Cardiac Muscle Activation Blunted by a Mutation to the Regulatory Component, Troponin T*

Background: Striated muscle contraction is regulated by Ca2+ binding to troponin. Results: A mutation at position 205 of cardiac troponin T drastically attenuates the Ca2+ activation of the thin filament by altering its properties. Conclusion: The main effect of this mutation is on the cooperativity in the thin filament activation. Significance: Our study sheds light on how mutations in troponin T affect the thin filament regulation. The striated muscle thin filament comprises actin, tropomyosin, and troponin. The Tn complex consists of three subunits, troponin C (TnC), troponin I (TnI), and troponin T (TnT). TnT may serve as a bridge between the Ca2+ sensor (TnC) and the actin filament. In the short helix preceding the IT-arm region, H1(T2), there are known dilated cardiomyopathy-linked mutations (among them R205L). Thus we hypothesized that there is an element in this short helix that plays an important role in regulating the muscle contraction, especially in Ca2+ activation. We mutated Arg-205 and several other amino acid residues within and near the H1(T2) helix. Utilizing an alanine replacement method to compare the effects of the mutations, the biochemical and mechanical impact on the actomyosin interaction was assessed by solution ATPase activity assay, an in vitro motility assay, and Ca2+ binding measurements. Ca2+ activation was markedly impaired by a point mutation of the highly conserved basic residue R205A, residing in the short helix H1(T2) of cTnT, whereas the mutations to nearby residues exhibited little effect on function. Interestingly, rigor activation was unchanged between the wild type and R205A TnT. In addition to the reduction in Ca2+ sensitivity observed in Ca2+ binding to the thin filament, myosin S1-ADP binding to the thin filament was significantly affected by the same mutation, which was also supported by a series of S1 concentration-dependent ATPase assays. These suggest that the R205A mutation alters function through reduction in the nature of cooperative binding of S1.

The striated muscle thin filament comprises actin, tropomyosin, and troponin. The Tn complex consists of three subunits, troponin C (TnC), troponin I (TnI), and troponin T (TnT). TnT may serve as a bridge between the Ca 2؉ sensor (TnC) and the actin filament. In the short helix preceding the IT-arm region, H1(T2), there are known dilated cardiomyopathy-linked mutations (among them R205L). Thus we hypothesized that there is an element in this short helix that plays an important role in regulating the muscle contraction, especially in Ca 2؉ activation. We mutated Arg-205 and several other amino acid residues within and near the H1(T2) helix. Utilizing an alanine replacement method to compare the effects of the mutations, the biochemical and mechanical impact on the actomyosin interaction was assessed by solution ATPase activity assay, an in vitro motility assay, and Ca 2؉ binding measurements. Ca 2؉ activation was markedly impaired by a point mutation of the highly conserved basic residue R205A, residing in the short helix H1(T2) of cTnT, whereas the mutations to nearby residues exhibited little effect on function. Interestingly, rigor activation was unchanged between the wild type and R205A TnT. In addition to the reduction in Ca 2؉ sensitivity observed in Ca 2؉ binding to the thin filament, myosin S1-ADP binding to the thin filament was significantly affected by the same mutation, which was also supported by a series of S1 concentration-dependent ATPase assays. These suggest that the R205A mutation alters function through reduction in the nature of cooperative binding of S1.
Muscle is found in a wide range of phyla in animals (1,2). Muscle contraction results from the relative sliding between myosin filaments (thick filaments) and actin filaments (thin filaments). The contraction is driven by the molecular motions of myosin in a process coupled to the hydrolysis of ATP, converting chemical energy from ATP to mechanical energy and, in the case of the striated muscle, is regulated by the actin-binding proteins troponin (Tn) 2 and tropomyosin (Tm) in a Ca 2ϩ -dependent manner. Tn is a protein complex comprising TnT, TnI, and TnC: TnC is the Ca 2ϩ sensor, TnI interacts with actin and is able to inhibit actin-activated myosin ATPase activity, and TnT anchors the whole Tn complex onto tropomyosin (Tm). The Tn complex, along with Tm, is located on the actin filament with a stoichiometry of actin:Tm:Tn ϭ 7:1:1.
The Tn complex can be cleaved into two parts, the tail and the core domains, by mild chymotryptic digestion (3). The tail domain of the Tn complex consists of the N-terminal part of TnT (TnT1), and the core domain of the Tn complex consists of the C-terminal part of TnT (TnT2), TnI, and TnC (Fig. 1). Thus the core domain of the Tn complex is generally thought to function as a Ca 2ϩ -sensing domain. Upon Ca 2ϩ binding to the regulatory site(s) of TnC, a hydrophobic patch is exposed (4) and binds the switch region of TnI; a series of conformational transitions occurs among components in the Tn complex, which causes the release of TnI from the actin surface and the movement of Tm to the inner domain of actin (5). This leads to a partial exposure of the high affinity myosin-binding site(s) on the actin.
One of the most puzzling questions of muscle regulation is how Tm senses the changes that occur in the regulatory site (6), because Tm and the regulatory head of the Tn complex are located quite distantly from each other, with no direct linkage between them. The structural change that occurs in the regulatory head upon Ca 2ϩ binding to the regulatory site must somehow be transmitted to the tail domain of the Tn complex, because the tail domain is involved in the cooperative Ca 2ϩ activation of actin-activated myosin ATPase (7,8). Although the N-terminal part of TnT, TnT1, has been shown to interact with the C-terminal portion of Tm and covers the head-to-tail interacting region of the Tm strand, there is also evidence indicating that TnT2 is close to Cys-190 of Tm (9 -12). The Ca 2ϩ sensitivity of the actomyosin interaction requires the presence of TnT, revealing its critical role in regulating actomyosin inter-action (13)(14)(15). The core domain of the Tn complex without the tail domain can inhibit in the absence of Ca 2ϩ , yet it cannot activate actomyosin ATPase in the presence of Ca 2ϩ , suggesting that the mere removal of inhibition by TnI is not enough for full Ca 2ϩ activation above the level of ATPase activation achieved by actin-Tm (e.g. Ref. 8). These findings highlight the importance of this region, but it is still unclear how Ca 2ϩ -induced structural changes in the core domain are relayed to the tail domain.
TnT is the Tm-binding subunit and serves as a bridge between the Ca 2ϩ sensor (TnC) and the actin filament (16). The crystal structures of the core domains of the human cardiac Tn complex reveal that a proposed interaction site of TnI and TnT (thus named the IT arm region) forms a coiled-coil (17). The C-lobe of TnC integrates with the IT-arm in the crystal structure. In addition to the TnC-TnI interaction, TnT makes contact with TnC through Ca 2ϩ sites III and IV in TnC and TnT residues in the C-terminal portion of the IT-arm. A high-resolution structure of the thin filament was unavailable, making it difficult to account for structural changes that underlie Ca 2ϩ activation in the thin filament. Thus it is important to identify elements within TnT that play a critical role in the Ca 2ϩ activation of muscle.
