Rapid disappearance of the medial epithelial seam during palatal fusion occurs by multifocal breakdown that is preceded by expression of alpha smooth muscle actin in the epithelium.

Breakdown of the medial epithelial seam (MES) is essential to allow bridging of the mesenchyme during palatal fusion. Evidence exists for three mechanisms for this breakdown that are incompatible at the level of individual cells in the seam. To determine if breakdown of the seam was regionally restricted, 3-dimensional reconstructions were generated using volume rendering software from 1 micron serial sections in the sagittal plane of rat palates fixed during the process of fusion. The earliest break detected in electron micrographs was cell separation and in reconstructions was a discrete defect, with a rounded outline, nearer to the nasal than to the oral margin of the seam. Further breakdown produced a pattern of rounded defects along the nasal margin of the seam resulting in interconnected columns of cells preferentially attached to the oral epithelium. Computer generated slicing of reconstructed seams showed that groups of cells evident in cross-sections as islands at this stage of breakdown of the MES could be artifacts. Unequivocal islands of epithelial cells formed later in fusion had a rounded outline, an incomplete basal lamina and a halo of cells containing phagocytosed apoptotic debris. The pattern of breakdown indicated that the MES breaks down under tension. Laser confocal microscopy of sections and whole-mounts of palates demonstrated alpha-smooth muscle actin preferentially localized in the epithelial cells of the palatal shelves immediately before and during formation of the seam. Expression in epithelial cells of the isoform of actin normally restricted to smooth muscle cells engaged in tonic contraction supported an interpretation that the epithelial cells of the seam may be capable of generating tension during the palatal fusion event.

Midline fusion in the formation of the secondary palate anticipated by upregulation of keratin K5/6 and localized expression of vimentin mRNA in medial edge epithelium.

Secondary palatal fusion is dependent on targeted removal of the epithelium between the palatal shelves. Aseptically delivered rat embryos 15 through 18 days post coitum (dpc) were probed with DIG-labeled antisense and sense ssDNA probes for spliced exon sequences flanking intron E of cytokeratins K5/6 and spliced exon sequences flanking intron F of vimentin. Cytokeratin K5/6 expression was upregulated in the medial edge epithelium (MEE) prior to rotation of the palatal shelves and in the vomerine epithelium in the region of fusion with the palate. K5/6 expression continued in the medial epithelial seam (MES) and in epithelial islands during breakdown of the MES. Vimentin expression was not detected in the MEE prior to rotation but was specifically upregulated in the MEE following rotation and prior to midline contact and continued in the MES and in epithelial cells identifiable during the breakdown of the MES. Initiation of vimentin upregulation in the MEE prior to contact of the palatal shelves was tested by serum-free organ culture of palates from embryos at 15.5 dpc with the shelves separated by a biocompatible membrane. Vimentin upregulation occurred in the epithelium specifically in the region of anticipated contact. These results are interpreted as indicating that i) cytokeratin K5/6 expression may play a critical role in the integration of the epithelial layers of the MES to ensure subsequent merging of the mesenchyme and ii) epithelial cells in the MEE are specifically 'primed' to upregulate expression of mesenchymal genes prior to integration into and breakdown of the MES.