
difficult because groups investigating these properties of- nucleus including cofilin (56, 203) (which has a nuclear
ten use different (yeast, chick, human) isoforms. There localization peptide) and actin (which lack this se-
may also be differences between isoforms within the quence). We do not know the functions of these proteins
cofilins and ADFs, and these will have subtle but impor- or how they are regulated.
tant functional effects.
G. Disease and the Cytoskeleton
C. Ternary Complexes of Monomeric Actin
With ABPs The role of the cytoskeleton and ABPs in disease is
an emerging story. There can be little doubt that several
We have reviewed the evidence for a ternary complex cytoskeletal proteins are positively associated with signif-
between actin monomers and two ABPs, cofilin and icant human disease. Sequenom of San Diego and Gemini
DNase I. Although there is good evidence for a ternary Genomics PLC of Cambridge recently discovered three of

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


466 DOS REMEDIOS ET AL.

these proteins, but their identities have not yet been tion of cardiac cytoarchitectural organization, dilated cardiomyop-
athy, and heart failure. Cell 88: 393– 403, 1997.
disclosed. However, it is possible that identification of 9. ASAKURA S, KASAI M, AND OOSAWA F. The effects of temperature on
specific disease-associated genes may not produce a com- the equilibrium state of actin solutions. J Polymer Sci 44: 35– 49,
plete understanding of the disease process for we know 1960.
10. AZUMA T, WITKE W, STOSSEL TP, HARTWIG JH, AND KWIATKOWSKI DJ.
that a single phenotype like hypertrophic cardiomyopathy Gelsolin is a downstream effector of rac for fibroblast motility.
can be the product of mutations in almost any one of the EMBO J 17: 1362–1370, 1998.
myofibrillar proteins. 11. BAINS NPS, GORBATYUK VY, NOSWORTHY NJ, ROBSON SA, MACIEJEWSKI
MW, DOS REMEDIOS CG, AND KING GF. Letter to the Editor: backbone
and side chain 1H, 15N, and 13C assignments for chick cofilin.
J Biomol NMR 22: 193–194, 2002.
H. Prokaryotic Cytoskeletal Elements 12. BALLWEBER E, GIEHL K, HANNAPPEL E, HUFF T, JOCKUSCH BM, AND
MANNHERZ HG. Plant profilin induces actin polymerization from
actin: beta-thymosin complexes and competes directly with beta-
It seems likely that the ancestor of eukaryotic actin is thymosins and with negative co-operativity with DNase I for bind-
the bacterial protein MreB. This protein can self-assemble ing to actin. FEBS Lett 425: 251–255, 1998.
and contributes to structures that resemble straightened 13. BALLWEBER E, HANNAPPEL E, HUFF T, AND MANNHERZ HG. Mapping the
binding site of thymosin beta4 on actin by competition with G-actin
actin rails. This discovery prompts questions like: Are binding proteins indicates negative co-operativity between binding
there MreB-binding proteins, equivalent to ABPs, that sites located on subdomains 1 and 2 of actin. Biochem J 327:
control filament assembly? Do bacteria have motor pro- 787–793, 1997.
14. BALLWEBER E, HANNAPPEL E, HUFF T, STEPHAN H, HAENER M,
teins that employ the MreB “rails”? TASCHNER N, STOFFLER D, AEBI U, AND MANNHERZ HG. Polymerisation
of chemically cross-linked actin:thymosin !4 complex to filamen-
We are grateful to Roger Craig, Roberto Dominguez, tous actin: alteration in helical parameters and visualisation of
Shin’ichi Ishiwata, Glenn King, Steve Marston, Takashi Obinata, thymosin !4 binding on F-actin. J Mol Biol 315: 613– 625, 2002.
15. BALLWEBER E, HANNAPPEL E, NIGGEMEYER B, AND MANNHERZ HG.
Tetsura Fujisawa, Alla Kostyukova, Yuichiro Maeda, Bob Rob- Induction of the polymerization of actin from the actin:thymosin
inson, Dan Safer, Fusinita van den Ent, and Willy Wriggers for beta 4 complex by phalloidin, skeletal myosin subfragment 1,
many of the figures used here and for valuable contributions on chicken intestinal myosin I and free ends of filamentous actin. Eur
the manuscript and for discussions. J Biochem 223: 419 – 426, 1994.
16. BAMBURG JR. Proteins of the ADF/cofilin family: essential regulators
We are grateful for the support of the National Health and of actin dynamics. Annu Rev Cell Dev Biol 15: 185–230, 1999.
Medical Research Council of Australia, the Australian Research 17. BAMBURG JR AND BERNSTEIN BW. Actin and actin-binding proteins in
Council, the National Heart Foundation of Australia, and the neurons. In: The Neuronal Cytoskeleton, edited by Burgoyne RD.
Medical Foundation of the University of Sydney. New York: Wiley-Liss, 1991, p. 121–160.
Address for reprint requests and other correspondence: C. 18. BAMBURG JR AND BRAY D. Distribution and cellular localization of
actin depolymerizing factor. J Cell Biol 105: 2817–2825, 1987.
G. dos Remedios, Institute for Biomedical Research, Muscle 19. BAMBURG JR AND DRUBIN DG. Actin depolymerizing factor (ADF)/
Research Unit, Dept. of Anatomy and Histology, F13, Univ. of cofilin. In: Guidebook to the Cytoskeletal and Motor Proteins, ed-
Sydney, Sydney 2006, Australia (E-mail: crisdos@anatomy. ited by Kreis T and Vale R. Oxford, UK: Oxford Univ. Press, 1999,
usyd.edu.au). p. 19 –24.
20. BAMBURG JR, HARRIS HE, AND WEEDS AG. Partial purification and
characterization of an actin depolymerizing factor from brain.
FEBS Lett 121: 178 –182, 1980.
REFERENCES 21. BAMBURG JR, MCGOUGH A, AND ONO S. Putting a new twist on actin:
ADF/cofilins modulate actin dynamics. Trends Cell Biol 9: 364 –370,
1. ABE H, ENDO T, YAMAMOTO K, AND OBINATA T. Sequence of cDNAs 1999.
encoding actin depolymerizing factor and cofilin of embryonic 21a.BAMBURG JR AND WIGGON OP. ADF/cofilin and actin dynamics in
chicken skeletal muscle: two functionally distinct actin-regulatory disease. Trends Cell Biol 12: 598 – 605, 2002.
proteins exhibit high structural homology. Biochemistry 29: 7420 – 22. BAO L, LODA M, JANMEY PA, STEWART R, ANAND-APTE B, AND ZETTER
7425, 1990. BR. Thymosin beta 15: a novel regulator of tumor cell motility
2. ABE H, OBINATA T, MINAMIDE LS, AND BAMBURG JR. Xenopus laevis upregulated in metastatic prostate cancer. Nat Med 2: 1322–1328,
actin-depolymerizing factor/cofilin: a phosphorylation-regulated 1996.
protein essential for development. J Cell Biol 132: 871– 885, 1996. 23. BARRANS JD, STAMATIOU D, AND LIEW C. Construction of a human
3. AEBI U, FOWLER WE, ISENBERG G, POLLARD TD, AND SMITH TD. Crys- cardiovascular cDNA microarray: portrait of the failing heart. Bio-
talline actin sheets their structure and polymorphism. J Cell Biol chem Biophys Res Commun 280: 964 –969, 2001.
91: 340 –351, 1981. 24. BECK T, DELLEY PA, AND HALL MN. Control of the actin cytoskeleton
4. ALLEN ML, DOBROWOWOSKI JM, MULLER H, SIBLEY LD, AND MANSOUR by extracellular signals. Res Prob Cell Motil 2002 32: 231–262, 2002.
TE. Cloning and characterization of actin depolymerization factor 25. BENNETTS BH, BURNETT L, AND DOS REMEDIOS CG. Differential expres-
from Toxoplasma gondii. Mol Biochem Parasitol 88: 43–52, 1997. sion of alpha-actin genes within the human heart. J Mol Cell Car-
5. AMBERG DC, BASART E, AND BOTSTEIN D. Defining protein interac- diol 18: 993–996, 1986.
tions with yeast actin in vivo. Struct Biol 2: 38 –35, 1995. 26. BERNARD O, GANIATSAS S, KANNOURAKIS G, AND DRINGEN R. Kiz-1, a
6. AMPE C, MARKEY F, LINDBERG U, AND VANDEKERCKHOVE J. The primary protein with LIM zinc finger and kinase domains, is expressed
structure of human platelet profilin: reinvestigation of the calf mainly in neurones. Cell Growth Differ 5: 1159 –1171, 1994.
spleen profilin sequence. FEBS Lett 228: 17–21, 1988. 27. BERNSTEIN BW, PAINTER WB, CHEN H, MINAMIDE LS, ABE H, AND
7. ARBER S, BARBAYANNIS FA, HANSER H, SCHNEIDER C, STANYON CA, BAMBURG JR. Intracellular pH modulation of ADF/cofilin proteins.
BERNARD O, AND CARONI P. Regulation of actin dynamics through Cell Motil Cytoskeleton 47: 319 –336, 2000.
phosphorylation of cofilin by LIM-kinase. Nature 393: 805– 809, 28. BLANCHOIN L, POLLARD TD, AND HITCHOCK-DEGREGORI SE. Inhibition
1998. of the Arp2/3 complex-nucleated actin polymerization and branch
8. ARBER S, HUNTER JJ, ROSS J JR, HONGO M, SANSIG G, BORG J, PERRIARD formation by tropomyosin. Curr Biol 11: 1300 –1304, 2001.
JC, CHIEN KR, AND CARONI P. MLP-deficient mice exhibit a disrup- 29. BLANCHOIN L, ROBINSON RC, CHOE S, AND POLLARD TD. Phosphoryla-

