Essential Roles of an Intercalated Disc Protein, mXinβ, in Postnatal Heart Growth and Survival
Rationale: The Xin repeat–containing proteins mXinα and mXinβ localize to the intercalated disc of mouse heart and are implicated in cardiac development and function. The mXinα directly interacts with β-catenin, p120-catenin, and actin filaments. Ablation of mXinα results in adult late-onset cardiomyopathy with conduction defects. An upregulation of the mXinβ in mXinα-deficient hearts suggests a partial compensation.
Objective: The essential roles of mXinβ in cardiac development and intercalated disc maturation were investigated.
Methods and Results: Ablation of mXinβ led to abnormal heart shape, ventricular septal defects, severe growth retardation, and postnatal lethality with no upregulation of the mXinα. Postnatal upregulation of mXinβ in wild-type hearts, as well as altered apoptosis and proliferation in mXinβ-null hearts, suggests that mXinβ is required for postnatal heart remodeling. The mXinβ-null hearts exhibited a misorganized myocardium as detected by histological and electron microscopic studies and an impaired diastolic function, as suggested by echocardiography and a delay in switching off the slow skeletal troponin I. Loss of mXinβ resulted in the failure of forming mature intercalated discs and the mislocalization of mXinα and N-cadherin. The mXinβ-null hearts showed upregulation of active Stat3 (signal transducer and activator of transcription 3) and downregulation of the activities of Rac1, insulin-like growth factor 1 receptor, protein kinase B, and extracellular signal-regulated kinases 1 and 2.
Conclusions: These findings identify not only an essential role of mXinβ in the intercalated disc maturation but also mechanisms of mXinβ modulating N-cadherin–mediated adhesion signaling and its crosstalk signaling for postnatal heart growth and animal survival.
- N-cadherin–mediated adhesion signaling
- Xin repeat-containing protein
- intercalated disc maturation
- diastolic dysfunction
- postnatal heart growth
A regulatory network of transcription factors is known to control cardiac morphogenesis. Although the core players in this network are highly conserved, from organisms with simple heart-like cells to those with complex four-chambered hearts, it has been theorized and proven that expansion of this regulatory network by adding new transcription factors is a major force for the heart to evolve new structures.1,2 However, the addition of new transcription factors can only be a part of the mechanism underlying the formation of complex hearts. The transcription factors must act through their downstream targets, which are directly involved in cardiac morphogenesis, growth and function. However, our inventory of such downstream targets remains incomplete.
The Xin repeat–containing proteins from chicken and mouse hearts (cXin and mXinα, respectively) were first identified as a target of the Nkx2.5-Mef2C pathway.3,4 Another mouse Xin protein, mXinβ (or myomaxin), has been subsequently identified as a Mef2A downstream target.5 Evolutionary studies suggest that Xin may be one of the factors that arose when the heart evolved from simple heart-like cells to the complex true-chambered hearts.6 Functional studies reveal that Xin proteins are involved in heart chamber formation and cardiac function in vertebrates.3,4,7 The striated muscle-specific Xin family of proteins are defined by the presence of 15 to 28 copies of the conserved 16-aa Xin repeats and originated just before the emergence of lamprey, coinciding with the appearance of the true-chambered heart.6 The Xin repeats are responsible for binding actin filaments,8–10 whereas a highly conserved β-catenin binding domain overlapping with the Xin repeats is responsible for localizing Xin to the intercalated discs.6,9 Supporting the roles of Xin in heart chamber formation and function, we have previously shown that knocking down the sole cXin in the chicken embryo collapses the wall of heart chambers and leads to abnormal cardiac morphogenesis.3
In mammals, however, a pair of paralogous Xinα and Xinβ genes exists. Ablation of the mouse mXinα gene does not affect heart development. Instead, the mXinα-deficient mice show cardiac hypertrophy and cardiomyopathy with conduction defects during adulthood. In the mXinα-deficient mice, mXinβ is upregulated at both message and protein levels, suggesting a compensatory role of mXinβ.7 Consistent with this idea, both mXin proteins have highly conserved Xin repeats and β-catenin–binding domain, as well as other functionally undefined domains located in the N termini.6 On the other hand, the C termini of both proteins are more diverged, suggesting that they also have distinct functions.6 Because mXinβ is more conserved than mXinα with the ancestral lamprey Xin that demarked the emergence of heart chamber, we hypothesized that mXinβ might play an essential role in heart morphogenesis. To test this hypothesis, we generated and characterized mXinβ knockout mice. The mXinβ-null mice died before weaning and showed abnormal heart shape, ventricular septal defects (VSDs), misorganized myocardium, and diastolic dysfunction. The mechanisms underlying these cardiac defects involved dysregulation of the N-cadherin–mediated signaling pathway and its crosstalk via abnormally activated Stat3 (signal transducer and activator of transcription 3) and depressed Rac1, insulin-like growth factor (IGF)-1 receptor (IGF-1R), protein kinase B (Akt), and extracellular signal-regulated kinase (Erk)1/2 activities.
