Analysis of cell migration, transdifferentiation and apoptosis during mouse secondary palate fusion

Malformations in secondary palate fusion will lead to cleft palate, a common human birth defect. Palate fusion involves the formation and subsequent degeneration of the medial edge epithelial seam. The cellular mechanisms underlying seam degeneration have been a major focus in the study of palatogenesis. Three mechanisms have been proposed for seam degeneration: lateral migration of medial edge epithelial cells; epithelial-mesenchymal trans-differentiation; and apoptosis of medial edge epithelial cells. However, there is still a great deal of controversy over these proposed mechanisms. In this study, we established a [Rosa26<-->C57BL/6] chimeric culture system, in which a Rosa26-originated ;blue' palatal shelf was paired with a C57BL/6-derived ;white' palatal shelf. Using this organ culture system, we observed the migration of medial edge epithelial cells to the nasal side, but not to the oral side. We also observed an anteroposterior migration of medial edge epithelial cells, which may play an important role in posterior palate fusion. To examine epithelial-mesenchymal transdifferentiation during palate fusion, we bred a cytokeratin 14-Cre transgenic line into the R26R background. In situ hybridization showed that the Cre transgene is expressed exclusively in the epithelium. However, beta-galactosidase staining gave extensive signals in the palatal mesenchymal region during and after palate fusion, demonstrating the occurrence of an epithelial-mesenchymal transdifferentiation mechanism during palate fusion. Finally, we showed that Apaf1 mutant mouse embryos are able to complete palate fusion without DNA fragmentation-mediated programmed cell death, indicating that this is not essential for palate fusion in vivo.     

Palatogenesis: morphogenetic and molecular mechanisms of secondary palate development

Mammalian palatogenesis is a highly regulated morphogenetic process during which the embryonic primary and secondary palatal shelves develop as outgrowths from the medial nasal and maxillary prominences, respectively, remodel and fuse to form the intact roof of the oral cavity. The complexity of control of palatogenesis is reflected by the common occurrence of cleft palate in humans. Although the embryology of the palate has long been studied, the past decade has brought substantial new knowledge of the genetic control of secondary palate development. Here, we review major advances in the understanding of the morphogenetic and molecular mechanisms controlling palatal shelf growth, elevation, adhesion and fusion, and palatal bone formation.     

Bmpr1a signaling plays critical roles in palatal shelf growth and palatal bone formation

Cleft palate, including submucous cleft palate, is among the most common birth defects in humans. While overt cleft palate results from defects in growth or fusion of the developing palatal shelves, submucous cleft palate is characterized by defects in palatal bones. In this report, we show that the Bmpr1a gene, encoding a type I receptor for bone morphogenetic proteins (Bmp), is preferentially expressed in the primary palate and anterior secondary palate during palatal outgrowth. Following palatal fusion, Bmpr1a mRNA expression was upregulated in the condensed mesenchyme progenitors of palatal bone. Tissue-specific inactivation of Bmpr1a in the developing palatal mesenchyme in mice caused reduced cell proliferation in the primary and anterior secondary palate, resulting in partial cleft of the anterior palate at birth. Expression of Msx1 and Fgf10 was downregulated in the anterior palate mesenchyme and expression of Shh was downregulated in the anterior palatal epithelium in the Bmpr1a conditional mutant embryos, indicating that Bmp signaling regulates mesenchymal-epithelial interactions during palatal outgrowth. In addition, formation of the palatal processes of the maxilla was blocked while formation of the palatal processes of the palatine was significantly delayed, resulting in submucous cleft of the hard palate in the mutant mice. Our data indicate that Bmp signaling plays critical roles in the regulation of palatal mesenchyme condensation and osteoblast differentiation during palatal bone formation.     

