Deficiency of Myo18B in mice results in embryonic lethality with cardiac myofibrillar aberrations
Communicated by: Fuyuki Ishikawa
Abstract
Myo18B is an unconventional myosin family protein expressed predominantly in muscle cells. Although conventional myosins are known to be localized on the A-bands and function as a molecular motor for muscle contraction, Myo18B protein was localized on the Z-lines of myofibrils in striated muscles. Like Myo18A, another 18th class of myosin, the N-terminal unique domain of the protein and not the motor domain and the coiled-coil tail is critical for its localization to F-actin in myocytes. Myo18B expression was induced by myogenic differentiation through the binding of myocyte-specific enhancer factor-2 to its promoter. Deficiency of Myo18B caused an embryonic lethality in mice accompanied by disruption of myofibrillar structures in cardiac myocytes at embryonic day 10.5. Thus, Myo18B is a unique unconventional myosin that is predominantly expressed in myocytes and whose expression is essential for the development and/or maintenance of myofibrillar structure.
Introduction
Myosins constitute a large superfamily of actin-based motor proteins that use ATP to play fundamental roles in many forms of eukaryotic cell motility, such as cell migration, cytokinesis, phagocytosis, and trafficking. A total of 40 myosin genes have been identified in humans to date; however, physiological and biological functions of more than half of their products are still unknown. (Mermall et al. 1998; Wu et al. 2000; Berg et al. 2001). We previously isolated a novel unconventional myosin, MYO18B, from a homozygously deleted region at chromosome 22q12.1 in a human lung cancer cell line (Nishioka et al. 2002). Inactivation of this gene by genetic and/or epigenetic alterations in human cancer cells suggests that the MYO18B gene functions as a tumor suppressor in human carcinogenesis (Nishioka et al. 2002; Tani et al. 2004; Yanaihara et al. 2004; Nakano et al. 2005). Like other myosins, MYO18B has a myosin head motor domain with an ATP-binding motif and a GPA actin-binding motif. It also contains a neck region with an IQ motif that is thought to participate in binding to myosin light chains or calmodulin and a C-terminal tail with a coiled-coil domain that likely mediates dimer formation. Furthermore, MYO18B has unique N- and C-terminal extensions (Fig. 1A), which should give a distinction against other classes of myosins. However, physiological functions of MYO18B are largely unknown. Here, we showed the N-terminal unique domain of MYO18B is crucial for its localization to F-actin, and analysis of Myo18B gene targeted mice revealed that Myo18B protein is crucial for the development and/or maintenance of myofibrillar structure in myocytes.
Intracellular localization of endogenous and deletion mutants of Myo18B protein in C2C12 cells. (A) Structure of MYO18B protein, positions of antigens for antibodies and EGFP-fusion protein structures used below are shown. (B) Western blot analysis of C2C12 cells with induction of myogenic differentiation using antibodies, anti-MYO18B-N1 (upper panel), anti-MYO18B-C2 (upper middle panel), anti-Myogenin (lower middle panel), and anti-α tubulin (lower panel). Confocal images of immunocytochemistry of differentiated C2C12 cells (C–J) using anti-MYO18B-N1 (C and green in F) and anti-MYO18B-C2 (G and green in J) are shown. Arrows in F and J indicate differentiated and multinucleated C2C12 cells. Localization of EGFP-fusion deletion mutants of MYO18B is shown (K–Z). MYO18B-EGFP (K–N), Nter-EGFP (O–R), EGFP-moter (S–V) and EGFP-ΔNter (W–Z) were expressed in C2C12 cells. The cells were co-stained with phalloidin (D, H, L, P, T, X and red in F, J, N, R, V, Z) and TOTO-3 (E, I, M, Q, U, Y and blue in F, J, N, R, V, Z). Bars = 50 µm (F), 20 µm (J, N, R, V, Z).
Results
Contribution of N-terminal extension of Myo18B protein to its localization on F-actin
To characterize the functions of Myo18B, we used C2C12 cells because the Myo18B transcription is increased by myogenic differentiation in C2C12 mouse myoblast cells (Salamon et al. 2003). Western blot analysis reveals that endogenous Myo18B protein level increases along with C2C12 cells differentiation (Fig. 1B). Then, we carried out immunofluorescence analysis of C2C12 cells using two independent antibodies against Myo18B. Endogenous Myo18B protein stained very faintly in undifferentiated C2C12 cells, but better in multinucleated differentiated C2C12 cells that were positive for myoglobin and skeletal myosin (data not shown). This result is consistent with the result of Western blot analysis. In differentiated C2C12 cells, Myo18B was predominantly distributed on F-actin with a punctate pattern (Fig. 1C–J). Previously Myo18B protein was reported to be localized in the nucleus of differentiated C2C12 cells and striated muscles (Salamon et al. 2003); however, the antibodies did not indicate a nuclear distribution of Myo18B protein in both differentiated C2C12 cells and striated muscles.
