What makes tropomyosin an actin binding protein? A perspective

Abstract

Tropomyosin is a two-chained alpha-helical coiled coil that binds along the length of the actin filament and regulates its function. The paper addresses the question of how a "simple" coiled-coil sequence encodes the information for binding and regulating the actin filament, its universal target. Determination of the tropomyosin sequence confirmed Crick's predicted heptapeptide repeat of hydrophobic interface residues and revealed additional features that have been shown to be important for its function: a 7-fold periodicity predicted to correspond to actin binding sites and interruptions of the canonical interface with destabilizing residues, such as Ala. Evidence from published work is summarized, leading to the proposal of a paradigm that binding of tropomyosin to the actin filament requires local instability as well as regions of flexibility. The flexibility derives from bends and local unfolding at regions with a destabilized coiled-coil interface, as well as from the dynamic end-to-end complex. The features are required for tropomyosin to assume the form of the helical actin filament, and to bind to actin monomers along its length. The requirement of instability/flexibility for binding may be generalized to the binding of other coiled coils to their targets.

Figure 1. Structure of striated muscle α-tropomyosin

The side chains of the alanine clusters (magenta) and consensus residues in proposed actin binding sites (cyan) (Phillips, 1986) are illustrated on a ribbon model of the 7 Angstrom structure of tropomyosin (Whitby and Phillips, 2000). The numbers correspond to the seven periodic repeats, 1-7. Note that the spacing of the periodic repeats is not perfectly regular. The Ala clusters (Brown et al., 2001) are positioned within consensus residues only in periods 1 and 5, but do not have a regular relationship to the other periods. Modified from (Singh and Hitchcock-DeGregori, 2006).

Figure 2. Solution structure of peptide models of the striated α-tropomyosin junctional complex and implications for actin binding

Modified from (Greenfield et al., 2006). a. The conformation of the C-terminal domain in complex with the N-terminal domain (cyan) compared to unbound C-terminal domain (brown). In the free form the two chains are almost parallel. Upon complex formation the chains splay apart following residue I270. b. Structure of the junctional complex between the C-terminal domain (cyan) and N-terminal domain (brown). The figure shows the overlay of two ribbon models of the complex to illustrate the maximal variation between the calculated structures due to the flexibility of the complex interface. c. Top: A model of the junctional complex was modeled into the 7 Angstrom structure of tropomyosin (Whitby and Phillips, 2000), joining two full-length tropomyosin molecules (Greenfield et al., 2009). The arrow indicates the junction; the side chains of consensus residues are shown in black. The orientations of the consensus residues have similar azimuthal relationships on sequential tropomyosin molecules. This point is emphasized in a model showing only one of the seven consensus residues (middle). Bottom: The model of the joined tropomyosin molecules was docked on a model of the actin filament (tan). The axial position is completely arbitrary, but K238 of actin (red) serves as a position marker on the actin monomers. The image shows the comparable relationship of the periods on sequential tropomyosin molecules to actin subunits in the filament. This would not be the case if the planes of the coiled coils in the junctional complex were parallel.

Figure 3. Model of tropomyosin showing consensus residues and Ala clusters highlighting regions selected for mutagenesis

a. The Ala clusters in period 2 and period 5 are in red boxes. Changing the alanines (magenta) into canonical interface residues (LVL) resulted in loss of function; mutation to QNQ had little effect on function or stability (Singh and Hitchcock-DeGregori, 2003; Singh and Hitchcock-DeGregori, 2006). b. Period 5, boxed in red, contains an Ala cluster within the region of the consensus residues. Extensive mutagenesis showed that both the surface consensus residues and a destabilizing interface are required for tropomyosin function (Singh and Hitchcock-DeGregori, 2006). c. Diagram illustrating the replacement mutants. Period 5 (hatched a red box) was replaced with period 1 (solid red box), period 2 with the consensus residues (solid red box) or the Ala cluster C-terminal to the period 2 consensus residues that contains an incomplete consensus sequence (magenta, hatched box). Function was retained only when period 5 was replaced with sequence that contained both consensus residues and a destabilized interface (Singh and Hitchcock-DeGregori, 2007). d. Diagram illustrating replacement of period 3 with period 5 sequence. Replacement of period 3 that lacks an Ala cluster, with the period 5 sequence increased the cooperativity of binding (Singh and Hitchcock-Degregori, 2009). Mutations to destabilize period 2 and period 3 consensus regions increased actin affinity and/or the cooperativity of binding.