Superhelical architecture of the myosin filament-linking protein myomesin with unusual elastic properties

Abstract

Active muscles generate substantial mechanical forces by the contraction/relaxation cycle, and, to maintain an ordered state, they require molecular structures of extraordinary stability. These forces are sensed and buffered by unusually long and elastic filament proteins with highly repetitive domain arrays. Members of the myomesin protein family function as molecular bridges that connect major filament systems in the central M-band of muscle sarcomeres, which is a central locus of passive stress sensing. To unravel the mechanism of molecular elasticity in such filament-connecting proteins, we have determined the overall architecture of the complete C-terminal immunoglobulin domain array of myomesin by X-ray crystallography, electron microscopy, solution X-ray scattering, and atomic force microscopy. Our data reveal a dimeric tail-to-tail filament structure of about 360 Å in length, which is folded into an irregular superhelical coil arrangement of almost identical α-helix/domain modules. The myomesin filament can be stretched to about 2.5-fold its original length by reversible unfolding of these linkers, a mechanism that to our knowledge has not been observed previously. Our data explain how myomesin could act as a highly elastic ribbon to maintain the overall structural organization of the sarcomeric M-band. In general terms, our data demonstrate how repetitive domain modules such as those found in myomesin could generate highly elastic protein structures in highly organized cell systems such as muscle sarcomeres.

Figure 1. Overall filament architecture of the dimeric myomesin IgH domain array My9–My10–My11–My12–(My13)₂–My12′–My11′–My10′–My9′.

(A) Schematic representation of the complete myomesin dimer. Those My domains that have been structurally investigated are shown in violet (first molecule) and blue (second molecule). (B) Ribbon representation of the complete myomesin tail-to-tail filament structure, in two different orientations, rotated around a horizontal axis by 90°. The helical linkers are shown in green. A ruler, providing an overall length estimate of the filament, is shown below. The conserved My domain/helix interface areas, shown in Figure 2B, are boxed.

Figure 2. Conserved, repetitive IgH modules of My9, My10, My11, and My12.

(A) Structure-based sequence alignment of My9, M10, My11, My12 IgH modules and My13. The residue numbers of each of the five sequences are indicated on top. The approximate locations of secondary structural elements are shown at the bottom (for further details see Figure S2). Highly conserved residues (:) and identical residues (*) are indicated in the consensus sequence line. Those residues that are involved in My domain/helix interfaces are highlighted in complementary colors (dark colors for specific hydrogen bonds, light colors for remaining interactions). The two residues (K1457 and Y1551) that have been mutated for SAXS studies (cf. Figure 3C) are boxed. (B) Structurally conserved My domain/helix interface areas in My9, My10, My11, and My12. Interacting helix residues are labeled; residues are boxed if involved in specific hydrogen bonds.

Figure 3. Limited flexibility of My–My domain arrangements, estimated from multiple crystal structures of identical My(n)–My(n+1) domain tandems.

The number of available structures, the length of connecting helices (number of residues), and the estimated tilt and twist angles , defining the arrangement of adjacent My domains, are listed. The standard deviations of these angles provide an estimate of the level of My–My domain flexibility observed. Each superposition uses the C-terminal My domain as the basis for superposition. The template structure is color-coded as in Figure 1, and the remaining superimposed structures are grey. The N- and C-termini are labeled.

Figure 4. Analysis of the repetitive structural features of the My9–My13 tail-to-tail filament.

(A) Surface presentation of the complete My9–My13 filament, with the centers of gravity indicated by spheres for each My domain. (B) Center of gravity distances, calculated for all My domain neighbor categories, from first (n, n+1) to eighth (n, n+8). (C) SAXS distance distribution plot of the wild-type My9–My13 filament (red) and two mutants K1457P (violet) and Y1551P (blue). The SAXS distance distribution plot calculated for the composite My9–My13 X-ray model is shown for comparison (thin red line). Matching additional maxima at about 60 Å, 115 Å, and 165 Å distances are indicated by dashed vertical lines.

Figure 5. Cross-validation of the overall structure of the dimeric tail-to-tail My9–My13 filament.

(A) Two-fold symmetry class average with superimposed iso-density contours. Seven stained domains can be recognized, with the two distal ones corresponding to the two terminal tagged MBP domains. The five central domains have been interpreted to be associated with the My9–My10 tandems (peaks 1 and 5), the two My11–My12 tandems (peaks 2 and 4), and the central My13 dimerization modules (peak 3). My9–My10 and My11–My12 are connected by shorter helices and therefore are less resolved as separate entities than My10–My11 and My12–My13, which are connected by longer helices (cf. Figure 1). (B) 2-D forward projection of the X-ray composite model of the dimeric My9–My13 filament (cf. Figure 1) low pass filtered to 30 Å in an orientation matching that of the EM class average. (C) Surface representation of an ab initio SAXS model of the My9–My13 filament. The My9–My13 dimer, as indicated on the right, exhibits a consistent arrangement in all data derived from EM, X-ray, and SAXS.

Figure 6. Atomic force microscopy measurements.

(A) Typical force–extension traces of My9–My13 unfolding. Domain unfolding events are marked by circles. Fits to the worm-like chain model (black traces) provide contour length increases, ΔL, from single domains unfolding. Regions of plateau force are indicated by arrows. (B) Histograms of measured contour length increases (left, black trace is Gaussian fit) and domain unfolding forces (right). (C) Single force–extension measurement of My9–My13 (bottom) with slow pulling between 0 and 50 nm extensions. The zoom into the plateau region exhibits a substructure; black traces are worm-like chain fits. The conversion of the plateau data points to contour lengths leads to the histogram shown on top, which is fitted by Gaussians (black traces). (D) Sample traces at the plateau region when stretch (red) and relaxation (blue) cycles were introduced into the otherwise continuous stretching of My9–My13. Both stretch and relaxation cycles feature the force plateau and show no hysteresis.