Considering the above mentioned actin-myosin interaction modulated by Tm and the cooperative activation of the thin filaments by Ca 2ϩ , we hypothesized that there was a key element in the linker between the tail and the core of the Tn complex for the cooperative activation of the thin filaments and Ca 2ϩ -dependent regulation of muscle contraction. This linker includes a short helix preceding the IT-arm region known as the H1(T2) helix (Fig. 1). The importance of this region is highlighted by the presence of DCM-linked mutations such as R205L, R205W (18 -20), and ⌬K210 (21) and a posttranslational phosphorylation site, Thr-203 in the H1(T2) (22). Furthermore, it has also been shown that the mutations introduced to the N terminus of this helix, Thr-203 of mouse cardiac TnT, attenuates the Ca 2ϩ activation of skinned fibers (22). Although several other studies on these modifications to the H1(T2) helix are reported, identification of the detailed molecular level mechanisms, in particular how the thin filament is Ca 2ϩ -modulated through this region of TnT, is still unclear. Therefore, we have focused on the region within and near the H1(T2) helix, which encompasses Arg-205, a site associated with the known DCM-linked mutation (R205L or R205W). First, we replaced several amino acid residues in this region, including Arg-205, with alanine, and then we compared the effects of R205A with those of the other mutated residues in this region of TnT. Second, we sought to elucidate the underlying molecular mechanism whereby a part of this linker functions as a "Ca 2ϩ activator."

EXPERIMENTAL PROCEDURES
Materials and Muscle Proteins-Anilinonapthalenesulfote iodoacetamide, N-(1-pyrenyl)iodoacetamide, and phalloidin were purchased from Invitrogen. Rabbit fast skeletal myosin S1 was a gift from Dr. Larry Tobacman. Carboxypeptidase A was purchased from Sigma-Aldrich. Bovine cardiac Tm and actin were prepared from dried powders according to the methods described elsewhere (23,24). Briefly, actin was extracted from the acetone-dried powder in the form of G-actin, and then other contaminating muscle proteins including Tm and Tn were removed by a series of ultracentrifugation after the removal of unwanted materials and subsequent polymerization of actin via high salt treatment. The purified actin was stored at 4°C as monomeric form and used within 1-1.5 months.
For the in vitro motility assays described below, both myosin and actin were isolated from chicken pectoralis tissue as described previously (25). Actin was purified from chicken pectoralis muscle using the methods described by Pardee and Spudich (26) with minor modifications as described previously (25). The isolated filamentous actin was stabilized and labeled fluorescently by incubating it overnight at 4°C with TRITC/ phalloidin (Sigma-Aldrich) as described previously (25).
Non-polymerizable tropomyosin (NPTM) was prepared by treating bovine cardiac Tm with carboxypeptidase A in a manner similar to that described elsewhere (27) with several minor modifications. Instead of diisopropyl fluorophosphate, PMSF was used to inhibit contaminating proteases. In the last step after removal of any contamination by polymerizable Tm, NPTM was separated on Resource-Q (GE Healthcare) using a linear salt gradient.
Mutagenesis and Recombinant Protein Preparation-Mutagenesis of mouse cardiac TnT (cTnT) was carried out using QuikChange (Qiagen). Each mutant cTnT gene inserted in a pSBET or pET17 vector was expressed in BL21(DE3) Escherichia coli in the presence of kanamycin or ampicillin, respectively. The expressed proteins were purified according to the method described elsewhere (28). Typically, a recombinant cTnT was purified as follows. The protein was extracted from a bacterial pellet suspended with buffer C (6 M urea, 1 mM EDTA, and 20 mM Tris/HCl, pH 8) plus protease inhibitors. After sonication and centrifugation, the supernatant containing cTnT was subjected to a series of ammonium sulfate fractionation followed by dialysis using the same buffer (buffer C) as used for extraction without the protease inhibitors. Then the protein of interest was separated on a DEAE column using a linear gradient from 0 to 0.4 M NaCl. If necessary, the fractions containing cTnT were pooled and dialyzed against buffer C plus 0.1 M NaCl. Then the protein was further purified on the DEAE column with a linear gradient from 0.1 to 0.3 M NaCl. Prior to the measurements described here, a ternary complex with cTnI, cTnC, and cTnT was prepared by mixing an appropriate amount of each component in a buffer containing 6 M urea, 1 M NaCl, 5 mM MgCl 2 , and 20 mM Tris/HCl, pH 8.0, followed by a series of dialysis against buffers without urea, sequentially decreasing NaCl concentration from 1 to 0.1 M. The final dialysis buffer was 0.1 M NaCl, 5 mM MgCl 2 , and 20 mM Tris/HCl, pH 8.0 plus 0.01% NaN 3 (buffer A). Then the complexed protein was separated from unbound components by Resource-Q using a linear gradient from 100% buffer A to 100% buffer B (0.5 M NaCl, 5 mM MgCl 2 , and 20 mM Tris/HCl, pH 8.0, plus 0.01% NaN 3 ). Finally the isolated Tn complex was dialyzed against an appropriate buffer for each experiment. The expression plasmid of recombinant ␣-Tm with Ala-Ser at the N terminus was a gift from Dr. Larry Tobacman. The recombinant ␣-Tm was purified in the same fashion as described by Landis et al. (29).
Sequence Comparison-Sequence comparison was carried out using BLAST at the NCBI Web site and ClustalW at the European Bioinformatics Institute Web site (30) followed by a visual examination. Then conservation of sequences was assessed by the category described by Mirny and Shakhnovich (31). Briefly, amino acid residues were classified into six groups: aliphatic (Ala, Val, Leu, Ile, Met, and Cys), aromatic (Phe, Trp, Tyr, and His), polar (Ser, Thr, Asn, and Gln), positive (Lys and Arg), negative (Asp and Glu), and special (Gly and Pro). Thus an amino acid exchange within a group is considered conservative.
Solution S1-ATPase Assays-To assess a change in the S1-ATPase enzymatic rates in the presence of Tn complex with each cTnT mutant or the wild-type cTnT, phosphate (a product of the reaction) release was measured by the malachite green procedure (32). In brief, 0.2 M myosin S1, 5 M F-actin, 2 M bovine cardiac Tm, an indicated amount of each Tn complex, and either 2 mM EGTA or 100 M CaCl 2 were mixed in 35 mM NaCl, 5 mM MgCl 2 , and 20 mM MOPS, pH 7.0. Throughout the reaction the temperature was maintained at 25°C. The reaction was initiated by the addition of ATP (to a final concentration of 1 mM) and was terminated by the addition of perchloric acid solution. Then the phosphate amount was quantitated by absorbance at wavelength 655 nm after staining with molybdate-malachite green followed by incubation at 25°C for 20 min. The released phosphate amount was calculated with reference to a series of known amounts of phosphate used as a standard. Typically ATPase rates were obtained from 5 to 6 time points during which phosphate release was linear (less than 10% of the ATP molecules were hydrolyzed).
