week ending

PRL 108, 238107 (2012) PHYSICAL REVIEW LETTERS 8 JUNE 2012

Contractile Units in Disordered Actomyosin Bundles Arise from F-Actin Buckling

Martin Lenz,1 Todd Thoresen,2 Margaret L. Gardel,1,2,3 and Aaron R. Dinner1,2,4,*
1
James Franck Institute, University of Chicago, Chicago, Illinois 60637, USA
2
Institute for Biophysical Dynamics, University of Chicago, Chicago, Illinois 60637, USA
3
Department of Physics, University of Chicago, Chicago, Illinois 60637, USA
4
Department of Chemistry, University of Chicago, Chicago, Illinois 60637, USA
(Received 19 October 2011; published 8 June 2012)
Bundles of filaments and motors are central to contractility in cells. The classic example is striated
muscle, where actomyosin contractility is mediated by highly organized sarcomeres which act as
fundamental contractile units. However, many contractile bundles in vivo and in vitro lack sarcomeric
organization. Here we propose a model for how contractility can arise in bundles without sarcomeric
organization and validate its predictions with experiments on a reconstituted system. In the model, internal
stresses in frustrated arrangements of motors with diverse velocities cause filaments to buckle, leading to
overall shortening. We describe the onset of buckling in the presence of stochastic motor head detachment
and predict that buckling-induced contraction occurs in an intermediate range of motor densities. We then
calculate the size of the ‘‘contractile units’’ associated with this process. Consistent with these results, our
reconstituted actomyosin bundles show contraction at relatively high motor density, and we observe
buckling at the predicted length scale.

DOI: 10.1103/PhysRevLett.108.238107 PACS numbers: 87.16.Ln, 87.16.A
, 87.16.Ka, 87.16.Nn

Contractility arising from interactions between myosin behavior is unfortunately lacking, and it is thus important
molecular motors and polar actin filaments (F-actin) is to investigate alternative routes to contractility.
used ubiquitously by cells to build tension and drive mor- In considering such mechanisms, it is important to rec-
phological changes [1]. Such force transmission from mo- ognize that actomyosin interactions can a priori elicit
lecular to cellular length scales is well understood in extension just as well as contraction. Figure 1(c) illustrates
striated muscle, where it critically relies on highly organ- this using two elementary bundles, each made of two polar
ized structures known as sarcomeres [2]. In sarcomeres, filaments—representing F-actin—and one motor—repre-
myosin motors are restricted to the pointed end of F-actin, senting a whole myosin thick filament, itself comprising
while passive actin cross-linkers are present at the barbed numerous individual myosin heads; we use this definition
end. This arrangement is crucial to their contraction of a ‘‘motor’’ throughout. These elementary bundles con-
mechanism, as shown in Fig. 1(a). However, many con- tract when the motor is located in the vicinity of the
tractile actomyosin bundles found in vivo, such as smooth filament pointed ends, but extend when it is close to the
muscle fibers [3], graded polarity bundles [4], and the barbed ends. Overall contractility in nonsarcomeric bun-
contractile ring [5], lack a sarcomeric organization. Most dles requires that the symmetry between these two com-
recently, we have shown that in vitro bundles lacking peting tendencies be broken. We show in Ref. [15] that this
apparent sarcomeric organization can also contract [6] necessitates (1) a dispersion of unloaded velocities to be
[e.g., Fig. 1(b)]. In these disparate systems, contraction present among the motors (as observed experimentally
occurs with a well-defined contraction velocity per unit [16]) and (2) an asymmetric response of the filaments to
length, suggesting that contractile bundles can be mean- longitudinally applied stresses, e.g., a tendency to yield
ingfully divided into elementary units that are arranged in under compression while resisting extension.
series [5–7]. The mechanisms giving rise to such units in In this Letter, we use theory and experiments to demon-
the absence of sarcomeric organization are not understood. strate a mechanism for nonsarcomeric contractility
Much theoretical work on nonsarcomeric actomyosin compatible with these constraints. We first show experi-
assemblies posits contractility as a fundamental assump- mentally that contraction in reconstituted actomyosin bun-
tion, and predicts larger-scale effects such as polarity dles is accompanied by F-actin buckling, an instance of the
organization [9], the appearance of topological defects asymmetric filament response discussed above. We then
[10], active stiffening [11], and oscillatory behavior in cells investigate the general consequences of asymmetric fila-
[12]. Models that address the microscopic origin of con- ment response theoretically by considering the buildup of
tractility assume that myosin motors dwell at the barbed forces in a bundle with randomly arranged motors. We
ends of F-actin, thus acting as transient static cross-linkers predict that buckling yields contraction, and occurs in an
[13]. This generates sufficient sarcomerelike organization intermediate range of motor density. We also calculate a
to elicit contraction [14]. Experimental evidence for this characteristic length scale between two buckles, which

