Author + information
- Received May 19, 2004
- Revision received July 26, 2004
- Accepted August 25, 2004
- Published online December 7, 2004.
- Takeharu Hayashi, MD, PhD*,
- Takuro Arimura, DVM, PhD*,
- Manatsu Itoh-Satoh, MD, PhD*,
- Kazuo Ueda, MD*,
- Shigeru Hohda, MD, PhD*,
- Natsuko Inagaki, MD*,
- Megumi Takahashi, MS*,
- Hisae Hori, PhD*,
- Michio Yasunami, MD, PhD*,
- Hirofumi Nishi, MD, PhD†,
- Yoshinori Koga, MD, PhD†,
- Hiroshi Nakamura, MD, PhD‡,
- Masunori Matsuzaki, MD, PhD‡,
- Bo Yoon Choi, MS§,
- Sung Won Bae, PhD§,
- Cheol Woon You, MD§,
- Kyung Hoon Han, MD∥,
- Jeong Euy Park, MD§,
- Ralph Knöll, MD, PhD¶,
- Masahiko Hoshijima, MD, PhD¶,
- Kenneth R. Chien, MD, PhD¶ and
- Akinori Kimura, MD, PhD*,* ()
- ↵*Reprint requests and correspondence:
Dr. Akinori Kimura, Department of Molecular Pathogenesis, Medical Research Institute, Tokyo Medical and Dental University, 2-3-10 Kandasurugadai, Chiyoda-ku, Tokyo 101-0062, Japan
Objectives We sought to explore the relationship between a Tcap gene (TCAP)abnormality and cardiomyopathy.
Background Hypertrophic cardiomyopathy (HCM) and dilated cardiomyopathy (DCM) cause severe heart failure and sudden death. Recent genetic investigations have revealed that mutations of genes encoding Z-disc components, including titin and muscle LIM protein (MLP), are the primary cause of both HCM and DCM. The Z-disc plays a role in establishing the mechanical coupling of sarcomeric contraction and stretching, with the titin/Tcap/MLP complex serving as a mechanical stretch sensor. Tcap interacts with the calsarcin, which tethers the calcineurin to the Z-disc.
Methods The TCAPwas analyzed in 346 patients with HCM (236 familial and 110 sporadic cases) and 136 patients with DCM (34 familial and 102 sporadic cases). Two different in vitro qualitative assays—yeast two-hybrid and glutathion S-transferase pull-down competition—were performed in order to investigate functional changes in Tcap's interaction with MLP, titin, and calsarcin-1 caused by the identified mutations and a reported DCM-associated mutation, R87Q.
Results Two TCAPmutations, T137I and R153H, were found in patients with HCM, and another TCAPmutation, E132Q, was identified in a patient with DCM. It was demonstrated by the qualitative assays that the HCM-associated mutations augment the ability of Tcap to interact with titin and calsarcin-1, whereas the DCM-associated mutations impair the interaction of Tcap with MLP, titin, and calsarcin-1.
Conclusions These observations suggest that the difference in clinical phenotype (HCM or DCM) may be correlated with the property of altered binding among the Z-disc components.
Cardiomyopathy is one of the major causes of sudden death and/or progressive heart failure, and cases of cardiomyopathy are mainly classified into hypertrophic cardiomyopathy (HCM) and dilated cardiomyopathy (DCM). Hypertrophic cardiomyopathy is characterized by left ventricular (LV) hypertrophy accompanied by myofibrillar disarray and manifests as diastolic dysfunction of the cardiac ventricles, whereas DCM is characterized by a dilated ventricular cavity with systolic dysfunction (1). Recent genetic investigations have revealed that HCM is caused by mutations in the genes for sarcomeric proteins, such as MYH7(cardiac beta-myosin heavy chain), TNNT2(cardiac troponin T), and ACTC(cardiac alpha-actin) (2). Mutations in the sarcomere genes also cause DCM (3,4). In addition, mutations in the genes for Z-disc or costameric proteins, such as DMD(dystrophin), DES(desmin), and SGCD(δ-sarcoglycan), cause DCM (5–7).