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


ACTIN AND ACTIN BINDING PROTEINS 467

tion of Acanthamoeba actophorin (ADF/cofilin) blocks interaction NH2-terminal domains of the alpha-1 and beta subunits are not
with actin without a change in atomic structure. J Mol Biol 295: required for actin capping and alpha 1 beta and alpha 2 beta
203–211, 2000. heterodimers bind differentally to actin. J Biol Chem 269: 6992–
30. BORK P, SANDER C, AND VALENCIA A. An ATPase domain common to 6998, 1994.
prokaryotic cell cycle proteins, sugar kinases, actin, and hsp70 heat 54. CHEN H, BERNSTEIN BW, AND BAMBURG JR. Regulating actin-filament
shock proteins. Proc Natl Acad Sci USA 89: 7290 –7294, 1992. dynamics in vivo. Trends Biochem Sci 25: 19 –23, 2000.
31. BORTHS EL AND WELCH MD. Turning on the Arp2/3 complex at 55. CHEN J AND CHIEN KR. Complexity in simplicity: monogenic disor-
atomic resolution. Structure 10: 131–135, 2002. ders and complex cardiomyopathies. J Clin Invest 103: 1483–1485,
32. BOWMAN GD, NODELMAN IM, HONG Y, CHUA NH, LINDBERG U, AND 1999.
SCHUTT CE. A comparative structural analysis of the ADF/cofilin 56. CHHABRA D, BAO S, AND DOS REMEDIOS CG. The distribution of cofilin
family. Proteins 41: 374 –384, 2000. and DNase I in vivo. Cell Res 12: 207–214, 2002.
33. BRITTON FC, HATTON WJ, ROSSOW CF, DUAN D, HUME JR, AND HORO- 57. CHHABRA D, NOSWORTHY NJ, AND DOS REMEDIOS. The role of ATP, ADP
WITZ B. Molecular distribution of volume-regulated chloride chan- and divalent cations in the formation of binary and ternary com-
nels (ClC-2 and ClC-3) in cardiac tissues. Am J Physiol Heart Circ plexes of actin with cofilin and DNase I. Electrophoresis 17: 3863–
Physiol 279: H2225–H2233, 2000. 3870, 2000.
34. BROWN SS. Cooperation between microtubule- and actin-based mo- 58. CHIEN KR. Genomic circuits and the integrative biology of cardiac
tor proteins. Annu Rev Cell Dev Biol 15: 63– 80, 1999. diseases. Nature 407: 227–232, 2000.
35. BURTNICK LD, KOEPF EK, GRIMES J, JONES EY, STUART DI, MCLAUGHLIN 59. CHIEN KR. Stress pathways and heart failure. Cell 98: 555–558, 1999.
PJ, AND ROBINSON RC. The crystal structure of plasma gelsolin: 60. CHIK JK, LINDBERG U, AND SCHUTT CE. The structure of an open state
implications for actin severing, capping, and nucleation. Cell 90: of beta-actin at 2.65 Å resolution. J Mol Biol 263: 607– 623, 1996.
661– 670, 1997. 61. COLLINS JF, PAWLOSKI-DAHM C, DAVIS MG, BALL N, DORN GW II, AND
36. BURTNICK LD, ROBINSON RC, AND CHOE S. Structure and function of WALSH RA. The role of the cytoskeleton in left ventricular pressure
gelsolin. Res Prob Cell Differ 32: 201–211, 2001. overload hypertrophy and failure. J Mol Cell Cardiol 28: 1435–1443,
37. BUSS FC, TEMM-GROVE C, HENNING S, AND JOCHUSH B. Distribution of 1996.
profilin in fibroblasts correlates with the presence of highly dy- 62. CONLEY CA, FRITZ-SIX KL, ALMENAR-QUERALT A, AND FOWLER VM.
namic actin-filaments. Cell Motil Cytoskeleton 22: 51– 61, 1992. Leiomodins: larger members of the tropomodulin (Tmod) gene
38. CALDWELL JE, HEISS SG, MERMALL V, AND COOPER JA. Effects of CapZ, family. Genomics 73: 127–139, 2001.
an actin capping protein of muscle, on the polymerization of actin. 63. COOKE R. The role of the bound nucleotide in the polymerization of
Biochemistry 28: 8506 – 8514, 1989. actin. Biochemistry 14: 3250 –3256, 1975.
39. CAMERON LA, FOOTER MJ, VAN OUDENAARDEN A, AND THERIOT J. 64. COOKE RK, BLAKE WT, AND RUBENSTEIN PA. Removal of the amino
Motility of ActA protein-coated microspheres driven by actin poly- terminal acidic residues of yeast actin. J Biol Chem 267: 9430 –9436,
merization. Proc Natl Acad Sci USA 96: 4908 – 4913, 1999. 1992.
40. CARLIER MF, DIDRY D, ERK I, LEPAULT J, VAN TROYS ML, VANDE- 65. COOPER JA AND SCHAFER DA. Control of actin assembly and disas-
KERCKHOVE J, PERELROIZEN I, YIN H, DOI Y, AND PANTALONI D. T!4 is
sembly at filament ends. Curr Opin Cell Biol 12: 97–103, 2000.
not a simple G-actin sequestering protein and interacts with F-actin
66. COOPER JA AND SCHAFER DA. Control of actin assembly and disas-
at high concentration. J Biol Chem 271: 9231–9239, 1996.
sembly at filament ends. Curr Opin Cell Biol 12: 66 –97–103, 1984.
41. CARLIER MF, JEAN C, RIEGER KJ, LENFANT M, AND PANTALONI D.
67. COPPIN CM AND LEAVIS PC. Quantitation of liquid crystalline order-
Modulation of the interaction between G-actin and thymosin beta 4
ing in F-actin solutions. Biophys J 63: 794 – 807, 1992.
by the ATP/ADP ratio: possible implication in the regulation of
68. CORBETT JM, WHY HJ, WHEELER CH, RICHARDSON PJ, ARCHARD LC,
actin dynamics. Proc Natl Acad Sci USA 90: 5034 –5038, 1993.
YACOUB MH, AND DUNN MJ. Cardiac protein abnormalities in dilated
42. CARLIER MF, LAURENT V, SANTOLINI J, MELKI R, DIDRY D, XIA GX, HONG
cardiomyopathy detected by two-dimensional polyacrylamide gel
Y, CHUA NH, AND PANTALONI D. Actin depolymerizing factor (ADF/
cofilin) enhances the rate of filament turnover: implication in actin- electrophoresis. Electrophoresis 19: 2031–2042, 1998.
based motility. J Cell Biol 136: 1307–1322, 1997. 69. COSSART P. Actin-based motility of pathogens: the Arp2/3 complex
43. CARLIER MF AND PANTALONI D. Direct evidence for ADP-Pi-F-actin as is a central player. Cell Microbiol 2: 195–205, 2000.
the major intermediate in ATP-actin polymerization. Rate of disso- 70. COX PR AND ZOGHBI HY. Sequencing, expression analysis, and map-
ciation of Pi from actin filaments. Biochemistry 25: 211–214, 1986. ping of three unique human tropomodulin genes and their mouse
44. CARLIER MF AND PANTALONI D. Actin polymerization and ATP hydro- orthologs. Genomics 63: 97–107, 2000.
lysis. Adv Biophys 76: 51–73, 1990. 71. CRAMER LP, BRIGGS LJ, AND DAWE HR. Use of fluorescently labelled
45. CARLIER MF, PANTALONI D, EVANS JA, LAMBOOY PK, KORN ED, AND deoxyribonuclease I to spatially measure G-actin levels in migrat-
WEBB MR. The hydrolysis of ATP that accompanies polymerization ing and nonmigrating cells. Cell Motil Cytoskeleton 51: 27–38, 2002.
is essentially irreversible. FEBS Lett 235: 211–214, 1988. 72. CUNNINGHAM CC, STOSSEL TP, AND KWIATKOWSKI DJ. Enhanced motil-
46. CARLIER MF, PANTALONI D, AND KORN ED. Evidence for an ATP cap ity in NIH 3T3 fibroblasts that overexpress gelsolin. Science 251:
at the ends of actin filaments and its aggregation of the F-actin 1233–1236, 1991.
steady state. J Biol Chem 259: 9983–9986, 1984. 73. CURMI PMG, BARDEN JA, AND DOS REMEDIOS CG. Actin tube forma-
47. CARLIER MF, PANTALONI D, AND KORN ED. The mechanism of ATP tion: effects of variation in commonly used solvent conditions.
hydrolysis accompanying the polymerization of Mg-actin and Ca- J Muscle Res Cell Motil 5: 423– 430, 1984.
actin. J Biol Chem 262: 3052–3059, 1987. 74. CURMI PMG, VAN DONKELAAR A, TULLOCH PA, AND DOS REMEDIOS CG.
48. CARPINTERO P, ANADON R, DIAZ-REGUEIRA S, AND GOMEZ-MARQUEZ J. A comparison of the unit cell of dried and hydrated actin micro-
Expression of thymosin 4 messenger RNA in normal and kainate- crystals. In: Actin Structure and Function in Muscle and Non-
treated rat forebrain. Neuroscience 90: 1433–1444, 1999. muscle Cells, edited by dos Remedios CG and Barden JA. Sydney,
49. CARPINTERO P, FRANCO DEL AMO F, ANADÓN R, AND GÓMEZ-MÁRQUEZ J. Australia: Academic, 1983, p. 45–52.
Thymosin beta10 mRNA expression during early postimplantation 75. CZISCH M, SCHLEICHER M, HORGER S, VOELTER W, AND HOLAK TA.
mouse development. FEBS Lett 394: 103–106, 1996. Conformation of thymosin beta 4 in water determined by NMR
50. CARRON E. Regulation of Wiscott-Aldrich syndrome protein and spectroscopy. Eur J Biochem 218: 335– 44, 1993.
related molecules. Curr Opin Cell Biol 14: 82– 87, 2000. 76. DALAKAS MC, PARK KY, SEMINO-MORA C, LEE HS, SIVAKUMAR K, AND
51. CASELLA JF, CRAIG SW, MAACK DJ, AND BROWN AE. CapZ(36/32), a GOLDFARB LG. Desmin myopathy, a skeletal myopathy with cardio-
barbed end actin-capping protein, is a component of the Z-line of myopathy caused by mutations in the desmin gene. N Engl J Med
skeletal muscle. J Cell Biol 105: 371–379, 1987. 342: 770 –780, 2000.
52. CASELLA JF, MAACK DJ, AND LIN S. CapZ(36/32), a barbed end actin- 77. DE LA CRUZ EM, MANDINOVA A, STEINMETZ MO, STOFFLER D, AEBI U,
capping protein, is a component of the Z-line of skeletal muscle. AND POLLARD TD. Polymerization and structure of nucleotide-free
J Biol Chem261: 10915–10921, 1986. actin filaments. J Mol Biol 295: 517–526, 2000.
53. CASELLA JF AND TORRES MA. Interaction of Cap Z with actin. The 78. DE LA CRUZ EM, OSTAP EM, BRUNDAGE RA, REDDY KS, SWEENEY HL,