Ca 2ϩ Binding Measurements-The Tn complex containing anilinonapthalenesulfote iodoacetamide-labeled cTnC (C35S/ C84S/T53C) was titrated with Ca 2ϩ while subjected to constant stirring (28,33). The steady-state fluorescence measurements were carried out using a model 2000-4 Spectrofluorometer equipped with two 814 PMT photon-counting detectors (Photon Technology International). The fluorescence intensity changes of the fluorophore were monitored as aliquots of CaCl 2 solution were added. The solution conditions were 100 mM NaCl, 5 mM MgCl 2 , 2 mM EGTA, and 20 mM MOPS, pH 7.0, and the titration was carried out at 25°C. The free Ca 2ϩ concentrations were calculated using the WinMAXC standard version 2.10 program (34). S1-ADP Binding Assays-To obtain fluorescently labeled F-actin, F-actin was reacted with N-(1-pyrenyl)iodoacetamide according to the method described by Criddle et al. (35) with minor modifications. Briefly, F-actin was incubated overnight in the dark at ambient temperature with an excess amount of N-(1-pyrenyl)iodoacetamide (typically 1.5-fold molar excess relative to the protein amount) dissolved with dimethylformamide. The reaction was quenched by the addition of DTT to 1 mM. The resulting pyrene-labeled F-actin was separated from denatured protein by centrifugation at 2000 ϫ g and then sedimented by ultracentrifugation at 140,000 ϫ g. After dissolving the pellet (F-actin) with buffer G (2 mM Tris/HCl, 0.1 mM ATP, 0.1 mM CaCl 2 , and 0.01% NaN 3 ) the labeled actin was dialyzed against a buffer containing 0.15 M NaCl, 5 mM MgCl 2 , 20 mM MOPS, pH 7.0, and 0.01% NaN 3 . A mixture of 80 nM pyrenelabeled F-actin/phalloidin with 80 nM ␣-Tm or ⌬23␣-Tm and 160 nM Tn complex (containing either the wild type or R205A TnT) was titrated with myosin S1 in the presence or absence of Ca 2ϩ . In some sets of binding experiments with ␣-Tm, a large excess amount of ␣-Tm (160 nM) was used, which did not affect the results. The buffer condition used was: 0.15 M NaCl, 5 mM MgCl 2 , and 20 mM MOPS, pH 7.0, containing 0.2 mg/ml bovine serum albumin, 2 mM ADP, 1 mM DTT, 1 mM D-glucose, 1 unit/ml hexokinase, 0.2 mM P1,P5-di(adenosine-5Ј)pentaphosphate (Ap5A), and either 0.1 mM CaCl 2 or 2 mM EGTA. The measurements were carried out in a fashion similar to that for the Ca 2ϩ binding measurements described above. The changes of the pyrene fluorescence were recorded at wavelength 407 nm with excitation at 365 nm. The cell holder was maintained at 25°C throughout the experiment. To determine the maximum binding, titration curves were obtained without ADP (rigor state) but otherwise in the same fashion (shown in Fig. 7).
Analysis of the S1-ADP Binding Data-The collected data points were fitted to the Hill model (15,36) as well as the McKillop-Geeves model (37,38). According to the Hill model, each Tm-Tn unit in the thin filament containing seven actin sites exists either in state 1 (inactive) or state 2 (active); S1 binding to actin in state 1 is weak, whereas that of state 2 is strong. The association constants of S1 binding to actin in state 1 and state 2 are K 1 and K 2 , respectively. A cooperative parameter between units is expressed as Y, and LЈ is the equilibrium constant defining the transition from the inactive state (state 1) to the active state (state 2), where L is the equilibrium constant for transition of an isolated Tm-Tn unit with no neighboring units, no bound Ca 2ϩ , and no bound S1. The cooperative unit Y can be expressed as a function of interaction energies between adjacent units, i.e. Y 11 , Y 22 , Y 12 , and Y 21 . The fraction of units in each state is written as p 1 or p 2 . The following set of equations (Equations 1-7) was used to fit the data,

Muscle Thin Filament Regulation Modulated by Troponin T
where C is free S1 concentration. For simplicity, it was assumed that Y 12 ϭ Y 21 , Y 1 ϭ Y 11 /Y 12 , and Y 2 ϭ Y 22 /Y 12 in the above equations (39). Then Y and LЈ can be written as Y ϭ Y 1 ⅐Y 2 and LЈ ϭ L⅐(Y 1 /Y 2 ), respectively. For each point, the degree of saturation of actin sites occupied by S1, theta (), was calculated from the equation f⅐(F max Ϫ F)/(F max Ϫ F min ) and plotted against free S1 concentration, [S1] added Ϫ ⅐[actin] total , where F max and F min are the maximum and minimum values of fluorescence, respectively, and f is a factor obtained from the rigor experiment mentioned above (typically f is about 1.25). The subscript "i" in the above equations can be either 1 or 2, corresponding to state 1 or state 2, respectively. Finally, n is the number of actin monomers in 1 unit (for the full-length Tm, n is assumed to be 7). The association constant of S1 binding to actin in state 1 (K 1 ) was fixed at 1/20th of K 2 during the fitting procedure, as K 1 is very small compared with K 2 (15,36).
To fit the McKillop-Geeves model to the equilibrium S1 binding data, the following equations were used (Equations 8 -10) (37,38). In their three-state model, five parameters, K 1G , K 2G , K B , K T , and n, were used to define the equation. The association constant of S1 binding to actin is K 1G , and the equilibrium constant for S1 isomerization from the A state to the R state is K 2G (37). The two constants, c and K T , describe the blocked-closed equilibrium and the closed-open equilibrium of the thin filaments, respectively. The apparent number of actins switched between states is denoted by n (or it is referred to as apparent cooperative unit size).
where C is free S1 concentration. Instead of letting all of the parameters vary, some of the parameters were taken from Refs. 40 (for K 2G ) and 37 (for K B ) and fixed to reduce the complexity of the fitting procedure. For both models, the fitting procedure was carried out using Excel Solver (Microsoft), thus minimizing the sum of squared residuals. After the minimization process, R 2 , the coefficient of determination, was calculated as a mea-sure of goodness of fit. All of the R 2 values were above 0.98 (most of them were above 0.99), indicating a good fit to the models. The parameters obtained by the regression were then analyzed using a Student's t test.
In Vitro Motility Assay and Analysis-Expressed wild-type and mutant Tn and Tm were stored in 70% ammonium sulfate saturation for preservation. Therefore, prior to the in vitro motility experiments Tn and Tm were dialyzed against an actin buffer (100 mM KCl, 25 mM imidazole, 1 mM EGTA, 4 mM MgCl 2 , and 1 mM DTT) to remove the ammonium sulfate. The dialyzed Tn and Tm were then used to produce reconstituted thin filaments (RTF) under rigor conditions (i.e. in the absence of ATP) within the flow cell used for the motility assays (41).
The in vitro motility assay was performed as described previously (42) with minor modifications (43). Briefly, purified fulllength myosin was adhered to 100 g/ml myosin in myosin buffer (0.3 M KCl, 25 mM imidazole, 1 mM EGTA, 4 mM MgCl 2 , and 10 mM dithiothreitol, pH 7.4) was added to a nitrocellulosecoated flow cell followed by 0.5 mg/ml BSA to block areas of the coverslip surface not coated by myosin from interacting with the thin filaments. Fluorescently labeled actin (60 l of 20 nM TRITC) was then introduced into the flow cell, in the absence of ATP, and allowed to incubate with the myosin on the coverslip surface for 1 min. Following this step, 0.35 M Tm and 0.75 M Tn were added and allowed to incubate in the flow cell for 7 min, resulting in the formation of RTF. This method, which had been used previously, creates RTF capable of fully regulating the actomyosin interaction, as evidenced by the lack of filament movement at pCa levels above 9, while also showing a strong Ca 2ϩ -dependent increase in velocity, from pCa ϳ7 to 4 (41). A solution containing 100 nM excess Tn and Tm was also added to the final assay buffer to ensure full and complete activation regulation of the RTFs as described previously (41,44). The motion of the labeled regulated actin filaments was visualized and recorded using a Nikon Ti-U inverted microscope, with a 100ϫ, 1.4 NA oil-coupled objective with a thermometer (20/20 Technologies, Wilmington, NC) to maintain the temperature at 30.0°C for all experiments. The coverslip surface was imaged using an intensified digital camera (Stanford Photonics, Palo Alto, CA) with a frame grabber (Epix, Buffalo Grove, IL) to capture the images using PIPER Control software (version 2.5.11, Stanford Photonics). Three 30-s videos were captured at 10 frames/s at three different locations in the flow cell.
