Cardiomyopathic Tropomyosin Mutations That Increase Thin Filament Ca 2 (cid:1) Sensitivity and Tropomyosin N-domain Flexibility*

The relationship between tropomyosin thermal stability and thin filament activation was explored using two N-domain mutants of (cid:2) -striated muscle tropomyosin, A63V and K70T, each previously implicated in familial hypertrophic cardiomyopathy. Both mutations had prominent effects on tropomyosin thermal stability as monitored by circular dichroism. Wild type tropomyosin unfolded in two transitions, separated by 10 °C. The A63V and K70T mutations decreased the melting temperature of the more stable of these transitions by 4 and 10 °C, respectively, indicating destabilization of the N-domain in both cases. Global analysis of all three pro-teins indicated that the tropomyosin N-domain and C-domain fold with a cooperative free energy of 1.0–1.5 kcal/mol. The two mutations increased the apparent affinity of the regulatory Ca 2 (cid:1) binding sites of thin filament in two settings: Ca 2 (cid:1) -dependent sliding speed of unloaded thin filaments in vitro (at both pH 7.4 and 6.3), and Ca 2 (cid:1) activation of the thin filament-myosin S1 ATPase rate. Neither mutation had more than small effects on the maximal ATPase rate in the presence of saturating Ca 2 (cid:1) or on the maximal sliding speed. Despite the increased tropomyosin flexibility implied by destabilization of Tropomyosin Mutations on Thermal Stability—

nent of regulation consists of tropomyosin sterically interfering with myosin binding to actin. In the absence of Ca 2ϩ , much of the myosin-binding site on actin is obscured by tropomyosin. Calcium binding to troponin causes tropomyosin to shift position, exposing much of the myosin-binding site. Strong actinmyosin binding requires a further repositioning of tropomyosin. These findings do not imply that steric interference fully explains regulation. For example, addition of troponin and tropomyosin to bare actin filaments increases actin-myosin affinity, force production, and sliding speed in the presence of Ca 2ϩ (5)(6)(7). Nevertheless, as was first proposed 30 years ago (8), most recent reports (albeit not all (e.g. Refs. 9 and 10)) point to the shifting position of tropomyosin on actin as a critical aspect of regulation (3)(4)(5)(11)(12)(13)(14)(15)(16)(17).
Each muscle tropomyosin binds to one troponin and spans seven actin monomers on the thin filament. This stoichiometry implies that regulation is extended in space, with calcium ions directly affecting a regulatory unit that is seven actins long. Also, tropomyosin polymerizes in solution, and successive troponin-tropomyosin complexes form a continuous strand along the actin filament. Therefore, the thin filament not only undergoes kinetic transitions, but also can have spatial transitions. Any local shift in tropomyosin position on the actin surface, for example caused by a lone, strongly bound cross-bridge, implies spatial transition(s) between actins with tropomyosin in one position, and actins with tropomyosin in another. The highly cooperative behavior of muscle fibers and of isolated thin filaments suggests that these spatial transitions have considerable physiological importance. Thus, the flexibility of tropomyosin on the actin filament is significant for full appreciation of its regulatory function.
Recently, a new approach to understanding tropomyosin has been provided by the discovery that tropomyosin missense mutations can cause the autosomal dominant disorder familial hypertrophic cardiomyopathy (HCM). Several studies have appeared describing functional effects of these mutations in solution (18 -20), in fibers or cells (21)(22)(23), and in whole animals (22,24,25). Most of this work (with the exception of Ref. 23) concerns the first two mutations detected, E180G and D175N, located near where tropomyosin interacts (weakly) with the globular, Ca 2ϩ binding domain of troponin. In examining cardiomyopathic mutations A63V and K70T (26,27), located instead in the N-terminal domain of tropomyosin, the current report combines with the earlier evidence to imply that a characteristic result of the cardiomyopathic tropomyosin mutations is to decrease thermal stability, which presumably reports an increased tropomyosin flexibility. Despite these findings, the cooperativity of thin filament activation, either by myosin or by Ca 2ϩ , was not altered by the mutations. The data suggest that the flexibility of the tropomyosin N-domain influences Ca 2ϩ sensitivity, but has an unexpectedly small effect on cooperative changes in tropomyosin position on actin.

