Wild-type C57BL/6 mice, transgenic mice expressing enhanced yellow fluorescent protein (EYFP) under the control of the CD11c promoter, enhanced green fluorescent protein (EGFP) under the control of the human ubiquitin C promoter, and enhanced cyan fluorescent protein (ECFP) under the control of chicken β-actin promoter were obtained from Jackson Laboratories (Margate, UK). Fcγr2b-deficient mice on a C57BL/7 background (4) were provided by Jeff Ravetch (Rockefeller University, New York) and Silvia Bolland (National Institutes of Health, NIAID, Bethesda, MD). For all experiments, both male and female mice were used. For in vivo experiments, mice between the ages of 6 and 12 weeks were used. Mice were maintained in specific pathogen-free conditions at a Home Office-approved facility, with all procedures conducted in accordance with the United Kingdom Animals (Scientific Procedures) Act of 1986 and the GSK Policy on the Care, Welfare, and Treatment of Animals.

I-BET151 (iBET) was supplied by GSK. For in vitro experiments, a stock solution of 10mM was prepared in DMSO solvent (Sigma) and was diluted and added to cell culture at the specified concentrations. For RNA sequencing experiments, BMDMs and BMDCs were treated with iBET at 5.0 and 3.3 μM respectively, added 30 minutes prior to immune complex stimulation. For in vivo experiments, iBET was administered at a concentration of 30 mg/kg, prepared by dissolving in normal saline containing 5% (v/v) DMSO and 10% (w/v) Kleptose HPB. For quantification of FcγR expression in vivo, mice were treated with intraperitoneal injection of iBET every 24 hours over 72 hours, followed by flow cytometry of harvested spleens.

Endotoxin-free ovalbumin (1 mg/mL, Hyglos) was opsonized with a polyclonal rabbit anti-ovalbumin antibody (3.7 mg/mL, Sigma-Aldrich) at 37°C for 1 hour, at a 1:10 (v/v) ratio, to form insoluble immune complexes (IC), for stimulation of cells in culture or administration in vivo. Cells were harvested at defined time points and washed to remove free ICs. For RNA sequencing experiments, BMDMs and BMDCs were stimulated with immune complexes of ovalbumin for 4 hours. For phagocytosis assays, Alexa Fluor (AF)647-conjugated ovalbumin was used (Thermo Fisher), and conjugated with antibody as described above. For in vivo phagocytosis, 0.33 g/kg AF647-conjugated ovalbumin was opsonized with 3.2 g/kg polyclonal rabbit anti-ovalbumin antibody before injection. Phagocytic degradation products were identified by substituting Ova-647 with DQ^™ Ovalbumin (Invitrogen), a conjugate that exhibits fluorescence only after proteolytic degradation.

For murine bone marrow-derived macrophages (BMDMs), bone marrow from femurs and tibias of mice were obtained by flushing with ice-cold sterile PBS and washed in ice-cold PBS. Bone marrow cells were then incubated in complete RPMI (cRPMI; 10% Hyclone FBS, 1% penicillin/streptomycin in RPMI-1640) supplemented with 100 ng/mL murine macrophage colony-stimulating factor (M-CSF, 100 ng/mL, PeproTech). Cells were cultured at 37°C over 5 days, with culture media being replaced on day 3. Adherent cells were harvested for downstream experiments and replated in cRPMI with M-CSF.

For murine bone marrow-derived dendritic cells (BMDCs), bone marrow cells were cultured in cRPMI supplemented with granulocyte-macrophage colony stimulating factor (GM-CSF, 40 ng/mL, PeproTech) over 8 days, with culture media being replaced on day 3 and 6. For BMDMs and BMDCs, cells were incubated with endotoxin-free ovalbumin (Hyglos) or immune-complexed ovalbumin for 24h at 37°C, on day 6 and 9 respectively, together with iBET at the specified concentration or DMSO control.

For murine peritoneal macrophages, the peritoneal cavity was flushed with 5 mL ice-cold sterile PBS. Cells harvested from peritoneal lavage were plated in cRPMI for 1 hour and adherent cells were used for downstream experiments.

Human monocyte-derived macrophages (moMac) were generated from the peripheral blood of healthy volunteers, obtained with informed consents under an IRB/EC approved protocol. Peripheral blood mononuclear cells (PBMCs) were isolated from Histopaque 1077 (Sigma-Aldrich) density separation, and monocytes enriched by negative selection using a MACS-based monocyte purification kit (Miltenyi Biotec). Purified PBMCs were cultured for 5-10 days in cRPMI with human M-CSF (0.1 μg/mL, PeproTech), supplemented every 72 hours, and subsequently stimulated with IC for downstream experiments.