The Z-disc plays various roles in sarcomeric organization, force transduction, and force transmission (8). Because the Z-disc is important in establishing the mechanical coupling of the sarcomere, defects in the Z-disc proteins may lead to either HCM or DCM. Titin, also known as connectin, is a component of the sarcomere and contains distinct motifs that sequentially unfold as the muscle is stretched (9). Titin forms a complex with alpha-actinin and alpha-actin at the Z-disc (Fig. 1)(10,11). We previously reported on two TTN(titin) mutations at the actinin-interaction domain, which increases and decreases the ability of titin to interact with alpha-actinin: an HCM-associated mutation that increases the ability of titin to interact with alpha-actinin and a DCM-associated mutation that decreases it (12,13). Titin also interacts with another Z-disc component, titin-cap (Tcap/telethonin) (14,15), and we found another DCM-associated TTNmutation at the Tcap-interaction domain, which impaired the interaction with Tcap (13). In addition, other DCM-associated TTNmutations not located in the Z-disc region have been reported (13,16), although the functional changes caused by these mutations remain unknown.
Tcap is a 19-kd protein required for sarcomerogenesis in striated muscles (14). It is phosphorylated by titin kinase at Ser157 and is suggested to play a key role in sarcomere assembly (17). Tcap interacts with calsarcin and minK at the N-terminal half and the C-terminal regions, respectively (18,19). Tcap also interacts with muscle LIM protein (MLP), and a DCM-associated MLPmutation abolishes the interaction (20). In addition, another DCM-associated MLPmutation was reported to impair the interaction of Tcap with alpha-actinin-2, whereas a DCM-associated alpha-actinin-2 mutation impaired its interaction with MLP (21). On the other hand, four MLPmutations were reported in HCM, one of which was shown to decrease the binding to alpha-actinin (22).
To further explore the relationship between an abnormality in Z-disc proteins and cardiomyopathy, we searched for mutations in the Tcap gene (TCAP)in a cohort of patients with HCM and DCM. Two HCM-associated TCAPmutations and a novel DCM-associated TCAPmutation were identified in this study. We investigated the interaction of Z-disc components titin, MLP, and calsarcin-1 (CS-1) with Tcap in the presence of the TCAPmutations. We report here that the HCM-associated and DCM-associated mutations showed different changes in Tcap's interaction properties with these Z-disc components.
A total of 346 and 136 genetically unrelated patients with HCM and DCM, respectively, were chosen as subjects. Among them, 332 HCM and 135 DCM patients were Japanese, and 14 HCM patients and 1 DCM patient were Korean. A family history of the disease (i.e., at least one blood relative with the same disease and/or unexplained sudden cardiac death <45 years of age) was observed in 236 patients with HCM and 34 with DCM. The patients with HCM were diagnosed based on medical history, physical examination, electrocardiogram, and echocardiogram (LV wall thickness >14 mm), whereas patients with DCM were diagnosed based on a medical history, physical examination, electrocardiogram, and echocardiogram (impaired LV function and fractional shortening <25%), in the absence of coronary artery disease, as evaluated by coronary angiography (12,13,23). When a patient was found to have a TCAPmutation, blood relatives were examined for the mutation, if available. Control subjects consisting of 240 Japanese and 70 Korean healthy individuals were tested for the mutations identified in this study. Blood samples were obtained from each subject after obtaining informed consent. The study protocol was approved by the Ethics Reviewing Committee of Medical Research Institute, Tokyo Medical and Dental University, and Institutional Reviewing Board of the Samsung Medical Center.
