Circadian BMAL1 regulates mandibular condyle development by hedgehog pathway

Abstract Objective Chondrogenesis and endochondral ossification in mandibular condyle play crucial roles in maxillofacial morphogenesis and function. Circadian regulator brain and muscle arnt‐like 1 (BMAL1) is proven to be essential for embryonic and postnatal development. The goal of this study was to define the functions of BMAL1 in the embryonic and postnatal growth of mandibular condylar cartilages (MCC). Materials and Methods Micro‐CT, TUNEL staining and EdU assay were performed using BMAL1‐deficient mice model, and in vitro experiments were performed using rat chondrocytes isolated from MCC. RNA sequencing in mandibular condyle tissues from Bmal1 ‐/‐ mice and the age‐matched wild‐type mice was used for transcriptional profiling at different postnatal stages. Results The expression levels of BMAL1 decrease gradually in MCC. BMAL1 is proved to regulate sequential chondrocyte differentiation, and its deficiency can result in the impairment of endochondral ossification of MCC. RNA sequencing reveals hedgehog signalling pathway is the potential target of BMAL1. BMAL1 regulates hedgehog signalling and affects its downstream cascades through directly binding to the promoters of Ptch1 and Ihh, modulating targets of hedgehog signalling which is indispensable for endochondral ossification. Importantly, the short stature phenotypes caused by BMAL1 deficiency can be rescued by hedgehog signalling activator. Conclusions Collectively, these results indicate that BMAL1 plays critical roles on chondrogenesis and endochondral ossification of MCC, giving a new insight on potential therapeutic strategies for facial dysmorphism.


| INTRODUC TI ON
Mandibular condyle plays a pivotal role in the development of oro-facial complex, since it functions as a regional adaptive growth site. 1 Mandibular condylar cartilage (MCC) caps the surface of the mandibular condyle and serves as a site of articulation with the mandible, as well as a locus for chondrogenesis and endochondral ossification that contributes to morphogenesis and growth of the mandible in length and height. 2 Unlike other articular cartilage within limb joints, MCC is considered as a secondary cartilage and its growth is categorized into two phases, maturation and mineralization. 3,4 Deviations in the growth of MCC affect the function of occlusion, leading to severe facial dysmorphism. 5 The growth process of MCC is not only strictly regulated by a variety of local autocrine/paracrine factors including intermittent parathyroid hormone (I-PTH) and insulin-like growth factors (IGFs), but also under the control of complex transcriptional networks, such as transforming growth factor β (TGF-β)/bone morphogenic protein (BMP) and circadian clock transcription outputs. 1,6,7 The circadian clock is an endogenous rhythm and is of considerable interest due to its potential effects on tissue growth and structural development. 8,9 The transcription factor brain and muscle arnt-like 1 (BMAL1) is the core driver of the circadian pacemaking in mammals and orchestrates organisms to adapt to the 24-hours light-dark cycle. [10][11][12] Accumulating evidences suggest that BMAL1 plays pivotal roles in embryonic and postnatal development, including oocyte fertilization, follicle development, angiogenesis and neurodevelopment. [13][14][15][16] Our recent study has indicated that BMAL1 controls osteogenic differentiation and osteoclast differentiation in mandible. 17 However, the roles of BMAL1 in chondrogenesis and endochondral ossification of mandibular condyle and its molecular mechanism are not been explained clearly.
In this study, we have verified that BMAL1 plays vital roles in chondrogenesis and endochondral ossification of mandibular. Our findings provide new insights into the pathogenesis of deformities in MCC and present potential therapeutic strategies for facial dysmorphism.

| Micro-CT
The mandibles of experimental mice were harvested and fixed.
Scanning was performed at a voxel size of 9 μm by Micro-CT (SkyScan 1176, Broker). The images were used to reconstruct the tomography scans and quantify. Bone mass was evaluated according to bone mineral density (BMD), bone volume/total volume (BV/TV), trabecular thickness (Tb.Th), bone surface area/bone volume (BS/ BV), trabecular number (Tb.N) and trabecular separation (Tb.Sp).