EXPERIMENTAL PROCEDURES
Protein Purification and Construct Design-Bovine cardiac whole troponin was purified as previously described from a heart muscle ether powder (28,29). Actin (30) and myosin S1 (31) were obtained from rabbit fast skeletal muscle. Recombinant control and mutant bovine tropomyosins were expressed in DE3 cells using vector pET3d, and purified to homogeneity as described (32). A63V or K70T mutations were introduced into cDNA encoding rat striated muscle ␣-tropomyosin by the same PCR-based approach used previously to create other missense mutations (33). Both wild type and mutant constructs were designed to include an N-terminal Ala-Ser dipeptide, added to functionally compensate for lack of acetylation of bacterially expressed tropomyosin (34,35). The coding sequences of the expression plasmids were confirmed by automated DNA sequencing at the University of Iowa DNA Facility.
Protein Folding and Circular Dichroism-The circular dichroism of tropomyosin ( 222 ) was monitored using an Aviv DS65 circular dichroism spectrometer, recording as a function of increasing temperature, beginning at 5°C. To correct for minor differences in protein concentration, data were normalized so that the average ellipticity between 5 and 9°C was the same for all curves. Conditions were: 0.05 mg/ml tropomyosin, 50 mM NaH 2 PO 4 (pH 6.5), and either 50 or 300 mM KCl. Data were fit to a model in which N-terminal and C-terminal regions of tropomyosin fold independently, except for a stabilizing cooperative interaction free energy, Ϫ⌬G i , for molecules with both regions folded (36). Mutations were assumed to affect the unfolding of the N-domain, i.e. N-domain melting temperature (T mN ) and unfolding enthalpy at this temperature (⌬H mN ). The mutations were assumed to have no effect on ⌬G i , C-domain melting temperature (T mC ), Cdomain unfolding enthalpy (⌬H mC ), or the ellipticity of any partially or fully folded state. For each tropomyosin, the partition function describing the relative probabilities of the fully folded, N-domain folded, C-domain folded, and unfolded states takes the form, shown by Equation 1, and Ϫ RTlnK i ϭ ⌬G i (⌬G i approximated as temperature (T)-invariant).
(Eq. 4) Better fits were obtained with the assumption that the ellipticity of the folded state varied linearly with temperature, but this did not qualitatively change the results of the analysis.
Motility Measurements and Analysis-Rhodamine-phalloidin-labeled F-actin was prepared as described previously (41) and used within 2 weeks. Nitrocellulose-coated cover slips and flow cells were prepared as previously described in Ref. 42. A heavy meromyosin solution (300 g/ml) was injected into the chamber. After 2 min of incubation, the solution was replaced by 50 l of 1 mg/ml bovine serum albumin in assay buffer (25 mM MOPS, 25 mM KCl, 2 mM MgCl 2 , 2 mM EGTA (pCa 9), 5 mM dithiothreitol, pH 7.4). One minute later 50 l of 20 nM rhodamine-phalloidin-labeled actin was introduced and allowed to incubate for 2 min. This solution was then followed by two 50-l washes of assay buffer to remove unbound actin. Next an assay buffer solution containing 200 nM tropomyosin and 200 nM troponin was introduced into the chamber and allowed to incubate for 2 min to permit regulatory protein binding to the thin filaments and to confer calcium regulation. Finally, to initiate and sustain motility, 50 l of a motility solution was introduced into the flow cell. It contained 25 mM MOPS (pH 7.4 or 6.3, depending on the experiment), 2 mM MgCl 2 , 2 mM EGTA (containing various ratios of K 2 EGTA and CaKEGTA depending on the desired pCa), 1 mM Na 2 ATP, 20 mM dithiothreitol, and photo-bleaching protective agents: 14 mM glucose, 240 units of glucose oxidase/ml, and 9 ϫ 10 3 units of catalase/ml (both enzymes from Sigma). Troponin and either wild type or mutant tropomyosin were also included in the motility solution at 200 nM concentrations. The ionic strength and pCa of the activating solution were calculated as previously described (42). The flow cell was placed on a temperature-controlled stage (25°C) of a Leica fluorescence microscope. The rhodamine-phalloidin-labeled regulated actin filaments were visualized using epifluorescence illumination (100watt