This study received approval from the Central Adelaide Local Health Network Human Research Ethics Committee (Reference: 17814). The inclusion criteria for participants with sarcoidosis were established based on clinical features and fluorodeoxyglucose (FDG) PET/CT scans. All patients participating in this study are diagnosed with chronic sarcoidosis, exhibiting persistent symptoms for more than two years. These individuals display clinical manifestations associated with pulmonary conditions, and some additionally demonstrate manifestations in the cardiovascular and ocular systems. FDG PET/CT scans were employed for identifying patients with active disease (Table 1). Prior to their participation, all individuals provided informed written consent. For all in vivo experiments, female hβcTg mice aged 10 to 16 weeks were used, and all mice were bred in-house at the RMIT Animal Facility (Bundoora, Australia). Experiments were conducted in compliance with the ethical guidelines of the National Health and Medical Research Council of Australia and ARRIVE, with approval from RMIT University (ethics approval #AEC26548).

The in vitro model of granuloma formation associated with sarcoidosis has been thoroughly described in prior research and closely mirrors the molecular characteristics of granulomas observed in the tissues of individuals affected by sarcoidosis (19–21). Peripheral blood mononuclear cells (PBMCs) were isolated from blood samples of patients with sarcoidosis through Ficoll gradient separation. The isolated PBMCs, at a concentration of 2 x 10⁶ cells/mL, were seeded on collagen I plus fibronectin-coated wells and equilibrated in RPMI media supplemented with 10% human AB serum (Sigma-Aldrich, US) and Penicillin-Streptomycin (Gibco) for 30 minutes at 37°C in an incubator with 5% CO₂. Subsequently, the PBMCs were stimulated with either purified protein derivative (PPD, AJ Vaccines, Denmark)-coated beads (Polysciences, US) or uncoated beads. Granuloma size was evaluated through light microscopy on day 4. Following this evaluation, human monoclonal anti-βc antibody CSL311 (100 µg/mL) or isotype control antibodies (BM4, 100 µg/mL) were added to the appropriate treatment groups, and 50 µL of human AB serum was added to each well. This antibody treatment was repeated on day 5. On day 7, after assessing the granuloma size, the supernatants were collected, snap-frozen in liquid nitrogen, and subsequently stored at -80°C. The cells were then harvested for immunofluorescence staining or RNA extraction.

Bright field microscopy images (100x magnification) were captured using a Nikon Eclipse Ts2 inverted microscope (Nikon, Japan). The granuloma area in each image was measured using Fuji ImageJ (1.54p, National Institutes of Health). Two experimenters were assigned the responsibility of independently capturing and analyzing images, ensuring a rigorous and objective approach to the analysis. Total granuloma area was presented as a percentage of the image capture area. For each condition in every independent experiment, we evaluated a minimum of ten fields of view. The data are presented as biological replicates.

RNA from PBMCs (2 x 10⁶ cells) or granuloma cultures was extracted using TRIzol reagent (Life Technologies, USA) according to the manufacturer’s instructions. 1 µg of RNA was used for complementary DNA (cDNA) synthesis using the QuantiTect reverse transcription kit (QIAGEN, Netherlands). Quantitative real-time PCR was performed using a 1:5 dilution of cDNA with the QuantiTect SYBR Green PCR System (QIAGEN) and specific primers (Supplementary Table S1) on a Rotor-Gene 6000 PCR machine (QIAGEN). PCR assays were performed for 50 cycles (95°C for 15 s, 60°C for 30 s, and 72°C for 30 s). ct values were obtained from the Rotor-Gene Series 6000 Software (QIAGEN), and relative expression levels of mRNA were normalized to GAPDH or 18s RNA control and expressed using the 2^-ΔΔCt method.