0031-9007=12=108(23)=238107(5) 238107-1 Ó 2012 American Physical Society

 week ending
PRL 108, 238107 (2012) PHYSICAL REVIEW LETTERS 8 JUNE 2012

myosin density (‘0 ¼ 540 nm) to shorten by ’ 10% rap-

idly (100–600 nm  s
1 ) [Fig. 1(d) and movie S1 [18]]. In
contrast, contraction does not occur at low myosin density
(‘0 ¼ 1:5
m) [Fig. 1(e) and movie S1 [18]]. A sharp
transition between those two behaviors is observed at
‘0 ¼ 1:3
m [Fig. 1(f)].
To better understand the underpinnings of contractility
in this system, we next examine individual bundles and
observe F-actin buckling coincident with contraction
[Fig. 2(a); movie S2 [18]]. Prior to ATP addition, compact
bundles with aligned F-actin are observed. Upon ATP
addition, the frequency of buckles increases rapidly during
contraction, and then diminishes once contraction stops
[Fig. 2(b)]. These F-actin buckles are dynamic, with their
amplitude, curvature, and location changing over time.
Qualitatively, the relationship between buckling and
contraction can be understood as follows. Consider two
FIG. 1. Contraction in actomyosin bundles. (a) Sarcomeric antiparallel filaments interacting through several different
structure as in striated muscle. As motors tend to move toward motors with distinct speeds [Fig. 2(c)]. As motors start to
the filament barbed ends, the sarcomeric structure imposes that
move relative to the filaments, stresses build in sections of
each contractile unit (sarcomere) contracts. (b) Bundle devoid of
sarcomeric organization or passive cross-linkers, as in our ex- the filament flanked by motors with different speeds. When
periments. (c) Motors and polar filaments induce local contrac- the flanking motor proximal to the barbed end is faster than
tion or extension depending on the geometry of their assembly that proximal to the pointed end, compression arises. When
(filament polarity always dictates the direction of motion [8]). it is slower, tension arises. Filament buckling breaks the
(d) Time-lapse images of a bundle comprised of F-actin and symmetry between these respective tendencies to contrac-
fluorescent myosin thick filaments (inverted contrast) with tion and extension. Indeed, following buckling of the com-
‘0 ¼ 540 nm. The initially wavy bundle becomes taut following pressed filament sections, fast motors are free to move
the addition of 1 mM ATP at t ¼ 0 s, indicating contraction. quickly while the others move slowly. This results in the
Scale bar, 5
m. (e) Similar experiment with ‘0 ¼ 1:5
m, growth of the compressed sections and shrinkage of the
showing no contraction. Scale bar as in (d). See also movie S1 extended ones, and thus in overall bundle contraction
[18]. (f) Bundle contraction as a function of ‘0 . Bars indicate
standard deviation (n  25).

