Tropomyosin Must Interact Weakly with Actin to Effectively Regulate Thin Filament Function

Abstract

Elongated tropomyosin, associated with actin-subunits along the surface of thin filaments, makes electrostatic interactions with clusters of conserved residues, K326, K328, and R147, on actin. The association is weak, permitting low-energy cost regulatory movement of tropomyosin across the filament during muscle activation. Interestingly, acidic D292 on actin, also evolutionarily conserved, lies adjacent to the three-residue cluster of basic amino acids and thus may moderate the combined local positive charge, diminishing tropomyosin-actin interaction and facilitating regulatory-switching. Indeed, charge neutralization of D292 is connected to muscle hypotonia in individuals with D292V actin mutations and linked to congenital fiber-type disproportion. Here, the D292V mutation may predispose tropomyosin-actin positioning to a myosin-blocking state, aberrantly favoring muscle relaxation, thus mimicking the low-Ca²⁺ effect of troponin even in activated muscles. To test this hypothesis, interaction energetics and in vitro function of wild-type and D292V filaments were measured. Energy landscapes based on F-actin-tropomyosin models show the mutation localizes tropomyosin in a blocked-state position on actin defined by a deeper energy minimum, consistent with augmented steric-interference of actin-myosin binding. In addition, whereas myosin-dependent motility of troponin/tropomyosin-free D292V F-actin is normal, motility is dramatically inhibited after addition of tropomyosin to the mutant actin. Thus, D292V-induced blocked-state stabilization appears to disrupt the delicately poised energy balance governing thin filament regulation. Our results validate the premise that stereospecific but necessarily weak binding of tropomyosin to F-actin is required for effective thin filament function.

Figure 1

Molecular diagrams and energy landscapes. Landscapes were produced computationally from a grid of 290 points, each one to represent the interaction energy of a different actin-tropomyosin conformation. When generating each conformer, tropomyosin was moved as a rigid body in incremental axial and azimuthal shifts over adjacent actin subunits along an F-actin helical strand (17). (A) Although tropomyosin can be translated lengthwise along its superhelical path on F-actin without hindrance, the cartoon indicates that azimuthal rotations of tropomyosin are limited by steric barriers imposed by protruding surfaces at the extreme outer and inner edges of actin (3), where loops extend from subdomain-1 residues 23–28 (spheres, right) and also from subdomain-4 residues 231–237 (spheres, left); a single actin monomer is shown in (A), where, for clarity, the figure is not diagrammed as a full F-actin model. (A number of other key actin residues are also highlighted as spheres and colored by residue type; actin subdomains are numbered.) Landscape measurement was performed over the relatively flat surface of F-actin between the two loops mentioned (shaded gray), an area that is fully accessible to tropomyosin (3). This region of actin was thoroughly sampled by tropomyosin over multiple actin subunits along the entire F-actin model by translating tropomyosin along its superhelical axis (i.e., in the direction indicated by obliquely slanted double arrows in (A)) in 2.5 Å steps and rotating it azimuthally in 2.5° increments across F-actin (i.e., in the direction of the horizontal double arrows). (B–E) After longitudinal translation and azimuthal rotation of tropomyosin, the structures were minimized (see detailed methods in the Supporting Material) and the electrostatic interaction energy between tropomyosin and F-actin was calculated and plotted (B–E). The values plotted are the composite Coulombic interaction terms derived for tropomyosin over the actin subunits at each grid position. In these plots, the origin set at 0,0 is the longitude and azimuth for the previously determined energy minimum representing the blocked state position of wild-type tropomyosin on wild-type actin (15). The electrostatic interaction energy landscapes are shown for tropomyosin and wild-type (B and D) and D292V actin (C and E), calculated using all actin surface residues (B and C) or only for actin residues R147, D292, K326, and K328 (D and E). Plots are contoured with isolines between −2600 and −600 kcal/mol in steps of 300 kcal/mol. Note the broader and deeper minimum for the mutant compared to the wild type (compare C versus B) for both the primary 0,0 (_∗) and secondary 0,22.5 (+) minima, as well as the same trend when judged exclusively over only actin residues R147, D292, K326, and K328 (E versus D).

Figure 2

Influence of tropomyosin on the motility of wild-type (black) and mutant D292V actin (gray). The myosin-driven velocity of D292V F-actin was not affected appreciably by the mutation. The addition of tropomyosin to D292V actin abolished filament movement yet had no significant effect on wild-type actin velocity. Data points are mean ± SE. N = 6 and 10 for wild-type actin with and without tropomyosin. N = 32 and 44 for D292V actin with and without tropomyosin.

Figure 3

Tropomyosin binding to mutant and wild-type F-actin. Several concentrations of tropomyosin (0.1–3.5 μM) were incubated with 5 μM mutant D292V or wild-type F-actin for 20 min at room temperature. Tropomyosin bound to F-actin was collected by cosedimentation at 100,000 g for 20 min. (Any potential aggregated tropomyosin was removed before binding experiments by centrifugation.) Fractional saturation of F-actin by tropomyosin in pellets was calculated from the relative band densities of 12% SDS-PAGE gels using the software ImageJ and plotted as a function of free tropomyosin concentration for wild-type (black circles) and D292V (red squares) actin. Bound and free tropomyosin was determined by comparing pellet and supernatant band densities with controls of known amount in four experiments. K_50% for both sample types was ∼0.1–0.2 μM. (Inset) SDS-PAGE shows an example of concentration-dependent tropomyosin binding to F-actin in one cosedimentation experiment. Actin, black arrows; tropomyosin, gray arrows. Solution conditions: 15 mM BES buffer, 5 mM MgCl₂, 55 mM KCl, pH 7.0 at room temperature (20–25°C). To see this figure in color, go online.

Figure 4

Ca²⁺-dependence of thin filament motility. Myosin-driven motility of thin filaments reconstituted from either wild-type or mutant D292V F-actin and added troponin-tropomyosin. (A) Plots of filament velocity as a function of Ca²⁺ concentration are sigmoidal. Mutant thin filaments (gray squares) are less Ca²⁺ sensitive than are wild-type filaments (black circles) (pCa₅₀ is 6.9 for wild-type and 6.5 for D292V actin, whereas the relationship plotted yields comparable Hill coefficients of 2.6 ± 0.5 and 3.0 ± 0.7, respectively). (B) Given here are plots of the fraction of thin filaments that are motile in these preparations. Here, a low percentage of thin filaments containing D292V actin are motile (57 vs. 77% for wild-type) and this tabulation shows mutant filaments with corresponding reduced Ca²⁺ sensitivity (pCa₅₀ is 6.8 for wild-type and 6.0 for D292V actin with Hill coefficients of 3.6 ± 0.8 and 1.2 ± 0.2). The relatively low Hill coefficient for the mutant filaments in the plot in (B) is consistent with a reduced attachment rate of cycling myosin heads to actin filaments caused by the D292V substitution. Data points are means and error bars are mean ± SE calculated for 4–12 measurements.