Loss of actomyosin regulation in distal arthrogryposis myopathy due to mutant myosin binding protein-C slow

Abstract

Myosin binding protein C (MyBP-C) is expressed in striated muscles, where it plays key roles in the modulation of actomyosin cross-bridges. Slow MyBP-C (sMyBP-C) consists of multiple variants sharing common domains but also containing unique segments within the NH₂ and COOH termini. Two missense mutations in the NH₂ terminus (W236R) and COOH terminus (Y856H) of sMyBP-C have been causally linked to the development of distal arthrogryposis-1 (DA-1), a severe skeletal muscle disorder. Using a combination of in vitro binding and motility assays, we show that the COOH terminus mediates binding of sMyBP-C to thick filaments, while the NH₂ terminus modulates the formation of actomyosin cross-bridges in a variant-specific manner. Consistent with this, a recombinant NH₂-terminal peptide that excludes residues 34–59 reduces the sliding velocity of actin filaments past myosin heads from 9.0 ± 1.3 to 5.7 ± 1.0 μm/s at 0.1 μM, while a recombinant peptide that excludes residues 21–59 fails to do so. Notably, the actomyosin regulatory properties of sMyBP-C are completely abolished by the presence of the DA-1 mutations. In summary, our studies are the first to show that the NH₂ and COOH termini of sMyBP-C have distinct functions, which are regulated by differential splicing, and are compromized by the presence of missense point mutations linked to muscle disease.—Ackermann, M. A., Patel, P. D., Valenti, J., Takagi, Y., Homsher, E., Sellers, J. R., Kontrogiannni-Konstantopoulos, A. Loss of actomyosin regulation in distal arthrogryposis myopathy due to mutant myosin binding protein-C slow.

Myosin binding protein C (MyBP-C), a family of accessory proteins of striated muscles, contributes to the assembly and stabilization of thick filaments and modulates the formation of actomyosin cross-bridges via direct interactions with both thick and thin filaments (1–6). Early studies have associated MyBP-C with the pathogenesis of hypertrophic cardiomyopathy, while recent reports have also implicated it in the development of skeletal myopathies (7–9). Three distinct MyBP-C isoforms have been identified, slow skeletal (sMyBP-C), fast skeletal (fMyBP-C), and cardiac (cMyBP-C), encoded by separate genes, namely MYBPC1, MYBPC2, and MYBPC3.

The core structure of MyBP-C is composed of 7 immunoglobulin (Ig) domains and 3 fibronectin-III (Fn-III) motifs, numbered from the NH₂ terminus as C1–C10. The C1 domain is preceded by a sequence of ∼50 aa that contains a high percentage of proline and alanine residues, referred to as Pro/Ala-rich motif, and followed by a conserved linker region of ∼100 aa, termed M motif. Unlike the cardiac and fast proteins, which exist as single isoforms, the slow protein consists of multiple variants that result from extensive alternative splicing (10–13). To date, at least 14 slow variants have been identified in humans, and at least 5 have been speculated in rodents (Fig. 1A). These share structural and sequence homology but also contain short unique segments located in the NH₂ terminus, the C7 domain, and the extreme COOH terminus (10). The slow variants are coexpressed in variable amounts and combinations in slow- and fast-twitch skeletal muscles and can coexist in the same myofiber (10, 14).

Figure 1.

sMyBP-C is a complex subfamily of proteins. A) Schematic representation of the 5 variants of sMyBP-C expressed in mouse skeletal muscle: v1, v2, v3, v4, and v002. sMyBP-C consists of tandem Ig (open ovals) and Fn-III (gray ovals) domains with the Pro/Ala-rich region and M motif shown as dark gray and open horizontal rectangles, respectively. Alternatively spliced regions are indicated by colored vertical lines. Point mutations linked to DA-1 myopathy are highlighted in yellow. B) Variant distribution of sMyBP-C in mouse FDB, TA, and soleus muscles is shown as a percentage of total sMyBP-C as calculated by averaging the percentage of variant expression obtained from RT-PCR. C) Wild-type and mutant DA-1 recombinant proteins used in this study containing the NH₂ terminus, domain C7, and COOH terminus.

In addition to being alternatively spliced, the NH₂ terminus of sMyBP-C undergoes phosphorylation, mediated by PKA and PKC (15). Five phosphorylation sites have been identified, including Ser-59 and Ser-62, which are substrates of PKA; Ser-83 and Thr-84, which are substrates of PKC; and Ser-204, which is a substrate of both PKA and PKC (15). Among these residues, Ser-59 is located within an NH₂-terminal insertion present only in select slow variants.

Recent genomic linkage analysis has demonstrated that mutations in MYBPC1 are causally linked to the development of distal arthrogryposis type 1 (DA-1), a skeletal muscle disorder characterized by contractures of the hands and feet (16). Two dominant missense mutations have been identified to date, W236R and Y856H (W, tryptophan; R, arginine; Y, tyrosine; H, histidine), located in the M motif and C8 domain, respectively. The congenital nature of the DA-1 contractures has been attributed to the early embryonic expression of sMyBP-C, which precedes that of fMyBP-C (7). Moreover, a homozygous nonsense mutation in the C2 domain of sMyBP-C (R318Stop) has been associated with the development of lethal congenital contractural syndrome (LCCS) type 4, the most severe form of arthrogryposis, resulting in neonatal lethality (9). The molecular defects that underlie the development of either of these types of arthrogryposis are though still elusive.

MyBP-C modulates actomyosin cross-bridge formation through its interaction with both thick myosin and thin actin filaments; via its COOH terminus, it binds directly to light meromyosin (LMM), while through its NH₂ terminus, it interacts weakly with subfragment 2 (S2) of sarcomeric myosin (17–19). In addition to binding S2, the NH₂ terminus of cMyBP-C, containing C0, the Pro/Ala-rich motif, C1 and the M motif, also mediates binding to actin in a Ca²⁺ dependent manner (20). However, this finding was recently challenged by Rybakova et al. (21), who reported that the COOH terminus of cMyBP-C supports binding to actin in a direct and saturable manner.