Actin filament velocity was determined by manual tracking of filament displacement divided by time from the frame rate using MTrackJ, a plug-in for ImageJ (NIH Image). In the absence of free Ca 2ϩ (pCa 10) the wild-type RTF of any of the filaments reconstituted with mutant TnT did not demonstrate myosin-directed movement. At pCa 4.0, the paths of eight individual filaments were tracked for each 30-s time block with 25-35 points typically needed to trace the path of a filament accurately over the entire recording. The velocity for each filament for a field, at pCa 4.0, was used to determine the average velocity for each field. The average velocity of the three fields was used to determine the overall velocity for each condition. Each condition was repeated 4 -8 times for the wildtype RTF and each mutant c. Thus the velocity of the filaments was calculated based only on the moving filaments, where moving was defined by the criteria below.
To gain more insight into the effect of the mutations on RTF function, we also quantified the percentage of filaments moving within a given field of view using LiTrack, a plug-in for ImageJ (45). For this analysis a 10-s sample of each data set was used to determine the fraction moving under each condition, with filaments that moved less than 1.3 m over the course of the 10-s time considered not to be moving. A 10-s segment of each 30-s video was used to minimize computer computational time. The LiTrackv7 program calculated the percentage of moving filaments by dividing the filaments defined as moving by the total number of filaments in a field of view. We also determined the fraction of filaments moving for the WT and the R205A mutation versus pCa using criteria described previously (43).
The in vitro motility data were analyzed using the following statistical method. A one-way analysis of variance was performed in SigmaPlot 11 (Systat Software, San Jose, CA) to determine whether there were differences in RTF velocity across the different mutations at saturating levels of free Ca 2ϩ (pCa 4.0), and a Tukey's HSD post hoc test was used to locate differences. The alpha level for this and all subsequent statistical tests was set at 0.05. For the wild-type cTn and cTn R205A we determined that RTF velocity was a function of pCa level, and these data were fit with a Hill equation, where V is RTF velocity, V max is the extrapolated maximum velocity, n is the Hill coefficient, and pCa 50 is the free [Ca 2ϩ ] (in Ϫlog units) at which RTF velocity is half the maximum value. The curve fits to the data were generated for the velocity versus pCa and fraction moving versus pCa using the curve-fitting feature within SigmaPlot 11. The goodness of fit was evaluated using the R 2 value. To compare the parameters of the fit across conditions, 95% confidence intervals were determined based on the number of observations and the standard error of the estimate of the parameter and differences, with non-overlap indicating significance.

RESULTS
Sequence Comparison-We compared 106 sequences from the H1(T2) region of vertebrate TnT (12 of the 106 aligned sequences are shown in Fig. 2A). First, we focused on Arg-205 because R205L and R205W are known DCM mutants (19,20). As expected, Arg-205 was a highly conserved residue ( Fig. 2A), when only vertebrate sequences were concerned. Then we chose four positions from the region containing the H1(T2) helix based on the sequence alignment in order to identify critical elements. Arg-205 and Arg-216 were chosen as highly conserved basic residues that are located on the same face of the helix. However, the position of Arg-216 is closer to the C terminus of the helix, whereas Arg-205 is near the N terminus of the same helix (Fig. 2B). As a comparison, Glu-204 was selected as one of the least conserved residues; its position immediately precedes Arg-205 and comes right after the proposed phosphorylation site, Thr-203. In addition to the above residues, His-223, a conservative residue outside the helix, was chosen because its side chain has no contact with other residues in the crystal structure and may serve as a control. We replaced each of the four selected amino acid residues with alanine to simplify the comparison between mutants. Replacement with alanine can be viewed as a side chain deletion, which is less likely to cause undesirable rearrangements within the protein and is considered a superior method to the addition of a larger side chain (46).
Solution S1-ATPase-In the solution S1-ATPase, Ca 2ϩ -induced activation was markedly impaired in the thin filaments containing cTnT R205A in the presence of Ca 2ϩ , whereas the same mutation had no effect on inhibitory ability (Fig. 3). At 1 M Tn, the thin filaments with wild-type Tn activated S1-ATPase up to 0.55 s Ϫ1 in the presence of Ca 2ϩ , whereas that of R205A only activated it up to 0.27 s Ϫ1 (to almost half of the wild type). The effect of mutation was not overcome even at 2 M Tn concentration. The specific activity for this mutant remained similar, i.e. 0.31 versus 0.59 s Ϫ1 for the wild type. Although another basic residue within the H1(T2) helix, Arg-216, is highly conserved, and its replacement with alanine, R216A, had only minor effects on the solution S1-ATPase activity (0.48 and 0.47 s Ϫ1 for 1 and 2 M Tn, respectively). The specific activity for 1 M R216A Tn was almost the same as that obtained with the thin filaments containing the H223A mutation (0.48 s Ϫ1 for 1 M Tn); its position is located outside of the H1(T2) helix. At 2 M Tn concentration, the specific activity for the H223A mutation caught up with that of the wild-type Tn. As expected from the sequence alignment ( Fig. 2A), the thin filaments containing E204A mutation stimulated the ATPase activity to a similar level as that obtained for the wild-type thin filament for all of the Tn concentrations tested.
In the absence of Ca 2ϩ , all of the mutants were able to inhibit S1-ATPase over the same Tn concentration range (Fig. 3). With no Tn in the system, the specific activity of S1-ATPase was 0.17 s Ϫ1 . The thin filament with the R205A mutation inhibited the S1-ATPase activity down to 0.072 s Ϫ1 for 2 M Tn, to a similar level as for the thin filament with the wild-type Tn (0.078 s Ϫ1 for 2 M Tn). For the thin filaments containing the other TnT mutants (2 M), i.e. R204A, R216A, and R223A, the specific activities were 0.090, 0.069, and 0.074 s Ϫ1 , respectively. Thus it is clear that Ca 2ϩ -induced activation beyond the level achieved by actin-Tm was significantly attenuated in the thin filaments with the R205A mutation in cTnT, whereas there was little effect in inhibition of the S1-ATPase. Similar results were obtained from the solution ATPase activity assays using recombinant ␣-Tm (data not shown).
In Vitro Motility Assay-RTF reconstituted with the wildtype cTnT and all of the mutated cTnT constructs were able to fully regulate actomyosin binding, as evidenced by a lack of motion in the absence of Ca 2ϩ (pCa 10) (data not shown). At pCa 4.0 the RTFs reconstituted with the mutant TnT moved at a velocity that was statistically similar to RTFs with the wildtype TnT, except for the RTFs reconstituted with R205A TnT, which moved at only about one-third of the velocity of the wildtype RTF (Fig. 4A). The effect of the R205A mutation is also evident in a histogram of the individual filament velocities (Fig.  4D). Both WT and R205A filaments were well fitted by a single Gaussian distribution, suggesting one population of filament velocities. In addition to a slowed velocity, the percentage of moving filaments was also significantly lower for the filaments containing the R205A mutation versus the wild type (Fig. 4B). In fact, only about 20% of the filaments with the R205A mutation moved even at pCa 4 compared with 50 -70% for filaments reconstituted with the wild-type cTn and those with mutations in neighboring regions. Given that 80% of the filaments did not move, the velocity data may not represent the full impact of the R205A mutation on muscle function. Thus, consistent with the solution ATPase results, the R205A mutation had the most drastic effect on the actin filament velocity, with all of other substitutions exerting nonsignificant effects, suggesting that not all the conserved residues in the region are critical for Ca 2ϩ activation.