To reveal which domain of Myo18B contributes to the localization on F-actin, we prepared expression vectors for EGFP-fused MYO18B full-length (MYO18B-EGFP) protein and for deletion mutants (Fig. 1A). Exogenously expressed MYO18B-EGFP in C2C12 cells was predominantly distributed on actin stress fibers as endogenous Myo18B, and localized in cytoplasm with a punctate pattern (Fig. 1K–N). When cells had a protruded region, which is a region where F-actin is actively polymerized, MYO18B-EGFP co-localized to these sites as well (data not shown) consistent with results shown previously in NIH3T3 cells (Ajima et al. 2007). Analysis of an EGFP-fused N-terminal half (1–1357 a.a.) and C-terminal half (1357–2567 a.a.) of MYO18B protein revealed that the N-terminal half of MYO18B is necessary, but dimerization with a coiled-coil domain in the C-terminal half of MYO18B, is not necessary for localization on F-actin (data not shown). The N-terminal half of MYO18B was further divided into the N-terminal extension (Nter-EGFP) and the motor region (EGFP-moter) and both were expressed in C2C12 cells. EGFP-moter was homogenously distributed throughout the cytoplasm (Fig. 1S–V), whereas Nter-EGFP was localized clearly on F-actin (Fig. 1O–R). A vector of MYO18B without the N-terminal extension (EGFP-ΔNter) was also expressed. EGFP-ΔNter was localized in the cytoplasm with a punctate pattern as MYO18B-EGFP, but its localization on actin stress fibers was unclear (Fig. 1W–Z). Taking these results together, it was concluded that the N-terminal extension of MYO18B protein greatly contributes to the localization of MYO18B on actin stress fibers.
Intracellular localization of Myo18B protein in striated muscle cells
To reveal the localization of Myo18B protein in skeletal and cardiac muscles, we carried out immunohistochemistry using anti-MYO18B antibodies. Myo18B showed a stripe pattern in muscle cells, suggesting that Myo18B is localized in either the A-bands or the Z-lines. The cells were co-stained with an α-actinin antibody, a marker of the Z-lines. α-actinin showed a stripe pattern in the same region, but the bands for α-actinin were thinner and sharper than those for Myo18B. The result indicated that Myo18B is localized in the Z-lines and in the region adjacent to the Z-lines in both skeletal and cardiac muscles (Fig. 2A,B,C and D,E,F). Co-staining of Myo18B with type II myosin showed that Myo18B did not co-localize with type II myosin in adult skeletal muscles (Fig. 2G,G′,H,H′,I,I′) or in cardiac muscles of E10.5 wild-type embryos, either (Fig. 2J,K,L). A confocal X-Z scan image confirmed that Myo18B was localized in myofibrils between the A-bands (Fig. 2G′,H′,I′). These results suggest that Myo18B is not used as a molecular motor in the A-bands for muscle contraction and plays a distinct role from conventional myosins.
Intracellular localization of endogenous Myo18B protein in striated muscles. Confocal images of frozen sections of skeletal muscle (A, B, C and G, G′, H, H′, I, I′), cardiac muscle (D, E, F) of adult mice and sections of paraffin embedded cardiac muscle of an E10.5 embryo (J, K, L) that are immunohistochemically stained with anti-MYO18B-N1 (A, D, G, G′, J and green in C, F, I, I′, L), anti-α-actinin (B, E and red in C, F), anti-skeletal myosin (H, H′ and red in I, I′) and anti-myosin heavy chain β (K and red in L) antibodies are shown. The localization of Myo18B is indicated by arrows and that of the A-bands is indicated by triangles (G–L). X–Z scan at the doted line in G, H, I, along the longitudinal axes of myocytes are shown in G′, H′, I′. Bars = 5 µm.