| RNA sequencing
Total RNA was extracted from distally mandibular condyle tissues of Bmal1 -/mice and wild-type mice (each 3 pairs of 4-, 6-, 8-and 10-week-old). The samples were sent to BGI Tech. for quantification, RNA-Seq library construction, sequencing and bioinformatics analysis (Appendix S1).

| Other methods
Alcian blue and alizarin red staining, Edu staining assay, etc, were performed. Please see the Appendix S1 for more details.

| Statistical analysis
Unless otherwise stated, all data were shown as mean ± standard deviation. The GraphPad Prism 7.0 software was used for statistical analysis. Statistically significant differences were determined by Student's t test or ANOVA. A P-value <.05 was considered with statistical significance.

| Circadian regulator BMAL1 decreases gradually in the development of MCC
To determine the functions of circadian clock in the growth of MCC, we measured the expression patterns and levels of core clock genes in MCC from various week-old mice. We revealed that the phases of circadian rhythms were consistent across different developmental stages ( Figure 1A). Meanwhile, we detected that the expression levels of circadian regulator BMAL1 decreased gradually in MCC, rather than other circadian proteins, including circadian locomotor output cycles kaput (CLOCK), period homologue drosophila 1/2 (PER1/2),  Figure S1).
Traditionally, the thicknesses of proliferative layer zone (PZ) and hypertrophic layer zone (HZ) in MCC is gradually decreased in wildtype mice ( Figure 2E-G). These results suggested that circadian regulator BMAL1 is closely correlated with the development of MCC.

| Loss of BMAL1 delays and reduces chondrogenesis and endochondral ossification in mandibular condyle
To further confirm the role of BMAL1 in development and growth of MCC, we constructed global BMAL1-deficiency (Bmal1 -/-) mice ( Figure   S2A). Here, we observed that the chondrogenesis of Bmal1 -/mice was slower and lesser compared with wild-type mice at E14.5, E16.5 and E18.5 stages ( Figure S2B, C). Consistently, the mandibular condyles of Bmal1 -/mice were smaller and shorter in height than those of the age-matched wild-type mice (Figure 2A). Micro-CT analysis showed that the bone mineral density (BMD), bone volume/total volume (BV/ TV), trabecular thickness (Tb.Th) and trabecular number (Tb.N) of Bmal1 -/mice significantly decreased and trabecular separation (Tb. Sp) increased relative to those wild-type mice in the mandibular condylar regions ( Figure 2B-D). Morphological evaluation of MCC further verified that the height of PZ and HZ was diminished obviously, and the number of chondrocytes per column in PZ was decreased significantly compared to the age-matched wild-type mice ( Figure 2E-G). The basic components of cartilage matrix type II collagen (COL2α1), aggrecan (ACAN) and type X collagen (COL10α1) were dramatically less in Bmal1 -/mice at prenatal and postnatal development stage ( Figure 2H and Figure S2D). Immunohistochemical staining for Ki67 revealed that chondrocyte proliferation was significantly attenuated in PZ ( Figure 2H and Figure S2E). Moreover, TUNEL staining showed that cell apoptosis of chondrocyte was increased in MCC of Bmal1 -/mice compared with the age-matched wild-type mice ( Figure 2I and Figure S2F).
To test whether BMAL1 has intrinsic functionality in chondrocyte proliferation and cartilage matrix productions, we constructed conditional Bmal1 knockout mice in mesenchymal stem cells (MSCs) (Bmal1 fl/fl ; Twist2-Cre) ( Figure 3A). The results showed that the chondrogenesis and endochondral ossification of the conditional Bmal1 knockout mice were significantly reduced, suggesting that BMAL1 in MSCs was closely associated with chondrogenesis and endochondral ossification ( Figure 3B). To further verify the function of BMAL1 in chondrocytes, we extracted primary chondrocytes from mandibular condylar tissues of wild-type rat and knocked down Bmal1 using CRISPR/CAS9 system ( Figure 3C). We found that BMAL1 deficiency Collectively, these findings verified that BMAL1 deficiency inhibits chondrogenesis and endochondral ossification in mandibular condyle by reducing sequential chondrocyte differentiation.