These findings suggest that the presence of the mutation R205A in TnT prevented Ca 2ϩ activation of RTF. To test  whether this reduced activation was due to the ability of myosin to bind strongly to the filaments, we quantified RTF velocity under conditions in which the presence of rigor-like species can activate RTF (10 M ATP) (47,48). Interestingly, filaments with the R205A mutation moved at the same velocity as those with the wild-type cTn (Fig. 4C), suggesting that the increased inhibition caused by this mutation was not due to myosin being prevented from forming a strong attachment to actin.
Ca 2ϩ Binding Measurements-We also tested whether impaired Ca 2ϩ activation caused by some of the mutations could be due to reduced Ca 2ϩ sensitivity of the system by measuring Ca 2ϩ binding to the RTF as well as to Tn complexes alone (Fig. 5). In the absence of thin filaments, none of the mutations had an effect on the ability of Tn to bind to Ca 2ϩ , with all of the Tn complexes having almost the same affinity for Ca 2ϩ (pCa 50 ϭ 6.6), corresponding to dissociation constants (K d ) between 2.5 ϫ 10 Ϫ7 M and 2.8 ϫ 10 Ϫ7 M (Fig. 5D). On the other hand, the thin filament containing the R205A mutation in TnT had a significantly lower affinity for Ca 2ϩ (pCa 50 ϭ 5.4 or K d ϭ 3.9 ϫ 10 Ϫ6 M) than that with the wild-type TnT did (pCa 50 ϭ 5.6 or K d ϭ 2.3 ϫ 10 Ϫ6 M), suggesting attenuated Ca 2ϩ sensitivity by the mutation (⌬pCa 50 ϭ Ϫ0.2); those of the other mutants exhibited smaller ⌬pCa 50 values, which were not statistically significant. Thus the R205A mutation affects the affinity of TnC for calcium only in the presence of actin-Tm. Interestingly, all of the mutations within H1(T2) appeared to significantly reduce the Hill coefficient (down to 1.5ϳ2.0 from 2.5 for the wild type), whereas a mutation outside of H1(T2), i.e.
H223A, had no apparent effect either on the Hill coefficient or the pCa 50 (n ϭ 2.8 and pCa 50 ϭ 5.7). This may suggest an impairment in propagation of Ca 2ϩ -induced changes, because n Ͼ 1 is associated with more than one Ca 2ϩ -binding site, which can be explained only by cooperative Ca 2ϩ binding through the thin filaments (the N-domain of cTn has only one Ca 2ϩ -binding site, and all the cTn had Hill coefficient values that were closer to 1 in the absence of the thin filaments, as indicated in the legend of Fig. 5). The presence of both actin and Tm is required for cooperative Ca 2ϩ binding to the regulatory Ca 2ϩ site in cTn, as we demonstrated previously (49). It is noteworthy that the dissociation constant of the wild-type thin filaments was 2.3 ϫ 10 Ϫ6 M, which is higher than that of the wildtype Tn alone, 2.6 ϫ 10 Ϫ7 M (⌬pCa 50 ϭ Ϫ0.94), suggesting Ca 2ϩ desensitization. When the thin filaments were reconstituted with a cTn complex containing the R205A mutation in TnT, the mutation further decreased the affinity of the regulatory site on cTnC for Ca 2ϩ (K d ϭ 3.9 ϫ 10 Ϫ6 M; ⌬pCa 50 ϭ Ϫ0.23).
The Ca 2ϩ Concentration Dependence in the Motility Assay-As shown above, there was no change between the wild-type and R205A TnTs in the motility assay under rigor-like conditions (Fig. 4C), and a shift in pCa 50 was detected in the Ca 2ϩ binding measurement in the presence of the thin filaments (Fig.  5D). These results suggested that the R205A mutation somehow affects the step or steps believed to be regulated by Ca 2ϩ . Therefore we characterized the impact of this mutation on the velocity-pCa relationship in the motility assay (Fig. 6A). This experiment revealed that the RTF velocity of filaments recon-stituted with the R205A mutation moved more slowly at nearly every pCa level, with the difference being equal to at least 50% from pCa 4.0 to 6.0. The difference was smaller at pCa 6.5, and neither set of filaments moved at pCa 7 or 10. Despite the large drop in maximal velocity, the presence of the R205A mutation did not affect the pCa 50 value. The Hill coefficient (n) was reduced by roughly 50%, although the latter difference only trended toward significance at p ϭ 0.06. This suggests that the mutation has little impact on Ca 2ϩ sensitivity in the motility but may alter the cooperative nature of activation, as well as the maximal Ca 2ϩ activation.
The percent moving as a function of pCa was also calculated, and the data indicate that the R205A mutant velocities at pCa 6.5 and 6.0 reflect the velocity of only a few filaments (Fig. 6B). Thus the impact of the mutation was revealed to be much more severe considering the reduction in percent moving. S1-ADP Binding Assays-How does R205A in the H1(T2) attenuate the Ca 2ϩ -induced activation of the myosin ATPase? Is the reduced Ca 2ϩ affinity of the thin filaments solely responsible for the impaired Ca 2ϩ -induced activation? Or does the R205A mutation in the linker between the core and tail domains concurrently introduce other changes in the actomyosin filaments, such as alterations in the interaction between actin and myosin? It is possible that the observed reduction in Ca 2ϩ affinity is the cause of the impairment. It is also quite likely that the mutation causes both reduced Ca 2ϩ affinity and other changes such as reduced myosin binding to the thin filaments. In a myofilament system, Ca 2ϩ is thought to regulate the transition from weak to strong binding and therefore activation (50,51). To further elucidate the possible mechanism of Ca 2ϩinduced activation modulated through TnT, a series of assays to measure myosin S1-ADP binding to the thin filament with the wild-type or R205A TnT was carried out in the presence or absence of Ca 2ϩ (Fig. 7). There was a detectable difference in the binding profiles at low S1-ADP concentrations between thin filaments containing the wild type and the R205A TnT, especially when Ca 2ϩ was present (Fig. 7A). This result was further supported by shifts observed in 13 C, 1 H-HSQC NMR spectra for the thin filaments composed of reductively 13 Cmethylated actin reconstituted with cTn either with the wildtype or R205A TnT in the presence of both Ca 2ϩ and S1-ADP. 3 As a control, the same S1-ADP binding assay was performed on H223A. The difference in the binding profile between H223A and the wild type was not as much as that observed between R205A and the wild type (Fig. 7, A and B).
