BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
ARTICLE NO.
221, 491–497 (1996)
0624
Elastic Properties of Single Titin Molecules Made Visible through
Fluorescent F-Actin Binding
Miklós S. Z. Kellermayer and Henk L. Granzier1
Department of Veterinary Comparative Anatomy, Pharmacology and Physiology, Washington State University,
Pullman, Washington 99164-6520
Received March 14, 1996
Titin (also known as connection) is a giant filamentous protein that spans the distance between the Z- and
M-lines of the vertebrate muscle sarcomere [1-4]. Several indirect observations have implicated titin as playing
a fundamental role in the generation of passive force of muscle [5,6], driven by titin’s elastic properties. A direct
observation of the mechanical properties of titin, however, has not been demonstrated. Here we have used the
recently shown strong actin-binding property of titin [7-9] to indirectly visualize and manipulate single molecules of titin. Titin molecules were immobilized on a microscope coverslip by attaching them to anti-titin
antibody. The titin molecules were detected by attaching fluorescent actin filaments to them. The titin molecules
were subsequently stretched by manipulating the free end of the attached actin filaments with a glass microneedle. Titin is shown here to possess a high degree of torsional and longitudinal flexibility. The molecule
can be repetitively stretched at least fourfold, followed by recoil. Titin’s unloaded elastic recoil proceeded in two
stages: an initial rapid process (15 ms time constant) was followed by a slower one (400 ms time constant). The
force necessary to fully extend titin—estimated by observing the breakage of the titin-bound actin filaments—
may reach above z100 pN (longitudinal tensile strength of actin [10]). Attachment of fluorescent actin filaments
to titin provides a useful tool to further probe titin’s molecular properties. © 1996 Academic Press, Inc.
Passive force develops when a non-activated (passive) muscle is stretched, and this is the force
that restores the muscle length following release. During muscle contraction, passive force limits
sarcomere-length inhomogeneity along the muscle cell [11], and limits A-band asymmetry within
the sarcomere [12]. Several observations have indicated that in the generation of passive force an
important role is played by titin/connection. Titin is a z3.5-million-dalton protein—the largest
protein known to date—that constitutes about 10% of the total muscle protein mass [1-4]. This
giant molecule spans the half sarcomere, from the Z-line to the M-line. It is anchored to the Z-line
and to the thick filaments of the A-band (via strong myosin-binding property). Upon stretch of the
sarcomere, passive tension is generated by the extension of the I-band segment of titin [5,6], by
virtue of the protein’s elastic nature. The elastic property of titin has been indirectly shown by
immunoelectron microscopic and mechanical experiments. Immunoelectron microscopic studies
have shown—by labelling several titin epitopes along the molecule, following varying degrees of
muscle stretch—that the I-band domain of titin can indeed be extended [13,14]. Mechanical
experiments have shown—by measuring passive muscle force following selective removal of
titin—that it is indeed titin that develops passive force in muscle [5,6]. A drawback of these studies
is that they are indirect. Immunoelectron microscopic experiments provide a snapshop of titin
epitope positions following varying degrees of muscle stretch; however, they fail to provide
information about the concomitant mechanical events. Mechanical experiments, on the other hand,
lack specificity; components other than titin might significantly affect the observations. Therefore,
1
To whom correspondence should be addressed at: Department of VCAPP, Wegner Hall, Room 205, Washington State
University, Pullman, WA, 99164-6520. Fax: (509) 335-4650. E-mail: [email protected].
Abbreviations: BSA, bovine serum albumin; DTT, DL-dithiothreitol; EGTA, ethylene glycol-bis(b-aminoethyl ether)N,N,N8,N9-tetraacetic acid; F-actin, filamentous actin.
491
0006-291X/96 $18.00
Copyright © 1996 by Academic Press, Inc.
All rights of reproduction in any form reserved.
Vol. 221, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
it would be necessary to establish the mechanical properties of titin by using purified preparations,
ideally single titin molecules.
In this study, we demonstrate the elastic nature of purified, single titin molecules by taking
advantage of titin’s recently shown strong actin-binding property [7-9]. We have attached titin
molecules to a microscope coverslip coated with an anti-titin antibody, followed by the addition of
fluorescent actin filaments. By mechanically manipulating the titin-bound actin filament, we succeeded in visualizing the repetitive stretch and recoil of titin. Our results revealed that titin
possesses a high degree of torsional and longitudinal flexibility.
