Distinct regions of troponin I regulate Ca2+-dependent activation and Ca2+ sensitivity of the acto-S1-TM ATPase activity of the thin filament.

The regions of troponin I (TnI) responsible for Ca2+-dependent activation and Ca2+ sensitivity of the actin-myosin subfragment 1-tropomyosin ATPase (acto-S1-TM) activity have been determined. A colorimetric ATPase assay at pH 7.8 has been applied to reconstituted skeletal muscle thin filaments at actin:S1:TM ratios of 6:1:2. Several TnI fragments (TnI-(104-115), TnI-(1-116), and TnI-(96-148)) and TnI mutants with single amino acid substitutions within the inhibitory region (residues 104-115) were assayed to determine their roles on the regulatory function of TnI. TnI-(104-115) is sufficient for achieving maximum inhibition of the acto-S1-TM ATPase activity and its importance was clearly shown by the reduced potency of TnI mutants with single amino acid substitutions within this region. However, the function of the inhibitory region is modulated by other regions of TnI as observed by the poor inhibitory activity of TnI-(1-116) and the increased potency of the inhibitory region by TnI-(96-148). The regulatory complex composed of TnI-(96-148) plus troponin T-troponin C complex (TnT.C) displays the same Ca2+ sensitivity (pCa50) as intact troponin (Tn) or TnI plus TnT.C while those regulatory complexes composed of TnT.C plus either TnI-(104-115) or TnI-(1-116) had an increase in their pCa50 values. This indicates that the Ca2+ sensitivity or responsiveness of the thin filament is controlled by TnI residues 96-148. The ability of Tn to activate the acto-S1-TM ATPase activity in the presence of calcium to the level of the acto-S1 rate was mimicked by the regulatory complex composed of TnI-(1-116) plus TnT.C and was not seen with complexes composed with either TnI-(104-115) or TnI-(96-148). This indicates that the N terminus of TnI in conjunction with TnT controls the degree of activation of the ATPase activity. Although the TnI inhibitory region (104-115) is the Ca2+-sensitive switch which changes binding sites from actin-TM to TnC in the presence of calcium, its function is modulated by both the C-terminal and N-terminal regions of TnI. Thus, distinct regions of TnI control different aspects of Tn's biological function.

Muscle contraction occurs through the interaction between myosin and the thin filament which is composed of actin, tropomyosin (TM), 1 and troponin (Tn). The Tn regulatory complex consists of three proteins: troponin I (TnI), the inhibitory protein; troponin T (TnT), which binds to TM; and troponin C (TnC), which binds Ca 2ϩ . In the absence of Ca 2ϩ , TnI interacts with actin-TM and inhibits the acto-S1-TM ATPase activity while in the presence of Ca 2ϩ , TnI interacts with TnC and the inhibition is released. Full biological function which includes inhibition, Ca 2ϩ -dependent release (neutralization), activation of the ATPase activity, and Ca 2ϩ sensitivity requires all three Tn components. The activation of the acto-S1-TM ATPase activity to the acto-S1 ATPase rate (unregulated actin) can occur in the presence of Tn and Ca 2ϩ or at high ratios of S1 to actin by TM alone (1). Full regulation by Tn may be due to changes in the interactions between TnI and actin, TM, TnT, or TnC. Therefore, it is critical to identify and determine the contributions of different regions of TnI toward the various aspects of Tn regulation.
Studies of skeletal TnI, using synthetic peptides, proteolytic and/or recombinant fragments have identified several regions of TnI that interact with both actin-TM and TnC (2)(3)(4). Syska et al. (5) demonstrated that the cyanogen bromide fragment of skeletal TnI, residues 96 -116, possessed all the inhibitory properties of TnI. In fact, a synthetic peptide, TnI-(104 -115), comprised the minimum sequence required for the maximum level of inhibition of the actomyosin or actomyosin fragment S1 ATPase activity (4,6,7). TnI-(104 -115) binds TnC resulting in the Ca 2ϩ -dependent release of inhibition (8,9). As well, TnI-(104 -115) has been shown to replace intact TnI in regulating the Ca 2ϩ -dependent contraction and relaxation of skinned 1 The abbreviations used are: TM, rabbit cardiac ␣␣-tropomyosin; HPLC, high performance liquid chromatography; S1, myosin subfragment 1 with the associated alkali light chains, A1 and A2; Tn, rabbit skeletal troponin; TnC, rabbit skeletal troponin C; rrTnI, rabbit skeletal recombinant TnI; crTnI, chicken skeletal recombinant TnI; TnI-(104 -115), synthetic peptide of rabbit skeletal TnI residues 104 -115 (TnI-(104 -115) has the same amino acid sequence in both chicken and rabbit); TnI-(96 -148), chicken skeletal recombinant TnI residues 96 -148, with the exception that Ser-96 has been mutated to Asn which is normally found in the rabbit sequence (the primary sequence of chicken TnI-(96 -148) differs from rabbit skeletal TnI-(96 -148) by three amino acid substitutions, S96N, R123K, and N133C); TnI 1-116, chicken skeletal recombinant TnI residues 1-116; TnI (L111G), rabbit skeletal recombinant TnI with a single glycine substitution at leucine 111; TnI (K105G), rabbit skeletal recombinant TnI with a single glycine substitution at lysine 105; TnI⅐C, rabbit skeletal TnI plus rabbit skeletal TnC complex formed at a 1:1 mol ratio; TnT⅐C⅐I, a complex formed from renaturing the individual rabbit skeletal TnI, TnT, and TnC components at a 1:1:1 molar ratio from urea/KCl; TnT; rabbit skeletal troponin T; TnT⅐C, rabbit skeletal TnT plus rabbit skeletal TnC complex formed at approximately a 1:2 molar ratio. muscle fibers (10). There are other regions within TnI that are also known to modulate TnI function. For example, the Nterminal synthetic peptide TnI-  binds to TnC and prevents the Ca 2ϩ -dependent release of TnI or TnI-(104 -115) inhibition (9). A TnI mutant in which residues 1-57 are deleted, TnI-(del57), has a modified interaction with TnC and does not interact with TnT (11,12). The N terminus of skeletal TnI is critical for the incorporation of TnI into the ternary complex with TnT and TnC. In addition, the C terminus of TnI in conjunction with the inhibitory region is critical for the Ca 2ϩdependent reconstitution of Tn activity in actomyosin ATPase assays (13). TnI-(96 -116) displays an 8-fold increase in affinity for TnC compared with TnI-(104 -115) (14) and by extending the sequence toward the C terminus TnI-(96 -148) displays an even larger increase in the binding affinity for TnC (15).