Analysis of known disease genes for HCM and DCM
Patients with HCM were screened for mutations in MYH7, TNNT2, MYBPC3(cardiac myosin-binding protein C), TNNI3(cardiac troponin I), MYL2(myosin regulatory light chain), MYL3(myosin essential light chain), TPM1(alpha-tropomyosin), TTN, MLP,and ACTC,whereas patients with DCM were analyzed for mutations in ACTC, DES, DMD, LMNA(lamin A/C), TTN, MYH7, TNNT2, MYBPC3,and TNNI3,by using the polymerase chain reaction (PCR) single-strand conformation polymorphism (SSCP) method (24–26), and the mutations were confirmed by sequencing the PCR products. The SSCP method we employed usually detects a point mutation with an efficacy of about 90% (25,26). The conditions of the PCR and primers used in the analyses are available on request. The screening analysis revealed 36 MYH7, 24 MYBPC3, 20 TTNT2, 6 TNNI3, 2 MYL2, 1 TPM1, 1 TTN, and 2 MYL3mutations in 91 familial HCM cases, and 5 MYBPC3, 3 TNNT2, 3 TNNI3, 1 MYL3, 1 TTN, and 2 MYH7mutations in 15 sporadic HCM cases. In addition, we found 1 DESand 3 TTNmutations in four familial DCM cases and 1 TTNand 3 DMDmutations in four sporadic DCM cases (details will be reported elsewhere).
Mutational analysis of TCAP
The primers used in this study are listed in Table 1.Both TCAP-1F and -1R for exon 1 and TCAP-2F and -2R for exon 2 were used in the analysis of TCAP. The PCR products from the patients were sequenced from the primers on both strands. Controls were analyzed only for exon 2.
Analysis of the cardiac myosin-binding protein C gene (MYBPC3)in a Korean family
Exon 6 of MYBPC3was amplified by PCR with MYBPC3-Ex6F and -Ex6R. The PCR products showed an abnormal SSCP pattern in patients with HCM, and the sequencing analysis demonstrated that this was due to an A-to-G change in codon 236. This change destroyed an AluI site. In order to detect the mutation in the blood relatives, the PCR products were digested by AluI and electrophoresed in 2.0% agarose gel.
Yeast two-hybrid (Y2H) assay
A human TCAPcomplementary deoxyribonucleic acid (cDNA) fragment (nucleotides 11–514; GenBank AJ000491, corresponding to amino acids 1–167) was amplified from cDNA generated from total cardiac RNA by the reverse transcription (RT)-PCR method using primers TCAP-EcoF and -BamR. Cardiomyopathy-associated mutations were introduced into the cDNA fragment by the primer-mediated mutagenesis method, as described previously (13). The following primers were used: R87Q: TCAP-EcoF, -87MR, -87MF, and -BamR; E132Q: TCAP-EcoF, -132MF, -132MR, and -BamR; T137I: TCAP-EcoF, -137MF, -137MR, and -Bam; R153H: TCAP-EcoF, -153MR, -153MF, and -Bam. Both S157A and S157D were constructed from cDNA clones obtained from Dr. S. Labeit (15). The PCR fragments were cloned into pCRII (Invitrogen, Carlsbad, California) for sequence confirmation. In order to obtain bait plasmids, normal and mutant TCAPcDNA fragments were cleaved by digestion with EcoRI and BamHI, gel-purified, and inserted into pGBKT7 (Clontech, Palo Alto, California). A human MLPcDNA fragment (nucleotides 58–642; GenBank U49837, corresponding amino acids 1–194) and a CS-1 cDNA fragment (nucleotides 86–424; GenBank BC005195, corresponding amino acids 1–113) were also obtained by the RT-PCR method with primers MLP-EcoF and -BamR, and CS-1-EcoF and -BamR, respectively. A human titin Z1-Z2 cDNA fragment (nucleotides 123–731; GenBank NM-003319, corresponding amino acids 1–203) was obtained by the RT-PCR method with primers TTN-NcoIF and -BamR. The MLP, the Z1-Z2 region of titin, and CS-1 cDNA fragments were cloned into pGADT7 (Clontech), and the nucleotide sequences were confirmed. The TCAPconstructs and either MLP, titin, or CS-1 constructs were used for co-transformation of Saccharomyces cerevisae(strain Y187) on agar plates lacking leucine and tryptophan. Several independent transformants (n = 8 to 10) were tested for beta-galactosidase activity, as described previously (12,13). Transformants carrying either bait plasmids or activator plasmids did not show any significant beta-galactosidase activity, demonstrating no self-activation. The data are presented as the mean value ± SD. The Student ttest was used to estimate statistical significance.