MATERIALS AND METHODS
Preparation of proteins. Actin was purified according to established methods [15]. F-actin was fluorescently labelled with
molar excess of tetramethyl-rhodamine-isothiocyanate-phalloidin (Molecular Probes, Eugene, OR). Titin was prepared from
rabbit back muscle (longissimus dorsi) essentially according to the method of Soteriou et al [16]. Purity of titin was
determined by SDS-polyacrylamide gel electrophoresis using 2.35-12% gels [17].
Antibodies. The 9D10 monoclonal anti-titin antibody used in our experiments, developed by Dr. Marion L. Greaser, was
obtained from the Developmental Studies Hybridoma Bank maintained by the Department of Pharmacology and Molecular
Sciences, Johns Hopkins University School of Medicine, Baltimore, MD 21205, and the Department of Biological Sciences,
University of Iowa, Iowa City, IA 52242, under contract NO1-HD-2-3144 from the NICHD. The antibody was used at
50-fold dilution in assay buffer (25 mM imidazole-HCl, pH 7.4, 200 mM KCl, 4 mM MgCl2, 1 mM EGTA, 1 mM DTT).
Mounting and visualization of single titin molecules. A flow-through microchamber (identical to the one used for the in
vitro motility assay, internal volume z10ml [18]) was first filled with 9D10 antibody at 50-fold dilution. The antibody was
allowed to bind to the nitrocellulose-coated surface of the microchamber for one minute. Unbound antibody was washed
out by the infusion of 100 ml blocking solution [5% (w/v) BSA, 1% (w/v) gelatin, 0.2% (v/v) Tween-20 in assay buffer].
Blocking was carried out for 10 minutes at room temperature. Titin was subsequently added to the microchamber at a
concentration of 10 mg/ml and allowed to bind to the 9D10 antibody for one minute. Unbound titin was washed out by the
infusion of 100 ml 0.5 mg/ml BSA in assay buffer. Fluorescent actin filaments were then added, at a concentration of 70
ng/ml (in low-ionic strength assay buffer, 25 mM imidazole-HCl, pH 7.4, 25 mM KCl, 4 mM MgCl2, 1 mM EGTA, 1 mM
DTT, 0.5 mg/ml BSA) and allowed to bind to titin for one minute. Unbound actin filaments were washed out by the infusion
of low-ionic strength assay buffer containing 100 mM b-mercaptoethanol and an oxygen scavenger enzyme system to
reduce photobleaching [10]. Fluorescent actin filaments were visualized by using a Nikon Diaphot 300 inverted epifluorescence microscope equipped with rhodamine interference filter set (Omega Optical, Brattleboro, VT) and a 64x, 1.4 NA
oil-immersion objective (Nikon), and using a microchannel-plate intensified CCD camera (ICCD-100F, VideoScope International, Ltd., Sterling, VA). The detected images were recorded on a Hi8mm VCR (SONY EV-S7000), and subsequently digitized by an LG-3 frame grabber board (Scion Corporation, Frederick, MD) in an Apple Power Macintosh
6100/60 computer using image analysis software (Scion Image Version 1.57c, based on NIH Image, Wayne Rasband, NIH,
Bethesda, MD).
Mechanical manipulation of single titin molecules. Single molecules of titin were stretched by using a nitrocellulosecoated glass microneedle attached to the titin-bound fluorescent actin filament. Microneedles were pulled from borosilicate
FIG. 1. Binding of actin filaments to titin molecules immobilized on anti-titin antibody-coated surface. Bar graph
comparing the number of actin filaments per field of view bound to either the titin-antibody complex, or in the absence of
titin, or in the absence of antibody.
492
Vol. 221, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. (A) Time-lapse images of a fluorescent actin filament bound to a titin molecule attached to anti-titin antibody.
The filament is pivoting around a single point of attachment. The time between the frames is 1s. Scale bar equals 5 mm.
(B) Superimposed picture of three images of the pivoting actin filament taken at different time points. The overlap of the
images of the filament in different orientations marks the single point of attachment (arrowhead). Scale bar equals 5 mm.
microcapillaries (O.D. 1 mm, I.D. 0.5 mm, Sutter Instrument Co., Novato, CA) using a Model 730 needle/pipette puller
(David Kopf Instruments, Tujunga, CA). The needles were coated with 1% nitrocellulose in iso-amyl acetate (Fullam,
Latham, N.Y.). A microneedle was mounted in a MX630L hydraulic micromanipulator (Newport Corp., Irvine, CA), and
inserted in the microchamber. Actin filaments and the microneedle were simultaneously visualized by using epifluorescence
and brightfield illuminations, respectively, with a 16ND filter placed in the brightfield optical path to reduce exposure of
the intensified CCD camera.