Cells grown on coverslips were carefully washed with PBS. Sections were blocked in normal goat serum in CAS-Block (blocking buffer, Thermo Fisher Scientific, USA) for at least 30min in a humidified chamber at room temperature before mouse anti-CD68 antibodies (Invitrogen USA, KP1, 1:200 in blocking buffer) were added to the section and incubated overnight at 4°C. Sections were washed in distilled water, and Alexa Fluor^®-555 goat anti-mouse secondary antibody (Thermo Fisher Scientific USA, 1:400 in blocking buffer) was added to the section and incubated for 1h at room temperature in the humidified chamber in the dark. Sections were washed in distilled water and incubated with BODIPY™ 493/503 (Thermo Fisher Scientific USA, 1 µM) in PBS for 20 min. After a final wash with distilled water, sections were mounted in Fluoroshield mounting medium with DAPI (Abcam, USA). Images were acquired using 40x (PlanApo 40x/1.0 W DIC) objective on a Zeiss LSM 700 confocal system and recorded using Z.E.N. 2011 (Black Edition) software (Carl Zeiss Microscopy, Germany). To ensure unbiased quantification, a second experimenter anonymized the images before the analysis was conducted using Fuji ImageJ (1.54p, National Institutes of Health), with thresholding applied to the corresponding no primary antibodies or isotype control. For each condition in every independent experiment, we evaluated a minimum of three fields of view. The data are presented as biological replicates.

To assess the cytokines released by granulomas, we measured the supernatants collected using an ELISA kit for human GM-CSF, human IFN-γ, human TNF-α, and human IL-1β (eBioscience™) following the manufacturer’s instructions. Each sample was measured in duplicate, and the absorbance was recorded using an Epoch microplate reader (BioTek Instruments, USA).

Granulomas were collected and centrifuged at 180g for 5 min at 4°C. Cells were then lysed in 100 µl ice-cold lysis buffer containing 50 mM Tris-base, 100 mM NaCl, 5 mM EDTA, 10 mM Na4P2O7, 1% Triton X-100, 2 mM NaF, 1X complete protease inhibitors cocktail (Roche). Total cell lysates were separated with SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to nitrocellulose membranes. Membranes were blocked in 5% non-fat dry milk in Tris-buffered saline that contained 0.1% Tween buffer; they were then probed with antibodies raised in rabbit against the phosphorylated form of mTOR(Ser2448, #5536), p70-S6K (Thr389, #9234) and 4E-BP1 (Thr37/46, #2855) at 1:1000 dilution. Membranes were then probed with horseradish peroxidase (HRP)–conjugated antibody against rabbit IgG (1:3000 dilution), and bands were visualized with ECL reagent (Bio-Rad Laboratories) with a LAS4000 imaging system (Fujifilm). Membranes were then stripped and re-probed with antibodies against the total form of mTOR (#2983) and β-actin (#4967) at 1:1000 dilution. All antibodies used were from Cell Signaling Technology. The band intensity was quantified using Fuji ImageJ (1.54p, National Institutes of Health), and data were normalized to the total amount of mTOR or β-actin, then to the untreated control.

Expression and purification of recombinant mouse vimentin was carried out according to (22) with several modifications. A cDNA encoding recombinant mouse vimentin (NCBI RefSeq: NP_035831.2), comprising residues 2–466 fused to an N-terminal 8×His tag, was cloned into the pET41a expression vector (Merck-Millipore, Germany). The construct was transformed into E. coli ClearColi™ cells (Research Corporation Technologies, USA) for LPS-free protein expression. Following overnight expression at 16°C, vimentin predominantly accumulated in inclusion bodies. Inclusion bodies were washed with buffer containing 50 mM Tris-HCl (pH 8.0), 50 mM NaCl, 0.5% Triton X-100, and 1 mM DTT, then solubilized in 50 mM Tris-HCl (pH 8.0), 50 mM NaCl, 8 M urea, and 1 mM DTT. The solubilized protein was purified using IMAC followed by anion exchange chromatography. Refolding was achieved by stepwise reduction of the urea concentration. The refolded vimentin was stored frozen at -80°C in 1x DPBS (Lonza, Switzerland) containing 2 mM TCEP. For immobilization, vimentin was first bound in its denatured state to sterilized Ni-NTA agarose beads (30–40 µm; Cube Biotech, Germany) in PBS supplemented with 8 M urea and 2 mM TCEP. On-column refolding was then performed using the method described above. The immobilization efficiency was estimated at ~1 mg of vimentin per 100 µl of beads. The final product (His-Vim) was stored at 4°C as a 50% slurry in 1x DPBS (Lonza, Switzerland) containing 2 mM TCEP. Abbreviations: DPBS, Dulbecco’s Phosphate Buffered Saline; DTT, Dithiothreitol, IMAC, Immobilized metal affinity chromatography; LPS, Lipopolysaccharide; Ni-NTA, Nickel-Nitrilotriacetic acid; TCEP,Tris(2-carboxyethyl)phosphine.

