Reconstitution of contractile actomyosin bundles

Abstract

Contractile actomyosin bundles are critical for numerous aspects of muscle and nonmuscle cell physiology. Due to the varying composition and structure of actomyosin bundles in vivo, the minimal requirements for their contraction remain unclear. Here, we demonstrate that actin filaments and filaments of smooth muscle myosin motors can self-assemble into bundles with contractile elements that efficiently transmit actomyosin forces to cellular length scales. The contractile and force-generating potential of these minimal actomyosin bundles is sharply sensitive to the myosin density. Above a critical myosin density, these bundles are contractile and generate large tensile forces. Below this threshold, insufficient cross-linking of F-actin by myosin thick filaments prevents efficient force transmission and can result in rapid bundle disintegration. For contractile bundles, the rate of contraction decreases as forces build and stalls under loads of ∼0.5 nN. The dependence of contraction speed and stall force on bundle length is consistent with bundle contraction occurring by several contractile elements connected in series. Thus, contraction in reconstituted actomyosin bundles captures essential biophysical characteristics of myofibrils while lacking numerous molecular constituents and structural signatures of sarcomeres. These results provide insight into nonsarcomeric mechanisms of actomyosin contraction found in smooth muscle and nonmuscle cells.

Figure 1

Templated assembly of tethered actomyosin bundles. (a) Schematic illustrating the sequential process used for templated bundle assembly (1). Biotinylated-bovine serum albumin is coupled to the surface of a PAA gel affixed to a glass coverslip. Neutravidin beads (gray circles) bind to the biotinylated-bovine serum albumin (2). Biotinylated F-actin (chevrons) is introduced and bind to beads. A dilute suspension of F-actin remains (3). Myosin thick filaments suspended in nucleotide free Assay buffer (black) are introduced (4). F-actin cross-linking by myosin filaments mediates bundle formation. (b) Inverted contrast image of F-actin asters visualized with Alexa 568-phalloidin before myosin perfusion. Dark circles are F-actin-coated beads. Asterisks indicate free F-actin ends. Scale bar is 5 μm; see Movie S1. (c) Inverted contrast images of F-actin visualized with Alexa 568-phalloidin (left) and OG-labeled myosin (right) illustrating network of bundles formed after 30 min incubation of F-actin asters with myosin thick filaments. Scale bar is 5 μm. (d) Schematic diagram illustrating transverse, u, and longitudinal, s, directions along a bundle. Plots of OG-myosin intensities in transverse and longitudinal line scans are shown below. For transverse line scans, the location, u_peak, and intensity, I_peak, of the peak are determined. (e) Number of F-actin per bundle cross section as a function of myosin concentration. Error bars indicate standard deviations, (n = 7–28 bundles for each data point). (f) The mole ratio of myosin heavy chains to actin in the bundles determined from quantitative fluorescence imaging (triangles) and densitometry (open circles) as a function of myosin concentration. Error bars for imaging indicate standard error (n = 20 bundles for each data point).

Figure 2

A critical myosin density is required to stabilize bundles and facilitate contraction. (a) Inverted contrast images of OG-myosin in bundles formed with R_M:A = 0.44, 0.64 and 1.9 in nucleotide-free (NN) Assay buffer (top row) and at three times after addition of Assay buffer containing 1 mM (R_M:A = 0.44 or 0.64) or 0.1 mM (R_M:A = 1.9) ATP (bottom rows). All times are indicated in seconds either before (negative times) or after (positive times) ATP addition. For R_M:A = 0.44, inverted contrast images of F-actin, visualized with Alexa 568-phalloidin, is also shown to illustrate the dissociation of F-actin from bundles and the reappearance of F-actin asters by +40s (see Movie S3). Scale bars are 5 μm. (b) Amplitude of transverse fluctuations of bundle contour, measured by the variance of the bundle midpoint position in direction normal to the bundle contour, u_peak, as a function of R_M:A in both nucleotide-free (solid black squares, NN) or 1 mM ATP (open triangles, ATP) conditions. Data shown are mean ± SE. (n = 12–15 bundles for all conditions). (c) Incidence, reported as percentage, of the states observed after ATP perfusion: contraction, bundles remain stable without contraction (no contraction) or bundles disintegrate as a function of R_M:A (n > 48 bundles for each data point). Data points obtained within 1 min of ATP perfusion. (d) Percent decrease in myosin intensity after the first 45 s of ATP addition. Data shown are mean ± SD (n = 5 bundles for each data point).

