Familial hypertrophic cardiomyopathy mutations from different functional regions of troponin T result in different effects on the pH and Ca2+ sensitivity of cardiac muscle contraction.

To understand the molecular function of troponin T (TnT) in the Ca(2+) regulation of muscle contraction as well as the molecular pathogenesis of familial hypertrophic cardiomyopathy (FHC), eight FHC-linked TnT mutations, which are located in different functional regions of human cardiac TnT (HCTnT), were produced, and their structural and functional properties were examined. Circular dichroism spectroscopy demonstrated different secondary structures of these TnT mutants. Each of the recombinant HCTnTs was incorporated into porcine skinned fibers along with human cardiac troponin I (HCTnI) and troponin C (HCTnC), and the Ca(2+) dependent isometric force development of these troponin-replaced fibers was determined at pH 7.0 and 6.5. All eight mutants altered the contractile properties of skinned cardiac fibers. E244D potentiated the maximum force development without changing Ca(2+) sensitivity. In contrast, the other seven mutants increased the Ca(2+) sensitivity of force development but not the maximal force. R92L, R92W, and R94L also decreased the change in Ca(2+) sensitivity of force development observed on lowering the pH from 7 to 6.5, when compared with wild type TnT. The examination of additional mutants, H91Q and a double mutant H91Q/R92W, suggests that mutations in a region including residues 91-94 in HCTnT can perturb the proper response of cardiac contraction to changes in pH. These results suggest that different regions of TnT may contribute to the pathogenesis of TnT-linked FHC through different mechanisms.

To understand the molecular function of troponin T (TnT) in the Ca 2؉ regulation of muscle contraction as well as the molecular pathogenesis of familial hypertrophic cardiomyopathy (FHC), eight FHC-linked TnT mutations, which are located in different functional regions of human cardiac TnT (HCTnT), were produced, and their structural and functional properties were examined. Circular dichroism spectroscopy demonstrated different secondary structures of these TnT mutants. Each of the recombinant HCTnTs was incorporated into porcine skinned fibers along with human cardiac troponin I (HCTnI) and troponin C (HCTnC), and the Ca 2؉ dependent isometric force development of these troponin-replaced fibers was determined at pH 7.0 and 6.5. All eight mutants altered the contractile properties of skinned cardiac fibers. E244D potentiated the maximum force development without changing Ca 2؉ sensitivity. In contrast, the other seven mutants increased the Ca 2؉ sensitivity of force development but not the maximal force. R92L, R92W, and R94L also decreased the change in Ca 2؉ sensitivity of force development observed on lowering the pH from 7 to 6.5, when compared with wild type TnT. The examination of additional mutants, H91Q and a double mutant H91Q/R92W, suggests that mutations in a region including residues 91-94 in HCTnT can perturb the proper response of cardiac contraction to changes in pH. These results suggest that different regions of TnT may contribute to the pathogenesis of TnTlinked FHC through different mechanisms.
Vertebrate striated muscles contract in response to increasing intracellular Ca 2ϩ concentration where Ca 2ϩ -bound troponin (Tn) 1 together with tropomyosin (Tm) activates the thin filaments and leads to interactions between the thick and thin filaments (1,2). Tn, one of the regulatory proteins of striated muscle, is composed of three subunits, a Ca 2ϩ binding subunit troponin C (TnC), an inhibitory subunit troponin I (TnI), and a Tm binding subunit troponin T (TnT). Although the activation and inhibition of muscle contraction are primarily achieved by TnC and TnI, respectively, these events cannot occur without TnT. In other words TnT not only anchors the Tn complex to the thin filament but also contributes to the Ca 2ϩ -dependent regulation of muscle contraction (3). Therefore, any functional and structural defects in these Tn subunits may cause alteration of the Ca 2ϩ regulation of muscle contraction.