Female hβcTg mice were subcutaneously injected with 50 μg of free vimentin in conjunction with complete Freund’s adjuvant (CFA, Sigma-Aldrich, US) on day 0. Subsequently, the mice underwent an intravenous challenge with histidine-tagged vimentin bound to Ni-NTA agarose beads (His-Vim, ~40,000 beads), on days 9 and 12. To investigate the anti-βc signaling related to lung sarcoidosis, mice were randomly assigned to either the CSL311 or the isotype control treatment groups. The treatment involved intravenous administration (i.v., 50 mg/kg) to the vimentin-treated hβcTg mice on experimental days 0, 8, and 11. All mice were euthanized on experimental day 15 by intraperitoneal injection of pentobarbitone (1,625 mg/kg) to evaluate lung inflammation and pathology. Details on the allocation of animals to different analyses are provided in Supplementary Table S2.

Lung lobes were fixed, paraffin-embedded, and coronally sectioned at 4 μm thickness. Tissue sections were stained with Hematoxylin and Eosin (H&E). Whole-slide imaging was performed using an Olympus Slide Scanner (Olympus, Japan). Lung granulomas associated with the vimentin-bound beads were manually counted. The number of lung granulomas per mm² of tissue area was calculated as the number of granulomas identified on a given lung section divided by the lung section area. CellSens software (Olympus, Japan) was used for morphometric assessments of mean area (µm²) of detected granulomas. A region of interest (ROI) was manually defined to delineate the granuloma boundary, and the total area and Circle Equivalent Diameter of each granuloma were quantified using automated image analysis.

To measure lung inflammatory cells, BAL was performed. Total cells recovered from the BAL were enumerated using a hemocytometer. Cytospins were prepared and stained with a Hemacolor^® Rapid staining kit (Merck Millipore, USA) for differential cell quantification. BAL cells were pelleted by centrifugation and snap-frozen in liquid nitrogen prior to -80 °C storage. Lungs were perfused free of blood with ice-cold PBS. Lung lobes were snap-frozen in liquid nitrogen prior to -80 °C storage and subsequent RNA-sequencing analysis.

RNA was extracted from crushed frozen lung tissue using a RNeasy Plus kit (Qiagen, Germany) according to manufacturer’s instructions. The extracted RNA was subsequently used for bulk RNA-seq by the Australian Genome Research Facility (AGRF, Melbourne, Australia). RNA purity and integrity were assessed using an Agilent TapeStation System (Agilent Technologies, US) prior to library construction with the Illumina Stranded mRNA Prep kit (Illumina, US). Twenty million 150-bp paired end reads were performed on the Illumina NovaSeq X plus platform, and primary sequence data was then generated with the Illumina DRAGEN BCL Convert 07.021.645.4.0.3 pipeline. The raw sequencing data was analyzed in FastQC and trimmed using TrimGalore-0.6.7 to remove low-quality reads and adaptor sequences. The STAR aligner (v 2.7.4a) was used to map reads against the Mus musculus genome (Build version mm39) and featureCounts from the Subread package (v2.0.3) was used to generate the raw gene counts. Low gene counts were removed in ‘edgeR (v 3.38.4)’, followed by the trimmed mean of M-values (TMM) normalization and differential gene analysis. Differentially expressed genes (DEGs) were defined as having a false discovery rate (FDR) < 0.05. DEGs were visualized using ‘ggplot2 (v 3.4.4)’ and ‘VennDiagram (v 1.7.3)’. Gene Set Enrichment Analysis (GSEA) and Gene Set Variation Analysis (GSVA) were conducted using established gene sets from the Human Phenotype Ontology (HPO) database, the Kyoto Encyclopedia of Genes and Genomes (KEGG) database and the Molecular Signatures Database (MSigDB). ‘ClusterProfiler (v 4.10.1)’, ‘GSVA (v 1.50.5)’, ‘enrichplot (v 1.14.2)’ and ‘ggplot2’ were used for the ORA/GSEA/GSVA analysis and data visualization. RNA-seq data utilized in this study were from 6 control (SAL) mice, 5 vimentin-treated mice injected with isotype (VIM-ISO), and 5 vimentin-treated mice injected with CSL311 (VIM-CSL311).