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


468 DOS REMEDIOS ET AL.

AND SAFER D. Thymosin-beta(4) changes the conformation and (His321)gelsolin isolated from a flat revertant of ras-transformed
dynamics of actin monomers. Biophys J 78: 2516 –2527, 2000. cells. Eur J Biochem 229: 615– 620, 1995.
79. DE LA CRUZ EM AND POLLARD TD. Nucleotide-free actin: stabilization 103. FUJIWARA I, SUETSUGU S, UEMURA S, TAKENAWA T, AND ISHIWATA S.
by sucrose and nucleotide binding kinetics. Biochemistry 34: 5452– Visualization and force measurement of branching by Arp2/3 com-
5461, 1995. plex and N-WASP in actin filament. Biochem Biophys Res Com-
80. DANCKER P, LÖW I, HASSELBACK W, AND WIELAND T. Interaction of mun 293: 1550 –1555, 2002.
actin with phalloidin: polymerization and stabilization of F-actin. 104. GALKIN VE, ORLOVA A, LUKOYANOVA N, WRIGGERS W, AND EGELMAN EH.
Biochim Biophys Acta 400: 407– 414, 1975. Actin depolymerizing factor stabilizes an existing state of F-actin
81. DEDOVA I, DEDOV V, NOSWORTHY NJ, HAMBLY BD, AND DOS REMEDIOS and can change the tilt of F-actin subunits. J Cell Biol 153: 75– 86,
CG. Cofilin and DNase I affect the conformation of the small 2001.
domain of actin. Biophys J 82: 3134 –3143, 2002. 105. GEISTERFER-LOWRANCE AA, KASS S, TANIGAWA G, VOSBERG HP, MC-
82. DEHOSTOS EL, BRADTKE B, LOTTSPEICH F, AND GERISCH G. Coactosin, KENNA W, SEIDMAN CE, AND SEIDMAN JG. A molecular basis for
a 17 kDa F-actin binding protein from Dictyostelium discoideum. familial hypertrophic cardiomyopathy: a beta cardiac myosin heavy
Cell Motil Cytoskeleton 26: 181–191, 1993. chain gene missense mutation. Cell 62: 999 –1006, 1990.
83. DE TOMBE PP. Altered contractile function in heart failure. Cardio- 106. GOLD JS, BAO L, GHOUSSOUB RA, ZETTER BR, AND RIMM DL. Localiza-
vasc Res 37: 367–380, 1998. tion and quantitation of expression of the cell motility-related
84. DEVINENI N, MINAMIDE LS, NIU M, SAFER D, VERMA R, BAMBURG JR, protein thymosin beta15 in human breast tissue. Mol Pathol 10:
AND NACHMIAS VT. A quantitative analysis of G-actin binding pro- 1106 –1112, 1997.
teins and the G-actin pool in developing chick brain. Brain Res 823: 107. GOLDSCHMIDT-CLERMONT PJ, FURMAN MI, WACHSSTOCK D, SAFER D,
129 –140, 1999. NACHMIAS VT, AND POLLARD TD. The control of actin nucleotide
85. DIDRY D, CARLIER MF, AND PANTALONI D. Synergy between actin exchange by thymosin beta 4 and profilin. A potential regulatory
depolymerizing factor cofilin and profilin in increasing actin fila- mechanism for actin polymerization in cells. Mol Biol Cell 3: 1015–
ment turnover. J Biol Chem 273: 25602–25611, 1998. 1024, 1992.
86. DIEFFENBACH CW, SENGUPTA DN, KRAUSE D, SAWZAK D, AND SILVERMAN 108. GOLDSCHMIDT-CLERMONT PJ, MACHESKY LM, BALDASSARE JJ, AND POL-
RH. Cloning of murine gelsolin and its regulation during differen- LARD TD. The actin-binding protein profilin binds to PIP2 and inhib-
tiation of embryonal carcinoma cells. J Biol Chem 264: 13281– its its hydrolysis by phospholipase C. Science 247: 1575–1578, 1990.
13288, 1989. 109. GOLDSCHMIDT-CLERMONT PJ, MACHESKY LM, DOBERSTEIN SK, AND POL-
87. DOS REMEDIOS CG AND DICKENS MJ. Actin microcrystals and tubes LARD TD. The mechanism of interaction of human platelet profilin
formed in the presence of gadolinium ions. Nature 276: 731–733, with actin. J Cell Biol 113: 1081–1089, 1991.
1978. 110. GOLLA R, PHILP N, SAFER D, CHINTAPALLI J, HOFFMAN R, COLLINS L, AND
88. DOS REMEDIOS CG AND THOMAS DD. (Editors). Molecular Interac- NACHMIAS VT. Co-ordinate regulation of the cytoskeleton in 3T3
tions of Actin. Actin Structure and Actin-Binding Proteins. Hei- cells overexpressing thymosin-!4. Cell Motil Cytoskeleton 38: 187–
delberg, Germany: Springer-Verlag, 2001, p. 32. 200, 1997.
89. DU J AND FRIEDEN C. Kinetic studies on the effect of yeast cofilin on 111. GRAZI E AND TROMBETTA G. Effects of temperature on actin poly-
yeast actin polymerization. Biochemistry 37: 13276 –13284, 1998. merized by Ca2". Biochem J 232: 297–300, 1985.
90. EDWARDS DC AND GILL GN. Structural features of LIM kinase that 112. GREMM D AND WEGNER A. Gelsolin as a calcium-regulated actin
control effects on the actin cytoskeleton. J Biol Chem 274: 11352– filament-capping protein. Eur J Biochem 267: 4339 – 4345, 2000.
11361, 1999. 113. GUNNING P, PONTE P, BLAU H, AND KEDES L. Alpha-skeletal and
91. EDWARDS KA, MONTAGUE RA, SHEPARD S, EDGAR BA, ERIKSON RL, AND alpha-cardiac actin genes are co-expressed in adult human skeletal
KIEHART DP. Identification of Drosophila cytoskeletal proteins by muscle and heart. Mol Cell Biol 3: 1985–1995, 1983.
induction of abnormal cell shape in fission yeast. Proc Natl Acad 114. HALL AK. Molecular interactions between G-actin, DNase I and the
Sci USA 91: 4589 – 4593, 1994. beta-thymosins in apoptosis: an hypothesis. Med Hypotheses 43:
92. ELLIOTT K, WATKINS H, AND REDWOOD CS. Altered regulatory proper- 125–131, 1994.
ties of human cardiac troponin I mutants that cause hypertrophic 115. HALL AK. Thymosin beta-10 accelerates apoptosis. Cell Mol Biol Res
cardiomyopathy. J Biol Chem 275: 22069 –22074, 2000. 41: 167–180, 1995.
93. ERICKSON-VITANEN S, RUGGIERI S, NATALINI P, AND HORECKER BL. 116. HANSON J AND LOWY J. The structure of F-actin and of actin filaments
Thymosin beta 10, a new analog of thymosin beta 4 in mammalian isolated from muscle. J Mol Biol 6: 46 – 60, 1963.
tissues. Arch Biochem Biophys 22: 570 –576, 1983. 117. HART MC, KORSHUNOVA YO, AND COOPER JA. Vertebrates have con-
94. ESTES JE, SELDEN LA, KINOSIAN HJ, AND GERSHMAN LC. Tightly-bound served capping protein alpha isoforms with specific expression
divalent cation of actin. J Muscle Res Cell Motil 13: 272–284, 1992. patterns. Cell Motil Cytoskeleton 38: 120 –132, 1997.
95. FEDOROV AA, LAPPALAINEN P, FEDOROV EV, DRUBIN DG, AND ALMO SC. 118. HATANAKA H, OGURA K, MORIYAMA K, ICHIKAWA S, YAHARA I, AND
Structure determination of yeast cofilin. Nat Struct Biol 4: 366 –369, INAGAKI F. Tertiary structure of destrin and structural similarity
1997. between 2 actin-regulating protein families. Cell 85: 1047–1055,
96. FEINBERG J, MERY J, HEITZ F, BENYAMIN Y, AND ROUSTAN C. Correla- 1996.
tions between biological activity and structural properties for two 119. HAWKINS M, POPE B, MACIVER SK, AND WEEDS AG. Human actin
short homologous sequences in thymosin beta4 and gelsolin. Int J depolymerizing factor mediates a pH-sensitive destruction of actin
Pept Protein Res 47: 62– 69, 1996. filaments. Biochemistry 32: 9985–9993, 1993.
97. FOWLER VM AND CONLEY CA. Tropomodulin. In: Guidebook to the 120. HAYDEN SM, MILLER PS, BRAUWEILER A, AND BAMBURG JR. Analysis of
Cytoskeletal and Motor Proteins, edited by Kreis T and Vale R. the interactions of actin depolymerizing factor with G- and F-actin.
Oxford, UK: Oxford Univ. Press, 1999, p. 154 –159. Biochemistry 32: 9994 –10004, 1993.
98. FOWLER VM, SUSSMAN MA, MILLER PG, FLUCHER BE, AND DANIELS PM. 121. HEGYI G, SZILAGYI L, AND BELAGYI J. Influence of the bound nucleo-
Tropomodulin is associated with the free (pointed) ends of the thin tide on the molecular dynamics of actin. Eur J Biochem 175:
filaments in rat skeletal muscle. J Cell Biol 120: 411– 420, 1993. 271–274, 1988.
99. FOWLER WE AND AEBI U. Polymorphism of actin paracrystals in- 122. HEINKE MY, WHEELER CH, YAN JX, AMIN V, CHANG D, EINSTEIN R,
duced by polylysine. J Cell Biol 93: 452– 458, 1982. DUNN MJ, AND DOS REMEDIOS CG. Changes in myocardial protein
100. FRANKEL S, CONDEELIS J, AND LEINWAND L. Expression of actin in expression in pacing-induced canine heart failure. Electrophoresis
Escherichia coli. J Biol Chem 265: 17980 –17987, 1990. 20: 2086 –2093, 1999.
101. FUJISAWA T, KOSTYUKOVA A, AND MAEDA Y. The shapes and sizes of 123. HEISS SG AND COOPER JA. Regulation of CapZ, an actin-capping
two domains of tropomodulin, the P-end-capping protein of actin- protein of chicken muscle, by anionic phospholipids. Biochemistry
tropomyosin. FEBS Lett 498: 67–71, 2001. 30: 8753– 8758, 1991.
102. FUJITA H, LAHAM LE, JANMEY PA, KWIATKOWSKI DJ, STOSSEL TP, BANNO 124. HELING A, ZIMMERMANN R, KOSTIN S, MAENO Y, HEIN S, DEVAUX B,
Y, NOZAWA Y, MULLAUER L, ISHIZAKI A, AND KUZUMAKI N. Functions of BAUER E, KLOVEKORN WP, SCHLEPPER M, SCHAPER W, AND SCHAPER J.