All animal procedures were approved and performed in accordance with institutional guidelines. The mXinβ-null line has been backcrossed to and maintained in C57BL/6J. All of phenotypes observed earlier remained the same.
Cloning of mXinβ cDNAs and genomic fragments; construction of a targeting vector; Southern, Northern, and Western blot analyses; tissue collection; DNA and RNA isolations; histological staining and immunofluorescence; electron microscopy; echocardiography; proliferation; apoptosis; Rho GTPase assays; and data analysis are described in the Online Data Supplement, available at http:// circres.ahajournals.org.
Generation of mXinβ-Null Mice
To construct a targeting vector, we cloned full-length mXinβ cDNAs and the corresponding genomic fragments. Alignment of these sequences revealed that the mXinβ gene contains 9 exons and encodes 3 mRNA species (mXinβ-A, mXinβ-B, and mXinβ-C) in adult heart through alternative splicing of exon 8 and alternative usage of polyA signals (Online Figure I, A). Both mXinβ-A and mXinβ-B encode a polypeptide of 3283-aa residues (termed mXinβ), whereas mXinβ-C is predicted to encode a protein of 3300-aa residues (termed mXinβ-a) (Online Figure I, B). By sequencing 24 randomly picked transformants generated from 3′ rapid amplification of cDNA ends, we found 23 clones representing either mXinβ-A or mXinβ-B, suggesting that mXinβ is the major isoform. Force expression of the cloned mXinβ-B cDNAs in CHO cells confirmed that mXinβ-B encoded the protein having the same size as endogenous mXinβ and reacting to anti-mXin antibody (Online Figure I, C). Furthermore, force-expressed mXinβ colocalized with actin filaments to stress fibers and cell cortex (Online Figure I, D).
Using multiple tissue Western blot, mXinβ was detected only in the striated muscles such as tongue, heart, and diaphragm (Online Figure II, A). During postnatal heart development, the expression of mXinβ increased at least 3-fold from postnatal day (P)0.5 to P13.5 (Online Figure II, C). The timing of this upregulation of mXinβ coincides with the period for intercalated disc maturation.11,12
A targeting vector was designed to delete the genomic region that encodes the highly conserved β-catenin–binding domain and Xin repeats (Figure 1A). After electroporation and selection, resistant embryonic stem clones were screened by Southern blot analysis (Figure 1B). The positive embryonic stem clone was used to generate chimeric founders. After confirming germline transmission, the heterozygous progeny were further crossed to obtain mXinβ-null mice. The genotypes of the resulting littermates were determined with tail DNAs by Southern blot and by PCR genotyping (Figure 1C).
All mXinβ-Null Mice Die Before Weaning
Northern blot analysis revealed a complete loss of mXinβ message in homozygote and a reduction in heterozygote (Figure 1D). Western blot analyses with antibody U10137 recognizing both mXinα and mXinβ (Figure 1E, top blot) or with antibody U1040 recognizing C terminus of mXinβ (data not shown) verified a complete loss of mXinβ in homozygotes and a reduced level in heterozygotes. The mXinβ-null hearts expressed the similar amounts of mXinα-a, mXinα, α-actinin, and α-tropomyosin (α-TM) as their age-matched counterparts (Figure 1E).
At birth (P0.5), the number of the mXinβ-null pups from heterozygous crosses was smaller than the expected number (Online Table I); however, this reduction was not statistically significant (P=0.17; χ2 test). In contrast, from P3.5 and on, the number of viable mXinβ−/− mice was significantly lower than expected. No viable mXinβ−/− mice could be observed at weaning stage. Thus, these observations suggest that mXinβ is essential for postnatal mouse survival.
