The Six1 knockout homozygous (Six1 ^−/− ) and Six1 conditional knockout (Six1 ^f/f ) mouse models were generated using the CRISPR/Cas9 system on a C57BL/6J mouse background by GemPharmatech Co., Ltd (Nanjing, China). The mouse strain creation strategy involved the knockout of exon1-2 of the Six1-201 (ENSMUST00000050029.7) transcript region. Wnt1-Cre mice were obtained from the Jackson Laboratory (Bar Harbor, ME, United States). Six1 ^f/f mice were crossed with Six1 ^f/+ ; Wnt1-cre mice to generate Six1 ^f/f ; Wnt1-Cre embryos. (Supplementary Figure 1; Supplementary Table S1).

Embryos were obtained for subsequent experiments at E18.5, E16.5, and E14.5 days. The day of the appearance of a vaginal plug was defined as E0.5, and the embryos were obtained at 12:30 on each day in question. All mice were maintained and used in experiments according to the guidelines approved by the Institutional Animal Care and Use Committee (IACUC) of the Shanghai Ninth People’s Hospital affiliated to Shanghai Jiao Tong University School of Medicine.

Skin and soft tissue were carefully removed from E18.5 embryos and the embryos were treated in 95% ethanol overnight, followed by staining with Alcian blue for 48 h at 37°C. Embryos were washed twice with 95% ethanol for 2 h each, treated with 1% KOH for 1 h, and stained with Alizarin red for 2 h. The embryonic bone tissue was soaked in a gradient mixture of 1% KOH in glycerol (75%, 50%, 25%) and photographed.

The heads of the embryos were surgically isolated and fixed overnight with 4% PFA at 4°C, followed by gradient dehydration using an ethanol solution, embedded using paraffin. Hematoxylin–Eosin (HE) and Alcian blue staining were performed on 7 µm-thick paraffin sections. Immunofluorescence staining was performed with anti-Osteopontin Polyclonal antibody (22952-1-AP, Proteintech, Rosemont, IL, United States; 1:50), anti-Osterix antibody (ab209484, Abcam, Cambridge, United Kingdom; 1:200), or anti-Ki67 antibody (ab16667, Abcam; 1:100) followed by goat secondary antibody to rabbit IgG(A-11008, Thermo Fisher Scientific, Waltham, MA, United States; 1:500) following a previously described protocol (Ha et al., 2022). TUNEL staining was performed using In Situ Cell Death Detection Kit (11684795910, Roche, Mannheim, Germany) according to the manufacturer’s instructions. Nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI). Images were captured using an Olympus IX83 inverted microscope (Olympus, Tokyo, Japan).

C3H10 T1/2 cells were purchased from the Cell Bank of the Chinese Academy of Sciences, Shanghai. The cells were maintained at 37°C with 5% CO₂, and were cultured in MEM-α containing 10% FBS (10099141C, Gibco™, Thermo Fisher Scientific), 1% penicillin/streptomycin, 1% Non-Essential Amino Acids, and 1% GlutaMAX™ Supplement, and the culture medium was replaced every 2 days. Osteogenic induction medium (MUXMT-90021, Cyagen Biosciences Inc., Guangzhou, China) was used in the process of cell osteogenic differentiation (Ma et al., 2021). C3H10 T1/2 cells were seeded into 6-well plates at 50,000 cells per well and incubated at 37°C, 5% CO₂ for 24 h. Then the regular cell culture medium was replaced with osteogenic induction medium. The osteogenic induction medium was changed every 48 h. ALP staining and RNA extraction were performed 7 days after osteogenic induction. The cells were fixed for 30 min with 4% paraformaldehyde, and ALP staining was performed using the BCIP/NBT Alkaline Phosphatase color development kit following the manufacturer’s instructions (C3206, Beyotime Biotechnology, Beijing, China). Cell proliferation was analyzed using the Cell Counting Kit-8 according to the manufacturer’s instructions (C0037, Beyotime Biotechnology). Cells were seeded in a 96-well plate at a density of 1,000 cells per well and incubated at 37°C, 5% with complete MEM-α. At 24, 48, 72, 96, and 120 h, the medium was removed and added CCK-8 solution was added to the medium and incubated for 1 h. Then, the reaction solution was read with a multimode reader (BioTek, Winooski, VT, United States) to obtain the absorbance at 450 nm.

