Please cite this article in press as: Kremneva et al., Cofilin-2 Controls Actin Filament Length in Muscle Sarcomeres, Developmental Cell (2014), http://dx.doi.org/10.1016/j.devcel.2014.09.002

Developmental Cell

Article

Cofilin-2 Controls Actin Filament Lengthin Muscle SarcomeresElena Kremneva,1 Maarit H. Makkonen,1 Aneta Skwarek-Maruszewska,1 Gergana Gateva,1 Alphee Michelot,2

Roberto Dominguez,3 and Pekka Lappalainen1,*1Institute of Biotechnology, University of Helsinki, P.O. Box 56, 00014 Helsinki, Finland2Laboratoire de Physiologie Cellulaire et Vegetale, Institut de Recherches en Technologies et Sciences pour le Vivant, iRTSV,CNRS/CEA/INRA/UJF, 38054 Grenoble, France3Department of Physiology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA 19104, USA

*Correspondence: pekka.lappalainen@helsinki.fi

http://dx.doi.org/10.1016/j.devcel.2014.09.002

SUMMARY

ADF/cofilins drive cytoskeletal dynamics by promot-ing the disassembly of ‘‘aged’’ ADP-actin filaments.Mammals express several ADF/cofilin isoforms, buttheir specific biochemical activities and cellularfunctions have not been studied in detail. Here, wedemonstrate that the muscle-specific isoform cofi-lin-2 promotes actin filament disassembly in sarco-meres to control the precise length of thin filamentsin the contractile apparatus. In contrast to other iso-forms, cofilin-2 efficiently binds and disassemblesbothADP- andATP/ADP-Pi-actin filaments.Wemap-ped surface-exposed cofilin-2-specific residuesrequired for ATP-actin binding and propose thatthese residues function as an ‘‘actin nucleotide-statesensor’’ among ADF/cofilins. The results suggestthat cofilin-2 evolved specific biochemical andcellular properties that allow it to control actin dy-namics in sarcomeres, where filament pointed endsmay contain amixture of ADP- andATP/ADP-Pi-actinsubunits. Our findings also offer a rationale for whycofilin-2 mutations in humans lead to myopathies.

INTRODUCTION

The myofibrils of striated muscle cells are composed of contrac-

tile units called sarcomeres, where contractile force is generated

by the sliding of thin actin filaments past thick, bipolar myosin fil-

aments. The actin filaments in sarcomeres are uniform in length

and precisely oriented with their barbed ends facing the Z-disc

and their pointed ends directed toward the M-band (Huxley,

1963). Whereas actin filaments in nonmuscle cells typically un-

dergo rapid turnover, the paracrystalline actin filament arrays

of muscle sarcomeres were traditionally considered to be very

stable. However, recent studies have provided evidence that

also sarcomeric actin filaments undergo dynamic turnover,

involving at least two distinct mechanisms. A subset of actin fil-

aments in sarcomeres was shown to undergo rapid turnover as a

result of myosin contractility (Skwarek-Maruszewska et al.,

2009). Additionally, the length of actin filaments in mature sarco-

meres is precisely regulated by actin subunit exchange at fila-

De

DEVCEL

ment barbed and pointed ends (Littlefield et al., 2001; Gokhin

and Fowler, 2013). Because, tropomodulin that caps the pointed

end is more dynamically associated with the filament than CapZ

that caps the barbed end, subunit exchange in sarcomeres is

faster at pointed ends (Littlefield et al., 2001; Cooper and Sept,

2008). This contrasts with nonmuscle cells where actin dynamics

occur mainly via addition of actin monomers to free barbed ends

and severing of ‘‘aged’’ filaments toward their pointed ends

(Pollard and Cooper, 2009).

Actin filaments in sarcomeres are additionally decorated

along their length by tropomyosin (Yamashiro et al., 2012) and

a large number of actin-binding proteins, which contribute

to maintaining sarcomere structure and organization. These

include the giant protein titin that provides an elastic link be-

tween Z-discs and M-bands, nebulin, which may regulate the

length of sarcomeric actin filaments and the actin nucleating

protein leiomodin (Labeit and Kolmerer, 1995; Agarkova and

Perriard, 2005; Chereau et al., 2008; Lange et al., 2006; Pappas

et al., 2010; Sanger and Sanger, 2008). Additionally, many ubiq-

uitous actin-binding proteins, which regulate actin dynamics in

nonmuscle cells, are also expressed in muscles. In most cases,

muscle-specific isoforms of these proteins have evolved to fulfill

the specific requirements of actin dynamics in muscle cells. In

mammals, these include the muscle-specific formin (FHOD3)

(Taniguchi et al., 2009; Iskratsch et al., 2010), cyclase-associ-

ated protein (CAP2) (Bertling et al., 2004), twinfilin-2b (Neva-

lainen et al., 2009), and cofilin-2 (Ono, 2010). However, the

specific biochemical properties distinguishing muscle from

nonmuscle cytoskeletal protein isoforms are largely unknown.

Defects in the regulation of the length or organization of actin fil-

aments in sarcomeres due to mutations or misregulated expres-

sion of cytoskeletal proteins are a hallmark of many heart and

skeletal muscle disorders, including several myopathies (Kho

et al., 2012).

ADF/cofilins are small (�15–20 kDa), ubiquitous actin-binding

proteins, which are essential regulators of actin dynamics in all

eukaryotes from yeasts to animals (Bernstein and Bamburg,

2010). ADF/cofilins bind preferentially to ‘‘aged’’ ADP-actin fila-

ments and promote actin dynamics by severing the actin fila-

ments, thereby replenishing the pool of monomers for new

rounds of filament assembly (Andrianantoandro and Pollard,

2006; Michelot et al., 2007). Cooperative binding of ADF/cofilins

alters the conformation of ADP-actin filaments (McGough et al.,

1997), resulting in increased fragmentation at the boundaries be-

tween cofilin-decorated and bare filaments (Andrianantoandro

velopmental Cell 31, 1–12, October 27, 2014 ª2014 Elsevier Inc. 1

3086

Figure 1. Localization of Cofilin-1 and Cofi-

lin-2 in Cardiomyocytes

(A) Western blot analysis showing the specificity of

cofilin-1 and cofilin-2 antibodies generated in

guinea pig and rabbit, respectively. One pmol:s of

recombinant cofilin-1 and cofilin-2 proteins were

run on gel, blotted, and detected with cofilin-1 (left)

and cofilin-2 (right) antibody. In addition to the

protein of expected mobility on SDS-gel, cofilin-1

antibody also cross-reacts with a protein of an

apparent molecular weight of �70 kDa.

(B) Western blot analysis of cofilin-1 and cofilin-2

expression during cardiomyocyte maturation. Cell

lysates containing 60 mg of proteins were run on

gel, blotted and detected with cofilin-1 and cofilin-

2 antibody. The blots were probed with antitubulin

antibody as a loading control.

(C) Localization of cofilin-1 and cofilin-2 in mature

cardiomyocytes plated for 4 days on fibronectin.

Cofilin-1 displays punctuate cytoplasmic localiza-

tion, which does not overlap with myofibrils. Cofi-

lin-2 concentrates on a striated pattern between

a-actinin-2 rich stripes in cardiomyocyte sarco-

meres. Bar represents 10 mm.

(D) Line profile quantification of cofilin-1 (in red),

cofilin-2 (in green), and a-actinin-2 (in blue) locali-

zation in myofibrils from the zoomed regions of (C).

Cofilin-1 does not display a regular localization

pattern, whereas cofilin-2 and a-actinin-2 display

regular alternative localization pattern along myo-

fibrils.

Developmental Cell

Cofilin-2 Promotes Actin Dynamics in Sarcomeres

Please cite this article in press as: Kremneva et al., Cofilin-2 Controls Actin Filament Length in Muscle Sarcomeres, Developmental Cell (2014), http://dx.doi.org/10.1016/j.devcel.2014.09.002

and Pollard, 2006; Elam et al., 2013; Suarez et al., 2011). ADF/

cofilins also bind actin monomers (G-actin), displaying higher

affinity for ADP-G-actin, and inhibit nucleotide exchange on

G-actin (Carlier et al., 1997; Hawkins et al., 1993; Hayden

et al., 1993).