To date there are at least 26 known mutations in the human cardiac TnT (HCTnT) gene that are linked to FHC, including 23 missense mutations, one deletion mutation, and one splicing donor site mutation (14). Patients who have TnT mutations often show no or mild hypertrophy, myocardial disarray, and have a malignant phenotype associated with a high incidence of SCD compared with the other FHC genes (15). These phenotypes, however, that are caused by the different TnT mutations, can vary, and the phenotype-genotype correlation is not currently fully understood. Many studies have been attempted to exploit the molecular pathogenesis of FHC caused by each TnT mutation from molecular to physiological levels and have shown that most FHC-TnT mutations alter the contractile properties of cardiac muscle, especially the Ca 2ϩ sensitivity of force development and ATPase activity in vitro and in vivo (14,16,17). Our previous studies have demonstrated that the FHC-TnT mutants I79N, R92Q, F110I, E163K, and R278C increased Ca 2ϩ sensitivity of force development when they were incorporated into porcine skinned cardiac fibers together with human CTnI and CTnC (18). Moreover, our transgenic mouse model, expressing the I79N mutation, has demonstrated similar properties (i.e. increase in the Ca 2ϩ sensitivity of force development and ATPase activity) as well as other phenotypic properties of FHC, such as impaired diastolic function (19,20). Similar results have been seen at both the molecular and physiological levels, implying that altered Ca 2ϩ regulation of muscle contraction by TnT mutations might be the primary mechanism for TnT-linked FHC.
To further characterize the functional consequences of the FHC-TnT mutations on muscle contraction we have examined eight different HCTnT mutations, R92L, R92W, R94L, A104V, R130C, E163R, S179F, and E244D. They are located in functionally and structurally different regions of the HCTnT molecule. R92W, R94L, and A104V were shown to cause a severe form of FHC characterized by a high incidence of SCD (15,(21)(22)(23). S179F has a relatively mild or benign phenotype (24). However, a rare homozygous S179F mutation had a malignant prognosis. The R92L mutation has been linked to a severe phenotype, with a higher level of hypertrophy but less SCD, although the incidence of disease-related death is high (15,25). There are insufficient clinical data for R130C, E163R, and E244D; therefore, the phenotypic features caused by these mutations are unclear (26,27). To investigate effects of these TnT mutations, recombinant HCTnT mutants were incorporated into porcine cardiac skinned fibers together with HCTnI and HCTnC using the Tn displacement method followed by the measurement of Ca 2ϩ -dependent isometric force development at pH 7.0 and 6.5. Circular dichroism (CD) spectroscopy was utilized to determine the secondary structural variations among the isolated TnT mutations. Results from fiber studies have shown a correlation between the location of the FHC mutation in the TnT molecule and its effect on Ca 2ϩ -regulated force. Mutations near the N terminus of TnT seemed to impair its affinity for Tm, whereas mutations in the middle and near the C-terminal end of the molecule altered steady state crossbridge kinetics. Nevertheless, most mutants showed similar effects on Ca 2ϩ -dependent force development, resulting in increased Ca 2ϩ sensitivity. These results suggest that different TnT mutations alter the Ca 2ϩ regulation of muscle contraction through different mechanisms and that these differences may contribute to the molecular pathogenesis of various forms of FHC caused by the TnT mutations.

Cloning and Mutagenesis of Human Cardiac Troponin cDNAs-
Cloning of wild-type (WT) human cardiac TnT, TnI, and TnC was described elsewhere (18). To create the FHC-TnT mutations R92L, R92W, R94L, A104V, R130C, E163R, S179F, and E244D as well as the additional TnT mutations H91Q and H91Q/R92W, a sequential overlapping polymerase chain reaction-based method was utilized, and all clones were constructed in the pET-3d expression vector. Mutations in the TnT cDNA were confirmed by DNA sequencing.
Expression and Purification of Recombinant Human Cardiac Troponin-Recombinant human cardiac Tn subunits were expressed in Escherichia coli strain BL21 (DE3), and standard methods were used for the purification of the HCTnTs, HCTnI, and HCTnC (18). Briefly, recombinant HCTnTs were purified from a sonicated bacterial cell suspension using a Fast Flow S-Sepharose ion exchange column followed by Q-Sepharose column chromatography. Depending on the purity, the samples were further purified using a DE-52 column. TnTs were eluted with a 0-0.6 M NaCl gradient in a buffer containing 20 mM citric acid (pH 7.0), 6 M urea, 2 mM EDTA, 1 mM dithiothreitol for the S-Sepharose column, a 0 -0.5 M NaCl gradient in a buffer containing 20 mM Tris-HCl (pH 7.8), 6 M urea, 1 mM EDTA, 1 mM dithiothreitol for the Q-Sepharose column, and 0 -0.3 M KCl in a buffer containing Tris-HCl (pH 8.0), 6 M urea, 1 mM EDTA, 1 mM dithiothreitol gradient for the DE-52 column.