The RNA was extracted using a total RNA extraction kit (LS1040, Promega, Madison, WI, United States). Following the manufacturer’s instructions, 1 µg of total RNA was reverse transcribed into cDNA using Hifair first Strand cDNA Synthesis SuperMix (11141ES10, Yeasen Biotech, Shanghai, China) for RT-qPCR analysis. Quantitative PCR was performed on a Lightcycler 96 (Roche, Basel, Switzerland) with Hieff qPCR SYBR Green Master Mix (No Rox) (11201ES03; Yeasen Biotech). The relative expression was calculated for each gene by the 2^−ΔΔCT method, normalized against GAPDH expression, and presented as fold changes relative to the control. The sequences of all the primers used in this study are shown in Supplementary Table S2.

Short hairpin RNA (shRNA) targeting Six1 (NM_009189.3) was designed with the following sequence: GCT​CAT​GTC​CAG​CTC​AGA​AGA. The shRNA was transfected into the pLV-shRNA-EGFP(2A) Puro vector. Six1-shRNA lentiviruses were packaged and amplified by co-transfecting recombinant vector together with pSPAX2 and pMD2G into 293T cells with lipo8000 and culturing for 48 h. Then the cell culture supernatants were collected and concentrated using a Universal Virus Concentration Kit (C2901M, Beyotime Institute of Biotechnology, Jiangsu, China). The virus concentrate was added to C3H10 T1/2 cells at an MOI of 50 with polybrene(12 μg/mL) and cultured for 6 h. Subsequently, the cell medium was changed and cells were cultured for a further 72 h. To obtain stably transfected cells, C3H10 T1/2 cells were cultured in MEM-α supplemented with puromycin dihydrochloride (10 μg/mL, Beyotime Institute of Biotechnology) for 7 days.

An RNA sequencing library was prepared using a NEBNext Ultra™ RNA Library Prep Kit for Illumina and was sequenced on an Illumina novaseq6000. Differentially-expressed genes (DEGs) were determined with log2 expression fold change(log2FC) > 1 and a p-value (padj) < 0.05. Gene Ontology (GO) enrichment analysis of DEGs was performed using the clusterProfiler package in R.

Cleavage under targets and tagmentation (CUT&Tag) libraries were prepared using an In-Situ ChIP Library Prep Kit (TD901, Vazyme Biotech, Nanjing, China). Adherent cultured cells were digested with 0.25% trypsin, and 50,000 C3H10 T1/2 cells per sample were used in two biological replicates. After centrifugation at 600 g for 5 min at room temperature, the cells were washed twice with Wash Buffer. Cells were bound to conA beads and incubated with anti-Six1 antibody (#12891, CST, Danvers, MA, United States; 1:50) overnight at 4°C. The primary antibody was removed, and then a secondary antibody (Goat Anti-Rabbit IgG, Vazyme) was diluted (1:100) in DIG Wash buffer and incubated with cells at room temperature in a shaker for 1 h. Next, cells were incubated with pA-Tn5 transposon complex (0.04 µM) at room temperature for 1 h. DNA was extracted and then purified using Hieff NGS^® DNA Selection Beads (12601ES03, Yeasen Biotech). The libraries were sequenced on the Novaseq-150PE platform and 150-bp pair-end were generated. The sequencing depth was 10G base pair raw data.

GraphPad Prism v.9.0 for Windows (GraphPad Software, LaJolla, CA, United States) was used for statistical analysis. All numerical data are presented as means ± SD. Independent two-tailed Student’s t-tests were used for comparisons between two groups, and differences were considered statistically significant at a p-value < 0.05.

To explore the role of Six1 in craniofacial development, we generated Six1 knockout mice using the CRISPR/Cas9-based approach. To ensure the efficiency of Six1 knockout in Six1 ^−/− mice, we examined the RNA-seq data at E18.5 and verified the Six1 expression level by RT-qPCR (Supplementary Figure S5). By comparing gross images and skeletal staining of Six1 ^+/+ (n = 4) and Six1 ^+/− (n = 4) embryos at E18.5, we found that the heterozygous embryos had no craniofacial deformities and no differences in mandibular length, and that heterozygous mice survived and reproduced normally (Figures 1A–D). Hence, we used Six1 ^+/− to mate with each other to obtain Six1 ^−/− pups, and the Six1 ^−/− birth probability conformed to the Mendelian ratio (14/55). All the Six1 ^−/− mice that died at birth exhibited a wide range of craniofacial deformities, including microsomia, high arched palate, and a small tongue. Morphologic observation and examination of skeletal preparations at E18.5 revealed that the mandible length of Six1 ^−/− mice (n = 3) was significantly shorter than wild-type or heterozygous littermates. Analyses of HE staining of E16.5 embryos revealed that Six1 ^−/− (n = 3) mice exhibited a high palate and small tongue with ankyloglossia (Figure 1B). The volume of tongue muscle of Six1 ^−/− mice was significantly reduced. We also found that mice (1/3) exhibited bifurcated ribs, characterized by abnormal fusion between the upper and lower rib cartilage (Figure 1D). The mandibular length of Six1 ^−/− (n = 3) embryos at E18.5 was significantly shorter than that of Six1 ^+/+ (n = 4) (p < 0.0001) (Figures 1C, E). Six1 knockout mice exhibited a stable phenotype of a short jaw with 100% penetrance (14/14). The craniofacial phenotype of Six1 knockout mice was similar to that reported in the literature (Liu et al., 2019), and the skeletal deformities were also as reported in the literature (Li et al., 2003). These results demonstrate that Six1 knockout mice were successfully constructed, and this model is suitable for studying the causes of the short mandible.