Loss of mXinβ Leads to Severe Growth Retardation
The mXinβ-null mice had severely retarded growth and reduced activity. The skin of newborn mXinβ−/− mice was apparently paler than their littermates, suggesting a systemic circulation defect. Great vessels in the newborn mXinβ-null mice were normal (Online Figure III). The body weight (BW) of P0.5 mXinβ−/− mice was ≈14.3% lighter than wild-type or heterozygous littermates (Figure 2A). From birth to P12.5, the mXinβ-null mice also gained weight more slowly than their littermates (Figure 2A). At P12.5, mXinβ-null mice weighed only ≈45% of wild-type or heterozygous mice. The loss of just one copy of mXinβ in heterozygotes had neither effect on BW nor on viability. Neonatal mXinβ−/− pups apparently breathed normally, and milk was always visible in their stomach, suggesting that a weakness in skeletal muscles is unlikely to be the major cause for the growth retardation and lethality.
The heart weights (HWs) of newborn wild-type and mXinβ-null pups were similar. However, from P3.5 to P12.5, the wild-type hearts grew much faster than mXinβ-null hearts. As a result, both P7.5 and P12.5 wild-type hearts were significantly larger than mXinβ-null counterparts (Figure 2B), suggesting that mXinβ is required for postnatal heart growth. Similar to its effects on BW, the loss of one copy of mXinβ in heterozygotes did not affect their heart size (data not shown). The HW/BW ratio of mXinβ−/− mice at most of postnatal stage except P3.5 was significantly higher than that of wild-type mice (Figure 2C) because of significantly smaller BW in mXinβ-null mice. Similar to the hearts, non–mXinβ-expressing organ such as liver (Figure 2D) of mXinβ-null mice also became significantly smaller between P3.5 and P7.5. However, the liver weight (LW) to BW (LW/BW) ratios of wild-type and mXinβ-null mice remained no difference (Figure 2E). Thus, the loss of mXinβ affected the mXinβ-expressing and non–mXinβ-expressing organs differently.
Loss of mXinβ Results in VSDs, Abnormal Heart Shape, and Misorganized Myocardium
Approximately 15% (5/33) of newborn mXinβ-null hearts had abnormal shape (Figure 3B). VSD was detected in 58% (7/12) of newborn mXinβ-null hearts analyzed by serial section analysis (arrow in Figure 3D). The VSD could be found in any locations within the muscular septum, and could be small, large or multiple. However, the VSD could not be the cause of postnatal lethality, because 42% of mXinβ-null mice without VSD also became small and weak and died before weaning. Misorganized myocardium (noncompaction in right ventricle) could be detected in mXinβ-null hearts as early as embryonic day 14.5 (Figure 3F and 3F′). Thus, mXinβ-null embryo may already have a defect in heart function, leading to a slight but significant reduction in BW at birth (Figure 2A). However, this defect may not be enough to cause embryonic lethality, because no significant loss of newborn mXinβ-null mice was found (Online Table I). Furthermore, all mXinβ-null neonatal hearts examined showed various degrees of misorganized myofibers within myocardium (an example shown in Figure 3H and 3H"). Electron microscopic (EM) analysis of P15.5 mXinβ-null hearts detected no sarcomere disorganization within each myocytes (Figure 3J), suggesting no myofilament disarray in mutant hearts.
Developing mXinβ-Null Hearts Exhibit Diastolic Dysfunction
Because all mXinβ−/− mice exhibited misorganized myocardium, we next analyzed chamber size, wall thickness, and cardiac function by echocardiography. Because echocardiographic results from wild types and heterozygotes were very similar, we treated them as a control group for the comparison to mXinβ-null group (Online Table II). We observed a reduction in left ventricular internal dimension and volume of both P3.5 and P12.5 mXinβ-null hearts during diastole and systole (Online Table II). In contrast, there was no difference between control and mXinβ-null mice in heart rate, left ventricular posterior wall thickness and interventricular septum thickness (Online Table II). Furthermore, the mXinβ-null hearts had normal or slightly higher systolic function, as determined by the ejection fraction and the fraction shortening (Online Table II). Using pulsed wave Doppler recordings, we found that mXinβ-null hearts exhibited abnormal ventricular filling. In mXinβ−/− mice, the mitral inflow E-wave (early filling) but not A-wave (atrial contraction) peak velocity was reduced (Figure 4), and the E/A ratios were also significantly smaller (Online Table II). These results suggest a diastolic dysfunction in mutant hearts as early as P3.5. However, this diastolic dysfunction was not attributable to increased fibrosis that could stiffen the myocardium, because trichrome staining detected no increase in fibrosis at P11.5 (Online Figure IV).