Glutathion S-transferase (GST) pull-down competition assay.
Each cDNA fragment used in the GST pull-down competition assay was generated by the PCR from each Y2H construct as a template, using combinations of the following primers: full Tcap fragment: Tcap-EcoF and -XhoT7R; short Tcap fragment: Tcap-EcoF and -120XhoR; titin Z1-Z2 fragment: TTN-NcoIF and -BamR; MLP: MLP-EcoF and -Xho1R; CS-1: CS-1-EcoF and -BamR. The TCAPfragments were cloned into pBAD/Thio-TOPO (Invitrogen) encoding the C-terminal polyhistidine tag to obtain a thioredoxin (Thio), T7, and His-tag fusion constructs. The titin, MLP, and CS-1 fragments were cloned into pGEX5X-1 (Amersham, Piscataway, New Jersey) to obtain GST-fusion constructs. All constructs were sequenced to ensure that no PCR errors were introduced. The GST-fusion constructs were used for the transformation of BL21-Gold (Stratagene, La Jolla, California). The transformants were collected in PBS with 1% Triton X-100 and sonicated. The GST-fusion proteins were purified and fixed on glutathione-sepharose beads as per the manufacturer's instructions. Thio-fusion TCAPconstructs were introduced into TOP10 (Invitrogen). Thio-fusion proteins were prepared under the denaturing condition (6-M urea) and purified by binding with Ni2+-charged resi, as per the manufacturer's instructions. A GST-only construct was used as a negative control. Binding assays were performed as described in the literature (14). Briefly, 5 μl each of Thio-fusion full Tcap protein and short Tcap protein (each 0.5 μg/μl) were added to 5 μl of either GST-fusion titin, MLP, or CS-1 (each 0.5 μg/μl) in 100 μl of binding buffer (20 mmol/l Tris, 50 mmol/l NaCl, pH 8.0). The ratio of Thio-fusion proteins to each GST-fusion protein used in the assay was determined in preliminary experiments, in which fixed amounts of Thio-fusion proteins were mixed with various amounts of GST-fusion proteins. The mixture was incubated for 1 h at 4°C, washed three times with 400 μl of binding buffer, suspended in sodium dodecyl sulfate (SDS) sample buffer, separated by SDS-polyacrylamide gel electrophoresis (PAGE) in 10% polyacrylamide gels, and transferred to duplicated nitrocellulose membranes. The membranes were blocked with 3% nonfat dry milk, incubated overnight at 4°C with anti-T7 tag monoclonal antibody (Novagen, San Diego, California) (1:7500) or goat anti-GST antibody (Amersham) (1:1000) as the first antibody. After washing with Tris-buffered saline (50 mmol/l Tris, 150 mmol/l NaCl, pH 7.5) containing 0.05% Tween 20, the membranes were incubated with anti-mouse immunoglobulin G conjugated with alkaline phosphatase (AP) (Promega, Madison, Wisconsin) (1:7500) or anti-goat immunoglobulin G conjugated with AP (Chemicon, Temecula, California) (1:7500), respectively, as the second antibody for 1 h at room temperature, and the immunocomplex was stained by using an Alkaline Phosphatase Conjugate Substrate Kit (Bio-Rad, Hercules, California). To confirm the specificity of the assay, we used affinity-purified rabbit polyclonal anti-Tcap antibody (kindly provided by Dr. S. Labeit) as the first antibody, substituting for anti-T7 antibody, and obtained similar results. The experiments were repeated at least five times, and each blot was analyzed by National Institutes of Health imaging in order to measure the contents of wild-type (WT) or mutant full Tcap protein in relation to that of WT short Tcap. The data are represented as the mean value ± SD. The Student ttest was used to estimate statistical significance.