Regulation of Myo18B expression in myogenic differentiation of C2C12 mouse myoblast cells
Because Myo18B protein is induced by myogenic differentiation, Myo18B gene expression could be strictly regulated by myogenic transcriptional factors. To analyze the transcriptional regulation of the Myo18B gene, we compared the genomic structure in the 5′ region of the human MYO18B and mouse Myo18B genes using the ClustalW multiple sequence alignment program (http://align.genome.jp/). The nucleotide sequence between the –350 nucleotide and a putative transcription start site (+1) of the human gene was highly conserved in the mouse gene. Furthermore, we screened repetitive elements upstream of the promoter region of the Myo18B gene using the RepeatMasker program (http://www.repeatmasker.org). Repetitive elements started roughly 500 nucleotides upstream from the putative transcription start site (Fig. 3A). Therefore, to define the critical elements for regulation of Myo18B mRNA expression in myocytes and during their differentiation, a series of MYO18B promoter deletion constructs were cloned with reporter vectors for the luciferase assay (Fig. 3A). These vectors were introduced into C2C12 cells and the differentiation was induced. The MYO18B-456 promoter activity in differentiated C2C12 cells was 2–3 times higher than that in undifferentiated cells. The induction of promoter activity was kept with MYO18B-310 and MYO18B-196 constructs, but was decreased with the MYO18B-131 construct and diminished with MYO18B-105 or –69 constructs (Fig. 3B). So, the segments of nucleotides –196 to –131 and –131 to –105 of the MYO18B gene were suggested to bear an enhancer element for myocyte differentiation. A putative AP2-response element was present between nucleotides –196 and –131, and an AT-rich putative MEF2-response element was between nucleotides –131 and –105, and both element were conserved in human and mouse (Fig. 3A). Therefore, reporter vectors with a mutation in each putative binding site of the MYO18B promoter were further constructed. A reporter construct with a mutation in the putative AP2-response element showed similar promoter activity as MYO18B-456 promoter. In contrast, a reporter construct with a mutation in the putative MEF2-response element showed slight induction by myogenic differentiation (Fig. 3C). These results suggest that MEF2 plays an important role in the induction of Myo18B expression in myocytes.
Regulation of Myo18B expression during differentiation of C2C12 myoblasts. (A) Schematic representation of the MYO18B gene promoter region and deletion constructs. The positions of transcriptional factor-response elements that are conserved in humans and mice and the position of the transcriptional start site (+1) are shown above a wild type MYO18B promoter construct, MYO18B-456. (B and C) Fold luciferase activities of several MYO18B promoter constructs compare with undifferentiated and differentiated C2C12 cells. The luciferase activities of MYO18B-456, six promoter deletion constructs (–310, –196, –131, –105, and –69 in B), and mutant constructs for AP2 and MEF2 responsible elements (mAP2 and mMEF2, respectively in C) in C2C12 cells were determined by transfecting the cells with each construct. Fold activities were calculated by dividing differentiated C2C12 cells activity with undifferentiated C2C12 cells activity. Constructs were used more than 3 times and showed similar results.
Generation of Myo18B gene targeting mice and embryonic lethality of homozygous knockouts
A targeting vector was constructed by insertion of the bacterial nls-LacZ and NeoR genes into exon 2 containing the translation initiation codon of the mouse Myo18B gene (Fig. 4A). Two independent Myo18+/–ES cell clones, 109 and 153, were established, used to make Myo18 mutant chimeric and heterozygous mice. Targeting disruption of the gene in ES clones and heterozygous mice was confirmed by PCR and Southern blot analyses (Fig. 4B). The mice appeared healthy and fertile, with no indication of physical abnormalities.
Targeted disruption of the mouse Myo18B gene and its expression. (A) Targeting strategy with the targeting vector (top), a restriction map of the wild-type Myo18B allele covering exon 2 to exon 4 (middle), and the targeted allele disrupted by a homologous recombination (bottom) are shown. ATG, SA-F + SA-R; PCR primer sets for ES screening and genotyping for short arm, F + R; PCR primer sets for genotyping wild-type allele, and F + LacZ; PCR primer sets for genotyping targeted allele, P; probe for Southern blot analysis, and 15-kb/13-kb of the expected fragment lengths are indicated. E, EcoRI; S, SphI; nls-LacZ; nuclear localization signal-β-galactosidase gene, NeoR; neomycin resistant gene, DTA; Diphtheria toxin A gene. Arrows above gene cassettes indicate the direction of the genes. (B) Representative results of PCR and Southern blot analyses for genotyping of ES cells and F1 mice are shown. The 2.2-kb PCR fragments represent a disrupted allele (upper), and 15 kb and 13 kb EcoRI/SphI genomic DNA fragments represent the wild-type and the disrupted alleles, respectively (lower). (C) Representative results of PCR-mediated genotyping of embryos are shown. The primers are shown by arrows in A. +/+, Myo18+/+; +/–, Myo18+/–; and –/–, Myo18−/–. (D) Results of Western blot analysis for Myo18B protein in cultured whole embryo are shown. Upper left and right panels are blots by anti-MYO18B-N1 and anti-MYO18B-C2 antibodies, respectively. Both lower panels are blots by anti α-tubulin antibody. Myo18B protein is indicated by arrows, and non-specific bands are indicated by*.
