Face morphogenesis is promoted by Pbx-dependent EMT via regulation of Snail1 during frontonasal prominence fusion.

Human cleft lip with or without cleft palate (CL/P) is a common craniofacial abnormality caused by impaired fusion of the facial prominences. We have previously reported that, in the mouse embryo, epithelial apoptosis mediates fusion at the seam where the prominences coalesce. Here, we show that apoptosis alone is not sufficient to remove the epithelial layers. We observed morphological changes in the seam epithelia, intermingling of cells of epithelial descent into the mesenchyme and molecular signatures of epithelial-mesenchymal transition (EMT). Utilizing mouse lines with cephalic epithelium-specific Pbx loss exhibiting CL/P, we demonstrate that these cellular behaviors are Pbx dependent, as is the transcriptional regulation of the EMT driver Snail1. Furthermore, in the embryo, the majority of epithelial cells expressing high levels of Snail1 do not undergo apoptosis. Pbx1 loss- and gain-of-function in a tractable epithelial culture system revealed that Pbx1 is both necessary and sufficient for EMT induction. This study establishes that Pbx-dependent EMT programs mediate murine upper lip/primary palate morphogenesis and fusion via regulation of Snail1. Of note, the EMT signatures observed in the embryo are mirrored in the epithelial culture system.

Pbx loss in cranial neural crest, unlike in epithelium, results in cleft palate only and a broader midface.

Orofacial clefting represents the most common craniofacial birth defect. Cleft lip with or without cleft palate (CL/P) is genetically distinct from cleft palate only (CPO). Numerous transcription factors (TFs) regulate normal development of the midface, comprising the premaxilla, maxilla and palatine bones, through control of basic cellular behaviors. Within the Pbx family of genes encoding Three Amino-acid Loop Extension (TALE) homeodomain-containing TFs, we previously established that in the mouse, Pbx1 plays a preeminent role in midfacial morphogenesis, and Pbx2 and Pbx3 execute collaborative functions in domains of coexpression. We also reported that Pbx1 loss from cephalic epithelial domains, on a Pbx2- or Pbx3-deficient background, results in CL/P via disruption of a regulatory network that controls apoptosis at the seam of frontonasal and maxillary process fusion. Conversely, Pbx1 loss in cranial neural crest cell (CNCC)-derived mesenchyme on a Pbx2-deficient background results in CPO, a phenotype not yet characterized. In this study, we provide in-depth analysis of PBX1 and PBX2 protein localization from early stages of midfacial morphogenesis throughout development of the secondary palate. We further establish CNCC-specific roles of PBX TFs and describe the developmental abnormalities resulting from their loss in the murine embryonic secondary palate. Additionally, we compare and contrast the phenotypes arising from PBX1 loss in CNCC with those caused by its loss in the epithelium and show that CNCC-specific Pbx1 deletion affects only later secondary palate morphogenesis. Moreover, CNCC mutants exhibit perturbed rostro-caudal organization and broadening of the midfacial complex. Proliferation defects are pronounced in CNCC mutants at gestational day (E)12.5, suggesting altered proliferation of mutant palatal progenitor cells, consistent with roles of PBX factors in maintaining progenitor cell state. Although the craniofacial skeletal abnormalities in CNCC mutants do not result from overt patterning defects, osteogenesis is delayed, underscoring a critical role of PBX factors in CNCC morphogenesis and differentiation. Overall, the characterization of tissue-specific Pbx loss-of-function mouse models with orofacial clefting establishes these strains as unique tools to further dissect the complexities of this congenital craniofacial malformation. This study closely links PBX TALE homeodomain proteins to the variation in maxillary shape and size that occurs in pathological settings and during evolution of midfacial morphology.