The circadian regulator Bmal1 in joint mesenchymal cells regulates both joint development and inflammatory arthritis
The circadian clock plays a crucial role in regulating physiology and is important for maintaining immune homeostasis and responses to inflammatory stimuli. Inflammatory arthritis often shows diurnal variation in disease symptoms and disease markers, and it is now established that cellular clocks regulate joint inflammation. The clock gene Bmal1 is critical for maintenance of 24-h rhythms and plays a key role in regulating immune responses, as well as in aging-related processes. Fibroblast-like synoviocytes (FLS) are circadian rhythmic joint mesenchymal cells which are important for maintenance of joint health and play a crucial role in the development of inflammatory arthritis. The aim of this study was to investigate the importance of the joint mesenchymal cell circadian clock in health and disease.
Mice were generated which lack Bmal1 in Col6a1-expressing cells, targeting mesenchymal cells in the ankle joints. Joints of these animals were assessed by X-ray imaging, whole-mount staining and histology, and the composition of the synovium was assessed by flow cytometry. Arthritis was induced using collagen antibodies.
Bmal1 deletion in joint mesenchymal cells rendered the FLS and articular cartilage cells arrhythmic. Targeted mice exhibited significant changes in the architecture of the joints, including chondroid metaplasia (suggesting a switch of connective tissue stem cells towards a chondroid phenotype), reductions in resident synovial macrophages and changes in the basal pro-inflammatory activity of FLS. Loss of Bmal1 in FLS rendered these resident immune cells more pro-inflammatory in response to challenge, leading to increased paw swelling, localised infiltration of mononuclear cells and enhanced cytokine production in a model of arthritis.
This study demonstrates the importance of Bmal1 in joint mesenchymal cells in regulating FLS and chondrocyte development. Additionally, we have identified a role for this core clock component for restraining local responses to inflammation and highlight a role for the circadian clock in regulating inflammatory arthritis.
KeywordsBMAL1 Circadian Fibroblast-like synoviocytes Chondrocyte Arthritis Macrophage Synovium
Enzyme-linked immunosorbent assay
Macrophage colony-stimulating factor
Major histocompatibility complex
Receptor activator of nuclear factor κ ligand
Tumour necrosis factor-α
The circadian clock enables organisms to regulate their physiology in synchrony with the changing 24-h environment. This biological timer regulates a wide range of physiological processes, including metabolism and immunity [1, 2]. Multiple different cell types possess the clockwork machinery, a network of genes forming a transcriptional-translational feedback loop. In mammals the core clock genes include Bmal1, Clock, Cryptochrome (Cry) and Period (Per), where Bmal1 is the only non-redundant gene. Mice lacking Bmal1 from pre-natal development are behaviourally arrhythmic in the absence of an entraining light/dark cycle, and show loss of rhythmic physiology . In addition to co-ordinating circadian rhythms, Bmal1 regulates other physiological functions, and these global knockout mice have reduced lifespan and fertility and exhibit pathologies affecting the eyes, brain and bone [4, 5]. This includes a progressive non-inflammatory arthropathy resulting in joint ankylosis . Interestingly, this phenotype persists if Bmal1 is rescued in brain or muscle , but is absent if Bmal1 is only deleted after birth .
In the present study, we explored the contribution of Bmal1 in fibroblast-like synoviocytes (FLS) to joint architecture. FLS are found within the lining of the synovium, the thin organised membrane located between the joint cavity and joint capsule. They are stromal cells of mesenchymal origin, producing a range of extracellular matrix components and secreted factors essential to maintaining the normal environment of the synovial fluid and articular surface . FLS play a critical role in the pathogenesis of inflammatory arthritis, producing inflammatory mediators which contribute to the recruitment and activation of leucocytes, cartilage breakdown and joint remodelling . It is well established that the core clock proteins (PERIOD1/2, BMAL1 and CLOCK) are expressed by FLS [10, 11], and we and others have shown that these immunoregulatory cells are circadian rhythmic [11, 12, 13]. Intriguingly, there is mounting evidence that under chronic inflammatory conditions, such as rheumatoid arthritis, these intrinsic timers are disrupted [10, 11, 13, 14, 15].