Heterozygous mice were then mated to obtain mice with Myo18B deficiency. However, no live Myo18B−/– offspring was born from either line, implying that Myo18B−/– embryos had died during embryogenesis. To determine the time of death, embryos from different developmental stages were genotyped by PCR (Fig. 4C) and macroscopically inspected (Table 1). Up to E10, the expected Mendelian ratio of viable Myo18B−/– embryos was observed. However, at E10.5 to E11, 12 of 31 Myo18B−/– embryos were abnormal and 4 empty deciduas were observed. At E11.5 to E13.5, none of the remaining 9 Myo18B−/– embryos were normal and the number of empty deciduas increased. Myo18B−/– dead embryos were shown to have either dilated pericardial cavities, internal hemorrhage or resorption.
| Total | Myo18B +/+ | Myo18B +/– | Myo18B −/– (D/V)* | Empty decidua | Aspects of Myo18B−/– dead embryos† | ||
|---|---|---|---|---|---|---|---|
| Newborn | line 153 | 112 | 50 | 62 | 0 | ||
| line 109 | 107 | 40 | 67 | 0 | |||
| Embryo (line 153) | E11.5-13.5 | 55 | 14 | 24 | 9 (9/0) | 8 | 1D; 1H; 7R |
| E10.5-11 | 139 | 37 | 67 | 31 (12/19) | 4 | 1D; 7H; 4R | |
| E9.5-10 | 74 | 13 | 37 | 19 (0/19) | 0 |
- * The number of dead (D) and viable (V) embryos is shown as D/V in parentheses.
- † The number of Myo18B dead embryos with dilated pericardiac cavity (D), internal hemorrhage (H), and resorption (R) is shown.
To verify the knockout of the targeted region, Western blot analysis with lysates of cultured whole embryos (Fig. 4D) were carried out. A prominent 290-kD band was detected in the lysate of wild-type (Myo18B+/+) embryos. The intensity of the band decreased in Myo18B+/– embryos, and the band disappeared in Myo18B−/– embryos.
Myo18B expression probed by the knocked-in bacterial LacZ gene in Myo18+/– mice
β-Galactosidase, the gene product of nls-LacZ, generates blue signals by hydrolyzing X-Gal. Therefore, the knocked-in nls-LacZ gene under the transcriptional control of the endogenous Myo18B promoter signals the pattern of Myo18B mRNA expression in Myo18+/– embryos and mice. The whole mount X-Gal staining of Myo18+/– embryos revealed that nls-LacZ was expressed in heart and somites at E9.5 and E11.5 (Fig. 5A–C). At E13.5, nls-LacZ expression was prominent in muscles of the cardiac atriums and ventricles (Fig. 5D left and lower right panel), and was also detected in myotomes and muscle mass (Fig. 5D left and upper right panel). In Myo18+/– adult mice, nls-LacZ was expressed prominently in cardiac and skeletal muscles (Fig. 5E,F) as well as in smooth muscles of several tissues, and in some parts of testis and brain (data not shown). The pattern of Myo18B expression indicated by X-gal staining was similar to that previously shown by tissue Northern blot analysis (Salamon et al. 2003).
Myo18B expression probed by LacZ after X-Gal staining. Whole-mount staining of E9.5 (A) and E11.5 (B, C) Myo18+/–embryos. Staining of cross sections at the level of the forelimb (D) of an E13.5 Myo18+/– embryo. High magnification view of the boxed regions is shown in right panels. Staining of sections of cardiac (E) and skeletal muscle (F) of an adult Myo18+/– mouse. Bars = 1 mm (A–C), 0.2 mm (D), 0.1 mm (E, F). Abbreviations: h, heart; s, somites; mt, myotome; mm, muscle mass.
Cardiac defects in Myo18B−/– embryos
Macroscopically, the Myo18B−/– embryos appeared to be normal in early developmental stages, with no detectable abnormality up to E9.5. Abnormal embryos were found as early as E10.5. From E10.5 to E13.5, the number of resorbed or growth-retarded embryos (Fig. 6D) increased, and some unresorbed Myo18B−/– embryos showed severe internal hemorrhage in the cardiac and ventral body wall regions (Fig. 6B) or pericardiac cavity dilation (Fig. 6H).
Analysis of embryonic development in control and Myo18B mutant embryos. (A–D) Control (A, C) and Myo18−/– embryos (B, D) at E10.5 and E12.5 are shown. Myo18−/– embryos displayed an internal hemorrhage at E10.5 (B) and were resorbed at E12.5 (D). (E–H) Comparative histological analysis of heart structures in control and Myo18−/– embryos at E10.5 and E12.5. Mid-level sections from control (E, G) and Myo18−/– hearts (F, H) are shown. Myo18−/– embryos show a cardinal veinal and atrial enlargement (F, H) and dilated pericardiac cavity (H) compared with control (E, G). Myo18B-153 line mice were used for all the data in this figure. Abbreviations: rc, right cardinal vein; lc, left cardinal vein; ra, right atrium; la, left atrium; rv, right ventricle; lv, left ventricle; pc, pericardiac cavity. Bars = 0.1 mm (A, B, E, F), 1 mm (C, D), 0.5 mm (G, H).