Circular Dichroism of HCTnT Mutants-The CD spectra of HCTnT WT and the mutants were collected using a Jasco J-720 spectropolarimeter with a cell path length of 0.1 cm at ambient temperature (20°C). Samples were dialyzed against 10 mM phosphate buffer (pH 7.0), including 0.5 or 0.3 M NaF. The concentrations of the HCTnT mutants were determined using the Bradford assay with wild type HCTnT (WT) as a standard. The final protein concentration was set to 0.25 mg/ml. Analysis and processing of data were carried out using the Jasco system software (Windows Standard Analysis, Version 1.20). Ten scans were averaged, and the spectrum of the buffer was subtracted for a base-line offset.
Mean Preparation of Porcine Skinned Cardiac Fibers and Measurement of Force Development-Skinned papillary muscles were prepared from the left ventricles of pig hearts. Small bundles of fibers were isolated and treated with the relaxing solution (10 Ϫ8 M Ca 2ϩ , 1 mM Mg 2ϩ , 7 mM EGTA, 5 mM MgATP, 20 mM imidazole (pH 7.0), 20 mM creatine phosphate, and 15 units/ml creatine phosphokinase, I ϭ 150 mM) containing 1% Triton X-100 for ϳ1 h and relaxing solution including 50% glycerol at Ϫ20°C. For measurement of force development, small fibers (ϳ120 m in diameter) were dissected and then mounted between a force transducer and a fixed needle. Fibers were skinned with 1% Triton X-100 in relaxing solution for 30 min at room temperature. After maximum force was measured at pCa 4 the fibers were exposed to increasing Ca 2ϩ concentration ranging from pCa 8 to pCa 4, and the force-pCa relationship was recorded at pH 7.0 and then at pH 6.5 on the same fiber. The data were analyzed with the Hill equation as described previously (18).
Tn Displacement and Reconstitution in Skinned Cardiac Fibers-The Tn displacement/replacement method, which we have modified from that of Hatakenaka and Ohtsuki (29), was utilized to replace the Tn present in the porcine fibers with human subunits (18,29). After the control Ca 2ϩ -dependent force development was measured the fibers were incubated with an excess amount of WT or HCTnT mutants (0.8 or 1.0 mg/ml) in a solution containing 250 mM KCl, 20 mM MES-KOH (pH 6.2), 5 mM MgCl 2 , 5 mM EGTA, and 0.5 mM dithiothreitol for 1 h. Fibers were then incubated in the same, fresh TnT solution for another hour. Displaced fibers were washed with the same solution without protein and tested for Ca 2ϩ -unregulated force at pCa 8 and pCa 4. This Ca 2ϩunregulated force development is due to the absence of the endogenous porcine TnI and TnC. To evaluate the extent of exchange by different TnT mutants the equation %Ca 2ϩ regulation ϭ (F pCa 4 Ϫ F pCa 8 )/F pCa 4 ϫ 100 was employed, where F pCa 8 and F pCa 4 are the force development of TnT-displaced fibers at pCa 8 and pCa 4, respectively (see Fig. 1). F pCa 4 of the TnT mutants-reconstituted fibers was also compared with that of HCTnT WT-incorporated fibers for the direct effect of TnT mutations on isometric force development. After TnT-displaced fibers were incubated with an HCTnI⅐C complex for 1.5 h, Ca 2ϩ -dependent force development of Tn-displaced fibers was determined at pH 7.0 and then at pH 6.5.
Data Analysis-Data were expressed as the mean value Ϯ S.D. The statistical significance of the differences between the mean value for WT versus HCTnT mutants was determined by one-way analysis of variance followed by a post-hoc test (Dunnett's multiple comparison test). Turky-Kramer's multiple-comparison test was utilized to evaluate differences among TnT mutations. For the effect of TnI⅐TnC reconstitution on the maximum force development, the paired t test was applied. Experimental protocol for isometric force development and TnT treatment followed by TnI⅐TnC reconstitution is shown. Treatment with HCTnT for 2h at room temperature resulted in a loss of the Ca 2ϩ -dependent regulation of force development. Subsequent TnI⅐C reconstitution recovered the Ca 2ϩ sensitivity of force development. (Fig. 2). The estimated ␣-helical content of WT TnT was ϳ38%. Both of the Arg to Leu mutations, R92L and R94L, resulted in increased ␣-helical content to ϳ60 and 59%, respectively (Fig.  2). R92W, A104V, R130C, and E163K also showed increased ␣-helical content, whereas the ␣-helical content of S179F and E244D were similar to that of WT. In the presence of 0.3 M NaF, all TnTs, especially A104V, showed less ␣-helical content compared with 0.5 M NaF (data not shown).