The present study was undertaken to determine the effect of the N-and C-terminal regions of skeletal TnI (situated adjacent to the inhibitory region 104 -115) on the Ca 2ϩ -dependent control of the acto-S1-TM ATPase activity and their ability to modulate this inhibitory region. Recombinant TnI fragments corresponding to residues 1-116 (TnI-(1-116)) and residues 96 -148 (TnI-(96 -148)), a synthetic peptide corresponding to residues 104 -115 (TnI-(104 -115)), and TnI mutants with single amino acid substitutions within the inhibitory region (TnI (K105G) and TnI (L111G)) display different binding affinities for actin-TM and/or the complex of troponin T and troponin C (TnT⅐C). Therefore, we designed a general protocol for the reconstitution and analysis of regulatory complexes composed of TnI mutants or TnI fragments. This protocol allows for rapid and easy evaluation of TnI fragments or mutants showing different affinities for either actin-TM and/or TnT and TnC. The manipulation of the molar ratio of the TnI fragments or mutants with respect to the other thin filament proteins allows for the reconstitution of a functional regulatory complex in the acto-S1-TM ATPase assay. Under the conditions of maximum Ca 2ϩ -response, the composition of the various regulatory complexes reconstituted on the thin filament are similar, even though the concentrations in solution may differ. We have found that the N-and C-terminal regions of TnI work in conjunction with TnT⅐C to influence different regulatory aspects of the reconstituted thin filament. That is, the C-terminal region of TnI plays a role in controlling the Ca 2ϩ sensitivity or responsiveness of the thin filament and the N-terminal region of TnI is critical for activation of the ATPase activity.

MATERIALS AND METHODS
Preparation of Proteins, Peptides, and Assay Buffers-Rabbit skeletal muscle Tn was purified according to the protocol of Ebashi et al., (16) with the following modifications: the wash buffer contained 20 mM KCl and 2 mM potassium carbonate; the extraction buffer contained 0.6 M lithium chloride and 50 mM sodium acetate, pH 4.5; contaminating TM was removed by isoelectric precipitation, pH 4.5; and the ammonium sulfate cuts were 0 -40, 40 -50, and 50 -60%. Rabbit skeletal troponin T (TnT) and rabbit skeletal troponin C (TnC) subunits were isolated according to the protocol of Pan et al. (17) with minor modifications; the column buffer contained 8 M urea, 50 mM Tris, 3 mM EGTA, 1 mM dithiothreitol, and 0.1 mM phenylmethylsulfonyl fluoride; and the gradient was 0 -0.5 M sodium chloride. Rabbit skeletal troponin I was further purified according to Wilkinson (18) with the following modifications: the column buffer contained 8 M urea, 100 mM sodium acetate, and 1 mM dithiothreitol, pH 4.5; and the gradient used was 0 -0.25 M sodium chloride. Recombinant rabbit skeletal muscle TnI (rrTnI) was expressed as described previously (19) and the TnI (K105G) and TnI (L111G) mutants were prepared according to Strauss et al. (20). Chicken skeletal muscle recombinant TnI (crTnI) and TnI-(1-116) were prepared and purified according to Farah et al. (13). Chicken skeletal recombinant TnI-(96 -148) was prepared according to Pearlstone et al. (15). The synthetic TnI peptide, TnI-(104 -115), was synthesized and purified by HPLC as described by Van Eyk and Hodges (4). The purity and authenticity of the peptides or fragments were verified by amino acid analysis and mass spectrometry.
Skeletal actin was prepared according to Spudich and Watts (21) and the final polymerization was carried out in either 6.5 mM KCl/ATPase buffer or 100 mM KCl binding buffer as described below. Cardiac ␣␣tropomyosin was prepared according to Smillie (22). Skeletal myosin subfragment 1 (S1(A1)(A2)) was prepared according to Weed and Taylor (23). The potassium-EDTA ATPase activity of the S1 was 7.2 nmol of phosphate released s Ϫ1 nmol Ϫ1 S1. Myosin S1(A1)(A2) and TM were dialyzed against 6.5 mM KCl/ATPase buffer. The potassium chloride (KCl) concentrations of the assay buffers were varied during the protein preparation and for different ATPase experiments. Therefore, when listing the buffers used for the experiments, the KCl concentration is listed in parentheses. For example, the ATPase (6.5 mM KCl) buffer refers to the assay buffer consisting of 20 mM Tris, 6.5 mM KCl, 3.5 mM MgCl 2 , 0.5 mM EGTA, 1 mM dithiothreitol, and 0.01% sodium azide (pH 7.8) while the binding buffer (100 mM KCl) contained 20 mM Tris, 100 mM KCl, 5.0 mM MgCl 2 , 0.5 mM EGTA, 1 mM dithiothreitol, and 0.01% sodium azide (pH 7.8). The KCl concentration for the ATPase assay was 6.5, 13, or 19 mM for 320 l or 19 mM KCl for a 200-l assay volume. The salt concentration of the assay buffer or the assay volume did not affect the relative acto-S1-TM ATPase activity, the pCa 50 or n H values of the ATPase/pCa curves (Table I) but did change the absolute ATPase activity. The buffers used to dissolve and dialyze Tn, TnC, or the TnT⅐C complex contained 50 M CaCl 2 and no EGTA and are identified as ATPase minus EGTA buffers.