By deleting Bmal1 in Col6a1-expressing cells we rendered joint mesenchymal cells (FLS and articular chondrocytes) arrhythmic. This targeted deletion had profound effects on joint architecture, homeostasis and inflammatory joint disease, highlighting the critical importance of the joint mesenchymal cell clock in health and disease.
B6.Cg-Tg(Col6a1-cre)1Gkl/Flmg mice, referred to hereinafter as Col6a1cre/+ mice, were purchased from the European Mutant Mouse Archive repository as frozen embryos and re-derived in-house. These mice, generated by Prof. G. Kollias , express Cre recombinase under the control of a collagen VI promoter cassette known to drive gene expression in mesenchymal cells in the ankle joints, mainly fibroblast-like cells but also articular chondrocytes [16, 17]. Bmal1flox/flox PER2::luc mice (as described previously ) were bred with Col6a1Cre/+ mice to produce Bmal1flox/flox PER2::luc Col6a1 Cre+/− mice (Col6a1-Bmal1−/−) and Bmal1flox/flox PER2::luc Col6a1 Cre−/− mice (wild-type counterparts). Global Bmal1−/− mice  were bred as heterozygous pairs to produce Bmal1−/− and wild-type littermates. All animal procedures were carried out in accordance with the United Kingdom Animals (Scientific Procedures) Act 1986 and were subject to local ethical review by the University of Manchester Animals Welfare and Ethical Review Board. Mice were maintained under standard 12-h/12-h light/dark lighting with ad libitum access to standard chow. Unless stated otherwise, male mice aged 8–20 weeks were used.
Murine FLS cultures
Mouse FLS were cultured as described previously . In brief, mice were killed, and their skin was removed from the hind paws, which were then cut 1 mm above the ankle joint and placed into Hanks’ balanced salt solution. After a wash, the paws were dissected between the toe and ankle joints using a scalpel blade, and the dissected tissue was placed into DMEM (containing 1% penicillin-streptomycin and 10% FBS) with 10 mg/ml collagenase (from Clostridium histolyticum type IV; Sigma-Aldrich, Gillingham, UK) and placed into a shaking incubator for 1.5 h at 37 °C. Cells (and some tissue debris) were pelleted by centrifugation, re-suspended in DMEM and plated out. Cells were cultured in DMEM to passage 3 before further experimental use, in order to obtain a purified population of FLS.
Bead sorting of FLS
Cultured mouse FLS (passage 3) were purified further by labelling them with a phycoerythrin (PE)-CD90.2 antibody (clone 30-H12) and positively selecting PE-labelled cells (EasySep PE Positive Selection Kit, as per kit instructions; STEMCELL Technologies, Vancouver, BC, Canada) for further culture. Bead-sorted FLS were used for Western blotting of the BMAL1 protein, as well as for qPCR analysis and in tumour necrosis factor (TNF)-α stimulation experiments (see below).
Mice were killed, and the peritoneal cavity was lavaged with two washes of 5 ml of RPMI (containing 1% penicillin-streptomycin and 10% FBS). The recovered media was centrifuged (432 × g for 10 min). For RNA analysis, the resultant pellet of peritoneal exudate cells was lysed immediately, and RNA extraction was carried out (as described below). For photomultiplier tube recordings, the pellet was re-suspended in RPMI, and cells were plated out into a 35-mm dish. After 2-h incubation at 37 °C, non-adherent cells were removed via three washes in warmed RPMI. For bone marrow-derived macrophages, mice were killed and the femur and tibia were removed. The bone marrow was flushed out with DMEM (containing 1% penicillin-streptomycin and 10% FBS), and the effluent was centrifuged (432 × g for 10 min). The pelleted cells were re-suspended in media containing 50 ng/ml macrophage colony-stimulating factor (M-CSF) (eBioscience, San Diego, CA, USA). The media were replaced after 3 days, and on day 6, the bone marrow-derived macrophages were scraped from the flask and lysed to extract protein for Western blot analysis (described below).