Mammals express three ADF/cofilin isoforms: cofilin-1 ubiqui-

tously, ADF in epithelial and neuronal tissues, and cofilin-2

mainly in muscles. Cofilin-2 is the only isoform present in mature

skeletal muscles (Ono et al., 1994; Vartiainen et al., 2002).

Despite partially overlapping expression patterns, ADF/cofilin

DEVCEL 3086

2 Developmental Cell 31, 1–12, October 27, 2014 ª2014 Elsevier Inc.

isoforms cannot functionally compensate

for each other in animals; cofilin-1-defi-

cient mice display early embryonic

lethality and defects in actin-dependent

morphogenic processes, whereas ADF

inactivation leads to corneal defects (Bel-

lenchi et al., 2007; Gurniak et al., 2005;

Ikeda et al., 2003). Mutations in the cofi-

lin-2 gene lead to myopathies in humans

(Agrawal et al., 2007; Ockeloen et al.,

2012), and cofilin-2-deficient mice show

progressive disruption of the sarcomeric

architecture and abnormal accumulation

of F-actin in skeletal muscles. Primary

myofibril assembly and muscle develop-

ment are unaffected in these mice, but

postnatal animals display defects in sar-

comeric actin exchange and muscle

maintenance (Agrawal et al., 2012; Gur-

niak et al., 2014). Cofilin-2 was reported to partially localize to

myofibrils in muscle cells, suggesting that it may be involved in

regulation of actin dynamics in sarcomeres (Nakashima et al.,

2005; Gurniak et al., 2014). However, the mechanism by which

cofilin-2 contributes to the assembly and/or maintenance of

actin filament structures in myofibrils is unknown. Furthermore,

apart from its higher affinity for ATP-actin monomers as

compared to other ADF/cofilin isoforms (Vartiainen et al.,

2002), the biochemical properties that account for the specific

function of cofilin-2 in muscle cells are unknown. Thus, the

Figure 2. Cofilin-2 Depletion by RNAi Dis-

rupts the Periodic Actin Filament Pattern

of Cardiomyocyte Myofibrils

(A) F-actin and a-actinin-2 localization in 4-day-old

cardiomyocytes incubated for 72 hr with control

(upper panel) or cofilin-2-specific (lower panel)

RNAi oligonucleotides. Actin filaments were visu-

alized with fluorescently-labeled phalloidin, and

a-actinin-2 antibody was used as a Z-disk marker.

Insets show zoomed areas of the cells. Control

cells display regular, periodic F-actin staining in

myofibrils, whereas the ‘‘F-actin free’’ regions at

the M-band are lost in majority of myofibrils in

cofilin-2 knockdown cells. Bars represent 10 mm.

Line profile quantification of actin (in red) and

a-actinin-2 (in blue) localization in myofibrils from

zoomed regions of (A).

(B) Western blot analysis of cardiomyocyte lysates

after incubation with control (line 1) or cofilin-2-

specific RNAi (line 2). GAPDH antibody was used

as a control for equivalent loading.

(C) Quantification of cofilin-2 levels of car-

diomyocyte lysates from control and cofilin-2

siRNA transfected cells. Cofilin-2 levels were

>50% lower in knockdown compared control

cells. The data are from four and five independent

experiments for control and cofilin-2-specific

RNAi, respectively. The error bars represent SEM.

See also Figure S1.

Developmental Cell

Cofilin-2 Promotes Actin Dynamics in Sarcomeres

Please cite this article in press as: Kremneva et al., Cofilin-2 Controls Actin Filament Length in Muscle Sarcomeres, Developmental Cell (2014), http://dx.doi.org/10.1016/j.devcel.2014.09.002

functional and mechanistic principles for why tissue-specific

ADF/cofilin isoforms evolved in vertebrates have not been stud-

ied in detail.

Here, we report that cofilin-2 localizes between Z-discs in car-

diomyocytes, close to the pointed ends of the actin filaments,

and its depletion leads to uncontrolled actin filament growth

in sarcomeres. We further show that cofilin-2 disassembles

ADP,BeFx-F-actin more efficiently than other ADF/cofilin iso-

forms and identified a cluster of residues at the surface of cofi-

lin-2 responsible for high-affinity binding to ATP-G-actin. Based

on our findings, we propose that cofilin-2 specifically evolved to

control the length of actin filaments in sarcomeres, which un-

dergo faster monomer exchange at their pointed ends.

RESULTS

Cofilin-2 Localizes near M-Bands in MyofibrilsTo study the role of cofilin-2 in muscle sarcomeres, we examined

its localization and function in cultured neonatal rat cardiomyo-

cytes, which provide a good model system to explore the func-

DEVCEL 3086

Developmental Cell 31, 1–

tion of muscle-specific actin-binding pro-

teins (McElhinny et al., 2005; Skwarek-

Maruszewska et al., 2010). Because

none of the available cofilin-2 antibodies,

including those used in an earlier study

(Nakashima et al., 2005), displayed spe-

cific localization in cardiomyocytes, we

generated isoform-specific antibodies

against rat cofilin-1 and cofilin-2, which

share �80% sequence identity. We used

two peptides, corresponding to surface-

exposed regions that differ between the two isoforms (Experi-

mental Procedures). Immunization of guinea pigs and rabbits

with these peptides yielded antibodies that specifically recog-

nized cofilin-1 or cofilin-2 (Figure 1A) and displayed good spec-

ificity in western blots of cell lysates (Figure 1B). Using these

antibodies, we found that cofilin-1 was present at equal levels

in both newly plated cardiomyocytes, lacking properly organized

sarcomeres (Skwarek-Maruszewska et al., 2009), and in day 3

cardiomyocytes having mature sarcomeres. In contrast, cofilin-

2 protein levels increased steadily during sarcomere maturation

(Figure 1B).

Immunofluorescence microscopy of day 4 cardiomyocytes

displaying striated sarcomeric organization, revealed mostly

irregular, punctuate localization of cofilin-1 in the cytoplasm.

While cofilin-2 also displayed irregular, punctuate localization,

a regular striated pattern was also observed (Figure 1C). Line

scans along myofibrils demonstrated that cofilin-2 concentrated

between the a-actinin-2 peaks (Figure 1D). These results show

that cofilin-2 expression increases during cardiomyocyte matu-

ration and that it localizes between Z-discs in sarcomeres.

12, October 27, 2014 ª2014 Elsevier Inc. 3

Figure 3. Cofilin-2 Depletion Results in Diffuse Localization of

Pointed End Capping Protein Tmod1

(A) F-actin and myomesin localization in 4-day-old control and cofilin-2

knockdown cardiomyocytes. TheM-band proteinmyomesin displayed regular

striated localization pattern also in knockdown cells lacking the F-actin free

region at the M-band.

(B) Tmod1 displays regular striated localization pattern in cells transfected with

control RNAi oligonucleotides, whereas it is mostly diffuse in cofilin-2 knock-

down cells. Bars represent 10 mm.

Developmental Cell

Cofilin-2 Promotes Actin Dynamics in Sarcomeres

Please cite this article in press as: Kremneva et al., Cofilin-2 Controls Actin Filament Length in Muscle Sarcomeres, Developmental Cell (2014), http://dx.doi.org/10.1016/j.devcel.2014.09.002

Cofilin-2 Regulates Actin Filament Length inCardiomyocyte SarcomeresTo address the role of cofilin-2 in sarcomeres, we transfected

day 1 cardiomyocytes with three different cofilin-2-specific

RNAi oligonucleotides, resulting in >50% reduction of cofilin-2

protein levels 72 hr after transfection (Figures 2B and 2C). Immu-

DEVCEL 308

4 Developmental Cell 31, 1–12, October 27, 2014 ª2014 Elsevier Inc

nofluorescence microscopy confirmed that cofilin-2 levels were

strongly reduced, especially in fluorescently conjugated RNAi-

transfected cells, where the oligonucleotides displayed punc-

tuate cytoplasmic localization, although �20%–40% of the cells

still expressed detectable levels of cofilin-2 (Figure S1A available

online). The depletion of cofilin-2 was accompanied by an in-

crease in the intensity of a-actinin-2 staining (Figure S1A),

consistent with previous results that cofilin-2-deficient mice

display a significant increase in the expression levels of other

cytoskeletal proteins, including a-actinin-2 (Agrawal et al.,

2012). All three RNAi oligonucleotides resulted in a similar

phenotype, suggesting that the observed alterations in actin or-

ganization do not result from off-target effects (Figure S1B).