Preparation of Tn, TnT⅐C, and Individual Tn Subunits-TnT and TnC were dissolved separately in ATPase (1 M KCl) minus EGTA buffer containing 6 M urea at approximately 5-10 mg/ml. The TnT and TnC were then dialyzed against ATPase (1 M KCl) minus EGTA buffer containing 2 M urea. TnT and TnC were combined in a 1:1 molar ratio based on peak area obtained from the elution profile at 210 nm from reversed-phase high performance liquid chromatography (for more details, see the centrifugation studies section). The TnT⅐C complex was subsequently dialyzed against the following series of buffers consisting of ATPase minus EGTA buffer containing 1, 0.5, 0.25, and 0.1 M KCl. TnT⅐C was further dialyzed to 6.5 mM KCl when TnT⅐C was used to reconstitute a regulatory complex with rrTnI, TnI (K105G), or TnI (L111G). The mole ratio of TnT to TnC in the final solution ranged between 1:1.5 and 1:1.8. This is the mole ratio required to keep TnT in solution when the concentration of KCl is 0.1 M and below. All TnT⅐C concentrations listed in the article are based on the TnT concentration.
CrTnI), TnI-(1-116), rrTnI, TnI (L111G), and TnI (K105G) were prepared using the same dialysis procedure as described for the TnT⅐C complex except the ATPase buffer contains 0.5 mM EGTA and no CaCl 2 . The final dialysis buffer for rrTnI, TnI (K105G), TnI (L111G), and crTnI was ATPase (100 mM KCl) buffer and for TnI-(1-116) the final buffer was ATPase (6.5 mM KCl) buffer. Rabbit skeletal TnI was either prepared in the same manner as the two recombinant TnI proteins (final buffer ATPase (100 mM KCl) with final concentration of approximately 1 mg/ml) or dissolved directly in H 2 O (pH ϳ 3.5) with a final concentration of approximately 4 mg/ml. Due to the higher solubility of TnI at low pH, the volume of TnI added to the assay was small and did not alter the pH of the assay buffer. TnI-(104 -115), TnI-(96 -148), and rabbit skeletal Tn were dissolved directly in ATPase (6.5 mM KCl) buffer or binding (100 mM KCl) buffer. Prior to use, the proteins and peptides were either centrifuged at 15,000 rpm at 4°C for 5 or 30 min if they were to be used in the centrifugation experiments. The concentrations of the proteins, peptides, and fragments were determined by amino acid analysis except for S1 which was determined by absorbance.
ATPase Assays-A colorimetric method was used to quantitate the amount of P i liberated by myosin S1. The colorimetric method is based on Heinonen and Lahti (24) except Tris was used in the ATPase buffer and the pH increased from 7.0 to 7.8. The assay volume was either 200 or 320 l. The amount of protein added to each assay volume was: 1.5 nmol of F-actin, 0.25 nmol of S1, and 0.429 nmol of TM (dimeric). The reaction was preincubated for 10 min at 25°C and then initiated with the addition of ATP (2 mM). The reaction was allowed to proceed for 10 -35 min depending on the assay. Control experiments demonstrated that the ATPase rate under these conditions were linear. The reaction was stopped and the quantity of liberated inorganic phosphate (P i ) determined by the addition of 2.4 ml of a 2:1:1 (v/v/v) acetone, 5 N sulfuric acid, 10 mM ammonium molybdate solution followed by the addition of 240 l of a 1 M citric acid solution. The absorbance of the individual tubes were determined at 355 nm. The endogenous S1 ATPase rate was subtracted from the data (i.e. 0.065 nmol phosphate released s Ϫ1 nmol Ϫ1 S1, 6.5 mM KCl, 320-l assay volume). Each ATPase data point was determined in triplicate. In addition, each pCa versus ATPase curve was repeated twice. The standard deviation between these curves were plotted.
In the inhibition assays, increasing quantities of TnI or TnI fragments were added until maximum inhibition of the acto-S1-TM ATPase activity was reached. For the Ca 2ϩ -dependent release assays the concentration of inhibitor required to reach maximum inhibition was used and increasing quantities of a performed TnT⅐C complex or TnC was added until maximum Ca 2ϩ activation of the acto-S1-TM ATPase activity was achieved. In the presence of 0.429 nmol of Tn (mole ratio of 1:1 with TM) the acto-S1-TM ATPase activity was determined in both the absence (pCa 8.0) and presence of Ca 2ϩ (pCa 3.5). The pCa versus ATPase curves were obtained using the concentration of TnT⅐C and TnI, TnI fragments, or TnI mutants that induce the maximum Ca 2ϩ -dependent ATPase activity. Increasing quantities of 2.64 or 5.40 mM standardized calcium chloride was added to achieve pCa values between 3.5 and 9.0. Precipitation occurred with TnI-(104 -115) and TnI-(1-116) unless there was a low concentration of Ca 2ϩ (pCa ϳ6.9) present in the assay prior to the addition of the TnT⅐C complex. At this low Ca 2ϩ concentration, the ATPase activity was similar for the regulatory complex and the TnI peptide or fragment alone. The ATPase/pCa curves ( Fig. 4) for TnI-(104 -115) and TnI-(1-116) were done in 320 l (13 mM KCl); rrTnI, TnI (K105G), and TnI (L111G) plus TnT⅐C were done in 320 l (19 mM KCl); while crTnI and TnI-(96 -148), were carried out at 200 l (19 mM KCl). The change in the assay volume from 200 to 320 l or changes in the salt concentration of the assay buffer did not alter the pCa or n H value of either rabbit skeletal Tn or rabbit skeletal TnI plus TnT⅐C (Fig.  4A, Table I). The increase in the volume did lower the overall turnover time of the complex due to the decrease in the concentration of S1 and the individual thin filament components (compare the acto-S1-TM ATPase activities of 0.455 Ϯ 0.015 to 0.656 Ϯ 0.004 nmol of phosphate released s Ϫ1 nmol Ϫ1 S1 in 320 and 200 l at 6.5 mM KCl, respectively).
The precautionary measures taken to control the calcium concentration have been described (25). Free calcium concentrations were calculated taking into account ionic strength and pH as described by Chandra et al. (14). The log of the individual binding constants used in the present work were as follows: EGTA ϩ H º EGTA⅐H, 9. 9.831. The pCa/ ATPase data were curve fitted to a sigmoidal curve using the kaleidograph program, Apple Macintosh.