Amniotic fluid induces rapid epithelialization in the experimentally ruptured fetal mouse palate--implications for fetal wound healing

Cleft of the secondary palate is one of the most common congenital birth defects in humans. The primary cause of cleft palate formation is a failure of fusion of bilateral palatal shelves, but rupture of the once fused palate has also been suggested to take place in utero. The possibility of post-fusion rupture of the palate in humans has hardly been accepted, mainly because in all the cleft palate cases, the cleft palatal edge is always covered with intact epithelium. To verify whether the intrauterine environment of the fetus plays roles in wound healing when the once fused palate is torn apart, we artificially tore apart fetal mouse palates after fusion and cultivated them in culture medium with or without mouse or human amniotic fluid. We thereby found that the wounded palatal edge became completely covered with flattened epithelium after 36 hours in culture with amniotic fluid, but not in culture without amniotic fluid. Using histological and scanning electron microscopic analyses of the healing process, it was revealed that the epithelium covering the wound was almost exclusively derived from the adjacent nasal epithelium, but not from the oral epithelium. Such actions of amniotic fluid on the fetal wound were never simulated by exogenous epidermal growth factor (EGF), albumin, or both. In addition, the rapid epithelialization induced by amniotic fluid was not prevented by either PD168393 (an inhibitor of the EGF receptor-specific tyrosine kinase) or SB431542 (a specific inhibitor of TGFbeta receptor type I/ALK5). The present study provides new insights into the unique biological actions of amniotic fluid in the repair of injured fetal palate.     

The inductive role of Wnt-β-Catenin signaling in the formation of oral apparatus

Proper patterning and growth of oral structures including teeth, tongue, and palate rely on epithelial–mesenchymal interactions involving coordinated regulation of signal transduction. Understanding molecular mechanisms underpinning oral–facial development will provide novel insights into the etiology of common congenital defects such as cleft palate. In this study, we report that ablating Wnt signaling in the oral epithelium blocks the formation of palatal rugae, which are a set of specialized ectodermal appendages serving as Shh signaling centers during development and niches for sensory cells and possibly neural crest related stem cells in adults. Lack of rugae is also associated with retarded anteroposterior extension of the hard palate and precocious mid-line fusion. These data implicate an obligatory role for canonical Wnt signaling in rugae development. Based on this complex phenotype, we propose that the sequential addition of rugae and its morphogen Shh, is intrinsically coupled to the elongation of the hard palate, and is critical for modulating the growth orientation of palatal shelves. In addition, we observe a unique cleft palate phenotype at the anterior end of the secondary palate, which is likely caused by the severely underdeveloped primary palate in these mutants. Last but not least, we also discover that both Wnt and Shh signalings are essential for tongue development. We provide genetic evidence that disruption of either signaling pathway results in severe microglossia. Altogether, we demonstrate a dynamic role for Wnt-β-Catenin signaling in the development of the oral apparatus.     

The fate of medial edge epithelial cells during palatal fusion in vitro: an analysis by DiI labelling and confocal microscopy.

Fusion of bilateral shelves, to form the definitive mammalian secondary palate, is critically dependent on removal of the medial edge cells that constitute the midline epithelial seam. Conflicting views suggest that programmed apoptotic death or epithelial-mesenchymal transformation of these cells is predominantly involved. Due in part to the potentially ambiguous interpretation of static images and the notable absence of fate mapping studies, the process by which this is achieved has, however, remained mechanistically equivocal. Using an in vitro mouse model, we have selectively labelled palatal epithelia with DiI and examined the fate of medial edge epithelial (MEE) cells during palatal fusion by localisation using a combination of conventional histology and confocal laser scanning microscopy (CLSM). In dynamic studies using CLSM, we have made repetitive observations of the same palatal cultures in time-course investigations. Our results concurred with the established morphological criteria of seam degeneration; however, they provided no evidence of MEE cell death or transformation. Instead we report that MEE cells migrate nasally and orally out of the seam and are recruited into, and constitute, epithelial triangles on both the oral and nasal aspects of the palate. Subsequently these cells become incorporated into the oral and nasal epithelia on the surface of the palate. We hypothesize an alternative method of seam degeneration in vivo which largely conserves the MEE population by recruiting it into the nasal and oral epithelia.     