Identification of TCAPmutations in HCM
TCAPwas analyzed by direct sequencing in a cohort of patients with HCM. There were three variations not accompanied by amino acid replacement: a CAC-to-CAT change in codon 52, a GGC-to-GGA change in codon 75, and a GCC-to-GCA change in codon 151. They were found in both patients and healthy controls at a similar frequency (variant allele frequencies in the patients and controls were 0.004 and 0.003, 0.003 and 0.003, and 0.432 and 0.440, respectively), indicating that these variations are polymorphisms not associated with the disease. In contrast, two missense mutations—a C-to-T transition in codon 137 replacing ACA (Thr) with ATA (Ile) (T137I) and a G-to-A transition in codon 153 replacing CGT (Arg) with CAT (His) (R153H)—were found in Japanese patients. The clinical characteristics of the patients and their blood relatives carrying these mutations are shown in Table 2.
A T137I mutation was found in a female patient (JH945 II-2) (Figs. 2a and 2b)and her son, both affected with HCM (JH945 III-1) (Table 2). On the other hand, an R153H mutation was found in a male patient (JH14 II-1) (Figs. 2c and 2d) and his sister (JH14 II-3) (Fig. 2c), both with HCM; this mutation was inherited by his sons, who did not suffer from HCM at the time of the examination (JH14 III-1 and III-2, respectively) (Table 2). These missense mutations were not found in unrelated controls. In addition, the patient cohort was screened for mutations in the known disease genes for HCM, and the patients carrying the TCAPmutations did not show any mutations in the known disease genes.
Identification of a novel TCAPmutation in DCM
While analyzing a cohort of DCM patients, we found a novel TCAPmutation in a Korean patient with DCM (SM728 II-5) (Fig. 3a).The patient had suffered from heart failure at the age of 34 years (Table 2) and undergone heart transplantation at the age of 35 years. The analysis demonstrated a missense mutation in codon 132, a replacement of GAG (Glu) with CAG (Gln) (E132Q) (Fig. 3b). This mutation was not found in the controls. A family study revealed that his mother (SM728 I-3) and maternal uncle (SM728 I-4) were affected by HCM, but no TCAP mutation was found in either of them (Fig. 3a). Instead, the TCAPmutation was transmitted from the father (SM728 I-2) to his son (SM728 III-4), who did not manifest a DCM phenotype at the time of the examination (Fig. 3a) (Table 2).
We searched for mutations in the known HCM-causing genes in his mother, and a missense mutation of MYBPC3, S236G, which is a replacement in codon 236 of AGC (Ser) with GGC (Gly), was identified (Fig. 3c). This mutation was also found in the affected uncle (Fig. 3a) and in another unrelated Japanese patient (data not shown), but not in the controls, suggesting that it may be a disease-causing mutation. As shown in Figure 3a, the MYBPC3mutation was not inherited by the proband, but was found in his brother (SM728 II-3) and his brother's offspring (SM728 III-1 and III-2), none of whom had HCM at the time of the examination (Table 2). Because both mother and uncle developed HCM after the age of 50 years, it is possible that the MYBPC3mutation is associated with late-onset HCM (2). These data demonstrate that this family has two different gene mutations: one is the TCAPmutation associated with DCM and the other is the MYBPC3mutation associated with HCM. In addition, the DCM cohort was screened for mutations in the known disease genes for DCM, and the patient carrying the TCAPmutation did not have any mutations in the known disease genes thus far tested.
As shown in Figure 4,the TCAPmutations identified in this study were found in the evolutionary conserved residues (except for E132Q, but this position is occupied by an acidic residue in the rat and mouse) at or near the binding domains to titin (14), MLP (20), and calsarcin-1 (19).
Alterations in the interaction of Tcap with MLP, titin and CS-1 caused by TCAPmutations, as found in the Y2H assay
In order to investigate the functional consequences of TCAPmutations for Tcap's interaction with MLP, titin, and CS-1, we performed Y2H assays. Wild-type and mutant Tcap's were constructed in an activator plasmid and co-transformed with a bait plasmid containing cDNA for MLP, titin (Z1-Z2 domains), or CS-1. Because the R153H mutation was found near the phosphorylation site, Ser157, and because this site is important in Tcap's interaction with minK (18), we investigated two other constructs, S157A (a model of the unphosphorylated form) and S157D (a model of the phosphorylated form).