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


ACTIN AND ACTIN BINDING PROTEINS 469

Increased expression of cytoskeletal, linkage, and extracellular 146. KINOSIAN HJ, SELDEN LA, ESTES JE, AND GERSHMAN LC. Kinetics of
proteins in failing human myocardium. Circ Res 86: 846 – 853, 2000. gelsolin interaction with phalloidin-stablized F-actin. Rate con-
125. HERRMANNSDOERFER AJ, HEEB PJ, FEUSTEL PJ, ESTES JE, KEENAN CJ, stants for binding and severing. Biochemistry 35: 16550 –16556,
MINNEAR FL, SELDEN L, GIUNTA C, FLOR JR, AND BLUMENSTOCK FA. 1996.
Vascular clearance and organ uptake of G- and F-actin in the rat. 147. KINOSIAN HJ, SELDEN LA, ESTES JE, AND GERSHMAN LC. Nucleotide
Am J Physiol Gastrointest Liver Physiol 265: G1071–G1081, 1993. binding to actin. Cation dependence of nucleotide dissociation and
126. HIGGS HN AND POLLARD TD. Regulation of actin polymerization by exchange. J Biol Chem 268: 8683– 8691, 1993.
Arp2/3 complex and WASp/Scar proteins. J Biol Chem 274: 325531– 148. KOFFER A, EDGAR AJ, AND BAMBURG JR. Identification of two species
32534, 1999. of actin depolymerizing factor in cultures of BHK cells. J Muscle
127. HITCHCOCK SE. Actin DNase I interaction. Depolymerization and Res Cell Motil 9: 320 –328, 1988.
nucleotide exchange. J Biol Chem 255: 5668 –5673, 1980. 149. KORN ED, CARLIER MF, AND PANTALONI D. Actin polymerization and
128. HOLMES KC, POPP D, GEBHARD W, AND KABSCH W. Atomic model of ATP hydrolysis. Science 238: 638 – 644, 1987.
the actin filament. Nature 347: 44 – 49, 1990. 150. KOTHAKOTA S, AZUMA T, REINHARD C, KLIPPEL A, TANG J, CHU K,
129. HUFF T, ZERZAWY D, AND HANNAPPEL E. Interactions of beta-thy- MCGARRY TJ, KIRSCHNER MW, KOTHS K, KWIATKOWSKI DJ, AND WIL-
mosins, thymosin beta 4-sulfoxide, and N-terminally truncated thy- LIAMS LT. Caspase-3-generated fragment of gelsolin: effector of
mosin beta 4 with actin studied by equilibrium centrifugation, morphological change in apoptosis. Science 278: 294 –298, 1997.
chemical cross-linking and viscometry. Eur J Biochem 230: 650 – 151. KREIS T AND VALE R. (Editors). Guidebook to the Cytoskeletal and
657, 1995. Motor Proteins (2nd ed.). Oxford, UK: Oxford Univ. Press, 1999.
130. HUG C, MILLER TM, TORRES MA, CASELLA JF, AND COOPER JA. Identi- 152. KREUDER V, DIECKHOFF J, SITTIG M, AND MANNHERZ HG. Isolation,
fication and characterization of an actin-binding site of CapZ. J Cell characterisation and crystallization of deoxyribonuclease I from
Biol 116: 923–931, 1992. bovine and rat parotid gland and its interaction with rabbit skeletal
131. HUXLEY HE. Electron microscope studies on the structure of natural muscle actin. Eur J Biochem 139: 389 – 400, 1984.
and synthetic protein filaments from striated muscle. J Mol Biol 7: 153. KRIEGER I, KOSTYUKOVA AS, YAMASHITA A, NITANAI Y, AND MAEDA Y.
281–308, 1963. Crystal structure of the C-terminal half of tropomodulin and struc-
132. ICHETOVKIN I, HAN J, PANG KM, KNECHT DA, AND CONDEELIS JS. Actin tural basis of actin filament pointed-end capping. Biophys J 83:
filaments are severed by both native and recombinant Dictyoste- 2716 –2725, 2002.
lium cofilin but to different extents. Cell Motil Cytoskeleton 45: 154. KUSANO K, ABE H, AND OBINATA T. Detection of a sequence involved
293–306, 2000. in actin-binding in the N-terminal side of cofilin. Mol Cell Biochem
133. ISENBERG G. Cytoskeleton Proteins: A Purification Manual. Berlin: 190: 133–141, 1999.
Springer-Verlag, 1995. 155. KWIATECK O, PAPA II, LEBART MC, AND ROUSTAN C. Interaction of actin
134. JEAN C, RIEGER K, BLANCHOIN L, CARLIER MF, LENFANT M, AND PAN- with the capping protein, CapZ from sea bass (Dicentrarchus
TALONI D. Interaction of G-actin with thymosin beta 4 and its
labrx) white skeletal muscle. Comp Biochem Physiol B Comp
variants thymosin beta 9 and thymosin beta met9. J Muscle Res Cell Biochem 127: 551–562, 2000.
Motil 15: 278 –286, 1994.
156. KWIATKOWSKI DJ. Functions of gelsolin: motility, signaling, apopto-
135. JIANG L, TSUBAKIHARA M, HEINKE MY, YAO M, DUNN NJ, PHILLIPS W,
sis, cancer. Curr Opin Cell Biol 11: 103–108, 1999.
DOS REMEDIOS CG, AND NOSWORTHY NJ. Heart failure and apoptosis:
157. KWIATKOWSKI DJ, STOSSEL TP, ORKIN SH, MOLE JE, COLTEN H, AND YIN
electrophoretic methods support data from micro- and macro-
HL. Plasma and cytoplasmic gelsolins are encoded by a single gene
arrays. A critical review of genomics and proteomics. Proteomics 1:
and contain a duplicate actin-binding domain. Nature 323: 455– 458,
1481–1488, 2001.
1986.
136. JONES LJF, CARBALLIDO-LÓPEZ R, AND ERRINGTON J. Control of cell
158. LAPPALAINEN P, FEDOROV EV, FEDOROV AA, ALMO SC, AND DRUBIN DG.
shape in bacteria: helical, actin-like filaments in Bacillus subtilis.
Essential functions and actin-binding surfaces of yeast cofilin re-
Cell 104: 913–922, 2001.
137. KABSCH W, MANNHERZ HG, SUCK D, PAI EF, AND HOLMES KC. Atomic vealed by systematic mutagenesis. EMBO J 16: 5520 –5530, 1997.
structure of the actin-DNase I complex. Nature 347: 37– 44, 1990. 159. LAZARIDES E AND LINDBERG U. Actin is the naturally occurring inhib-
138. KAISER DA AND POLLARD TD. Characterization of actin and poly-L- itor of deoxyribonuclease I. Proc Natl Acad Sci USA 71: 4724 – 4746,
proline binding sites of Acanthamoeba profilin with monoclonal 1974.
antibodies and by mutagenesis. J Mol Biol 256: 89 –107, 1996. 160. LECHLER T, JONSDOTTIR GA, KLEE SK, PELLMAN D, AND LI R. A two-
139. KANGAS H, ULMANEN I, PAUNIO T, KWIATKOWSKI DJ, LEHTOVIRTA M, tiered mechanism by which Cdc42 controls the localization and
JALANKO A, AND PELTONEN L. Functional consequences of amyloid- activation of an Arp2/3-activating motor complex in yeast. J Cell
osis mutation for gelsolin polypeptide: analysis of gelsolin-actin Biol 155: 261–270, 2001.
interaction and gelsolin processing in gelsolin knock-out fibro- 161. LEONARD SA, GITTIS AG, PETRELLA EC, POLLARD TD, AND LATTMAN EE.
blasts. FEBS Lett 454: 233–239, 1999. Crystal structure of the actin-binding protein actophorin from
140. KASAI M, ASAKURA S, AND OOSAWA F. The cooperative nature of the Acanthamoeba. Nat Struct Biol 4: 369 –373, 1997.
G-F transformation of actin. Biochim Biophys Acta 57: 22–31, 1962. 162. LIAO TH. Bovine pancreatic deoxyribonuclease D. J Biol Chem 249:
141. KASAI M, NAKANO E, AND OOSAWA F. Polymerization of actin free 2354 –2356, 1974.
from nucleotides and divalent cations. Biochim Biophys Acta 94: 163. LIN KM, MEJILLANO M, AND YIN HL. Ca2" regulation of gelsolin by its
494 –503, 1965. C-terminal tail. J Biol Chem 275: 27746 –27752, 2000.
142. KEKIC M, NOSWORTHY NJ, DEDOVA I, COLLYER CA, AND DOS REMEDIOS 164. LITTLEFIELD R, ALMENAR-QUERALT A, AND FOWLER VM. Actin dynamics
CG. Regulation of the cytoskeleton assembly: a role for a ternary at pointed ends regulates thin filament length in striated muscle.
complex with two actin-binding proteins. Res Prob Cell Differ 32: Nat Cell Biol 3: 544 –551, 2001.
165–180, 2001. 165. LITTLEFIELD R AND FOWLER VM. Defining actin filament length in
143. KELLEHER JF, ATKINSON SJ, AND POLLARD TD. Sequences, structural striated muscle: rulers and caps or dynamic instability? Annu Rev
models, and cellular localization of the actin-related proteins Arp2 Cell Dev Biol 14: 487–525, 1998.
and Arp3 from Acanthamoeba. J Cell Biol 131: 385–397, 1995. 166. LOISEL TP, BOUJEMAA R, PANTALONI D, AND CARLIER MF. Reconstitu-
144. KIMURA A, HARADA H, PARK JE, NISHI H, SATOH M, TAKAHASHI M, HIROI tion of actin-based motility of Listeria and Shigella using pure
S, SASAOKA T, OHBUCHI N, NAKAMURA KOYANAGI T, HWANG THI, CHOO proteins. Nature 401: 613– 616, 1999.
JA, CHUNG KS, HASEGAWA A, NAGAI R, OKAZAKI O, NAKAMURA H, 167. LORENZ M, POPP D, AND HOLMES KC. Refinement of the F-actin model
MATSUZAKI M, SAKAMOTO T, TOSHIMA H, KOGA Y, IMAIZUMI T, AND against X-ray by the use of a direction mutation algorithm. J Mol
SASAZUKI T. Mutations in the cardiac troponin I gene associated Biol 234: 826 – 836, 1993.
with hypertrophic cardiomyopathy. Nat Genet 16: 379 –382, 1997. 168. LOW TL, HU SK, AND GOLDSTEIN AL. Complete amino acid sequence
145. KINOSIAN HJ, SELDEN LA, ESTES JE, AND GERSHMAN LC. Actin filament of bovine thymosin beta 4: a thymic hormone that induces terminal
annealing in the presence of ATP and phalloidin. Biochemistry 32: deoxynucleotidyl transferase activity in thymocyte populations.
12353–12357, 1993. Proc Natl Acad Sci USA 78: 1162–1166, 1981.