Developmental changes in ventricular diastolic function correlate well with changes in myoarchitecture (compact versus trabecular areas in ventricles).13 In general, the peak E-wave velocity is exponentially correlated with the area of compact region per unit myocardium, whereas the peak A-wave velocity is correlated with the area of trabecular region per unit myocardium. Using similar measurement in newborn mXinβ-null mice, we found a significant reduction in the area of left ventricle compact myocardium and a trend of increase in the area of left ventricle trabecular myocardium in mutant hearts (Online Table III). Similar trends of decrease in compact area and increase in trabecular area were also observed for right ventricle (Online Table III). These results again support diastolic dysfunction associated with newborn mXinβ-null hearts.
The Delay in Switching off Slow Skeletal Troponin I Also Supports Diastolic Dysfunction Associated With mXinβ-Null Mice
Apparent preservation of systolic function and presence of diastolic dysfunction in mXinβ-null hearts led us to examine the expression levels and isoform switches of contractile and regulatory proteins. The observations of normal expression levels of α-actinin and α-TM (Figure 1E), as well as normal timing of switching from β-myosin heavy chain to α-myosin heavy chain (Online Figure V) and from embryonic cardiac troponin (cTn)T to adult cTnT (Figure 5C) largely support that mXinβ-null hearts having normal systolic function. In contrast, a significant delay in switching off slow skeletal troponin (ssTn)I was detected in P7.5 and P13.5 mXinβ-null hearts (Figure 5B). This delay may allow mutant hearts to gain increased Ca2+-activated myofilament tension to compensate for function, as suggested from a previous study comparing force generations between ssTnI- and cardiac troponin I (cTnI)-expressing cardiomyocytes.14 Nonetheless, transgenic mice ectopically expressing ssTnI in the heart exhibit impairments of cardiomyocyte relaxation and diastolic function.15 Together, the delay in switching off ssTnI also supports diastolic dysfunction in mXinβ-null heart. It should be noted that mXinβ-null hearts did not upregulate N-terminal truncated cTnI (Figure 5B), which has been previously shown to enhance ventricular diastolic function in transgenic mice.16
Developing mXinβ-Null Hearts Exhibit an Increased Apoptosis and a Decreased Proliferation
Apoptosis and proliferation contribute greatly to myocardial remodeling during postnatal development.17 Thus, we asked whether defects in these processes might contribute to the misorganization of mutant myocardium. The wild-type hearts had high apoptosis only at P0.5, as detected by anti–active caspase 3, which then rapidly declined to a minimal level at P7.5, similar to that of adult heart (Figure 6A, 6A′, and 6E).17 In contrast, the level of apoptosis in P0.5 mXinβ-null hearts decreased more slowly and remained significantly higher at P3.5 and P7.5 (Figure 6B, 6B′, and 6E). Using bromodeoxyuridine (BrdUrd) labeling, we found that there was no difference in proliferation rate in mXinβ-null and control hearts until P7.5, at which mutant hearts showed slightly reduced cell proliferation (Figure 6C, 6D, and 6F). Therefore, slightly decreased proliferation and increased apoptosis in mXinβ-null hearts postnatally may in part account for the smaller HW and the misorganized myocardium.
Cardiomyocyte organization was compared from cross-sections of individual cardiomyocytes of similar regions of littermate hearts. The cTnT-positive cardiomyocytes were outlined by anti-laminin for shape and width comparison. At P3.5, there was no detectable difference between wild-type and mXinβ-null cardiomyocytes in either cell shape or cell width (Online Figure VI, A, B, and E). In contrast, by P12.5, mXinβ−/− cardiomyocytes became more irregularly shaped (Online Figure VI, D) and smaller in cell width (Online Figure VI, E).