Therefore, Myo18B−/– embryos and control (Myo18B+/+ and Myo18B+/– embryos) littermates were histologically analyzed at various developmental stages. At E9.5, Myo18B−/– embryos showed a normal looping of the linear heart tube and correct cardiac morphology compared with control embryos (data not shown). At E10.5, the formation of the cardiac chambers was initiated correctly in Myo18B−/– embryos that were neither resorbed nor hemorrhagic. However, some Myo18B−/– embryos at E10.5 to E12.5 showed striking abnormalities in the myocardium (Fig. 6F,H). Variable degrees of pericardial effusion were observed (Fig. 6H). The lumens of both the right and left cardinal veins and the right atrium in Myo18B−/– embryos were morbidly enlarged (Fig. 6F,H) when compared with controls (Fig. 6E,G).
Abnormal myofibrillar organization in Myo18B−/– cardiac myocytes
Ultrastructural analysis of comparable areas from the left ventricular free walls of Myo18B+/+ and Myo18B−/– embryos revealed abnormalities in sarcomeric assembly in the Myo18B−/– embryos. Myo18B−/– cardiac myocytes had normal sarcomeres length with structures of intercalated disks and Z-lines, but displayed a disorganized alignment of parallel thick and thin filaments (Fig. 7B,C) when compared with Myo18B+/+ cardiac myocytes (Fig. 7A). Additionally, in transverse sections of myofibrils, Myo18B+/+ myocytes showed ordered alignment of thick filaments and of thin filaments surrounding thick filaments (Fig. 7D,D′); however, Myo18B−/– myocytes showed disordered alignment and unbalanced distribution of thick and thin filaments (Fig. 7E,E′,F,F′). These data indicate that Myo18B is necessary for the development and/or maintenance of normal cardiac myofibril structure.
Ultrastructural analysis of cardiac myocytes by TEM. Representative images were taken from comparable areas of ventricular free walls. Vertical sections (A, B, C) and transverse sections (D, E, F) of myofibrils of E10.5 Myo18+/+ (A, D) and Myo18−/–(B, C, E, F) embryos are shown. Arrows and arrowheads indicate intercalated disk and Z-line, respectively. Thick filaments and thin filaments are marked in yellow and red dots, respectively (D′, E′, F′) in the same fields of D, E, and F. Bars = 0.5 µm (A–C), 0.1 µm (D–F).
Discussion
We demonstrated here that Myo18B is a unique unconventional myosin that is predominantly expressed in cardiac and skeletal muscle cells and its expression is essential for the development and/or maintenance of myofibrillar structure in myocytes. Most unconventional myosins are known to be expressed ubiquitously in various types of cells, or specifically in non-muscle cells, such as MYO15A (Liang et al. 1999). Thus, Myo18B is a unique unconventional myosin that is predominantly expressed in myocytes like conventional (type II) myosins. Whereas conventional myosins, which consist of 15 genes in humans, are expressed in a time limited manner in specific types of myocytes, Myo18B is expressed in cardiac, skeletal and a part of smooth muscle cells, and is expressed from the embryonic stage to adulthood. Therefore, the expression pattern of Myo18B is distinct from conventional myosins. Promoter analyses of the Myo18B gene revealed that MEF2 plays an important role in Myo18B gene expression in myocytes. The expression patterns of the MEF2 genes in vivo are very similar to that of the Myo18B gene (Edmondson et al. 1994). These data support the regulation of Myo18B gene expression by MEF2.
Another unique aspect of Myo18B is its localization to the Z-lines and their peripheral regions in striated muscle cells, because most conventional myosins are localized to the A-bands. One member of conventional myosins, non-muscle myosin IIB (NMHCIIB), is also localized to the Z-lines in cardiac myocytes (Takeda et al. 2000). NMHCIIB deficiency in mice showed cardiac myocyte enlargement and binucleation (Tullio et al. 1997; Uren et al. 2000; Takeda et al. 2003). This phenotype is clearly different from the phenotype of Myo18B deficiency. Thus, physiological functions of these two myosins are likely to be distinct from each other. Because the N-terminal extension of Myo18B was localized on stress fiber, Myo18B is likely to interact with F-actin through not only the motor domain but also the N-terminal unique extension. Recently, MYO18A, another 18th class of myosin, was shown to be localized on F-actin with the domain of its N-terminal extension. The residues in the N-terminal extension that are conserved between MYO18A and MYO18B were further shown to be important for the physiological interaction of MYO18A to F-actin (Isogawa et al. 2005). This data also support the possibility that Myo18B interacts with F-actin through its N-terminal unique extension. However, Myo18B did not localize along with thin filaments in striated muscle cells (Fig. 2), and localized on stress fiber with a punctuated structure (Fig. 1). Furthermore, our previous study showed that Myo18B is localized in the region of membrane protrusion, but not of cell-cell contact, although both regions are known as ones with F-actin rich structure (Ajima et al. 2007). These results strongly indicate that Myo18B does not co-localize with F-actin itself, but co-localize with some actin-associated proteins, which are present in stress fibers or in the region surrounding the Z-lines.