Cardiomyocytes transfected with scrambled control oligo-

nucleotides showed a characteristic striated pattern of phalloi-

din-staining, with clearly defined M-bands devoid of F-actin

(Figure 2A). Depletion of cofilin-2 led to an abnormal F-actin

pattern, characterized by the lack of visible actin-free zones

along M-bands. Quantification of three independent experi-

ments showed that �75% of control cells displayed clear peri-

odic F-actin staining versus �35% in cofilin-2 depleted cells.

Even in knockdown cells that displayed periodic actin staining,

only a fraction of the myofibrils exhibited actin-free M-bands

and showed less uniform periodicity than control cells. Further-

more, the typical striated pattern of the pointed end capping pro-

tein tropomodulin 1 (Tmod1) observed in control cells was

severely diminished in cofilin-2 knockdown cells (Figure 3B),

further suggesting that in the absence of cofilin-2 the actin fila-

ments have irregular lengths. Alternatively, the loss of striated

Tmod1 pattern may result from a more complex interplay of

Tmod1 and cofilin-2. Yet, a-actinin-2 in Z-discs and myomesin

in M-bands showed normal striated localization (Figures 2A

and 3A), suggesting that sarcomere organization is mostly pre-

served in cofilin-2 knockdown cells.

Cofilin-2 Binds and Disassembles ATP-Actin FilamentsTo understand the biochemical features of cofilin-2 responsible

for its function in sarcomeres, we compared the interactions of

cofilin-1 and cofilin-2 with actin filaments. Cosedimentation as-

says revealed that cofilin-1 and cofilin-2 bind ADP-F-actin with

similar affinities (Figure 4A). TIRF microscopy using GFP-cofilin-

2 and mCherry-cofilin-1 showed that the two isoforms can bind

to the same filament, suggesting that their distinct localization

patterns in sarcomeres do not arise from incompatible coopera-

tive interaction with the actin filament (Figure S2). Furthermore,

both cofilin isoforms competed with tropomyosin for F-actin in-

teractions and did not display isoform-specific binding toward

skeletal muscle a-actin or nonmuscle b/g-actin (Figure S3). How-

ever, two observations led us to examine the interactions of the

two cofilin isoforms with ATP-F-actin. First, earlier studies

demonstrated that cofilin-2 interacts with ATP-actin monomers

with higher affinity compared to other ADF/cofilin isoforms (Var-

tiainen et al., 2002). Second, due to continuous exchange of actin

subunits, filament pointed ends in muscle sarcomeres might

contain a mixture of ATP/ADP-Pi- and ADP-actin subunits (Little-

field et al., 2001).

Due to rapid nucleotide hydrolysis and phosphate release, we

could not determine the binding of cofilin isoforms to ATP-actin

filaments. Therefore, filaments were saturated with the Pi analog

6

.

Figure 4. Interactions of Cofilin-1 and Cofi-

lin-2 with ADP- and ADP,BeFx-Actin Fila-

ments

(A and B) Equilibrium binding of mouse cofilin-1

and cofilin-2 to ADP- (A) and ADP,BeFx-actin fil-

aments (B). Cofilins (1 mM) were mixed with 0, 0.2,

0.5, 1, 2, 3, 4, 6, and 8 mM of prepolymerized actin

(A) or 0, 0.5, 1, 2, 4, 7, 10, 15, 20, and 25 mM

ADP,BeFx-actin filaments (B). Actin filaments

were sedimented, and the amounts of cofilins in

the supernatant and pellet fractions were quanti-

fied from six independent experiments.

(C–F) Disassembly of ADP- and ADP,BeFx-actin

filaments in the presence (E and F) and absence

(C and D) of Tmod1 (0.25 mM) and CapZ (0.05 mM).

Cofilin-1 or cofilin-2 (0.5 mM in C and E or 2.5 mM in

D and F) were incubated for 5 min with 2.5 mM of

prepolymerized ADP- (C and E) or ADP,BeFx- (D

and F) pyrene-actin filaments, followed by addition

of actin monomer sequestering vitamin D-binding

protein (4 mM). Error bars represent SEM from four

to six independent experiments. See also Figures

S2 and S3. Assays were carried out in 5 mM Tris

(pH 7.5) in the presence of 100 mM KCl and other

components specified in the Experimental Pro-

cedures.

Developmental Cell

Cofilin-2 Promotes Actin Dynamics in Sarcomeres

Please cite this article in press as: Kremneva et al., Cofilin-2 Controls Actin Filament Length in Muscle Sarcomeres, Developmental Cell (2014), http://dx.doi.org/10.1016/j.devcel.2014.09.002

BeFx to mimic the ADP-Pi-F-actin state (Combeau and Carlier,

1988; Muhlrad et al., 2006). Based on cosedimentation assays,

a small fraction (�20%) of both cofilin-1 and cofilin-2 appeared

to bind ADP,BeFx-F-actin with �1 mM affinity. Importantly, a

majority of cofilin-1 bound ADP,BeFx-F-actin only with low affin-

ity (Kd > 10 mM), whereas a majority of cofilin-2 bound

ADP,BeFx-F-actin with higher affinity (Kd �5 mM) (Figure 4B).

To estimate the effects of cofilin-1 and cofilin-2 on actin fila-

ment turnover, we monitored the decrease of pyrene-actin fluo-

rescence resulting from filament disassembly in the presence

of the two cofilin isoforms and vitamin D-binding protein (DBP)

(Gandhi et al., 2009). DBP binds in the hydrophobic cleft between

actin subdomains 1 and 3, competing with cofilin binding (Lees

et al., 1984; Otterbein et al., 2002), and was used in this ex-

periment to sequester the actin monomers and impede their rein-

corporation into filaments. While the binding of cofilin alters

the fluorescence of pyrene-actin (Carlier et al., 1997) (Figure S4A),

this effect does interfere with the interpretation of the data,

because cofilin was incubated with F-actin for 5 min before addi-

tion of DBP. At an actin:cofilin molar ratio of 5:1, both cofilin iso-

forms displayed high ADP-actin filament disassembly activity

(Figure 4C), albeit cofilin-1 had slightly higher rate than cofilin-2

(relative dissociation rates of 4.08 ± 0.13 versus 3.65 ±

0.03 ms�1, respectively). Under identical concentrations, both

isoforms displayed weak ADP,BeFx-F-actin disassembly activity

DEVCEL 3086

Developmental Cell 31, 1–1

(data not shown). However, at an actin:co-

filin molar ratio of 1:1, cofilin-2 increased

the disassembly rate of ADP,BeFx-F-actin by �10-fold, whereas cofilin-1 did

not significantly promote the disassembly

of ADP,BeFx-F-actin (Figure 4D). Impor-

tantly, cofilin-2 also promoted disas-

sembly of ADP- and ADP,BeFx-actin fila-

ments capped by CapZ and Tmod1 at their barbed and pointed

ends, whereas cofilin-1 could only disassemble capped filaments

in the ADP-bound state (Figures 4E and 4F). These data suggest

that, unlike other ADF/cofilin isoforms characterized so far, cofi-

lin-2 can disassemble of ADP-Pi-mimetic filaments.