Centrifugation Studies-To quantitate the amount of each regulatory protein bound to the actin thin filament, binding studies were carried out in binding (100 mM KCl) buffer in the presence of 3 mM CaCl 2 . The concentration of MgCl 2 (26 -29) and KCl (26) were increased compared with the ATPase buffer (5 and 100 mM, respectively) to ensure maximum binding of TM in the absence of Tn. In addition, the higher salt concentrations (100 mM KCl) prevented nonspecific binding of proteins to the centrifuge tubes and ensured quantitative recovery of the proteins. Since lower salt concentrations were used in the ATPase assays than in the binding experiments, it is important to note that the composition of the various regulatory complexes bound to actin-TM in the ATPase assays may be different than that observed in the binding experiments. The total assay volume used was 175 l. The concentration of actin was 1.31 M and the mole ratio of actin to TM was 7:2. Actin, TM, Tn, TnI, TnI-(1-116), and TnT⅐C were dialyzed against the binding buffer. Tn, TnI, or TnI-  in the presence and absence of the TnT⅐C were added to the actin and TM mixture. The samples were spun for 30 min at approximately 24 p.s.i. on a Beckman airfuge using a 18 Å-100 rotor. This resulted in 90 -99% of the actin being pelleted. The pellets were dissolved in 50 l of 0.05% aqueous trifluoroacetic acid. 40 l of this solution was injected on a Zorbax C8 SB300 reversed-phase column (4.6 mm inner diameter ϫ 250 mm) on a Hewlett Packard series 1090 LC coupled to a Hewlett Packard Vectra 486166 XM processor. The various proteins and fragments were eluted using a 2% B/min linear gradient where eluent A is 0.05% aqueous trifluoroacetic acid and eluent B is 0.05% trifluoroacetic acid in acetonitrile at a flow rate of 1 ml/min (40). The peak areas were determined at 210 nm and/or 280 nm and converted to nanomoles using a standard curve obtained for each protein or fragment. The standard deviation of the standard curves ranged from 5 to 8%. The amount of protein (TM, TnT, TnC, and TnI) or TnI fragment pelleted in the absence of actin (2-7.5%) was subtracted from the amount pelleted in the presence of actin. The experiments were done in triplicate and the standard deviations were calculated.

Inhibitory Action of Skeletal TnI and TnI Fragments on the
Acto-S1-TM ATPase-All of the skeletal TnI fragments except TnI-(1-116) were able to inhibit the acto-S1-TM ATPase activity to approximately the same degree as native skeletal muscle TnI (10%, Fig. 1) or intact Tn in the absence of calcium (18%, data not shown). However, the quantity required to achieve this level of inhibition markedly differed between the various skeletal TnI fragments (Fig. 1). For example, compared with rabbit skeletal TnI, TnI-(104 -115) is a less effective inhibitor requiring approximately three times the concentration to reach maximum inhibition (compare IC 50 values of 0.84 to 2.25, respectively; Table I). The importance of the inhibitory region (residues 104 -115) is also illustrated by comparing rrTnI to either TnI (K105G) or TnI(L111G). These mutants require a 2.2-3-fold higher concentration to induce the same maximum level of inhibition of the acto-S1-TM ATPase activity (compare the IC 50 of 0.62 with 1.88 and 1.37, respectively; Table I). This indicates that the single amino acid substitution of lysine at position 105 or leucine at position 111 with glycine in the intact protein reduces its apparent affinity for actin-TM. The frag-FIG. 1. Inhibition of the acto-S1-TM ATPase activity by TnI, TnI fragments, and TnI mutants. The acto-S1-TM ATPase activity was measured as a function of increasing quantities of TnI, TnI fragment (panel A), or TnI with single amino acid substitutions within the inhibitory region (panel B). Symbols: native rabbit skeletal TnI (rsTnI, q), chicken recombinant skeletal TnI (crTnI, ç), and TnI-(1-116) (Ç), TnI-(96 -148) (Ⅺ), TnI-(104 -115) (E), rabbit recombinant skeletal TnI (rrTnI, å), TnI (K105G) (É), and TnI (L111G) (छ). The acto-S1-TM activity is equal to 100%. The c indicates the 1:1 mol ratio of inhibitor to TM. Standard deviations were calculated from duplicate or triplicate data points. Where error bars are not shown, the standard deviation is small and lies under the symbol of the data point. All assays were performed in ATPase (13 mM KCl) buffer in 320 l as described under "Materials and Methods." ment, TnI-(96 -148), had similar potency as native TnI (IC 50 value of 0.77 and 0.84, respectively; Table I) which suggests that TnI-(96 -148) contains residues (other than residues 104 -115) that enhance the potency of this region.
TnI-(1-116) was the least effective inhibitor displaying no inhibition at a molar ratios of 1:1 TnI-(1-116) to TM and only 16% inhibition at a 3:1 mol ratio (Fig. 1A). Higher mole ratio could not be used since TnI-(1-116) precipitated. These results suggest that the N-terminal region of TnI antagonizes the interaction between the TnI inhibitory region (104 -115) and actin-TM either through reducing the quantity of protein bound to the actin filament or by reducing the effectiveness of the inhibitory region. To determine between these two possibilities, actin centrifugation studies were carried out as described below.
The Ca 2ϩ -dependent Release of Inhibition by TnT⅐C Complex- Fig. 2A shows the ability of rabbit skeletal TnT⅐C to release the inhibition of various TnI fragments (in the presence of Ca 2ϩ ). Rabbit skeletal TnI inhibition is released and the ATPase activity is activated to 224% by the addition of rabbit TnT⅐C at a 3:6:7 ratio of TnI:TnT⅐C:actin (Table I). This is similar to the ATPase activity induced by native rabbit Tn in the presence of Ca 2ϩ (pCa Ϸ 7.5). As well, regulatory complexes composed of either crTnI ( Fig. 2A) or rrTnI (Fig. 2B) plus rabbit skeletal TnT⅐C (compare 205 and 198%, respectively, Table I) or chicken recombinant TnI plus chicken skeletal TnT⅐C (187%, data not shown) also activated the ATPase activity. This data shows that it is possible to reconstitute a fully Ca 2ϩ -sensitive regulatory complex by addition of TnI followed by a preformed TnT⅐C complex.
