Enhanced Circadian Clock in MSCs-Based Cytotherapy Ameliorates Age-Related Temporomandibular Joint Condyle Degeneration

To identify the age-related morphological change in TMJ condyles, we focused on three specific life phases: mature young adult, middle-aged adult, and old-age adult. According to previous studies that determined the approximately comparable life phases between human beings and mice [19,20], TMJ condyles were isolated from 4-month-old (young adult), 10-month-old (middle-aged adult), and 20-month-old (old-age adult) mice. The microscope photographs of TMJ condyles (Figure 1A,B) showed significant morphology alterations in the geometric shape and condition of the osteochondral surface. The shape of TMJ condyles became longer in the sagittal direction and flatter on the condylar surface with aging, which might result from adaptation remodeling to continuous masticatory force. In the aged mice, the TMJ condylar surface became uneven with irregular erosion, which coincides with the age-related degeneration in human being joints [21]. To further understand the age-related change in the subchondral bone, micro-CTs were performed on each age group. The coronal-section images of the TMJ condyles showed a progressive increase in bone volume and enlarged subchondral cysts in TMJ condyles with aging (Figure 1C). The quantitative analysis of subchondral bone supported the previous finding (Figure 1D). The bone volume/tissue volume ratio (BV/TV) and bone mineral density (BMD) were significantly increased in 20-month-old TMJ condyles, while the trabecular number (Tb.N) and trabecular separation (Tb.Sp) were significantly diminished in aged TMJ condyles and the trabecular thickness (Tb.Th) was increased with aging. Overall, the radiographic manifestations of bone degeneration in aged TMJ condyles were consistent with signs detected in human aged joints.

Next, we sought to examine the histological changes in TMJ condyles with aging. H&E staining showed that the thickness of the proliferative layer zone and hypertrophic layer zone on TMJ condylar cartilage was significantly decreased in middle-aged and old-age mice (Figure 2A,B). Furthermore, the integrity and clear layer structure of cartilage were progressively damaged with aging. Safranin O staining and quantitative analysis revealed an obvious decline in safranin O positive area with the aging, which indicated a significant loss of glycosaminoglycan in the cartilage and subchondral bone in age-related TMJ condyle degeneration. The Osteoarthritis Research Society International (OARSI) histomorphology score of TMJ condyles was higher as the mice grew old. The osteoclasts in the subchondral bone were detected by tartrate-resistant acid phosphatase (TRAP) staining. Quantitative analysis of TRAP staining revealed a significant decline in osteoclasts activities in subchondral bone with aging.

We further examined the age-related molecular changes in the TMJ condyle. The qRT-PCR analysis was performed on the TMJ condyles in different age groups, to determine the expression of chondrogenic, osteogenic and inflammatory markers including Acan, MMP13, Col I, Sp7, TNF-α, and IL-6. The expression of Acan, the proteoglycan found in articular cartilage, increased from the young adult to the middle-aged periods and diminished in old age. On the contrary, the expression of matrix metallopeptidase 13 (MMP13), the collagenase related to extracellular matrix degeneration, decreased during middle age and was elevated in old age (Figure 1C). The osteogenic markers Col I and Sp7 were significantly declined with aging, while the inflammatory markers IL-6 and TNF-α became prominent in older age. Together with our previous histological finding, a significant decrease in osteogenesis and osteoclastic activities was observed in the subchondral bone of aged TMJ condyles, which indicated that the bone turnover activities in the subchondral bone were flattened with aging.

To investigate the role of circadian genes in age-related TMJ condylar degeneration, immunostaining was performed on condylar sections in different age groups (Figure 3A–D) to detect the expression alterations of core circadian genes. As shown in Figure 3A,B, core clock-gene BMAL1 and CLOCK were widely expressed and gradually decreased in TMJ cartilage with aging, especially on the hypertrophic layer zone and proliferative layer zone. This change was in parallel with the morphologic development of TMJ condyle cartilage with aging, which highlighted the potential crucial role of the circadian clock in age-related TMJ degeneration. PER2 and CRY1 were the typical circadian-controlled genes (CCGs), which were regulated by BMAL1/CLOCK heterodimers and mediated the transcriptional feedback loops system. Immunofluorescent staining showed a significant decrease in PER2 expression in the TMJ condyles of 20-month-old mice with a younger age, while the CRY1 expression substantially declined in the 4-, 10- and 20-month-old mice with aging (Figure 3C,D).