Given that sMyBP-C is expressed as many variants with distinct NH₂ and COOH termini, we herein explore how alternative splicing and the presence of the DA-1 mutations affects the actomyosin regulatory properties of sMyBP-C, using a combination of in vitro binding and motility assays. Our studies demonstrate that the NH₂ and COOH termini of sMyBP-C have distinct functional properties, which are regulated by the inclusion or exclusion of their respective novel insertions. While the NH₂ terminus directly binds to actin thin filaments and myosin head domains, modulating the formation of actomyosin cross-bridges in a variant-specific manner, the COOH terminus interacts with actin thin filaments in a variant-specific manner and with myosin thick filaments in a variant-independent manner. Notably, the DA-1 W236R mutation abolishes the binding and regulatory properties of the NH₂ terminus of all variants of sMyBP-C, while the DA-1 Y856H mutation eliminates the interactions of the COOH terminus of all variants of sMyBP-C with both thin and thick filaments. Collectively, our studies are the first to demonstrate that the activities of sMyBP-C are regulated via alternative splicing and to provide a mechanistic interpretation of the pathogenesis of DA-1.

MATERIALS AND METHODS

Characterization of the NH₂-terminal, C7-domain, and COOH-terminal spliced products of mouse sMyBP-C

Total RNA was isolated with Trizol reagent (Invitrogen, Carlsbad, CA, USA) from adult mouse flexor digitorum brevis (FDB), soleus, and tibialis anterior (TA) muscles. Approximately 5 μg of freshly prepared RNA was used to synthesize cDNA with the Superscript First Strand Synthesis System for RT-PCR (Invitrogen). Newly synthesized cDNA was used as template for PCR amplification of the 3 alternatively spliced regions, as described previously (15). Primers flanking the regions of interest are listed in Table 1. The resulting amplicons included the following: the NH₂ terminus, containing the Pro/Ala-rich motif, domain C1, and the M motif (accession numbers AEK21993.1, NP_780627.2, and NP_001239301.1); the C7 domain (NP_001239301.1 and AEK21994.1); and the COOH terminus containing the C10 domain and downstream amino acid residues (NP_780627.2 and NP_001239301.1). At least 10 PCR products for each alternatively spliced region per muscle type were analyzed and sequenced. The percentage expression of each amplicon per muscle type was calculated. In addition, the percentage expression of each variant per muscle type was estimated using the information obtained from the RT-PCR analysis of the different amplicons.

Table 1.

Oligos used in cloning

Amplicon or mutation	Primers
	Sense	Antisense
Directional cloning
Amplicon
NH2-terminal region	ATGCCAGAACCCACTAAG	ATCGAGAATTTTTGCAAA
C7 domain	AGCCCTCCTACTCTTCG	GGGCTCACTAGCCCCAGC
COOH-terminal region	CCCATGTTTACTCAACT	CGACTGTTGCTGCCCCTC
Mutagenesis
Mutation
NH2-W236R	GAGCCTGAGATAGACCTGCGGGAGCTGCTGAAAAATG	CATTTTTCAGCAGCTCCCGCAGGTCTATCTCAGGCTC
C8-Y856H	GATCCCCGGAATTCCAAACTCACATCCGCAGA	TCTGCGGATGTGAGTTTGGAATTCCGGGGATC

Generation of recombinant proteins

Recombinant proteins encoding the NH₂ terminus, the C7 domain, and the COOH terminus of sMyBP-C were designed as described above, using the primer sets listed in Table 1. The 3 NH₂-terminal constructs (aa 1–285) included the Pro/Ala-rich motif, Ig domain C1, and the M motif and differed from one another by inclusion of aa 21–59 (NH_2aa1–285) or exclusion of aa 21–59 (NH₂Δ21–59_aa1–285) or aa 34–59 (NH₂Δ34–59_aa1–285). The C7 domain constructs (aa 720–823) included the Fn-III domain C7 with (C7_aa720–823) and without (C7Δ763–780_aa720–823) aa 763–780. The COOH-terminal constructs (aa 1016–1127) contained Ig domain C10 with (C10_{aa1016–1127}) and without (C10Δ1107–1127_{aa1016–1127}) residues 1107–1127. Lastly, the C8-C10 COOH-terminal constructs (854–1127) contained Ig domain C8, Fn-III domain C9 and Ig domain C10 with (C8-C10_aa854–1127) and without (C8-C10Δ1107–1127_aa854–1127) residues 1107–1127.

Following amplification of all constructs, they were introduced into the pGEX4T-1 vector at EcoRI/XhoI sites (Amersham Pharmacia, Piscataway, NJ, USA) to generate GST-fusion proteins. Recombinant polypeptides were expressed by induction with 1 mM isopropyl-β-d-thioglycopyranoside (IPTG) for 4 h at 30°C and purified by affinity chromatography on glutathione-agarose columns. The DA-1 W236R and Y856H point mutations were introduced into the NH₂- and COOH-terminal constructs, respectively, via site-directed mutagenesis using the Quickchange mutagenesis kit (Stratagene, La Jolla, CA, USA), as previously reported (22). Sense and antisense oligonucleotide sets are listed in Table 1. The authenticity of all constructs was verified by sequence analysis. Mutant proteins were expressed as GST-fusions, as described above.