mXinβ-Null Hearts Fail to Develop Mature Intercalated Discs
At the first 2 weeks of age, mXinα, N-cadherin, and β-catenin progressively coalesce to the termini of aligned cardiomyocytes to form mature intercalated discs.12 We asked whether mXinβ plays a role in the intercalated disc maturation. In P16.5 wild-type hearts, the majority of mXinβ, N-cadherin, and mXinα (Figure 7A, 7C, and 7E), as well as β-catenin and p120-catenin (data not shown), was already localized to the mature intercalated discs. In contrast, most N-cadherin (Figure 7D) and β-catenin (data not shown) found in the P16.5 mXinβ-null hearts remained as small puncta along the lateral contacts of cardiomyocytes, whereas p120-catenin (data not shown) and mXinα (Figure 7F) puncta became dispersed throughout the cardiomyocytes. These results suggest that mXinβ is essential for promoting and maintaining the localization of adherens junctional components and mXinα to the mature intercalated discs.
The wild-type and mXinβ-null mice from newborn to 2 to 3 weeks of age appeared to express comparable amounts of N-cadherin and connexin 43 (Online Figure V; Online Figure VII, A; and some data not shown). Both N-cadherin and connexin 43 continued to accumulate to the myocyte termini of the wild-type mice from P15.5 to P18.5, whereas the terminal distribution of both molecules remained unchanged in mutants (Online Figure VII, B), again suggesting a defect in the maturation of intercalated discs. EM analysis revealed that the developing intercalated disc at the cell termini of P15.5 mXinβ-null hearts was smaller than the wild-type counterparts (arrows in Figure 7G and 7H). At higher magnification, the membranes at the maturing intercalated disc of mXinβ-null cardiomyocytes were less convoluted and less wavy (Figure 7J), suggesting a depressed membrane activity at the termini of mXinβ-null cells. At the lateral membrane contacts, developing T-tubules could be detected in both mutant and control cells (asterisk in Figure 7G and 7H), and less difference in the membrane activity was observed.
The mXinβ-Null Hearts Increased Stat3 Activity but Decreased Rac1, IGF-1R, Akt, and Erk1/2 Activities
Accumulated lines of evidence suggest that N-cadherin–mediated adhesion signaling is critical for intercalated disc integrity and cardiac function.18 Cadherin and its associated catenins are also known to interact with many signaling molecules, providing the ability to cross-talk with other signaling pathways such as receptor tyrosine kinase–, cytokine receptor–, and G protein coupled receptor–mediated signaling. We asked whether impairing intercalated disc maturation by the loss of mXinβ could lead to abnormal activities of Rho GTPase, Stat, protein kinase B (Akt), and Erk, important effectors in relaying signaling for postnatal heart development.
Using GST-Pak PBD and GST-Rhotekin RBD beads to pull-down active forms of Rac1 and RhoA, respectively, we found that relative GTP-bound Rac1 in P7.5 mXinβ-null hearts was reduced to ≈65% of the control, whereas the active RhoA level in mXinβ-null hearts did not change significantly (Figure 8A). A reduction of Rac1 activity may result in less dynamic membranes at the termini of mutant cells, which was indeed suggested by the EM observation. Using phospho-specific antibodies to assess the activation of key signaling molecules, we found an increased Stat3 activity, as suggested by increased level (Figure 8B) and nuclear localization (data not shown) of p-Stat3(Y705) (tyrosine-phosphorylated Stat3 at no. 705), persistently in mXinβ-null hearts starting from P0.5. This Stat3 activation was not correlated to the activation of Jak2 (Janus kinase 2) (one member of nonreceptor tyrosine-protein kinases upstream of Stat3) (Figure 8B), suggesting that other Jaks and/or c-Src may be involved in the activation of Stat3. Alternatively, defects in negative regulators of Stat3, such as SOCS3 (suppressor of cytokine signaling 3) or tyrosine phosphatases, may participate in the abnormal activation of Stat3 in mutant hearts. Moreover, the activations/phosphorylations of Akt, glycogen synthase kinase 3β (a downstream target of Akt), Erk1/2, and IGF-1R were significantly depressed in mutant hearts starting from P7.5, whereas the total proteins of Akt and Grb2 (growth factor receptor-bound protein 2) in mutant and control hearts remained the same (Figure 8B). The persistent activation of Stat3, although not 100% penetrant, precedes the reductions in the activations of growth-related signaling molecules.
