The circadian clock protein REVERBα inhibits pulmonary fibrosis development

Significance The circadian clock plays an essential role in energy metabolism and inflammation. In contrast, the importance of the clock in the pathogenesis of fibrosis remains poorly explored. This study describes a striking alteration in circadian biology during pulmonary fibrosis where the relatively arrhythmic alveolar structures gain circadian but asynchronous rhythmicity due to infiltration by fibroblasts. Disruption of the clock in these cells, which are not widely implicated in circadian pathophysiology, results in a profibrotic phenotype. Translation of these findings in humans revealed previously unrecognized important circadian risk factors for pulmonary fibrosis (sleep length, chronotype, and shift work). In addition, targeting REVERBα repressed collagen secretion from human fibrotic lung tissue, making this protein a promising therapeutic target.

pulmonary fibrosis as well as control subjects. Patients were classified as having clinically significant pulmonary fibrosis based on reduced lung function (FVC<80%predicted) and a pathological diagnosis of UIP.

In vivo Bleomycin model of pulmonary fibrosis
Male mice (aged 10-14 weeks) were dosed intratracheally with bleomycin (Sigma) (50µL) or vehicle (saline) under anaesthesia (isoflurane). Bleomycin and vehicle treated animals were analysed 28 days after exposure unless stated otherwise. Hydroxyproline was measured on snap frozen lungs, histological analysis was performed on lungs inflated with 4% paraformaldehyde (PFA) that were then left to fix overnight. For precision cut lung slices, lungs were harvested 14-21 days after bleomycin.

Hydroxyproline measurement of tissue collagen
Hydroxyproline was measured using a colourimetric assay as previously described (8).
Briefly, whole lungs were dissected, weighed and homogenised in a mortar and pestle under liquid nitrogen. ~50mg of tissue was dried (80°C for 1 hour) and then hydrolysed in 6N HCl at 100°C for 20 hours. Samples were then oxidised in Citrate-Acetate Buffer with Chloramine T allowing colourimetric determination of hydroxyproline concentration with pdiamethylaminobenzaldehyde in Isopropanol/perchloric acid.

Bioluminescence Microscopy
Precision-cut organotypic lung slices (PCLS) were prepared as described before (9). Briefly, the trachea was cannulated with flexible polyethylene tubing of 0.58 mm in diameter. The lungs were inflated with 2 ml of freshly prepared agarose (2% w/v in HBSS media at 38°C) or until the accessory lobe was fully inflated. The trachea was tied off with 2/0 nylon and the lungs were then placed on ice until the agarose had set (approx. 15mins). The lungs were then fixed onto the stage of a vibrating microtome (Campden instruments integraslice 7550mm) and sectioned at 250μm at 4°C.
After sectioning the slices underwent sequential washes of DMEM to remove agarose and were then placed on a cell culture insert (MilliCell cell culture inserts 0.04μm pores) in luciferin containing culture media (phenol red-free DMEM with 3.5 mg/mL glucose, 25 U/mL penicillin, 25 μg/mL streptomycin, 0.1 mM luciferin, 10 mM HEPES-NaOH) (9). Brightfield imaging was captured on a Zeiss Axiovert 100 with a Plan Achromat 5× 0.16 NA objective equipped with an XL incubator to maintain the cells at 37°C, 5% CO 2 . Bioluminescence images were obtained using a 2.5x objective (Zeiss) and captured using a cooled Andor iXon Ultra camera over a 30 minute integration period acquired using Micro-Manager software (Version 1.4).
Captured images were analysed using MATLAB R2018a to define amplitude and phase of oscillations. In brief a grid was placed across the image where each square represented 1500μm 2 . The same analysis grid was placed on subsequent images captured 30 minutes apart allowing phase and amplitude to be calculated for each defined area. Spikes over time were removed via Hampel outlier identification (10) in order to reduce the influence of random photons and then baseline subtraction was performed for each square using a 24 hour moving average. The acrophase was then calculated by locating the local maxima of the time series. The prominence and relative timing of the acrophase were used to calculate respectively amplitude (in arbitrary units) and phase differences (in hours) per square.
Amplitude and phase was then plotted into a heatmap where the location of each squares corresponds with its morphological position in the original lung slice.

Hoechst staining
After bioluminescent imaging the tissue was fixed in 4% PFA for 2 hours and stored in 70% ethanol at 4°C until Hoechst staining. PCLS were then incubated in PBS with 2 µM Hoechst 33342 for 15 minutes. Staining solution was removed and replaced by PBS. Nuclei were imaged using a Leica TCS SP8 AOBS upright confocal microscope to obtain a 3D optical stack of the whole lung slice. The image presented in the results section is a maximum intensity projection of this 3D stack. Using morphological landmarks, the field of view of this image was matched with the field of view of the previous bioluminescence recording.

Lumicycle Bioluminescence Analysis
Precision cut lung slices and primary lung fibroblasts were also analysed with a lumicycle bioluminescence recorder after being placed on cell culture inserts in a 35mm dish containing media with luciferin as previously described (11). To investigate the effects of lung inflation, lungs were either inflated with 2% agarose solution and then set at 4°C or kept uninflated at 4°C. Sections of lung were then cut and placed in the lumicycle.

Histology and Immunohistochemistry
Paraformaldehyde fixed whole lungs or precision cut lung slices were paraffin embedded and cut sectioned at 5µm. To visualise collagen, sections were dewaxed, rehydrated and stained with Picrosirius Red/Fast Green. For αSMA IHC dewaxed and rehydrated sections were processed for antigen retrieval followed by blocking for endogenous peroxidase with

Generation of primary mouse lung fibroblasts
Whole lungs were dissected to ~2mm 3 pieces and washed in Hank's Balanced Salt Solution (HBSS, Sigma-Aldrich). This tissue was then Collagenase (1.5mg/ml in HBSS, Sigma-Aldrich) digested at 37°C for 2 hours. After Collagenase incubation the homogenate was passed through a 40µm cell strainer, washed in HBSS and finally re-suspended in culture media (DMEM +10% FBS) for fibroblast cell expansion.