BMDMs or BMDCs in culture were lysed using RLT plus buffer (QIAGEN), vortexed, snap frozen on dry ice, and stored at -80°C. To extract RNA from cell lysates, RNeasy plus micro kit (QIAGEN) were used according to the manufacturer’s instructions. Genomic DNA contamination was removed using Optimal DNA depletion columns (QIAGEN). Purified RNA was eluted in nuclease free water (Ambion) and stored at -80°C. To assess the quality and concentration of purified RNA, RNA pico chip (Applied Biosystems) on Bioanalyzer 2000 (Applied Biosystems) was used according to the manufacturer’s instructions. For all RNAseq experiments, samples had an RNA integrity number of >8. For library preparation, SMARTer^® stranded total RNAseq mammalian pico input kit (Takara) was used according to the manufacturer’s instructions. 5ng of total RNA was used for production of libraries. Library size was assessed using a High Sensitivity DNA chip (Applied Biosystems) on Bioanalyzer 2000 (Applied Biosystems) according to the manufacturer’s instructions. Concentration of the library was determined by qPCR using ROX low KAPA library quantification kit (Roche). Libraries were pooled at an equimolar concentration with up to 10 libraries per pool.

For investigation of the effect of iBET treatment and immune complex stimulation on the transcriptome of murine BMDMs and BMDCs, we performed bulk RNA-sequencing (RNA-seq) using Hiseq 4000 (Illumina) on a 2 x 150bp sequencing run. Sequencing was performed by Genewiz. Pooled libraries were de-multiplexed by Genewiz using Casava (Illumina) before transfer of the data to the University of Cambridge. The Fastq files from libraries prepared using the Takara library prep were trimmed of the first 3 nucleotides of the R1 strand. Contaminating adaptor sequences and poor-quality bases removed (bases with a Phred 33 score of < 30) using Trim Galore! (Babraham bioinformatics). The Illumina library preps were only trimmed for quality. Sequencing quality of the resulting files was assessed using FastQC (Babraham bioinformatics). Fastq files were aligned to the mm10 genome using HISAT2.

Subsequent RNA-seq analysis was performed in the R statistical environment. Resulting data is available on GEO under accension numbers GSE200033 and GSE200226. Reads were counted and assigned to genes using the Featurecount function from the Rsubread package and differential expression analysis was performed using DESeq2 with an appropriate design matrix according to the default workflow, and batch effects removed using the sva package. Figures were plotted using the ggplot2, pheatmap, and EnhancedVolcano packages. Gene Set Enrichment Analysis (GSEA, https://www.gsea-msigdb.org/gsea) was conducted using GSEA 4.1.0 according to developers’ instruction, using the pre-ranked option and classic setting. The rank metric for pre-ranked GSEA was calculated according to the following formula:

Hallmark and Kegg gene sets were downloaded from Molecular Signature Database (MSigDB). Migratory DC gene sets were obtained from differentially expressed genes from previously published single cell RNA-sequencing analysis (GEO: GSE131957 and GSE137710) (36, 37).

RNA extraction was performed using Ambion RNA PureLink kit (Life Technologies) according to the manufacturer’s instructions, from lysates of BMDMs. RNA concentration and purity were determined by Nanodrop spectrophotometry (Thermo Fisher) and subjected to cDNA synthesis using a High Capacity RNA-to-cDNA kit (Life Technologies). Quantitative polymerase chain reaction (qPCR) samples were performed in triplicate using Taqman reagents and the following TaqMan Gene Expression Assay primers and probes (Thermo Fisher). Primers used: Gapdh (Mm99999915_g1), Hprt (Mm03024075_m1), Fcgr1 (Mm00438874_m1), Fcgr2b (Mm00438875_m1), Fcgr3 (Mm00438882_m1), Fcgr4 (Mm00519988_g1), Il1b (Mm00434228_m1), Il6 (Mm00446190_m1), Tnf (Mm00443258_m1). qPCR was performed using the Viia 7 PCR machine (Life Technologies) and gene expression normalised to Gadph or Hprt using the 2^-ΔCt method.