RESULTS AND DISCUSSION
Visualization of single titin molecules. Single titin molecules were indirectly visualized by using
the actin-binding property of titin. The nitrocellulose-covered surface of a sample chamber was
coated with anti-titin antibody. The surface was then blocked with a mixture of 5% (w/v) BSA, 1%
FIG. 3. Angular distribution of the pivoting actin filament. The angle (in degrees) was measured between an arbitrary
reference line and the long axis of the actin filament.
493
Vol. 221, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 4. Sequence of images showing the stretching of a single titin molecule by micromanipulation of the actin filament.
A fluorescent actin filament, bound with its tip to a single titin molecule on the anti-titin antibody, was pulled by help of
a glass microneedle (the needle is outside of the image area). First, the actin filament was pulled to the right; upon returning
the microneedle, the tip of the actin filament returned to the starting position (indicated by arrowhead). Subsequently, the
filament was pulled to the left; upon returning the microneedle, the tip of the actin filament repeatedly returned to the
starting position, driven by titin’s elastic recoil. Scale bar equals 10 mm.
494
Vol. 221, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
(w/v) gelatin, and 0.2% (v/v) Tween-20 to prevent nonspecific binding of titin. Control experiments
indicated that such an extensive blocking was necessary to prevent the nonspecific binding of titin
to the nitrocellulose surface. Following the blocking procedure, titin was added, and allowed to
bind to the antibody. Fluorescent actin filaments were then added, which bound to titin. When
either the antibody or titin was omitted from the control experiments, no actin filaments were seen,
indicating that actin recognized the antibody-bound titin specifically (Figure 1). Most of the actin
filaments were focally attached, as evidenced by their pivoting around a single point (Figure 2).
Occasionally, an actin filament was found to be tethered at more than one point of attachment. Such
an actin filament did not exhibit pivoting motion. The pivoting movement of the actin filaments
indicated that they were attached to single titin molecules. Although the pivoting motion of the
actin filaments is a good indication for the presence of single titin molecules at the tether, it is
conceivable that the tether is formed by not single molecules but by dimers or even trimers. At the
high ionic strengths used, however, titin has been shown to be mostly monomeric in solution [19].
Thus, it is likely that the pivoting actin filaments are indeed tethered to the antibody by single titin
molecules.
The angular distribution of the pivoting actin filaments had a range of 620 degrees (Figure 3).
The profile of the angular distribution was found to be multimodal. The maxima of the distribution
are separated by 200 degrees. The multimodal nature of the distribution profile indicates that the
pivoting actin filaments are found in preferred angular positions separated by 200 degrees. Provided that neither the actin-titin bond nor the titin-antibody bond allow for slippage (that is, they
do not let go under the thermally driven torsional load), the presence of preferred angular positions
of the actin filament represents preferred torsional configurations of the titin molecule.
Mechanical manipulation of single titin molecules. Using the actin filament bound to titin, we
carried out experiments to mechanically manipulate a single titin molecule. The freely moving end
of the actin filament bound to a titin molecule was attached to a nitrocellulose-coated glass
microneedle. The microneedle was then translated by a hydraulic micromanipulator. By pulling the
actin filament with the microneedle, the titin molecule was stretched. Upon moving the microneedle back, the end of the actin filament bound to titin returned to its starting position (Figure
4). The procedure could be repeated many times and in different directions; the tip of the actin
filament attached to titin was always pulled back (by the elastic titin molecule) to its starting
position. Figure 5 shows the deviation of the tip of the actin filament from the starting position. The
maximal deviation in this experiment was 4.2 mm. Considering the position of 9D10 epitope along
the titin molecule [14,20], and that the actin filament could in principle bind anywhere along titin,
FIG. 5. Absolute distance (mm) of the actin-filament tip from the starting point as a function of time (s). The fluorescent
actin filament was repeatedly pulled away from and returned to the starting point. The maximal deviation of the actinfilament tip from the starting point was 4.2 mm.
495
Vol. 221, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 6. Velocity of a short fragment of actin filament driven under unloaded conditions by titin’s elastic recoil.
Following the stretching of a titin molecule, the actin filament broke and was observed to snap back, driven by titin’s recoil.