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


470 DOS REMEDIOS ET AL.

169. LU M, WITKE W, KWIATKOWSKI DJ, AND KOSIK KS. Delayed retraction the actin-monomer subunit interface to prevent polymerization.
of filopodia in gelsolin null mice. J Cell Biol 138: 1279 –1287, 1997. Nat Cell Biol 2: 376 –379, 2000.
170. LUECK A, YIN HL, KWIATKOWSKI DJ, AND ALLEN PG. Calcium regulation 194. MULLENS RD, STAFFORD WB, AND POLLARD TD. Structure subunit
of gelsolin and adseverin: a natural test of the helix latch hypoth- topology and actin-binding activity of the Arp2/3 complex from
esis. Biochemistry 39: 5274 –5279, 2000. Acanthamoeba. J Cell Biol 136: 331–343, 1997.
171. MACHESKY LM AND MAY RC. Arps: actin-related proteins. Res Prob 195. MULLINS RD, HEUSER JA, AND POLLARD TD. The interaction of Arp2/3
Cell Differ 32: 213–229, 2001. complex with actin: nucleation, high affinity pointed end capping,
172. MACIVER SK, TERNENT D, AND MCLAUGHLIN PJ. Domain 2 of gelsolin and formation of branching networks of filaments. Proc Natl Acad
binds directly to tropomyosin. FEBS Lett 473: 71–75, 2000. Sci USA 95: 6181– 6186, 1998.
173. MAEKAWA S, NISHIDA E, OHTA Y, AND SAKAI H. Isolation of low 196. NACHMIAS VT. Small actin-binding proteins: the beta-thymosin fam-
molecular weight actin-binding proteins from porcine brain. J Bio- ily. Curr Opin Cell Biol 5: 56 – 62, 1993.
chem 95: 377–385, 1984. 197. NAGAOKA R, ABE H, KUSANO K, AND OBINATA T. Concentration of
174. MALINDA KM, GOLDSTEIN AL, AND KLEINMAN HK. Thymosin 4 stimu- cofilin, a small actin-binding protein, at the cleavage furrow during
lates directional migration of human umbilical vein endothelial cytokinesis. Cell Motil Cytoskeleton 30: 1–7, 1995.
cells. FASEB J 11: 474 – 481, 1997. 198. NAGAOKA R, ABE H, AND OBINATA T. Site-directed mutagenesis of the
175. MANIOTIS AJ, CHEN S, AND INGBER DE. Demonstrations of mechanical phosphorylation site of cofilin: its role in cofilin-actin interaction
connections between integrins, cytoskeletal filaments, and nucleo- and cytoplasmic localization. Cell Motil Cytoskeleton 35: 200 –209,
plasm that stabilize nuclear structure. Proc Natl Acad Sci USA 94: 1996.
847– 854, 1997. 199. NAGAOKA R, MINAMI N, HAYAKAWA K, ABE H, AND OBINATA T. Quanti-
176. MANNHERZ HG, KABSCH W, AND LEBERMAN J. Crystals of skeletal tative analysis of low molecular weight G-actin-binding proteins,
muscle actin: pancreatic DNase I complex. FEBS Lett 73: 141–143, cofilin, ADF and profilin, expressed in developing and degenerating
1977. chicken skeletal muscles. J Muscle Res Cell Motil 17: 463– 473,
177. MANNHERZ HG, GOODY RS, KONRAD M, AND NOWAK E. The interaction 1996.
of bovine pancreatic deoxyribonuclease I and skeletal muscle ac- 200. NEIDL C AND ENGEL J. Exchange of ADP, ATP and 1,N-ethenoad-
tin. Eur J Biochem 104: 367–379, 1980. enosine 5'-triphosphate on G-actin. Equilibrium and kinetics. Eur
178. MARSTON SB AND HODGKINSON JL. Cardiac and skeletal myopathies: J Biochem 101: 163–169, 1979.
can genotype explain phenotype? J Muscle Res Cell Motil 22: 1– 4, 201. NEWMAN J, ZANER KL, SCHICK KL, GERSHMAN LC, SELDEN LA, KINOSIAN
2001. HG, TRAVIS JL, AND ESTES JE. Nucleotide exchange and rheometric
179. MARUYAMA K, KUROKAWA H, OOSAWA M, SHIMOKA S, YAMAMOTO H, ITO studies with F-actin prepared from ATP- or ADP-monomeric actin.
M, AND MARUYAMA K. Beta-actinin is equivalent to CapZ protein. Biophys J 64: 1559 –1566, 1993.
J Biol Chem 265: 8712– 8715, 1990. 202. NISHIDA E. Opposite effects of cofilin and profilin from porcine
180. MATSUDAIRA P, JAKES R, AND WALKER JE. A gelsolin-like Ca2"-depen- brain on rate of exchange of actin-bound adenosine 5'-triphos-
phate. Biochemistry 24: 1160 –1164, 1985.
dent actin-binding domain in villin. Nature 315: 248 –250, 1985.
203. NISHIDA E, IIDA K, YONEZAWA N, KOYASU S, YAHARA I, AND SAKAI H.
181. MCGOUGH A, CHIU W, AND WAY M. Determination of the gelsolin
Cofilin is a component of intranuclear and cytoplasmic actin rods
binding site on F-actin: implications for severing and capping.
induced in cultured cells. Proc Natl Acad Sci USA 84: 5262–5266,
Biophys J 74: 764 –772, 1998.
1987.
182. MCGOUGH A, POPE B, CHIU W, AND WEEDS A. Cofilin changes the
204. NIWA R, NAGATA-OHASHI K, TAKEICHI M, MIZUNO K, AND UEMURA T.
twist of F-actin: implications for actin filament dynamics and cel-
Control of actin reorganization by Slingshot, a family of phospha-
lular function. J Cell Biol 138: 771–781, 1997.
tases that dephosphorylate ADF/cofilin. Cell 108: 233–246, 2002.
183. MCGOUGH A, POPE B, AND WEEDS A. The ADF/cofilin family: accel-
205. NOSWORTHY NJ, KEKIC M, AND DOS REMEDIOS CG. The affinity of chick
eration of actin reorganization. Res Prob Cell Differ 32: 135–154,
cofilin for actin increases when actin is complexed with DNase I:
2001. formation of a cofilin-actin-DNase I ternary complex. Proteomics 1:
184. MCGRATH JL, OSBORN EA, TARDY YS, DEWEY CF JR, AND HARTWIG JH. 1513–1519, 2001.
Regulation of the actin cycle in vivo by actin filament severing. Proc 206. OBINATA T, NAGAOKA-YASUDA R, ONO S, KUSANO K, OHTAKA Y, YA-
Natl Acad Sci USA 97: 6532– 6537, 2000. MASHIRO S, OKAADA K, AND ABE H. Low molecular weight G-actin
185. MCKIM KS, MATHESON C, MARRA MA, WAKARCHUK MF, AND BAILLIE DL. binding proteins involved in the regulation of actin assembly during
The Caenorhabditis elegans Unc-60 gene encodes proteins homol- myofibrillogenesis. Cell Struct Funct 22: 181–189, 1997.
ogous to a family of actin-binding proteins. Mol Gen Genet 242: 207. ODA T, MAKINO K, YAMASHITA I, NAMBA K, AND MAEDA Y. The helical
346 –357, 1994. parameters of F-actin precisely determined from X-ray fiber diffrac-
186. MCLAUGHLIN PJ, GOOCH JT, MANNHERZ HG, AND WEEDS AG. Structure tion of well-oriented gels. Res Prob Cell Differ 32: 43–58, 2001.
of gelsolin segment-1-actin complex and the mechanism of filament 208. OGAWA Y, TASHIMA M, YUMOTO Y, OKUDA SAWADA T, H, OKUMA M, AND
severing. Nature 364: 685– 692, 1993. MARUYAMA Y. Coding sequence of human placenta cofilin cDNA.
187. MIZUNO K, OKANO I, OHASHI K, NUNOUE K, KUMA K, MIYATA T, AND Nucleic Acid Res 18: 7169, 1990.
NAKAMURA T. Identification of a human cDNA clone encoding a 209. OLSON TM, MICHELS VV, THIBODEAU SN, TAI YS, AND KEATING MT.
novel protein kinase with two repeats of the LIM/double zinc finger Actin mutations in dilated cardiomyopathy, a heritable form of
motif. Oncogene 9: 1605–1612, 1994. heart failure. Science 280: 750 –752, 1998.
188. MOORE PB, HUXLEY HE, AND DEROSIER DJ. Three-dimensional recon- 210. OOSAWA F AND ASAKURA S. Thermodynamics of the Polymerization
struction of F-actin, thin filaments and decorated thin filaments. J of Protein. New York: Academic, 1975.
Mol Biol 50: 279 –295, 1970. 211. OOSAWA F, ASAKURA S, HOTTA K, AND OOI T. G-F transformation of
189. MOORE S. Pancreatic DNase. Enzymes 14: 281–296, 1981. actin as a fibrous condensation. J Polymer Sci 37: 323–326, 1959.
190. MORGAN TE, LOCKERBIE RO, MINAMIDE LS, BROWNING MD, AND BAM- 212. ORIOL C, DUBORD C, AND LANDON F. Crystallization of native striated-
BURG JR. Isolation and characterization of a regulated form of actin muscle actin. FEBS Lett 73: 89 –91, 1977.
depolymerizing factor. J Cell Biol 122: 623– 633, 1993. 213. ORLOVA A, CHEN X, RUBENSTEIN PA, AND EGELMAN EH. Modulation of
191. MORIYAMA K, NISHIDA E, YONEZAWA N, SAKAI H, MATSUMOTO S, IIDA K, yeast F-actin structure by a mutation in the nucleotide-binding
AND YAHARA I. Destrin, a mammalian actin-depolymerizing protein, cleft. J Mol Biol 271: 235–243, 1997.
is closely related to cofilin. Cloning and expression of porcine brain 214. ORLOVA A AND EGELMAN EH. Structural dynamics of F-actin. I.
destrin cDNA. J Biol Chem 265: 5768 –5773, 1990. Changes in the C terminus. J Mol Biol 245: 582–597, 1995.
192. MORIYAMA K, YONEZAWA N, SAKAI H, YAHARA I, AND NISHIDA E. Muta- 215. OSBORNE M AND WEBER K. (Editors). Cytoskeletal Proteins in Tu-
tional analysis of an actin-binding site of cofilin and characteriza- mor Diagnosis. Current Communications. Cold Spring Harbor,
tion of chimeric proteins between cofilin and destrin. J Biol Chem NY: Cold Spring Harbor, 1989.
267: 7240 –7244, 1992. 216. OTTERBEIN L, GRACEFFA P, AND DOMINGUEZ R. The crystal structure of
193. MORTON WM, AYSCOUGH KR, AND MCLAUGHLIN PJ. Latrunculin alters G-actin (Abstract). Biophys J 80: 172a, 2001.