provides a natural size for a contractile unit. These pre-

dictions are consistent with experimental observations,
suggesting that buckling underlies contractility in nonsar-
comeric actomyosin bundles.
To form reconstituted actomyosin bundles, we follow
the protocol described in Ref. [6]. We incubate F-actin with
length ‘f ’ 5
m with smooth muscle myosin thick fila-
ments of length ’ 300 nm in a buffer lacking adenosine
triphosphate (ATP) such that thick filaments cross-link
F-actin with high affinity. In this system, the dispersion
of motor velocities necessary for contraction likely arises
from the variation in number of myosin heads in the thick
filaments. While flexible motors have been considered as a
basis for contraction [17], this is unlikely to apply here as FIG. 2 (color online). Buckling in nonsarcomeric contractile
thick filaments are significantly more rigid than F-actin. actomyosin bundles. (a) Time-lapse images of fluorescent actin
The bundle lengths range from 10 to 100
m with 4–6 (inverted contrast) showing F-actin buckling (arrowheads) follow-
F-actin per bundle cross-section, and no sarcomeric orga- ing the addition of 1 mM ATP at t ¼ 0 s. Scale bar, 5
m. See
also movie S2 [18]. (b) Relative contraction (filled squares) and
nization is observed. By varying the concentration of my-
number of F-actin buckles (open circles) as a function of time.
osin filaments, the average spacing ‘0 between two Data show mean sd averaged over n ¼ 3 bundles with ‘0 ’
consecutive myosin filaments can be varied from 390 nm 1
m. (c) The presence of fast (gray) and slow (white) motors
to 5:3
m. generically induce compressive (solid red) and extensile (hatched
Once the bundles are formed, we perfuse a buffer con- blue) stresses in filaments. (d) Buckling of the compressed fila-
taining 1 mM ATP, which causes bundles formed with high ments leads to an overall shortening of the bundle.

238107-2
 week ending
PRL 108, 238107 (2012) PHYSICAL REVIEW LETTERS 8 JUNE 2012

[Fig. 2(d)]. The region centered around each buckle thus where c > 0 is the filament compliance. We refer to the
plays the role of a contractile unit, whose typical size is motor flanked by filament sections i
1 and i as ‘‘motor
equal to the distance ‘B between two buckles. i,’’ and describe its operation by the simplified force-
In this picture, the contractile behavior of the bundle velocity relationship
hinges on the ability of the motors to induce filament
fi
1
fi ¼ Fi
vi : (2)
buckling. At high motor density, we expect the bundle to
be so strongly cross-linked that buckling becomes impos- Here vi denotes the local velocity of the filament at the
sible despite the sizable stresses induced by a large number location of motor i and  > 0 is the motor susceptibility.
of motors. At low motor density, we expect that stochastic Equations (1) and (2) yield a local relaxation time scale
detachment of the motors undermines stress buildup and r ¼ c=2. The time-independent stall force of motor i is
thus prevents buckling. Here we present a mathematical denoted by Fi in Eq. (2), and is drawn from a random
model to predict the range of myosin densities enabling distribution satisfying
buckling and the contractile unit length ‘B . These results
are then compared with the observations in Figs. 1 and 2 to F
i ¼ FS and Fi Fj
F
i F
j ¼ FS2 ij ; (3)
validate the proposed contraction mechanism. where bars denote averages over the motor distribution.