mercury arc lamp and a 100ϫ, numeric aperture ϭ 1.3 objective), were imaged on a VE 1000 SIT camera (DAGE-MTI, Michigan City, IN) and were recorded without enhancement on a Panasonic VHS VCR. Quantification of thin filament sliding speed was performed using a Motion Analysis System (Motion Analysis Corporation, Santa Rosa, CA). Data were acquired as previously described (42). Typically, the motions of Ͼ200 filaments were recorded for 20 -30 s in each motility assay. The mean sliding speed for each filament was computed. Those moving at speeds less than the noise recorded from non-moving objects were assigned a value of zero, and the average sliding speed under each condition was determined by averaging all the recorded filaments. To further analyze the speed-pCa data, the speeds of every recorded filament (ϳ 200 filaments/condition) as a function of pCa were fitted to the Hill equation in the form shown in Equation 5.
V is the sliding speed at a given pCa, V max is maximal sliding speed at saturating calcium, n is the Hill coefficient, pCa 50 is the pCa at which sliding speed is half-maximal, and pCa is the pCa of the measurement. As reported previously (41), the number of filaments sliding at speeds greater than zero increased as the pCa was reduced. Typically, the fraction of filaments sliding at pCa 9 was Ͻ0.5%, whereas at pCa Ͻ 6 the fraction sliding was Ͼ90%.
FIG. 1. Thermal denaturation of control and mutant tropomyosins. Increasing temperature caused a biphasic change in the circular dichroism signal produced by control tropomyosin (filled circles), consistent with two unfolding transitions. Mutations A63V (squares) and K70T (open circles) destabilized the N-terminal portion of tropomyosin, plainly evident by shifts in midpoint of the higher temperature, N-domain transition. However, note that the full thermal denaturation curves were affected by mutational N-domain destabilization, not just the higher temperature transitions, implying cooperative interactions between the N-and C-regions of tropomyosin. Solid lines are best-fit curves corresponding to parameters in Table I (50 mM  KCl data).

Effects of Tropomyosin Mutations on Thermal Stability-
Tropomyosin folding stability is weakened by HCM mutation E180G, and by mutation D175N in pyrene-labeled tropomyosin. Both mutations are located in a tropomyosin region that interacts with troponin, albeit weakly (1). Similar assessments of folding stability have not been reported for mutations in the N-terminal half of tropomyosin, far from residues directly binding to troponin. However, structural reports suggest that such mutations might alter protein folding. Mutation V95A, located one third of the way from N to C terminus, inserts an alanine in the coiled-coil dimerization core, and clusters of such alanines alter core packing and induce tropomyosin bending (43). Lysine 70, eliminated in patients with mutation K70T, forms a stabilizing intrahelical salt bridge with Asp-66 in the x-ray structure of tropomyosin fragment 1-81 (43). Of the four identical strands in the unit cell (on two molecules), the one (helix B) with this salt bridge is the most ordered locally, i.e. its C-terminal portion has the lowest refined temperature factors. 2 Against this background, circular dichroism was employed to monitor protein folding as a function of temperature for wild type tropomyosin and for HCM mutants K70T and A63V ( Fig.  1). At intermediate temperatures, each mutation significantly altered 222 , a measure of ␣-helical content. Wild type tropomyosin unfolded in two distinct transitions, separated by ϳ10°C. The lower temperature transition can be assigned to the more C-terminal portion of tropomyosin, both because previous work indicates that tropomyosin is more stable toward its N terminus (44,45), and also because these N-terminal mutations primarily alter the higher temperature transition. The N-terminal transition is smaller than the C-terminal transition: ϳ98 versus 185 peptide bonds, respectively, based on fitting to the model described below. A63V decreases the Nterminal transition midpoint (T mN ) by 3.7°C (Table I). For K70T tropomyosin, the shift in T mN is more pronounced (10.3°C), and the C-terminal transition also is plainly affected.