In this study, we demonstrate that an intercalated disc–associated and Xin repeat–containing protein, mXinβ, is required for postnatal heart development. First, the postnatal upregulation of mXinβ coincides with the maturations of the intercalated disc,11,12 T-tubule, and sarcoplasmic reticulum,19 as well as diastolic function.20 Second, ablation of mXinβ leads to abnormal heart shape, VSD, diastolic dysfunction, severe growth retardation, and postnatal lethality. Third, loss of mXinβ results in failure of forming mature intercalated disc. Our data further identify that the proper clustering of N-cadherin to form intercalated disc regulates the Stat3 activity and activates the Rac1, IGF-1R, Akt, and Erk1/2 activities, which are required for postnatal heart growth/hypertrophy.21–23
How Does the Intercalated Disc Mature?
Postnatal maturation of intercalated discs is characterized by gradual clustering of N-cadherin complexes/puncta from lateral localization to termini of aligned cardiomyocytes. Such a clustering process likely involves modulating the interaction between cadherins and underlining actin cytoskeleton. In a classic view, the actin bundling protein α-catenin binds β-catenin to organize the adhesion complex that links to actin cytoskeleton.24 However, this stable linkage role for α-catenin has not been proven; instead, compelling evidence suggests α-catenin as a molecular switch that modulates actin cytoskeleton.25 Consistent with this notion, 2 types of cadherin-mediated intercellular contacts are recently detected in the adherens junctions of epithelia: a mobile and α-catenin–dependent contact associated with a dynamic actin network, as well as a stable and α-catenin–independent contact associated with a stable actin patch.26 The existence of this stable contact suggests that an unidentified X protein has to link the cadherin/catenin complex to actin patches. In the heart, the role of this unidentified X protein may be served by the Xin repeat–containing proteins. We propose that developmental upregulation and functional hierarchy of mXinβ initiate the formation of mature intercalated discs. The mXinα further reinforces the stability of intercalated discs. In support of this role, loss of mXinβ leads to failure of forming mature intercalated discs and mislocalizations of mXinα and N-cadherin. On the other hand, mature intercalated discs form normally in the mXinα-null heart (Online Figure VIII) but eventually lose close membrane apposition between cardiomyocytes at young adult. This structural defect progressively worsens by older age.7
Diastolic Dysfunction May Be Responsible for Heart Failure and Lethality in mXinβ-Null Mice
The mXinβ-null hearts have normal systolic function and heart rate but exhibit a significant delay in switching off ssTnI and significant reductions in mitral early filling (E-wave) peak velocity and E/A ratio, suggesting diastolic dysfunction. Impaired diastolic function was also suggested by the left ventricle internal dimension and left ventricle volumes being smaller in mXinβ-null mice. The detection of a significant reduction in the compact areas of ventricles in newborn mutant hearts (Online Table III) further supported a reduction in E-wave velocity.13 The diastolic dysfunction would lead to diminished cardiac output (stroke volume×heart rate) of mutant hearts and could contribute in part to heart failure and postnatal lethality. The mXinβ-null cardiomyocytes after P15.5 exhibited a significant reduction in the terminal connexin 43 localization (Online Figure VII), which may cause arrhythmic sudden death. However, this spatial connexin 43 alteration cannot be the cause for the loss of mXinβ-null mice at earlier age (Online Table I).
mXinβ Regulates Postnatal Cardiac Growth
In the heart, the Rac1 activation is essential for rearranging cytoskeleton to align cardiomyocytes23 and for regulating mitogen-activated protein kinases21 and NADPH oxidase activity22 for cardiac hypertrophy. Moreover, transgenic mice expressing constitutively active Rac1 in the heart develop dilated myocardium with high postnatal mortality.27 Most transgenic mice die within 2 to 3 weeks after birth, suggesting that postnatal heart development requires an intricate regulation of Rac1 activity. It is also known that classic cadherin engagement activates Rac1 through c-Src–phosphatidylinositol 3-kinase–Vav2, and Vav2 is a guanine nucleotide exchange factor capable of binding to p120-catenin.28,29 The loss of mXinβ may dysregulate this signaling, leading to a downregulation of Rac1 activity and forming less convoluted, less wavy, and less stable intercalated discs. The loss of mXinβ may also dysregulate cytokine/angiotensin II/growth hormone-mediated signaling, leading to a persistent activation of Stat3 (Online Figure IX). The activation of Stat can promote IGF-1 production,30 which would facilitate postnatal heart growth. However, the lack of mature intercalated discs in mutant hearts reduced the activities of IGF-1R, Akt, and Erk1/2, resulting in severely retarded growth.