Identification of Cofilin-2 Residues Required for BindingADP-Pi-Actin Filaments and ATP-Actin MonomersSequence analysis reveals only a few clusters of residues

conserved in cofilin-2 but not cofilin-1 (Figure 5A). Among these

residues only Thr6, Asn8, Glu10, Trp135, Val137, Asp141, Asp142,

and Ile143 are located on the G/F- and F-actin-binding surface

(Figure 5B). To identify cofilin-2-specific residues responsible

for its higher affinity for ATP-actin, we designed four cofilin-1mu-

tants by replacing clusters of surface-exposed residues with

their cofilin-2 counterparts and testing the ability of the hybrid

mutants to disassemble ADP,BeFx-F-actin. Out of the four mu-

tants analyzed, mutant cof-1/cof-2141,142,143, in which residues

141–143 of cofilin-1 were replaced by cofilin-2 residues Asp141,

Asp142, Ile143, disassembled ADP,BeFx-actin filaments signifi-

cantly faster than wild-type cofilin-1 (Figures 5C and S4B). The

inverse mutant, cof-2/cof-1141,142,143, in which the same resi-

dues in cofilin-2 were replaced by their cofilin-1 counterparts,

displayed significantly lower ADP,BeFx-actin filament disas-

sembly activity compared to cofilin-2 (Figure 5C).

2, October 27, 2014 ª2014 Elsevier Inc. 5

(legend on next page)

DEVCEL 3086

Developmental Cell

Cofilin-2 Promotes Actin Dynamics in Sarcomeres

6 Developmental Cell 31, 1–12, October 27, 2014 ª2014 Elsevier Inc.

Please cite this article in press as: Kremneva et al., Cofilin-2 Controls Actin Filament Length in Muscle Sarcomeres, Developmental Cell (2014), http://dx.doi.org/10.1016/j.devcel.2014.09.002

Developmental Cell

Cofilin-2 Promotes Actin Dynamics in Sarcomeres

Please cite this article in press as: Kremneva et al., Cofilin-2 Controls Actin Filament Length in Muscle Sarcomeres, Developmental Cell (2014), http://dx.doi.org/10.1016/j.devcel.2014.09.002

Cofilin-2 also binds ATP-actin monomers with higher affinity

compared to cofilin-1 (Vartiainen et al., 2002). To examine

whether the same residues that increase the affinity of cofilin-2

for ADP,BeFx-F-actin also increase its affinity for ATP-G-actin,

we used a fluorometric assay with NBD-labeled actin. The bind-

ing of cofilin-1 and cofilin-2 to NBD-G-actin resulted in a fluores-

cence increase, with saturation at higher cofilin concentrations,

enabling us to measure the Kd of the interactions (Figure 5D).

Mutant cof-1/cof-2141,142,143 displayed �2.5-fold higher affin-

ity for ATP-G-actin than wild-type cofilin-1 (Figures 5D and S5).

The affinity was further increased by introducing the cofilin-2-

specific residue Val137 into this mutant. The inverse mutant,

cof-2/cof-1141,142,143 had significantly lower affinity for ATP-

G-actin compared to cofilin-2 (Figure 5D).

ATP-Actin Binding Residues Are Essential for CellularFunction of Cofilin-2To assess the importance of cofilin-20s ability to bind ATP-actin

for its function in cardiomyocyte sarcomeres, we performed

RNAi-rescue experiments with wild-type and mutant cofilins.

HA-tagged cofilin-2 displayed mainly diffuse cytoplasmic locali-

zation when expressed in cardiomyocytes, but also accumu-

lated in the M-band and Z-disc regions as detected by costain-

ing with myomesin, Tmod1 and a-actinin antibodies (Figure S6

and data not shown). Moreover, HA-cofilin-2 expression rescued

the loss of a striated F-actin pattern resulting from depletion

of endogenous cofilin-2. In contrast, neither HA-cofilin-1 nor

HA-cof-2/cof-1141,142,143, which cannot efficiently disas-

semble ADP,BeFx-F-actin or bind ATP-G-actin with high affinity,

rescued the cofilin-2 depletion phenotype. Whereas the majority

of control and cofilin-2 rescue cells displayed a periodic F-actin

staining pattern, most cofilin-1 and cof-2/cof-1141,142,143 ex-

pressing cofilin-2 knockout cells lost the striated F-actin pattern

along myofibrils (Figures 6A and 6B).

DISCUSSION

Three tissue-specific ADF/cofilin isoforms are expressed in

mammals, but the biochemical properties and cellular functions

distinguishing these isoforms have remained elusive. Here, we

demonstrate that the muscle-specific isoform cofilin-2 localizes

between Z-discs in myofibrils and is an essential regulator of

proper actin filament length in sarcomeres. We provide evidence

that, unlike other ADF/cofilin isoforms, cofilin-2 also disassem-

bles ADP,BeFx-actin filaments. Finally, we identify a cluster of

residues on the surface of cofilin-2 that is responsible for its

high affinity for ATP-actin and propose that this region functions

as a nucleotide state sensor in ADF/cofilin.

Figure 5. Identification of Cofilin-2 Residues Responsible for High-Affi

(A) Sequence alignment of mammalian cofilin-1 and cofilin-2 isoforms. A blue b

servation. Colored frames highlight cofilin-2 residues introduced into cofilin-1 (m

orange: cof-1/cof-2141,142,143).

(B) Location of mutated residues on the surface of the cofilin-1 structure (Pope et

to G-actin (green) and F-actin (blue and green).

(C) Disassembly of 2.5 mM ADP,BeFx-F-actin in the presence of 4 mM DBP and

mutants. The error bars represent SEM from four to six independent experiment

(D) Bar diagram showing the dissociation constants of wild-type and mutant co

pendent experiments. See also Figures S4 and S5. Assays were carried out in 5 m

in the Supplemental Experimental Procedures.

DEVCEL

De

The ability of cofilin-2 to bind and disassemble both ADP- and

ADP-Pi-F-actin specifically adapts it for the disassembly of actin

filaments near M-bands in sarcomeres. In nonmuscle cells, actin

monomer addition occurs primarily at the barbed ends due to

rapid association of ATP-G-actin subunits to free barbed ends.

This and the differences in the affinities of the two ends for Pi in-

crease the likelihood that barbed ends will contain ATP- or ADP-

Pi-actin, whereas ADP-actin accumulates toward the pointed

end (Fujiwara et al., 2007; Pollard, 2007; Bugyi and Carlier,

2010). In sarcomeres, the majority of the barbed ends are tightly

capped by CapZ, and subunit exchange occurs primarily at the

pointed end (Littlefield and Fowler, 2008). Because of this

reversal of monomer exchange kinetics in sarcomeres, rapidly

elongating filaments (at least uncapped filaments) may contain

a mixture of ADP- and ATP/ADP-Pi-actin at their pointed ends.

This might create the necessity for a muscle-specific ADF/cofilin

isoform with the ability to promote the disassembly of both ADP-

and ATP/ADP-Pi-actin filaments. Filament disassembly by cofi-

lin-2 may be particularly necessary if rapidly elongating filament

extensions do not contain tropomyosin. While not addressed

here, cofilin-2 may collaborate with tropomodulin for this activity

(Yamashiro et al., 2012). Our results (Figure S3) and an earlier

study (Nakashima et al., 2005) also suggest that cofilin-2 is

slightly more effective than cofilin-1 at competing with tropomy-

osin for F-actin binding/disassembly. It is also unclear what the

exact mechanism is by which cofilin-2 localizes to the M-band

region. We suggest that interactions with M-band-specific pro-

teins may be responsible for cofilin-2 localization to this region.

In this regard, it is interesting to note that in nonmuscle cells

ADF/cofilins interact with CAP1 (Makkonen et al., 2013; Mor-

iyama and Yahara, 2002), whose muscle-specific isoform

CAP2 localizes to M-bands (Peche et al., 2007).