Displacement of TnT Mutants in Skinned Cardiac
Fibers-We have examined whether mutations in TnT affect the affinity of TnT for the thin filament. Alterations in the affinity of TnT for Tm or other thin filament proteins could affect the efficiency of Tn replacement in skinned fibers. We have previ-ously found that the FHC-TnT F110I-mutant showed poor incorporation into the porcine skinned cardiac muscle preparation (18). This result is consistent with another report in which fragments, including the N-terminal Tm binding region of TnT were used, and a F110I fragment showed a decreased affinity for Tm (30,31). To estimate the exchangeability of TnT mutants we have calculated the %Ca 2ϩ regulation of force development after fibers were treated with each TnT mutant (see "Materials and Methods"). The amount of endogenous Tn complex that remained in the fibers after TnT treatment correlated well with the %Ca 2ϩ regulation remaining after TnT incubation (18,32). These values, derived from the measurement of Ca 2ϩ -unregulated force development, can be used as an index for the extent of Tn displacement. We have previously demonstrated that endogenous porcine cardiac Tn can be replaced by recombinant HCTnT WT nearly completely under certain conditions, and HCTnT WT-incorporated fibers showed almost no Ca 2ϩ regulation unless they were reconstituted with HCTnI and TnC (18). Consistent with this previous observation, we observed very little Ca 2ϩ regulation of force development after HCTnT WT treatment (mean %Ca 2ϩ regulation ϭ 6%) in the present study (Table I). The %Ca 2ϩ regulation for A104V, E163R, and S179F-displaced fibers was comparable with the WT fibers, suggesting that the exchangeability of HCTnT with porcine Tn was not altered by these mutations. The value of %Ca 2ϩ regulation of E244D-treated fibers was also not significantly different from that of WT fibers, and we did not observe any Ca 2ϩ regulation of force development after the fibers were treated with this mutant. The R92L-displaced fibers showed slightly increased %Ca 2ϩ regulation but were not significantly different when they were compared with WT fibers. In contrast, the %Ca 2ϩ regulation of force development was significantly increased when fibers were reconstituted with R92W, R94L, or R130C. This result indicates that the amino acid substitutions Arg to Trp at position 92, Arg to Leu at position 94, and Arg to Cys at position 130 impaired the exchangeability of HCTnT. The latter mutants were able to exchange about 80% with the endogenous Tn, as judged by the remaining Ca 2ϩ regulation of force development.
Effect of Mutants on the Maximum Force Development before and after HCTnI⅐C Reconstitution-As we have observed previously the maximum force development (F pCa 4 ) was decreased by HCTnT WT treatment (by ϳ25%) compared with the original maximum force of porcine cardiac fibers. This was in part due to a difference between porcine and human cardiac TnTs since the extent of maximum force reduction dramatically decreased when porcine fibers were treated with native porcine cardiac TnT or when human cardiac fibers were treated with recombinant HCTnT WT (18,33). The maximum force development of HCTnT WT-displaced fibers, however, was not affected by the subsequent HCTnI⅐C reconstitution, and the maximum force of fully reconstituted fibers was 69.4 Ϯ 4.1% of that of untreated fibers. We have examined the effect of mutants on the maximum force development after TnT treatment as well as after HCTnI⅐C reconstitution, and the amount of force development is expressed as the percentage of the maximum force development relative to HCTnT⅐I⅐C WT-displaced fibers, which was set as 100% (Table I, Fig. 3). F pCa 4 values (i.e. the maximum force development of TnI⅐TnC-depleted fibers by TnT treatment) of R92L-, R94L-, A104V-, and R130C-incorporated fibers were very similar to that of WT-displaced fibers, whereas the F pCa 4 values of E163R, S179F, and E244D fibers were significantly higher, especially the E244D fibers. Because the %Ca 2ϩ regulation was nearly 0 after fibers were treated with E163R, S179F, or E244D, a potentiation of the maximum force development could be a direct effect of these TnT mutations. R92W seemed to increase F pCa 4 , although the increase was not significantly different from WT. Interestingly, the maximum force development of all the TnT mutant-incorporated fibers except the E244D fibers became similar to that of WT fibers after reconstitution with HCTnI⅐C. The E244D fibers reconstituted with TnI⅐C showed increased maximal force development compared with WT fibers, although the TnI⅐TnC reconstitution significantly reduced the maximum force development. All TnTs, including WT did not affect the base line of force development (i.e. force development at pCa 8) of TnI⅐C-reconstituted fibers.