The key assumptions of our model are that (1) motors As a result, different motors have different unloaded
have a dispersion in their unloaded velocities, (2) a section velocities Fi = as required for contraction. Owing to the
of filament between two motors buckles above a certain conservation of filament mass,
threshold force FB , and (3) motors intermittently detach
dLi
from the filaments, thus allowing local stress relaxation. ¼ vi
viþ1 : (4)
We consider a bundle of weakly deformed filaments and dt
ask whether the forces developing within it are sufficient to Finally, a motor bound to several filaments as in Fig. 3(a)
induce buckling [Fig. 3(a)]. can transiently detach from one while still holding onto the
To this end, we focus on a single filament of length ‘f others. We thus let each motor i randomly detach from the
and approximate its surroundings by an effective medium filament with a constant rate 1=d . Following detachment,
composed of evenly spaced pointlike motors separated by a local filament stresses relax instantaneously, yielding fi ¼
distance ‘0  ‘f [Fig. 3(b)]. This divides the filament into fi
1 ¼ ðfi þ fi
1 Þ=2. The motor then reattaches after a
discrete sections, which we label by i ¼ 0; . . . ; ‘f =‘0 . We time much shorter than r and d . Since a motor in a dense
take into account the possibility that the filaments are not bundle is typically close to several filaments, the probabil-
straight, but bend away from the x axis, implying that the ity that it detaches from all filaments at the same time and
contour length Li of filament section i can be larger than leaves the bundle is negligible. We denote by h. . .i the
‘0 . Defining fi as the tension of filament section i (fi < 0 average over the Poisson process of motor detachment.
for a compressed filament section), we expand its force- We obtain the space and time evolution of the filament
extension relationship for small deformations, tension fðx; tÞ in the continuum limit i ! x=‘0 by combin-
ing Eqs. (1)–(4) and averaging over motor detachment,
Li ¼ Li ðfi ¼ 0Þ
cfi ; (1) @t hfi
D@2x hfi ¼ ð‘0 =2r Þ@x F; (5)
where D ¼ ‘20 ð
1
1
r þ d Þ=2. The right-hand side of
Eq. (5) involves the spatial gradient of the stall force
FðxÞ, reflecting the fact that nonidentical motors lead to
force buildup. This effect competes with the relaxation of
filament forces through motor detachment, which enters
through the diffusion term D@2x hfi.
An initially relaxed filament [fðx; t ¼ 0Þ ¼ 0] experien-
ces a vanishing average force hfiðx;
tÞ ¼ 0 throughout its
dynamics. To quantify the magnitude of the motor-induced
stress, we use Eqs. (1)–(4) and average over motor detach-
ment to calculate the rms filament force in the continuum
FIG. 3 (color online). Stress buildup in bundles with nonident- limit,
ical motors. (a) In a bundle with motors having nonidentical
1=2 X 1=2
3
n2 Dt=ð2‘f Þ2 Þ2
velocities (shades of gray), filaments of lengths  ‘f are sub- hf2 i ¼ f1 ð1
e : (6)
jected to random motor forces at points  ‘0 apart distributed n2Z
n2  2
throughout their length. (b) Prior to contraction, the environment
of a filament of interest (red, delimited by dotted lines) can be This force increases monotonically from zero at t ¼ 0 to
approximated by a collection of evenly spaced motors (shades of f1 ¼ ð‘f =12‘0 Þ1=2  FS =ð1 þ r =d Þ at t ¼ 1 [Figs. 4(a)
gray). and 4(b)]. We next estimate the dependence of the ratio