GST pulldown assays

GST pulldown assays were performed as described previously (14) with minimal modifications. Briefly, equivalent amounts of recombinant proteins bound to glutathione beads were incubated overnight at 4°C with 0.5 mg of adult mouse FDB muscle lysates, washed, eluted, and prepped for analysis via SDS-PAGE and Western blotting. Pulldown assays for myosin were performed in the presence of buffer A (20 mM imidazole, 180 mM KCl, 1 mM MgCl₂, 1 mM DTT, 10 mM NaN₃, and 0.25% Tween-20) and probed with antibodies for myosin (1:1000, M1570; Sigma, St. Louis, MO, USA). Pulldown assays for actin were carried out in buffer A with the addition of 0.2% Nonidet P-40 and probed with antibodies specific for sarcomeric actin (1:400, A2066; Sigma). Immunoreactive bands were detected with the Tropix Chemiluminescence kit (Applied Biosystems, Carlsbad, CA, USA). Assays were performed ≥3 times, and results were consistent.

Direct binding assays

Two types of direct binding assays were performed, far-Western blots and slot blots. The far-Western blot assays were performed as described previously (22) with minimal modifications. Briefly, myosin and actin were purified from rabbit psoas skeletal muscle as reported by Spudich et al. (23), and heavy meromyosin (HMM) was enzymatically prepared as described by Margossian and Lowey (24) and Kron et al. (25). Equivalent amounts (0.75 μg) of purified actin and HMM were separated by SDS-PAGE and transferred electrophoretically to nitrocellulose. Equivalent loading and transfer were confirmed by staining with Ponceau Red. Blots were incubated in buffer B (50 mM KCl, 20 mM MOPS, 4 mM MgCl₂, 0.1 mM EGTA, 2 mM DTT, 3% BSA, 10 mM NaN₃, 0.5% Tween-20, and 0.5% Nonidet P-40) for 8 h at room temperature, followed by incubation with buffer B containing 0.5 μg/ml of GST fusion proteins for 10 h at room temperature. Blots were washed extensively with buffer B and PBS and subsequently probed with antibodies to GST (1:1000; GenScript, Piscataway, NJ, USA). Immunoreactive bands were detected with the Tropix Chemiluminescence kit (Applied Biosystems).

The slot blot assays were performed similar to the far-Western blot assays. In brief, 500 ng of purified actin or HMM was spotted on nitrocellulose, allowed to dry, and incubated with buffer B for 8 h at room temperature. Following blocking in buffer B, 0.5 μg/ml of GST-fusion proteins was added for 10 h at room temperature. Blots were washed with buffer B and PBS and probed with anti-GST (1:1000; GenScript). Immunoreactive spots were detected as above.

Both types of in vitro binding assays were performed ≥3 times, and results were consistent.

The far-Western blots were analyzed using densitometry and ImageJ software (U.S. National Institutes of Health, Bethesda, MD, USA) to quantify the relative binding of the recombinant NH₂-terminal proteins to HMM and actin. For comparison purposes, binding of GST-NH_2aa1–285 to HMM and actin was arbitrarily set to 100%, and binding of GST-NH₂Δ34–59_aa1–285 and GST-NH₂Δ21–59_aa1–285 was calculated relatively to GST-NH_2aa1–285. Statistical significant differences in relative binding were calculated with Student's t test (P<0.01).

Actin gliding assays

Motility assays were performed as previously reported (26–29) using purified actin labeled with rhodamine-phalloidin (Invitrogen) and HMM. Assays were performed in a buffer containing 50 mM KCl, 20 mM MOPS (pH 7.4), 4 mM MgCl₂, 0.1 mM EGTA, 1 mM ATP, 50 mM DTT, 25 μg/ml glucose oxidase, 45 μg/ml catalase, and 2.5 mg/ml glucose, at 30°C with 0.15 mg/ml HMM. Before the addition of the final motility buffer, the HMM-covered surface was blocked with 1mg/ml of BSA, and unlabeled actin was used to block any inactive myosin heads. Recombinant sMyBP-C proteins were added at a concentration of 0, 0.1, 0.5, 1, 2.5, or 5 μM to the final motility buffer. Due to aggregation and precipitation of the NH₂Δ21–59 protein at concentrations ≥5 μM, we were only able to use it up to 2.5 μM.

The velocity of actin filaments' movement over an HMM-coated surface was analyzed using the Cell-Trak system (Motion Analysis, Santa Rosa, CA, USA; ref. 30). The reported mean ± se values were obtained from 3 different experiments following analysis of ≥500 actin filaments. Average velocities were normalized to sham conditions (i.e., in the absence of recombinant sMyBP-C proteins) and reported as percentage change in velocity. The fluorescence intensity of still frames at 5 and 75 s after addition of sMyBP-C was used to calculate the percentage of actin present on the motility surface. The average amount of actin on the surface was normalized to sham-treatment conditions and reported as percentage difference. Significance was calculated via Student's t test (P<0.01).