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


ACTIN AND ACTIN BINDING PROTEINS 471

217. OTTERBEIN L, GRACEFFA P, AND DOMINGUEZ R. The crystal structure of ternary complex of DNase I, actin and thymosin beta4. FEBS Lett
uncomplexed actin in the ADP state. Science 293: 708 –711, 2001. 387: 132–136, 1996.
218. OWEN C AND DEROSIER D. A 13-angstrom map of the actin-scruin 240. RESSAD F, DIDRY D, XIA GX, HONG Y, CHUA NH, PANTALONI D, AND
filament from the Limulus acrosomal process. J Cell Biol 123: CARLIER MF. Kinetic analysis of the interaction of actin-depolymer-
337–344, 1993. izing factor (ADF)/cofilin with G- and F-actins. J Biol Chem 273:
219. PANTALONI D AND CARLIER MF. How profilin promotes actin filament 20894 –20902, 1998.
assembly in the presence of thymosin beta 4. Cell 75: 1007–1014, 241. RICKARD JE AND SHETERLINE P. Cytoplasmic concentrations of inor-
1993. ganic phosphate affect the critical concentrations for actin assem-
220. PANTALONI D, LE CLAINCHE C, AND CARLIER MF. Mechanism of actin- bly in the presence of cytochalasin D or ADP. J Mol Biol 191:
based motility. Science 292: 1502–1506, 2001. 273–280, 1986.
221. PARDEE JD, SIMPSON PA, STRYER L, AND SPUDICH JA. Actin filaments 242. RIDLEY A. What initiates actin polymerization? Genome Biol 1: 1– 6,
undergo limited exchange at physiological salt conditions. J Cell 2000.
Biol 94: 316 –324, 1982. 243. ROBINSON RC, MEJILLANO M, LE VP, BURTNICK LD, YIN HL, AND CHOE
222. PAUDEL HK AND LIAO TH. Comparison of the three primary struc- S. Domain movement in gelsolin: a calcium-activated switch. Sci-
tures of deoxyribonuclease isolated from bovine, ovine, and por- ence 286: 1939 –1942, 1999.
cine pancreas. Derivation of the amino acid sequence of ovine 244. ROBINSON RC, TURBEDSKY K, KAISER DA, MARCHAND JB, HIGGS HN,
DNase and revision of the previously published amino acid se- CHOE S, AND POLLARD TD. Crystal structure of Arp2/3 complex.
quence of bovine DNase. J Biol Chem 261: 16012–16017, 1986. Science 294: 1679 –1684, 2001.
223. PEITSCH MC, POLZAR B, STEPHAN H, CROMPTON T, MACDONALD HR, 245. RODAL AA, TRETREAULT JW, LAPPALAINEN P, DRUBIN DG, AND AMBERG
MANNHERZ HG, AND TSCHOPP J. Characterization of the endogenous DC. Aip 1p interacts with cofilin to disassemble actin filaments.
deoxyribonuclease involved in nuclear DNA degradation during J Cell Biol 145: 1251–1264, 1999.
apoptosis (programmed cell death). EMBO J 12: 371–377, 1993. 246. RUBENSTEIN PA. The functional importance of multiple actin iso-
224. PERELROIZEN I, CARLIER MF, AND PANTALONI D. Binding of divalent forms. Bioessays 12: 309 –315, 1990.
cation and nucleotide to G-actin in the presence of profilin. J Biol 247. SAFER D AND CHOWRASHI PK. Beta-thymosins from marine inverte-
Chem 270: 1501–1508, 1995. brates: primary structure and interaction with actin. Cell Motil
225. PERELROIZEN I, DIDRY D, CHRISTENSEN H, CHUA NH, AND CARLIER MF. Cytoskeleton 38: 163–171, 1997.
Role of nucleotide exchange and hydrolysis in the function of 248. SAFER D, ELZINGA M, AND NACHMIAS VT. Thymosin beta 4 and Fx, an
profilin in actin assembly. J Biol Chem 271: 12302–12309, 1996. actin-sequestering peptide, are indistinguishable. J Biol Chem 266:
226. POETTER K, JIANG H, HASSANZADEH S, MASTER SR, CHANG A, DALAKAS 4029 – 4032, 1991.
MC, RAYMENT I, SELLERS JR, FANANAPAZIR L, AND EPSTEIN ND. Muta- 249. SAFER D AND NACHMIAS VT. ! Thymosins as actin binding peptides.
tions in either the essential or regulatory light chains of myosin are Bioessays16: 473– 479, 1994.
associated with a rare myopathy in human heart and skeletal 250. SAFER D AND NACHMIAS VT. Thymosin !4. In: Guidebook to the
muscle. Nat Genet 13: 63– 69, 1996.
Cytoskeletal and Motor Proteins, edited by Kreis T and Vale R.
227. POLLARD TD. Polymerization of ADP-actin. J Cell Biol 99: 769 –777,
Oxford, UK: Oxford Univ. Press, 1999, p. 149 –151.
1984.
251. SAFER D, SOSNICK TR, AND ELZINGA M. Thymosin beta 4 binds actin
228. POLLARD TD. Rate constants for the reactions of ATP- and ADP-
in an extended conformation and contacts both the barbed and
actin with the ends of actin filaments. J Cell Biol 103: 2747–2754,
pointed ends. Biochemistry 36: 5806 –5816, 1997.
1986.
252. SAKIYAMA S, FUJIMURA S, AND SAKIYAMA H. Absence of gamma-isoac-
229. POLLARD TD. Introduction. In: Guidebook to the Cytoskeletal and
tin expression in the mouse fibroblast cell line. J Biol Chem 256:
Motor Proteins, edited by Kreis T and Vale R. Oxford, UK: Oxford
31–33, 1981.
Univ. Press, 1999, p. 3–11.
253. SATOH M, TAKAHASHI M, SAKAMOTO T, HIROE M, MARUMO F, AND
230. POLLARD TD. Profilins. In: Guidebook to the Cytoskeletal and Motor
Proteins, edited by Kreis T and Vale R. Oxford, UK: Oxford Univ. KIMURA A. Structural analysis of the titin gene in hypertrophic
Press, 1999, p. 117–120. cardiomyopathy: identification of a novel disease gene. Biochem
231. POLLARD TD AND COOPER JA. Actin and actin-binding proteins: a Biophys Res Commun 262: 411– 417, 1999.
critical evaluation of mechanisms and function. Annu Rev Biochem 254. SCHAFER DA AND COOPER JA. Control of actin assembly at filament
55: 987–1035, 1986. ends. Annu Rev Cell Dev Biol 11: 497–518, 1995.
232. POPE B, MACIVER S, AND WEEDS A. Localization of the calcium- 255. SCHAFER DA, KORSHUNOVA YO, SCHROER TA, AND COOPER JA. Differ-
sensitive actin monomer binding site in gelsolin to segment 4 and ential localization and sequence analysis of capping protein beta-
identification of calcium binding sites. Biochemistry 34: 1583–1588, subunit isoforms of vertebrates. J Cell Biol 127: 453– 465, 1994.
1995. 256. SCHEVZOV G, LLOYD C, HAILSTONES D, AND GUNNING P. High level
233. POPP D, LEDNEV VV, AND JAHN W. Methods of preparing well oriented expression of transfected beta- and gamma-actin genes differen-
sols of F-actin containing filaments suitable for X-ray diffraction. J tially impacts on myoblasts cytoarchitecture. J Cell Biol 117: 775–
Mol Biol 197: 679 – 684, 1987. 785, 1993.
234. PUIUS YA, FEDOROV EV, EICHINGER L, SCHLEICHER M, AND ALMO SC. 257. SCHOEPPER B AND WEGNER A. Rate constants and equilibrium con-
Mapping the functional surface of domain 2 in the gelsolin super- stants for binding of actin to the 1:1 gelsolin-actin complex. Eur
family. Biochemistry 39: 5322–5331, 2000. J Biochem 202: 1127–1131, 1991.
235. QUINLAN R, HUTCHISON C, AND LANE B (Editors). Intermediate Fila- 258. SCHUTT CE, LINDBERG U, MYSLIK J, AND STRAUSS N. Molecular packing
ment Proteins (3rd ed.). Oxford, UK: Oxford Univ. Press. In press. in profilin:actin crystals and its implications. J Mol Biol 209: 735–
236. QUIRK S, MACIVER SK, AMPE C, DOBERSTEIN SK, KAISER DA, VANDAMME 746, 1989.
J, VANDEKERCKHOVE JS, AND POLLARD TD. Primary structure of and 259. SCHUTT CE, MYSLIK JC, ROZYCKI MD, GOONESKERE N, AND LINDBERG U.
studies on Acanthamoeba actophorin. Biochemistry 32: 8525– The structure of crystalline profilin-beta-actin. Nature 365: 810 –
8533, 1993. 816, 1993.
237. RATNASWAMY G, KOEPF E, BEKELE H, YIN HL, AND KELLY JW. The 260. SEGURA M AND LINDBERG U. Separation of non-muscle isoactins in
amyloidogenicity of gelsolin is controlled by proteolysis and pH. the free form or as profilin complexes. J Biol Chem 259: 3949 –3954,
Chem Biol 6: 293–304, 1999. 1984.
238. REDWOOD C, LOHMANN K, BING W, ESPOSITO GM, ELLIOTT K, ABDUL- 261. SELDEN LA, ESTES JA, AND GERSHMAN LC. High affinity divalent
RAZZAK H, KNOTT A, PURCELL I, MARSTON S, AND WATKINS H. Investi- cation binding to actin. Effect of low affinity salt binding. J Biol
gation of a truncated cardiac troponin T that causes familial hy- Chem 264: 9271–9277, 1989.
pertrophic cardiomyopathy: Ca2" regulatory properties of recon- 262. SELDEN LA, KINOSIAN HJ, NEWMAN J, LINCOLN B, HURWITZ C, GERSHMAN
stituted thin filaments depend on the ratio of mutant to wild-type LC, AND ESTES JE. Severing of F-actin by the amino-terminal half of
protein. Circ Res 86: 1146 –1152, 2000. gelsolin suggests internal cooperativity in gelsolin. Biophys J 75:
239. REICHERT A, HEINTZ D, ECHNER H, VOELTER W, AND FAULSTICH H. The 3092–3090, 1998.

