Drivers of phenotypic variation in cartilage: Circadian clock genes

Abstract Endogenous homeostasis and peripheral tissue metabolism are disrupted by irregular fluctuations in activation, movement, feeding and temperature, which can accelerate negative biological processes and lead to immune reactions, such as rheumatoid arthritis (RA) and osteoarthritis (OA). This review summarizes abnormal phenotypes in articular joint components such as cartilage, bone and the synovium, attributed to the deletion or overexpression of clock genes in cartilage or chondrocytes. Understanding the functional mechanisms of different genes, the differentiation of mouse phenotypes and the prevention of joint ageing and disease will facilitate future research.


| INTRODUC TI ON
Osteoarthritis (OA) is a highly prevalent rheumatic musculoskeletal disorder, encompassing progressive, inflammatory and immunological changes affecting joint structures. 1 The clinical features of OA include cartilage loss, increased subchondral bone thickness, tidemark replication, decreased subchondral trabecular bone mass and bone marrow lesions (BML). 2,3 Endogenous biological clocks determine daily, monthly or annual rhythms in biological processes. Cells with endogenous selfexcited oscillations, caused by molecular fluctuations produced by a series of clock genes, form the basis of circadian rhythms. The classical biological clock pathway is comprised of three transcription-translation feedback loops (TTFL) in mammals ( Figure 1). A heterodimer is the core element of TTFLs, which is formed by BMAL1 (brain and muscle ARNTlike 1/Arntl) and NPAS2 (Neuronal PAS domain protein 2) and/or CLOCK (Circadian locomotor output cycles kaput). 4,5 Additional proteins in the loops such as Cryptochrome (CRYs), Period (PERs), REV-ERBs, retinoic acid receptor-related orphan receptors (RORs) and D-box binding protein (DBP) are synthesized at specific times of the day, accumulate and degrade in the cytoplasm.
A vast array of studies has elucidated specific genes or external factors that disrupt the body's natural clock. The systemic or specific deletion of classic clock genes in the biological clock's TTFLs and the artificial disruption of the sleep-wake cycle, causes significant alterations in the human body, including changes in body mass and composition, impaired mental status, growth retardation, premature ageing and cardiovascular disease. [7][8][9][10] In this review, we summarize the integral articular phenotypic variations and the disordered expression of molecular oscillations caused by the disruption of representative clock genes throughout the whole body or within cartilage. These insights provide valuable clues regarding the biological clock's involvement in maintaining cartilage and joint developmental, ageing, and metabolic processes, and healthy matrix homeostasis.

| E VIDEN CE OF CIRC AD IAN CLO CK IN C ARTIL AG E
The circadian rhythm of peripheral tissues can be regulated by specific inputs from external factors, such as eating, temperature, light and physical exercise. These factors provide coordinated input to circadian rhythmicity through crosstalk between metabolic process and neural circuits. [11][12][13] In 1962, Simmons described diurnal variations in cell proliferation and bone growth plate metabolism. 14,15 Soon after, the metabolic rhythms of the epiphysis and condylar cartilage were also studied in detail. [16][17][18][19][20][21][22] Additionally, research reported that, in the absence of alterations in environmental factors, the thickness and volume of articular cartilage also vary from day to night. 23,24 Collectively, these findings suggest the possibility of an autonomously controlled clock in cartilage. A recent series of PER2::Luc clock reporter mouse model studies more directly demonstrated the autonomic functioning of the circadian clock in cartilage and cartilage explants in vivo. [25][26][27] With the emergence of DNA microarrays, mass spectrometry and other advanced molecular biological technologies, more clock genes have been detected in cartilage. More than 600 genes and 145 proteins expressed in cartilage have periodic rhythmicity controlled by clock genes. 28-31