First, the binding curves obtained were fit according to the model described by Hill et al. (36) to account for the observed difference. In this model, an S1 molecule can interact with actin weakly (state 1) or in a strongly bound state (state 2) where each Tm-Tn unit in the thin filaments is either in state 1 (inactive) or state 2 (active) with cooperative behavior governing the transition from state 1 to state 2. The cooperative transition from state 1 to state 2 occurs in two ways: 1) by a change within each Tm unit (each Tm-Tn unit contains seven actin sites and they change state as a group.); and 2), by nearest-neighbor interactions between units, an important feature of this model. The association constants of S1 binding to actin in state 1 and state 2 are K 1 and K 2 , respectively. A cooperative parameter between units is expressed as Y, and LЈ is the equilibrium constant defining the transition from the inactive state (state 1) to the active state (state 2). Note that K 1 was fixed at 5% of K 2 during the fitting procedure (15,36).
The effect of the R205A mutation was most prominent as a reduction in the number of cooperative units (Y) in the presence of Ca 2ϩ ( Table 1). The average value of Y for the wild type is 33, and that for R205A is 15, which is significantly smaller than the wild type, with the relative ratio of Y between the wild type and R205A mutant at 2.2-fold, indicating a dramatic effect by the R205A mutation. The average Y value for H223A is 35, and is not significantly different from that of the wild type, as expected. Conversely, the other parameters are not significantly different between the wild type and R205A (regardless of Ca 2ϩ concentration). The association constant representing strong binding, K 2 , is 1.6 ϫ 10 6 M Ϫ1 for the wild type and 2.0 ϫ 10 6 M Ϫ1 for R205A in the presence of Ca 2ϩ . The average value of the equilibrium constant, LЈ, for the wild type is 2.1, and that for R205A is 3.6. However, the difference in LЈ between the wild type and R205A is not statistically significant. Therefore, the result can be interpreted as a decrease in the number of acti-3 M. Kobayashi, unpublished data.  . Myosin S1-ADP binding to the thin filaments. The degree of saturation of actin with S1 () is expressed as a function of free S1 concentration in the absence or presence of Ca 2ϩ . The magnified portion at the lower free S1-ADP concentrations is shown on the left panels of A-C. A, S1-ADP binding to the thin filament containing either the wild-type or R205A TnT in the presence of Ca 2ϩ (square, wild type; circle, R205A). B, S1-ADP binding to the thin filament containing either the wild-type or H223A TnT in the presence of Ca 2ϩ (square, wild type; triangle, H223A). C, S1-ADP binding to the thin filament containing either the wild-type or R205A TnT in the absence of Ca 2ϩ (open square, wild type; open circle, R205A). D, a typical set of rigor S1 binding experiments (without ADP or ATP) using the thin filament containing the full-length Tm with the wild-type TnT (square) or with R205A TnT (circle). This set of experiments determines the maximal fluorescent change necessary to calculate . vated actin molecules, mainly because of the lower cooperativity in the thin filaments with high Ca 2ϩ concentration.
On the other hand, although there was a similar trend in cooperative unit Y between the wild type and R205A, i.e. 23 and 15, respectively, the difference was not significant in the absence of Ca 2ϩ . As in the presence of Ca 2ϩ , the other parameters of R205A were not statistically different compared with those of the wild type (Table 1).
To confirm the above results experimentally, we carried out S1-ADP binding assays utilizing ⌬23␣-Tm (an internal deletion mutant of ␣-Tm (29,52), which is capable of forming the headto-tail interaction between units as well as retaining TnT-binding sites) in addition to the binding experiments with the fulllength ␣-Tm described above. Thin filaments with ⌬23␣-Tm exhibited a more pronounced difference in the S1-ADP binding curves between the wild type and R205A in the presence of Ca 2ϩ with more than a 3-fold decrease in Y by the R205A mutation (data not shown), further supporting the above finding on the cooperative parameter. The binding curves could also be interpreted within the framework of the McKillop-Geeves model of thin filament activation (37). In the McKillop-Geeves three-state model, the filament can exist in three different states, i.e. the blocked state, the closed state, and the open state. Myosin S1 does not bind to the blocked state but binds weakly in the closed state; isomerization of S1 from the A state to the R state is induced in the open state to form strong binding. In their three-state model, five parameters, K 1G , K 2G , K B , K T , and n, are used to define the fraction of total actin sites occupied (38). The association constant of S1 binding to actin is K 1G and the equilibrium constant for S1 isomerization from the A state to the R state is K 2G . The two equilibrium constants, K B and K T , describe the blocked-closed equilibrium and the closed-open equilibrium of the thin filaments, respectively. The apparent cooperative unit size is denoted by n. Two of the parameters, K 2G and K B , are obtained from independent experiments utilizing stopped-flow kinetics for K 2G (40) and K B (37, 53) by their method.
To fit the McKillop-Geeves model to our equilibrium S1 binding data without performing independent stopped-flow experiments, the assumption was made that K 2G is not dependent on either Ca 2ϩ concentration or the mutations; thus it was fixed for all the data sets. To further reduce complexity in the fitting procedure and interpretation of the results, we proceeded to analyze the data with these additional assumptions. Scenario 1: the thin filaments either with the wild-type TnT or mutant TnTs have the same K B value but can have different K T values. Scenario 2: alternatively, K B for the thin filaments with the mutant TnT(s) can be different from that of the wild-type TnT, but the thin filaments either with the wild-type c or mutant TnTs have the same K T values. For Scenario 1, the K B values were taken from Ref. 37, and 100 and 0.3 were used for the K B value in the presence and absence of Ca 2ϩ , respectively. The other three parameters were allowed to float during the fitting. For Scenario 2, the average K T values obtained for the wildtype K T value from Scenario 1 were used to obtain the other parameters including c by fitting. Table 2 summarizes the results of the fitting to the McKillop-Geeves model. In Scenario 1, both the equilibrium constant, K T , and the cooperative parameter, n, were affected by the R205A mutation in the presence of Ca 2ϩ . The K T value was 0.13 for the wild type and 0.05 for R205A, and n was 30 for the wild type and 22 for R205A. In Scenario 2, the value of n for R205A was significantly smaller than that of the wild type (n was 31 for the wild type and 18 for R205A). Although the average K B value of R205A was about half that of the wild type, the difference was not statistically significant due to large errors. It is noteworthy that in both scenarios the interdependency in resolving into

TABLE 2 Equilibrium constants obtained by fitting the S1-ADP binding data to the McKillop-Geeves model
The values shown in parentheses were fixed during the fitting procedure. Values are means Ϯ S.E.; # and *, indicate the levels of significance of the statistical test, p Ͻ 0.1 and p Ͻ 0.05, respectively, for the mutants when compared with the wild-type Tn. In Scenario 1, K 2G and K B values were fixed, and the other three (K 1G , K T , and n) were allowed to vary. Alternatively, in Scenario 2, K 2G and K T values were fixed, and the other three were allowed to vary. c The K T values were taken from those obtained for the wild type in Scenario 1. d In Scenario 1, n⅐ T is shown, and in Scenario 2, n⅐K B is shown. two components, n and the equilibrium constant(s), should be taken into account, as suggested elsewhere (40). Therefore, we expressed the interdependency as the product of n and one of the equilibrium constants that was allowed to vary, n⅐K T or n⅐K B , and included their values in Table 2. In the absence of Ca 2ϩ , none of the parameters obtained here was statistically different between the wild type and R205A. For both scenarios, significant differences were not observed in the binding constant K 1G as in the case of the Hill model, regardless of Ca 2ϩ concentration. Hence, it can be stated that S1-ADP binding to the thin filaments containing the R205A mutation is altered through a reduction in cooperative unit size n, possibly in combination with a decrease in the equilibrium constant(s) in the presence of Ca 2ϩ .