The proposed role of cofilin-2 as a regulator of actin filament

length in mature sarcomeres is consistent with its spatial and

temporal expression pattern in cardiomyocytes and mouse tis-

sues and with published genetic data in mice and humans. In

mice, cofilin-2 mRNA expression increases during myogenesis

and is the only ADF/cofilin isoform expressed in mature skeletal

muscles (Ono et al., 1994; Vartiainen et al., 2002). Here, we also

demonstrate maturation-dependent expression of cofilin-2 at

the protein level. In agreement with its specific function inmature

myofibrils, cofilin-2 depletion inmice did not severely affectmus-

cle development, but resulted instead in progressive sarcomeric

disruption and accumulation of abnormal F-actin structures

(Agrawal et al., 2012; Gurniak et al., 2014). Similarly, mutations

in the human cofilin-2 gene that decrease the expression/stabil-

ity of cofilin-2 result in nemaline myopathy, which is character-

ized by the abnormal organization of actin filaments in striated

nity Binding to ADP,BeFx-F-Actin

ackground highlights conservation, with darker blue representing higher con-

agenta: cof-1/cof-26,8,10, red: cof-1/cof-2135, cyan: cof-1/cof-2137, and

al., 2004) color-coded as in (A). Ovals highlight surfaces participating in binding

2.5 mM of cofilin-1, cofilin-2, cof-2/cof-1141,142,143, or cof-1/cof-2141,142,143s.

filin-1 and cofilin-2 for ATP-G-actin. Error bars indicate SEM from three inde-

M Tris (pH 7.5) in the presence of 100 mMKCl and other components specified

3086

velopmental Cell 31, 1–12, October 27, 2014 ª2014 Elsevier Inc. 7

Figure 6. Rescue of Cofilin-2 Depletion

(A) Distributions of HA-tagged cofilins (left) and F-actin in 4-day-old car-

diomyocytes incubated for 72 hr with cofilin-2-specific RNAi oligonucleotides

and transfected with plasmids expressing RNAi-refractory HA-cofilin-1 (upper

panel), HA-cofilin-2 (middle panel) or HA-cof2/cof1141,142,143 DNA (lower

panel). Expression of HA-cofilin-2 rescued the loss of striated F-actin pattern

resulting from coflin-2 depletion, whereas majority of HA-cofilin-1 and HA-

cof2/cof1141,142,143 expressing cofilin-2 knockdown cells did not display

periodic F-actin pattern along their myofibrils.

(B) Quantification of percentage of control, cofilin-2 knockdown and cofilin-2

knockdown-rescue cells displaying periodic F-actin pattern. Error bars indi-

cate SEM from four to five independent experiments.

See also Figure S6.

DEVCEL 308

Developmental Cell

Cofilin-2 Promotes Actin Dynamics in Sarcomeres

8 Developmental Cell 31, 1–12, October 27, 2014 ª2014 Elsevier Inc

Please cite this article in press as: Kremneva et al., Cofilin-2 Controls Actin Filament Length in Muscle Sarcomeres, Developmental Cell (2014), http://dx.doi.org/10.1016/j.devcel.2014.09.002

muscle myofibrils (Agrawal et al., 2007; Wallgren-Pettersson

et al., 2011). We propose that compromised cofilin-2 activity

leads to uncontrolled elongation of sarcomeric actin filaments,

which in turn impacts the maintenance and correct organization

of the sarcomeric apparatus in mature myofibrils.

Cofilin-1 and cofilin-2 diverged late in evolution, after the

emergence of vertebrates, and thus the invertebrate muscle-

specific ADF/cofilin isoforms are not functionally equivalent to

mammalian cofilin-2 (Poukkula et al., 2011). For example, inCae-

norhabditis elegans the muscle-specific ADF/cofilin isoform

UNC-60B appears to be involved in muscle development, not

maintenance (Ono et al., 2003), and the biochemical differences

between the muscle and nonmuscle isoforms are distinct from

those revealed here for mammalian cofilin-1 and cofilin-2 (Yama-

shiro et al., 2005).

By introducing cofilin-2-specific residues in cofilin-1, we were

able to increase its affinity for ATP-actin monomers and its effi-

ciency to promote the disassembly of ADP,BeFx-F-actin. Recip-rocally, replacing the same residues in cofilin-2 by their cofilin-1

equivalents diminished the affinity of cofilin-2 for ATP-actin

monomers. These results point to a specific cluster of residues

(Asp141, Asp142, and Ile143) on the surface of cofilin-2 responsible

for its higher affinity for ATP-actin monomers. However, because

these mutants only partially increased and decreased the ATP-

actinmonomer affinity of cofilin-1 and cofilin-2, respectively (Fig-

ure 5), other regions of cofilin-2 are also likely to contribute to

ATP-actin monomer interactions. Importantly, the higher affinity

of cofilin-2 for skeletal muscle ATP-G-actin does not result from

actin isoform specificity, because cofilin-2 also binds b/g -ATP-

actin monomers with higher affinity compared to cofilin-1

(Figure S3).

The residues required for high-affinity ATP-actin binding in co-

filin-2 are located at the barbed end region of actinmonomer that

is sensitive to the bound nucleotide. The actin monomer consists

of two major domains on each side from the nucleotide-binding

cleft that, according to their relative position in the filament, are

referred to as the outer (subdomains 1 and 2) and inner (subdo-

mains 3 and 4) domains (Dominguez and Holmes, 2011). The

main difference between monomeric ATP-actin and ADP-actin

subunits in the filament is a propeller-twist rotation of the outer

domain relative to the inner domain, which renders the actin sub-

units in the filament flatter compared to monomeric actin (Fujii

et al., 2010; Oda et al., 2009) (Figure 7A). In the actin monomer,

nucleotide-dependent conformational changes are subtler, but

they also involve a propeller-twist rotation, albeit far less pro-

nounced than in the filament (Otterbein et al., 2001). Opposite

to the nucleotide-binding cleft, actin has a second cleft, known

as the hydrophobic target-binding cleft or the barbed end

groove, where most actin-binding proteins interact (Dominguez,

2004). This cleft also mediates intersubunit contacts in the fila-

ment (Dominguez and Holmes, 2011; Fujii et al., 2010; Oda

et al., 2009). In this way, nucleotide- and polymerization-de-

pended conformational changes in actin are transmitted to the

target-binding cleft, such that proteins that bind there can sense

the state of the nucleotide. Structural studies have shown that

cofilin family members interact in this cleft of actin, both in the

monomeric (Paavilainen et al., 2008) and filamentous (Galkin

et al., 2011) states (Figures 7B–7D). The cofilin surface involved

in interactions with the actin monomer is also implicated in

6

.

Figure 7. The Mechanism of Cofilin-2 Interaction with ATP-Actin

(A) Actin in the filament (red) undergoes a propeller-twist rotation, resulting in a flatter conformation compared tomonomeric actin (blue) (Dominguez andHolmes,

2011; Fujii et al., 2010; Oda et al., 2009). A similar, but less pronounced propeller-twist rotation occurs between the ATP- and ADP-bound conformations of the

actin monomer (Otterbein et al., 2001).

(B) Illustration of the nuclear magnetic resonance (NMR) structure of human cofilin-1 (Pope et al., 2004), showing the side chains of all the amino acids that differ

between cofilin-1 and cofilin-2 (Figure 5A).

(C)Model of cofilin-1 bound toG-actin obtained by superimposing its NMR structure (Pope et al., 2004) onto the crystal structure of the C-terminal ADF-homology

domain 2 of twinfilin in complex with actin (Paavilainen et al., 2008).

(D) A ribbon diagram and space-fillingmodel of cofilin-1 bound to the actin filamentmade by fitting atomic structures into a relatively low resolution reconstruction

from EM structure (Galkin et al., 2011), showing only two actin subunits of the long-pitch helix. The Ca backbone of cofilin-1 is shown in gray, except for regions

implicated in binding to G-actin (green Ca backbone) and F-actin (green and cyan Ca backbones). Side chains with yellow Ca atoms correspond to a cluster of

cofilin-2-specific residues that when introduced into cofilin-1 (mutant Cof-1141,142,143) increase its affinity for ATP-actin. These residues face Lys291, Lys326, and

Lys328 in subdomain 3 of actin (actin subdomains are numbered 1–4) and are expected to form salt bridges that would be affected in different ways for cofilin-1

and cofilin-2 by the propeller-twist rotation in actin (see main text).