Medial epithelial seam cell migration during palatal fusion

The mammalian secondary palate forms from two shelves of mesenchyme sheathed in a single-layered epithelium. These shelves meet during embryogenesis to form the midline epithelial seam (MES). Failure of MES degradation prevents mesenchymal confluence and results in a cleft palate. Previous studies indicated that MES cells undergo features of epithelial-to-mesenchymal transition (EMT) and may become migratory as part of the fusion mechanism. To detect MES cell movement over the course of fusion, we imaged the midline of fusing embryonic ephrin-B2/GFP mouse palates in real time using two-photon microscopy. These mice express an ephrin-B2-driven green fluorescent protein (GFP) that labels the palatal epithelium nuclei and persists in those cells through the time window necessary for fusion. We observed collective migration of MES cells toward the oral surface of the palatal shelf over 48 hr of imaging, and we confirmed histologically that the imaged palates had fused by the end of the imaged period. We previously reported that ephrin reverse signaling in the MES is required for palatal fusion. We therefore added recombinant EphA4/Fc protein to block this signaling in imaged palates. The blockage inhibited fusion, as expected, but did not change the observed migration of GFP-labeled cells. Thus, we uncoupled migration and fusion. Our data reveal that palatal MES cells undergo a collective, unidirectional movement during palatal fusion and that ephrin reverse signaling, though required for fusion, controls aspects of the fusion mechanism independent of migration.     

Overexpression of Smad2 in Tgf-beta3-null mutant mice rescues cleft palate

Transforming growth factor (TGF)-beta3 is an important contributor to the regulation of medial edge epithelium (MEE) disappearance during palatal fusion. SMAD2 phosphorylation in the MEE has been shown to be directly regulated by TGF-beta3. No phospho-SMAD2 was identified in the MEE in Tgf-beta3-null mutant mice (Tgf-beta3-/-), which was correlated with the persistence of the MEE and failure of palatal fusion. In the present study, the cleft palate phenotype in Tgf-beta3-/- mice was rescued by overexpression of a Smad2 transgene in Keratin 14-synthesizing MEE cells following mating Tgf-beta3 heterozygous mice with Keratin 14 promoter directed Smad2 transgenic mice (K14-Smad2). Success of the rescue could be attributed to the elevated phospho-SMAD2 level in the MEE, demonstrated by two indirect evidences. The rescued palatal fusion in Tgf-beta3-/-/K14-Smad2 mice, however, never proceeded to the junction of primary and secondary palates and the most posterior border of the soft palate, despite phospho-SMAD2 expression in these regions at the same level as in the middle portion of the secondary palate. The K14-Smad2 transgene was unable to restore all the functional outcomes of TGF-beta3. This may indicate an anterior-posterior patterning in the palatal shelves with respect to TGF-beta3 signaling and the mechanism of secondary palatal fusion.     

Differentiation of the avian secondary palate

The avian secondary palate exhibits the unique feature of a midline cleft. Cryostat sections indicated that although extensive contact between homologous shelves was present, chick palatal medial edge epithelium (MEE) failed to fuse. The failure of fusion and subsequent clefting of the avian palate were correlated with continued proliferation of the avian MEE, a failure of selective MEE cell death, and an absence of elevated levels of intracellular cAMP. Moreover, immunohistochemical staining for cAMP and microspectrophotometric quantitation of staining intensity indicated that staining of chick MEE was significantly (p less than .01) less than murine MEE at comparable gestational ages. These data indicate that differentiation of the avian secondary palate is fundamentally different than reported for the mammalian palate in that many developmental events known to be associated with normal mammalian palate formation (cessation of MEE proliferation, MEE cell death, elevated levels of MEE cAMP) fail to occur in the chick. The developing avian secondary palate, with its midline cleft, thus provides an interesting and useful model system with which to compare mammalian palate formation where the palate is normally fused in the midline.     

The Etiology of Cleft Palate Formation in BMP7-Deficient Mice

Palatogenesis is a complex process implying growth, elevation and fusion of the two lateral palatal shelves during embryogenesis. This process is tightly controlled by genetic and mechanistic cues that also coordinate the growth of other orofacial structures. Failure at any of these steps can result in cleft palate, which is a frequent craniofacial malformation in humans. To understand the etiology of cleft palate linked to the BMP signaling pathway, we studied palatogenesis in Bmp7-deficient mouse embryos. Bmp7 expression was found in several orofacial structures including the edges of the palatal shelves prior and during their fusion. Bmp7 deletion resulted in a general alteration of oral cavity morphology, unpaired palatal shelf elevation, delayed shelf approximation, and subsequent lack of fusion. Cell proliferation and expression of specific genes involved in palatogenesis were not altered in Bmp7-deficient embryos. Conditional ablation of Bmp7 with Keratin14-Cre or Wnt1-Cre revealed that neither epithelial nor neural crest-specific loss of Bmp7 alone could recapitulate the cleft palate phenotype. Palatal shelves from mutant embryos were able to fuse when cultured in vitro as isolated shelves in proximity, but not when cultured as whole upper jaw explants. Thus, deformations in the oral cavity of Bmp7-deficient embryos such as the shorter and wider mandible were not solely responsible for cleft palate formation. These findings indicate a requirement for Bmp7 for the coordination of both developmental and mechanistic aspects of palatogenesis.     