We measured beta-galactosidase activity in several independent transformants in order to investigate qualitatively the strength of protein-protein interaction (Table 3).It was demonstrated that the DCM-associated mutations, R87Q and E132Q, significantly impaired the interaction (both were p < 0.01). On the other hand, the HCM-associated mutations, T137I and R153H, showed no statistical difference (p = 0.29 and 0.54, respectively). The unphosphorylated Tcap model (S157A) and the phosphorylated model (S157D) showed virtually no change (p = 0.49 and 0.70, respectively).
We next investigated Tcap's interaction with titin. The DCM-associated Tcap mutations showed statistically significant impairment (both were p < 0.001). In clear contrast, the HCM-associated mutations significantly augmented the interaction (both were p < 0.001). On the other hand, the unphosphorylated model and phosphorylated model showed no change in the interaction (p = 0.48 and p = 0.15, respectively).
Because calsarcin-3, a skeletal muscle-specific calsarcin, was reported to interact with Tcap, and cardiac muscle expresses another calsarcin, CS-1 (19), we examined the interaction between Tcap and CS-1. Co-transformants of Tcap and CS-1 constructs showed that the beta-galactosidase activity of WT Tcap colonies was 1.04 ± 0.03 units, whereas no beta-galactosidase activity was observed in colonies carrying either Tcap or CS-1. These data suggest that CS-1 could interact with Tcap. The DCM-associated mutations significantly impaired the interaction (both were p < 0.001). The HCM-associated mutations, in clear contrast, augmented the interaction (both were p < 0.001). Unphosphorylated or phosphorylated Tcap models showed no change (p = 0.58 and p = 0.19, respectively).
These observations demonstrate that the DCM-associated mutations impair the interaction of Tcap not only with MLP but also with titin and CS-1. On the other hand, the HCM-associated mutations augment Tcap's interaction with titin and CS-1 without significant changes in the interaction with MLP. In addition, the phosphorylation status of Tcap does not affect its interaction with these Z-disc components.
GST pull-down competition assay of the interaction between Tcap and the other Z-disc components, MLP, titin, and CS-1
In order to investigate the altered interaction between Tcap and the other Z-disc components by an independent method, we performed GST pull-down competition assays. The MLP, titin, and CS-1 proteins were expressed as GST-fusion proteins. The WT and mutant Tcap proteins were expressed as Thio-fusion proteins. These recombinant Tcap proteins were designated as test-Tcap-WT or test-Tcap-mutant (e.g., test-Tcap-R87Q). A short form of Tcap (amino acids 1–120) protein was also constructed as a Thio-fusion protein, to be used as a competitor (designated as comp-Tcap). Equal amounts of test-Tcap and comp-Tcap (Fig. 5a)were complexed with each GST-fusion protein. After incubation and washing, the Tcap proteins were eluted from the complex, separated by SDS-PAGE, and immunoblotted with anti-GST antibody and anti-T7 antibody so that the amounts of GST-fusion proteins complexed with the test-Tcap in the presence of comp-Tcap could be measured. Test-Tcap-WT did not form a complex with GST-only proteins (data not shown).
As shown in Figure 5b, the interaction between GST-MLP and test-Tcap-R87Q was weak, compared with that between GST-MLP and test-Tcap-WT. Similarly, the interaction between GST-MLP and test-Tcap-E132Q was weak. In contrast, the interactions between the GST-MLP and HCM-associated mutant Tcap, test-Tcap-T137I or -R153H, were comparable with that with test-Tcap-WT. Neither S157A nor S157D showed remarkable changes. The results obtained with GST-titin (Fig. 5c) and GST-CS1 (Fig. 5d) demonstrated that the DCM-associated mutations impaired the interaction with titin and CS-1, respectively. On the other hand, the HCM-associated mutations showed an augmented interaction with titin and CS1 (Figs. 5c and 5d, respectively).