RESULTS AND DISCUSSION

sMyBP-C is a complex subfamily of proteins

Unlike cMyBP-C and fMyBP-C that exist as single isoforms, sMyBP-C is a subfamily of proteins that consists of multiple variants sharing common domains but also containing unique sequences. It was previously shown that human and rat MYBPC1 are alternatively spliced in at least 3 regions, encoding the NH₂-terminal Pro/Ala-rich motif, the Fn-III domain C7, and the extreme COOH terminus (10, 14). Given the high identity shared by the respective regions of the human and mouse genes (85, 84, and 95%, respectively), we set forth to examine whether these splicing events also take place in mouse skeletal muscles. To this end, we used primers (Table 1) that flanked the 3 regions of interest to amplify cDNA obtained from 3 different adult mouse skeletal muscles, including the relatively slow soleus (∼50% type I and ∼50% type IIA), the fast TA (∼75% type IIB and ∼25% type IIA) and the mixed fast FDB (∼70% type IIX, ∼13% type IIA, and ∼17% type I) muscle. In particular, the NH₂-terminal region included the Pro/Ala-rich motif, the Ig domain C1, and the M motif (aa 1–285; accession number AEK21993.1), the C7 region contained the entire Fn-III domain C7 (aa 720–823; AEK21994.1), and the COOH-terminal region included the C10 domain and residual COOH-terminal residues (aa 1016–1127; NP_001239301.1). We found that ≥3 alternative splicing events take place within the NH₂ terminus of sMyBP-C: one that results in inclusion of aa 21–59 (NH_2aa1–285) contained within variants 1 and 2 (v1 and v2), a second that results in exclusion of residues 34–59 (NH₂Δ34–59_aa1–285) missing from v3 and v4, and a third that results in skipping of aa 21–59 (NH₂Δ21–59_aa1–285) absent from v002 (Fig. 1A). Of the 3 amplicons analyzed, NH₂Δ21–59_aa1–285 was moderately expressed in TA (25%) but was undetectable in FDB (0%) and soleus (0%) muscles (Table 2). On the contrary, soleus mainly contained the NH_2aa1–285 amplicon (70%) and low amounts of the NH₂Δ34–59_aa1–285 amplicon (30%), and FDB possessed similar amounts of the NH_2aa1–285 (50%) and NH₂Δ34–59 (50%) amplicons (Table 2).

Table 2.

Percentage expression of amplicons from RT-PCR

Amplicon	Skeletal muscle type (%)
	FDB	TA	Soleus
NH2aa1–285	50	25	70
NH2Δ34–59aa1–285	50	50	30
NH2Δ21–59aa1–285	−	25	−
C7aa720–823	25	50	25
C7Δ763–780aa720–823	75	50	75
C10aa1016–1127	45	25	60
C10Δ1107–1127aa1016–1127	55	75	40

The splicing events occurring at the region encoding Fn-III C7 and the extreme COOH terminus were considerably less complex than those described for the NH₂-teminus, as they only involved the exclusion or inclusion of a short segment of amino acids. Specifically, we found inclusion of 18 residues (aa 763–780; accession number AEK21994.1) in the C7 Fn-III domain. While v3 contains this novel insertion (amplicon C7_aa720–823), v1, v2, v4, and v002 skip it (C7Δ763–780_aa720–823; Fig. 1A). The C7_aa720–823 amplicon corresponding to v3 was primarily present in TA muscle (50%), whereas the C7Δ763–780_aa720–823 amplicon expressed in v1, v2, v4, and v002 was predominantly expressed in soleus (75%) and FDB (75%) muscles (Table 2). Moreover, splicing at the extreme COOH terminus of MYBPC1 results in loss of 60 nucleotides encoding aa 1107–1127 (accession number NP_001239301.1). This causes a frameshift and use of a premature stop codon generating the COOH terminus of v2, v3, and v4 that is 20 aa shorter in length (aa 1107–1127; C10Δ1107–1127_{aa1016–1127}) compared with the COOH terminus of v1 and v002 (C10_{aa1016–1127}) (Fig. 1A). While soleus muscle mainly contains transcripts carrying these 20 aa (60%), and thus expresses the C10_{aa1016–1127} amplicon corresponding to v1 and v002, TA and FDB muscles predominantly express transcripts that exclude these 20 residues and thus contain the C10Δ1107–1127_{aa1016–1127} amplicon (75 and 55%, respectively) present in v2-v4 (Table 2).

Given the restricted expression and select combinations of these inserts in specific variants, we were able to estimate the percent expression of each variant in the 3 skeletal muscles examined. Thus, v1, v3, and v4 are abundantly expressed in the fast-twitch FDB muscle, v1 and v3 in the slow-twitch soleus muscle, and v2, v3, and v002 in the mixed TA muscle, although lower levels of other variants may also be present (Fig. 1B). Taken together, our results suggest a high level of complexity for sMyBP-C involving the differential coexpression of multiple isoforms in distinct skeletal muscles.

Role of alternative splicing within the NH₂ terminus of sMyBP-C

Given the presence of extensive differential splicing in the NH₂ terminus of sMyBP-C, we sought to examine how splicing affects its regulatory properties. We therefore produced a series of recombinant proteins containing the 3 different NH₂-terminal mouse variants that included the Pro/Ala-rich motif, the C1 domain, and the M motif (NH_2aa1–285, NH₂Δ34–59_aa1–285, and NH₂Δ21–59_aa1–285; Fig. 1C).

We performed GST pulldown assays to assess the ability of each recombinant protein to interact with native myosin or actin using protein lysates prepared from adult mouse FDB muscle (Fig. 2A). All 3 NH₂-terminal proteins were able to efficiently and specifically precipitate endogenous myosin, although to different extents (Fig. 2A, top panel). Thus, binding was significantly enhanced in the absence of residues 34 to 59 (NH₂Δ34–59_aa1–285) but not of 21 to 59 (NH₂Δ21–59_aa1–285). On the contrary, only the NH_2aa1–285 protein, including aa 21 to 59, was able to efficiently interact with endogenous actin (Fig. 2A, middle panel).

Figure 2.