FLS (passage 3) and peritoneal macrophages were plated out on 35-mm dishes. Primary murine lung fibroblasts were cultured as described elsewhere  and plated out on 35-mm dishes. Femoral head tissue was placed directly onto tissue culture inserts within the dishes. Cells and tissues were synchronised (200 nM dexamethasone, 1 h), and the media were replaced with recording media containing luciferin . Dishes were sealed over with a glass coverslip using vacuum grease . Bioluminescence from FLS, femoral head tissue and lung fibroblasts was recorded every minute using photomultiplier tubes. Bioluminescence data from peritoneal macrophages were collected and analysed using a LumiCycle (ActiMetrics, Wilmette, IL, USA). Data were plotted using Prism software (GraphPad Software, La Jolla, CA, USA).
Western blot analysis
Cultured FLS and bone marrow-derived macrophages were lysed using a radioimmunoprecipitation assay buffer (Sigma-Aldrich) containing cOmplete Mini, EDTA-free Protease Inhibitor Cocktail Tablets (Roche Diagnostics, Mannheim, Germany). Protein concentrations of cell lysates were measured with a bicinchoninic acid protein quantification kit (Merck, Kenilworth, NJ, USA), and an equal amount of protein was separated on SDS-polyacrylamide gels and transferred to nitrocellulose membranes (Thermo Fisher Scientific, Waltham, MA, USA). Membranes were incubated with primary antibody against BMAL1 (D2L7G; Cell Signaling Technology, Danvers, MA, USA) and β-ACTIN (ab8226; Abcam, Cambridge, UK), followed by horseradish peroxidase-conjugated secondary antibody, and imaged using enhanced chemiluminescence (Clarity; Bio-Rad Laboratories, Hercules, CA, USA).
TNF-α stimulation of FLS
Bead-sorted FLS were stimulated with 10 ng/ml murine TNF-α (Sigma-Aldrich). Eight hours later the supernatant was removed. Cytokine levels in supernatants were quantified using the Bio-Plex Pro Mouse Cytokine 23-plex assay (Bio-Rad Laboratories) as per kit instructions, and run on a Bio-Plex 200 System. CXCL5 and M-CSF levels in supernatant were quantified separately using DuoSet enzyme-linked immunosorbent assay (ELISA) kits as per kit instructions (R&D Systems, Minneapolis, MN, USA).
Collagen antibody-induced arthritis
Male mice (aged 10 weeks) were administered a cocktail of collagen antibodies (ArthritoMab; MD Biosciences, Egg, Switzerland) intravenously (4 mg/mouse) on day 0, and lipopolysaccharide (100 μg/mouse) intraperitoneally on day 3. Both treatments were administered during the mid-light phase (Zeitgeber time [ZT] 6). Each paw was scored for disease severity using a 4-point scale (1 = inflammation of one digit; 2 = inflammation of more than one digit; 3 = inflammation and swelling of the footpad with or without involvement of one or more digits; 4 = severe inflammation and swelling of the footpad and ankle with or without involvement of one or more digits). Each paw score was totalled to produce an animal score. Paw thickness was measured in the hind limbs using a thickness gauge. Mice were killed on day 7, when serum was harvested along with the left hind paw for RNA analysis and the right hind paw was harvested for flow cytometry. ELISAs for interleukin (IL)-6 in mouse serum were performed using DuoSet ELISAs (R&D Systems) as per kit instructions.