Developmental Cell

Cofilin-2 Promotes Actin Dynamics in Sarcomeres

Please cite this article in press as: Kremneva et al., Cofilin-2 Controls Actin Filament Length in Muscle Sarcomeres, Developmental Cell (2014), http://dx.doi.org/10.1016/j.devcel.2014.09.002

interactions with an actin subunit in the filament, while additional

contacts are established through other surfaces with a second

actin subunit of the long pitch filament helix (Figures 5B and

7B–7D). In this way, cofilin can sense nucleotide-dependent

conformational changes in actin, both in the monomer and in

the filament. The cofilin-2 residues implicated here in high-affin-

ity binding to ATP-actin are all located on the surface that

contacts subdomain 3 of actin. Specifically, cofilin-2 residues

Asp141 and Asp142 face actin residues Lys291, Lys326, and

Lys328, likely forming salt bridges. In cofilin-1 these two positions

are occupied by longer glutamic acid side chains (Figure 5A),

which probably also form salt bridges with the three lysine

residues on actin in the ADP state (following the propeller-twist

rotation), but would be too close to actin in the ATP state for

favorable interactions (i.e., their longer side chains may sterically

hinder the interaction with ATP-actin).

In summary, we have identified residues on the surface of

cofilin-2 that account for its higher affinity for ATP-actin mono-

mers than cofilin-1. Although cofilin-2 may display also other

biochemical differences to cofilin-1 and ADF, our RNAi-rescue

experiments provided evidence that the residues responsible

for ATP-G-actin-binding and ADP,BeFx-actin filament disas-

sembly are essential for the function of cofilin-2 in regulating

the length and/or organization of actin filaments in muscle

sarcomeres.

DEVCEL

De

EXPERIMENTAL PROCEDURES

Generation of Cofilin-1 and Cofilin-2 Antibodies

Two guinea pigs and two rabbits were immunized with appropriate mixture of

two peptides conjugated to Keyhole Limpet Hemocyanin. Peptide synthesis

and immunizations were performed in Inbiolabs. The peptides corresponding

to surface exposed regions were (NHCOCH3)CASGVAVSDGVIK(CONH2) and

(NHCOCH3)CLSEDKKNIILEEGK(CONH2), for cofilin-1 and (NHCOCH3)CASG

VTVNDEVIK(CONH2) and (NHCOCH3)CWQVNGLDDIKDRSTLG(CONH2) for

cofilin-2. The serums were collected after four immunizations, and the anti-

bodies were affinity purified as described previously (Vartiainen et al., 2000).

Isolation of Neonatal Rat Cardiomyocytes

Neonatal rat cardiomyocytes were isolated and cultured as described before

(Skwarek-Maruszewska et al., 2009). Briefly, neonatal (1–3 days) rat hearts

were dissected and enzymatically digested. After plating for �70 min to

discard fibroblasts, the cardiomyocytes were replated on fibronectin-coated

dishes. After culturing for 24 hr, the plating medium was exchanged for ‘‘main-

taining medium’’ (Skwarek-Maruszewska et al., 2009).

Cell Biological Methods

Western blotting, immunofluorescence microscopy, siRNA treatment, and

rescue experiments were performed as described in the Supplemental Exper-

imental Procedures.

Protein Expression and Purification

Recombinant wild-type and mutant cofilin-1, cofilin-2, CapZ, and Tmod1 were

produced as described in the Supplemental Experimental Procedures. Rabbit

muscle actin was prepared from acetone powder as described in Spudich and

3086

velopmental Cell 31, 1–12, October 27, 2014 ª2014 Elsevier Inc. 9

Developmental Cell

Cofilin-2 Promotes Actin Dynamics in Sarcomeres

Please cite this article in press as: Kremneva et al., Cofilin-2 Controls Actin Filament Length in Muscle Sarcomeres, Developmental Cell (2014), http://dx.doi.org/10.1016/j.devcel.2014.09.002

Watt (1971), and pyrene-actin was from AP05, Cytoskeleton). Rabbit muscle

actin and human platelet actin (APHL99-E, Cytoskeleton) were labeled with

NBD (7-chloro-4-nitrobenz-2-oxa-1,3-diazole-Cl) as described previously

(Detmers et al., 1981).

ADP- and BeFx-Actin Disassembly Assays

The steady-state rate of actin filament disassembly was monitored by the

decrease in the pyrene fluorescence with excitation at 365 nm and emission

at 407 nm after addition of 4 mM vitamin D binding protein (DBP) (Human

DBP, G8764, Sigma) using modified protocol (Gandhi et al., 2009). To prepare

ADP-pyrene actin, a mixture of G-actin and 5% pyrene-actin was polymerized

in G-buffer by the addition of 5 mMMgCl2 in the presence of 100 mM KCl and

1 mM EGTA. To prepare BeFx-pyrene actin, polymerization was performed as

described above in the presence of 0.2 mM ADP, 5 mM NaF, 600 mM BeCl2,

100 mM KCl, and 1 mM EGTA. Samples of ADP-pyrene or BeFx-pyrene actin

(2.5 mM) were preincubated in the presence of 0.5 or 2.5 mM cofilin-1/2 for

5 min and the reaction was initialized by the addition of DBP. To analyze the

effect of Tmod1 and CapZ on disassembly activity of cofilins ADP-pyrene or

BeFx-pyrene actin were additionally incubated with 0.25 mM Tmod1 and

0.05 mM CapZ for 5–6 min at room temperature before mixing with cofilins.

During the experiments, salt conditions were constant: 100 mM KCl, 2 mM

MgCl2, and 1 mM EGTA. All measurements were carried out using Perkin

Elmer fluorometer. Origin 7.5 software (OriginLab) was used to analyze the

data. To calculate initial relative actin filament disassembly rates for cofilin-1

and cofilin-2, first 80 s of each curve was fitted with linear equation y = a 3

x + b, where coefficient ‘‘a’’ represents speed of the process.

Actin-Binding Assays

The protocols for actin monomer binding and actin filament binding assays are

available in the Supplemental Experimental Procedures.

Statistical Analysis

Mann-Whitney nonparametric test was used to check significance of the

observed effects.

SUPPLEMENTAL INFORMATION

Supplemental Information includes Supplemental Experimental Procedures

and six figures and can be found with this article online at http://dx.doi.org/

10.1016/j.devcel.2014.09.002.

ACKNOWLEDGMENTS

We thank Carol Gregorio for Tmod1 antibody and Minna Poukkula, Stanislav

Rozov, and Maria Vartiainen for critical reading of the manuscript. Sari

Tojkander is acknowledged for help in cardiomyocyte isolation. This study

was supported by grants from the Academy of Finland and Finnish Foundation

for Cardiovascular Research to (P.L). M.H.M. and G.G. were supported by fel-

lowships from GPBM and VGSB graduate programs. R.D. was supported by

NIH grant R01 GM073791.

Received: June 7, 2014

Revised: July 3, 2014

Accepted: September 3, 2014

Published: October 27, 2014

REFERENCES

Agarkova, I., and Perriard, J.-C. (2005). The M-band: an elastic web that cross-

links thick filaments in the center of the sarcomere. TrendsCell Biol.15, 477–485.

Agrawal, P.B., Greenleaf, R.S., Tomczak, K.K., Lehtokari, V.-L., Wallgren-

Pettersson, C., Wallefeld, W., Laing, N.G., Darras, B.T., Maciver, S.K.,

Dormitzer, P.R., and Beggs, A.H. (2007). Nemaline myopathy with minicores

caused bymutation of the CFL2 gene encoding the skeletal muscle actin-bind-

ing protein, cofilin-2. Am. J. Hum. Genet. 80, 162–167.

Agrawal, P.B., Joshi, M., Savic, T., Chen, Z., and Beggs, A.H. (2012). Normal

myofibrillar development followed by progressive sarcomeric disruption with

DEVCEL 308

10 Developmental Cell 31, 1–12, October 27, 2014 ª2014 Elsevier In

actin accumulations in a mouse Cfl2 knockout demonstrates requirement of

cofilin-2 for muscle maintenance. Hum. Mol. Genet. 21, 2341–2356.

Andrianantoandro, E., and Pollard, T.D. (2006). Mechanism of actin filament

turnover by severing and nucleation at different concentrations of ADF/cofilin.

Mol. Cell 24, 13–23.

Bellenchi, G.C., Gurniak, C.B., Perlas, E., Middei, S., Ammassari-Teule, M.,

andWitke, W. (2007). N-cofilin is associated with neuronal migration disorders

and cell cycle control in the cerebral cortex. Genes Dev. 21, 2347–2357.

