|Home | About | Journals | Submit | Contact Us | Français|
Mutations in MYH11 cause autosomal dominant inheritance of thoracic aortic aneurysms and dissections. At the same time, rare, non-synonymous variants in MYH11 that are predicted to disrupt protein function but do not cause inherited aortic disease are common in the general population and the vascular disease risk associated with these variants is unknown.
To determine the consequences of the recurrent MYH11 rare variant, R247C, through functional studies in vitro and analysis of a knock-in mouse model with this specific variant, including assessment of aortic contraction, response to vascular injury, and phenotype of primary aortic smooth muscle cells (SMCs).
The steady state ATPase activity (actin-activated) and the rates of phosphate and ADP release were lower for the R247C mutant myosin than for the wild-type, as was the rate of actin filament sliding in an in vitro motility assay. Myh11R247C/R247C mice exhibited normal growth, reproduction, and aortic histology but decreased aortic contraction. In response to vascular injury, Myh11R247C/R247C mice showed significantly increased neointimal formation due to increased SMC proliferation when compared with the wild-type mice. Primary aortic SMCs explanted from the Myh11R247C/R247C mice were de-differentiated compared with wild-type SMCs based on increased proliferation and reduced expression of SMC contractile proteins. The mutant SMCs also displayed altered focal adhesions and decreased Rho activation, associated with decreased nuclear localization of myocardin-related transcription factor-A. Exposure of the Myh11R247C/R247C SMCs to a Rho activator rescued the de-differentiated phenotype of the SMCs.
These results indicate that a rare variant in MYH11, R247C, alters myosin contractile function and SMC phenotype, leading to increased proliferation in vitro and in response to vascular injury.
Thoracic aortic aneurysms leading to acute ascending aortic dissections (TAAD) are a common cause of premature deaths. Up to 20% of the patients who present with TAAD do not have an identified genetic syndrome, such as Marfan syndrome, but do have first degree relatives similarly affected, termed familial TAAD (FTAAD).1 Heterozygous mutations in genes encoding the major proteins in the smooth muscle cell (SMC) contractile filaments, smooth muscle-specific isoforms of α-actin (ACTA2) and myosin heavy chain (MYH11), respectively, are responsible for disease in 10 – 14% of FTAAD families.2–4 In addition to TAAD, ACTA2 mutations also predispose individuals to occlusive vascular disease, including early onset coronary artery disease, stroke and Moyamoya disease.5, 6 The vascular pathology in the occluded arteries of patients with ACTA2 mutations is characterized by increased numbers of SMCs in the media or neointima, and aortic SMCs explanted from these patients proliferate more rapidly in culture than control SMCs. These observations have raised speculation that the occlusive vascular diseases associated with ACTA2 mutations result from excessive proliferation of SMCs in response to vascular injury.
MYH11 mutations are a rare cause of FTAAD and are identified primarily in families with TAAD inherited in association with a patent ductus arteriosus (PDA)4. Similar to patients with ACTA2 mutations, individuals with MYH11 mutations can also present with early onset occlusive vascular disease and stenotic arteries in the vasa vasorum of the aorta due to increased numbers of SMCs.4 These observations have raised speculation that MYH11 mutations may predispose to both TAAD and occlusive vascular diseases, i.e., confer a predisposition to a similar range of vascular diseases as ACTA2 mutations. The myosin superfamily is a large class of motor molecules that interact with actin filaments to generate force or exert movement using energy generated through ATP hydrolysis. The vertebrate smooth muscle myosin is a hexameric complex composed of two myosin heavy chains (SM-MHC, encoded by MYH11), two essential light chains, and two regulatory light chains (RLCs). Myosin heavy chains consist of a globular head domain that includes the ATPase and actin binding subdomains, a linker region with Ig repeats that bind light chains, and a variable tail region that specializes each myosin for its cellular functions. SM-MHC’s coiled-coil domain in the tail region enables the protein to first dimerize and then polymerize to form thick filaments that interact with actin in thin filaments. The few MYH11 mutations identified in FTAAD families are deletions and missense mutations primarily located in the coiled-coil domain of the protein.2, 4 MYH11 mutations causing FTAAD differ from mutations in the cardiac-specific myosin heavy chain, encoded by MYH7, that cause hypertrophic cardiomyopathy. MYH7 mutations are primarily missense mutations involving the globular motor domain of the protein.7 Additionally, MYH7 is a gene that is commonly mutated in familial hypertrophic cardiomyopathy, whereas MYH11 mutations are a rare cause of FTAAD.8