Insight into tropomyosin folding thermodynamics per se, plus further analysis of the mutations, was obtained by global fitting of an equilibrium model to 222 data from all three molecules. Because two transitions are evident for wild type tropomyosin, the model adopts the approximation that tropomyosin has two folding domains under the examined conditions. The A63V and K70T mutations affect only the more N-terminal domain in the model, i.e. the unfolding enthalpy of the domain (⌬H mN) and/or melting temperature (T mN ). The CD results indicate major changes in these parameters (Table I), implying that the mutations affect the structure and dynamics of a large region of tropomyosin. Furthermore, both mutations also affect (indirectly) the behavior of the more C-terminal portion of tropomyosin, as is particularly evident for K70T. The entire circular dichroism curve is shifted toward lower temperatures, not just the portion of the curve above 40°C. This shows that cooperative interactions within tropomyosin span virtually the full length of the coiled-coil, and that a tropomyosin point mutation can affect the entire molecule in this sense. Fig. 1 was more precisely analyzed by assuming there is a cooperative interaction free energy (36) between two folding domains (Ϫ⌬G i ), which is lost when either of them unfolds. This is equivalent to stating that the N-and C-terminal regions can fold separately, but that the results imply that folding together is accompanied by a facilitating interaction.  (Table I). These calculations should be interpreted cautiously, because other experiments (46 -48) imply that tropomyosin folding is more complex than accounted for by this two-domain model. Additionally, the present data do not permit separate assessment of dimerization, which is omitted from the model. However, the measurement of a significant value for ⌬G i implies long range 2 J. H. Brown and C. Cohen, personal communication.

FIG. 2. Effects of pH and of tropomyosin K70T mutation on in vitro motility.
Rhodamine-phalloidin-labeled thin filaments were applied to heavy meromyosin-coated surfaces, and sliding speed monitored in the presence of ATP. Acidic pH weakened the apparent Ca 2ϩ affinity by an order of magnitude, and decreased the maximal sliding speed by ϳ25%. The K70T mutation increased the apparent Ca 2ϩ affinity 2.1-fold in the presence of pH 7.4, and 1.4-fold at acidic pH. Data points are averages of all continuously moving filaments (ϳ200) at a given pH. All such filaments were used to determine the best-fit lines, calculated using parameters included in Table II. cooperativity within tropomyosin. This is consistent with the folded tropomyosin structure having a continuous ␣-helix, as is likely from its amino acid sequence. Effects on in Vitro Motility-Actin filament sliding over a heavy meromyosin-coated surface is in several respects a suitable, purified protein correlate of the unloaded shortening velocity of muscles fibers (41,42). This correlation is now shown to apply to the complex effects of pH on troponin-regulated thin filaments. Consistent with muscle fiber results in the absence of thin filament mutations (49 -52), acidic pH decreased maximal sliding speed, decreased the apparent Ca 2ϩ affinity markedly, and increased the cooperativity of activation by Ca 2ϩ (Fig.  2 and Table II).