Next, we further examined the circadian rhythmic expression in TMJ condylar cartilage in young and old-age adult mice. TMJ condylar cartilage at 4 months old and 20 months old was harvested at 6 h intervals according to zeitgeber time (ZT0, ZT6, ZT12, ZT18, ZT24), and qRT-PCR was performed to detect the rhythmic expression of core clock genes. The significantly blunted circadian rhythms of Bmal1, Clock, and Cry1 expression were observed in the old group compared with the mature adult group (Figure 3E). The molecular rhythmic expression of Per2 presented obvious postponement.

Bmal1, as the core circadian gene, has been reported to orchestrate the multiple differentiation of mesenchymal stem cells. In another work, we generated Bmal1 conditional knockout mice (Prx1-cre; Bmal1^fl/fl) from early limb bud mesenchyme and carried out RNA sequencing (RNA-seq) to examine the expression profile of BMSCs isolated from 4-month-old Prx1-cre, Bmal1^fl/fl and Bmal1^fl/fl mice. The volcano map showed the most obvious differentially expressed genes from Prx1-cre, Bmal1^fl/fl BMSCs, and Bmal1^fl/fl controls, from which the Prg4 was the most significant down-regulated one by Bmal1 deletion (Figure 4A). PRG4 is a large proteoglycan which synthesized by cells located at the surface of articular cartilage and is involved in elastic absorption and energy exchange of synovial fluid. Moreover, the expression heatmap further confirmed the down-regulation of chondrogenic marker genes, such as Col7a1, Col14a1, Matns2 and Matns3 (Figure 4B). We also performed a gene set enrichment analysis (GSEA) on our sequence data, which showed a global down-regulation of chondrogenic differentiation genes (Figure 4C). The protein–protein interaction network visualization using STRING revealed that the PRG4-related function might have a strong bearing on chondrogenesis, and the local network cluster and associated biological process analysis reported close links with extracellular matrix organization and cartilage development (Figure 4D,E). This result suggested that Prg4 and chondrogenesis-related genes were the potential downstream targets of Bmal1-mediated osteochondral development and chondrogenic differentiation.

Since age-related TMJ condyle degeneration was strongly linked to Bmal1 depletion on articular cartilage, we next examined the reparative effects of circadian-enhanced MSCs transplantation on the aged TMJ region. Primary BMSCs were flushed out from the femur and tibia of 2-month-old mice and cultured as previously described [22]. Flow cytometry analysis showed that the isolated cells were positive for BMSCs surface markers CD29 and Sca-1, and a lack of hematopoietic stem cells surface markers CD45 and CD34. The spindle- and triangular-shaped BMSCs at passage 2 were observed under phase contrast microscopy and multiple differentiation capability was proved by Alizarin Red staining, Oil Red staining and Alcain Blue staining after conditional medium induction (Figure 5A,B). BMSCs overexpressing Bmal1 (Bmal1-OE) and GFP control (GFP-NC) were constructed by lentivirus transfection. The transfection efficiencies were visualized under fluorescence microscopy and determined by qRT-PCR and Western blot, as shown in Figure 5C,D. The original Western blot acquisitions for cropped images were shown in Figure S2. The migration capability of donor cells into the target tissue is an essential condition in MSCs-based therapies. Therefore, transwell assays were performed to investigate the effects of Bmal1 overexpression on MSCs migration. As shown in Figure 5E, the number of migrated BMSCs was significantly increased in Bmal1-OE in the control group. This result revealed that BMSCs overexpressing Bmal1 presented greater migration capability than GFP-NC.

The workflow diagram of the MSCs therapy is shown in Figure 6A. To verify the successful transplantation of donor cells, the immunostaining analysis revealed that the implanted GFP-positive BMSCs in aged TMJ cartilage were differentiated into chondrocyte-like cells in vivo (Figure 6B). The microscope photographs of the treated TMJ condyles showed that both Bmal1-OE and GFP-NC MSCs-treated TMJ condyles presented a smoother osteochondral surface than the sham group, indicating the restoration of degenerated cartilages after cytotherapy (Figure 7C). The representative micro-CT images further showed an obvious decline in osteophytes around the margin of the osteochondral surface and lessening of subchondral cysts below the articular surface. The micro-CT analysis of the MSCs-treated subchondral bone showed slight but not significant increase in BV/TV and BMD with the sham group, accompanied with an obvious rise in Tb. N and decrease in Tb. Sp and Tb. Th (Figure 6D,E).