The regulatory complex composed of TnI-(1-116) plus TnT⅐C reached approximately the same degree of activation of the ATPase activity as TnT⅐C and intact rabbit skeletal TnI (compare 240 to 224%, respectively, Table I). This result is in stark comparison to the regulatory complexes composed of TnT⅐C plus TnI-(104 -115) or TnI-(96 -148) which released inhibition but did not activate the ATPase activity (compare 108 and 119%, respectively, Table I). Taken together, this data suggest that the N-terminal region of TnI in conjunction with TnT⅐C is required to activate the ATPase activity to rates above acto-S1-TM. Farah et al. (13) showed that a regulatory complex formed by the renaturation of TnI-(1-116), TnT, and TnC at a 1:1:1 molar ratio from urea/KCl activates the actomyosin ATPase activity in the presence of Ca 2ϩ to 133%. This value is similar to the acto-S1-TM ATPase activity induced at a 1:1 molar ratio of TnT⅐C:TnI-(1-116) of 150% ( Fig. 2A). The large ratio of TnT⅐C:TnI-(1-116) (4:1) required for maximal Ca 2ϩ -dependent activation of the acto-S1-TM ATPase activity ( Fig. 2A) and the poor inhibition of the acto-S1-TM ATPase activity by TnI-(1-116) (16% inhibition at a molar ratio of 3:1 TnI-(1-116): TM, Fig. 1A) indicates that additional binding sites in the C-terminal region of TnI for TnT and/or TnC are needed to duplicate the effectiveness of TnI as a whole. The low ratio of TnT⅐C:TnI-(96 -148) required for maximal Ca 2ϩ dependent activity (2:1) compared with the high ratios required by TnI-(1-116) and TnI-(104 -115) (4:1 and 6:1, respectively) suggests that additional TnI residues within 117-148 are involved in TnT⅐C binding ( Fig. 2A).
Approximately a 1:1 molar ratio of TnT⅐C to rrTnI, TnI (K105G), or TnI (L111G) is required to reach the maximum Ca 2ϩ -dependent ATPase activity (Fig. 2B). Interestingly, the regulatory complex composed of TnI (L111G) plus TnT⅐C reached a higher level of ATPase activity (293%) compared with TnI (K105G) and rrTnI (221 and 198%, respectively; Table  I). Taken together, these results indicate that single amino acid substitutions in the inhibitory region do not appear to alter the efficacy of TnI binding to TnT⅐C but can modulate the extent of activation.
Centrifugation Studies-To determine whether differences in efficacy are a result of differences in binding affinity, we determined the composition of the thin filament proteins bound to actin at the various mole ratios of these proteins that were required for maximum inhibition and Ca 2ϩ -dependent activation of the ATPase activity. Actin centrifugation experiments a The assay consisted of 0.428 nmol of TM (dimeric) and 1.5 nmol of actin (ratio 2:7) and 0.249 nmol S1. The values in brackets show the ratio of skeletal TnI or Tnl fragment: TnT ⅐ C:TM:actin or Tn:TM:actin when actin is taken as 7.
b n H is the Hill coefficient determined from the ATPase/pCa curves for each regulatory complex determined from Fig. 4. c pCa 50 is the pCa (-log concentration of Ca 2ϩ ) required to induce half of the Ca 2ϩ -dependent change in ATPase activity. The pCa 50 value is determined from the ATPase/pCa curves for each regulatory complex from Fig. 4. d The IC 50 value, the concentration (M) of inhibitor to induce 50% of the inhibition of the acto-S1-TM ATPase activity, was determined from curve fitting the data in Fig. 1.
e The calculated maximum Ca 2ϩ -dependent ATPase activity determined from curve fitting the data in Fig. 2 (acto-S1-TM ATPase ϭ 100%). The acto-S1-TM ATPase activity for TnT ⅐ C at a 6:2:7 mole ratio of TnT ⅐ C:TM:actin is approximately 110%.
f The volume of assay or the salt concentration of the assay buffer, 6.5 mM KCl (low salt), or 13 mM KCl (high salt, 320 l), or 19 mM KCl (high salt, 200 l), did not significantly alter pCa 50 or n H values.
g TnI ⅐ T ⅐ C is a complex formed by renaturing from urea/KCl the individual rabbit skeletal TnI, TnT, and TnC components at a 1:1:1 molar ratio. h NA, not applicable.
were performed under conditions similar to the ATPase assay with the exception that a higher concentration of MgCl 2 and KCl was used to promote binding of TM to actin in the absence of Tn (26 -29) as well as to remove nonspecific binding of the proteins to the centrifuge tubes (see "Materials and Methods").
To monitor all of the thin filament proteins simultaneously after centrifugation, the pellet was analyzed by reversed-phase HPLC. Fig. 3 shows sample elution profiles for the pellet obtained from centrifugation studies of actin, TM, and various components of the regulatory complexes. From the HPLC elution profiles, the peak areas of the various proteins or peptides can be determined at 210 or 280 nm and converted to nanomoles using the standard curve for each protein or peptide (Table II). The binding of TM to actin is not enhanced by TnI alone but requires the presence of TnT⅐C. This is clearly demonstrated in Table II (30) has shown that TnT is responsible for the enhanced binding of TM to actin. At ratios of Tn or TnT⅐C plus TnI to TM (1:1 or 2:1:1, respectively) required for maximum Ca 2ϩ -dependent change in the acto-S1-TM ATPase activity, the amounts of TnI and TM bound to actin in the pellet are similar (experiments 4 and 5, Table II). However, 40% more TnT⅐C is bound to the thin filament when the regulatory complex is reconstituted as TnI plus TnT⅐C than when native Tn is used (experiments 4 and 5, Table II). The reason for the higher quantity of TnT⅐C bound is unclear but may reflect additional TnT-binding sites on the thin filament. Heeley and Smillie (31) have shown that TnT binding to actin does not saturate at a 1:7 ratio of TnT to actin. Interestingly, at an equivalent mole ratio of TnT⅐C to TnI (3:3:2:7 ratio of TnT⅐C: TnI:TM:actin), only about 67% of the maximum Ca 2ϩ -dependent change has occurred ( Fig. 2A) suggesting that the TnT⅐C bound is not all associated with TnI. A ratio of 2:1 TnT⅐C:TnI was required for maximum Ca 2ϩ -dependent change ( Fig. 2A).