The identification MYH11 mutations that cause FTAAD is complicated by the fact that rare, non-synonymous variants occur in MYH11 in the general population; 0.6% of the individual exomes in the Exome Rare Variant database have such a variant (http://evs.gs.washington.edu/EVS/). A recurrent MYH11 rare variant, R247C, is located in the head region of the myosin heavy chain and disrupts an arginine that is completely conserved across species. This alteration has been identified in patients with thoracic aortic disease without a family history but not in patients with FTAAD (unpublished data). A mutation in the corresponding amino residue in MYH7, R249Q, causes familial hypertrophic cardiomyopathy. The mutation in both the cardiac and smooth muscle isoforms lies near the ATP binding domain, and in vitro assays of MYH7 R249Q have confirmed decreased ATPase activity and filament velocity.9, 10 We sought to determine if the recurrent rare MYH11 variant R247C disrupts myosin function and alters SMC phenotype to understand its potential for contributing to aortic or occlusive vascular diseases.
To study wildtype (WT) and mutant R247C myosins, in vitro and in vivo studies were pursued. All experimental protocols are described in the Supplemental Methods section, available online at http://circres.ahajournals.org. Animals were cared for according to the NIH Guide for the Care and Use of Laboratory Animals. All animal breeding and experiments were performed under protocols approved by the University of Texas Health Science Center at Houston and University of Texas Southwestern Medical Center and in accordance with NIH guidelines.
To determine if the MYH11 R247C rare variant impacted the kinetics of the myosin motor, soluble, two-headed fragments (HMM fragments) of SM-MHC (SM1A isoform of human smooth muscle myosin) were expressed with and without the R247C mutation. As shown in Table 1, the mutation resulted in a lowering of the maximal actin-activated ATPase activity when the myosin was phosphorylated on the RLC. However, regulation was not affected in that the mutant, like the WT protein, had no activity in the absence of RLC phosphorylation.
We then examined the rate of inorganic phosphate (Pi) release when the myosin was allowed to bind and hydrolyze ATP, and then mixed with actin. This step is rate-limiting for the overall ATPase cycle in WT SM-MHC and is the entry point into the force generating states on actin (see Sweeney and Houdusse for review;11 Online Figure I). As shown in Table 1, Pi release was lower for the mutant myosin and was only slightly faster than the overall ATPase cycle rate for both mutant and WT proteins, consistent with it being the rate limiting step in the overall cycle. We next measured the rate of release of ADP from mutant and WT myosin. The release of ADP from SM-MHC limits the rate of myosin detachment from actin as shown in Online Figure I, and also limits the maximal shortening velocity in the absence of load.11 The rate of ADP release was lower for the mutant than for the WT as shown in Table 1, as was the rate of actin filament sliding in an in vitro motility assay.
The overall time that myosin spends bound strongly to actin during its kinetic cycle is known as the duty ratio and will ultimately be a determinant of the force that the myosin produces. From the Pi and ADP release rates we can calculate the unloaded duty ratio, making the assumption that all other rates are much faster. This assumption is warranted given that the Pi release rates are only slightly faster than the overall actin-activated ATPase rate for both mutant and WT SM-MHC. As shown in Online Figure I, the rate of Pi release (k+4′) controls exit from the non-force generating states and actin-myosin ADP release (k+5′) controls exit from the force-generating states. Thus the duty ratio = k+4′ / (k+4′ + k+5′). In the absence of load, the calculated duty ratio of the R247C mutant is almost half of the WT protein (Table 1). As load increases, k+4′ and the duration of the non-force generating states remains unchanged, while the duration of the force-generating states increases (k+5′ decreases). Thus it is likely that even under isometric conditions; the duty ratio of the mutant will be decreased as compared to WT due to the reduction in the rate of Pi release. Thus force production and shortening velocity are both likely decreased in vivo by the MYH11 R247C mutation.