The NH₂ terminus of sMyBP-C differentially modulates actomyosin interactions in a variant-specific manner. A) Equivalent amounts of NH₂-terminal sMyBP-C wild-type fusion constructs, including NH_2aa1–285, NH₂Δ34–59_aa1–285, and NH₂Δ21–59_aa1–285, were bound to glutathione matrices, incubated with protein homogenates prepared from adult mouse FDB muscle, and examined for their ability to retain endogenous myosin and actin. Total lysates were also included in the Western blot. Coomassie blue staining of the input proteins is shown in the bottom panel to indicate equal loading. B, C) Far-Western blot (B) and slot blot (C) assays were performed to assay direct interactions between the NH₂-terminal wild-type sMyBP-C constructs and myosin or actin. For the far-Western blot assays, actin (1 μM) and the HMM portion of myosin (1 μM) were separated by SDS-PAGE, transferred to nitrocellulose, and incubated with equivalent amounts of sMyBP-C NH₂-terminal proteins (NH_2aa1–285, NH₂Δ34–59_aa1–285, and NH₂Δ21–59_aa1–285). For the slot blot assays, actin and HMM (500 ng) were blotted onto nitrocellulose and incubated with equivalent amounts of sMyBP-C NH₂-terminal proteins. Positive interactions in both assays were determined by immunoprobing for the GST moiety of recombinant sMyBP-C proteins. D) Relative binding of recombinant NH₂-terminal proteins to HMM (solid bars) and actin (open bars) was quantified by densitometry of the immunoreactive bands detected in the far-Western blots; percentage (%) binding is reported relative to NH_2aa1–285, which was set to 100%. Percentages were calculated from 3 independent experiments. Error bars = sd (t test, P<0.01). E) In vitro motility assays were used to measure the sliding velocity of actin filaments over an HMM-coated surface in the presence and absence of NH₂-terminal sMyBP-C wild-type constructs (NH_2aa1–285, NH₂Δ34–59_aa1–285, NH₂Δ21–59_aa1–285). sMyBP-C constructs were added to the final motility buffer at various concentrations, including 0, 0.1, 0.5, 1, 2.5, and 5 μM. F) Bar graphs depicting the percentage of actin filaments available at the gliding surface 75 s following addition of the indicated proteins. Insets: representative snapshots of the gliding surface 75 s following the addition of the wild-type proteins at 1 μM. Error bars = sem. *P < 0.01 vs. sham treatment; t test.

To assess the ability of the different NH₂-terminal fragments to directly interact with actin and the head region of myosin or heavy meromyosin (HMM; this fragment of myosin was also used in our motility assays), we performed far-Western (Fig. 2B) and slot blot (Fig. 2C) assays. In agreement with the pulldown assays, all 3 NH₂-terminal variants were able to directly bind to HMM albeit to different degrees (Fig. 2B, C, top panels), and the binding was enhanced in the absence of residues 34 to 59 (NH₂Δ34–59_aa1–285). Similarly, all 3 recombinant proteins directly bound actin to different extents (Fig. 2B, C, bottom panels), with the presence of aa 21–59 (NH_2aa1–285) dramatically augmenting the interaction.

To assess the relative strength of binding between the 3 NH₂-terminal variants and HMM or actin, we used Image J software to perform densitometry of the immunoreactive bands detected in the far-Western blot assays (Fig. 2D). Binding of GST-NH_2aa1–285 to HMM and actin was arbitrarily set to 100%, and binding of GST-NH₂Δ34–59_aa1–285 and GST-NH₂Δ21–59_aa1–285 was normalized to that of GST-NH_2aa1–285. Both GST-NH₂Δ34–59_aa1–285 and GST-NH₂Δ21–59_aa1–285 exhibited ≤10% binding to actin relatively to GST-NH_2aa1–285. On the contrary, GST-NH₂Δ34–59_aa1–285 and GST-NH₂Δ21–59_aa1–285 showed ∼200 and ∼90% binding to HMM, respectively, compared with GST-NH_2aa1–285.

We next used actin-gliding assays to explore how alternative splicing may affect the ability of sMyBP-C proteins to modulate actomyosin binding and sliding. In vitro motility assays were performed in the presence of HMM and F-actin labeled with rhodamine-phalloidin. Consistent with the binding assays, the NH₂-terminal constructs were able to modulate the sliding velocity of actin filaments past myosin heads, albeit to different extents and in a variant-specific manner (see Supplemental Table S1). The NH₂Δ34–59_aa1–285 protein exhibited the highest degree of modulation among the 3 NH₂-terminal proteins by significantly reducing the sliding velocity of actin filaments past myosin heads even at the lowest concentration tested (0.1 μM; Fig. 2E, thin hatched bars, and Supplemental Movie S1). Notably, the reduction in sliding velocity was concentration dependent; as the concentration of NH₂Δ34–59_aa1–285 increased, the sliding velocity of actin filaments inversely decreased (Supplemental Table S1). On the contrary, the NH₂Δ21–59_aa1–285 protein elicited a similar response only at 2.5 μM, the highest concentration used in our assay for this recombinant protein (please see Materials and Methods, Fig. 2E, thick hatched bars). Interestingly, the NH_2aa1–285 protein failed to exert any regulatory effect on the sliding velocity of actin filaments (Fig. 2E, solid bars) and behaved similarly to sham control (8.3–10.1 ± 1.0–1.3 μm/s; Supplemental Table S1). It did, however, significantly reduce the amount of actin filaments available to interact with myosin heads at the gliding surface in a time- and concentration-dependent manner, even at the lowest concentration used (0.1 μM; Fig. 2F, solid bars). This phenomenon was not observed in the presence of either the NH₂Δ34–59_aa1–285 or the NH₂Δ21–59_aa1–285 proteins (Fig. 2F, thin and thick hatched bars, respectively). Notably, the NH_2aa1–285 protein is the only one that interacts efficiently with both myosin and actin filaments in binding assays (Fig. 2A–C).

Taken together, the binding studies and the in vitro motility assays suggest that sMyBP-C proteins that include aa 21–59 (e.g., v1 and v2) are capable to bind both actin and myosin filaments via their NH₂ terminus. Such slow variants may therefore reduce the availability of either or both filaments to interact with each other, preventing the formation of actomyosin cross-bridges. Conversely, sMyBP-C proteins lacking aa 34–59 or 21–59 (e.g., v3 and v4 or v002, respectively) can interact with myosin but not actin through their NH₂ terminus and may compete with actin by occupying its binding site on myosin heads. Our studies therefore demonstrate that alternative splicing at the NH₂ terminus of MYBPC1 regulates the ability of the different slow variants to modulate actomyosin binding and sliding.