Cells in culture, peritoneal lavage, splenocytes or lymph node cell suspensions were blocked with 0.5% heat-inactivated mouse serum for 15 minutes, followed by extracellular staining for 1 hour at 4°C using a combination of fluorophore-conjugated antibodies. Staining for CCR7 was performed at room temperature. Viability staining was performed using LIVE/DEAD™ Fixable Aqua Dead Cell Stain Kit (Life Technologies) or Zombie™ UV/Aqua Fixable Viability Dye (Biolegend) according to manufacturer’s instructions. Fells were fixed in PBS containing 1% formaldehyde, 0.02% sodium azide, 2% glucose. Cell counting was performed using 123count eBeads (eBioscience). Flow cytometry data collection was performed on a LSRFortessaTM (Becton Dickinson) flow cytometer and data analysed using FlowJo software (BD, version 10.6). Antibodies used are listed in Supplementary Information .

Murine BMDMs and BMDCs were seeded onto glass cover slips sterilized with 70% ethanol prior to placement in 24 well culture plates, prior to addition of IC or iBET where appropriate. After incubation, cells were washed to remove free IC and fixed to cover slips by incubation with FACS FICS for 15 minutes. Cells were blocked and permeabilized in blocking buffer (0.1M TRIS, 1% BSA, 0.1% Triton X-100, 1% normal goat serum), and subsequently incubated with primary antibodies diluted in blocking buffer for 2 hours. Following washing with 0.1M TRIS, cells were incubated for 2 hours in appropriate secondary antibodies. Where relevant, slides were incubated with phalloidin diluted in blocking buffer for 1 hour. Cells were washed and mounted onto slides using VectaShield Hardset Mounting Medium with DAPI (Vector Labs). Images were acquired using a Zeiss 710 or Zeiss 780 confocal microscope and analysed using Imaris software (Oxford instruments). Antibodies used are listed in Supplementary Information .

Thymi of C57BL/6 were harvested and passed through a 70 μm cell trainer. Thymocytes were labelled with cell tracker orange (ThermoFisher) according to manufacturer’s instructions, then rendered apoptotic via treatment with 1 μM dexamethasone (Sigma) for 17 hours (38). Murine BMDMs were pre-treated for 17 hours with iBET (0.5 μM) or DMSO control. Phagocytosis was assessed by flow cytometry, 4 hours after incubation of BMDMs with apoptotic thymocytes.

Wild type C57BL/6 mice were injected with iBET (30 mg/kg.) or solvent control intraperitoneally, prepared as described above, with a second dose given 24 hours later. 1 hour later, Alexa Fluor 647-conjugated immune complexes or ovalbumin control were injected intraperitoneally. After 6 hours, mice were sacrificed, and peritoneal cavity resident immune cells were obtained by flushing with 3 mL of ice-cold sterile PBS (with 3% FBS). Peritoneal lavage was stained for subsequent flow cytometry analysis.

Wild-type C57/BL6 mice were injected intraperitoneally with iBET (30mg/kg) or solvent control, with a second dose given 24 hours later, followed by an intravenous dose of Alexa Fluor 647-conjugated immune complexes or ovalbumin control, injected via tail vein at a dose of 500 ng/g. After 60 minutes, mice were sacrificed and organs flushed with ice-cold PBS. For tissue dissociation of murine kidneys, organs were finely minced, and digested in RPMI-1640 medium containing 10mM Hepes, 1 mg/mL collagenase A (Roche), 0.1 mg/mL DNaseI (Roche) and 2% heat inactivated FCS (Sigma Aldrich) for 20 minutes. Tissue pieces were mechanically dissociated through a 70μm cell strainer and washed with PBS containing 2% FCS, and red blood cell lysis was performed using distilled H₂O containing 0.83% (w/v) NH₄Cl, 0.1% (w/v) NaHCO₃, 100 mM EDTA. Single cell suspensions were subjected to a 44% v/v Percoll gradient (Sigma Aldrich) and washed thoroughly in ice-cold PBS for downstream analysis. For RNA extraction for qPCR, murine kidneys were submerged after harvest in RNAlater™ Stablisation Solution (Invitrogen) for storage, and later homogenised in lysis buffer using Precellys^® homogenisation technology and RNA extracted using PureLink RNA Mini Kit (Invitrogen).