Lung microarray datasets GSE16538 and GSE19976, as well as the skin RNA-Seq dataset GSE169146, were downloaded from the Gene Expression Omnibus (GEO). Datasets GSE16538 and GSE19976 were merged and inter-study batch effects were removed with ComBat using sva (v3.50.0) (23). Samples were categorized as normal (n=6), sarcoidosis (n=6), self-limiting sarcoidosis (n=8) or progressive sarcoidosis (n=7) as designated in the original papers. Differential expression was analyzed with limma (v3.58.1), and transcripts with FDR < 0.05 were deemed differentially expressed genes (DEGs). Gene set enrichment analysis (GSEA) was performed on pre-ranked DEG list using clusterProfiler (v4.10.1) and mapped against gene sets in MSigDB, KEGG and HPO database. Single-sample pathway activity was quantified by GSVA (v1.50.5), and core-enriched genes were visualized as row-scaled heatmaps using ComplexHeatmap (v2.18.0).

All data were statistically analyzed using GraphPad Prism 10.0 (GraphPad, US) or R Studio. Data are presented as the mean ± SEM of biological replicates. Where detailed and appropriate, one-way ANOVA or two-way ANOVA with Tukey or Bonferroni post-tests was used for multiple comparisons. Asterisks above a group indicate that group’s level of significance relative to the control (*P <.05, **P <.01, ***P <.001). GSEA was conducted with ‘ClusterProfiler (v 4.22)’ (21). GSVA was performed using ‘GSVA (v1.50.5)’ (22).

To investigate βc cytokine signaling in sarcoidosis, we conducted an extensive analysis of 12 publicly available “omic” sarcoidosis datasets. This analysis encompassed both single-cell and bulk transcriptomics, to examine the relevance of the β-common pathway to the activity and severity of the disease. Notably, our findings identified a consistent correlation between the activation of the βc signaling pathway and granuloma samples from sarcoidosis patients across various organs (Table 2). We next performed an in-depth analysis of the transcriptomic datasets, comparing sarcoidosis lung tissues with normal lung tissues (GSE16538, PMID: 19218196) (24) and lung tissues from patients with nodular, self-limiting disease versus those with progressive, fibrotic disease (GSE19976, PMID: 20194811) (25). The two datasets were combined for co-analysis, and batch effects were removed. Gene set enrichment analysis (GSEA) revealed significant enrichment of βc cytokine signaling, both collectively and individually, in progressive sarcoidosis compared to self-limiting disease or normal lung tissue (Figures 1A, B). A heatmap analysis further demonstrated an upregulation of key genes involved in these pathways, particularly in the progressive sarcoidosis group, including CSF2RB, CSF2RA, IL3RA, STAT5A, and STAT1 (Supplementary Figure S1). Single-sample gene set variation analysis (GSVA) further confirmed that the enrichment score (ES) for βc cytokine signaling was significantly higher in the progressive sarcoidosis group compared to the self-resolving group and normal controls (Figures 1A, B). Additionally, the cutaneous sarcoidosis dataset (GSE169146, PMID: 35668129) provided valuable insights by comparing skin samples from patients with those from a control group (26). The analysis highlighted a significant enhancement in the activation of all βc cytokine pathway signatures (Figure 1C). Furthermore, it revealed an increase in expression levels of specific receptor components, including CSF2RB, CSF2RA, and IL3RA in the diseased tissues (Figure 1D). These findings suggest that signaling activation via the βc receptor is associated with sarcoidosis and disease severity.