Bernstein, B.W., and Bamburg, J.R. (2010). ADF/cofilin: a functional node in

cell biology. Trends Cell Biol. 20, 187–195.

Bertling, E., Hotulainen, P., Mattila, P.K., Matilainen, T., Salminen, M., and

Lappalainen, P. (2004). Cyclase-associated protein 1 (CAP1) promotes cofi-

lin-induced actin dynamics in mammalian nonmuscle cells. Mol. Biol. Cell

15, 2324–2334.

Bugyi, B., and Carlier, M.-F. (2010). Control of actin filament treadmilling in cell

motility. Annu. Rev. Biophys. 39, 449–470.

Carlier, M.F., Laurent, V., Santolini, J., Melki, R., Didry, D., Xia, G.X., Hong, Y.,

Chua, N.H., and Pantaloni, D. (1997). Actin depolymerizing factor (ADF/cofilin)

enhances the rate of filament turnover: implication in actin-based motility.

J. Cell Biol. 136, 1307–1322.

Chereau, D., Boczkowska, M., Skwarek-Maruszewska, A., Fujiwara, I., Hayes,

D.B., Rebowski, G., Lappalainen, P., Pollard, T.D., and Dominguez, R. (2008).

Leiomodin is an actin filament nucleator in muscle cells. Science 320, 239–243.

Combeau, C., and Carlier, M.F. (1988). Probing the mechanism of ATP hydro-

lysis on F-actin using vanadate and the structural analogs of phosphate BeF-3

and A1F-4. J. Biol. Chem. 263, 17429–17436.

Cooper, J.A., and Sept, D. (2008). New insights into mechanism and regulation

of actin capping protein. Int. Rev. Cell. Mol. Biol. 267, 183–206.

Detmers, P., Weber, A., Elzinga, M., and Stephens, R.E. (1981). 7-Chloro-4-ni-

trobenzeno-2-oxa-1,3-diazole actin as a probe for actin polymerization.

J. Biol. Chem. 256, 99–105.

Dominguez, R. (2004). Actin-binding proteins—a unifying hypothesis. Trends

Biochem. Sci. 29, 572–578.

Dominguez, R., and Holmes, K.C. (2011). Actin structure and function. Annu.

Rev. Biophys. 40, 169–186.

Elam, W.A., Kang, H., and De la Cruz, E.M. (2013). Biophysics of actin filament

severing by cofilin. FEBS Lett. 587, 1215–1219.

Fujii, T., Iwane, A.H., Yanagida, T., and Namba, K. (2010). Direct visualization

of secondary structures of F-actin by electron cryomicroscopy. Nature 467,

724–728.

Fujiwara, I., Vavylonis, D., and Pollard, T.D. (2007). Polymerization kinetics of

ADP- and ADP-Pi-actin determined by fluorescence microscopy. Proc. Natl.

Acad. Sci. USA 104, 8827–8832.

Galkin, V.E., Orlova, A., Kudryashov, D.S., Solodukhin, A., Reisler, E.,

Schroder, G.F., and Egelman, E.H. (2011). Remodeling of actin filaments by

ADF/cofilin proteins. Proc. Natl. Acad. Sci. USA 108, 20568–20572.

Gandhi, M., Achard, V., Blanchoin, L., and Goode, B.L. (2009). Coronin

switches roles in actin disassembly depending on the nucleotide state of actin.

Mol. Cell 34, 364–374.

Gokhin, D.S., and Fowler, V.M. (2013). A two-segment model for thin filament

architecture in skeletal muscle. Nat. Rev. Mol. Cell Biol. 14, 113–119.

Gurniak, C.B., Perlas, E., andWitke,W. (2005). The actin depolymerizing factor

n-cofilin is essential for neural tubemorphogenesis and neural crest cell migra-

tion. Dev. Biol. 278, 231–241.

Gurniak, C.B., Chevessier, F., Jokwitz, M., Jonsson, F., Perlas, E., Richter, H.,

Matern, G., Boyl, P.P., Chaponnier, C., Furst, D., et al. (2014). Severe protein

aggregate myopathy in a knockout mouse model points to an essential role

of cofilin2 in sarcomeric actin exchange and muscle maintenance. Eur. J.

Cell Biol. 93, 252–266.

Hawkins, M., Pope, B., Maciver, S.K., and Weeds, A.G. (1993). Human actin

depolymerizing factor mediates a pH-sensitive destruction of actin filaments.

Biochemistry 32, 9985–9993.

6

c.

Developmental Cell

Cofilin-2 Promotes Actin Dynamics in Sarcomeres

Please cite this article in press as: Kremneva et al., Cofilin-2 Controls Actin Filament Length in Muscle Sarcomeres, Developmental Cell (2014), http://dx.doi.org/10.1016/j.devcel.2014.09.002

Hayden, S.M., Miller, P.S., Brauweiler, A., and Bamburg, J.R. (1993). Analysis

of the interactions of actin depolymerizing factor with G- and F-actin.

Biochemistry 32, 9994–10004.

Huxley, H.E. (1963). Electron microscope studies on the structure of natural

and synthetic protein filaments from striated muscle. J. Mol. Biol. 7,

281–308.

Ikeda, S., Cunningham, L.A., Boggess, D., Hawes, N., Hobson, C.D.,

Sundberg, J.P., Naggert, J.K., Smith, R.S., and Nishina, P.M. (2003).

Aberrant actin cytoskeleton leads to accelerated proliferation of corneal

epithelial cells in mice deficient for destrin (actin depolymerizing factor).

Hum. Mol. Genet. 12, 1029–1037.

Iskratsch, T., Lange, S., Dwyer, J., Kho, A.L., dos Remedios, C., and Ehler, E.

(2010). Formin follows function: a muscle-specific isoform of FHOD3 is regu-

lated by CK2 phosphorylation and promotes myofibril maintenance. J. Cell

Biol. 191, 1159–1172.

Kho, A.L., Perera, S., Alexandrovich, A., and Gautel, M. (2012). The sarcomeric

cytoskeleton as a target for pharmacological intervention. Curr. Opin.

Pharmacol. 12, 347–354.

Labeit, S., and Kolmerer, B. (1995). Titins: giant proteins in charge of muscle

ultrastructure and elasticity. Science 270, 293–296.

Lange, S., Ehler, E., and Gautel, M. (2006). From A to Z and back?

Multicompartment proteins in the sarcomere. Trends Cell Biol. 16, 11–18.

Lees, A., Haddad, J.G., and Lin, S. (1984). Brevin and vitamin D binding pro-

tein: comparison of the effects of two serum proteins on actin assembly and

disassembly. Biochemistry 23, 3038–3047.

Littlefield, R.S., and Fowler, V.M. (2008). Thin filament length regulation in stri-

ated muscle sarcomeres: pointed-end dynamics go beyond a nebulin ruler.

Semin. Cell Dev. Biol. 19, 511–519.

Littlefield, R., Almenar-Queralt, A., and Fowler, V.M. (2001). Actin dynamics at

pointed ends regulates thin filament length in striated muscle. Nat. Cell Biol. 3,

544–551.

Makkonen, M., Bertling, E., Chebotareva, N.A., Baum, J., and Lappalainen, P.

(2013). Mammalian and malaria parasite cyclase-associated proteins catalyze

nucleotide exchange on G-actin through a conserved mechanism. J. Biol.

Chem. 288, 984–994.

McElhinny, A.S., Schwach, C., Valichnac,M., Mount-Patrick, S., andGregorio,

C.C. (2005). Nebulin regulates the assembly and lengths of the thin filaments in

striated muscle. J. Cell Biol. 170, 947–957.

McGough, A., Pope, B., Chiu, W., and Weeds, A. (1997). Cofilin changes the

twist of F-actin: implications for actin filament dynamics and cellular function.

J. Cell Biol. 138, 771–781.

Michelot, A., Berro, J., Guerin, C., Boujemaa-Paterski, R., Staiger, C.J.,

Martiel, J.-L., and Blanchoin, L. (2007). Actin-filament stochastic dynamics

mediated by ADF/cofilin. Curr. Biol. 17, 825–833.