The assays were repeated and each blot was analyzed by National Institutes of Health imaging to measure semi-quantitatively the intensity of each test-Tcap in comparison with that of comp-Tcap. The altered interaction between Tcap and the Z-disc components, as assessed by the relative intensity of the bands, is shown in Table 4. Regarding the interaction with MLP, the DCM-associated mutations showed significantly lower values (both were p < 0.001), whereas the HCM-associated mutations showed no change (p = 0.78 and 0.30, respectively). As for the interaction with titin, the DCM-associated mutations showed lower values (both were p < 0.001). On the other hand, the HCM-associated mutations clearly showed higher values (both were p < 0.001). The interaction with CS-1 was at a comparable level. The DCM-associated mutations gave lower values (p < 0.001), whereas the HCM-associated mutations gave higher values (p < 0.001). Both S157A and S157D showed virtually no change in the interaction with MLP, titin, and CS-1.
In this study, we identified for the first time, TCAPmutations associated with HCM. The HCM-associated mutations were found in the evolutionary conserved residues and were absent in healthy controls. In addition, we identified another DCM-associated mutation. The penetrance of the TCAPmutations may not be high, as there were several unaffected blood relatives carrying the mutations. In this cohort study, 2 of 236 familial HCM patients and 1 of 34 familial DCM patients were found to carry the TCAPmutations, suggesting that TCAPmutations may be a rare cause of cardiomyopathy.
The Z-disc of MLPknock-out mice showed widened, dispersed, and irregular structures, and the cardiomyocytes of MLPknock-out mice did not exhibit a hypertrophic response to mechanical stress. On the other hand, as a completely preserved Z-disc structure was observed in MLPand PLB(phospholamban gene) double knock-out mice, it is presumed that the promotion of calcium cycling can prevent Z-disc defects by enhancing cardiac relaxation and reducing wall stress (20). These observations suggest an important role of the Z-disc complex, titin/Tcap/MLP, in response to cardiac muscle stretching. Through the Y2H and GST pull-down competition assays, we demonstrated the functional alterations caused by the TCAPmutations. Due to the impaired interaction of Tcap with MLP, titin, and CS-1 caused by DCM-associated mutations, mutation-prone cardiomyocytes are more easily extended than the normal cardiomyocytes in the presence of mechanical stretch. The Z-disc with mutant Tcap could affect the stretch response and the calcium signals via CS-1 and the calcineurin complex.
In clear contrast, due to the augmented interaction of Tcap with titin and CS-1 caused by HCM-associated mutations, the Z-disc with mutant Tcap might exert an increased passive tension. If the length of the muscle fiber is constant, the increased passive tension leads to increased calcium sensitivity in muscle contraction (27). Therefore, it is presumed that TCAPmutations in HCM increase calcium sensitivity. The HCM-associated mutations in TPM1, MYL2, TNNT2, MYH7, and TNNI3are reported to increase calcium sensitivity (28–32), and the increased calcium sensitivity may be a common fundamental change caused by HCM-associated sarcomere mutations. It may be worth noting that Mohapatra et al. (21), by using the immunoprecipitation method, found a DCM-associated MLPmutation leading to decreased interaction with alpha-actinin, whereas Geier et al. (22), by means of an Y2H assay, revealed that an HCM-associated MLPmutation decreases the binding to alpha-actinin. It is not clear why DCM-associated and HCM-associated mutations exhibited similar functional changes, but it is difficult to compare the results because different assay systems were used in these studies; in particular, the Y2H assay might not be adequate for the protein-protein interaction in mammalian cells. In our study, we used the pull-down assay in addition to the Y2H assay in order to demonstrate the functional changes.