To investigate the specific physiological and vascular effects of the mutation Myh11 R247C, we generated a targeting vector to flox the Myh11 R247C allele between two loxP sites of the backbone, which would knock in the R247C mutation and allow for Cre-mediated conditional inactivation of the Myh11 gene. The targeting vector contains a 4.76 Kb 5′ targeting arm, a floxed 2.04 Kb DNA fragment spanning exons 7 and 8, a neo expression cassette for positive selection, a 6.84 Kb 3′ targeting arm and a tk expression cassette for negative selection (Figure 1A). The floxed exons 7 and 8 contain the mutation from C to T at nucleotide 846, which corresponds to the amino acid change R247C. In addition, the deletion of exons 7 and 8 would cause a coding-frame shift between exons 6 and 9 and result in disruption of the Myh11 gene (Figure 1A). After the mouse ES cells were electroporated with the targeting vector DNA and cultured in selection medium containing G418 and FIAU, 288 surviving clones were isolated and DNA samples were prepared from these clones for Southern blot screening. Four clones were positive for recombination by restriction digest analysis (Figure 1B). DNA sequencing confirmed 3 of 4 positive clones containing the mutant Myh11 R247C sequence. Since this targeted allele contains an Frt-flanked neo cassette, this allele is designated as Myh11R247C(f-neo).
Three targeted ES clones were injected into the blastocysts of C57BL/6J mice to produce chimeric mice. A total of 8 chimeric founder mice were identified to have germ line transmission capability. Two of these founder mice (lines A and B) were crossed with female FLPeR mice20 to remove the Frt-flanked neo cassette in their pups with an Myh11 R247Cf-neo/+ FLPeR/+ genotype. As shown in Figure 1, PCR analyses detected WT and targeted alleles in FLPeR-negative Myh11R247C(f-neo)/+ mice and WT and floxed alleles in FLPeR-positive and neo-negative Myh11R247C/+ mice (Figure 1C). Subsequently, Myh11R247C/+ breeding pairs were used to generate WT and Myh11R247C/R247C mice that were negative of both neo cassette and FLPeR. DNA sequencing was performed to confirm the presence of the knock-in mutation (Figure 1D). The established mouse colony had a mixed 129SvEv and C57BL/6J strain background. Assessment of the mRNA and protein expression of myosin (Myh11), α-actin (Acta2) and calponin (Cnn1) in aortic tissues using quantitative PCR and immunoblot analyses showed similar expression and protein levels for these contractile proteins between WT, and Myh11R247C/R247C aortas (Figure 1E and F, Online Figure II).
The predicted genotypes based on Mendelian inheritance were observed with breeding of mice heterozygous for the mutation. The average body weights of 1, 2, and 8 month old male and female mice of various genotypes were not statistically different (Online Table II). Up to age 18 months, Myh11R247C/R247C and Myh11R247C /+ mice were indistinguishable from WT mice with regard to life expectancy. Finally, the reproductive capability of Myh11R247C /+ and Myh11R247C/R247C mice with WT mice was not altered (Online Table I). Together, these data suggest that introduction of the R247C alteration into SM-MHC did not affect mouse survival, development, or reproduction.
Contractile force development in response to phenylephrine was attenuated in ascending aortic rings from Myh11R247C/R247C mice compared to littermate control mice at 5 months (Figure 2A). Similar differences were found in response to 90 mM KCl (data not shown). In contrast, there was no attenuation in RLC phosphorylation (Figure 2B). Therefore, the defective contractile response does not result in changes to signaling modules acting on myosin light chain kinase but appears to be intrinsic to the properties of the mutated myosin.