Effect of TnT Mutations on the Force-pCa Relationship at Different pH Levels-The pCa-force development relationships were determined at both pH 7.0 and 6.5 in the fibers that had been reconstituted with HCTnT WT or the HCTnT mutants and HCTnI⅐TnC complex. Fig. 4 shows the pCa-force development curves of FHC mutants and WT-incorporated fibers at pH 7.0. The mean values of pCa 50 , Hill coefficient (n H ), and the maximum force for each curve are summarized in Table II. All mutants, with the exception of E244D, increased the Ca 2ϩ sensitivity of force development in the following order of pCa 50 values: S179F Ͼ R92W Ͼ E163R Ͼ R94L Ͼ R92L Ͼ R130C Ͼ A104V. The Hill coefficients of force-pCa curves were also in-creased by R92W, A104V, R130C, and E163R. In contrast, E244D did not change pCa 50 or the (n H ) but increased maximal force development (115.8 Ϯ 4.7%) as described above.
The Ca 2ϩ sensitivity and the maximum force development of WT-incorporated fibers were reduced when pH was lowered from 7.0 to 6.5 (Table II). Similar to WT fibers, all mutantreconstituted fibers showed decreased Ca 2ϩ sensitivity and maximal force development at pH 6.5 compared with pH 7.0. The Ca 2ϩ sensitivity of force development of seven of the eight mutant (i.e. R92L/W, R94L, A104V, R130C, E163R, and S179F) fibers were significantly higher than that of WT fibers. The Ca 2ϩ sensitivity of the E244D-reconstituted fibers was not different from that of WT fibers. However, the maximal force development was increased by incorporation of E244D. Interestingly, the effect of lowering pH on the Ca 2ϩ sensitivity of each TnT-reconstituted fibers was different. R92L, R92W, and R94L fibers exhibited less sensitivity to changes in Ca 2ϩ upon lowering of the pH and had smaller ⌬pCa 50 values (⌬pCa 50 ϭ pCa 50 at pH 7.0 Ϫ pCa 50 at pH 6.5), 0.42 Ϯ 0.06, 0.35 Ϯ 0.04, and 0.36 Ϯ 0.06, respectively, indicating that these mutants have greater acidic pH resistance than WT in terms of Ca 2ϩ sensitivity. In contrast, the ⌬pCa 50 of A104V and R130C fibers FIG. 3. Effect of mutants on the maximum isometric force development before and after TnI⅐TnC reconstitution. The maximum force development of skinned fibers measured after each TnT treatment (black bar) was compared with the maximum force of the same fibers but after being reconstituted with the TnI⅐TnC complex (white bar). Those values from TnT mutants-treated fibers were also compared with that of WT-treated fibers. *, p Ͻ 0.05; **, p Ͻ 0.01; ***, p Ͻ 0.001 (paired t test). (0.51 Ϯ 0.02 and 0.53 Ϯ 0.06 respectively) were nearly identical to WT fibers (0.54 Ϯ 0.03). The acidic pH resistance of E163R, S179F, and E244D fibers were slightly less than that of WT fibers.

DISCUSSION
A number of studies has been attempted to elucidate the molecular mechanisms of Ca 2ϩ regulation of muscle contraction. However, the molecular function of TnT is not understood as well as that of TnC or TnI. At least 26 different mutations in the HCTnT gene have been identified (14) since Thierfelder et al. (9) reported the first FHC-related TnT mutations in 1994. Because FHC that is caused by TnT mutations might be a consequence of an altered molecular function of TnT, that would lead to abnormal cardiac muscle contraction, the functional analyses of these disease-causing TnT mutations could help our understanding of the molecular function of TnT as well as the pathogeneses of FHC. Lin et al. (34) first reported the functional consequences of FHC-I79N TnT mutation in 1996, and then Morimoto et al. (32) found that the FHC-TnT mutants, I79N and R92Q, increased Ca 2ϩ sensitivity of isometric force development of skinned cardiac fibers when they were exchanged with the endogenous Tn complex. Since then many studies including those from our lab demonstrates that FHC-TnT mutants can alter Ca 2ϩ regulation of muscle contraction (14,16,17). In this study we have also observed that all eight FHC-TnT mutants examined (i.e. R92L, R92W, R94L, A104V, R130C, E163R, S179F, and E244D) altered the contractile properties of skinned cardiac fibers in various ways. E244D potentiated the maximum force development without changing Ca 2ϩ sensitivity, whereas the other seven mutants increased Ca 2ϩ sensitivity but not maximal force development. Our results may lead to an understanding of why different mutations in TnT can cause the same pathological disease, i.e. FHC. We suggest that the altered Ca 2ϩ regulation of force production could be a common physiological functional consequence of TnT-linked pathogenesis of FHC.