Palatogenesis in hamster. II. Ultrastructural observations on the closure of palate

Formation of secondary palate in hamster was studied with electron microscopy. Prior to assuming horizontal position, the palatal shelves were covered by a two to three cell layer thick epithelium which was separated from the underlying mesenchyme by an intact basal lamina. Epithelial cells were attached to each other by desmosomes. Early hemidesmosomes could be identified as thickenings of the cytoplasmic membrane opposing the basal lamina. Epithelial cells, like other embryonic cells, contained only few organelles but were rich in polyribosomes. As the horizontal shelves approached each other towards the midline, lysosomes and tonofilaments appeared in the superficial and basal cells of the epithelia. Superficial cells showed degeneration and eventual lysis. Fusion of the opposing epithelia occurred between the deeper cells by means of newly formed desmosomes. The epithelial seam resulting from fusion of the epithelia was limited on each side by a continuous basal lamina. Its subsequent thining and eventual fragmentation resulted from the loss of cells by autophagy. There was no evidence of mesenchymal invasion of the epithelial seam. Mesenchymal macrophages appeared in the later stage of palatogenesis and were responsible for phagocytosis of cellular debris. Formation of the soft palate was basically similar to that of the secondary hard palate and occurred by fusion of the opposing shelves. Similarly, anterior closure of the palate occurred by fusion of the lower end of the nasal septum to the primary and secondary palates. Hyperplasia of the opposing epithelia, prior to their fusion, was often seen. It is suggested that formation of the palate occurs in predictable and coordinated fashion and that timely appearance of lysosomes causing lysis of intervening epithelia is of great significance in normal palatogenesis.     

Regulation of the Epithelial Adhesion Molecule CEACAM1 Is Important for Palate Formation

Cleft palate results from a mixture of genetic and environmental factors and occurs when the bilateral palatal shelves fail to fuse. The objective of this study was to search for new genes involved in mouse palate formation. Gene expression of murine embryonic palatal tissue was analyzed at various developmental stages before, during, and after palate fusion using GeneChip® microarrays. Ceacam1 was one of the highly up-regulated genes during palate formation, and this was confirmed by quantitative real-time PCR. Immunohistochemical staining showed that CEACAM1 was present in prefusion palatal epithelium and was degraded during fusion. To investigate the developmental role of CEACAM1, function-blocking antibody was added to embryonic mouse palate in organ culture. Palatal fusion was inhibited by this function-blocking antibody. To investigate the subsequent developmental role of CEACAM1, we characterized Ceacam1-deficient (Ceacam1 ^−/−) mice. Epithelial cells persisted abnormally at the midline of the embryonic palate even on day E16.0, and palatal fusion was delayed in Ceacam1 ^−/− mice. TGFβ3 expression, apoptosis, and cell proliferation in palatal epithelium were not affected in the palate of Ceacam1^−/−mice. However, CEACAM1 expression was retained in the remaining MEE of TGFβ-deficient mice. These results suggest that CEACAM1 has roles in the initiation of palatal fusion via epithelial cell adhesion.     

Cell autonomous requirement for Tgfbr2 in the disappearance of medial edge epithelium during palatal fusion

Palatal fusion is a complex, multi-step developmental process; the consequence of failure in this process is cleft palate, one of the most common birth defects in humans. Previous studies have shown that regression of the medial edge epithelium (MEE) upon palatal fusion is required for this process, and TGF-beta signaling plays an important role in regulating palatal fusion. However, the fate of the MEE and the mechanisms underlying its disappearance are still unclear. By using the Cre/lox system, we are able to label the MEE genetically and to ablate Tgfbr2 specifically in the palatal epithelial cells. Our results indicate that epithelial-mesenchymal transformation does not occur in the regression of MEE cells. Ablation of Tgfbr2 in the palatal epithelial cells causes soft palate cleft, submucosal cleft and failure of the primary palate to fuse with the secondary palate. Whereas wild-type MEE cells disappear, the mutant MEE cells continue to proliferate and form cysts and epithelial bridges in the midline of the palate. Our study provides for the first time an animal model for soft palate cleft and submucous cleft. At the molecular level, Tgfb3 and Irf6 have similar expression patterns in the MEE. Mutations in IRF6 disrupt orofacial development and cause cleft palate in humans. We show here that Irf6 expression is downregulated in the MEE of the Tgfbr2 mutant. As a recent study shows that heterozygous mutations in TGFBR1 or TGFBR2 cause multiple human congenital malformations, including soft palate cleft, we propose that TGF-beta mediated Irf6 expression plays an important, cell-autonomous role in regulating the fate of MEE cells during palatogenesis in both mice and humans.     