In summary, we have identified that mXinβ, as a critical component for the intercalated disc maturation, is essential for postnatal heart development. Our findings provide the first insights into its function of transducing the N-cadherin–mediated adhesion and crosstalk signaling by regulating the activities of Stat3, Rac1, Erk1/2, and Akt. Ablation of mXinβ leads to VSDs, cardiac diastolic dysfunction, and severe growth retardation. The human ortholog, cardiomyopathy-associated 3 (CMYA3), of mXinβ is mapped to 2q24.3. Human patients with chromosome band 2q24 deletion also exhibit severe growth retardation and VSDs (http://www.orpha.net/data/patho/GB/uk-2q24.pdf). The genome-wide linkage analysis of a large Kyrgyz family also reveals candidate genes on 2q24.3-q31.1 conferring susceptibility to premature hypertension.31 Further studies are warranted to characterize the involvement of mXinβ in cardiac development, function, and disease.
We thank Keyu Chen for excellent assistance.
Sources of Funding
This work was supported, in part, by NIH grants HL075015 (to J.J.-C.L.), HL088883 (to T.D.S.), and HL078773 (to J.-P.J.) and by National Science Council (Taiwan) grant NSC98-2320-B015-010-MY3 (to C.-I.L.).
Olson EN. Gene regulatory networks in the evolution and development of the heart. Science. 2006; 313: 1922–1927.
Wang DZ, Reiter RS, Lin JL, Wang Q, Williams HS, Krob SL, Schultheiss TM, Evans S, Lin JJ. Requirement of a novel gene, Xin, in cardiac morphogenesis. Development. 1999; 126: 1281–1294.
Huang HT, Brand OM, Mathew M, Ignatiou C, Ewen EP, McCalmon SA, Naya FJ. Myomaxin is a novel transcriptional target of MEF2A that encodes a Xin-related alpha-actinin-interacting protein. J Biol Chem. 2006; 281: 39370–39379.
Gustafson-Wagner EA, Sinn HW, Chen YL, Wang DZ, Reiter RS, Lin JL, Yang B, Williamson RA, Chen J, Lin CI, Lin JJ. Loss of mXinα, an intercalated disk protein, results in cardiac hypertrophy and cardiomyopathy with conduction defects. Am J Physiol Heart Circ Physiol. 2007; 293: H2680–H2692.
Pacholsky D, Vakeel P, Himmel M, Lowe T, Stradal T, Rottner K, Furst DO, van der Ven PFM. Xin repeats define a novel actin-binding motif. J Cell Sci. 2004; 117: 5257–5268.
Choi S, Gustafson-Wagner EA, Wang Q, Harlan SM, Sinn HW, Lin JL, Lin JJ. The intercalated disc protein, mXinα, is capable of interacting with β-catenin and bundling actin filaments. J Biol Chem. 2007; 282: 36024–36036.
Ishiwata T, Nakazawa M, Pu WT, Tevosian SG, Izumo S. Developmental changes in ventricular diastolic function correlate with changes in ventricular myoarchitecture in normal mouse embryos. Circ Res. 2003; 93: 857–865.
Westfall MV, Metzger JM. Troponin I isoforms and chimeras: tuning the molecular switch of cardiac contraction. News Physiol Sci. 2001; 16: 278–281.
Barbato JC, Huang QQ, Hossain MM, Bond M, Jin JP. Proteolytic N-terminal truncation of cardiac troponin I enhances ventricular diastolic function. J Biol Chem. 2005; 280: 6602–6609.
Li J, Patel VV, Radice GL. Dysregulation of cell adhesion proteins and cardiac arrhythmogenesis. Clin Med Res. 2006; 4: 42–52.
Hirakow R, Gotoh T. Ontogenetic implication of the myocardial ultrastructure in the development of mammalian heart. In: van Praagh R, Takao A, eds. Etiology and Morphogenesis of Congenital Heart Disease. Mount Kisco, NY: Futura Publishing Co; 1980: 99–108.