Myo18B−/– embryos were lethal during embryogenesis, and dead embryos had either dilated pericardial cavities or internal hemorrhage. Furthermore, cardiac myofibrillar aberrations were observed in Myo18B−/– myocytes. Therefore, it is likely that Myo18B functions in the Z-lines during the development and/or maintenance of highly ordered structures of thick and thin filaments, and that this function of Myo18B cannot be compensated for by other myosins in vivo. Such myofibril aberrations in Myo18B−/– myocytes could lead to their reduced contractility that results in the occurrence of heart failure and embryonal death at E10.5–11.5 of Myo18B−/– embryos. However, whether or not cardiac myofibrillar disorganization in Myo18B−/– embryos disrupts cardiac myocyte function and leads embryonic lethality remains to be clarified, because pericardial effusion could occur not only by myocardial dysfunction itself but also by several other circulation defects. Therefore, to resolve this question, further functional studies of Myo18B−/– myocytes will be needed.
The Z-lines and the region adjacent to the Z-lines, where Myo18B locates, is known to be important not only to keep the structure of the sarcomeres but also to regulate the function of striated muscles. Protein complexes in the peripheral region of the Z-lines physically interact with costameres, which transmit contractile forces from the sarcomere across the sarcolemma to the extracellular matrix, and mutations of genes for several proteins in these protein complexes cause diseases of muscle (Reviewed in Ervasti 2003; Frank et al. 2006). For instance, germ-line mutations of genes encoding cardiac α-actin, muscle LIM-domain protein (MLP), desmin and Cypher/Z ASP have been found in patients with familial dilated cardiomyopathy (Chien 2000; Knoll et al. 2002; Mohapatra et al. 2003; Arimura et al. 2004). Mice deficient in these genes, which are considered as being models of human dilated cardiomyopathy, resulted in the formation of disordered myofibrillar structures (Milner et al. 1996; Arber et al. 1997; Honda et al. 1998; Zhou et al. 2001; Chu et al. 2003; Abdelwahid et al. 2004). Thus, these proteins in the Z-lines could also be essential for the development and/or maintenance of highly ordered myofibrillar structures in cardiac myocytes, and their defects may induce dilated cardiomyopathy. The phenotypes of Myo18B deficient mice, such as enlargement of the right atrium and disordered myofibrillar structures, are similar to dilated cardiomyopathy developed in these model mice. Therefore, functional and physiological interaction of Myo18B with these proteins will help us further elucidate the function of Myo18B in myocytes and clarify the molecular mechanism of the onset of dilated cardiomyopathy. It will be also worth investigating whether germ-line MYO18B mutations are present in a subset of familial dilated cardiomyopathy.
Experimental procedures
Preparation of anti-Myo18B antibody
Glutathione S-transferase (GST)-MYO18B C-terminal (human MYO18B 2448–2518 a.a. region) fusion protein was produced from pGEX-6T (Amersham Pharmacia Biotech), and recombinant protein was isolated by a glutathione–Sepharose 4B column (Amersham Pharmacia Biotech), cleaved by 150 U/mL PreScission Protease (Amersham Pharmacia Biotech), and were used to immunize two rabbits. Antisera from the immunized rabbits were affinity purified using the antigenic column, and immunoabsorbed against GST and E. coli proteins. Polyclonal antibody (pAb) against MYO18B C-terminal proteins was designated as anti-MYO18B-C2.
Western blot analysis
Cells were lysed with TNE buffer [10 mm Tris-Hcl (ph7.5), 150 mm NaCl, 1 mm EDTA, 1% NP-40 and complete Mini protease inhibitor tablet (Roche)], and lysates were resolved by SDS-PAGE and electroblotted to PVDF membranes (Millipore). The membranes were blocked with 0.05% Tween 20, 3% Skim milk in TBS. Anti-MYO18B-N1 (Ajima et al. 2007) and anti-MYO18B-C2 antibodies (1:2000) were used with anti-rabbit IgG-HRP (1:3000; Santa-Cruze Biothecnology). Monoclonal antibodies (mAbs) against myogenin (1:1000; Santa-Cruze Biotechnology) and α-tubulin (1:10 000; Sigma-Aldrich) were used with anti-mouse IgG-HRP (1:3000; Santa-Cruz Biotechnology). Bands were detected by chemiluminescence (PIERCE).