RNA was extracted from snap-frozen limbs and Peyer’s patches using TRIzol reagent (Life Technologies, Carlsbad, CA, USA). Frozen limbs were first pulverised using a pestle and mortar containing liquid nitrogen, and the resulting material was transferred to a Lysing Matrix D tube (MP Biomedicals, Santa Ana, CA, USA) and homogenised in TRIzol reagent using a FastPrep-24 machine (MP Biomedicals). Peyer’s patches were added directly to the Lysing Matrix D tubes containing TRIzol reagent for homogenisation. RNA was extracted using the standard TRIzol procedure. RNA was extracted from cultured primary cells using RNeasy kits (Qiagen, Hilden, Germany). After DNase treatment (DNase I, Thermo Fisher Scientific), RNA was converted to complementary DNA (cDNA) using the High-Capacity RNA-to-cDNA Kit (Thermo Fisher Scientific). qPCR was performed using a StepOne Real Time PCR system (Thermo Fisher Scientific) with TaqMan primers (see Additional file 1: Table S1). The housekeeping gene Gapdh was used to normalise data.
Whole limbs were dissected at the joints, and the tissue was digested in collagenase (as described earlier). Isolated cells were counted using a nucleocounter NC-250 (ChemoMetec, Allerod, Denmark). A sample (5 μl) of the cell suspension was diluted in PBS (1:4), and a solution of acridine orange and 4′,6-diamidino-2-phenylindole (Solution 18; ChemoMetec) was added to provide counts of total and dead cells, respectively. For flow cytometric analysis, the remaining cells were subjected to live/dead staining (LIVE/DEAD Fixable Blue Dead Cell Stain Kit; Thermo Fisher Scientific) for 20 min. ArC Amine Reactive Compensation Beads (Thermo Fisher Scientific) were used to obtain compensation controls for live/dead staining. After washing off the live/dead stain, an Fc block (CD16/CD32) was applied for 15 min, and then cells were stained with a cocktail of antibodies (Additional file 1: Tables S2 and S3). Cells were analysed on a BD LSR II flow cytometer (BD Biosciences, San Jose, CA, USA) after compensation set-up with OneComp eBeads (Thermo Fisher Scientific). Gating strategies are outlined in Additional file 1: Figures S4 and S5. Data were analysed using FlowJo software (FlowJo, Ashland, OR, USA). In some experiments FLS (CD45−CD90.2+) were sorted from these samples (BD Influx cell sorter; BD Biosciences) into lysis buffer, and RNA was prepared using RNeasy Plus Micro kits (Qiagen).
The skin was removed from hind limbs, which were then placed into formalin overnight. The next day, tissue was transferred to OSTEOSOFT (VWR, Radnor, PA, USA) for 2 weeks to allow de-calcification. De-calcified tissue was processed and embedded into paraffin wax. Sections (5 μm) were cut and mounted onto glass slides. After rehydration, slides were subjected to H&E staining or Safranin O staining using standard protocols. Slides were imaged on a Leica DM2000 microscope, and images were captured with a Leica DFC295 camera (Leica Microsystems, Wetzlar, Germany).
Mice were killed and X-rayed on a Faxitron MX-20 (32 kV for 55 s) (Faxitron Bioptics, Tucson, AZ, USA). Scanned X-ray films were analysed using ImageJ software. Using these images, pixel intensity was quantified in a standardised size ROI superior to the calcaneus or between vertebral disks. This value was adjusted for background intensity.
Whole-mount staining was performed as described elsewhere . In brief, skin and adipose tissue were removed from the hind limbs, and limbs were detached at the pelvis. Limbs were fixed in 95% ethanol for 48 h, then placed into 100% acetone (2 days). They were stained with Alcian blue solution for 3 days (0.03% wt/vol Alcian blue, 80% ethanol, 20% glacial acetic acid). After a brief wash (70% ethanol) they were de-stained overnight in 95% ethanol. Limbs were then placed in potassium hydroxide (1% wt/vol) for 4 h before staining with Alizarin red solution (0.005% wt/vol Alizarin red in 1% wt/vol KOH) for 4–5 days. After a wash in distilled water they were placed in 1% KOH for 2 days to begin clearing. Clearing solution was changed weekly, with decreasing amounts of KOH and increasing amounts of glycerol, until after 3 weeks limbs were placed in 100% glycerol for imaging and storage.