Migration of BMDCs to a chemokine gradient was assessed as previously described by Haessler et al. (39). Briefly, BMDCs derived from transgenic mice expressing a fluorescent tag were stimulated with IC for 24 hours in the presence of iBET or DMSO control. iBET-treated cells were ECFP labelled and DMSO-treated cells EGFP labelled. After washing, cells were resuspended to a concentration of 10 × 10⁶ cells/mL and incorporated within 300 μL of a 1.5 mg/mL bovine collagen gel (20 μL 10x Minimum Essential Media Eagle (Sigma), 20 μL water, 10 μL 7.5% sodium bicarbonate solution (Sigma), 50 μL RPMI 1640, 150 μL 3mg/mL bovine collagen I (Sigma), 50 μL cell suspension in RPMI) loaded into a μ-Slide Chemotaxis chamber (Ibidi), according to the manufacturer’s instructions. The chamber was incubated for 15 minutes at 37°C, inverting once to ensure uniform cell distribution during collagen gelatination. Following incubation and solidification of the gel matrix, CCL19 (100 ng/mL, Peprotech) was applied to the chemoattractant solution reservoir and the sink reservoir was maintained base media. Migratory behaviour was imaged using time lapse confocal microscopy on an inverted Zeiss LSM 780 inverted confocal microscope with a 20X 0.9 NA objective, with stacked images taken every 45 to 60 seconds, over 2 to 4 hours.

CD11cEYFP mice were anaesthetised with isoflurane and footpads were imaged at 915nm using a Zeiss 510 microscope, as previously described previously (40, 41). Briefly, a 70 μm-thick section of the dermis containing lymphatic vessels were scanned at 3 μm Z-steps every 45 seconds for 60 minutes, to generate time-lapse sequences. Imaging was alternated between left and right footpads over an 8 hour period to compare footpads in which 50 μL of ovalbumin or immune complexes had been injected 18 hours prior. Ovalbumin (0.1 mg/mL) was administered subcutaneously into the right footpad and immune complexes (50 μg Ova; 150 μg rabbit anti-Ova) to the left, so that animals acted as their own controls for immune complex stimulation. iBET or DMSO control was administered intraperitoneally two hours prior to ovalbumin or immune complex injection. Qdot^® A655 (Invitrogen Molecular Probes) was administered intravenously prior to imaging, to delineate dermal vasculature, where successful. The investigator performing the imaging was not blinded to information regarding treatment or stimulation.

Dermal DC migration was assessed using a FITC sensitization model, as previously described (20, 42). Ovalbumin or immune-complex ovalbumin was administered subcutaneously to the base of the tail, flank, or groin on the right and left side respectively of wildtype C57BL/6 mice, allowing for comparison of two experimental conditions within the same animal. At the same time, FITC (8 mg/mL, Sigma) was dissolved in equal volumes of acetone and dibutyl phthalate (Sigma) and 25 μL was applied to shaved skin over the site of IC administration. iBET (30 mg/kg) or solvent control was administered intraperitoneally thrice over a 48 hour period beginning at the time of IC administration. 48 hours after FITC application, draining inguinal lymph nodes and non-draining brachial lymph nodes were harvested for comparison. Lymph nodes were homogenized, passed through a 70 μm cell strainer, and digested in collagenase A (1 mg/mL, Roche) DNaseI (1 mg/mL, Roche), and 2% FCS in PBS for 20 minutes. The number of FITC+ cells were quantified by flow cytometry in harvested lymph nodes.

Both in vitro and in vivo migration sequences were processed using Imaris software (Oxford Instruments, version 7.4). The Snapshot tool was used to generate time-lapse movies, still images and track histories. To characterize migration behaviour of individual DCs and obtain track statistics, all DC migration data was analysed using the surface tracking feature in Imaris. Specifically, x- and y-coordinates for each cell tracked were recorded to plot spatial trajectories, to obtain values for displacement and distance covered. Speed values were recorded over the entire sequence. The standalone version of Ibidi’s Chemotaxis and Migration Tool (Ibidi) was used to visualize tracks for in vitro movies. Track displacement was calculated in the x- and y-axes for the in vitro chemotaxis assay. Track length refers to the accumulated distance travelled by each cell over the observed period. The centre of mass represents the averaged point of all cell end points respective to the x- and y-axes.