Effect of the NH₂-terminal W236R DA-1 mutation

Given the differential effects that the NH₂ terminus of sMyBP-C exerts on actomyosin binding and sliding, we next examined how the presence of the W236R DA-1 mutation affects its regulatory properties. We introduced the W236R mutation into the 3 NH₂-terminal constructs (Fig. 1C) and tested their ability to interact with endogenous myosin and actin and regulate the gliding velocity of actin filaments. Pulldown assays demonstrated that the presence of the W236R DA-1 mutation completely abolished the ability of all 3 NH₂-terminal proteins (NH₂W236R_aa1–285, NH₂Δ34–59W236R_aa1–285, and NH₂Δ21–59W236R_aa1–285) to interact with native myosin or actin (Fig. 3A). Consistent with these findings, the W236R DA-1 mutation nearly eliminated the regulatory effect of the NH₂Δ34–59W236R_aa1–285 (Supplemental Movie S2) and NH₂Δ21–59W236R_aa1–285 proteins. Only at the highest concentration used (5 μM) was either recombinant protein able to significantly reduce the sliding velocity of actin filaments past myosin heads, minimizing the physiological importance of the observed effect (Fig. 3B). Although the NH₂W236R_aa1–285 protein appears to reduce the amount of actin available on the gliding surface, its effect is nonsignificant (P > 0.01) and nonsaturating, suggesting that it may be an artifact of the protein binding nonspecifically to the gliding surface (Fig. 3C, solid bars). Similar to their wild-type counterparts, NH₂Δ34–59W236R_aa1–285 and NH₂Δ21–59W236R_aa1–285 proteins did not reduce the amounts of available actin at the gliding surface (Fig. 3C, thin and thick hatched bars, respectively).

Figure 3.

The NH₂ terminus of sMyBP-C fails to modulate actomyosin interaction and sliding in the presence of the W236R DA-1 mutation. A) GST pulldown assays were used to assess the ability of the sMyBP-C NH₂-terminal variants (NH₂W236R_aa1–285, NH₂Δ34–59W236R_aa1–285, and NH₂Δ21–59W236R_aa1–285) to interact with endogenous myosin and actin in the presence of the W236R DA-1 mutation. B) In vitro motility assays were used to measure the sliding velocity of actin filaments over a surface coated with HMM in the presence or absence of the NH₂-terminal variants containing the W236R mutation (NH₂ W236R_aa1–285, NH₂Δ34–59W236R_aa1–285, and NH₂Δ21–59W236R_aa1–285), as described in Fig. 2. Bar graphs indicate the percentage change in the sliding velocity of actin filaments normalized to sham conditions. C) Percentage ± se of available actin at the gliding surface 75 s following addition of the mutant NH₂-terminal proteins. Insets: representative snapshots of the gliding surface 75 s following the addition of the mutant proteins at 1 μM. *P < 0.01 vs. sham treatment, ^#P < 0.01 vs wild-type proteins; t test.

Despite the extensive alternative splicing that takes place at the NH₂ terminus of MYBPC1, the region that includes the DA-1 W236R mutation is constitutively expressed. Therefore, all known sMyBP-C variants carry this mutation in DA-1 affected muscles and are unable to regulate the formation of actomyosin cross-bridges. Interestingly, the presence of the W236R mutation does not seem to affect the expression levels and proper distribution of the mutant slow variants at C-zones, as reported by Vydyanath et al. (31). Our studies therefore indicate that failure of mutant sMyBP-C proteins to regulate actomyosin binding and sliding may underlie the pathogenesis of DA-1 myopathy associated with the W236R conversion.

Role of alternative splicing within domain C7 and the COOH terminus of sMyBP-C

Examination of recombinant proteins containing the Fn-III domain C7 in pulldown assays demonstrated that regardless of the inclusion or exclusion of the unique insertion (aa 763–780), they were unable to precipitate native myosin or actin from total lysates (Fig. 4A, top and middle panels, respectively).

Figure 4.

Alternative splicing within the C7 domain has no effect on actomyosin interactions; however, alternative splicing within the C10 domain affects its ability to interact with thick filaments. A, D) C7 (A) and C10 (D) domain constructs of sMyBP-C (C7_aa720–823 and C7Δ763–780_aa720–823, A; and C10_{aa1016–1127} and C10Δ1107–1127_{aa1016–1127}, D) were assessed for their ability to interact with endogenous myosin and actin in GST pulldown assays. E, F) Far-Western blot (E) and slot blot (F) assays were also performed with the C10 domain constructs (C10_{aa1016–1127} and C10Δ1107–1127_{aa1016–1127}) to assay for direct interactions with actin and HMM. Although the COOH terminus of sMyBP-C is complexed with myosin and actin filaments in a variant-specific manner (D), it is not able to directly interact with HMM or actin (B). B, C, G, H) In vitro motility assays were used to measure the sliding velocity of actin filaments over a surface coated with HMM in the presence or absence of the C7 (C7_aa720–823 and C7Δ763–780_aa720–823; B, C) and C10 (C10_{aa1016–1127} and C10Δ1107–1127_{aa1016–1127}; G, H) domain proteins. B, G) Bar graphs indicate the percentage change in the sliding velocity of actin filaments normalized to sham conditions. C, H) Bar graphs depict the percentage ± se of available actin at the gliding surface 75 s following addition of the sMyBP-C proteins. Insets: representative snapshots of the gliding surface 75 s following the addition of the proteins at 1 μM. *P < 0.01 vs. sham treatment; t test.