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


472 DOS REMEDIOS ET AL.

263. SELLERS J. Myosin. In: Protein Profile (2nd ed.), edited by Sheterline ulatory proteins to mixed lipid bilayers. Eur J Biochem 263: 85–92,
P. Oxford, UK: Oxford Univ. Press, 1999. 1999.
264. SELVE N AND WEGNER A. Rate of treadmilling of actin filaments in 286. VANCOMPERNOLLE K, GOETHALS M, HUET C, LOUVARD D, AND VANDE-
vitro. J Mol Biol 187: 627–231, 1986. KERCKHOVE J. G- to F-actin modulation by a single amino acid
265. SEPT D, XU J, POLLARD TD, AND MCCAMMON JA. Annealing accounts substitution in the actin binding site of actobindin and thymosin
for the length of actin filaments formed by spontaneous polymer- beta 4. EMBO J 11: 4739 – 4746, 1992.
ization. Biophys J 77: 2911–2919, 1999. 287. VANCOMPERNOLLE K, VANDEKERCKHOVE J, BUBB MR, AND POLLARD TD.
266. SHETERLINE P, CLAYTON J, AND SPARROW JC (Editors). Actin. In: The interfaces of actin and Acanthamoeba actobindin. J Biol Chem
Protein Profile (4th ed.). Oxford, UK: Academic, 1998. 266: 15427–15431, 1991.
267. SIMENEL C, VAN TROYS M, VANDEKERCKHOVE J, AMPE C, AND DELE- 288. VANDEKECKHOVE J AND WEBER K. At least six different actins are
PIERRE M. Structural requirements for thymosin beta4 in its contact expressed in a higher mammal: an analysis based on the amino acid
with actin. An NMR-analysis of thymosin beta4 mutants in solution sequence of the amino-terminal tryptic peptide. J Mol Biol 126:
and correlation with their biological activity. Eur Biophys J 267: 783– 802, 1978.
3530 –3538, 2000. 289. VAN DEN ENT F, AMOS LA, AND LÖWE J. Prokaryotic origin of the actin
268. SOUTHWICK FS. Gelsolin and ADF/cofilin enhance the actin dynam- cytoskeleton. Nature 413: 39 – 44, 2001.
ics of motile cells. Proc Natl Acad Sci USA 97: 6936 – 6938, 2000. 290. VAN TROYS M, DEWITTE D, VERSCHELDE JL, GOETHALS M, VANDEKERCK-
269. STANTON LW, GARRARD LJ, DAMM D, GARRICK BL, LAM A, KAPOUN AM, HOVE J, AND AMPE C. Analogous F-actin binding by cofilin and
ZHENG Q, PROTTER AA, SCHREINER GF, AND WHITE RT. Altered pat- gelsolin segment 2 substantiates their structural relationship. J Biol
terns of gene expression in response to myocardial infarction. Circ Chem 272: 32750 –32758, 1997.
Res 86: 939 –945, 2000. 291. VARTIAINEN MK, MUSTONEN T, MATTILA P, OJALA PJ, THESLEFF MK,
270. STRZELECKA-GOLASZEWSKA H. Divalent cations, nucleotides and actin PARTANEN J, AND LAPPALAINEN P. The three mouse actin-depolymer-
structure. Res Prob Cell Differ 32: 23– 42, 2001. izing factor/cofilins evolved to fulfil cell-type-specific requirements
271. STRZELECKA-GOLASZEWSKA H, PROCHNIEWICZ E, AND DRABIKOWSKI W. for actin dynamics. Mol Biol Cell 13: 183–194, 2002.
Interaction of actin with divalent cations. I. The effect of various 292. VARTIAINEN N, PYYKONEN I, HOKFELT T, AND KOISTINAHO J. Induction of
cations on the physical state of actin. Eur J Biochem 88: 219 –227, thymosin !(4) mRNA following focal brain ischemia. Neuroreport
1978. 7: 1613–1616, 1996.
272. SUCK W, OEFNER C, AND KABSCH W. Three-dimensional structure of 293. VOLKMANN N, AMANN KJ, STOILOVA-MCPHIE S, EGILE C, WINTER DC,
bovine pancreatic DNase I at 2.5 Å resolution. EMBO J 3: 2423– HAZELWOOD L, HEUSER JE, LI R, POLLARD TD, AND HANEIN D. Structure
2430, 1984. of Arp2/3 complex in its activated state and in actin filament branch
273. SUN HQ, KWIATKOWSKA K, AND YIN HL. !-Thymosins are not simple junctions. Science 293: 2456 –2459, 2001.
actin monomer buffering proteins. Insights from overexpression 294. WANG F, SAMPOGNA RV, AND WARE B. pH dependance of actin
studies. J Biol Chem 271: 9223–9230, 1996. self-assembly. Biophys J 55: 293–298, 1989.
274. SUN HQ, LIN KM, AND YIN HL. Gelsolin modulates phospholipase C 295. WATAKABE A, KOBAYASHI R, AND HELFMAN DM. N-tropomodulin: a
activity in vivo through phospholipid binding. J Cell Biol 13: 811– novel isoform of tropomodulin identified as the major binding
820, 1997. protein to brain tropomyosin. J Cell Sci 109: 2299 –2309, 1996.
275. SUN HQ, WOOTEN DC, JANMEY PA, AND YIN HL. The actin side-binding 296. WATKINS H, CONNER D, THIERFELDER L, JARCHO JA, MACRAE C, MC-
domain of gelsolin also caps actin filaments. Implications for actin KENNA WJ, MARON BJ, SEIDMAN JG, AND SEIDMAN CE. Mutations in the
filament severing. J Biol Chem 35: 5229 –5237, 1994. cardiac myosin binding protein-C gene on chromosome 11 cause
276. SUNG LA, FAN Y, AND LIN CC. Gene assignment, expression, and familial hypertrophic cardiomyopathy. Nat Genet 11: 434 – 437,
homology of human tropomodulin. Genomics 34: 92–96, 1996. 1995.
277. SUSSMAN MA, BAQUE S, UHM CS, DANIELS MP, PRICE RL, SIMPSON D, 297. WEBER A, PENNISE CR, BABCOCK GG, AND FOWLER VM. Tropomodulin
TERRACIO L, AND KEDES L. Altered expression of tropomodulin in caps the pointed ends of actin filaments. J Cell Biol 127: 1627–1635,
cardiomyocytes disrupts the sarcomeric structure of myofibrils. 1994.
Circ Res 82: 94 –105, 1998. 298. WEBER A, PENNISE CR, AND PRING M. DNase I increases the rate
278. SUSSMAN MA, WELCH S, GUDE N, KHOURY PR, DANIELS SR, KIRKPATRICK constant of depolymerization at the pointed ($) end of actin fila-