Flow cytometry data were analysed using FlowJo version 10 software. All other data were analysed using Prism 7 software. Data are presented as mean ± SEM. Where statistical tests were used, statistical significance is annotated as follows: *P < 0.05, **P < 0.01, ***P < 0.005.
Bmal1 deletion renders FLS arrhythmic
Loss of Bmal1 in FLS significantly affects joint structure
Bmal1 in FLS represses inflammatory arthritis
This study addresses the role of the core clock component Bmal1 in defining musculoskeletal structure. By deleting Bmal1 in a targeted cell population, we were able to identify the cell lineage responsible for a previously recognised joint arthropathy. Bmal1 deletion in joint mesenchymal cells rendered FLS and articular cartilage arrhythmic and led to disruption of ankle joint architecture. This was evidenced by focal chondroid metaplasia, decreased resident synovial macrophage numbers, and altered basal secretion of a subset of cytokines by FLS in the native state.
Global Bmal1-knockout mice show a progressive ectopic mineralisation with new bone formation at tendon and ligament insertion sites . Our findings suggest this ossification may be a consequence of a change in mesenchymal stem cell differentiation towards a bone-depositing chondroid phenotype. These pathological chondrocytes drive calcification of the tendon sheath to form a calcaneal spur. Global deletion of Bmal1 in adult mice does not result in this abnormal calcification ; therefore, it is likely a consequence of the absence of Bmal1 in mesenchymal cells during embryogenesis/early post-natal development. Articular chondrocytes and synovial fibroblasts are derived from common joint progenitor cells, with the action of different factors driving them towards a chondrogenic or anti-chondrogenic lineage . Furthermore, Col6a1 is present in synovial membrane and articular cartilage in 16.5-day murine embryos . We suggest that BMAL1 expression in joint mesenchymal cells may be essential for regulating the balance between FLS and chondrocytes of the developing embryo. In support of this, higher expression levels of chondrocyte-associated genes were detected in the hind limbs of Col6a1-Bmal1−/− mice. This includes genes encoding for type II collagen and aggrecan (major structural components of cartilage), proteoglycan 4 (synthesised by chondrocytes) and Indian hedgehog (associated with chondrocyte differentiation, proliferation and maturation). Intriguingly, others have shown that targeted ablation of Bmal1 in mouse chondrocytes causes progressive degeneration of articular cartilage within the knee, with loss of chondrocytes and extracellular matrix evident from 3 months of age . Taken together, it is clear that Bmal1 within mesenchymal cells is critical for joint health throughout development and later life.
Col6a1-Bmal1−/− animals demonstrated further alterations in the cellular composition of the synovial tissue. Tissue-resident MHC II− macrophages were significantly reduced within the joints. Resident synovial macrophages play critical roles in immune surveillance, maintenance of tissue integrity and limiting inflammation. It should be noted that human and murine macrophages express Col6a1 and secrete collagen VI protein [33, 34]. Macrophages derived from Col6a1-Bmal1−/− mice express low levels of Col6a1 and Cre (data not shown). However, Cre expression was insufficient to drive recombination and did not affect expression of the targeted exon of Bmal1 (exon 8) within the macrophage (Additional file 1: Figure S6a). In keeping with this, expression of BMAL1 protein was unchanged (Additional file 1: Figure S6b). Furthermore, Col6a1-Bmal1−/− macrophages showed robust circadian rhythms in PER2::luc bioluminescence (Additional file 1: Figure S6c). (Previous work demonstrated that efficient deletion of Bmal1 renders these cells arrhythmic .) Thus we suggest that the observed change in numbers of MHC II− macrophages within the joints is a consequence of loss of BMAL1 in the mesenchymal cell population. The differentiation and function of resident synovial macrophages is governed by (as yet undetermined) tissue-specific cues most likely derived from FLS . Our data suggest that Bmal1 deletion in mesenchymal joint cells indirectly affects the joint macrophage population by disrupting signalling between FLS and synovial macrophages. Indeed, we saw that in stimulated FLS the expression of granulocyte-macrophage colony-stimulating factor, which regulates the proliferation and differentiation of macrophages, was significantly decreased in the absence of Bmal1 (Additional file 1: Table S4). Conversely, M-CSF levels released from FLS lacking Bmal1 were heightened. Taken together, this suggests that Bmal1 is important for the regulation of cytokine signals from FLS which facilitate the proliferation, differentiation and survival of macrophages within the synovium. The disordered macrophage population may contribute to the destructive joint phenotype, and may also affect the propagation and resolution of inflammatory arthritis.