Although an even higher mole ratio of TnT⅐C was required for maximum Ca 2ϩ -dependent release in the ATPase assays for TnI-(1-116) (4:1), Fig. 2A, the amount of the regulatory proteins bound to actin in the centrifugation experiments were similar (experiments 5 and 7, Table II). This means that under conditions of maximum calcium response, the composition of the thin filament is similar for the different regulatory complexes even though their concentrations may differ in solution.
The influence of TnT⅐C on binding of TnI-(1-116) to actin-TM is also shown in Table II. More TnI-(1-116) is bound to actin in the presence than in the absence of TnT⅐C (experiments 6 and 7 to experiment 3, ratio of 1.34 or 1.27 versus 0.08). A change in the TnT⅐C to TnI-(1-116) ratio from 1:1 to 4:1 (experiments 6 and 7, Table II) did not result in any further increase in TnI-(1-116) bound to actin-TM. The increase in the ATPase activity when the ratio of TnT⅐C increases from 1:1 to 4:1 ( Fig. 2A) is due to an increase in TnT⅐C bound to actin-TM as TnI-(1-116) does not affect the amount of TnT⅐C bound (the same amount of TnT⅐C is bound when TnI-(1-116) is not present in the assay (compare experiments 7 and 8, Table II)). One cautionary comment should be made concerning these results, although in the absence of calcium, TnI, TnT, and TnC were bound to TM-actin in similar mole ratios as in the presence of calcium, 2 this does not mean that some of the regulatory complexes composed of TnI fragments (which have different affinities for actin and TnC) and TnT⅐C may have different ratios bound to actin-TM in the absence of calcium.
Ca 2ϩ Affinity and Ca 2ϩ -dependent Cooperativity-The contribution of TnI and TnI fragments to the calcium affinity (pCa 50 ) and the calcium-dependent cooperativity (n H ) of reconstituted thin filaments was determined from the ATPase/pCa relationship. One method for reconstitution is to renature the individual subunits, TnI, TnT, and TnC at a molar ratio of 1:1:1 from urea/KCl slowly through a series of buffers containing no 2 J. E. Van Eyk, L. T. Thomas, and R. S. Hodges, unpublished data.

FIG. 2. The effect of TnT⅐C (plus Ca 2؉ ) on the inhibition of the acto-S1-TM ATPase activity by TnI, TnI fragments, or TnI mutants with single amino acid substitutions within the inhibitory region.
Increasing quantities of the TnT⅐C (TnT⅐C concentration based on TnT concentration) was added to the acto-S1-TM assay, which was maximally inhibited by TnI, TnI mutant, or TnI fragment (preformed as described under "Materials and Methods"). Panel A shows the effect of increasing quantities of TnT⅐C in the presence of calcium (pCa Х 3.5) on the inhibitory activity of native rsTnI (q), crTnI (ç), TnI-(1-116) (Ç), TnI-(96 -148) (Ⅺ), or TnI-(104 -115) (E). The various inhibitors were present at the concentration required for maximal inhibition in the absence of calcium (mole ratios of TnI or TnI fragment to actin were: rsTnI ϭ 3:7; crTnI ϭ 3:7; TnI-(1-116) ϭ 6:7, TnI-(104 -115) ϭ 6:7, and TnI-(96 -148) ϭ 3:7). Panel B shows the effect of increasing quantities of TnT⅐C in the presence of calcium (pCa Х 3.5) on the ATPase activity with rrTnI (å), TnI (K105G) (É), or TnI (L111G) (छ). The inhibitors were present at the concentration required for maximal inhibition (mole ratio of TnI or TnI mutant to actin were: rrTnI ϭ 4:7; TnI (K105G) ϭ 9:7, and TnI (L111G) ϭ 9:7). The actin and TM concentrations were 7.5 and 2.14 M, respectively. 100% is equal to the acto-S1-TM ATPase rate. The average with the standard deviation of at least three independent determinations is shown. Where error bars are not shown, the standard deviation is small and lies under the symbol of the data point. All assays were performed in ATPase (13 mM KCl) buffer in 320 l as described under "Materials and Methods." urea and decreasing concentration of KCl. A regulatory complex reconstituted in this manner (TnT⅐C⅐I) has similar pCa 50 and n H values to native Tn (Fig. 4A, Table I). However, an alternative method is required when dealing with TnI mutants or fragments that display reduced affinity for TnT, TnC, or actin. We have chosen to reconstitute the regulatory complex by the addition of TnI to the assay solution followed by the addition of a preformed TnT⅐C complex at a mole ratio of these proteins that produces the maximum Ca 2ϩ -dependent change in the acto-S1-TM-ATPase activity (Fig. 4A). For example, in Fig. 1B a 1.5:1 mol ratio of TnI to TM is required for maximum inhibition. At this concentration of TnI, a 2:1 mol ratio of TnT⅐C to TnI (equivalent to a 1.5:3:1 ratio of TnI:TnT⅐C:TM) is required for maximum ATPase activity ( Fig. 2A). At these mole ratios, the regulatory complex behaves like native Tn displaying similar Ca 2ϩ sensitivity (Fig. 4A, Table I). This reconstitution method should also work for TnI fragments which have reduced affinities for TnT⅐C and actin. To achieve maximum Ca 2ϩ response of the regulatory complex for TnI- (1-116), a molar ratio of 3:12:1 of TnI-(1-116):TnT⅐C:TM was required (Fig. 4B). Importantly the binding experiments show that the amount of protein bound to actin with the regulatory complexes composed of TnT⅐C and either TnI or TnI-(1-116) were equivalent, even though the concentration in solution differed. Thus, under these conditions where the various regulatory complexes display maximum Ca 2ϩ change in the acto-S1-TM ATPase activity, the pCa/ATPase relationship should be comparable. In fact, it is the use of the high concentrations of TnI-(1-116) and TnT⅐C required for full effect and binding which can explain the differences between the ATPase/pCa curve obtained in this study and that reported by Farah et al. (13). The difference lies with the extent of Ca 2ϩ -dependent activation by the TnI-(1-116) regulatory complexes which are prepared to have different mole ratios with respect to both TnT⅐C and actin-TM. Farah et al. (13) reconstituted the regulatory complex from a mixture containing TnI- (1-116), TnT, and TnC at a 1:1:1 mol ratio and the assay contained a 1:1 mol ratio with respect to TM. This regulatory complex displayed no Ca 2ϩ -dependent response. In this study, the regulatory complex was reconstituted on the thin filament using high mole ratios of TnI-(1-116) and TnT⅐C. Under these conditions there was full Ca 2ϩ -dependent activation (Fig. 2) and hence a ATPase/pCa curve was observed (Fig. 4).