Although the detailed mechanisms by which single point mutations in TnT may cause Ca 2ϩ -sensitizing effects on force development are not clear at this time, the observed changes in the increased Ca 2ϩ sensitivity of force development might be explained by one of the following ways or a combination of them; 1) increased Ca 2ϩ affinity of TnC; 2) reduced inhibitory action of TnI, and 3) activation of the actomyosin interaction that indirectly leads to increased Ca 2ϩ affinity of TnC. For example, we have previously demonstrated that the deletion mutant, ⌬E160, which increased Ca 2ϩ sensitivity of both myofibrillar ATPase activity and force development of skinned fibers, did not affect the inhibitory action of TnI but enhanced neutralizing action of TnC (35,36). Moreover, Tobacman et al. (37) show that the Tn⅐Tm complex containing the ⌬E160 TnT increased its affinity for Ca 2ϩ . These results strongly suggest that ⌬E160 increases Ca 2ϩ sensitivity of myofibrillar ATPase activity and force development of skinned fibers by increasing Ca 2ϩ affinity of TnC. Although we have not investigated in detail the effects of TnT mutations on the interaction between TnT and the other Tn subunits or Tm, we suggest that the specific amino acid substitutions and the region of TnT where the mutation occurs lead to distinct physiological consequences such as the increase in Ca 2ϩ sensitivity and maximum force development. CD spectroscopy demonstrated variable secondary structures of TnT mutants in the presence of both low (0.3 M) and high (0.5 M) concentrations of NaF (Fig. 2). The WT and all TnT mutants had lower ␣-helical content in the presence of 0.3 M NaF than in 0.5 M NaF (data not shown). This was not due to simple protein precipitation since centrifugation of these 0.3 M NaF samples at room temperature showed no pellets. The extent of ␣-helical content reduction by decreasing NaF concentration was much greater for A104V mutant (ϳ40%) than for other TnT proteins (Ͻ23%). Interestingly, two independent reports from different laboratories show that a TnT fragment FIG. 4. pCa-force relationship of HCTnT WT on FHC mutants reconstituted fibers at pH 7.0. After TnT-treated fibers were reconstituted with the TnI⅐TnC complex, the Ca 2ϩ dependent of force development of these fibers was increased at pH 7.0. Every panel contains the pCa-force curve of WT-reconstituted fibers (open circles) as a reference, and each panel represents the pCa-force relationship for the following mutant-treated fibers; A, R92L (triangles), R92W (inverted triangles) and R94L (squares); B, A104V (triangles) and R130C (squares); C, E163R (triangles), S179F (inverted triangles), and E244D (dot in squares and dashed lines).
containing the Val-104 residue exhibited poor thermal stability of secondary structure compared with WT or the other mutants (30,31). All TnT mutants showed similar secondary structure and the same amount of ␣-helical content (ϳ70%, data not shown) when 40% 2,2,2-trifluoroethanol was added to stabilize the ␣-helix and to closely mimic in vivo protein-protein interactions. This result implies that even though these mutations may not drastically alter the secondary structure of TnT alone under physiological conditions, the structural differences seen with TnT mutants under certain conditions may imply that mutations cause modified dynamic interactions between TnT and Tm, TnI, or TnC. However, it should be noted that these experiments are qualitative experiments since TnT would be unable to fold into its native conformation in the absence of the other troponin subunits.
Besides affecting Ca 2ϩ sensitivity of force development, the near N-terminal mutants, R92L, R92W, R94L, and R130C, were shown to have decreased ability to displace Tn complex in the fibers. Palm et al. (30) show that mutations occurring in the region within 92-110 impair Tm binding or TnT enhancement of Tm binding to actin. Our results showing that R92L, R92W, and R94L had impaired exchangeability in fibers agree with their findings. Similar effects have been reported for the F110I mutant, which did not exchange well with endogenous porcine cardiac Tn complex (18). We showed that A104V did not reduce its exchange ability, whereas Palm et al. (30) demonstrate that it reduced the enhancement of Tm binding to actin but not its Tm-biding affinity. The R130C mutant also showed decreased exchangeability. The affinity of R130C for Tm or the thin filament, however, is unknown. Arg-130 is located in the essential TnT region (residues 112-136) for Tm binding (31); hence, it is expected that an Arg to Cys substitution at this position could reduce the binding affinity for Tm, resulting in the poor exchangeability seen with the mutant in this study.