238107-3
 week ending
PRL 108, 238107 (2012) PHYSICAL REVIEW LETTERS 8 JUNE 2012

r =d on the experimentally accessible parameter ‘0 . A While the regime ‘0 ’ ‘
0 is not accessible experimentally,
wormlike chain model for filament elasticity yields the predicted value for ‘þ
0 is strikingly similar to the motor
c  ‘40 =kB T‘2p , where ‘p is the filament persistence length spacing at which the breakdown of contraction is observed
[19], and we approximate   FS =v, where v is a charac- in Fig. 1(f) (1:3
m), suggesting that the proposed mecha-
teristic motor velocity. This implies r =d  ð‘0 =‘0 Þ4 , nism is a good description of our experiments.
with ‘0 ¼ ðkB T‘2p vd =FS Þ1=4 . We can thus distinguish To characterize the contractile units resulting from this
two regimes for the steady-state force f1 [Fig. 4(a)]. For mechanism when ‘
þ
0 < ‘0 < ‘0 , we turn to the transient
‘0  ‘0 , detachment events are rare compared to the time regime leading up to filament buckling. The filament force
r needed for the force to recover from such an event, and profile as a function of x is initially flat, and subsequently
f1 is not affected by them. For ‘0 ‘0 , f1 quickly coarsens into a random walk for t ¼ þ1. According to
decreases with increasing ‘0 as detachment becomes Eq. (5), this coarsening occurs diffusively with diffusion
much faster than recovery. coefficient D. The typical filament forces at time t  ‘2f =D
pffiffiffiffiffiffi
Up to a prefactor of order one, contraction proceeds as in are thus of order f1 ð Dt=‘f Þ1=2 . We denote the time that
Figs. 2(c) and 2(d) if f1 > FB  kB T‘p =‘20 [19]. this force reaches the buckling threshold FB by tB , follow-
Comparing f1 to FB as in Fig. 4(a), we find a threshold ing which contraction proceeds as in Figs. 2(c) and 2(d)
stiffness above which buckling cannot occur (as and the coarsening dynamics is interrupted. The distance
exampled by the upper blue line). Reasonable values for between buckles at tB thus yields the contractile unit size
our actomyosin system are ‘p ’ 10
m, v ’ 200 nm  s
1 ,
FS  FS ’ 1 pN, and d ’ 200 ms based on the typical pffiffiffiffiffiffiffiffiffi ‘2p  kB T 2 
r 2
time scales involved in the myosin mechanochemical cycle. ‘B  DtB  3 1þ : (8)
‘0 FS d
Since r * d in our experiments and detached motors
reattach in  1 ms [20], our previous assump-
As illustrated in Fig. 4(c), ‘B is typically in the micrometer
tion of fast motor reattachment is justified. These values
range, in agreement with the observations of Fig. 2(a) and
put us in the soft filament regime defined by ‘p  the findings of Ref. [6].
FS4 L2f ðvd Þ3=2 =kB T 5=2 FS3=2 ’ 20 cm (lower red line). In Because of compensating effects between contractile
this regime, the lines representing FB and f1 intersect at and extensile motor-filament configurations, the familiar
framework involving rigid filaments and identical motors
commonly used to describe striated muscle contraction is
‘

0 ¼ ðkB T‘p =FS ‘f Þ
1=2 2=3
’ 70 nm; (7a)
not suited to study actomyosin bundles lacking sarcomeric
‘þ
0 ¼ ð‘f vd ‘p FS =FS Þ
1=2 2=5
’ 1
m; (7b) organization. Here, we put forward an alternative mecha-
nism based on our observation of buckling. The buckling
arises from the nonlinear elastic response of F-actin [21]
meaning that buckling and contraction occur for ‘

0 < ‘0 <
þ and dispersion in the speeds of myosin motors [22]. F-actin
‘0 . This range reflects the fact that strong cross-linking buckling has previously been invoked to explain contrac-
(‘0 < ‘

0 ) suppresses buckling while sparse motors tion qualitatively [23]. Addition of passive cross-linkers,
(‘0 > ‘þ
0 ) are undermined by stochastic detachment. which are formally equivalent to immobile motors, would
reinforce a dispersion of motor velocities and promote
contraction.
The order-of-magnitude agreement between theory and
experiments with respect to the size of contractile units and
the critical myosin concentration required for contraction
suggests that our current analysis offers a good description
of the onset of bundle contractility. Our conclusions are
robust to inclusion of features such as inhomogeneous
motor spacings ‘0 and force dependence of the motor
detachment rate (see Supplemental Material [18]). Our
mechanism is a general one and applies to any one-
dimensional system of polar filaments and motors. It is
FIG. 4 (color online). Model predictions for filament force also generalizable to any situation where filaments respond
buildup. (a) Black line: steady-state filament force f1 as a asymmetrically to compression and extension, even if
function of motor spacing ‘0 [Eq. (6)]. For ‘0  ‘0 and ‘0 buckling is not present. Further experiments and theory
‘0 , f1 / ‘0
1=2 and ‘
9=2
0 , respectively. Colored straight lines: are needed to better understand the molecular basis for
buckling force FB / ‘
20 . (b) Typical filament force ðhf iÞ
2 1=2 as a motor inhomogeneities and filament asymmetric response
function of time [Eq. (6)]. (c) Contractile unit size ‘B as a in the myriad of nonsarcomeric organizations found
function of ‘0 as in Eq. (8) (‘f ’ 5
m). in vivo.