The sliding speed of actin-troponin-tropomyosin filaments is highly sensitive to Ca 2ϩ (41), and such regulation was preserved for the K70T mutation (Fig. 2). Near-maximal speed was obtained in the presence of very low Ca 2ϩ concentrations (pCa 50 Ͼ 7 at pH 7.4), consistent with previous motility data showing that, for unloaded movement, only 15% Ca 2ϩ saturation of troponin is required for near-maximal speed (6). Similar regulation was obtained for A63V tropomyosin, and results for both mutations are summarized in Table II. Notably, lower Ca 2ϩ concentrations were required to activate sliding in the presence of either A63V or K70T tropomyosin. This finding agrees with previously observed effects on the force versus pCa relationship in cultured cells (23), where even larger effects were detected. The log of the apparent Ca 2ϩ affinity for sliding speed activation was increased relative that of thin filaments containing wt tropomyosin: ⌬pCa 50 ϭ 0.33 for the effect of K70T, and ⌬pCa 50 ϭ 0.30 for the effect of A63V, in each case in the presence of pH 7.4. Similar but smaller effects were detected at pH 6.3; ⌬pCa 50 was 0.15 for the K70T mutation, and 0.22 for the A63V mutation (Table II). Although the corresponding changes in apparent affinity are modest (1.4 -2.1fold), it can be seen from Fig. 2 that they correspond to large increases in sliding speed at intermediate Ca 2ϩ concentrations. Finally, Table II shows that the A63V mutation also had small to moderate effects on maximal sliding speed (3-7 and 13-28% increases at pH 7.4 and 6.3, respectively) and cooperativity of sliding speed activation (decrease of 0.7 in Hill coefficient). Interestingly, although the K70T decreased N-domain stability more than did A63V, the former had no effect on cooperativity. It also had no effect on maximal speed.
Effects on Thin Filament-Myosin S1 ATPase Regulation-Unlike sliding speed, the thin filament-myosin S1 MgATPase rate is activated nearly (albeit not strictly) in proportion to Ca 2ϩ binding to troponin (39,53). Furthermore, the effects of many contractile protein alterations on sliding speed differ from the effects on actin-myosin ATPase activity (reviewed in Ref. 54). Therefore, mutations A63V and K70T were examined for alterations in Ca 2ϩ -sensitive regulation of MgATPase ac-tivity. Similar to findings with many other HCM-causing thin filament mutations (33, 54 -60), regulation was preserved. Addition of Ca 2ϩ increased the ATPase rate more than 10-fold, regardless of the presence of either mutation (Fig. 3). However, Ca 2ϩ sensitivities were increased. The apparent Ca 2ϩ affinities were 3.10 Ϯ 0.11 ϫ 10 5 M Ϫ1 (pCa 50  ATPase activation by Ca 2ϩ was highly cooperative for all three tropomyosins, despite the fact that cardiac troponin has only one regulatory Ca 2ϩ binding site (1). For ligand binding to a series of sites (e.g. many troponins) along a linear lattice such as the thin filament, Y Ϸ n H 2 is the -fold increase in ligand affinity when binding is to adjacent as opposed to isolated sites (39,61). Attributing the cooperative ATPase activation to troponin-Ca 2ϩ binding of this form, Y ϭ 18 Ϯ 5, 14 Ϯ 3, and 15 Ϯ 4 for control, K70T, and A63V tropomyosins, respectively. These values do not differ significantly. Finally, the mutations had no significant effects on the ATPase rates in the presence of either saturating or very low Ca 2ϩ concentrations.