| Along with the development of mandibular condyle, the transcriptional effects of BMAL1 in MCC are gradually decreased
BMAL1 is a widely acknowledged transcriptional activator. 19 To determine the transcriptional effects of BMAL1 on the development of MCC, we performed a genome-wide RNA sequencing to acquire the transcriptional profile in mandibular condyles from 4-, 6-, 8-and 10-weekold Bmal1 -/mice and the age-matched wild-type mice ( Figure 4A).  Figure S3). Importantly, we found that the number of differentially expressed genes (DEGs) in the group of 4 weeks of age was the largest, with 206 differential genes, and the 10 weeks of age was the least, with 58 differential genes ( Figure 4B). GO analysis revealed that many DEGs were enriched in development process, differentiation and cell F I G U R E 2 Loss of BMAL1 delays and reduces chondrogenesis and endochondral ossification in mandibular condyle. A, Representative images of micro-CT reconstruction of mandibles (n = 3 per group). Red arrow indicated the mandibular condyle length. Scale bar, 2 mm. B-D, Representative images and analysis of length, BMD, BV/TV, Tb.N, Tb.sp and Tb.Th of mandibular condyles (n = 3 for each bar). Scale bar, 1 mm. Data represented as mean ± SD, * P < .05, ** P < .01, *** P < .001. E, F, Representative images of H&E and S-O staining of mandibular condyles at 4-wk-old and 8-wk-old (n = 3 per group). Scale bar, 20 μm. G, The height and the number of chondrocytes per column in PZ and HZ of mandibular condyle cartilages (n = 3 for each bar). Data represented as mean ± SD, *** P < .001. H, Immunohistochemistry of COL2α1, ACAN, COL10α1 and Ki67 in mandibular condyles at ZT10, without primary antibody as negative group (n = 3 per group). Scale bar, 20 μm. I, The cell apoptosis of mandibular condyles was detected by TUNEL staining (n = 3 per group). Scale bar, 50 μm death. Remarkably, the number of DEGs in development and differentiation during prepuberty and puberty (4 and 6 weeks of age) was almost twice as many as in young adulthood (8 and 10 weeks of age; Figure 4C). Then, we focused on the candidate gene set that related to chondrogenesis and endochondral ossification, including Gli1, Col2α1 and Col10α1. We found that these genes had the similar spatio-temporal expression patterns to Bmal1, with a higher expression level during prepuberty and puberty ( Figure 4D). Moreover, the expression changes of these genes after knocking down Bmal1 were greater during prepuberty and puberty than young adulthood ( Figure 4D). These results verified that the transcriptional effects of BMAL1 in MCC are gradually decreased which is along with the development of mandibular condyle.

| Hedgehog signalling is under the strict control of BMAL1 and is critical for endochondral ossification
Due to the higher expression levels of Bmal1 and the genes related to chondrogenesis and endochondral ossification in prepuberty and  20 To verify the effects of Hh signalling in F I G U R E 4 Along with the development of mandibular condyle, the transcriptional effects of BMAL1 in MCC are gradually decreased. A, Scatter plot of differentially expressed genes (DEGs) in Bmal1 -/and the age-matched control wild-type mice (n = 3 pairs per group). B, C, DEGs and associated Gene Ontology (GO) across 4, 6, 8 and 10 wk of Bmal1 -/and wild-type mice. D, Heatmap showed some DEGs that were related to cartilage development. E, Confirmation of the DEGs in the MCC from 4-wk-old wild-type and Bmal1 -/mice by qRT-PCR analysis (n = 3 for each bar). Data represented as mean ± SD, * P < .05, ** P < .01, *** P < .001. F, Immunohistochemistry of IHH, PTCH1 and GLI1 according to these studies [45][46][47] in the MCC from 4-week-old wild-type and Bmal1 -/mice (n = 3 per group), without primary antibody as negative group. Scale bar, 20 μm and bcl-2-like protein 1 (BCL-XL), chondrocyte proliferation markers SOX9 and CyclinD1 increased with SAG supplement ( Figure S4E, F), while the expression levels of all above genes decreased with GDC-0449 treatment ( Figure S4G, H). These results suggested that the loss of BMAL1 may decrease chondrocyte proliferation and cartilage matrix production by downregulating Hh signalling.