| BMAL1
3.1.1 | BMAL1 may not be necessary for embryonic limb cartilage development and bone growth Many studies have shown that BMAL1 deficiency significantly reduces early embryonic development and the implantation potential of female mice. [32][33][34] However, in normal mouse embryo development, BMAL1 might not be required for cartilage and long bone development. Yu et al using Alcian blue and Alizarin red staining analysis found insufficient cranial cartilage calcification and smaller and shorter mandibular condyle phenotypes in E14.5-E18.5 embryos of global BMAL1 deletion and BMAL1 fl/fl Twist-Cre mice. 35 This finding suggests that BMAL1 occupies a crucial position in the chondrogenesis and entochondrostosis processes of the mandibular condyles.
However, the abnormal progression of endochondral ossification in the limbs was not mentioned. To further understand its role, Ma et al specifically ablated BMAL1 from the cartilage of E13.5-E18 mice. They found no abnormalities in body mass, bone length (femur, tibia and humerus) and growth plates compared to wide-type (WT) mice, 36 suggesting that BMAL1 is not necessary for the elongation of long bones ( Table 1). The sharp contrast between the two studies may be attributed to the different stages of endochondral osteogenesis between the mandibular condyle and limb bones.
The primary synovial joint articular cartilage grows from chondrocytes in the central layer of the epiphyseal plate, gradually differentiated from E11.5. Its shape and biosynthetic properties remain invariable throughout its lifetime. 37,38 Inversely, condyle cartilage, as secondary cartilage, does not develop significantly until E15. Growth starts from mesenchymal tissue covering the prenatal or postnatal condyle, and it easily adapts to alterations in the environment. 39,40 In contrast to the significant effect of BMAL1 on the postnatal endochondral ossification of bone (detailed below), studies have shown that these gene products may not yet form a functional circadian feedback loop during the early stages of development. [41][42][43] However, more research is needed to support this proposal.

| BMAL1 significantly affects adolescent cartilage and bone development
Although the expression of BMAL1 in cartilage and bone gradually decreases with body development, its inverse role is indispensable. 44 Genome-wide RNA sequencing shows the differential expression of chondrogenesis-related genes, and Hedgehog pathway-related proteins were highest during adolescence (4-6 weeks) compared to postpubescence (8-10 weeks). 35 BMAL1 and CLOCK were highly expressed by proliferous zone (PZ) or prehypertrophic to hypertrophic zone (HZ) chondrocytes in the growth plates of WT mice, 35,45 which alluded that BMAL1 was closely related to chondrogenesis.
The study of mice with global BMAL1 deletion was pioneered by Bunger et al. 46 No significant abnormal changes were observed in body mass, body length, degenerated bone and joint pathology before 15 weeks of age in MOP3 −/− mice. Alternatively, Bunger et al suggested that MOP3 was critical for maintaining joint homeostasis in adulthood. In another study, body mass and length, as well as the longitudinal length of the tibia and femur, of BMAL1 −/− mice were significantly reduced from three weeks of age, indicating that BMAL1 influences the elongation of long bones, but it does not indicate obvious cartilage pathology. 45 However, in a study of the mandibular condyle, Yu et al 35 reported that cartilage thickness and matrix were significantly reduced across the whole-body BMAL1-deleted mice, accompanied by reduced chondrocyte proliferation and increased apoptosis. These findings are completely contrary to the previous conclusions, possibly due to the reasons proposed above. Using microcomputed tomography to evaluate intervertebral disc height in mice, the annulus fibrosus (AF) tissue of BMAL1 −/− mice was shown to be hyperplasic, the lumbar vertebrae and intervertebral discs were significantly smaller and thinner, and the vertebral bone parameters were also significantly reduced. 47  to articular cartilage and did not include the synovium or its affiliated ligaments. However, no distinct decreases in body mass and length were observed, which was inconsistent with another study of αⅠ(Ⅱ)-collagen-Cre; Bmal1 fl/fl mice. 45 The conspicuous discrepancy may be explained by differences in the type II collagen a1 (COL2A1) promoter targeting strategies of the two groups of COL2A1-CRE transgenic deletion strains. 48 Ma et al also reported improvements in skeletal lengths 36 45 Early in the mouse lifecycle, the hedgehog pathway activator, SAG, partially saves the mandible shortness phenotype caused by BMAL1 deficiency rather than in adulthood. 35 This is consistent with the rate of cartilage repair during development being much higher in the young compared to adults and the elderly, suggesting the reversibility of cartilage formation. Transforming growth factorβ (TGFβ) and the nuclear factor of activated T cells (NFAT) signalling also play a non-negligible role in maintaining stable growth and ageing of articular cartilage 49,50 by RNA sequencing in cKO mice. Consistently, TGFβ signalling has also been significantly altered by an in vitro RNA-seq of knock-down BMAL1 and NR1D1. 51