Effects of Removal of Tropomyosin Overlap-Because activation mechanisms through the nearest-neighbor units were suggested in the Ca 2ϩ activation, we investigated the role of the Tm overlap regions in cooperative behavior of thin filament activation by Ca 2ϩ . Myosin S1-ATPase assays utilizing NPTM were carried out at a saturating amount of Ca 2ϩ (pCa 4.0). NPTM has been used to study the cooperative behavior of thin filament through Tm (27,54). Although considerable cooperativity remains after the removal of 11 overlapping residues at the C terminus of Tm, most of the literature on the topic reports altered effects with overlap removal (55,56), including a change in cooperative behavior (27,54). As indicated in Fig. 8, the thin filaments with the intact Tm or NPTM were reconstituted with Tn containing either the wild-type TnT or R205A TnT, then ATPase rates were measured at various S1 concentrations. When compared with the intact Tm, the thin filaments reconstituted with NPTM significantly less stimulated S1-ATPase. At most of the S1 concentrations, the differences between the wild-type and R205A TnT were larger for the thin filament reconstituted with the intact Tm than with NPTM (Fig. 8). When the intact Tm was used, the S1-ATPase rate with the R205A mutation in TnT was less sensitive to S1 concentration than was the wild type, requiring a higher S1 concentration (ϳ3 mM S1) to reach its maximum rate. However, when NPTM was used in place of the intact Tm, the difference between the wild type and R205A almost disappeared, regardless of S1 concentrations. Thus we concluded that the overlap region of Tm is involved in increasing the degree of impairment caused by the R205A mutation, at least in an S1 concentration-dependent manner; however, a residual decrease by the mutation was still observed in the NPTM experiments, suggesting that elements other than the Tm overlap region also contribute to ATPase attenuation by the TnT mutation.

DISCUSSION
We were able to identify one of the critical residues in the short helix (the H1(T2) helix) region preceding the IT-arm, which is important for Ca 2ϩ activation of the thin filament. DCM-linked mutations have been reported to occur in this critical position 205 (R205L or R205W). Although we acknowledge that a number of previous studies have described the effects of DCM-linked mutations in this region of cTnT (18 -20), ours is the first mechanistic study linking the mutation in the H1(T2) region to the attenuation of the Ca 2ϩ activation of the thin filament by utilizing biochemical methods including a solution ATPase activity assay, Ca 2ϩ binding and S1-ADP binding measurements, and the in vitro motility assay.
We compared 106 sequences of the H1(T2) region of vertebrate TnT from different striated muscle tissues of various species. The finding suggests that Arg-205, a residue relevant to DCM, is highly conserved. The importance of this region is highlighted by the observation that R205A attenuated Ca 2ϩ activation to less than half of the wild type in both solution S1-ATPase and an in vitro motility assay (Figs. 3 and 4). However, our subsequent experiments revealed that being conserved and located in the helix is not sufficient to explain the marked impairment in Ca 2ϩ activation with the thin filament, because cTnT with an alanine replacement at Arg-216, another conserved basic residue, exhibited a much smaller effect on function in the presence of Ca 2ϩ both in the solution S1-ATPase activity assay and the in vitro motility assay when compared with the wild type. Thus we deduced that not all of the conserved residues are critical for function. The other alanine replacements, i.e. E204A and H223A, did not have a significant effect on either the solution S1-ATPase activity assay or the in vitro motility assay. Sumandea et al. (22) have demonstrated that specific phosphorylation of Thr-203 attenuates Ca 2ϩ activation of skinned fiber, affecting both the ATPase rate and isometric tension. A DCM-linked mutation, R205L, in cTnT has been shown to desensitize muscle function toward Ca 2ϩ stimulation (18). As for a deletion of lysine residue 210 (⌬K210), another known DCM mutant in the H1(T2) helix, most of the published results report that there is a significant reduction in pCa 50 (for review see Ref. 21). A subsequent biochemical study demonstrated that ⌬K210 caused Thr-203, located at the N terminus of the H1(T2) helix of TnT, to be more susceptible to phosphorylation by PKC-␣ when the TnT was complexed with TnI and TnC (57). These authors speculated that there might be a change in the electrostatic microenvironment near Thr-203 because of the deletion of Lys-210 (57). Taken together with our results, this suggests that the modification of the conserved residues located at the N-terminal side of the H1(T2) helix appears to have a more drastic, and possibly deteriorative, effect on the Ca 2ϩ -stimulated functions of cardiac muscle fibers, although it is yet to be delineated as to whether the N terminus of the helix is involved in interaction with another component of muscle proteins or is critical for the stability of the helix.
Interestingly, our in vitro motility assays indicate that the R205A mutation in cTnT only blunts Ca 2ϩ activation of the thin filament without affecting rigor activation. This is supported by the observation that under rigor-like conditions (10 M ATP, pCa 10) the RTFs with R205A TnT moved at a velocity that was statistically indistinguishable from RTF reconstituted with the wild-type TnT (Fig. 4C), suggesting that the slowed velocity of the R205A filaments is due to factors involved in the Ca 2ϩ -regulated steps of activation. This finding is further supported by the S1 binding assays, which suggest that the mutation does not affect the strong binding by myosin (Fig. 7). Therefore, these latter results suggest that increased inhibition caused by the presence of this mutation results from an inability of the RTF to be activated only through the Ca 2ϩ -mediated processes.
How then does the H1(T2) helix modulate Ca 2ϩ activation of the thin filament? In a myofilament system, Ca 2ϩ is thought to regulate, at least in part, the transition from weak to strong binding and therefore activation (50,51). Based on this principle, we hypothesized that the attenuated ATPase rate and the slowed velocity in filaments with the R205A mutation might be due to an effect related to this step of muscle activation (Figs. [3][4][5]. Ca 2ϩ binding to Tn was measured in isolation as well as complexed with a complete thin filament. In the absence of thin filament, all of the Tn complexes showed similar affinity to Ca 2ϩ (pCa ϳ6.6). In the presence of thin filament, however, the Tn complex with the R205A mutation in TnT had a significantly lower affinity for Ca 2ϩ (pCa 50 ϭ 5.4) than the wild-type Tn complex (pCa 50 ϭ 5.6) (Fig. 5). This closely matches the decrease in pCa 50 for the percentage of filaments moving in the motility assay, which with only 20% of filaments moving may be more similar to the large ensembles that were measured than the velocity-pCa relationship (Fig. 6B). Furthermore, the Hill coefficient for the R205A mutation was reduced in the Ca 2ϩ binding assays as well as in the motility assay (a roughly 50% reduction in the velocity-Ca 2ϩ relationship). These sets of experiments imply that the mutation alters this property in the thin filament. Similar observations have been made on the DCM-linked mutation, R205L, in cardiac TnT, where it is reported that a mixture of Tn complexes containing equimolar amounts of the R205L mutant and wild-type cTnT reduced Ca 2ϩ affinity in the presence of thin filament (18). Therefore Ca 2ϩ activation impairment can be explained in part by weakened binding to Ca 2ϩ .