There was no abnormality or enlargement aorta in these mice apparent by echocardiography and Doppler studies at 7 or 10 months of age (Figure 2C). The Myh11R247C/R247C mice did not show any dissection, elongation or tortuosity of the thoracic aorta, and their vessels were morphologically indistinguishable from their WT and Myh11R247C/+ littermates (Online Figure III). There were no differences in systolic or diastolic blood pressures between WT and Myh11R247C/R247C mice (systolic: WT, 98± 3 vs mutant, 95±7 mm Hg; diastolic, WT, 74±23 vs mutant, 70±7 mm Hg). Histological analysis of the ascending and descending aortas of the WT and Myh11R247C/R247C aortas assessed at various ages up to 10 months of age did not reveal any evidence of vascular pathology (Figure 2D, Online Figure IV and V). There was no evidence of aortic SMC proliferation as detected by anti PH3 and α-actin immunostaining at 6 months of age (Figure 2E and F). Additionally, RNA isolated from the ascending thoracic aorta at 10 months of age showed no significant increase in expression of lumican (p=0.81), decorin (p=0.46), or matrix metalloproteinase 2 (Mmp2; p=0.44) expression by Q-PCR (Figure 3G). Therefore, the homozygous Myh11 R247C allele decreases aortic contractility, but there was no evidence that the presence of mutant SM-MHC causes aortic pathology consistent with thoracic aneurysm formation.
To test the hypothesis that the Myh11 R247C alteration would lead to increased neointimal formation with vascular injury, we performed carotid artery ligation using the carotid flow cessation injury model and analyzed the vascular injury response in WT and Myh11R247C/R247C mice at 7, 14, and 21 days after injury. Myh11R247C/R247C mice exhibited enhanced formation of neointima as compared with WT mice (Figure 3A, Online Figure VI). Neointimal area and intima/media ratio were significantly larger in Myh11R247C/R247C mice than WT mice at 14, and 21 days (Figure 3B and C). Medial areas of injured vessels were increased as compared with uninjured vessels in both Myh11R247C/R247C and WT mice, but not significantly different from one another at any time point (data not shown). Myh11R247C/+ mice were also analyzed, but showed no increased neointimal formation compared with WT (Online Figure VII).
To evaluate cellular proliferation in vivo, we quantified proliferating cells in the injured arteries by using anti-PH3 staining, a marker of mitosis.12 There was a significant increase in the number of anti-PH3 positive cells in the arteries of Myh11R247C/R247C mice compared with WT (Figure 3D and E). Thus, the SMC response to vascular injury in the Myh11R247C/R247C mice is characterized by excessive SMC proliferation and increased neointimal formation.
Carotid artery remodeling in response to vascular injury has been associated with increased matrix metalloproteinase 2 levels.13 To determine the expression level of Mmp2 in the injured vessels from Myh11R247C/R247C mice, ligated and un-ligated carotid arteries were harvested at days 14 and 21 after carotid artery ligation, and the gene expression was measured by Q-PCR. Mmp2 mRNA levels were significantly increased in the ligated carotid arteries at days 14 and 21 after vascular injury in Myh11R247C/R247C mice (6-fold) as compared to the WT mice (3-4-fold, p<0.05; Figure 3F). In contrast, Mmp2 expression was not significantly different in the unligated arteries.
To further characterize the effect of the Myh11 R247C mutation on SMC phenotype and function, SMCs were explanted from the ascending thoracic aorta; these SMCs are neural crest-derived.14 Three separate explants were derived and the results described were consistent in all explants. The Myh11R247C/R247C aortic SMCs proliferated significantly more rapidly than the WT SMCs (Figure 4A). Myh11R247C/+ SMCs were also explanted; however they did not show increased proliferation compared with WT SMCs (Online Figure VII). Due to the lack of a proliferative response in vitro or in vivo of the heterozygous cells, further analyses were performed primarily on Myh11R247C/R247C and WT cells.
Increased proliferation of SMCs is typically associated with de-differentiated SMC phenotype characterized by decreased expression of SMC contractile proteins.15 Quantitative PCR analysis of message isolated from Myh11R247C/R247C versus WT SMCs showed decreased expression levels of Acta2, Cnn1, and Myh11 (Figure 4B). Immunoblot analysis performed on protein harvested from the Myh11R247C/R247C SMCs also demonstrated decreased protein levels of α-actin (Acta2), calponin (Cnn1) and SM-MHC (Myh11) (Figure 4C, Online Figure II). Myh11R247C/+ cells had decreased expression of contractile proteins (Online Figure VIII). These data suggested that the homozygous Myh11 R247C mutation alters the phenotype of SMC, leading to increased proliferation.