Histological staining and quantitative analysis of the condylar sections further revealed that Bmal1-enhanced MSCs-treated TMJ condyles were increased in cartilage thickness and decreased in numbers of excessive subchondral cysts. Safranin O-positive areas and OARSI scores revealed a significant recovery of proteoglycan loss in the condylar surface after Bmal1-enhanced MSCs treatment. The TRAP staining and quantitative analysis of osteoclast numbers showed a decrease in osteoclast activities in Bmal1-overexpressed MSCs-treated TMJ condyle, indicating the regenerative effects of circadian-enhanced therapy in the subchondral bone. The immunostaining and quantitative analysis of the expression levels of chondrogenesis-related markers ACAN and COL II further revealed that more active chondrogenesis was shown in the cartilage layer in the Bmal1-OE group than in the control groups. These results imply that transplantation of Bmal1-overexpressed BMSCs in aged TMJ cavity could effectively ameliorate age-related cartilage degeneration.

Male C57BL/6 mice (4-month-old, 10-month-old and 18~20-month-old) were purchased from Chengdu Dossy Biological Technology Co. Ltd. and housed in West China Hospital Animal Center under standard animal housing conditions with a 12 h light and 12 h night cycle. For the analysis of age-related changes in temporomandibular joints, animals were randomly selected and sacrificed in each age group (n = 8–14 per group). All procedures and management of the animals were approved by the animal ethics committee of West China School of Stomatology, Sichuan University and the State Key Laboratory of Oral Disease (WCHSIRB-D-2017-046).

The temporomandibular condyles were harvested and immediately fixed in 4% paraformaldehyde at 4 °C for 72 h. Collected samples were scanned using a micro-CT scanner (μCT50;SCANO, Switzerland) at a voltage of 70 kVp, a current of 200 μA, and resolution of 4.0 μm/pixel. For the analysis of subchondral bone, a semi-arched region of interest (ROI) was defined at 100μm below the osteochondral interface of the temporomandibular condyle head. The diagram of the selected ROI is shown in. The bone volume fraction (BV/TV), trabecular number (Tb.N), trabecular thickness (Tb. Th) and trabecular separation (Tb.Sp) were calculated. Supplementary Figure S1

Following micro-CT scanning, the samples were decalcified in PBS-buffered 10% EDTA (pH 7.4) for 14 days. The decalcified samples were processed, embedded in paraffin, and dissected into 5 μm-thick sections at the sagittal plane. Hematoxylin and eosin (H&E), safranin O staining, and tartrate-resistant acid phosphatase (TRAP) staining were performed according to the manufacturer’s instructions as in the previous study. The Osteoarthritis Research Society International (OARSI) scores were used to grade the cartilage degeneration of temporomandibular joint condyle [49].

Immunostaining was performed as previously described. For IHC staining, the anti-rabbit/mouse immunohistochemical secondary antibody kit (Absin Bioscience) was used following the manufacturer’s instruction. The following antibodies were used: Bmal1 (1:250, #NB100-2288, Novusbio), Clock (1:100, #ab3517, Abcam), GFP (1:100, #NB600-597, Novusbio), Acan (1:200, #13880-1-AP, Proteintech), Collagen II (1:100, #ab34712, Abcam). For IF staining, the following primary antibodies were used: Per2 (1:200, #180655, Abcam), Cry1 (1:200, #ab104736, Abcam). The sections were stained with DAPI to detect the nuclei. After mounting, the sections were photographed with a Leica DM2500 microscope (Leica, Germany). The immunoreactivity of each group was quantified in five different samples and three random areas using ImageJ software.

To determinate the rhythmic expression of circadian clock genes in 4-month-old and 20-month-old temporomandibular condylar, the total RNA of the temporomandibular condylar cartilage was extracted using TRIzol (Invitrogen) at Zeitgeber Time in 6 h intervals (ZT0, ZT6, ZT12, ZT18, ZT24) according to the manufacturer’s instruction. The RNA concentration was measured with the NanoDrop 2000 (Thermo Fisher Scientific) and reverse transcribed into cDNA using the PrimeScript RT Reagent Kit with gDNA Eraser (Takara). RT-qPCR was performed using SYBR Premix Ex Taq II (Takara) on an Applied Biosystems Quant Studio 7 (Thermo). Relative genes expression was performed by Gadph for mRNA expression using the 2^−ΔΔCt method. The primers of target genes are listed in Supplementary Table S1.