FLS are key players in inflammatory arthritis, releasing a range of pro-inflammatory mediators. Chondrocytes respond to these mediators (e.g., TNF-α and IL-1) by releasing matrix-degrading proteinases, resulting in joint damage. Chondrocytes do also have the capacity to release pro-inflammatory cytokines in response to stimulation , and so may also contribute as a cellular source of inflammatory mediators which damage the joint . Col6a1-Bmal1−/− mice showed enhanced localised inflammation after induction of arthritis. In comparison to wild-type counterparts, the affected limbs exhibited enhanced swelling, increased expression of pro-inflammatory cytokines (e.g., Ccl2, Cxcl1 and Cxcl5) and increased infiltration of neutrophils and Ly6Chi monocytes. Both Ly6Chi and Ly6Clo monocytes were recruited to the joints during collagen antibody-induced arthritis. Interestingly, we did not see significant differences in numbers of infiltrating Ly6Clo monocytes between the two genotypes, only Ly6Chi. Recent work has established that Ly6Clo monocytes have an increased potential to differentiate into osteoclasts and inflammatory macrophages, which play a key role in the destruction of articular bone during arthritis [26, 38]. Furthermore, given the increased levels of Rankl (which plays a role in bone erosion) in the arthritic joints of Col6a1-Bmal1−/− animals, it would be interesting to determine, in future studies, the effects of loss of Bmal1 in joint mesenchymal cells on bone erosion. CCL2 drives monocyte recruitment, whilst CXCL1 and CXCL5 are neutrophil chemoattractants and thus likely explain the observed increase in monocytes and neutrophils. This more severe inflammatory phenotype extended to a systemic response, because arthritic Col6a1-Bmal1−/− mice exhibited increased levels of circulating IL-6. Some stromal cells within secondary lymphoid organs (Peyer’s patches and isolated lymphoid follicles) express Col6a1 . In control experiments, qPCR analysis confirmed the expression of Col6a1 in Peyer’s patches of wild-type and Col6a1-Bmal1−/− animals (Additional file 1: Figure S1b) and Cre expression in Col6a1-Bmal1−/− animals only (data not shown). However, this did not affect expression of Bmal1 (Additional file 1: Figure S1b). Consequently, this increase in systemic IL-6 is most likely a consequence of targeting of joint-resident stroma rather than an extra-articular source.
In vitro studies demonstrated that loss of Bmal1 in FLS promotes enhanced secretion of pro-inflammatory cytokines and reduced secretion of anti-inflammatory cytokines after stimulation, suggesting that Bmal1 within FLS is critical for restraining joint inflammation. Cultured Col6a1-Bmal1−/− FLS demonstrated elevated release of a subset of pro-inflammatory cytokines under naïve conditions. This reflects observations in pulmonary Club cells , macrophages and monocytes , where BMAL1 regulates recruitment of the glucocorticoid receptor (Club cells) and the polycomb repressive complex (macrophages/monocytes) to target genes, resulting in repression of Cxcl5 and Ccl2 transcription, respectively. Given the role of Bmal1 in restraining joint inflammation, it is important to consider the effects of the chronic inflammatory environment on Bmal1 expression. It is reported that TNF-α induces the expression of Bmal1 in human synovial fibroblasts via calcium-dependent pathways [11, 15]. Similarly, studies have shown higher levels of BMAL1 protein in synovium from patients with rheumatoid arthritis than in synovium from patients with osteoarthritis . In light of our findings, we suggest that this enhanced BMAL1 expression in rheumatoid arthritis may in fact act to dampen inflammation.