The reconstitution of the thin filament with native rabbit skeletal TnI, rabbit recombinant or chicken recombinant skeletal TnI plus TnT⅐C complex shows similar pCa/ATPase relationships with pCa 50 values of 6.16, 6.14, and 6.08, respectively, and n H values of 4.7, 3.8, and 3.1, respectively (Table I,  part B). TnI-(96 -148) plus TnT⅐C displayed similar Ca 2ϩ sensitivity (pCa 50 ϭ 6.01) and cooperativity (n H ϭ 3.1) as that of intact TnI (Fig. 4C). However, the regulatory complexes composed of TnI-(104 -115) or TnI-(1-116) plus TnT⅐C displayed increased Ca 2ϩ affinity (pCa 50 ϭ 6.52 and 6.36, respectively) compared with TnI-(96 -148) plus TnT⅐C (pCa 50 ϭ 6.01) (Fig.  4B, Table I). The shift in pCa 50 of TnI-(1-116) and TnI-(104 -115) from TnI-(96 -148) is independent of the ability of these fragments to activate the ATPase activity, i.e. both 104 -115 and 1-116 shift the pCa 50 but only 1-116 activates. Interestingly, the largest difference in pCa 50 values occurs between peptides TnI-(104 -115) and TnI-(96 -148) (both of which do not activate the ATPase activity above the acto-S1-TM rate) suggesting that other residues outside the inhibitory region 104 -115 are responsible for the change in Ca 2ϩ sensitivity. Fig. 4D compares the pCa/ATPase curves obtained for the regulatory complexes composed of TnT⅐C plus rrTnI, TnI (L111G), and TnI (K105G) at a molar ratio of 4:6:2:7 of rrTnI: TnT⅐C:TM:actin and a 9:13.5:2:7 ratio of TnI mutant:TnT⅐C: TM:actin (the ratios required for maximum Ca 2ϩ -dependent change in the ATPase assay, Fig. 2B). The regulatory complex composed of the TnI (L111G) plus TnT⅐C has a similar pCa 50 value as rrTnI plus TnT⅐C (compare 6.22 to 6.14, respectively, Table I) while a small increase in the pCa value (6.35; Table I) is observed by TnI (K105G) plus TnT⅐C. This is in agreement with the small increase in Ca 2ϩ sensitivity (0.2 pCa units) when endogenous TnI is substituted with TnI (K105G) in cardiac skinned muscle fiber (20). In addition, skinned fibers reconstituted with TnI (K105G) had a slower rate of fiber relaxation than either fibers reconstituted with rrTnI or unextracted fibers which could reflect a decrease in the affinity of this TnI mutant for actin-TM. DISCUSSION We have outlined a general protocol for the reconstitution of regulatory complexes composed of TnI fragments and mutants which have different affinities for actin-TM, TnT, and/or TnC Following centrifugation, the pellet was dissolved in 50 l of 0.05% aqueous trifluoroacetic acid. 40 l of this sample was injected on a Zorbax C8 SB300 reversed-phase column. The proteins were eluted using a 2% B/min linear gradient where buffer A is 0.05% aqueous trifluoroacetic acid and buffer B is 0.05% trifluoroacetic acid in acetonitrile at a flow rate of 1 ml/min. TnT is eluted as a doublet as shown previously by Ingraham and Hodges (40). The retention time for the main peak of TnT, TnI, TM, TnC, and actin are approximately 19.9, 21.7, 23.5, 27.5, and 29.8 min, respectively. The nanomole of protein or fragment in the pellet was determined based on the peak area determined from standard curves generated for the various proteins and fragments. Since TnI-(1-116) has less amino acids than intact TnI, the peak area/nanomole of protein is less and thus the peak on the chromatogram is smaller than TnI for the intact Tn complex. than native TnI. The quantity of the TnI fragment or mutant required for maximum inhibition of the acto-S1-TM ATPase activity is initially determined. At this concentration of inhibitor, increasing quantities of a preformed TnT⅐C complex is added until maximum Ca 2ϩ -dependent release and/or activation of the ATPase activity occurs. The pCa/ATPase relationship of these regulatory complexes are then assayed at the molar ratio of these proteins required for maximum Ca 2ϩ -dependent ATPase activity. In this way, the pCa/ATPase relationship of these regulatory complexes can be directly compared with native Tn since the thin filament compositions are similar (even though the concentrations of the regulatory proteins differ in solution).
Using the colorimetric ATPase assay at pH 7.8 and at a molar ratio of actin:S1:TM of 6:1:2, native rabbit skeletal Tn, and the regulatory complexes composed of TnI plus TnT⅐C, showed high cooperativity (n H Ͻ 3), similar pCa values (between 6.1 and 6.2) and Ca 2ϩ -dependent activation of the ATPase activity (Ն198%) ( Table I). Thus, a functional regulatory complex can be reconstituted by the addition of TnI plus TnT⅐C and this reconstitution protocol allows for the study of complexes produced with TnI fragments or TnI mutants which have reduced affinities for actin-TM, TnT, or TnC when combined with a preformed TnT⅐C complex.