Immunofluorescence
For αSMA staining, cells in 35mm dishes were fixed in 4% PFA/0.2% Triton X, followed by ice cold methanol fixation (αSMA). For focal adhesions proteins cells were exposed to ice-cold cytoskeleton buffer (12) for 10 minutes followed by 4% PFA fixation for a further 10 minutes. Cells were then incubated with primary antibodies (indicated in the relevant figures) for 1 hour at room temperature and then washed prior to incubation with a species specific fluorescent secondary antibody (Alexa Flour) for one hour. Following this cell nuclei were then stained with DAPI before dishes were washed, mounted (Prolong Diamond) and cover slipped.
For ECM analysis cells were lysed in 20mM NH 4 OH, 0.5% Triton X-100 in PBS, the DNA was digested using DNAse1 (Roche) and the remaining protein was stained with collagen-1 antibody.
Fiji (13) was used to analyse the immunofluorescence images. αSMA and collagen-1 were quantified using a macro which measured intensity above a threshold which was set the same for all experimental conditions. Focal adhesions were quantified via subtracting the background (rolling ball 50), then a fft band pass filter was applied allowing a threshold to be applied which identified the focal adhesions. All samples were coded and images were quantified by a user blinded to the cell transfection conditions.

RNA-seq
siRNA transfected mLF-hT and Mrc5 cells were lysed and RNA was extracted using the ReliaPrep RNA miniprep system. RNA was sequenced on an Illumina HiSeq 4000. Analysis of these data was performed using the Ingenuity Pathway Analysis software (QIAGEN).

Human lung slice and fibroblast culture
Human lung tissue was collected from patients undergoing carcinoma resection (control) or for fibrotic tissue from human explants after transplantation. Tissue which the pathologist deemed suitable for research was then inflated with 2-3% low boiling point agarose, with the agarose being allowed to set at 4°C. Precision cut lung slices were then cut at 400µm on a vibrating microtome and cultured in DMEM media. TGFβ (2ng/ml), GSK4112 (10µM) or Vehicle (DMSO) treatments were then performed each day with the slices being lysed after 4 days for qPCR analysis or 7 days for supernatant analysis.
To prepare a lung fibroblast culture, fibrotic and non-fibrotic tissue was dissected to ~2mm 3 pieces and placed in culture media in a 15cm 2 dish scored to create points of adhesion. hours after which cells were lysed for qPCR analysis.

ELISA
Collagen 1a1 concentration in cell culture media from precision cut lung slices were assessed using a DuoSet® kit (R&D Systems) as per manufacturer's instructions.

Western immunoblotting
For whole cell analyses, protein lysates were generated by lysis in RIPA buffer (Cell Signalling) containing Protease Inhibitor Cocktail (Roche) and 1 mM PMSF. Extracellular

Statistics
For data from the biobank binary logistic regression was used to examine the relationships between pulmonary fibrosis and chronotype, shift work or sleep duration. Both unadjusted and adjusted analyses were performed taking into account the known epidemiological associations for pulmonary fibrosis (age, sex and smoking) (15) as well as BMI.
Other data was evaluated using Student's t-test, one-way ANOVA or two-way ANOVA for multiple comparisons as indicated. Significant values are shown as p < 0.05 (*), p < 0.01 (**).

Animals
All animals were maintained in 12h:12h light:dark (LD) with food and water supplied ad libitum. mPER2::luc transgenic mice were previously described (16). The Rev-erbα fl/fl mouse (Rev-erbαDBD m ) and Cre drivers targeting club cells (CCSP icre ) and myeloid cells (Lysm cre ) are as previously described (14). The PDGFRβ cre mouse was a kind gift from Henderson and has been previously described (17). The Bmal1 fl/fl mouse has been previously described (9).

Cell Lines
mLF-hT cells have been recently described (18) and were a kind gift from Hinz. Briefly the primary murine lung fibroblasts had telomerase reverse transcriptase virally introduced preventing them from becoming senescent. Human lung fibroblast MRC-5 cells were purchased from the ATCC.

Chemicals
Bleomycin (sulfate from Streptomyces verticillus) was obtained from Sigma. The REVERB agonist, GSK 4112 (19) has been previously described and were obtained from Sigma.

Study Approval
Ethical approvals for both human and animal studies were obtained. The UK Biobank study was approved by the National Health Service National Research Ethics Service (ref.     confidence interval, n= 90 controls and 98 patients with IPF, **=p<0.01 student's t-test).

Supplementary Videos:
Video S1: Per2 bioluminescence in WT mice after bleomycin Video showing bioluminescence over time in a precision cut lung slice in a WT mouse on a mPER2::luc background. The mouse was treated with intra-tracheal bleomycin 14 days before the lung slice was harvested.

Video S2: Per2 bioluminescence in CCSPicre-Bmal1 -/mice after bleomycin
Video showing bioluminescence over time in a precision cut lung slice in a CCSPicre-Bmal1 -/on a mPER2::luc background. The mouse was treated with intra-tracheal bleomycin 14 days before the lung slice was harvested.

Video S3: Per2 bioluminescence in Pdgfrb-Bmal1 -/mice after bleomycin
Video showing bioluminescence over time in a precision cut lung slice in a Pdgfrb-Bmal1 -/on a mPER2::luc background. The mouse was treated with intra-tracheal bleomycin 14 days before the lung slice was harvested.