Consistent with the observation that Myh11R247C/R247C SMCs are de-differentiated compared to WT cells, immunofluorescence staining of the mutant SMCs demonstrated α-actin filaments with an altered cellular structure compared to WT SMCs (Figure 4E, Online Figure IX). The Myh11R247C/R247C filaments were thinner and shorter than the WT filaments, and the mutant filaments crossed the cell body but typically did not extend to the periphery of the cell, whereas the WT filaments extended to the cell periphery. Additionally, pools of unpolymerized α-actin were present in a subset of mutant cells but not in the WT cells. An F/G actin assay confirmed that Myh11R247C/R247C cells had an increase in the ratio of monomeric α-actin (G) to filamentous α-actin (F), whereas WT cells had more F-actin than G-actin (Figure 4F).
In SMCs, the transcription factor serum response factor (SRF) regulates SMC phenotype via association with its cofactors, the myocardin-related transcription factors (MRTFs). MRTF-A is a potent transcriptional coactivator that links changes in actin filaments to gene expression to alter the phenotype of SMCs.16 Increased pools of monomeric -actin, such as those observed in Myh11R247C/R247C cells, can sequester MRTF-A in the cytoplasm and prevent its translocation to the nucleus to drive SMC differentiation and contractile gene expression. Assessment of MRTF-A localization in Myh11R247C/R247C by immunofluorescence revealed a shift from a primarily nuclear localization of MRTF-A in WT SMCs to predominantly cytoplasmic localization in Myh11R247C/R247C SMCs (Figure 4G, Online Figure IX). This differential localization was quantified by computing a Pearson coefficient for the correlation of blue and green staining in the same pixels, confirming a significant decrease in nuclear localization of MRTF-A in Myh11R247C/R247C SMCs when compared to WT SMCs.
Since focal adhesion (FA) composition and maturation is dependent on force generation across the cell by cellular type II myosin motors, we also assessed FAs in the Myh11R247C/R247C SMCs. Inhibition of these myosin motors in cells using blebbistatin disrupts the maturation of nascent FAs, altering the recruitment of small G protein signaling molecules and leading to an absence of RhoA activators and an enrichment of Rac1 activators.17 To determine if the FAs were altered in the Myh11R247C/R247C SMCs, we performed immunofluorescence using an antibody against the FA protein vinculin, which demonstrated more numerous FAs in Myh11R247C/R247C SMCs, with each individual adhesion appearing smaller and rounder than the rod-like adhesions seen in WT cells (Figure 5A). These changes were enhanced with exposure to TGF-β1 for 48 hours. Adhesion size was quantified using confocal images, confirming decreased FA size in Myh11R247C/R247C SMCs. Focal adhesion kinase, FAK, a major signaling molecule at FAs, has been shown to drive actin polymerization as well as proliferation in SMCs. Immunofluorescence using an antibody recognizing phosphorylated Y397 of FAK (pFAK) showed increased staining, and immunoblot analysis confirmed an increase in cellular pFAK (Figure 5B and C, Online Figure II). FAK activates multiple pathways, including Akt signaling, and phosphorylation of Akt was also increased in Myh11R247C/R247C SMCs (Figure 5C, Online Figure II). Blocking FAK activation using a specific inhibitor (PF537228) decreased proliferation in Myh11R247C/R247C SMCs (Figure 5D). However, the FAK inhibitor did not increase nuclear localization of MRTF-A, nor did it stimulate any change in expression of SMC contractile genes (Online Figure X).