However, it was stated above that the Ca 2ϩ activation was attenuated even at a saturating amount of Ca 2ϩ . The solution S1-ATPase activity assays were performed at a saturating concentration of Ca 2ϩ , i.e. pCa 4.0, and the enzymatic activity remained low even at a 2 M Tn concentration in the solution S1-ATPase activity assays. Hence, it is quite likely that factors other than Ca 2ϩ affinity play an important role in the propagation of Ca 2ϩ activation along the thin filament at subsaturating concentrations of Ca 2ϩ . To cause the displacement of a thin filament the myosin must be able to strongly bind to the actin filament, and our data suggest that this process is not affected. Therefore the principal effect of this mutation may be to prevent the steps following Ca 2ϩ binding to TnC but before Tm transitions to a more active state (such as state 2 in the Hill model). Thus, as hypothesized previously, this region of TnT may be crucial for communicating the Ca 2ϩ binding signal from TnC to Tm (58).
As such, the reduced affinity for Ca 2ϩ may not be the only reason for diminished Ca 2ϩ activation by the R205A mutation. For this reason, we performed myosin S1-ADP binding assays to see if there were additional factors involved in the attenuated Ca 2ϩ activation. The R205A mutation caused a significant reduction in the number of cooperative units (Y) when fit with the model described by Hill et al. (36). A similar trend was observed when the S1-ADP binding data were analyzed with the McKillop-Geeves model: the apparent cooperative unit size (n) was reduced by one-third by the R205A mutation in the presence of Ca 2ϩ . This may be the underlying mechanism of the decrease in the Hill coefficient in the velocity-pCa curve for the motility data, which also indicates decreased cooperativity (Fig. 6A). Moreover, the analysis with the McKillop-Geeves model suggested that there may be an alteration in the blockedclosed equilibrium and/or the closed-open equilibrium by the same mutation. However, it should be pointed out that a more accurate assessment of the effects (e.g. to determine whether the K B value for R205A is affected or not) would be beyond our experimental resolution without performing additional experiments (e.g. stopped-flow kinetics) independent of the steadystate S1-ADP binding. Nonetheless, our results indicate that propagation of Ca 2ϩ activation along the thin filament, evident in the cooperativity of Ca 2ϩ binding and myosin S1-ADP binding, is markedly attenuated by the mutation R205A in cTnT and that the impairment in myosin S1-ADP binding persists even at saturating Ca 2ϩ .
Because activation mechanisms through the nearest-neighbor units were suggested in the Ca 2ϩ activation, an involvement of the Tm overlaps was investigated in cooperative behavior of thin filament activation by Ca 2ϩ . Myosin S1-ATPase assays utilizing NPTM were carried out at pCa 4.0. At most of the S1 concentrations tested, greater attenuation in ATPase was caused by the R205A mutation with the Tm overlap than without the Tm overlap (Fig. 8). When the intact Tm was reconstituted in the thin filament, the S1-ATPase rate with the R205A mutation in TnT was less sensitive to S1 concentration than with the wild type in the presence of the Tm overlap. On the contrary, when NPTM was reconstituted instead of the intact Tm, the difference between the wild type and R205A almost disappeared at any S1 concentration. Thus we concluded that the overlap region of Tm is to some degree involved in the activation mechanisms through the nearest-neighbor units, although a residual decrease by the mutation was still observed in the NPTM experiments, suggesting that elements other than the Tm overlap region, such as through actin, also contribute to the ATPase attenuation by the TnT mutation.
Correlation between cross-bridge states and Ca 2ϩ binding of muscle fibers has been reported elsewhere (33, 51, 59 -62). Although a number of reports (36,37,39) as well as our current data show that Ca 2ϩ binding leads to an increase in myosin-ADP binding to the thin filament, the opposite is also true: an increase in actomyosin-ADP binding enhances Ca 2ϩ binding to the N-domain of TnC (59). Furthermore, Pinto et al. (62) have demonstrated that myosin-ADP-P i binding actually increases the affinity of TnC for the thin filament. One possible explanation of the attenuated Ca 2ϩ activation is that R205A weakens the affinity of Ca 2ϩ or TnC to the thin filament when S1-ADP is attached. Because the Ca 2ϩ activation of muscle thin filament involves two concurrent transitions, where Ca 2ϩ enhances the actomyosin interaction in the strongly bound cross-bridge state (ADP-bound) and at the same time the strong binding increases Ca 2ϩ binding to the thin filament, we propose that the linker region between the core domain and tail domain of cTn may play a role in modulating these concurrent events.
Then how is the interplay between the core domain and tail domain of cTn or between the core domain and the thin filament carried out? A study by Manning et al. (63) using molecular dynamics shows that a mutation in the TnT1 domain of cTnT (the tail domain of Tn) causes not only changes within TnT1 but also has a long-range effect, reaching up to the cTnT linker, the IT-arm, the regulatory domain of cTnI, the D-E linker of cTnC, and site II of cTnC. They also found that there is an inversely proportional correlation between the calculated flexibility of TnT1 and the cooperativity of Ca 2ϩ activation obtained from regulated in vitro motility experiments (63,64). We observed significant changes in the Ca 2ϩ binding experiments and a 50% reduction in the Hill coefficient in the velocity-pCa relationship, which indicates that the R205A mutation in the linker region of cTn alters the Ca 2ϩ binding property of cTnC when reconstituted in the thin filament. Therefore, structural changes may occur in site II of cTnC that are similar to those found in their simulations (63,64). It is also likely that R205A changes the flexibility of the TnT1 domain, which subsequently affects the cooperativity of Ca 2ϩ activation. The TnT1 domain is responsible for anchoring the Tn complex onto the thin filament through interaction with Tm. The TnT1-Tm interaction modulates the Tm-actin interaction (65,66), which may have a distant effect through the thin filament. The effect of deletion of the TnT1 region was studied by Geeves and colleagues (7), who found that the apparent number of actins switched on for the S1 binding to the thin filament reconstituted with Tn lacking TnT1 are decreased by half compared with that of the intact Tn complex. Our current finding may support the idea that the H1(T2) region plays an important role in the structural communication between the core domain and the tail domain (TnT1 region), because we also observed a decrease in the cooperativity parameters, which is similar to their observation (7). Recently, Jin and Chong (67) identified a novel Tm-binding site, different from TnT1 in TnT, which encompasses the H1(T2) region, providing new information about the mode of anchoring the Tn complex onto the thin filament (66). Moreover, it is known that removal of the head-to-tail overlap between Tm units reduces the cooperativity of myosin S1-ADP binding (27). It might be that the R205A mutation disrupts the Ca 2ϩ binding signal from TnC to Tm, which might also contribute to the breakdown of this communication of the Ca 2ϩ binding signal and affect the propagation of activation down the thin filament and thus cooperative binding. Thus we postulated that the R205A mutation may structurally perturb the junction region of Tm units either via a direct interaction with Tm or actin or through conformational changes in TnT1 domain, which then attenuates the cooperativity of Ca 2ϩ activation.
In conclusion, we have demonstrated that Ca 2ϩ activation was markedly impaired by a point mutation in a highly conserved basic residue, R205A, located near the N terminus in the short helix H1(T2) of cTnT and propose a possible mechanism by which the impairment occurs. The affinity for Ca 2ϩ decreased when R205A TnT was incorporated into the thin filament instead of the wild-type TnT. In addition to the reduction in Ca 2ϩ binding, myosin S1-ADP binding to the thin filament was also significantly affected by the same mutation, reducing the number of switched-on actin monomers, possibly by less cooperative activation of the thin filament. Further we have postulated that a similar attenuating mechanism may be at work for the dilated cardiomyopathy-linked mutation at this same site (R205L), which may initiate the cascade of events that precipitates the pathological remodeling of the myocardium.