Fluorescent labelled BMDCs were generated from transgenic EGFP or ECFP mice, as described above, and were stimulated with IC over 16 hours and treated with iBET or DMSO respectively. Cells were washed to remove free IC and injected subcutaneously into the flank or footpad of wild-type recipient C57BL/6 mice. 48 hours following BMDC transfer, draining and non-draining lymph nodes were harvested and fluorescently labelled DCs were quantified by flow cytometry.

With the expression of RNA-seq analyses, described above, statistical analysis was performed using Graphpad PRISM software (version 9.0). Data is expressed as mean ± SEM, mean with individual data points, or median with individual data points as indicated in figure legends. For comparisons, Wilcoxon matched-pairs signed rank test, nonparametric Mann-Whitney-U test or two-tail Student t-test were applied, as described in figure legends. * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001, NS = not significant. Illustration in Figure 6 was made in Affinity Designer (version 1.9).

To gain a global overview of the effect of iBET on FcγR-mediated MNP activation, we performed bulk RNA-sequencing (RNA-seq) on murine bone marrow-derived macrophages (BMDMs) following stimulation with insoluble IgG opsonized ovalbumin immune complexes (ICs) or ovalbumin control alone (Ova) for 4 hours (20), with iBET treatment or DMSO solvent control added 30 minutes prior to IC or Ova stimulation. 377 genes were significantly differentially expressed in IC+DMSO macrophages compared to 181 differentially expressed genes (DEGs) in IC+iBET cells ( Figure 1A ). Notably, the expression of several antimicrobial genes strongly driven by FcγR cross-linking, including Irg1 and Jag1 (43, 44), were abrogated by iBET treatment. Consistent with this, we observed a global dampening of top IC-inducible gene changes in iBET, including transcripts of known importance in inflammatory and NFκB signalling, such as Tnf, Nfkbia, Nfkbie, Cxcl2, Ikbke, etc. ( Figure 1B ). As anticipated, due to the extensive role of bromodomain proteins in transcriptional regulation (24, 25), treatment with iBET alone drove major changes in gene expression. 2263 genes were differentially expressed after just 4.5 hours of exposure to iBET, accounting for 9.3% of gene transcripts captured by RNA-seq ( Supplementary Figure 1A ).

To understand the biological significance of iBET-induced gene changes, we used gene set enrichment analysis (GSEA) ( Figure 1C and Supplementary Figure 1B ). iBET treatment down-regulated inflammatory signalling pathways, including interferon responses, ‘cytokine signalling’ and ‘toll-like receptor (TLR) signalling’ and ‘chemokine signalling’ in both IC and Ova-treated macrophages, with additional effects on genes relating to MNP phagocytosis (‘FcyR-mediated phagocytosis’) and morphology (‘regulation of actin cytoskeleton’). Gene Ontology (GO) biological processes enriched in DEGs unique to IC + iBET treatment ( Supplementary Figure 1C ) confirmed the anti-inflammatory effects of iBET, with negative enrichment of ‘response to IFNγ’ and ‘IL1 signalling pathway’ genes, but interestingly, induction of genes involved in phago-endocytic processes ( Figure 1D ).

Given effects on genes associated with FcyR-mediated phagocytosis, we next sought to profile whether iBET influenced the expression of FcγR genes themselves. RNA-seq of murine BMDMs showed reduced expression of activating receptors Fcgr1 and Fcgr3, as well as the inhibitory Fcgr2b in iBET treated macrophages ( Figure 1E ), effects validated by qPCR ( Figure 1F ). Similarly, in bone marrow-derived dendritic cells (DC), iBET treatment also reduced Fcgr transcripts ( Supplementary Figure 2A ). We confirmed reduced cell-surface FcγR protein expression, (using a FcγRIIb/III-binding antibody, 2.4G2), on BMDM ( Supplementary Figure 2B ) and murine peritoneal macrophages ( Supplementary Figure 2C ) following in vitro iBET treatment. To explore the effects of iBET on FcγR expression in vivo, we treated mice systemically with iBET for 3 days. This resulted in a significant reduction in FcγRIIb/III expression on splenic CD11b+ macrophages and CD11c+ dendritic cells (DCs), with minimal effects on FcγRIIb expression, suggesting that the main effect of iBET in vivo is to reduce activating FcγR expression and to shift the activating-to-inhibitory (A/I) FcγR ratio to promote a less inflammatory state ( Figure 1G and Supplementary Figure 2D ). To demonstrate relevance in humans, we generated macrophages from peripheral blood monocytes, and similarly found a reduction in activating FcγR expression, particularly FcγRIII, following iBET treatment ( Figure 1H and Supplementary Figure 2E ).