Altered FA maturation due to disruption of the type II myosin motors would be expected to alter Rac1 and RhoA signaling.17 As predicted, Rac1 activity was increased and RhoA activity decreased in Myh11R247C/R247C SMCs when compared to WT SMCs (Figure 5E and F). Interestingly, there was no increase in Rac1 activation or decrease of RhoA activation observed in the aortic tissue of Myh11R247C/R247C mice compared to WT mice (Figure 5G and H). We sought to determine if the altered Rac1 and RhoA signaling was responsible for the de-differentiated phenotype of the Myh11R247C/R247C SMCs. Inhibition of Rac1 using a specific inhibitor (NSC23766) had no effect on the proliferation of Myh11R247C/R247C SMCs or on the localization of MRTF-A (Online Figure X). Cellular RhoA activity levels can be activated using an activator derived from the bacterial cytotoxic necrotizing factor endotoxins, which converts RhoA to a constitutively active form by deaminating glutamine 63.18 We confirmed that when mutant and WT SMCs were exposed to this compound, CN03, RhoA activity levels were significantly increased (Fig. 6A). Further analyses determined that exposure to CN03 increased the nuclear localization of MRTF-A in the Myh11R247C/R247C SMCs (Figure 6B) and significantly increased expression of smooth muscle specific contractile genes after 4 hours of exposure (Figure 6C). Immunoblot also showed an increased in contractile protein levels in the Myh11R247C/R247C SMCs after 24 hours of exposure to CN03 (Figure 6D, Online Figure II). Finally, treatment of Myh11R247C/R247C cells with CN03 for 24 hours increased the polymerization of α-actin as assessed by immunofluorescence. Actin polymerization was quantified by computing a Pearson coefficient for colocalization of α-actin (green) and phalloidin (red), showing a dramatic increase in colocalization after treatment in the Myh11R247C/R247C cells (Figure 6E). Although twenty-four hours of CN03 treatment did not significantly decrease the proliferation of Myh11R247C/R247C cells, the difference in proliferation between Myh11R247C/R247C and WT SMCs was no longer significant after treatment (p=0.09, Online Figure X). Therefore, increasing Rho activity levels in the Myh11R247C/R247C SMCs rescued the de-differentiated SMCs phenotype based on increased nuclear localization of MRTF-A, increased contractile gene and protein expression, and increased polymerization of α-actin.
This study sought to determine if a recurrent rare variant, R247C, in the SMC-specific isoform of the myosin heavy chain disrupted myosin function and altered SMC phenotype. The R247C variant was chosen because it is a recurrent rare variant of unknown significance, and mutation of the paralogous amino acid in cardiac myosin heavy chain, MYH7, causes hypertrophic cardiomyopathy. This study demonstrates that MYH11 R247C leads to reduced force production and shortening velocity of SM-MHC in vitro and decreased aortic contractility in vivo. Despite decreased aortic contractility in the Myh11R247C/R247C aortas, RLC phosphorylation responses were not different between the mutant and WT aortas, indicating consistency in activation of signaling pathways leading to RLC phosphorylation. Thus, the defective contractile response does not result from changes to signaling modules acting on myosin light chain kinase and RLC phosphorylation, but appears to be due to intrinsic properties of the mutated myosin. The decrease in aortic contractility was not associated with aortic pathology in these mice and suggests that decreased aortic contractility alone may not be sufficient to cause thoracic aortic disease. In contrast, vascular injury induced significantly more neointimal formation in Myh11R247C/R247C mice than in WT mice and mutant SMCs proliferated more rapidly in culture than WT cells. SMC hyperplasia associated with this genetic variant may increase the risk of occlusive vascular diseases in individuals with this variant.
Aortic SMCs explanted from the Myh11R247C/R247C mice had an altered phenotype characterized by increased proliferation and de-differentiation as defined by decreased expression of contractile proteins. The SRF:MRTF axis is a well-characterized pathway linking actin filament formation to SMC proliferation and differentiation.19 With actin polymerization, MRTF-A moves from the cytoplasm to the nucleus and induces the transcription of SMC-specific contractile genes with the transcriptional co-activator SRF. If actin filaments are disrupted, MRTF-A moves out of the nucleus and binds to monomeric actin, leading to the downregulation of contractile gene expression and allowing SRF to bind to ternary complex factors (TCFs).20, 21 The SRF:TCF complex activates a subset of SRF-regulated growth responsive genes, leading to cell proliferation. Our results indicate that altering R247 in the myosin motor domain leads to de-differentiation of the SMCs via the SRF:MRTF axis, based on the decreased expression of contractile proteins, increased expression of c-fos, and increased cytoplasmic localization of MRTF-A when compared to WT SMCs.