| Lack of BMAL1 accelerates ageing and inflammation of limbs and joints in adult mice
Internal body factors, such as ageing, alter the circadian rhythm in cartilage manifested by classical CLOCK, BMAL1, NR1D1 expression, causing changes in amplitude inversely proportional to age (rather than other clock proteins). 25,35,44,51,52 As mentioned above, Bunger et al found that after twenty weeks of age, Mop3 −/− mice showed abnormal phenotypes, including reduced body mass, anorexia, dehydration and abnormal gait. 46 Typical non-inflammatory joint diseases were found through radiological and histological analysis, and there was enhanced heterotopic ossification of hind limb joints, but there were no significant changes in global bone density.
Inversely, the articular cartilage of the tibia and femur, and other soft structures (eg muscles), seemed to remain intact and uneroded.
Similarly, degeneration of the lumbar intervertebral disc (IVD) was observed in Col2a1-BMAL1 cKO mice after six months, and extensive denaturation, ecstatic ossification and increased fibrosis at twelve F I G U R E 1 Circadian clock transcription-translation feedback loops. In primary TTFL, phosphorylated BMAL1:CLOCK/NPAS2 is translocated into the nucleus and initiates the transcription of related genes by binding to the promoter regions E-box (5′-CACGTG-3′) of several genes, including the core clock genes Cryptochrome (CRYs), Period (PERs), REV-ERBs and retinoic acid receptor-related orphan receptors (RORs). The PER and CRY proteins accumulate in the cytoplasm, enter the nucleus and inhibit the transcription and phosphorylation of BMAL1 and CLOCK. Consequently, further production of clock proteins in inhibited. In another feedback pathway, RORs and REV-ERBs proteins competitively bind to ROR elements (ROREs) near the BMAL1 promoter. While REV-ERBs inhibit the transcription of BMAL1, RORs promote the process. As such, REV-ERBs show a strong circadian rhythm that is aligned with the rhythm of BMAL1. Additionally, D-box binding protein (DBP), regulated by E-box and the mammalian transcription factor E4 binding protein 4 (E4BP4; regulated by ROREs), accommodate the expression of PER gene through D-box and, in turn, affect E-box activity. 6 Overall, these three cis-acting elements coordinate a circadian cycle over approximately 24 h TA B L E 1 Summary of phenotypes of genome-editing mouse models in cartilage circadian clock disruption. Description of mouse models in the foetus, adolescence and adult stage in terms of alterations of body height/weight, cartilage/growth plate and bone compared with age-matched wide-type mice  55 Dudek

| Other clock genes
As another pivotal member of the BMAL1:CLOCK heterodimer, CLOCK also plays a noteworthy role in maintaining normal articular cartilage. Large-scale DNA microarray analysis showed that CLOCK might be a mechanically sensitive clock gene, as it was significantly up-regulated following mechanical stress. 30 According to Safranin-O staining, abnormal weight loss and cartilage degeneration phenotypes, similar to BMAL1 deletion, were found in CLOCK Δ19 mice, where the nuclear factor κB (NF-κB) signalling was activated to trigger chronic inflammation. 57,58 This was inconsistent with the vacant study of Kc et al, 11 who found no significant pathological differences in Safranin-O fast green staining in the cartilage of CLOCK and Taumutant mice. This might be due to age differences, as the genetic background was the same between the studies. The loss of CLOCK gene fragments may alter cartilage phenotypes in adult mice. In addition to cartilage degradation, shorter tibiae and lower bone mass could also be induced by CLOCK mutation. 59 The latest research demonstrated that the motor ability of 18-month-old mice, promoted by injecting lentiviral vectors encoding CLOCK, to enhance cartilage regeneration, decreased ageing markers (Cdkn1a and Cdkn2a), as well as activating genes involved in cartilage development. 60 These results imply that the CLOCK gene supports cartilage rejuvenation and alleviates ageing. However, the importance of other clock genes, such as CRYs and PERs, in maintaining joint phenotypes has not been studied.

| BMAL1
As a core clock factor in the circadian rhythm feedback loop, BMAL1 in antiphase to PER2::Luc reporter. 25,29 Remarkably, BMAL1 excision deleted both BMAL1 and BMAL2, which could not substitute BMAL1 to maintain the expression oscillation of most genes in the tissues. 63 These results allude to the consistent regulatory role of BMAL1 on other clock genes. In contrast, other studies have shown that BMAL1 was incapable of mediating RORA, RORC and RBX1. 51

| CLOCK
There have been no obvious rhythmic fluctuations in the CLOCK gene detected in mouse cartilage tissues, such as rib growth plates 45 and intervertebral discs. 64 It is possible that NPAS2 is combining with BMAL1 instead of the non-oscillating CLOCK.

| REV-ERBs
Although they are members of the secondary circadian clock path- reduce abnormal collagen fibre synthesis and collagen accumulation. 69 These findings have proved the importance of the circadian rhythm in balancing collagen fibre synthesis and degradation.

| FUTURE RE S E ARCH D IREC TI ON S
Currently

CO N FLI C T O F I NTE R E S T
The authors confirm that there are no conflicts of interest. Funding acquisition (lead); Supervision (lead); Validation (lead).

DATA AVA I L A B I L I T Y S TAT E M E N T
Data sharing is not applicable to this article.