D, WALSH RA, PRICE RL, LIM HW, AND MOLKENTIN JD. Pathogenesis of ments. Biochemistry 33: 4780 – 4786, 1994.
dilated cardiomyopathy: molecular, structural, and population 299. WEETMAN AP. Graves’ disease. N Engl J Med 343: 1236 –1248, 2000.
analyses in tropomodulin-overexpressing transgenic mice. Am J 300. WEGNER A. Head to tail polymerization of actin. J Mol Biol 108:
Pathol 155: 2101–2113, 1999. 139 –150, 1976.
279. SUSSMAN MA, WELCH S, WALKER A, KLEVITSKY R, HEWETT TE, WITT SA, 301. WEGNER A AND ISENBERG G. Twelvefold difference between the
KIMBALL TR, PRICE R, LIM HW, AND MOLKENTIN JD. Hypertrophic critical concentrations of the two ends of actin filaments in phys-
defect unmasked by calcineurin expression in asymptomatic tro- iological salt conditions. Proc Natl Acad Sci USA 80: 4922– 4925,
pomodulin overexpressing transgenic mice. Cardiovasc Res 46: 1983.
17–19, 2000. 302. WELCH MD, IWAMATSU A, AND MICHISON TJ. Actin polymerization is
280. TAGAWA H, WANG N, NARISHIGE T, INGBER DE, ZILE MR, AND COOPER induced by Arp2/3 protein complex at the surface of Listeria
G. Cytoskeletal mechanics in pressure-overload cardiac hypertro- monocytogenes. Nature 385: 265–269, 1997.
phy. Circ Res 80: 281–289, 1997. 303. WITKE W, SHARPE AH, HARTWIG JH, AZUMA T, STOSSEL TP, AND KWI-
281. TEUBNER A AND WEGNER A. Kinetic evidence for a readily exchange- ATKOWSKI DJ. Haemostatic, inflammatory, and fibroblast responses
able nucleotide at the terminal subunit of the barbed ends of actin are blunted in mice lacking gelsolin. Cell 81: 41–51, 1995.
filaments. Biochemistry 37: 7532–7538, 1998. 304. WRIGGERS W, TANG JX, AZUMA T, MARKS PW, AND JANMEY PA. Cofilin
282. THIERFELDER L, WATKINS H, MACRAE C, LAMAS R, MCKENNA W, VOS- and gelsolin segment-1: molecular dynamics simulation and bio-
BERG HP, SEIDMAN JG, AND SEIDMAN CE. Alpha-tropomyosin and chemical analysis predict a similar actin binding mode. J Mol Biol
cardiac troponin T mutations cause familial hypertrophic cardio- 282: 921–932, 1998.
myopathy: a disease of the sarcomere. Cell 77: 701–712, 1994. 305. XU J, CASELLA JF, AND POLLARD TD. Effect of capping protein, CapZ,
283. TIRION MM, BENAVRAHAM D, LORENZ M, AND HOLMES KC. Normal on the length of actin filaments and mechanical preperties of actin
modes as refinement parameters for the F-actin model. Biophys J filaments networks. Cell Motil Cytoskeleton 42: 73– 81, 1999.
68: 5–12, 1995. 306. YANG J, MORAVEC CS, SUSSMAN MA, DIPAOLA NR, FU D, HAWTHORN L,
284. TSUTSUI H, ISHIHARA K, AND COOPER G. Cytoskeletal role in the MITCHELL CA, YOUNG JB, FRANCIS GS, MCCARTHY PM, AND BOND M.
contractile dysfunction of hypertrophied myocardium. Science 260: Decreased SLIM1 expression and increased gelsolin expression in
682– 687, 1993. failing human hearts measured by high-density oligonucleotide
285. TUOMINEN EKJ, HOLOPAINEN JM, CHEN J, PRESTWICH MACHILLER GD, arrays. Circulation 102: 3046 –3052, 2000.
KINNUNEN PR, PKJ, AND JANMEY PA. Fluorescent phosphoinositide 307. YANG N, HIGUCHI O, OHASHI K, NAGATA K, WADA A, KANGAWA K,
derivatives reveal specific binding of gelsolin and other actin reg- NISHIDA E, AND MIZUNO K. Cofilin phosphorylation by LIM-kinase 1

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org


ACTIN AND ACTIN BINDING PROTEINS 473

and its role in rac-mediated actin reorganization. Nature 393: 809 – 313. YU FX, JOHNSTON PA, SUDHOF TC, AND YIN HL. gCap39, a calcium ion-
812, 1998. and polyphosphoinositide-regulated actin capping protein. Science
308. YONEZAWA N, NISHIDA E, IIDA K, KUMAGAI H, YAHARA I, AND SAKAI H. 250: 1413–1415, 1990.
Inhibition of actin polymerization by a synthetic dodecapeptide 314. YU FX, LIN SC, MORRISON-BOGORAD M, AND YIN HL. Effects of thy-
patterned on the sequence around the actin-binding site of cofilin. mosin ! 4 and thymosin beta 10 on actin structures in living cells.
J Biol Chem 266: 10485–10499, 1991. Cell Motil Cytoskeleton 27: 13–25, 1994.
309. YONEZAWA N, NISHIDA E, IIDA K, YAHARA I, AND SAKAI H. Inhibition of 315. ZALEVSKY J, LEMPERT L, KRANITZ H, AND MULLINS RD. Different WASP
family proteins stimulate different Arp2/3 complex-dependent ac-
the interactions of cofilin, destrin, and deoxyribonuclease I with
tin-nucleating activities. Curr Biol 11: 1903–1913, 2001.
actin by phosphoinositides. J Biol Chem 265: 8382– 8386, 1990.
316. ZALLEN JA, COHEN Y, HUDSON AM, COOLEY L, WIESCHAUS E, AND
311. YONEZAWA N, NISHIDA E, KOYASU S, MAEKAWA S, OHTA Y, YAHARA I, AND SCHEJTER ED. SCAR is a primary regulator of Arp2/3-dependent
SAKAI H. Distribution among tissues and intracellular localization of morphological events in Drosophila. J Cell Biol 156: 689 –701,
cofilin, a 21 kDa actin-binding protein. Cell Struct Funct 12: 443– 2002.
452, 1987. 317. ZALVIDE JB, ALVAREZ CV, VIDAL A, DIEGUEZ C, VEGA FV, AND
312. YONEZAWA N, NISHIDA E, AND SAKAI H. pH control of actin polymer- DOMINGUEZ F. Regulation of thymosin beta 4 mRNA levels during
ization by cofilin. J Biol Chem 260: 14410 –14412, 1985. cell proliferation. Cell Prolif 28: 85–91, 1995.

Physiol Rev • VOL 83 • APRIL 2003 • www.prv.org

You might also like