Signaling through Tgf-beta type I receptor Alk5 is required for upper lip fusion

Cleft lip with or without cleft palate is one of the most common congenital malformations in newborns. While numerous studies on secondary palatogenesis exist, data regarding normal upper lip formation and cleft lip is limited. We previously showed that conditional inactivation of Tgf-beta type I receptor Alk5 in the ectomesenchyme resulted in total facial clefting. While the role of Tgf-beta signaling in palatal fusion is relatively well understood, its role in upper lip fusion remains unknown. In order to investigate a role for Tgf-beta signaling in upper lip formation, we used the Nes-Cre transgenic mouse line to delete the Alk5 gene in developing facial prominences. We show that Alk5/Nes-Cre mutants display incompletely penetrant unilateral or bilateral cleft lip. Increased cell death seen in the medial nasal process and the maxillary process may explain the hypoplastic maxillary process observed in mutants. The resultant reduced contact is insufficient for normal lip fusion leading to cleft lip. These mice also display retarded development of palatal shelves and die at E15. Our findings support a role for Alk5 in normal upper lip formation not previously reported.     

Temporal and Spatial Expression of Hoxa-2 During Murine Palatogenesis

1. Mice homozygous for a targeted mutation of the Hoxa-2 gene are born with a bilateral cleft of the secondary palate associated with multiple head and cranial anomalies and these animals die within 24 hr of birth (Gendron-Maguire et al., 1993; Rijli et al., 1993; Mallo and Gridley, 1996). We have determined the spatial and temporal expression of the Hoxa-2 homeobox protein in the developing mouse palate at embryonic stages E12, E13, E13.5, E14, E14.5, and E15.2. Hoxa-2 is expressed in the mesenchyme and epithelial cells of the palate at E12, but is progressively restricted to the tips of the growing palatal shelves at E13.3. By the E13.5 stage of development, Hoxa-2 protein was found to be expressed throughout the palatal shelf. These observations correlate with palatal shelf orientation and Hoxa-2 protein may play a direct or indirect role in guiding the palatal shelves vertically along side the tongue, starting with the tips of the palatal shelves at E13, followed by the entire palatal shelf at E13.5.4. As development progresses to E14, the stage at which shelf elevation occurs, Hoxa-2 protein is downregulated in the palatal mesenchyme but remains in the medial edge epithelium. Expression of Hoxa-2 continues in the medial edge epithelium until the fusion of opposing palatal shelves.5. By the E15 stage of development, Hoxa-2 is downregulated in the palate and expression is localized in the nasal and oral epithelia.6. In an animal model of phenytoin-induced cleft palate, we report that Hoxa-2 mRNA and protein expression were significantly decreased, implicating a possible functional role of the Hoxa-2 gene in the development of phenytoin-induced cleft palate.7. A recent report by Barrow and Capecchi (1999), has illustrated the importance of tongue posture during palatal shelf closure in Hoxa-2 mutant mice. This along with our new findings of the expression of the Hoxa-2 protein during palatogenesis has shed some light on the putative role of this gene in palate development.     

Strain differences between C57BL/6 and SWV mice in time of palate closure and induction of palatal slit and cleft palate

Palatal slit, which occurs spontaneously in C57BL/6 (C57BL) mice, is increased in frequency among C57BL fetuses from dams treated with triamcinolone acetonide, but is not induced in SWV fetuses. On the other hand, C57BL is more resistant than SWV to cleft palate induction by triamcinolone. Using these C57BL and SWV mice, the relation of palate stage and chronological age was examined from 1 P.M. on day 14 to 9 A.M. on day 16 in untreated embryos, and the condition of the palate after triamcinolone treatment on day 12 was examined at 9 A.M. on day 16. In untreated embryos, horizontalization and fusion of the palatal shelves occurred earlier in C57BL than in SWV embryos, but fusion of the primary palate with the secondary palate occurred later. After triamcinolone treatment, the development of the palate was delayed in both C57BL and SWV embryos. These results suggest that the times of normal palate closure are related to the differences between C57BL and SWV mice in their susceptibilities to palatal slit and cleft palate induction and that triamcinolone produces palatal slit and cleft palate by delaying palate closure.     