In contrast, in the middle of the TnT molecule, mutants E163R and S179F and the C-terminal mutant E244D increased maximum force development of TnI⅐C-depleted fibers without reducing their TnT exchangeability. In the absence of TnI and TnC, E163R and S179F increased the maximum force development by 15ϳ20% compared with WT, whereas E244D increased maximum force development by Ͼ50%. Furthermore, the E244D-incorporated fibers showed higher maximum force development even after being reconstituted with the TnI⅐C complex than WT fibers, whereas the E163R and S179F fibers did not (or rather decreased the maximum force development slightly). These results together with the fact that these mutations occurred in the different structural and functional regions suggest that the mechanism by which E244D could potentiate the isometric force development might be different from the mechanism found for E163R or S179F. The region including residues Glu-163 and Ser-179 is near the end of the Tm binding rod region of TnT, and its detailed role in the regulation of force development is not known. It is possible that amino acid substitutions in this region alter thin filament activating states and/or the stability, although the observed potentiation of force occurred at sub-maximal Ca 2ϩ concentrations. Yanaga et al. (38) report that E244D increased the maximum myofibrillar MgATPase activity, whereas Nakaura et al. (39) have shown that partial reconstitution of E244D into rabbit skinned cardiac fibers also potentiated the maximum force development before and after TnI⅐C reconstitution. These results are consistent with our current observations. It has been suggested that the region including Glu-244 in TnT interacts with TnI (40). The crystal structure of human cardiac Tn partial complex, including the entire TnC molecule, the N-terminal truncated fragment of TnI (residues 31-163), and the C-terminal half of TnT (residues 183-288), recently revealed by Takeda et al. (41), has demonstrated that residue Glu-244 is located in the middle of helix composed of residues 225-271, which interacts with the helix of TnI, including residues 90 -136. Therefore, the mutation of Glu to Asp at position 244 would be expected to alter the interaction between TnT and TnI, leading to an abnormal contractile response to Ca 2ϩ . Nevertheless, we have observed the potentiated effect of this mutation on the isometric force development even in the absence or very little amount of TnI. These results indicate that the region including residue Glu-244 interacts with Tm and/or actin as well as TnI, and a substitution of Glu to Asp alters these interactions and possibly cross-bridge kinetics.
We have also studied the effect of these FHC-TnT mutations on the pH sensitivity of Ca 2ϩ -dependent force development. We have shown that all TnT-reconstituted fibers including WT decreased the Ca 2ϩ sensitivity of force development as well as its maximum value when the pH was lowered from 7.0 to 6.5. However, three mutants, R92L, R92W, and R94L showed significantly smaller differences in the Ca 2ϩ sensitivity of force measured at pH 7.0 and 6.5; their respective mean ⌬pCa 50 values were 0.42, 0.35, and 0.36 versus WT (mean ⌬pCa 50 ϭ 0.54) or versus the other mutants (mean ⌬pCa 50 ϭ 0.51ϳ0.64). This suggests that these mutants demonstrate increased resistance to pH change. It has been previously reported by Morimoto et al. (32) that the I79N-and R92Q-TnT mutants reduce the Ca 2ϩ desensitizing effect of pH. We have observed similar effects for the I79N-TnT mutation, which demonstrated an acidic pH resistance of the Ca 2ϩ sensitivity of force in our transgenic mouse model (19). Moreover, as shown by Solaro et al. (42) skinned cardiac fibers from the transgenic mouse model of R92Q showed enhanced effects of increased Ca 2ϩ sensitivity of force development at lower pH compared with that of nontransgenic mouse fibers. These results strongly indicated that a region including residues 79 -94 in the HCTnT molecule could be involved in the pH-sensitive Ca 2ϩ regulation of cardiac muscle contraction, and any mutations occurring in this region could lead to alterations in the acidic pH resistance. The His residue is often involved in pH-dependent functional perturbations because of its similar pK a value to that of imidazole at physiological pH. For instance it has been demonstrated that one histidine residue in fast skeletal TnI is responsible for the difference in acidic pH resistance of skeletal versus cardiac muscles (43). There are two histidine residues, His-91 and His-109, that could be involved in this mechanism. Both of them are adjacent to residues involved in TnT-linked FHC, Arg-92 and Phe-110. NMR studies have revealed that the pK a of the imidazole group, corresponding to the histidine residue in fast skeletal TnT (His-79) was around 7 in