The total RNA samples were extracted using Trizol reagent. A sequencing library was constructed using NEBNext Ultra RNA Library Prep Kit for Illumina (NEB, USA) following the manufacturer’s recommendations, and was then performed on the Illumina HiSeq 2500. The FastQC and FASTX toolkits were used to control the sequencing quality. Sequencing data were mapped to Mus musculus reference genomes by HISATS, and differential expression analysis was performed by Ballgown software. Genes were considered significantly differentially expressed if fold change ≥2.0 and p value < 0.05.

The GSEA was performed with GSEA software (http://www.broad.mit.edu/GSEA↗ accessed on 23April 2021) using a sequencing-derived gene list. The protein–protein interaction (ppi) network analysis was performed with STRING (https://www.string-db.org↗ accessed on 16 July 2021).

Primary BMSCs were isolated by flushing the bone marrow from the femurs and tibiae of 2-month-old mice as previously described [22,50]. Cells were cultured in α-MEM medium (Hyclone) supplemented with 10% fetal bovine serum (Gibco) and 1% penicillin and streptomycin (HyClone) at 37 °C and in a 5% CO₂ atmosphere. Cells at passage 2 were used to characterize BMSCs surface biomarkers. The flow cytometric analysis was performed as previously described. The following antibodies were used for FACS analysis: APC-Cyanine7 anti-mouse CD29 (#102225, Biolegend), PE anti-mouse CD34 (#119307, Biolegend), FITC anti-mouse CD45 (#10307, Biolegend), and FITC anti-mouse Sca-1 (#108105, Biolegend).

The lentivirus vectors Lv-GFP-Bmal1 and Lv-GFP-mock were purchased from Hanbio Co. Ltd., Shanghai, China. Prior to lentivirus transfection, primary BMSCs at passage 2 were seeded in 6-well plates. Cells were transfected with lentivirus vectors when cells confluence reached 40–50%, and experiments were performed 48 h after infection. The transfected cells were named Bmal1-OE and GFP-NC.

Cells were lysed in RIPA buffer (Sabbiotech). The protein concentrations were evaluated using a BCA protein assay kit (Beyotime). Equal amounts of proteins in each sample were separated by SDS-PAGE (Bio-Rad Laboratories), transferred to PVDF membranes (Millipore), and blocked in 5% milk. Following incubation with rabbit anti-BMAL1 (1:1000, #14020, Abcam), anti-beta-actin (1:1000, #8457S, CST) at 4 °C overnight, and horseradish peroxidase-conjugated secondary antibody (1:5000, #L3012-2, Sabbiotech), the target protein was exposed using a chemiluminescence kit (Bio-Rad Laboratories).

Cell migration assays using a transwell system (Corning, 3422) were performed as previously described [51]. Briefly, the lentivirus transfected BMSCs were seeded in the upper chamber at a density of 2 × 10⁵ cells/100 uL/well in a serum-free medium. The lower chamber was filled with a complete medium containing 10% FBS. Following the 36 h of incubation, the migrated cells on the lower chamber were fixed in 4% paraformaldehyde and stained with crystal violet. The number of migrated cells was counted under camera.

We investigated the treatment effect of circadian enhanced BMSCs on age-related TMJ degeneration. The 5000 Bmal1-OE and GFP-NC BMSCs were suspended in 20 μL phosphate-buffered saline (PBS) 48 h after lentivirus transfection and injected into the TMJ region of 18-month-old mice (n = 5–6) [18]. The aged mice were anesthetized by intraperitoneal injection of Ketamine and Xylazine (100 mg/mL, 2:1, 1 mL/kg body weight). The delivery of cells into the bilateral TMJ regions were performed as previously described [18,39]. Briefly, the mice were laid sidelong after deep anesthesia. We used 30-gauge needles in the most prominent bulge of the zygomatic arch, and along the superomedial of the zygomatic arch toward the TMJ region. Transfected BMSCs were injected into the bilateral TMJ region, respectively, at 0, 2, and 4 weeks successively. In the sham group, 20 μL PBS was injected into the TMJ region.

All data were presented as mean ± SD. Statistical differences of multiple comparisons were determined by one-way ANOVA followed by Tukey’s post hoc test with GraphPad Prism 7 software. The p value < 0.05 was considered to be statistically significant.