The Z-disc has a functional and mechanical coupling to the T-tubule system, a site of initial Ca2+-induced Ca2+release that links excitation and contraction. Calsarcin-1 binds calcineurin, a well-known mediator of hypertrophy. Overexpression of calsarcin can inhibit calcineurin activity (19), and CS-1 interacts with Cypher/ZASP, which binds protein kinase C (PKC) (Fig. 1) (19,33). Of interest is that we recently identified in a DCM family a Cypher/ZASP mutation that increases the binding to protein kinase C (33), and other Cypher/ZASP mutations were reported in several DCM families (34). These observations suggest that Z-disc proteins may play a role in monitoring local Ca2+signals at the Z-disc and regulating the activation of compensatory pathways (35). Our results suggest that the titin/Tcap/MLP/alpha-actinin complex plays an important role as a stretch sensor controlling cardiac hypertrophy, as well as cardiac failure. The modulation of the mechanical links between the stretch sensor and the Ca2+signals could become a therapeutic target in the future.
Scope of this study
We investigated the interaction of Tcap with titin, MLP, and CS-1, using two different methods. Because both methods are in vitro qualitative assays and recombinant proteins were used, the alterations in the interaction caused by the TCAPmutations might be different in vivo. A cohort of HCM and DCM patients was analyzed for mutations in the known disease genes by the SSCP method, which detects a point mutation with an efficacy of 90% or less (26). In addition, some disease genes, such as MYH6(cardiac alpha-myosin heavy chain) and MYLK2(myosin light chain kinase) for HCM and EMD(emerin) and G4.5(tafazzin) for DCM, were not analyzed in the patients carrying TCAPmutations. As such, the possibility remains that these patients might have additional disease-related mutations.
The authors are grateful to Drs. H. Toshima, K. Matsuyama, H. Kagiyama, T. Sakamoto, K. Kawai, K. Kawamura, M. Nagano, H. Yasuda, A. Matsumori, S. Sasayama, R. Nagai, and Y. Yazaki for their contributions to the clinical evaluation of and blood sampling from patients with HCM and DCM. The authors thank Dr. S. Labeit for providing us with Tcap constructs and affinity-purified anti-Tcap polyclonal antibody, as well as Dr. M. Yanokura and Ms. M. Emura for their technical assistance.
This work was supported in part by Grant-in-Aids for Scientific Research from the Ministry of Education, Culture, Sports, Science, and Technology, Japan; research grants from the Ministry of Health, Labour, and Welfare, Japan; and Mitsui Life Social Welfare Foundation. Dr. Jeffrey A. Towbin acted as Guest Editor for this paper.
- Abbreviations and acronyms
- dilated cardiomyopathy
- glutathion S-transferase
- hypertrophic cardiomyopathy
- left ventricle/ventricular
- muscle LIM protein
- polymerase chain reaction
- reverse transcription
- single-strand conformation polymorphism
- yeast two-hybrid
- Received May 19, 2004.
- Revision received July 26, 2004.
- Accepted August 25, 2004.
- American College of Cardiology Foundation
- Richardson P.,
- McKenna W.,
- Bristow M.,
- et al.
- Olson T.M.,
- Michels V.V.,
- Thibodeau S.N.,
- Tai Y.S.,
- Keating M.T.
- Li D.,
- Tapscoft T.,
- Gonzalez O.,
- et al.
- Young P.,
- Ferguson C.,
- Banuelos S.,
- Gautel M.
- Gregorio C.C.,
- Trombitas K.,
- Centner T.,
- et al.
- Frey N.,
- Olson E.N.
- Geier C.,
- Perrot A.,
- Ozcelik C.,
- et al.
- Orita M.,
- Iwahana H.,
- Kanazawa H.,
- Hayashi K.,
- Sekiya T.
- Cazorla O.,
- Wu Y.,
- Irving T.C.,
- Granzier H.
- Bottinelli R.,
- Coviello D.A.,
- Redwood C.S.,
- et al.
- Yanaga F.,
- Morimoto S.,
- Ohtsuki I.
- Tyska M.J.,
- Hayes E.,
- Giewat M.,
- et al.
- Elliott K.,
- Watkins H.,
- Redwood C.S.
- Arimura T.,
- Hayashi T.,
- Terada H.,
- et al.
- Vatta M.,
- Mohapatra B.,
- Jimenez S.,
- et al.