Unlike the C7 domain, but similar to the NH₂ terminus, the COOH terminus exhibited differential binding to actin filaments, depending on the presence or absence of the 20-aa-long novel insertion. Construct C10_{aa1016–1127} including Ig domain C10 and residues 1107–1127 was unable to interact with endogenous actin, whereas construct C10Δ1107–1127_{aa1016–1127} containing only Ig C10 efficiently precipitated native actin (Fig. 4D, middle panel). On the contrary, both constructs, regardless of the presence or absence of the additional 20 residues, were able to precipitate endogenous myosin, though to different extents (Fig. 4D, top panel). To assess the ability of the different domain C10 fragments to directly interact with the head region of myosin (HMM) and actin, we again performed far-Western blot (Fig. 4E) and slot blot (Fig. 4F) assays. Neither protein was able to bind directly to HMM or actin (Fig. 4E, F), indicating that C10_{aa1016–1127} and C10Δ1107–1127_{aa1016–1127} interact with endogenous myosin through its LMM portion, as previously shown (19), which is absent from our assays.

In accordance with the binding assays, neither the C7_aa720–823 nor the C7Δ763–780_aa720–823 protein was able to modulate actin-gliding velocity (Fig. 4B), while the extreme COOH terminus modestly affected the formation of actomyosin cross-bridges (Fig. 4G), yet in a variant-specific manner. The C10_{aa1016–1127} protein, including the unique COOH-terminal sequence, decreased the actin sliding velocity only at the highest concentration used (5 μM; Fig. 4G, solid bars). On the contrary, C10Δ1107–1127_{aa1016–1127} failed to exert any regulatory effect (Fig. 4G, thin hatched bars). A reduction in the availability of actin filaments at the gliding surface was not observed in the presence of either the C7 or the COOH-terminal proteins (Fig. 4C, H, respectively).

The inclusion or exclusion of the unique insertion within domain C7 does not seem to affect the domain's ability to interact with actin or myosin or its capability to modulate the formation of cross-bridges. To date, the functional significance of the alternative splicing within domain C7 is still undetermined. The alternative splicing at the extreme COOH terminus does, however, affect the ability of the C10 domain to interact with actin filaments. Inclusion of the unique insertion abolished binding to actin. On the contrary, exclusion of the unique insertion allowed binding to actin. Binding of the C10 domain to endogenous myosin was not dependent on the presence of the unique insertion, although the inclusion of the 20-residue-long insertion appears to enhance binding to native myosin. Thus, while all slow variants can bind to native myosin to different extents, only v2, v3, and v4, which lack the novel COOH-terminal insertion, may interact with actin filaments.

Effect of the COOH-terminal Y856H DA-1 mutation

Previous studies have demonstrated that the COOH terminus of MyBP-C interacts with both thick and thin filaments (17–23, 25–29). The COOH-terminal C10 domain binds to the LMM portion of the myosin rod, and this interaction is significantly enhanced by the presence of the C8 and C9 domains (18, 19, 32–36). Since the C10 domain alone (C10Δ1107–1127_{aa1016–1127}) minimally precipitated native myosin (Fig. 4D), we assessed the ability of domains C8–C10 to interact with endogenous myosin and actin filaments in the presence or absence of the 20-aa insertion. We therefore produced recombinant proteins containing aa 854–1127, including Ig domain C8, Fn-III domain C9, and Ig domain C10 with (C8–C10_aa854–1127) and without (C8–C10Δ1107–1127_aa854–1127) residues 1107–1127 (Fig. 1C). The addition of the C8 and C9 domains did not enhance binding of the C10 domain to either endogenous myosin or actin in the presence of the unique 20-aa insertion (Fig. 5A, top and middle panels). However, the inclusion of domains C8 and C9 did augment binding of the C10 domain to myosin, but not to actin, in the absence of the unique insertion (Fig. 5A, top panel).

Figure 5.

COOH terminus of sMyBP-C mediates binding to thick filaments but fails to support binding in the presence of the Y856H DA-1 mutation. A) GST pulldown assays were used to assess the ability of the extended COOH-terminal constructs of sMyBP-C (C8–C10_aa854–1127 and C8–C10Δ1107–1127_aa854–1127) to interact with endogenous myosin and actin. B) COOH-terminal constructs of sMyBP-C carrying the Y856H DA-1 mutation (C8–C10Y856H_aa854–1127 and C8–C10Δ1107–1127 Y856H_aa854–1127) were assessed for their ability to interact with endogenous myosin and actin. The presence of the Y856H DA-1 mutation abolishes all binding activity.

To examine how the presence of the Y856H DA-1 mutation localized to the C8 domain affects the ability of domains C8-C10 to interact with endogenous myosin and actin, we generated mutant recombinant proteins carrying the Y856H mutation (C8-C10Y856H_aa854–1127 and C8-C10Δ1107–1127Y856H_aa854–1127; Fig. 1C). Pulldown assays demonstrated that the presence of the DA-1 mutation completely abolished all binding activity with native myosin and actin filaments (Fig. 5B).

Notably, the region that includes the DA-1 Y865H mutation within domain C8 is not contained within an alternatively spliced region and is constitutively expressed. Therefore, all known sMyBP-C variants carry this mutation in DA-1 affected muscles and are unable to interact with native myosin or actin filaments through their COOH termini. Similar to the NH₂-terminal DA-1 mutation, the presence of the Y856H mutation does not seem to affect the expression levels of the mutant sMyBP-C variants (31). Therefore, the mechanism that may underlie the pathogenesis of DA-1 myopathy associated with the Y856H mutation may involve its inability to interact with myosin and actin filaments.

Physiological complexity of sMyBP-C

Using biochemical tools, we show that alternative splicing occurring in MYBPC1 modulates the actomyosin regulatory properties of sMyBP-C (Fig. 6A). We also provide a mechanistic understanding of the pathogenesis of DA-1 myopathy by demonstrating that the presence of single missense mutations in the M motif and domain C8 of MYBPC1 gives rise to nonfunctional slow proteins that fail to regulate the formation of actomyosin cross-bridges, and interact with myosin and actin filaments, respectively (Fig. 6B). Given that sMyBP-C undergoes extensive phosphorylation (15) in addition to consisting of multiple variants that are potentially coexpressed in the same thick filament, it becomes apparent that its regulation and roles are highly complex. Thus, the systematic and comprehensive characterization of sMyBP-C is warranted before we are able to assign specific roles to each variant. Herein we have begun this immense task by assessing the effects of alternative splicing on actomyosin interactions and cross-bridge formation.