The synthetic peptide, TnI-(104 -115), has been shown to contain the minimum amino acid sequence required for maximum inhibition of the acto-S1-TM ATPase activity (4,7,32,33) and elimination of the inhibitory region of TnI abolishes or greatly reduces the ability of this TnI mutant to inhibit the ATPase activity (13,34). In the present work, the importance of the inhibitory region is further demonstrated by the reduced efficacy of the TnI mutants, which contain selective single amino acid substitutions within the inhibitory region, for inhibiting the ATPase activity compared with native TnI (Fig.  1B). This confirms our earlier work, in which the corresponding single amino acid substituted analogs of the synthetic inhibitory peptide TnI-(104 -115) were less effective inhibitors of the acto-S1-TM ATPase activity compared with the native peptide (4,7). Although native TnI-(104 -115) peptide (as well as the single amino acid substituted analogs) is able to reach the same level of inhibition as intact TnI, the TnI-(104 -115)-peptide requires approximately three times the concentration of TnI or TnI-(96 -148) to reach maximum inhibition (Table I). This indicates that additional residues within the region 96 -148 enhances the apparent binding affinity of this fragment for actin-TM and suggests that there is a second actin-binding site within this region. On the other hand, the N-terminal region TnI-(1-116) is a poor inhibitor (Ref. 13; Fig. 1A), suggesting that the N terminus of TnI acts as a negative effector and hinders the interaction of the inhibitory region with actin-TM. This is supported by the actin binding studies which showed that TnI-(1-116) binds weakly to actin compared with intact TnI (Table II).
On the other hand, the ability of Tn to activate the ATPase activity above the acto-S1-TM ATPase activity resides in the interplay between the N terminus of TnI and TnT⅐C. This is shown by several experimental results. First, regulatory complexes reconstituted from rabbit TnT⅐C plus intact rabbit TnI activates the ATPase activity above acto-S1-TM rates (Ն198% versus 100%) while those formed with the N terminus of TnI deleted (TnI-(104 -115) or TnI-(96 -148)) had activities close to the acto-S1-TM activity (108 and 119%, respectively). Second, the regulatory complex reconstituted with chicken TnI-(1-116) plus TnT⅐C activated the ATPase activity to the same degree as rabbit skeletal TnI plus TnT⅐C. Third, the inhibitory region can modulate or influence the extent of activation since, a single amino acid substitution within the inhibitory region of TnI (L111G) had a significant increase in the level of activation compared with intact TnI (Fig. 2B). Fourth, TnT is essential for activation of the ATPase activity since TnI⅐C (8,12) and TnI-(1-116)⅐C (13) are unable to activate the ATPase activity. Thus, activation (and the degree of activation) of the ATPase activity in the presence of Ca 2ϩ is dependent on the presence of TnT, TnC, the N-terminal region of TnI, and amino acid residues within the TnI inhibitory region (residues 104 -115).
It has been shown that the N-terminal region of TnI contains a TnT-binding site since a skeletal TnI mutant with the first 57 amino acid residues deleted does not bind to TnT (11). The importance of this result is that a troponin regulatory complex composed of this TnI mutant, TnT, and TnC was able to potentiate the acto-S1-TM ATPase activity (12). Thus, the first 57 amino acid residues of TnI which contains the region comprising the regulatory peptide residues 1-40 which binds tightly to TnC (9), are not involved in activation of the ATPase activity. This suggests a region within TnI residues 57-96 interacts with TM, actin, or TnC and is responsible for activation of the ATPase activity. Ca 2ϩ sensitivity of the reconstituted thin filament is described by the concentration of calcium required to induce 50% of the maximum Ca 2ϩ -dependent change in the ATPase activity (pCa 50 ). The Ca 2ϩ sensitivity of the reconstituted thin filament is controlled by TnI-(96 -148) interaction with TnC. As shown in Fig. 4, the regulatory complex composed of TnI-(96 -148) plus TnT⅐C has the same Ca 2ϩ sensitivity and cooperativity as Tn or intact TnI plus TnT⅐C while the regulatory complexes reconstituted with TnI-(1-116) or TnI-(104 -115) plus TnT⅐C displayed an increased Ca 2ϩ sensitivity (Fig. 4B). This supports previous results which showed that binding of TnI-(104 -115) and TnI-(96 -115) to TnC, increases the Ca 2ϩ affinity of the complex compared with TnC alone (14,35,36). In addition, S1 is also present in our assay and could effect the Ca 2ϩ affinity of the thin filament (for reviews, see Refs. 5 and 37-39).
The Ca 2ϩ sensitivity may be increased by a reduction in the apparent affinity constant for TnI or TnI fragment with actin-TM or an increase in the apparent affinity constant between TnI or TnI fragment and TnT⅐C. However, the regulatory complexes in our assays are formed at concentrations of inhibitor and TnT⅐C that induce maximum inhibition and Ca 2ϩ -dependent change and therefore the composition of the thin filaments are similar for each regulatory complex (Table II). This implies that any effect from the different binding affinities of the various inhibitors for TnT⅐C or actin-TM should be minimized and the changes in Ca 2ϩ affinity between the various regulatory complexes should reflect a change in association constant of Ca 2ϩ binding to the thin filament.
In summary, distinct regions of TnI regulate the Ca 2ϩ -dependent response of the acto-S1-TM ATPase activity through its interactions with other thin filament proteins. For example, in the absence of Ca 2ϩ , TnI residues 96 -148 are sufficient for full inhibition of the acto-S1-TM ATPase activity. This data indicates that there must be an additional site of interaction in the region 96 -148 with actin-TM (Fig. 1A) other than the inhibitory site 104 -115. It is noteworthy that TnT does not contribute to the maximum level of inhibition of the acto-S1-TM ATPase activity since the TnI fragments, TnI-(96 -148) and TnI-(104 -115), are able by themselves to inhibit the ATPase activity to the same level as intact TnI or Tn in the absence of Ca 2ϩ (Fig. 1). On the other hand in the presence of Ca 2ϩ , TnI-(104 -115) or Tn-(96 -148) plus TnT⅐C are responsible for the partial release of the inhibition (to acto-S1-TM ATPase activity). It is the interaction between the N terminus of TnI and TnT⅐C that is responsible for the Ca 2ϩ -dependent activation of the ATPase activity. In addition, Ca 2ϩ sensitivity of the thin filament is controlled through the Ca 2ϩ -dependent switch of TnI-(96 -148) from multiple binding sites on actin to multiple binding sites on TnC(Ca 2ϩ ). Therefore, although the inhibitory region of TnI-(104 -115) is the Ca 2ϩ -sensitive switch (switching binding sites from actin-TM to TnC in the presence of Ca 2ϩ ) its function is modulated by the N and C terminus of TnI.