Zhou YQ, Foster FS, Parkes R, Adamson SL. Developmental changes in left and right ventricular diastolic filling patterns in mice. Am J Physiol Heart Circ Physiol. 2003; 285: H1563–H1575.
Clerk A, Pham FH, Fuller SJ, Sahai E, Aktories K, Marais R, Marshall C, Sugden PH. Regulation of mitogen-activated protein kinases in cardiac myocytes through the small G protein Rac1. Mol Cell Biol. 2001; 21: 1173–1184.
Satoh M, Ogita H, Takeshita K, Mukai Y, Kwiatkowski DJ, Liao JK. Requirement of Rac1 in the development of cardiac hypertrophy. Proc Natl Acad Sci U S A. 2006; 103: 7432–7437.
Pokutta S, Weis WI. The cytoplasmic face of cell contact sites. Curr Opin Struc Biol. 2002; 23: 255–262.
Noren NK, Liu BP, Burridge K, Kreft B. p120 catenin regulates the actin cytoskeleton via Rho family GTPases. J Cell Biol. 2000; 150: 567–580.
Honsho S, Nishikawa S, Amano K, Zen K, Adachi Y, Kishita E, Matsui A, Katsume A, Yamaguchi S, Nishikawa K, Isoda K, Riches DW, Matoba S, Okigaki M, Matsubara H. Pressure-mediated hypertrophy and mechanical stretch induces IL-1 release and subsequent IGF-1 generation to maintain compensative hypertrophy by affecting Akt and JNK pathways. Circ Res. 2009; 105: 1149–1158.
Kalmyrzaev B, Aldashev A, Khalmatov M, Polupanov A, Jumagulova A, Mamanova L, Wilkins MR, Town M. Genome-wide scan for premature hypertension supports linkage to chromosome 2 in a large Kyrgyz family. Hypertension. 2006; 48: 908–913.
Novelty and Significance
What Is Known?
The origin of Xin repeat–containing proteins coincides with the genesis of true heart chambers.
The mouse Xin (mXinα and mXinβ) proteins localize to the intercalated disc of the heart, and their human orthologs, CMYA1 and CMYA3, coexpress with the 13 known cardiomyopathy-associated genes, serving as potential diagnostic markers and drug targets for cardiac diseases.
Mice lacking mXinα upregulate mXinβ, develop adult-onset cardiac hypertrophy and cardiomyopathy with conduction defects, and attenuate the induction of atrial fibrillation in their left atrial-pulmonary vein tissues.
What New Information Does This Article Contribute?
Developmental upregulation of mXinβ coincides with intercalated disc and T-tubule maturation during postnatal cardiac remodeling.
Complete loss of mXinβ in mice leads to misorganized myocardium, abnormal heart shape, ventricular septal defect, cardiac diastolic dysfunction, severe growth retardation, and postnatal lethality.
The mXinβ in a functional hierarchy plays essential roles for the terminal localization of mXinα, N-cadherin, and connexin 43 and for mediating N-cadherin and its crosstalk signaling pathways for postnatal heart growth and animal survival.
Early evolutionary and functional studies reveal the critical requirement of Xin repeat–containing proteins in cardiac chamber formation and cardiac function. Here, we report for the first time that mice lacking mXinβ exhibit abnormal heart shape, ventricular septal defect, and misorganized myocardium. Impaired left ventricular relaxation and filling, as well as smaller diastolic volumes, may be responsible for severe growth retardation and premature death of all mutant mice. We also demonstrate for the first time that postnatal increase in the expression of mXinβ is required for the intercalated disc maturation. Mechanistically, mXinβ is involved in N-cadherin–mediated signaling and its crosstalk signaling pathways that are essential for intercalated disc formation and postnatal heart growth. Importantly, postnatal growth retardation, ventricular septal defects, progressive heart failure, and lethality have been reported in human infants missing chromosome band 2q24.3, which contains the human ortholog (CMYA3) of mXinβ. A genome-wide association analysis has also revealed candidate genes near 2q24.3 for premature hypertension. Moreover, the CMYA3 coexpresses with many known cardiomyopathy-associated genes and may serve as a useful diagnostic marker and therapeutic target for cardiac diseases. Thus, the mXinβ-deficient mice represent a novel model for studying heart development and diseases.
Original received November 8, 2009; revision received March 16, 2010; accepted March 18, 2010.