the presence of Tm (44), indicating that a fraction of the charged form of the imidazole group changes when the pH changes from pH 7.0 to 6.5 (ϳ50 and ϳ78%, respectively). This difference may be an important factor for the pH-dependent Ca 2ϩ regulation of force regulation. However, Nakaura et al. (39) report that F110I did not alter acidic pH resistance of Ca 2ϩ sensitivity of force (39). Furthermore, we did not observe altered pH-dependent Ca 2ϩ regulation by A104V, which is five amino acids away from His-109 and is closer to His-109 than Arg-92 or Arg-94. Therefore, we suspect that another residue, i.e. His-91, may be important as responsible for the increased acidic pH resistance by the TnT mutants, R92L, R92W, and R94L. To test this we created two new TnT mutants described below and examined their pH-dependent Ca 2ϩ regulation of isometric force development (Fig. 5A). One mutant has a glutamine residue at position 91 instead of histidine (H91Q). We have chosen glutamine to neutralize the charge effect at position 91 and because glutamine corresponds to residue His-91 in the fast skeletal isoforms of most species. Another mutant, H91Q/R92W, carries one more mutation besides H91Q. Because the H91Q mutant fibers showed increased Ca 2ϩ sensitivity of isometric force development at pH 7.0 (pCa 50 ϭ 5.65 Ϯ 0.01, n ϭ 3) and demonstrated acidic pH resistance (⌬pCa 50 ϭ 0.31 Ϯ 0.2) compared with WT it seems as if the His-91 residue is the one that is involved in the determination of pH-dependent Ca 2ϩ regulation. H91Q/ R92W exhibited a higher Ca 2ϩ sensitivity of force development (pCa 50 ϭ 5.84 Ϯ 0.04, n ϭ 3) and greater acidic pH resistance (⌬pCa 50 ϭ 0.25 Ϯ 0.4) than H91Q or R92W and indicated that the Ca 2ϩ -sensitizing effect and the pH-desensitizing effect caused by H91Q and R92W substitutions were additive. These results strongly suggest that the cluster of positively charged residues of 91-94 (His-Arg-Lys-Arg) rather than a difference in the fraction of the charged imidazole group of His-91 between pH 7.0 and 6.5 per se is responsible for the determination of pH-dependent Ca 2ϩ regulation of normal cardiac muscle contraction.
It is well known that the Ca 2ϩ -dependent cardiac muscle contraction is more sensitive to the acidosis effect than fast or slow skeletal muscles (45)(46)(47)(48). Such a large shift in the Ca 2ϩ sensitivity of force on the change in pH in cardiac muscle contraction is thought to protect the muscle when the heart undergoes ischemic processes. Morimoto et al. (32) suggest that the increased Ca 2ϩ sensitivity of contraction at low pH would cause a rapid decrease in intracellular ATP, resulting in the HCTnT WT and R92W was replaced by glutamine and the pCa-force relationship of these two mutants, H91Q (circles) and H91Q/R92W (triangles), were increased at pH 7.0 (opened symbol) and pH 6.5 (closed symbols) (A). pCa 50 at pH 7.0 (B) and ⌬pCa 50 values (C) were compared with those values from WT-and R92W-reconstituted fibers by Turky-Kramer's multiple-comparison test. *, p Ͻ 0.05; **, p Ͻ 0.01; ***, p Ͻ 0.001. n.s., not significant. promotion of up-regulation of cytokine-linked pathways by which several events involved in cell fate, such as apoptosis, could occur (32). The results seen with R92L, R92W, and R94L that the acidic pH resistance of Ca 2ϩ -dependent force development was increased seem to support this hypothesis and explain why these mutations could be linked with a poor prognosis of FHC (15,21,22,25). The phenotype of the FHC linked to the A104V mutation, which showed the least Ca 2ϩ -sensitizing effect (except E244D) at both pH 7.0 and 6.5 has been shown to be associated with a malignant form of FHC, including a high incidence of SCD (23). Based on our observation, it is difficult to explain why this mutation causes such a poor prognosis. Perhaps alterations in other contractile properties that were not determined in the present study such as ATPase activity, velocity of contraction, or the response to ␤-adrenergic stimulation could be responsible for this.
In conclusion, we have observed multiple functional consequences from different FHC-TnT mutations that occur in different functional regions of the HCTnT molecule. These results strongly suggest that various mutations alter contractile properties in different ways and that these differences may contribute to the molecular pathogenesis of TnT-linked FHC. Because of the limited amount of clinical data available, complicated backgrounds such as environmental factors or haplotype (49) or the possibility of additional mutations, including homozygous mutations (24,50,51), it is difficult to characterize phenotypes precisely. However, results from in vitro studies together with animal models might be helpful for elucidating the molecular mechanisms of the effect of these mutations on cardiac function.