Figure 3

Contraction of tethered and untethered bundles. (a) Time-lapse series of inverted contrast, OG-myosin images in a contracting bundle with $R_{\text{M}:\text{A}}^{\text{ATP}}$ = 1.4. Times are in seconds before (negative times) or after (positive times) addition of 0.1 mM ATP. Dashed line demarks changing contour of the tethered bundle of interest. A connection to a neighboring bundle breaks between 60 and 65 s (arrow), following which contraction of both the untethered bundle (asterisk) and tethered bundle (dashed line) resume. Scale bar is 5 μm; see Movie S6. (b) Contour length (left axis, solid circles) and contraction speed (right axis, open circles) of the bundle indicated by the dashed line in a. (c) Time-lapse series of inverted contrast OG-myosin images illustrating the contraction of an untethered bundle following the rupture of a taut bundle 85 s after 1mM ATP addition. Bundle shown contains $R_{\text{M}:\text{A}}^{\text{ATP}}$= 1.4. Asterisk indicates the free bundle end. Scale bar, 5 μm. (d) Bundle contour length (closed circles, left axis) and contraction speed (right axis, open circles) versus time for the contracting untethered bundle shown in (c).

Figure 4

Tension is built during contraction of tethered bundles. (a) Schematic illustrating how the displacement of a bead bound to the top surface of an elastic hydrogel can be used to determine the forces applied on the bead. A force, F, applied to the surface-bound bead is balanced by the elastic restoring force exerted by the underlying gel, resulting in a bead displacement Δx such that F = k_eff Δx, where k_eff is an effective spring constant determined by the gel elastic properties. Further details of this measurement are discussed in Fig. S6. (b) Images of OG-myosin (inverted contrast) in a bundle with $R_{\text{M}:\text{A}}^{\text{ATP}}$= 1.4. The underlying gel has a shear elastic modulus of 54 Pa. Time = 0 s delineates addition of 1 mM ATP. Scale bar is 5 μm; see Movie S7. (c) Tensile force (left axis, red squares) and contour length (right axis, open triangles) as a function of time for the bundle shown in (b). (d) Contraction speed versus tensile force for data shown in (b). The dashed line approximates the force-velocity relationship observed at long times and high loads. Arrows indicate times = 0 s, 5 s, and 90 s. (e) Histogram of maximum tensile force, or stall force, of contractile bundles with $R_{M:A}^{ATP}$= 1.4 and contracted with buffer containing 1 mM ATP. (f) Stall forces calculated in (e) plotted as a function of initial bundle contour length.

Figure 5

Unloaded contraction speed is proportional to bundle length. (a) Maximal contraction speed of untethered bundles plotted as a function of initial bundle length for $R_{\text{M}:\text{A}}^{\text{ATP}}$ = 1.4 (solid circles). Open squares indicate zero-load velocities extrapolated from the inverse force-velocity relationship observed in Fig. 4, d (dashed line). Dashed line indicates a linear fit to the data with slope $\dot{\gamma }$ = 0.04 s⁻¹ (R² = 0.68). (b) Maximal contraction speed of untethered bundles as a function of bundle length for $R_{\text{M}:\text{A}}^{\text{ATP}}$ = 0.49. Dashed lines indicate a linear fit to the data with slope $\dot{\gamma }$ = 0.02 s⁻¹ (R² = 0.87). (c) Inverted contrast images of OG-myosin during the untethered contraction following the rupture of a bundle with $R_{\text{M}:\text{A}}^{\text{ATP}}$ = 1.4. Scale bar is 5 μm. (d) Line scans of myosin fluorescence intensity along bundle length averaged over a width of 0.5 μm for the images shown in (c), showing variations in myosin intensity. Solid black lines indicate guides to observe the movement of fiduciary marks. (e) The ratio of the change in distance between two fiduciary marks relative to their initial distance for the fiduciary marks visualized in (d) for times between 50 and 60 s. This measure of the percent contraction shows variations in the extent of contraction along the bundle length.

Figure 6

Bundle contraction operates as a series of individual contractile units Cartoons illustrating our model of how myosin densities affect the length d of contractile elements within the bundle. F-actin (chevrons) is bundled through the cross-linking of myosin thick filaments (black). A series of contractile elements each contracting with a velocity v and stall force F will result in the observed length-dependent rate of bundle contraction Lv/d and length independent stall force F. If the speed of the contractile elements remains unchanged for different myosin densities, then the higher rate of contraction per bundle length observed with high myosin densities (Fig. 5, a and b) can be explained by a decrease in d.

Figure 7

Impact of myosin filament density on bundle stability and contraction. The number n of myosin filaments (black) per F-actin (chevrons) is estimated from measured stoichiometries (Methods in the Supporting Material). In the absence of nucleotide, bundles are stable over a wide range of thick filament density (top row). ATP addition reduces the affinity of myosin cross-bridges by initiating motor mechanochemistry (middle row). A portion of thick filaments dissociate, leading to loss of bundle structure under low (n < 1) myosin densities. Intermediate myosin densities (n =1–4) retain enough cross-linking within bundles to resist ATP-induced bundle disintegration, but no contraction is observed. At the highest myosin densities (n > 4), myosin-generated forces lead to bundle contraction (bottom row).