TGFbeta2 and TGFbeta3 have separate and sequential activities during epithelial-mesenchymal cell transformation in the embryonic heart.

Heart valve formation is initiated by an epithelial-mesenchymal cell transformation (EMT) of endothelial cells in the atrioventricular (AV) canal. Mesenchymal cells formed from cardiac EMTs are the initial cellular components of the cardiac cushions and progenitors of valvular and septal fibroblasts. It has been shown that transforming growth factor beta (TGFbeta) mediates EMT in the AV canal, and TGFbeta1 and 2 isoforms are expressed in the mouse heart while TGFbeta 2 and 3 are expressed in the avian heart. Depletion of TGFbeta3 in avian or TGFbeta2 in mouse leads to developmental defects of heart tissue. These observations raise questions as to whether multiple TGFbeta isoforms participate in valve formation. In this study, we examined the localization and function of TGFbeta2 and TGFbeta3 in the chick heart during EMT. TGFbeta2 was present in both endothelium and myocardium before and after EMT. TGFbeta2 antibody inhibited endothelial cell-cell separation. In contrast, TGFbeta3 was present only in the myocardium before EMT and was in the endothelium at the initiation of EMT. TGFbeta3 antibodies inhibited mesenchymal cell formation and migration into the underlying matrix. Both TGFbeta2 and 3 increased fibrillin 2 expression. However, only TGFbeta2 treatment increased cell surface beta-1,4-galactosyltransferase expression. These data suggest that TGFbeta2 and TGFbeta3 are sequentially and separately involved in the process of EMT. TGFbeta2 mediates initial endothelial cell-cell separation while TGFbeta3 is required for the cell morphological change that enables the migration of cells into the underlying ECM.

Production of the transforming growth factor-beta binding protein endoglin is regulated during chick heart development.

The early embryonic heart consists of two cell types. The cells form an inner epithelial tube of endocardium within an outer tube of myocardium separated by a cell-free extracellular matrix. A crucial process in heart development is the production of cushion mesenchyme in the atrioventricular (AV) canal and outflow tract (OT). Cushion mesenchyme differentiates from the endocardium in response to signaling molecules produced by the adjacent myocardium. In chicken hearts, both transforming growth factor-beta3 (TGF-beta3) and TGF-beta2 are present and have been identified as being important in the production of cushion mesenchyme. We were interested in how the signals from these two similar molecules may be differentiated during early heart development. To this end, we examined the expression of endoglin, a TGF-beta receptor molecule, in the developing chick heart. Endoglin is typically located on endothelial cell layers and binds tightly to TGF-beta1 and TGF-beta3 but not well to TGF-beta2. We show that during the formation of the primitive heart tube, endoglin is found at relatively high levels in both presumptive myocardium and endocardium. However, as myocardium differentiates and development proceeds, endoglin expression is progressively reduced. At stage 20 in the heart, endoglin expression is most readily seen in the AV canal and the OT. This pattern of expression is similar to the reported TGF-beta3 expression patterns in the heart.

A GATA-dependent nkx-2.5 regulatory element activates early cardiac gene expression in transgenic mice.

nkx-2.5 is one of the first genes expressed in the developing heart of early stage vertebrate embryos. Cardiac expression of nkx-2.5 is maintained throughout development and nkx-2.5 also is expressed in the developing pharyngeal arches, spleen, thyroid and tongue. Genomic sequences flanking the mouse nkx-2.5 gene were analyzed for early developmental regulatory activity in transgenic mice. Approximately 3 kb of 5' flanking sequence is sufficient to activate gene expression in the cardiac crescent as early as E7.25 and in limited regions of the developing heart at later stages. Expression also was detected in the developing spleen anlage at least 24 hours before the earliest reported spleen marker and in the pharyngeal pouches and their derivatives including the thyroid. The observed expression pattern from the -3 kb construct represents a subset of the endogenous nkx-2.5 expression pattern which is evidence for compartment-specific nkx-2.5 regulatory modules. A 505 bp regulatory element was identified that contains multiple GATA, NKE, bHLH, HMG and HOX consensus binding sites. This element is sufficient for gene activation in the cardiac crescent and in the heart outflow tract, pharynx and spleen when linked directly to lacZ or when positioned adjacent to the hsp68 promoter. Mutation of paired GATA sites within this element eliminates gene activation in the heart, pharynx and spleen primordia of transgenic embryos. The dependence of this nkx-2. 5 regulatory element on GATA sites for gene activity is evidence for a GATA-dependent regulatory mechanism controlling nkx-2.5 gene expression. The presence of consensus binding sites for other developmentally important regulatory factors within the 505 bp distal element suggests that combinatorial interactions between multiple regulatory factors are responsible for the initial activation of nkx-2.5 in the cardiac, thyroid and spleen primordia.

Mechanisms of cell transformation in the embryonic heart.

The process of cell transformation in the heart is a complex one. By use of the invasion bioassay, we have been able to identify several critical components of the cell transformation process in the heart. TGF beta 3 can be visualized as a switch in the environment that contributes to the initial process of cell transformation. Our data show that it is a critical switch in the transformation process. Even so, it is apparently only one of the factors involved. Others may include other TGF beta family members, the ES antigens described by Markwald and co-workers and additional unknown substances. Observing the sensitivity of the process to pertussis toxin, there is likely to be a G-protein-linked receptor involved, yet we have not identified a known ligand for this type of receptor. Clearly, there are several different signal transduction processes involved. The existence of multiple pathways is consistent with the idea that the target endothelial cells receive a variety of environmental imputs, the sum of which will produce cell transformation at the correct time and place. Adjacent endothelial cells of the ventricle that do not undergo cell transformation are apparently refractory to one or more of the stimuli. Figure 4 depicts a summary diagram of this invasion process with localization of most of the molecules mentioned in this narrative. As hypothesized here, elements of the transformation process may recapitulate aspects of gastrulation. Since some conservation of mechanism is expected in cells, it is not surprising that cells undergoing phenotypic change might reutilize mechanisms used previously to produce mesenchyme from the blastodisk. Though we have preliminary data to suggest this point, confirmation of the hypothesis by perturbation of genes such as brachyury, msx-1, etc. will be required to establish this point. The advantage of this hypothesis is that it provides, from the work of others in the area of gastrulation, a ready source of molecules and mechanisms that can be tested in the transforming heart. Whereas, perturbation of such mechanisms at gastrulation may be lethal to the embryo, such molecules and mechanisms may be responsible for the high incidence of birth defects in the heart.

Sense and antisense TGF beta 3 mRNA levels correlate with cardiac valve induction.

The formation of the valves in the heart is a spatially and temporally controlled process. A tissue interaction between the endothelium and its adjacent myocardium initiates the transformation of the endothelium into the mesenchymal precursors of the heart valve. One or more of the molecules implicated as critical for valve formation are members of the transforming growth factor beta family of molecules. Presented here is a spatial and temporal analysis of TGF beta 2 and TGF beta 3 in the chick heart during valve formation. We show that TGF beta 3 mRNA is concentrated in AV canal tissue where valve formation will occur, consistent with previous observations that TGF beta 3 production is critical during valve formation. Additionally, an RNA complementary to TGF beta 3 encoding mRNA is present in the heart. The temporally controlled appearance of RNA complementary to TGF beta 3 suggests that this molecule may play a role in the regulation of TGF beta 3 production in the heart.