In the present study, we demonstrated the critical importance of Bmal1 in joint mesenchymal cells in regulating FLS and chondrocyte development within the joints. Our data provide a mechanism to explain the abnormal bone phenotype and altered cellular composition of the joint associated with pre-natal deletion of Bmal1. Additionally, we identified a role for Bmal1 in FLS for restraining local responses to inflammation and highlighted a role for the circadian clock in regulating inflammatory arthritis.
We thank the Transgenic Core Facilities for re-deriving the Col6a1Cre mice, the Flow Cytometry Core Facility for assistance with fluorescence assisted cell sorting, and the Histology Core Facility for advice on tissue processing and staining. We acknowledge the kind gifts of Col6a1Cre mice from Prof. G. Kollias and global Bmal1-knockout mice from Prof A. Reddy. We thank Dr. M. Dudek and Prof. Q. J. Meng for their assistance with isolating articular cartilage and X-ray imaging, and Dr. James Early for assistance with culturing murine bone marrow-derived macrophages.
This work was supported by Arthritis Research UK (grant number 20629 awarded to JEG) and the Medical Research Council (grant number MR/L018640/1 awarded to JEG).
Availability of data and materials
The datasets used and analysed during this study are available from the corresponding author on reasonable request.
JEG and LEH designed the study. LEH, SHD and JEG carried out the experiments, data collection and analysis. JEG, DWR and AJF prepared the manuscript. JEG supervised all the experiments. All authors read and approved the final manuscript.
All animal procedures were carried out in accordance with the United Kingdom Animals (Scientific Procedures) Act 1986 and were subject to local ethical review by the University of Manchester Animals Welfare and Ethical Review Board.
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 15.Yoshida K, Nakai A, Kaneshiro K, Hashimoto N, Suzuki K, Uchida K, Hashimoto T, Kawasaki Y, Tateishi K, Nakagawa N, et al. TNF-α induces expression of the circadian clock gene Bmal1 via dual calcium-dependent pathways in rheumatoid synovial cells. Biochem Biophys Res Commun. 2018;495(2):1675–80.CrossRefGoogle Scholar
- 18.Gibbs JE, Blaikley J, Beesley S, Matthews L, Simpson KD, Boyce SH, Farrow SN, Else KJ, Singh D, Ray DW, et al. The nuclear receptor REV-ERBα mediates circadian regulation of innate immunity through selective regulation of inflammatory cytokines. Proc Natl Acad Sci U S A. 2012;109(2):582–7.CrossRefGoogle Scholar
- 21.Yoo SH, Yamazaki S, Lowrey PL, Shimomura K, Ko CH, Buhr ED, Siepka SM, Hong HK, Oh WJ, Yoo OJ, et al. PERIOD2::LUCIFERASE real-time reporting of circadian dynamics reveals persistent circadian oscillations in mouse peripheral tissues. Proc Natl Acad Sci U S A. 2004;101(15):5339–46.CrossRefGoogle Scholar
- 28.Williams AS, Richards PJ, Thomas E, Carty S, Nowell MA, Goodfellow RM, Dent CM, Williams BD, Jones SA, Topley N. Interferon-γ protects against the development of structural damage in experimental arthritis by regulating polymorphonuclear neutrophil influx into diseased joints. Arthritis Rheum. 2007;56(7):2244–54.CrossRefGoogle Scholar
- 40.Gibbs JE, Ince L, Matthews L, Mei J, Bell T, Yang N, Saer B, Begley N, Poolman T, Pariollaud M, Farrow S, Demayo F, Hussell T, Worthen GS, Ray D, Loudon A. An epithelial circadian clock controls pulmonary inflammation, and glucocorticoid action. Nat Med. 2014;20:919–26.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.