Vector construction
For reporter vector construction, the 5′ region (–456 to +79) of the human MYO18B gene was amplified by PCR and ligated into the pGL3 basic vector (Promega). Deletion constructs were generated from MYO18B-456 (–456 to +79) by serial deletions of the 5′ end by restriction enzyme digestion. Namely, the MYO18B-456 fragment in the vector was digested with FauI, NcoI, SacI, SalI, XcmI, and SmaI/HindIII to produce MYO18B-310 (–310 to +79), MYO18-196 (–196 to +79), MYO18-131 (–131 to +79), MYO18-105 (–105 to +79), and MYO18-69 (–69 to +79), respectively. Deletion constructs were then ligated into the corresponding restriction sites in the multiple cloning sites of the vector. To produce a construct with a mutation in the putative AP2- or MEF2-binding site, a site-directed mutagenic PCR was carried out with primer sets, mAP2 Forward (5′-GCTGCCG TCTACTGACCAGATCTGGATCTC ACAAGCTAAT-3′) and mAP2 Reverse (5′-ATTAGCTTGTGAGATCCAGATCTGGT CAGTAGACGGCAGC-3′), or mMEF2 Forward (5′-GACCTGG AGCTCACAAGCTGACGTTCGAAATGACTGGTCGACTC-3′) and mMEF2 Reverse (5′-GAGTCGACCAGTCATTTCGA ACGTCAGCTTGTGAGCTCCAGGTC-3′), using a QuickChangeTM Site-directed Mutagenesis Kit (Stratagene). Positions of the binding sites are underlined, and mutated nucleotides are indicated in italics.
For the construction of an EGFP-fusion full-length MYO18B expression vector, the stop codon of human MYO18B cDNA in the pcDNA3.1+ vector (Nishioka et al. 2002) was excluded by a PCR based method, and EGFP cDNA from the pEGFP-N3 vector (BD Clontech) was inserted into the 3′ end of MYO18B cDNA. For the construction of an Nter-EGFP expression vector, the 1–554 a.a. region of MYO18B cDNA was amplified by PCR and inserted into the pEGFP-N3 vector (BD Clontech). For the construction of an EGFP-moter expression vector, the 555–1357 a.a. region of MYO18B cDNA was amplified by PCR and inserted into the pEGFP-C1 vector (BD Clontech). An EGFP-ΔNter expression vector was constructed by insertion of the 1357–2567 a.a. region of MYO18B cDNA from pcDNA3-MYO18B to the EGFP-moter vector.
Cell culture
C2C12 cells (ATCC) were cultured in DMEM supplemented with 15% FBS. To induce differentiation of C2C12 cells, the medium was replaced by DMEM supplemented with 2% horse serum and further changed every 2 days for 1 week. E9.5 whole embryos were suspended in 150 µL matrigel (BD Biosciences) and dropped to the center of 6 well dishes. After 30 min incubation at 37 °C, a drop of embryo in matrigel was covered with DMEM supplemented with 10% fetal bovine serum (FBS). After 1-week's culture, cells were collected with BD cell recovery solution (BD Biosciences).
Transfection and luciferase assay
For luciferase assay, C2C12 cells were plated at 5 × 104 cells/well onto a 12-well dish one day before transfection, and the differentiation was induced by changing the medium just before transfection. The cells were then transfected with 0.5 µg of each reporter vector construct using FuGene6 (Roche) triplicate for each constructs, harvested 60 h after transfection, and cell extracts were prepared using a standard procedure. The transfection efficiency was determined by co-transfection of 5 ng of the Renilla luciferase expression vector (Promega) under the β-actin promoter, and relative luciferase activity was calculated as recommended by the manufacturer.
For immunocytochemistry, C2C12 cells were plated onto coverslips one day before transfection, and transfected with each expression vector construct using FuGene6 (Roche).
Immunohistochemistry and immunocytochemistry
For immunohistochemistry, sections were prepared as follows. For frozen sections, tissues were fixed overnight with 1% paraformaldehyde (PFA) at 4˚C, washed with ice-cold phosphate-buffered saline (PBS), equilibrated with 10, 20, 30% sucrose in PBS at 4˚C, embedded in Tissue-Tek OCT compound, and frozen using dry ice. The tissues were then sectioned at 10 mm thickness, mounted on glass slides, post-fixed with 0.2% PFA/PBS for 10 min, and permeabilized by 0.2% Triton X-100/PBS for 10 min. For paraffin sections, embryos were fixed overnight with 10% neutralized formalin, dehydrated, embedded in paraffin, sectioned at 5 mm thickness, and mounted on glass slides. The sections were then stained using a VECTOR M. O. M. Immunodetection Kit or a VECTASTAIN Elite ABC kit (VECTOR Laboratories) according to the manufacturer's protocol. For immunocytochemistry, C2C12 cells on coverslips were fixed with 4% PFA/PBS for 10 min, permeabilized by 0.2% Triton X-100/PBS for 10 min, and blocked with 2% BSA in PBS. Antibodies and reagents used were as follows: anti-MYO18B-N1 and anti-MYO18B-C1 rabbit pAbs (1:200); anti-α-actin mouse mAb (1:800; Sigma-Aldrich, EA-53); anti-skeletal myosin mAb (1:2000; Sigma-Aldrich, MY-32); anti-myosin heavy chain β (1:300; CHEMICON); Rhodamine-Phalloidin (1:100; Molecular Probe); Alexa Fluor 488 conjugate goat anti-rabbit IgG (1:500; Molecular Probe); Cy3 conjugate goat anti-mouse IgG (1:1000; Jackson ImmunoResearch); Alexa Fluor 546 conjugate streptavidin (1:500; Molecular Probe). Confocal images were obtained using a Carl-Zeiss Radiance2000 system.