238107-4
 week ending
PRL 108, 238107 (2012) PHYSICAL REVIEW LETTERS 8 JUNE 2012

We thank Yitzhak Rabin and Tom Witten for useful [11] F. C. MacKintosh and A. J. Levine, Phys. Rev. Lett. 100,
discussions. This work was supported by NSF 018104 (2008).
DMR-MRSEC 0820054, NIH P50 GM081892, and NIH [12] G. Salbreux, J. F. Joanny, J. Prost, and P. Pullarkat, Phys.
DP10D00354. Biol. 4, 268 (2007).
[13] K. Kruse and F. Jülicher, Phys. Rev. Lett. 85, 1778 (2000);
K. Kruse and K. Sekimoto, Phys. Rev. E 66, 031904
(2002); K. Kruse and F. Jülicher, Phys. Rev. E 67,
051913 (2003); T. B. Liverpool and M. C. Marchetti,
*[email protected] Phys. Rev. Lett. 90, 138102 2003; R. Peter, V. Schaller,
[1] J. Stricker, T. Falzone, and M. L. Gardel, J. Biomech. 43, 9 F. Ziebert, and W. Zimmermann, New J. Phys. 10, 035002
(2010). (2008).
[2] B. Alberts, A. Johnson, J. Lewis, M. Raff, K. Roberts, and [14] A. Zemel and A. Mogilner, Phys. Chem. Chem. Phys. 11,
P. Walter, Essential Cell Biology (Garland, New York, 4821 (2009).
1998). [15] M. Lenz, M. L. Gardel, and A. R. Dinner, New J. Phys. 14,
[3] F. S. Fay, K. Fujiwara, D. D. Rees, and K. E. Fogarty, J. 033037 (2012).
Cell Biol. 96, 783 (1983). [16] A. Yamada, N. Ishii, and K. Takahashi, J. Biochem.
[4] L. P. Cramer, M. Siebert, and T. J. Mitchison, J. Cell Biol. (Tokyo) 108, 341 (1990).
136, 1287 (1997). [17] T. B. Liverpool, M. C. Marchetti, J.-F. Joanny, and J. Prost,
[5] Ana Carvalho, Arshad Desai, and Karen Oegema, Cell Europhys. Lett. 85, 18007 (2009).
137, 926 (2009). [18] See Supplemental Material at http://link.aps.org/
[6] T. Thoresen, M. Lenz, and M. L. Gardel, Biophys. J. 100, supplemental/10.1103/PhysRevLett.108.238107 for a
2698 (2011). study of nonidentical motor spacings and force-dependent
[7] W. M. Bement and D. G. Capco, Cell Motil. Cytoskeleton filament detachment.
20, 145 (1991); A. M. Herrera, B. E. McParland, A. [19] Theo Odijk, Macromolecules 28, 7016 (1995).
Bienkowska, R. Tait, P. D. Paré, and C. Y. Seow, J. Cell [20] B. Gilboa, D. Gillo, O. Farago, and A. Bernheim-
Sci. 118, 2381 (2005). Groswasser, Soft Matter 5, 2223 (2009).
[8] J. R. Sellers and B. Kachar, Science 249, 406 (1990). [21] J. Berro, A. Michelot, L. Blanchoin, D. R. Kovar, and J.-L.
[9] K. Kruse, A. Zumdieck, and F. Jülicher, Europhys. Lett. Martiel, Biophys. J. 92, 2546 (2007).
64, 716 (2003); N. Yoshinaga, J.-F. Joanny, J. Prost, and P. [22] Y. Tanaka, A. Ishijima, and S. Ishiwata, Biochim.
Marcq, Phys. Rev. Lett. 105, 238103 (2010). Biophys. Acta 1159, 94 (1992).
[10] K. Kruse, J. F. Joanny, F. Jülicher, J. Prost, and K. [23] M. Soares e Silva, M. Depken, B. Stuhrmann, M. Korsten,
Sekimoto, Phys. Rev. Lett. 92, 078101 (2004); Eur. F. C. MacKintosh, and G. H. Koenderink, Proc. Natl.
Phys. J. E 16, 5 (2005). Acad. Sci. U.S.A. 108, 9408 (2011).

238107-5