Loss of Epigenetic Regulator ASH2L Causes Cleft Palate in Mice

Deleting a single epigenetic regulator from the connective tissue of the developing palate is enough to produce complete cleft palate in mice, according to a new study published in Epigenetics & Chromatin. Researchers report that conditional removal of the gene Ash2l from Six2-expressing palatal mesenchymal cells caused cleft palate in every knockout animal examined, with most newborn mice showing no midline contact between the two palatal shelves at all. The work traces the defect to a collapse of the H3K4me3 histone mark, which silences key developmental genes—most notably Wnt5b—and starves palatal cells of the signals they need to proliferate and form bone.

Orofacial clefts are among the most common congenital anomalies worldwide, affecting roughly one in 700 live births. Cleft palate without cleft lip accounts for about one-third of these cases but carries some of the most severe consequences, impairing speech articulation, chewing, and swallowing. While genetic and environmental risk factors such as maternal smoking, alcohol consumption, and nutritional deficiencies are well established, the contribution of epigenetic dysregulation—particularly aberrant histone modifications—has remained poorly defined. The new study addresses this gap by focusing on ASH2L, an essential core subunit of the WRAD complex that KMT2 family enzymes require to deposit H3K4me3, an activating histone mark typically enriched at gene promoters.

Because ASH2L is expressed widely throughout the embryo, the team used a cell-type-specific strategy. They crossed mice carrying floxed Ash2l alleles with a Six2-creGFP driver line, allowing deletion specifically in palatal mesenchymal cells. SIX2 mutations cause Frontonasal Dysplasia 1 in humans, and Six2-expressing cells localize to the frontonasal mesenchyme—including the palate—during embryogenesis, with minimal presence in the mandible and nasal epithelium. Mice carrying one floxed allele, or heterozygous Ash2lflox/+ controls, developed normally with no craniofacial abnormalities, but homozygous knockouts died shortly after birth with cyanotic body surfaces, a sign of neonatal lethality.

The palatal phenotype was striking. Among knockout newborns, 65.4 percent showed complete cleft palate with no midline contact between the shelves, 26.9 percent showed partial clefting with residual defects, and 7.7 percent displayed submucous cleft palate with intact epithelium but underlying structural faults. Scanning electron microscopy confirmed prominent midline clefts in all knockout mice examined, along with widespread surface wrinkling and irregular topography compared with the smooth, rugae-patterned surfaces of controls. Histological analysis at embryonic days 13.5, 14.5, and 15.5 showed that shelf orientation was initially normal: shelves grew vertically alongside the tongue and began elevating on schedule. The failure emerged later. By E15.5, control shelves had contacted and fused at the midline, whereas knockout shelves failed to make contact entirely, and where limited contact did occur, a persistent epithelial seam prevented fusion.

Searching for the cellular basis of the defect, the researchers found that proliferation of palatal mesenchymal cells was significantly reduced in knockouts at every stage examined, as measured by EdU incorporation, while TUNEL staining showed no increase in apoptosis. The reduced cellularity therefore reflected impaired proliferation rather than cell death. Bone formation was also devastated. Immunofluorescence at E15.5 revealed markedly reduced levels of osteogenic markers including ALP, DMP1, OSX, OCN, and OPN, and RT-qPCR confirmed downregulation of genes such as Alp, Sp7, Bglap, Spp1, Col1a1, Col1a2, and Runx2. Micro-computed tomography of newborn skulls showed a near-complete absence of palatal bone, shortened and narrowed maxillary bones, and disorganized trabecular architecture in knockouts, while mandibular development remained largely unaffected and birth weights showed no significant difference from littermate controls.

The molecular mechanism centered on H3K4me3. Immunofluorescence showed sustained, widespread H3K4me3 throughout control palatal mesenchyme from E13.5 to E15.5, but a near-absent signal in knockout tissue. CUT&Tag sequencing on FACS-sorted Six2-positive mouse embryonic palatal mesenchyme cells at E14.5 confirmed that the majority of H3K4me3 peaks localize to promoters and that modification levels were significantly diminished genome-wide in knockout cells. Occupancy was markedly reduced at the promoters of critical palatogenesis genes, including Wnt5b, Wnt2, and Bmp4, and pathway analysis of genes with reduced modification pointed strongly toward the WNT signaling pathway.

