Selective Requirement of Myosin Light Chain 2v in Embryonic Heart Function*

Two major myosin light chain 2 isoforms are coexpressed in the early stages of murine cardiogenesis, a cardiac ventricular isoform and a cardiac atrial isoform, each of which is tightly regulated in a muscle cell-type-specific manner during embryogenesis (Chien, K. R., Zhu, H., Knowlton, K. U., Miller-Hance, W., van Bilsen, M., O’Brien, T. X., and Evans, S. M. (1993) Annu. Rev. Physiol. 55, 77–95). We have disrupted myosin light chain 2v gene in mice and monitored in vivo cardiac function in living myosin light chain 2v −/− embryos. The mutant embryos die at approximately embryonic day 12.5. In mutant ventricles, the myosin light chain 2a protein level is increased and reaches levels comparable to the myosin light chain 2v in the ventricles of wild type littermates and is appropriately incorporated into the thick filaments of mutant embryonic hearts. However, despite the substitution of myosin light chain 2a, ultrastructural analysis revealed defects in sarcomeric assembly and an embryonic form of dilated cardiomyopathy characterized by a significantly reduced left ventricular ejection fraction in mutant embryos compared with wild type littermates. We conclude that myosin light chain 2v may have a unique function in the maintenance of cardiac contractility and ventricular chamber morphogenesis during mammalian cardiogenesis and that a chamber-specific combinatorial code for sarcomeric assembly may exist that ultimately requires myosin light chain 2v in ventricular muscle cells.

Muscle myosin, the highly conserved molecular motor, contains one pair of heavy chains and two pairs of light chains, the essential myosin light chains (MLC-1 or -3) 1 and the regulatory myosin light chain-2 (MLC-2) (1). The three-dimensional structure of myosin indicates that the light chains are arranged in tandem, with MLC-1/-3 in the amino-terminal half of the neck and MLC-2 in the neck/tail junction (2)(3)(4). MLC-2 plays an essential role in regulating vertebrate smooth muscle contraction. The phosphorylation of a single serine residue (Ser-19) of smooth muscle MLC-2 is the switch for turning on the actinactivated myosin ATPase and hence, contraction. However, the acto-myosin interaction in vertebrate striated muscle is mainly regulated through the troponin-tropomyosin complex, and MLC-2 is thought to have only a modulatory effect (5). It has been shown in an in vitro motility assay that removal of MLC-2 from myosin markedly reduces actin filament sliding velocity without significantly reducing myosin ATPase activity (6,7). Although mutation of the single MLC-2 gene results in a flightless phenotype in Drosophila (8), the precise in vivo physiological function of myosin light chains in vertebrate striated muscle is unclear. Furthermore, there is no direct evidence that the highly conserved structure of individual MLC-2 isoforms reflects a unique functional requirement in distinct muscle cell types. Recently, alterations in MLC-2 expression have been correlated with the onset of cardiac morphogenetic defects during embryogenesis (9 -11), point mutations in MLC-2v have been shown to be associated with a genetic form of human cardiomyopathy (12), and induction of the atrial MLC-2 isoform has been shown to occur in cardiac hypertrophy and failure (13). Our efforts have been directed toward understanding the biological and physiological roles of the MLC-2v, which is the ventricular isoform of MLC-2, by disrupting the gene through homologous recombination in mice.
Histological and Immunohistochemical Analysis-Histological and immunohistochemical studies were performed by a modification of previously described methods (19,20,21).
Transmission Electron Microscopy-Embryos were processed for transmission electron microscopy as described by Price et al. (22) and Reynolds (23).
In Vivo Videomicroscopy of Embryos-Mice pregnant at E11.5 and 12.5 were intraperitoneally anesthetized with ketamine (100 mg/kg) and xylazine (2.5 mg/kg), the trachea was intubated, and ventilation was provided with a respirator. Fetuses were visualized using intravital microscopy (11,24), but with images recorded in color from a chargecoupled device camera, as recently described (24). In brief, each embryo was delivered individually from the uterus with the umbilical circulation intact, placed on a heated stage with superfusion, and imaged by transillumination. The left ventricular chamber area (including the left ventricular wall, which is thin at this age) and the chamber long axis were measured using NIH IMAGE software. Left ventricular (LV) elliptical volumes were then calculated using the single plane area-length method (25). The LV ejection fraction was calculated as (LVEDV Ϫ LVESV)/LVEDV, where LVEDV is the left ventricular end diastolic volume and LVESV is the end systolic volume. RNase Protection Assay-The riboprobes for RNase protection assays were generated as described previously (26). Ventricular samples were dissected from individual day 12 embryonic hearts, and RNase protection assays were performed per manufacturer recommendations using the Direct Protect Lysate Ribonuclease protection assay kit from Ambion, Inc. (Austin, TX).