Altered FAs in the Myh11R247C/R247C SMCs also contributed to the altered phenotype of these cells. The generation of tension across cells with adhesion to the matrix requires functional cellular actin and myosin and this tension drives FA maturation, composition and localization.22 Nascent FAs form at the periphery of cells with the binding of integrins to the extracellular matrix. These nascent FAs recruit FAK, paxillin and other proteins and initiate the anchoring of actin stress fibers to integrins. The conversion of nascent to mature FAs is dependent upon further force generation, which in turn is dependent on cellular type II myosin motors.23 Recent studies have identified differential recruitment of proteins to FAs in the presence and absence of ATPase inhibitor blebbistatin, which prevents force generation by cellular type II myosin motors.17 In particular, RhoA activating proteins are absent in the immature focal adhesions of blebbistatin-treated cells, while Rac1 activators are enriched. RhoA-dependent regulation of the actin cytoskeleton has been previously established to regulate SMC differentiation marker gene expression by modulating SRF:MRTF-dependent transcription.24 Knockdown of RhoA signaling prevents transcription of contractile genes in SMCs, and constitutive activation of RhoA induces both actin polymerization and increases contractile gene expression in SMCs via the SRF:MRTF axis.24, 25 Our data suggest that the Myh11 R247C variant prevents maturation of FAs, thus decreasing Rho activation and leading to a de-differentiated SMC phenotype. In fact, activation of RhoA rescued the de-differentiated SMC phenotype in Myh11R247C/R247C SMCs as assessed by increased MRTF-A nuclear localization and increased contractile protein expression. The heterozygous Myh11R247C/+ cells had intermediate levels of SMC contractile gene expression but no increase in proliferation relative to WT SMCs, which may indicate that the presence of 50% WT myosin provides sufficient force generation to allow the FAs to mature and prevent proliferative pathways from being activated. Although the SMCs contain other cellular myosin motors, including nonmuscle myosin IIA and IIB (encoded by Myh9 and Myh10, respectively), these myosins were not able to compensate for the loss of SM-MHC function to generate sufficient cell tension to drive FA maturation or force generation in aortic SMCs. Therefore, our results demonstrate a previously uncharacterized role for myosin motors and FAs in determining SMC phenotype.
A similar SMC phenotype, characterized by decreased RhoA signaling and SMC de-differentiation via the SRF:MRTF axis, was reported in the aortas and explanted aortic SMCs from the SMC-specific integrin-linked kinase (ILK) knockout mice.26 ILK, like FAK, is a kinase that binds to the integrin β1 subunit and plays a critical role in the organization of the actin cytoskeleton through its association with parvin, paxillin and vinculin.27 Mice with SMC-specific reduction of ILK exhibit vascular pathologies similar to the vascular diseases observed in FTAAD patients with disease-causing MYH11 mutations, including aneurysms of the ascending aorta and PDA. Aortic tissue and explanted SMCs from ILK-deficient mice showed similar reductions in the expression of contractile proteins due to aberrant localization of MRTF-A to the cytoplasm, with concomitant decreased levels of RhoA activation.
Although we observed de-differentiation of SMCs in culture and increased proliferation of SMCs with injury, aortas from the Myh11R247C/R247C had RhoA activity and contractile protein levels similar to WT aortas. In the intact aorta, the interactions between the extracellular matrix (ECM) and intracellular actin-myosin contractile filaments occur through integrin receptors in dense plaques, which have a similar function to FAs. Unlike FAs in tissue culture, these dense plaques are under continuous biomechanical forces due to pulsatile blood flow. Therefore, other sources, like ILK, may be responsible for the normal levels of Rho activation in Myh11R247C/R247C aortas. Alternatively, Rho activation in vascular tissues may be enhanced by signaling brought about by GPCR agonists, such as norepinephrine, that couple to RhoA guanine nucleotide exchange factors.28
The similarity of the phenotype of the SMC-specific reduction of ILK mice and FTAAD patients with MYH11 mutations, specifically thoracic aortic aneurysms with PDA, could provide insight as to why the Myh11R247C/R247C mice do not have aortic pathology. As described above, the ILK-deficient SMCs are de-differentiated in vivo, while our data suggest the Myh11R247C/R247C SMCs in vivo are not de-differentiated. Furthermore, we identified no differences in RLC phosphorylation, suggesting that cell signaling leading to RLC phosphorylation is intact. The majority of the identified MYH11 mutations leading to FTAAD are predicted to disrupt SM-MHC polymerization into thick filaments, which could cause a more severe defect in myosin function and disrupt signaling through dense plaques in the aortic SMCs in vivo. The altered signaling through dense plaques could decrease ILK signaling, therefore predisposing to thoracic aortic aneurysms and PDA, which is the same phenotype identified in the ILK deficient mouse.