| BMAL1 regulates sequential chondrocyte differentiation through directly activating Ptch1 and Ihh transcription
As a classical transcription activator, BMAL1 plays its roles by combining the E-box site in the promoter of downstream target genes and activating their transcription. 11 To further clarify the underlying mechanisms in which BMAL1 controls Hh signalling, the ChIP assays were employed to examine BMAL1-binding site within Ptch1 and Ihh promoter regions.
Our results indicated that BMAL1 selectively bound to the promoter (−1800/-1542, 259 bp) of Ptch1 ( Figure 5A). We also found that BMAL1 could bind to Ihh promoter region which was identical with the similar report 21 ( Figure 5B). Dual-luciferase assay further verified that BMAL1 improved the activity of Ptch1 and Ihh promoters, but not those of the mutants with BMAL1-binding site being either deleted or mutated. In addition, we found that BMAL1 overexpression could further improve the activity of both Ptch1 and Ihh promoters ( Figure 5C Figure S5). Taken together, these findings indicated that the loss of BMAL1 inhibits sequential chondrocyte differentiation through directly activating Ptch1 and Ihh transcription.

| Phenotypes caused by BMAL1 deficiency can be rescued by Hh signalling activator during prepuberty and early puberty periods
To determine whether activating Hh signalling can rescue the short malformation phenotypes of mandible caused by BMAL1 deficiency, SAG was injected intraperitoneally into Bmal1 -/mice during the prepuberty and early puberty periods (2 weeks of age) and young adulthood (8 weeks of age), respectively (20μg/g, three times a week, 4 weeks). Immunohistochemical staining for GLI1 demonstrated significant activation of the Hh pathway as defined by increased expression of GLI1 after the application of SAG ( Figure 6A). Micro-CT results showed that skeletal development in the puberty-SAG-injected Bmal1 -/mice was substantially ameliorative (Figure 6B-E).
H&E staining showed a significant increase in the number of PZ chondrocytes ( Figure 6F and Figure S6A). Furthermore, there were more Ki67-positive chondrocytes in PZ and less TUNEL-positive chondrocytes in MCC from the Bmal1 -/mice at puberty-SAG-injected ( Figure 6G, H). Consistently, the basic components of cartilage matrix COL2α1 were significantly increased after injecting SAG ( Figure S6C). However, young adulthood-SAG-injected Bmal1 -/mice did not response effectively ( Figure 6B-E and Figure S6B). Taken together, our results revealed that Hh signalling activator plays potential roles in improving bone mass reduction caused by BMAL1 deficiency, which was effective in prepuberty and early puberty periods intervention, rather than young adulthood treatment. These findings provide a new strategy for the clinical treatment of facial dysmorphism.