Lack of MLC-2v Results in Embryonic Lethality at E12.5-
The mlc-2v gene targeting construct is depicted in Fig. 1. The linearized targeting construct was electroporated into J1 ES cells, and the cells were selected with G418. Of a total of 47 colonies screened, 3 were identified as homologous recombinants by Southern blot hybridization analysis of genomic ES cell DNA. Two of the three colonies were injected into C57BL/B6 blastocysts, both resulting in the generation of chimeras, one of which displayed germ-line transmission. Het-erozygous (MLC-2vϩ/Ϫ) offspring appeared normal in all respects and were crossed to generate MLC-2v homozygote (MLC-2v Ϫ/Ϫ) mice. Southern blot and PCR analyses were used to genotype offspring (Fig. 1, b and c). The absence of MLC-2v proteins in MLC-2v Ϫ/Ϫ embryos was confirmed by Western blot analysis, utilizing a polyclonal anti-MLC-2v antibody (Fig. 1d). No viable MLC-2v Ϫ/Ϫ offspring were obtained in litters from MLC-2v ϩ/Ϫ intercrosses (Table I), indicating that the MLC-2v Ϫ/Ϫ null mutation was embryonic lethal. To investigate the timing of the lethality, mice from E9.5 to 13.5 were genotyped. All of the MLC-2v Ϫ/Ϫ mutants were viable at E11.5; however, 30% of MLC-2v Ϫ/Ϫ mutants lost viability by E12.5, and all MLC-2v Ϫ/Ϫ died by E13.5 (Table I), thereby indicating that the MLC-2v Ϫ/Ϫembryos die at approximately

E12.5.
MLC-2v Ϫ/Ϫ Embryos Die from Heart Failure-All MLC-2v Ϫ/Ϫ embryos that died just before dissection displayed massive cardiac enlargement, wall thinning, chamber dilation, and pleural effusions. Histologic examination revealed hepatic congestion and an engorged vena cava consistent with congestive heart failure (data not shown). No significant abnormalities in other organs were found. Analysis of global cardiac function revealed that the left ventricular ejection fraction of MLC-2v Ϫ/Ϫ embryos averaged 33%, which was significantly reduced when compared with wild type and heterozygous littermates at E11.5 (46%, 49%) and at E12.5 (50%, 53%) (Fig. 2a). Left ventricular end diastolic volumes were similar among the three groups at E11.5, but by E12.5, were significantly elevated in MLC-2v Ϫ/Ϫ embryos (Fig. 2b). Collectively, these results suggest that global cardiac function is already severely impaired by E11.5 in MLC-2v Ϫ/Ϫ embryos and that cardiac dysfunction results in progressive embryonic heart failure around E12.5. The resulting phenotype is similar to mid-gestational embryonic heart failure seen in a wide variety of other gene-targeted murine embryos (14).
MLC-2v Ϫ/Ϫ Embryonic Hearts Display Abnormal Myofibrillar Organization-Ultrastructural analysis of comparable areas from the left ventricular free walls of wild type and mutant mice revealed abnormalities in sarcomeric assembly in the MLC-2v Ϫ/Ϫ embryos (Fig. 3). Myocytes from E12.5 MLC-2v Ϫ/Ϫ embryonic hearts displayed prominent interruptions and myofibrillar disorganization of the normal parallel alignment of thick and thin filaments (Fig. 3, b and c). Additionally, total fiber width was narrower within mutant ventricular myocytes, and overall distances between Z-bands was greater than in wild type specimens. There were also several areas in mutant hearts where the organization and alignment of Z-bands between sarcomeres was not conserved (Fig. 3c). These observations could also be observed in ventricular samples from younger MLC-2v Ϫ/Ϫ embryos (E10.5 and E11.5) during earlier stages of myofibrillogenesis (data not shown).