Our analysis of the effect of iBET on the transcriptome of BMDMs ( Figures 1C, D and Supplementary Figure 1B ) suggested that phagocytic processes may be enhanced in MNPs, despite the reduced expression of FcγRs we identified ( Figures 1E-H ). To test this functionally, we quantified in vitro phagocytosis of fluorescent-labelled Ovalbumin (Ova) or IgG opsonised Ova by BMDMs, and observed that iBET treatment significantly increased the per cell uptake of both Ova and Ova-IC, as evidenced by an increase in MFI ( Figure 2A ). Macrophage phagocytosis is not only important for clearing circulating immune complexes, but is also required to remove apoptotic cells. Indeed, reduced apoptotic cell clearance is thought to promote the breakdown of tolerance to nuclear antigens presented on the surface of apoptotic blebs, contributing to lupus pathogenesis (45, 46). We therefore generated apoptotic cell tracker-labelled thymocytes (38) and added them to BMDMs that were pre-treated with iBET or DMSO control. iBET treatment resulted in an increased accumulation of apoptotic cell debris within macrophages compared with DMSO-treated cells ( Figures 2B, C ).

Next, we sought to determine the effect of iBET on macrophage phagocytosis in vivo. Mice were pre-treated with iBET or solvent control, followed by the administration of Ova or Ova-IC intraperitoneally ( Supplementary Figure 3A ). 6 hours post-IC challenge there was a similar proportion of Ova-positive macrophages in iBET-treated mice compared with controls however, the per cell uptake of both Ova and Ova-IC was significantly higher in the iBET group ( Figure 2D ). iBET treatment did not affect the frequency of peritoneal cells ( Supplementary Figure 3B ), nor phagocytosis by peritoneal DCs ( Supplementary Figure 3C ).

Increased levels of Ova fluorescence may result from increased antigen uptake or delayed lysosomal degradation leading to accumulation of fluorescent material. To probe this point, we generated insoluble IgG immune complexes with fluorescent DQ-Ova, a compound that fluoresces only upon proteolytic degradation in the lysosome. This did not indicate any delayed appearance of DQ-Ova in the presence of iBET ( Supplementary Figure 3D ). Indeed, at 15 minutes, increased accumulation was already evident. Interestingly, at later timepoints, the difference in DQ signal between iBET and DMSO-treated macrophages did not increase, as one might expect if there was impaired lysosomal degradation. Altogether, this data supports an increase in Fc-mediated phagocytosis in the presence of iBET.

We sought to extend our observations to a tissue context relevant to SLE, specifically the effect of circulating IgG immune complex on kidney resident macrophages. Kidney macrophages may arise from yolk-sac progenitors or be monocyte-derived, characterised by F4/80^high CD11b^int and F4/80^int CD11b^high expression respectively (47), and differ in their functional characteristics (48). Mice were pre-treated with iBET or solvent control, followed by intravenous administration of Ova or Ova-IC ( Supplementary Figure 4A ), where both populations of kidney macrophages have been previously shown to uptake immune complex (49). The administration of intravenous IC led to an increased frequency of F4/80^int CD11b^high kidney monocyte-derived macrophages ( Supplementary Figure 4B ), robust phagocytosis of Ova-IC compared with Ova alone, with little effect of iBET treatment ( Supplementary Figures 4C, D ). Despite this avid phagocytosis of circulating IgG IC, administration of iBET attenuated the IC-associated increase in pro-inflammatory cytokine gene expression in the kidney, including Il1b and Tnf transcripts ( Figure 2E ). Altogether, these data suggest that iBET may enhance the removal of both IgG ICs and apoptotic cells, without a corresponding increase in pro-inflammatory signalling, both potentially beneficial to attenuate lupus pathogenesis.

Additional gene pathways regulated by iBET in BMDM included ‘regulation of actin cytoskeleton’ ( Figure 1C and Supplementary Figure 1B ). Consistent with, we observed morphological changes in iBET-treated BMDMs in vitro, with a reduction in dendritic processes ( Figure 3A ), a decrease in cell surface area and volume, and an increase in sphericity ( Figure 3B ). At higher concentrations, iBET treatment resulted in a decrease in adherent cells observed per high power field ( Figures 3C, D ). Similar effects on cell morphology were observed in murine bone marrow-derived dendritic cells (BMDCs) following iBET treatment, with decreased cell adhesion and increased sphericity ( Figures 3E, F ), but no significant effects on cell viability ( Supplementary Figure 4E ). Overall, these iBET-induced morphological changes might reasonably be expected to influence cell motility and chemotaxis, processes of particular importance for tissue DCs, enabling migration from peripheral tissues to draining lymph nodes, critical for antigen presentation to, and spatial co-localisation with, CD4 T cells (9, 10).