Myh11 deficient mice die shortly after birth from vascular complications, as well as from dysfunction of the bladder and intestine.29 However, Myh11R247C/R247C mice demonstrate no evidence of bowel or bladder dysfunction, normal growth and survival. Additionally, the reproductive capacity of these mice is similar to WT mice despite the fact that smooth muscle contraction is required for uterine function. Physiologically, myosin light chain kinase (MLCK) is not fully activated nor is RLC maximally phosphorylated in visceral smooth muscles.30, 31 The partial decrease in myosin function in the mutant mice may be compensated for by increased nervous and hormonal influences sufficient to maintain visceral functions. Compensation by other type II myosin motors may also explain the lack of contractile problems in other organs.
Mutations in MYH7, the cardiac β-myosin heavy chain, typically cluster in the functional subdomains of the myosin motor head domain, specifically in the ATP-binding and actin binding clefts.7 In fact, studies of the alteration of the paralogous amino acid in MYH7, R249Q, that causes hypertrophic cardiomyopathy have shown decreased ATPase activity and sliding filament velocity, similar to the results reported here for R247C in MYH11.9, 10 The MYH11 mutations causing FTAAD are not in the head domain, but rather located in the coiled-coil rod domain and predicted to destabilize myosin filament formation. The one exception is a missense mutation at amino acid 712 near the converter domain, which transduces force from the ATPase motor and allows the flexible movement of the myosin head along actin. Based on these data, it is tempting to speculate that cardiomyocytes are more sensitive to disturbances of myosin motor function than vascular SMCs.
In summary, these data show that the rare variant in MYH11, R247C, can alter SM-MHC function, aortic contraction and SMC phenotype. These findings raise the possibility that this variant may predispose to vascular diseases in conjunction with other environmental or genetic risk factors. It is clinically important to distinguish between MYH11 variants that cause Mendelian inheritance of TAAD from those variants that are benign or confer a low or no risk for vascular disease and additional studies are needed to fully understand which MYH11 variants confer a high risk for aortic disease. Finally, these data demonstrate a previously unrecognized role of FAs in influencing SMC phenotype.
This study sought to determine the role of rare variants in MYH11 in the pathogenesis of vascular disease, including aortic aneurysms and occlusive diseases. We generated a knock-in mouse model of the most frequently recurrent rare variant, R247C, and showed that this variant decreases force generation in the aorta but does not lead to aneurysms. Knock-in mice had increased neointimal formation in response to vascular injury in vivo and increased smooth muscle cell proliferation in vitro. Cultured smooth muscle cells homozygous for the R247C mutation were de-differentiated with immature focal adhesions and decreased Rho activation. A Rho activator rescued the de-differentiated phenotype of these cells. This study identifies for the first time a key role for focal adhesions in the regulation of smooth muscle cell phenotype, and also suggests that in smooth muscle cells the force generation specifically of smooth muscle myosin rather than cytoskeletal myosin is crucial for the maturation of focal adhesions. Additionally, the data suggests that rare variants in MYH11 in humans may increase the risk of developing vascular occlusive disease.
We would like to acknowledge the technical assistance of Liqiong Chen, Xiaoyan Liu, Alan B. Zong, and Zhaohui Yang in preparing and assaying recombinant SM-MHC.
Sources of Funding
The following sources provided funding for these studies: P50HL083794-01 (D.M.M.), R01 HL62594 (D.M.M.), R01 HL26043 (J.T.S.), Whitaker Foundation, Pasani Family Foundation. This work was supported in part by the Center for Clinical and Translational Sciences, which is funded by National Institutes of Health Clinical and Translational Award TL1 RR024147 from the National Center for Research Resources.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.