Effects on Cooperative Binding of Myosin S1 to the Thin Filament-Myosin S1 binds very cooperatively to the thin filament in the absence of both Ca 2ϩ and ATP. The full structural FIG. 3. Thin filament-myosin S1 ATPase rate regulation. Ca 2ϩsensitive ATPase regulation was similar for thin filaments containing control (triangles), K70T (circles), or A63V (diamonds) tropomyosin. The mutations increased the apparent Ca 2ϩ sensitivity only slightly, 1.2-fold for K70T and 1.5-fold for A63V (see "Results"). Notably, the cooperativity of ATPase activation was not significantly affected by the mutations. basis for this is not clear, but an important component may be a rigidity in tropomyosin position on actin. Tropomyosin sterically interferes with the myosin-binding site on actin in the thin filament B-state, observed in the absence of Ca 2ϩ or myosin. For filaments saturated with myosin (regardless of Ca 2ϩ ), the position of tropomyosin is shifted, and in this M-state position there is no steric interference between myosin and tropomyosin (3). According to this mechanism of cooperativity, an energetic cost results from shifts in multiple short sections of the tropomyosin polymer, resulting in multiple B to M transition points along a thin filament, relative to the free energy with fewer transition points and fewer but longer sections shifting position (5). A more weakly folded tropomyosin, such as the A63V and K70T mutants, might be more flexible, diminishing this contribution to the cooperativity of myosin binding. However, neither mutation affected cooperative binding of S1-ADP to regulated thin filaments in the absence of Ca 2ϩ at 25°C (Fig. 4A). (Additionally, they had minimal effect on less cooperative S1-thin filament binding in the presence of Ca 2ϩ (Fig.  4B).) For both control and mutant thin filaments, the binding curves in the absence of Ca 2ϩ were S-shaped, indicating high and indistinguishable cooperativity. Furthermore, this was not the result of the temperature of the experiment, 25°C. At 37°C there is significant unfolding of the N-domain of K70T tropomyosin in the absence of actin or troponin (see Fig. 1). Nevertheless, the K70T mutation had no effect on cooperative S1thin filament binding at this higher temperature (Fig. 4C). Similar results were obtained for A63V tropomyosin (data not shown).

DISCUSSION
Of the tropomyosin mutations implicated in cardiomyopathy, E180G and D175N were the first identified (62) and at present are the best characterized, both at the protein level (18 -20) and in transgenic mice (22,24,25). These mutations increase muscle Ca 2ϩ sensitivity (21,22,24,25,63), perhaps related to their location within a putative troponin-binding region of tropomyosin (see review in Ref. 1). More surprisingly, Michele et al. (23) have shown that muscle Ca 2ϩ sensitivity is also increased by mutations A63V and K70T, and we have shown a similar effect in solution studies of these molecules (present study) and of mutant V95A (55). Many thin filament HCM mutations increase Ca 2ϩ sensitivity, regardless whether in TnT, TnI, or tropomyosin (reviewed in Ref. 64). However, the mechanism(s) for these effects are poorly understood, and are particularly unapparent in the case of the tropomyosin N-domain mutations.
HCM missense mutations have not, as a general rule, been detected in residues that directly participate in binding interactions or, in the case of myosin, in catalysis. Exceptions may be found as structural information advances, but current data are consistent with malfunction arising from indirect effects on the dynamics, structure, and/or interactions of protein regions in which the mutations occur. The present manuscript is concordant with this pattern, establishing that the A63V and K70T mutations diminish tropomyosin Ndomain folding stability. Because E180G and D175N mutations also have destabilizing effects, this appears to be a characteristic property of HCM tropomyosins. Increased thin filament Ca 2ϩ sensitivity is equally characteristic, making it interesting to consider whether these two properties are causatively linked.