Figure 6.

sMyBP-C modulates the formation of actomyosin cross-bridges via its NH₂ terminus in a variant-specific manner, and binds to thick filaments through its COOH terminus. Both regulatory properties are abolished by single-point mutations linked to DA-1 myopathy. A) Proposed model depicting the different roles of the sMyBP-C variants within the sarcomeric A and M bands. B) Introduction of the single DA-1 point mutations within the NH₂ and COOH termini of MYBPC1 (W236R and Y856H, respectively) abolishes the ability of sMyBP-C to regulate actomyosin binding and sliding.

Earlier studies have demonstrated that the NH₂ and COOH termini of MyBP-C proteins interact with both thick and thin filaments (17–23, 25–29). Through its COOH-terminal C10 domain, MyBP-C supports binding to the LMM portion of the myosin rod (18, 19, 32–36). A weaker interaction between the NH₂-terminal M motif and the S2 fragment of myosin has also been reported (17). cMyBP-C also associates with filamentous actin in a phosphorylation-dependent manner (37). Several reports, focusing solely on cMyBP-C, have demonstrated that its NH₂ terminus (including Ig domain C0, present only in the cardiac isoform, the Pro/Ala-rich motif, C1, and the M motif) binds actin with micromolar affinity (37–40). Recently, however, it was shown that cMyBP-C interacts with actin via a single site located within the COOH-terminal C6–C10 domains, whereas domains C0–C5 bind actin weakly, and domains C0–C1 have no appreciable affinity for actin (21).

Our studies demonstrate that the NH₂ and COOH termini of sMyBP-C also mediate binding to myosin and actin filaments, however, in a variant-specific manner, while the C7 domain fails to bind either filament. The presence or absence of aa 21–59 does not affect the binding capacity of the NH₂ terminus for myosin filaments; however, exclusion of aa 34–59, leaving aa 21–33 intact, significantly enhances it. Therefore, v1 and v2, which express residues 21–59, or v002, which lacks residues 21–59, may exhibit weaker binding to native myosin via their NH₂ termini compared with v3 and v4, which exclude residues 34–59, but retain residues 21–33. On the contrary, inclusion of aa 21–59 is necessary for binding to actin, suggesting that while v1 and v2 can interact with actin filaments via their NH₂ termini, v002, v3, and v4 fail to do so.

In accordance with the binding studies, alternative splicing within the middle portion of the NH₂-terminal Pro/Ala-rich motif differentially regulates the ability of the slow variants to modulate the formation of actomyosin cross-bridges. Variants that include aa 21–59 (v1 and v2) may inhibit the formation of actomyosin cross-bridges by efficiently interacting with both myosin and actin filaments. However, variants that express residues 21–33 (v3 and v4) may reduce the rate of actomyosin cross-bridge formation by competing with actin filaments for binding to myosin heads. Lastly, variants that skip aa 21–59 (v002) modestly decrease the sliding velocity of actin filaments past myosin heads by weakly interacting with the latter. Therefore, alternative splicing of MYBPC1 is a mechanism through which the activities of sMyBP-C are highly regulated.

Previous work from our group has demonstrated that the presence of the 20-residue insertion at the very COOH terminus of sMyBP-C targets the protein at the periphery of the M band, where it interacts with the giant sarcomeric protein obscurin (14). Consequently, v1 and v002, which carry this insertion, are targeted to the M band, where they are able to interact with the LMM portion of myosin filaments and have a structural, rather than a regulatory, role (14, 19). On the contrary, v2, v3, and v4 localize to the C zones of the A band, where they are able to interact with both myosin and actin filaments and modulate the formation of actomyosin cross-bridges.

Taken together, our results support a model in which the NH₂ terminus of sMyBP-C strongly modulates the formation of actomyosin cross-bridges in a variant-specific manner, while the COOH terminus regulates its localization and ability to interact with myosin and actin filaments. It is therefore apparent that the regulation and roles of sMyBP-C within skeletal muscles are more complex than previously projected, especially given the presence of complex phosphorylation events that take place in the NH₂ terminus of the protein (15).

More important, our findings also demonstrate that the activities of sMyBP-C are completely abolished by the presence of single-missense mutations associated with arthrogryposis myopathy, further underscoring the importance of this multifaceted family of regulators in muscle physiology and pathophysiology. In particular, the presence of the W236R mutation eliminates the ability of the NH₂ terminus to interact with either myosin or actin, and thus modulate the formation of actomyosin cross-bridges. Whether tryptophan-236 contributes directly or indirectly to the binding of the NH₂ terminus to myosin and/or actin is currently unknown; however, it is tempting to speculate that substitution of the bulky hydrophobic tryptophan-236 by a positively charged arginine may induce an unfavorable conformational change that precludes binding of the NH₂ terminus to either filamentous system, leading to loss of regulation. Similarly, substitution of the polar, yet uncharged, tyrosine-856 by a positively charged histidine may also induce an unfavorable conformational change in the COOH terminus that eliminates binding to myosin. Interestingly, the minimal region on sMyBP-C that supports binding to myosin is domain C10, while domains C8 and C9 further enhance this interaction. Yet, the Y856H mutation is present in domain C8. It is therefore possible that the Y856H mutation alters the layout of the entire COOH terminus of sMyBP-C along thick filaments, rendering it inaccessible for binding. Further studies are needed in order to precisely determine the molecular alterations triggered by either mutation that lead to loss of binding and regulation.