Cytoskeletal restructuring is required for cell shape changes, for the movement of cells, and for the shaping of tissues throughout morphogenesis. Cells depend on cytoskeletal molecules and molecular motors to establish their asymmetrical shapes and to drive cell motility (Pollard,2003). Heart organogenesis requires cell migrations (Linask and Lash,1986,1988), cell shape changes (Linask,1992a, b; Linask et al., 1997), and the bending of epithelial sheets to form a tubular structure (Linask et al.,2005a). The straight heart tube bends and rotates forming the definitive four-chambered organ (Linask and Lash,1998; Linask,2003). The anterior–posterior and dorsal–ventral patterning of cell–cell and cell–matrix adhesion molecules and their activity in signal transduction pathways are key in coordinating the development of cardiac form with function (Linask,2003; Linask and VanAuker,2007).
Nonmuscle Myosin II Proteins and Heart Looping
On the basis of the present study, the cross-talk between the cell adhesion molecules, cell–cell and cell–matrix, during looping is facilitated by the actomyosin cytoskeletal complex and signaling, that involves nonmuscle myosin IIB, and possibly NMHC-IIA, in the myocardium. Knockout mice of NMHC-IIB and NMHC-IIA had been generated (Tullio et al.,1997; Conti et al.,2004) previous to our identification of flectin antibody recognizing NMHC-IIA and -IIB isoforms in the heart. NMHC-IIA transgenic embryos displayed embryonic lethality at ED7.5 (i.e., early heart compartmentalization stage). The NMHC-IIB null transgenic mice were described as displaying heart looping defects and usually surviving to mid-gestation (Tullio et al.,1997). However, in those studies an analysis of NMHC-IIA and -IIB expression during earlier, heart looping, stages was not done, nor was the effect on heart function defined. Indeed possibly because of level of detectability of their Western blot data, it was indicated previously that NMHC-IIA was not present in the heart during development, unless NMHC-IIB was knocked out and then was up-regulated (Tullio et al.,1997). Our data indicate that both NMHC-IIA and NMHC-IIB isoforms are present already during the earliest stages of normal heart development and during looping.
Flectin antibody seemingly is a pan-NMHC-II antibody, as it interacts with both NMHC-IIA and NMHC-IIB. It also appears as seen in Figure 5 to be localizing to possibly a secreted isoform of NMHC-II in the cardiac jelly or alternatively, to cellular footprints or cell processes, as matrix molecules are being secreted to form the cardiac jelly by the myocardial cells (Manasek,1976a). The commercial NMHC-IIA and -IIB antibodies are made against 12 amino acid peptides corresponding to the C-terminus of human nonmuscle myosin heavy chain isoform A and B that are also well conserved in the mouse, but differ in the chick isoforms. This finding also suggests the peptide antibodies may not bind to modified isoforms, as phosphorylated forms, or to secreted forms where cleavage of the secreted protein may have taken place. It is known that the phosphorylation status of myosin II is important in intracellular tension and is associated with stress fibers, whereby signaling can occur by means of the ROCK-RhoA-kinase pathway regulating the phosphorylation level of myosin II (Vemuri et al.,1999; Davis et al.,2001; Epstein and Davis,2003; Kawabata et al.,2004). Interestingly, inhibition of Rho kinases by Y27632 blocks normal heart development and normal left–right asymmetry (Wei et al.,2001,2002). We suggest this may occur by its leading downstream to an inhibition of nonmuscle myosin II phosphorylation. Flectin antibody on the other hand was made against cellular proteins and apparently recognizes a conserved region on both NMHC-IIA and -IIB, and it may bind to both phosphorylated and unphosphorylated isoforms. This allows detection of myosin II proteins in cells and cell processes when under tension or possibly even secreted under conditions where there is mechanical tension. The latter aspects are now being investigated in the looping heart. NMHC-IIB is reported to be in cell processes within lamellipodia/filopodia, and it is associated also with fibroblast collagen fiber transport and deposition during remodeling of the ECM (Meshel et al.,2005). Either tension fiber formation or matrix deposition could explain the seemingly extracellular and/or subepithelial localization initially reported for flectin-F22 protein (Lash et al.,1992). Myocyte cytoplasmic stress fibers have been previously reported in the matrix compartment of later stage embryonic hearts in the AP septum region and were described as morphological adaptations to mechanical tensions in the heart (Sumida et al.,1989). Both possibilities appear to exist for the embryonic looping heart stages.