The dynamic aspects of tropomyosin important for thin filament function almost certainly are distinct from the foldingunfolding reactions explored in the present study. Bound to actin, the tropomyosin N-domain does not unfold. Nevertheless, a more dynamic tropomyosin is likely in the setting of decreased folding stability. Furthermore, this may be functionally important, because tropomyosin flexibility, or rather, inflexibility, has been suggested as critical for cooperative regulation of myosin binding to the thin filament (see above). Surprisingly, the present data suggest that the mutant tropomyosins are more flexible, but this does not produce the ex- FIG. 4. Myosin S1-ADP binding to thin filaments containing control or mutant tropomyosins. Increasing myosin S1 was added to actin-troponin-tropomyosin filaments in the presence of either EGTA (25°C in A or 37°C in C) or 0.1 mM CaCl 2 (25°C in B). Binding was determined from the S1-induced decrease in fluorescence intensity of pyrene-labeled actin. Indistinguishable results were obtained, regardless whether the tropomyosin was wild type (filled circles in A and B; solid line in C), K70T (open circles in A and B; dashed line in C), or A63V (squares in A and B; not shown in C). Note that, regardless of temperature, the mutations did not alter the cooperative, S-shape of the binding curves in the presence of EGTA. Data points indicate means of three titrations Ϯ S.E. pected effect on cooperative aspects of regulation. With one exception (the effect of A63V on sliding speed), the mutations failed to decrease the cooperativity of thin filament activation in several different assays. Instead, it appears that the tendency of the tropomyosin strand to shift position on many rather than few adjacent actins (3) is preserved in the presence of the mutations. The explanation for this apparent discrepancy may be that the cooperative shifting of the tropomyosin strand depends only in part on tropomyosin (or troponin-tropomyosin) stiffness; it also depends on the shape of the free energy profile for tropomyosin contacting all possible positions across the actin filament surface (32,65), and this energy profile depends on actin-tropomyosin interactions (5). In the M-state position, these interactions are very tight, and tropomyosin binds with affinity Ͼ Ͼ 10 9 M Ϫ1 (66). Therefore, the data support the possibility that tropomyosin-actin interactions are more important for the cooperativity of the transition to the M-state, than is the intrinsic flexibility of the tropomyosin N-domain.
Tropomyosin N-domain dynamics could affect Ca 2ϩ sensitivity by a related mechanism. The rate of the Ca 2ϩ -regulated weak to strong binding transition of myosin (67, 68) may depend on tropomyosin shifting from its C-state to M-state position on actin. A more dynamic tropomyosin might lower the activation energy for this crossbridge step, increasing the apparent Ca 2ϩ affinity. This would not be detected in equilibrium binding measurements, consistent with Fig. 4. It would instead affect activation in the presence of cycling myosin (Figs. 2 and 3), and cause the largest change in pCa 50 in experiments involving major cross-bridge effects on activation. In agreement with this, A63V and K70T have larger effects on pCa 50 in force versus pCa experiments (23) than they do in the present ATPase and motility studies, in which fewer myosins attach to actin.
Alternatively, the increased Ca 2ϩ sensitivity could be a result of alteration of direct troponin-tropomyosin interactions, via effects on conformation that are propagated from the mutation sites to the troponin-binding site. We do not favor this possibility, because it would require that the mutations alter the Ca 2ϩ dependence of cardiac troponin-tropomyosin binding, and there is no such dependence for skeletal muscle tropomyosin-troponin affinity (69,70). However, the structure of the troponin-tropomyosin complex is not known, and long range conformational propagation within tropomyosin is shown both by Fig. 1 of the current study and by previous results (44, 46 -48, 71).
In summary, thermal denaturation studies of control and mutant tropomyosins show a cooperative interaction between folding of the N-domain and C-regions of tropomyosin, of ϳ1.5 kcal/mol. HCM mutations A63V and K70T prominently destabilized the tropomyosin N-domain, which indirectly affected cooperative interactions between N-and C-terminal regions of tropomyosin. Both A63V and K70T increased the apparent affinity of the regulatory Ca 2ϩ binding sites of thin filament in in vitro motility and ATPase experiments. Neither mutation had more than small effects on the maximal ATPase rate in the presence of saturating Ca 2ϩ or on the maximal sliding speed. Despite the increased tropomyosin dynamics implied by destabilization of the N-domain, neither the cooperativity of thin filament activation by Ca 2ϩ nor the cooperative binding of myosin S1-ADP to the thin filament was altered by the mutations. The current and previous results together suggest that HCM mutations locally destabilize tropomyosin, leading to an increased thin filament Ca 2ϩ sensitivity. A more dynamic tropomyosin N-domain has an unexpectedly small effect on cooperative changes in tropomyosin position on actin.