| D ISCUSS I ON
BMAL1 controls limb bone and knee cartilage homeostasis and maintains structural integrity, and its aberrant expression leads to osteopenia and predisposes knee cartilage to osteoarthritis-like damage. 21,22 William found that BMAL1-deficient mice display an earlier closure of growth plates than wild-type mice. 23 Takarada et al also found the body length measured from nose to anus and the longitudinal lengths of both tibia and femur decreased remarkably in BMAL1-deficient mice than wild-type mice. 21 Our previous study also demonstrated that osteogenic differentiation is reduced in limb and mandible from Bmal1 -/mice. 17 Unlike the limb bone and maxillary and mandibular body which mainly rely on intramembranous osteogenesis, MCC is considered as a secondary cartilage and has significant structural differences from primary cartilages. 24 F I G U R E 6 Phenotypes caused by BMAL1 deficiency are rescued by Hh signalling activator during prepuberty and early puberty periods. A, Immunohistochemistry of GLI1 in mandibular condyles at ZT10, without primary antibody as negative group. Scale bar, 20 μm. B, Images of micro-CT reconstruction of the mandibles of wild-type and Bmal1 -/mice with or without injection of SAG at 2 wk or 8 wk, 20 μg/g, three times a week for 4 wk (n = 3 per group). Red arrow indicated the mandibular condyle length. Scale bar, 2 mm. C-E, Representative images and analysis of length, BMD, BV/TV, Tb.N, Tb.sp and Tb.Th of mandibular condyles (n = 3 for each bar). Scale bar, 1 mm. Data represented as mean ± SD, * P < .05, ** P < .01. F, The height and the number of chondrocytes per column in PZ and HZ of mandibular condyle cartilages (n = 3 for each bar). Data represented as mean ± SD, *** P < .001. G and H, The cell proliferation and cell apoptosis in mandibular condyles were detected by Ki67 and TUNEL staining at ZT10 (n = 3 per group). Scale bar, 20 μm, 50 μm In our study, we found the body size was apparently smaller in BMAL1-deficient mice than that in wild-type mice as reported previously. 21,25,26 Importantly, we observed that foetuses in BMAL1deficient mice appear insufficient cranial cartilage calcification, and smaller and short mandibular condyle phenotypes.
Although the effects of BMAL1 on osteoblast differentiation and endochondral ossification of growth plate have been well established, its roles and the underlying mechanism in the regulation of chondrogenesis and endochondral ossification in MCC have not been explained explicitly. Here, we identified BMAL1 constitutively expressed in chondrocytes controls the sequential chondrocyte differentiation process in MCC through a mechanism related to regulation of the transactivation of Ptch1 and Ihh. 21  In this report, we verified that PTCH1 overexpression partially rescues Hh signalling in BMAL1-deficient chondrocytes. PTCH1 is essential for activating Hh downstream signalling, and Hh activation depends on the binding of hedgehog and PTCH1. Unbound PTCH1 will suppress Hh signalling in turn. 33,34 When Hh is sufficient, PTCH1 mild overexpression can recruit much more Hh from the extracellular matrix and then activate Hh signalling greatly. 35,36 Besides, the effect of PTCH1 also has tissue specificity, which might explain the reason that the mild overexpression of PTCH1 results in the "rescue" effect in BMAL1-deficient chondrocytes in this work. 37,38 As we know, SAG cannot directly increase the expression of IHH.
According to literature, the transcription factor Gli1 in hedgehog pathway enhances the direct induction of both Runx2 and Runx3 transcription and RUNX2 can subsequently regulate chondrocyte maturation and proliferation through the induction of Ihh expression. [39][40][41] This might be the reason that SAG increases Ihh expression though GLI1-RUNX2-Ihh signal axis in chondrocytes.
Orthopedic or orthodontic operation is the main clinical strategy to remedy facial dysmorphism. 42 But, the operation is complicated and the side effects are serious. Due to its complex aetiology, it is indeed difficult to develop an effective strategy to prevent the deformities. SAG, a specific Hh signalling activator, is reported to enhance bone healing in a bone defect model. 43 Supplemental SAG can also promote neural regeneration after ischaemic brain injury. 44 In our in vivo experiment, we found that SAG injection significantly improves mandibular bone mass reduction caused by Bmal1 knockout at prepuberty and early puberty periods rather than young adulthood, suggesting that SAG may be the potential candidates as novel therapeutic strategies for preventing the facial dysmorphism.
Collectively, our data demonstrate chondrogenesis and endochondral ossification in MCC are under the strict regulation of BMAL1 through a mechanism directly relevant to modulate Hh signalling and provide insights into the mechanism of facial dysmorphism. Moreover, this study presents a ponderable reference for juveniles to correct their undesirable life styles. Anbing Shi for valuable suggestions.

CO N FLI C T O F I NTE R E S T
The authors claim no competing interest.

AUTH O R CO NTR I B UTI O N S
LC conceived the study; SY, QT and LC involved in study design; SY and QT investigated the study; SY collected the data; SY, YL, YX, MX, XZ and FG analysed the data; SY and QT wrote the manuscript; LC supervised the study; and LC acquired funding.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available on request from the corresponding author.