Generation of Myo18B knockout mice
A 2.0-kb 5′ genomic fragment and a 5.2-kb 3′ fragment flanking exon 2 of the Myo18B gene were amplified by PCR from a mouse C57BL/6J BAC clone (Rosewell Park Cancer Institute) DNA, and linked to the nlsLacZ and NeoR genes at the 5′ and 3′ positions, respectively, in pBluescript SK + (Stratagene co.). A Poly-A signal sequence was inserted between the nlsLacZ and NeoR genes. The Diphtheria toxin A (DTA) gene was also linked to the 3′ end of the targeting cassette (Fig. 4A). The targeting construct was linearized and electroporated into cultured ES cells, and G418-resistant cell clones were screened for homologous recombination by PCR and Southern blot analyses. Two independent ES cell clones with heterozygous targeted disruption of the Myo18B gene (clones109 and 153) were injected into blastocysts of C57/BL6J mice, 3 days after their fertilization to create chimeric mice. The chimeric male mice and C57/BL6J female mice were used to generate Myo18B+/– mice. The Myo18B+/– mice were interbred to generate Myo18B−/– mice. The animal experiment protocols were approved by the Committee for Ethics in Animal Experimentation, and the experiments were conducted in accordance with the Guideline for Animal Experiments of the National Cancer Center.
Genotyping of Myo18B gene targeted mice
Genomic DNA was isolated from ES cells and mouse tails. Genotypes of ES cells were first determined by PCR with a set of primers (SA-F: 5′-TCAGCTGTCCTGGATTTCAATAT AAAGCTGAAGA-3′ and SA-R: 5′-TGCGCAACTGTTGGG AAGGGCGATC-3′) that specifically amplified the disrupted gene, and then confirmed by Southern blot analysis with a probe marked in Fig. 4A. Genotypes of offsprings were determined using tail or yolk sac DNA by PCR (primer sequences: F = 5′-GAGCACTCAGAAGCACAACGTCATCGC-3′, R = 5′-GGC CACAGCCCCTCTTGCCAG, LacZ = 5′-TGCGCAACTGTT GGGAAGGGCGATC-3′).
X-Gal staining
For the staining of whole embryos, embryos were fixed with 2% PFA and 0.2% glutaraldehyde in PBS for 30 min at 4˚C, rinsed with PBS, and incubated for 3–4 h at 37 °C or overnight at room temperature in the staining solution containing 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-Gal). For the staining of specimens, frozen sections were prepared as described above, post-fixed with 0.2% PFA/PBS, permeabilized by 0.01% sodium-deoxycolate, 0.02% NP-40 and 2 mm MgCl2 in PBS, and incubated for 3–4 h at 37 °C in the staining solution.
Histological analysis
Embryos were isolated, fixed overnight with 10% neutralized formalin, dehydrated, embedded in paraffin, sectioned, and stained with hematoxylin and eosin (H&E).
Transmission electron microscopy (TEM)
E10.5 embryos, whose hearts were still beating, were prefixed with a mixture of 4% paraformaldehyde and 0.3% glutaraldehyde in a 0.1 M cacodylate buffer for 12 h. Hearts were dissected from embryos, postfixed with 2% glutaraldehyde in a 0.1 M cacodylate buffer for 24 h and with 3% osmium tetroxide for another 3 h, dehydrated through an ethanol series, and embedded in Epon812. The hearts were then sectioned at 0.5 µm thickness, stained with toluidine blue, and examined by a light microscope. After careful examination for location and orientation of the sample, 80–90 nm sections were prepared, doubly stained with uranylacetate and lead citrate, and observed under a JEOL 2000EX electron microscope at 80 kV.
Acknowledgements
We thank T. Noda (Cancer Institute, Tokyo, Japan) for providing us with nls-LacZ plasmid. This work was supported in part by a grant-in-aid from the Ministry of Health, Labor and Welfare of Japan for the 3rd-term Comprehensive 10-year Strategy for Cancer Control. R. A. was an awardee of Research Fellowships of the Japan Society for the Promotion of Science for Young Scientists during this study. This manuscript was edited by NIH Fellows Editorial Board.