The mandible is derived from neural crest cell-derived tissues (Parada and Chai, 2015). To explicitly assess whether the craniofacial defects were caused by the loss of Six1 function in CNCC-derived ectomesenchyme, we generated Six1 ^f/f mice and crossed them with Wnt1-Cre mice to conditionally knockout Six1 in CNCCs (Six1 ^f/f ; Wnt1-Cre). Wnt1-Cre pups were born in accordance with the Mendelian ratio. However, the majority of Six1 ^f/f ; Wnt1-Cre pups died at birth. By morphological analysis and examination of skeletal preparations at E18.5, we found that all Six1 ^f/f ; Wnt1-Cre (n = 3) pups exhibited microsomia compared with control littermates, and the phenotype was similar to that of Six1 knockout mice (Figure 2A).

Interestingly, a new phenotype, cleft palate (70%, 7/10) with ankyloglossia, was found in Six1 ^f/f ; Wnt1-Cre mice without atrophy of tongue muscle. We further studied the Six1 ^f/f ; Wnt1-Cre mice at E16.5 (n = 3) and E14.5 (n = 3), and found that the tongue muscle occupied the development space of the palate at E14.5, which prevented palatal lifting and led to the development of cleft palate (Figures 2C, D). Tongue connective tissue is derived from CNCCs, whereas the skeletal muscles originate from the myoblasts (Noden and Francis-West, 2006). The reduction in oral volume was due to the lack of mandibular development, while the unaffected tongue muscle volume resulted in an increased tongue muscle height. Therefore, the cleft palate phenotype of conditional knockout Six1 mice might be a secondary cleft palate. These data indicate that Six1 plays a crucial role in the growth and differentiation of CNCC-derived mesenchyme during craniofacial development.

We reasoned that Six1 deletion might disrupt the complex gene expression pattern during craniofacial development. To reveal the key genes regulated by the transcription factor Six1 during mandibular development, we surgically isolated mandibular skeletal tissues and surrounding soft tissues from E18.5 Six1 ^−/− or littermate control wild-type Six1 ^+/+ mice and performed bulk RNA-seq on two independent biological replicates for each genotype. Analysis of the RNA-seq data revealed that the Six1 transcripts were completely absent in Six1 ^−/− . Comparing the results of Six1 ^−/− and Six1 ^+/+ mice revealed that 196 genes exhibited significant expression changes (log2FC > 1, padj < 0.05). Among these, 172 genes were downregulated, and 24 genes were upregulated (Figure 3A; Supplementary Table S1). Correlation analysis of RNA-seq showed that Six1 ^−/− and Six1 ^+/+ were significantly different (Supplementary Figure S2). The uniquely mapped reads were all greater than 85%, indicating high-quality sequencing data (Supplementary Table S3). Notably, Six1 ^−/− showed significantly downregulated expression of osteogenic and mineralization genes at E18.5, including Opn, Ocn, and Osx (Figure 3B). GO enrichment analysis of downregulated DEGs showed that multiple development-related biological processes were impacted, and the DEGs were significantly enriched in “ossification”, “biomineralization”, and “biomineral tissue development” (Figure 3C; Supplementary Table S2). Using immunofluorescence staining, we further verified that the level of Opn in the Six1 ^−/− mandible was significantly lower than that in heterozygous littermates at E16.5 and E18.5 (Figure 3D). We also found a moderate downregulation in the mandibular region of Six1 ^−/− mice by Osx immunofluorescence staining, which was consistent with the RT-qPCR results (Supplementary Figure S3). Six1 ^−/− knockout mice had no significant effect on the proliferation and apoptosis of the mandible at E16.5 (Supplementary Figure S4). Collectively, these data suggest that the knockout of Six1 impaired mandibular bone formation by regulating the expression of critical genes involved in osteogenesis.