Shox2-deficient mice exhibit a rare type of incomplete clefting of the secondary palate

The short stature homeobox gene SHOX is associated with idiopathic short stature in humans, as seen in Turner syndrome and Leri-Weill dyschondrosteosis, while little is known about its close relative SHOX2. We report the restricted expression of Shox2 in the anterior domain of the secondary palate in mice and humans. Shox2-/- mice develop an incomplete cleft that is confined to the anterior region of the palate, an extremely rare type of clefting in humans. The Shox2-/- palatal shelves initiate, grow and elevate normally, but the anterior region fails to contact and fuse at the midline, owing to altered cell proliferation and apoptosis, leading to incomplete clefting within the presumptive hard palate. Accompanied with these cellular alterations is an ectopic expression of Fgf10 and Fgfr2c in the anterior palatal mesenchyme of the mutants. Tissue recombination and bead implantation experiments revealed that signals from the anterior palatal epithelium are responsible for the restricted mesenchymal Shox2 expression. BMP activity is necessary but not sufficient for the induction of palatal Shox2 expression. Our results demonstrate an intrinsic requirement for Shox2 in palatogenesis, and support the idea that palatogenesis is differentially regulated along the anteroposterior axis. Furthermore, our results demonstrate that fusion of the posterior palate can occur independently of fusion in the anterior palate.     

The nasopalatine ducts and associated structures in the rhesus monkey (Macaca mulatta): topography, prenatal development, function, and phylogeny

Morphological and developmental characteristics of the rhesus monkey nasopalatine duct system and associated primary palatal structures are described along with functional and phylogenetic considerations. Examination of five adult palates and coronal sections of 13 fetal palates together with dissections of a sixth adult specimen and of a 119-day-old fetal palate reveal that the lateral lobes of the tripartate incisive papilla cover clefts leading into the ducts. The ducts pierce the bony palate to enter the nasal fossae in proximity to the incisive suture. The ontogenetic stability of the duct path reflects the retention of ancient duct and primitive choanae relationships and functionally maintains an optimal oral odorant-to-receptor channel. Sixteen timed pregnancy specimens (35-100 days) provided histological material for documenting rostral nasopalatal development. Duct primordia, identified at 35 days, had by 40 days formed the medial duct walls (conjoined septum-papilla from the primary medial palatal component), the lateral duct walls (maxillary processes), and the rostral walls (fused maxillary-intermaxillary components). The caudal walls derive from the fusion of palatal shelves with the papilla (45 days), thus distinguishing primary and secondary fusion modes. Duct epithelial maturation occurs between 70 and 100 days. The absence of a vomeronasal system is attributed to reduction of olfaction in reproductive behavior, while the presence of the coevolved nasopalatine ducts is linked to the persistence of epiglottal-velar valving. The ducts serve as oral food-odor conduits in otherwise functionally separated respiratory and digestive tracts.     

The fate of the expected fusion zone in rat fetuses with experimentally-induced cleft palate—An ultrastructural study

The ultrastructure of palatal processes from rat fetuses with cleft palate induced by Meclozine or amniotic sac puncture was studied from gestation days 16 to 20. Degenerative changes involving the cytoplasm of epithelial cells of the expected fusion zone began on Day 16, when palatal fusion occurs normally in the midline seam in untreated fetuses. By Day 17 the cells in the same region showed more advanced degeneration, degraded cytoplasm and occasional dead epithelial cells appearing to be removed via intercellular extrusion or by macrophages. In the expected fusion zone, the basal lamina became discontinuous while in the oral and nasal zones of the palatal processes a firm epithelium-connective tissue attachment developed. As gestation progressed, the zone became narrower until, by Day 20, no degenerating cells remained. Replacement of lost epithelium appeared to have occurred from the nasal side, the termination of the keratinizing oral epithelium having the same location throughout the period studied. The results support the concept of programmed cell degeneration in palatal processes but indicate that this event has the potential to extend over a relatively long period of fetal life.