GSEA of RNAseq of BMDCs treated with iBET ( Supplementary Figures 5A, B ) demonstrated negative enrichment of ‘chemokine signalling’ and ‘cell adhesion molecules’ gene sets with iBET treatment, in both Ova and Ova-IC treated DCs ( Supplementary Figure 5C ). Furthermore, in vivo migratory DC gene signatures curated from publicly available single-cell RNA-seq (scRNAseq) datasets of tissue cells (36, 37) were also negatively enriched in iBET-treated BMDCs compared with DMSO-treated cells ( Figure 4A and Supplementary Figure 5D ). Indeed, several genes known to be important for DC migration and maturation (36), including Dock8 (a Cdc42-specific guanine nucleotide exchange factor that is critical for interstitial DC migration) (50), several chemokine receptors, and co-stimulatory molecules (51), were down-regulated in iBET-treated DCs ( Figure 4B and Supplementary Figure 5E ).

A major mechanism by which tissue DCs migrate to lymph nodes is via CCR7-dependent chemotaxis to CCL19 and CCL21 (18, 19), a process stimulated by FcyR cross-linking on DCs (20). iBET treatment of IC-stimulated DCs embedded within a three-dimensional collagen matrix (52) led to reduced chemotaxis towards a CCL-19 gradient compared with DMSO-treatment, although iBET had no effect on overall DC movement or migration speed ( Figures 4C, D and Supplementary Figure 5F ). Consistent with our observations on the effect of iBET on MNP chemotaxis, we found that iBET impaired recruitment of monocyte-derived macrophages to the kidney in our model of systemic IC administration ( Supplementary Figure 4B ).

To determine if iBET inhibited IC-stimulated chemotaxis of endogenous DCs in vivo we used intravital two-photon microscopy to assess the movement of dermal DCs in WT and Fcgr2b ^-/- CD11cEYFP mice (20), the latter a model for enhanced activating FcyR signalling (3). Systemic iBET treatment significantly decreased IC-induced DC track displacement, reflecting directional mobilization, in both strains, with variable effects on DC movement and speed, consistent with an inhibition of chemotaxis but not chemokinesis ( Figures 5A-D ).

Next, we asked whether iBET could prevent dermal DCs from completing their migration to draining lymph nodes, using an established FITC paint model (20) to label dermal DCs, followed by local administration of IC and simultaneous systemic treatment the mice with iBET ( Supplementary Figure 6A ). At 48 hours post-IC stimulation, there was a significant reduction in the frequency of FITC⁺ MHC class II^high CD11c⁺ DCs in draining lymph nodes in iBET-treated mice compared to controls. This reduction was observed in both tissue-resident EPCAM+ Langerhans cell, as well as CD103+ conventional DCs ( Figures 6A, B and Supplementary Figure 6B ). We also noted a reduction in FITC-labelled DCs in non-draining lymph nodes, and that mice treated systemically with iBET had smaller lymph nodes. This was associated with a reduction in live cells in non-draining lymph nodes ( Figure 6C ), suggesting iBET may affect homeostatic leucocyte migration to lymph nodes, even in the absence of an inflammatory stimulus.

The reduction in DC migration observed in iBET-treated animals could be due to direct cell-intrinsic effects on DCs, or due to effects on cells involved in governing chemotaxis, such as chemokine-producing stromal cells. We therefore utilized a DC transfer model in which BMDCs from green fluorescent protein (GFP)-expressing mice and cyan fluorescent protein (CFP)-expressing mice were stimulated in vitro with IC, and simultaneously treated with iBET or DMSO respectively ( Figure 6D and Supplementary Figure 7A ). An equal mixture of iBET and control DCs ( Supplementary Figure 7B ) were administered subcutaneously to wild-type mice, and the number of DCs reaching draining lymph nodes quantified. We observed a significant reduction in the proportion of iBET-treated DCs reaching the draining lymph node compared with DMSO-treated DCs ( Figures 6E, F ).