Fibronectin (FN) and collagens localize to the myocardial basal lamina where we detect cardiomyocyte cell processes. Upon RGD-perturbation of integrin-mediated cell matrix interactions, heart development becomes abnormal (Lash et al.,1987; Yost,1992). As a result of RGD-perturbation, flectin was reduced in its expression in the cardiac jelly region of the heart (Linask and VanAuker,2007). This finding may be due to a weakening of integrin binding on the cell processes with the basal lamina FN, and leads to a retraction of the cell processes or filopodia. This possibility warrants further investigation. We hypothesize that NMHC-IIB in association with integrin-mediated signal transduction is involved in mechanotransduction of increasingly higher blood flow volumes that are feeding into the heart during looping. As an example, increasing blood pressure may be transmitted to the inflow part of the myocardial wall as seen in Figure 7 by the compression or displacement of the cardiac jelly. During the cardiac cycle, the forces of ECM compression/displacement seemingly can be detected and transmitted to the closely apposed myocardium by cell processes, even short ones that are extending into the cardiac jelly matrix during the cardiac cycle. Although degradation of the cardiac jelly by hyaluronidase allows looping to continue in the rat and chick models (Baldwin and Solursh,1989; Linask et al.,2003b), this enzyme exposure would be expected to break down proteoglycans, but would not be expected to degrade integrin-glycoprotein interactions within the myocardial basal lamina that becomes closely apposed to the endocardium in these experimentally treated embryos. The association of NMHC-II proteins with mechanotransduction is seen in other systems, as for example, in relation to cilia in the inner ear for transmission of sound waves and in relation to afferent and efferent arterioles in the kidney in detection of blood flow pressure (Shiraishi et al.,2003; Marigo et al.,2004). Recently, with regard to mesenchymal stem cells the nonmuscle myosins are shown to be involved in sensing matrix-elasticity to drive cell lineage specification (Engler et al.,2006). This feedback of local matrix changes on the cell state has important implications for development, differentiation, disease, and tissue regeneration (Discher et al.,2005; Linask and VanAuker,2007).
Reduced Asymmetric Nodal Activity Leads to Abnormal Looping and Reduced Levels of Flectin/NMHC-IIB in the Heart
A downstream read-out of misexpression of upstream laterality regulatory genes has been the dextral direction and bending of heart looping. We determined that direction of heart looping is specified at approximately the seven-somite stage in the chick (Linask et al.,2005a) and suggested that the asymmetry of cell proliferation and flectin in the anterior heart field may relate to asymmetric patterning of Pitx2c (Ai et al.,2006). The laterality gene Nodal in all vertebrates analyzed shows a left expression in the embryo and heart fields and is upstream of Pitx2c. Reducing the level of Nodal in the left lateral plate mesoderm in the NodalΔ600/Δ600 embryo led to a restricted field of expression, as well as delayed expression of Pitx2c and abnormal cardiac looping at ED9.5 (Norris et al., 2002b). As we report here, the reduced Nodal expression in these transgenic mouse embryos resulted in a reduced expression of flectin and similarly of NMHC-IIB in the myocardium at ED9.5, detectable on both the protein and mRNA level. This study confirms in the mouse model our previous observations in the avian model on modulation of flectin by laterality genes during cardiac looping (Linask et al.,2002,2003b).
Actomyosin Cytoskeleton and Heart Looping
The bending forces within the tubular heart appear to have a common underpinning related to tension and contractility of the cytoskeleton. Actin is a major component of looping (Manasek et al.,1972; Manasek,1976b; Itasaki et al.,1991; Latacha et al.,2005). Development of actin filaments considerably differ in inner and outer myocardial cell layers (Shiraishi et al.,1992). In the inner layer at the tubular heart stage when the tube begins to bend, myofibrils are aligned with their long axis circumferentially arranged in relation to the tubular structure. Myocytes in the outer layer are round. It was suggested that the circumferentially arranged actin filaments at the basal aspect of the inner layer of the myocardium promote cardiac looping (Shiraishi et al.,1992; Price et al.,1996). Extracellular matrix protein fibronectin (FN) becomes aligned parallel with the actin bundles early in looping (Shiraishi et al.,1995). It is also this inner layer of cardiomyocytes that detectably sends cell processes, as well as secretes proteins, into the subjacent cardiac jelly layer. Additionally, a member of a the Ca++-dependent family of adhesion molecules, N-cadherin, is responsible for the connection of actin and myofibrils between neighboring myocytes at cell–cell junctions (Linask,1992a; Kostetskii et al.,2005) and for the cell alignment and arrangement of the two layers of the developing heart tube (Shiraishi et al.,1993). Thus, during looping, as in the earlier period of cardiac compartment formation, cross-talk between two adhesion mediated-signaling pathways, integrin and N-cadherin, is evident (Linask and Lash,1990; Linask,1992b; Linask et al.,2005b) and as shown here, appears to be facilitated by the actin-nonmuscle myosin-II complex.
The identification of flectin as being more like nonmuscle myosin IIB localization during heart looping suggests the actomyosin cytoskeleton is involved in heart tube bending at a time that the myocardium is contracting and blood flow is present. We observed in the NMHC-IIB null embryo that heart function was compromised when compared with wild-type littermates, and may aid in the formation of the previously described cardiac anomalies (Tullio et al.,1997). NMHC-IIA apparently partially can compensate for NMHC-IIB, as a four-chambered heart does form, albeit abnormally. The abnormal cardiac function present in the heterozygous and homozygous embryos strengthens the hypothesis that there is a coordination of cardiac form with function during morphogenesis of the four-chambered heart (Linask,2003).
Nonmuscle myosin-II proteins are important in morphogenesis throughout phylogeny. Evidence has been presented that nonmuscle myosins function in Drosophila in organizing the cytoskeleton during morphogenesis and in controlling left–right asymmetry (Blake et al.,1998,1999; Speder et al.,2006). In summary, our study suggests that NMHC-IIB within the cytoskeleton has a role in laterality and mammalian mouse heart morphogenesis during looping.