Velocity is plotted as a function of time, and fitted with double exponential function (f 4 a−bx + c−dx).
the theoretically maximal unextended length of the titin molecule was z1 mm (distance from the
9D10 epitope to the M-line). Thus, at the extended length of 4.2 mm, the titin molecule was
stretched at least fourfold. The demonstrated fourfold extension of a single titin molecule supports
previous predictions of titin’s extensibility [21]. At the minimally fourfold extension of titin, the
protein domains (immunoglobulin, fibronectin type III, and PEVK [4]) from which titin is constructed are likely to be unfolded, as it has previously been proposed [21,22]. Figure 5 indicates that
the mechanically induced unfolding-folding of titin is reversible.
Upon stretching the actin-titin-antibody complex, the actin filament occasionally broke. In these
instances, the broken piece of actin filament was observed to snap back toward the attachment point
on the coverslip, driven by the elastic recoil of the titin molecule. The initial, resolvable velocity
of the recoil was 150 mm/s (Figure 6). This initial, very rapid recoil rate (15 ms time constant) was
followed by a slower process (400 ms time constant, Figure 6). Such a two-stage process has been
previously proposed [22], and is likely to be driven by the initial, rapid refolding of the domains
into a molten globule, followed by a consolidation of the native structure. The actin filament in our
system may be considered as a crude strain gauge. The tensile strength of a single actin filament
has been measured to be z100 pN [10]. The presence or absence of actin-filament breakage may
reveal the tension in the antibody-titin-actin complex. Since pulling on the actin filament occasionally led to the breakage of the filament, the tension in the antibody-titin-actin complex exceeded z100 pN, indicating that the passive force generated by stretching titin may exceed z100
pN per molecule. The forces involved in the high degree of titin unfolding are thus likely to be
higher than the 5 pN proposed by Erickson [22], ranging up to at least z100 pN (the tensile
strength of actin).
In summary, single molecules of titin were indirectly visualized and mechanically manipulated.
Titin exhibited a considerable degree of torsional and longitudinal flexibility. The elastic nature of
titin was directly and visually demonstrated at the molecular level. Such a single-molecule assay
may lead to a better understanding of titin’s function and the molecular mechanisms of passive
force generation in muscle.
ACKNOWLEDGMENTS
We thank B. Stockman for assistance. This work was supported by grants from the American Heart Association,
Washington Affiliate, the Whitaker Foundation, and by NIAMS (R29AR42652).
REFERENCES
1. Wang, K. (1985) Cell Muscle Motil. 6, 315–369.
2. Maruyama, K. (1994) Biophys. Chem. 50, 73–85.
496
Vol. 221, No. 3, 1996
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Trinick, J. (1991) Current Opinion Cell Biol. 3, 112–118.
Labeit, S., and Kolmerer, B. (1995) Science 270, 293–296.
Horowits, R., Kempner, E. S., Bisher, M. E., and Podolsky, R. J. (1986) Nature 323, 160–4.
Granzier, H. L., and Irving, T. C. (1995) Biophys. J. 68, 1027–44.
Jin, J. P. (1995) J. Biol. Chem. 270, 6908–16.
Li, Q., Jin, J.-P., and Granzier, H. L. (1995) Biophys. J. 69, 1508–1518.
Kellermayer, M. S. Z., and Granzier, H. L. (1996) FEBS Letters 380, 281–286.
Kishino, A., and Yanagida, T. (1988) Nature 334, 74–6.
Granzier, H. L., ter Keurs, H. E. D. J., and Akster, H. A. (1991) Am. J. Physiol. 260, C1060–70.
Horowits, R., and Podolsky, R. J. (1987) J. Cell Biol. 105, 2217–23.
Fürst, D. O., Osborn, M., Nave, R., and Weber, K. (1988) J. Cell Biol. 106, 1563–72.
Granzier, H. L., Helmes, M., and Trombitás, K. (1996) Biophys. J. 70, 430–42.
Pardee, J. D., and Spudich, J. A. (1982) Methods Enzymol. 85, 164–182.
Soteriou, A., Gamage, M., and Trinick, J. (1993) J. Cell Sci. 104, 119–123.
Granzier, H. L., and Wang, K. (1993) Electrophoresis 14, 56–64.
Kellermayer, M. S. Z., Hinds, T. R., and Pollack, G. H. (1995) Biochem. Biophys. Acta 1229, 89–95.
Nave, R., Fürst, D. O., and Weber, K. (1989) J. Cell Biol. 109, 2177–87.
Trombitás, K., Jin, J.-P., and Granzier, H. L. (1995) Circ. Res. 77, 856–61.
Soteriou, A., Clarke, A., Martin, S., and Trinick, J. (1993) Proc. R. Soc. Lond. B Biol. Sci. 254, 83–6.
Erickson, H. P. (1994) Proc. Natl. Acad. Sci. USA 91, 10114–8.
497