Interestingly, we also found that genes related to muscle development were significantly downregulated (Figure 3C). Observing the downregulated GO term “muscle organ development” revealed that their enriched genes include Etv1(Tenney et al., 2019), Tcap (Markert et al., 2010), Lbx1(Wang et al., 2022), Actn3 (Nicot et al., 2021), and Fos (Almada et al., 2021), which could explain the uvula deformity observed in Six1 ^−/− mice. RT-qPCR showed that Etv1 expression was significantly reduced in the mandibular tissues of Six1 ^−/− mice (Figure 3B). These results suggest that the craniofacial defects observed in Six1 ^−/− mice result from profound dysregulation of genes related to skeletal and muscle development.

To further explore the role of Six1 during mandibular osteogenesis, we performed osteogenic induction assay on the mouse embryonic mesenchymal stem cell line (C3H10 T1/2) to investigate the potential mechanisms in vitro. By performing RT-qPCR, we showed that the expression of Six1 could be readily detected in C3H10 T1/2 cells (Figure 4A). We then performed Six1 knockdown by infecting C3H10 T1/2 cells with lentivirus expressing an shRNA specifically targeting Six1, and verified that the Six1 mRNA was markedly depleted in Six1 knockdown cells (p < 0.0001). RT-qPCR analysis showed that several critical osteogenic genes, including Osx, Runx2, Alp, and Dlk1, were downregulated in Six1 knockdown cells (Figure 4B). We further compared the osteogenic differentiation capacity of control and Six1 knockdown C3H10 T1/2 cells after osteogenic induction for 7 days by quantifying alkaline phosphatase (ALP) staining as well as measuring the mRNA levels of Alp, Osx, Opn, and Ocn by RT-qPCR (Figure 4C). Six1 knockdown C3H10 T1/2 cells exhibited lower ALP activity and downregulation of Osx, Opn, and Ocn expression. The proliferation activity of Six1 knockdown C3H10 T1/2 cells was inhibited (Supplementary Figure S4). These results indicate that Six1 knockdown leads to the decline of osteogenic marker genes expression and reduced osteogenic differentiation in osteogenesis.

To investigate the underlying mechanism by which Six1 regulates osteogenic differentiation of C3H10 T1/2 cells, we analyzed the transcriptional effect of Six1 knockdown on C3H10 T1/2 cells by performing RNA sequencing on three biological replicates of control and Six1 knockdown cells. The knockdown and control groups showed a more significant correlation with each other, indicating good quality and repeatability of the RNA sequencing dataset (Figure 5B, Supplementary Figure S2). Analysis of the DEGs (log2FC > 1, padj<0.05) revealed that 662 genes were downregulated and 660 genes were upregulated in Six1 knockdown cells compared with that in control cells (Figure 5A; Supplementary Table S3). GO analysis of the downregulated DEGs showed that the knockdown of Six1 suppressed osteogenic differentiation through the regulation of biological processes associated with “ossification” and “muscle tissue development” (Figure 5C; Supplementary Table S4). We further validated the mRNA expression of several osteogenic differentiation related genes in Six1 knockdown C3H10 T1/2 cells. Consistent with the RNA-Seq results, the mRNA expression of Bmp4, Fat4, Fgf18, Fgfr2, and Runx1 significantly decreased (Figure 5D).

To further explore the mechanism by which Six1 regulates osteogenesis, we examined the genome-wide occupancy of Six1 in C3H10 T1/2 cells by performing CUT&Tag. The IDR consistency test was performed on the two sets of CUT&Tag data, and a total of 19,728 peaks were obtained (Figure 6A; Supplementary Table S5). Among the Six1 peaks, 40.26% were located in the promoter region (≤ 1 kb from the TSS), while 24.87% were located in the distal intergenic region (Figure 6B). These CUT&Tag peaks were annotated to the 10,788 closest genes. In addition, CUT&Tag assay showed that Six1 directly regulated the promoters of Bmp4, Fgfr2, Fgf18, and Fat4, all of which have been reported to play important roles in osteogenesis and were downregulated in Six1 knockdown C3H10 T1/2 cells (Figure 6C). Importantly, nearly 3/4 (2,157/3,027) of the DEGs from RNA-seq were associated with the Six1 peaks (Figure 6D). GO enrichment analysis of these 2,157 Six1-bound DEGs again showed that the ossification function was significantly enriched (Figure 6E). Six1 was also showed to bind to the promoter of Etv1, a gene involved in muscle development (Tenney et al., 2019). Taken together, our data strongly suggest that Six1 regulates the expression of a group of genes involved in bone and muscle development by binding to their promoters or cis-regulatory regions, thereby influencing craniofacial development and morphogenesis.