Moriyama, K., and Yahara, I. (2002). Human CAP1 is a key factor in the re-

cycling of cofilin and actin for rapid actin turnover. J. Cell Sci. 115, 1591–

1601.

Muhlrad, A., Ringel, I., Pavlov, D., Peyser, Y.M., and Reisler, E. (2006).

Antagonistic effects of cofilin, beryllium fluoride complex, and phalloidin on

subdomain 2 and nucleotide-binding cleft in F-actin. Biophys. J. 91, 4490–

4499.

Nakashima, K., Sato, N., Nakagaki, T., Abe, H., Ono, S., and Obinata, T.

(2005). Two mouse cofilin isoforms, muscle-type (MCF) and non-muscle

type (NMCF), interact with F-actin with different efficiencies. J. Biochem.

138, 519–526.

Nevalainen, E.M., Skwarek-Maruszewska, A., Braun, A., Moser, M., and

Lappalainen, P. (2009). Two biochemically distinct and tissue-specific twinfilin

isoforms are generated from the mouse Twf2 gene by alternative promoter us-

age. Biochem. J. 417, 593–600.

Ockeloen, C.W., Gilhuis, H.J., Pfundt, R., Kamsteeg, E.J., Agrawal, P.B.,

Beggs, A.H., Dara Hama-Amin, A., Diekstra, A., Knoers, N.V.A.M.,

Lammens, M., and van Alfen, N. (2012). Congenital myopathy caused by a

novel missense mutation in the CFL2 gene. Neuromuscul. Disord. 22,

632–639.

DEVCEL

Dev

Oda, T., Iwasa, M., Aihara, T., Maeda, Y., and Narita, A. (2009). The nature of

the globular- to fibrous-actin transition. Nature 457, 441–445.

Ono, S. (2010). Dynamic regulation of sarcomeric actin filaments in striated

muscle. Cytoskeleton (Hoboken) 67, 677–692.

Ono, S., Minami, N., Abe, H., and Obinata, T. (1994). Characterization of a

novel cofilin isoform that is predominantly expressed in mammalian skeletal

muscle. J. Biol. Chem. 269, 15280–15286.

Ono, K., Parast, M., Alberico, C., Benian, G.M., and Ono, S. (2003). Specific

requirement for two ADF/cofilin isoforms in distinct actin-dependent pro-

cesses in Caenorhabditis elegans. J. Cell Sci. 116, 2073–2085.

Otterbein, L.R., Graceffa, P., and Dominguez, R. (2001). The crystal structure

of uncomplexed actin in the ADP state. Science 293, 708–711.

Otterbein, L.R., Cosio, C., Graceffa, P., and Dominguez, R. (2002). Crystal

structures of the vitamin D-binding protein and its complex with actin: struc-

tural basis of the actin-scavenger system. Proc. Natl. Acad. Sci. USA 99,

8003–8008.

Paavilainen, V.O., Oksanen, E., Goldman, A., and Lappalainen, P. (2008).

Structure of the actin-depolymerizing factor homology domain in complex

with actin. J. Cell Biol. 182, 51–59.

Pappas, C.T., Krieg, P.A., and Gregorio, C.C. (2010). Nebulin regulates

actin filament lengths by a stabilization mechanism. J. Cell Biol. 189,

859–870.

Peche, V., Shekar, S., Leichter, M., Korte, H., Schroder, R., Schleicher, M.,

Holak, T.A., Clemen, C.S., Ramanath-Y, B., Pfitzer, G., et al. (2007). CAP2,

cyclase-associated protein 2, is a dual compartment protein. Cell. Mol. Life

Sci. 64, 2702–2715.

Pollard, T.D. (2007). Regulation of actin filament assembly by Arp2/3 complex

and formins. Annu. Rev. Biophys. Biomol. Struct. 36, 451–477.

Pollard, T.D., and Cooper, J.A. (2009). Actin, a central player in cell shape and

movement. Science 326, 1208–1212.

Pope, B.J., Zierler-Gould, K.M., Kuhne, R., Weeds, A.G., and Ball, L.J.

(2004). Solution structure of human cofilin: actin binding, pH sensitivity,

and relationship to actin-depolymerizing factor. J. Biol. Chem. 279, 4840–

4848.

Poukkula, M., Kremneva, E., Serlachius, M., and Lappalainen, P. (2011).

Actin-depolymerizing factor homology domain: a conserved fold performing

diverse roles in cytoskeletal dynamics. Cytoskeleton (Hoboken) 68,

471–490.

Sanger, J.M., and Sanger, J.W. (2008). The dynamic Z bands of striated mus-

cle cells. Sci. Signal. 1, pe37.

Skwarek-Maruszewska, A., Hotulainen, P., Mattila, P.K., and Lappalainen, P.

(2009). Contractility-dependent actin dynamics in cardiomyocyte sarcomeres.

J. Cell Sci. 122, 2119–2126.

Skwarek-Maruszewska, A., Boczkowska, M., Zajac, A.L., Kremneva, E.,

Svitkina, T., Dominguez, R., and Lappalainen, P. (2010). Different localizations

and cellular behaviors of leiomodin and tropomodulin inmature cardiomyocyte

sarcomeres. Mol. Biol. Cell 21, 3352–3361.

Spudich, J.A., and Watt, S. (1971). The regulation of rabbit skeletal muscle

contraction. I. Biochemical studies of the interaction of the tropomyosin-

troponin complex with actin and the proteolytic fragments of myosin. J. Biol.

Chem. 246, 4866–4871.

Suarez, C., Roland, J., Boujemaa-Paterski, R., Kang, H., McCullough,

B.R., Reymann, A.-C., Guerin, C., Martiel, J.-L., De la Cruz, E.M., and

Blanchoin, L. (2011). Cofilin tunes the nucleotide state of actin filaments

and severs at bare and decorated segment boundaries. Curr. Biol. 21,

862–868.

Taniguchi, K., Takeya, R., Suetsugu, S., Kan-O, M., Narusawa, M., Shiose, A.,

Tominaga, R., and Sumimoto, H. (2009). Mammalian formin fhod3 regulates

actin assembly and sarcomere organization in striated muscles. J. Biol.

Chem. 284, 29873–29881.

Vartiainen, M., Ojala, P.J., Auvinen, P., Peranen, J., and Lappalainen, P. (2000).

Mouse A6/twinfilin is an actin monomer-binding protein that localizes to the re-

gions of rapid actin dynamics. Mol. Cell. Biol. 20, 1772–1783.

3086

elopmental Cell 31, 1–12, October 27, 2014 ª2014 Elsevier Inc. 11

Developmental Cell

Cofilin-2 Promotes Actin Dynamics in Sarcomeres

Please cite this article in press as: Kremneva et al., Cofilin-2 Controls Actin Filament Length in Muscle Sarcomeres, Developmental Cell (2014), http://dx.doi.org/10.1016/j.devcel.2014.09.002

Vartiainen, M.K., Mustonen, T., Mattila, P.K., Ojala, P.J., Thesleff, I., Partanen,

J., and Lappalainen, P. (2002). The three mouse actin-depolymerizing factor/

cofilins evolved to fulfill cell-type-specific requirements for actin dynamics.

Mol. Biol. Cell 13, 183–194.

Wallgren-Pettersson, C., Sewry, C.A., Nowak, K.J., and Laing, N.G. (2011).

Nemaline myopathies. Semin. Pediatr. Neurol. 18, 230–238.

DEVCEL 308

12 Developmental Cell 31, 1–12, October 27, 2014 ª2014 Elsevier In

Yamashiro, S., Mohri, K., and Ono, S. (2005). The two Caenorhabditis elegans

actin-depolymerizing factor/cofilin proteins differently enhance actin filament

severing and depolymerization. Biochemistry 44, 14238–14247.

Yamashiro, S., Gokhin, D.S., Kimura, S., Nowak, R.B., and Fowler, V.M. (2012).

Tropomodulins: pointed-end capping proteins that regulate actin filament ar-

chitecture in diverse cell types. Cytoskeleton (Hoboken) 69, 337–370.

6

c.