RNA sequencing of the same sorted cells reinforced the picture. Among 26,501 expressed genes, the team identified 1,210 differentially expressed genes in knockouts—889 downregulated and 321 upregulated. Enrichment analyses flagged disruption of signal transduction, ion transport, metabolism, and developmental programs, including WNT signaling. Although classical proliferation pathways did not emerge as significantly enriched, several proliferation-associated genes, including Agtr1a, Epor, Lck, Plk1, Pparg, and Ret, were substantially downregulated, suggesting the proliferative defect arises indirectly through pathways such as WNT rather than direct disruption of canonical cell-cycle machinery. RT-qPCR validation confirmed decreased expression of Wnt5b, Wnt5a, Wnt2, Fzd1, Prkca, Camk2d, Nfatc2, and Ctnnb1, and immunofluorescence at E15.5 showed reduced WNT5B and FZD1 protein distribution in knockout palatal tissue.

To test whether suppressed WNT signaling was functionally responsible, the researchers treated knockout mesenchymal cells with two pharmacological agonists acting through distinct branches of the pathway: SKL2001, a canonical WNT/β-catenin agonist, and calcium chloride, which activates the WNT/Ca2+ branch. Cell viability assays showed significantly attenuated growth in knockout versus control cells at 24 and 48 hours, and both agonists significantly increased knockout cell proliferation at both time points without affecting control cells. Neither treatment, however, restored proliferation to control levels. ALP staining after seven days of osteogenic induction showed severely impaired differentiation in knockout cultures, with minimal stained area, and both agonists partially rescued this deficit without fully restoring it. The partial rescue indicates that WNT suppression is a principal—but not the sole—mechanism downstream of ASH2L-mediated H3K4me3 loss.

The identification of WNT5B as a downstream effector is a notable contribution. WNT5B primarily activates non-canonical signaling, partnering with WNT5A in the WNT/Ca2+ cascade and engaging the planar cell polarity pathway, with context-dependent effects on canonical signaling also reported. Zebrafish wnt5b mutants show convergence-extension defects and reduced palatal complex length, causative WNT5B mutations have been identified in a case of bilateral cleft lip and palate, and population studies in the United States, Poland, and China support WNT5B as a susceptibility gene for non-syndromic orofacial clefts. The authors state that their work provides the first functional evidence that WNT5B participates in mammalian palatal development, positioning it as a novel effector of the ASH2L–H3K4me3 axis.

The findings also offer a possible explanation for why ASH2L variants have not surfaced in human cleft palate cohorts. Heterozygous mice were viable and phenotypically normal with Ash2l expression comparable to wild type, while homozygous knockouts showed complete cleft palate and neonatal lethality. The authors propose that heterozygous ASH2L mutations in humans may not cause haploinsufficiency and thus may produce no recognizable phenotype, whereas homozygous or compound heterozygous mutations, if they occur, may be embryonically lethal and escape clinical detection. They caution, however, that conditional ablation of Wnt5b in palatal mesenchyme will be required to definitively establish its essential role, and the rescue experiments were conducted in cultured cells rather than in living embryos. As mouse-model work, the study establishes a mechanistic framework—epigenetic mark, gene silencing, signaling failure, failed shelf growth—rather than a therapy, though the authors suggest the ASH2L–H3K4me3–WNT5B axis identifies potential targets for future research into cleft palate.

Subject of Research: Epigenetic regulation of palatal development by ASH2L-mediated H3K4me3 and WNT5B in mice

Article Title: ASH2L-mediated H3K4me3 methylation governs palatal development by regulating WNT5B

Article References: Ling, Z., Zhu, X., Tan, X., Ma, Z., Xie, F., & Wang, J. (2026). ASH2L-mediated H3K4me3 methylation governs palatal development by regulating WNT5B. Epigenetics & Chromatin, 19(1), Article 43. https://doi.org/10.1186/s13072-026-00690-1

Image Credits: AI Generated

DOI: 10.1186/s13072-026-00690-1

Keywords: cleft palate, ASH2L, H3K4me3, WNT5B, palatogenesis, histone methylation, craniofacial development, mouse model, epigenetics, WNT signaling, osteogenic differentiation, palatal mesenchyme