Taken together, these data indicate that there is a selective requirement for MLC-2v in the normal development of ventricular cardiac myocyte structure and function. The observed disruption of myofibrillar organization in the absence of MLC-2v leads to reduced myocyte contractility and cardiac function that results in death at approximately E12.5 and suggests a specific requirement of MLC-2v in the interaction between thick and thin filaments during sarcomere assembly. In this regard, in vitro motility assays have previously shown that removal of MLC-2 reduces actin filament sliding velocity by about 63% without significantly reducing the myosin ATPase activity (6, 7). It has also been proposed that MLC-2 can act to stiffen the myosin neck (3,4). If each head of the thick filament is responsible for the arrangement of proper hexagonal packing of the thick filament and thin filament during myogenesis, then the disorganized and improperly functioning myosin head due to the lack of MLC-2v would result in improper packing and organization of thick and thin filaments during cardiac sarcomere assembly.
MLC-2a Cannot Compensate for the Deficiency of MLC-2v-Two different myosin light chain 2 genes, mlc-2v and mlc-2a, are abundantly expressed in the ventricular chamber at early stages of murine cardiogenesis, raising the issue as to whether there is a unique role for MLC-2v versus MLC-2a during the maintenance of heart function and morphogenesis. mlc-2v is initially expressed at about E8 and continues to be restricted to the ventricular chamber throughout embryonic development and into adulthood (27,28). The expression of mlc-2a is uniform in the E8 linear heart tube and then becomes downregulated at the RNA level in the ventricular chamber by E12.5 (29). To investigate whether MLC-2a can compensate for MLC-2v in mutant embryos, we examined MLC-2a mRNA and protein levels in MLC-2v Ϫ/Ϫ embryos at E12. Although there were no significant differences in the MLC-2a mRNA levels between wild type and MLC-2v Ϫ/Ϫ embryos (Fig. 4a), Western blot analysis (Fig. 4b) demonstrated that MLC-2a protein content in the ventricular chamber was dramatically increased in both total cellular protein and myofibril protein extracts of these embryos. A quantitative study of myofilament proteins isolated from E12 embryonic ventricles indicated that levels of MLC-2a protein in MLC-2v Ϫ/Ϫ ventricles were comparable to the MLC-2v protein levels in wild type litter mates (Fig. 4c). This result is consistent with the findings in a transgenic mouse model in which overexpression of cardiac MLC-2v does not result in an increase in ventricular MLC-2v protein levels despite a significantly higher level of MLC-2v mRNA in these transgenic mice (30). In both cases, the MLC-2 protein level appears to be regulated at the post-transcriptional level. Thus, the current study indicates that during early stages of cardiac chamber development, MLC-2a protein levels increase in the ventricles of mutant animals, implying that MLC-2a may par-  1-4) and purified myofilament proteins (lanes 5-8) from E12.0 embryonic ventricles using a polyclonal antibody for MLC-2a and a monoclonal antibody for tropomyosin. c, myofilamental proteins from E12 embryonic ventricle were analyzed by SDS-glycerol polyacrylamide gel electrophoresis and stained with Coomassie Blue. d-i, immunohistochemical analysis of MLC-2v (panels d and f) and MLC-2a (panels e, g, h, and i) in wild type (panels d, e, and h) and MLC-2v Ϫ/Ϫ embryos (panels f, g, and i) by confocal microscopy. Panels h and i were taken from the left ventricular free wall rough trabecular area, similar to that shown in Fig. 3. tially replace MLC-2v in the developing ventricular myocyte. In this regard, immunohistochemical analysis by confocal microscopy of MLC-2a expression in MLC-2v Ϫ/Ϫ embryos demonstrated that MLC-2a protein was indeed elevated in the ventricles of mutant embryos (Fig. 4g) and incorporated into nascent myofibrils (Fig. 4i). However, MLC-2a cannot completely compensate for the deficiency of MLC-2v, as shown by the lack of sarcomeric structure (Fig. 3) and depressed cardiac contractility (Fig. 2). We conclude that there is an important qualitative difference between MLC-2v and MLC-2a proteins, reflecting a unique requirement for MLC-2v during functional maturation of the ventricular chamber, thereby underscoring the potential importance of MLC-2 in the in vivo regulation of cardiac contractility. Furthermore, these studies provide the first direct evidence that alterations in MLC-2v can directly lead to overt heart failure, which will become of interest to explore in the post-natal setting of the adult heart.