We further validated the above findings in PBMC isolated from patients diagnosed with sarcoidosis. Our analysis revealed that the expression levels of the receptor α subunits were not altered across the groups. However, we observed a significant increase in βc receptor gene expression in PBMCs from sarcoidosis patients. Notably, PBMCs from those with active sarcoidosis exhibited the highest levels of βc receptor expression, suggesting a possible peripheral signature (Figures 1E–H). Furthermore, our comparative analysis revealed no significant differences in the mRNA expression levels of GM-CSF, IL-3, and IL-5 between healthy donors and patients with sarcoidosis (Supplementary Figure S2).

In previously reported in vitro granuloma models, the incubation of PBMCs obtained from patients with sarcoidosis with tuberculin PPD-coated beads led to the formation of granuloma-like aggregates (19–21). This model is a well-established system for studying human granuloma formation in vitro and was utilized in this study to evaluate the effects of βc receptor antagonism on granuloma formation. Following a 4-day incubation with PPD-coated beads, we carefully examined the granuloma-like cell aggregates under an inverted microscope, which enabled us to confirm the presence of granuloma aggregates and effectively assess both their area and size. Following the formation of granuloma-like aggregates, we administered either CSL311 or isotype control antibodies. By day 7, we observed that treatment with CSL311 resulted in a significant reduction in the area of the aggregates compared to both the untreated and isotype-treated control groups (Figures 2A, B). Furthermore, the granulomas treated with CSL311 showed a marked decrease in size, resulting in a predominance of smaller cell aggregates at the time of assessment (Supplementary Figure S3). Additionally, as part of our prophylactic treatment strategy, we incubated PBMCs with CSL311 or an isotype control antibody prior to their exposure to PPD. This approach demonstrated that CSL311 nearly completely inhibited the initiation of granuloma formation (Supplementary Figure S4).

Granulomatous inflammation involves the regulatory effects of several cytokines produced by local mononuclear phagocytes, T cells, and other cells within the inflammatory microenvironment. In the context of sarcoidosis, granulomatous inflammation is characterized by the dominant expression of the critical cytokines, GM-CSF, IL-1β and T helper 1 (Th1) cytokines, including IFN-γ and TNF-α. In this study, we evaluated the levels of these cytokines in the culture supernatant on Day 7. Our data show that the in-vitro PPD-induced granulomas produced all these cytokines, which may further facilitate granuloma expansion. Importantly, we found that treatment with CSL311 significantly reduced the release of GM-CSF and IFN-γ, while also demonstrating a trend toward lower levels of IL-1β and TNF-α (Figures 2C-F).

It is well-established that chronic signaling through the metabolic checkpoint kinase mTOR promotes macrophage granuloma formation and is an indicator of sarcoidosis progression (27). In this study, we assessed the activities of the mTOR kinase and its principal substrates, eIF4E-binding protein (4E-BP1) and p70 S6 kinase (S6K), utilizing cell lysates obtained from granulomas at Day 7. Immunoblot analysis demonstrated an increase in the phosphorylation of mTOR, 4E-BP1, and S6K in granulomas induced by PPD stimulation in comparison to untreated and control beads controls. Notably, treatment with CSL311 led to a reduction in mTOR pathway activity (Figure 3A).

Human sarcoid granulomas are characterized by the presence of lipid-laden CD68⁺ macrophages that exhibit aberrant lipid metabolism. The βc cytokines, particularly IL-3 and GM-CSF, play a crucial role in lipid metabolism and lipid droplet (LD) dynamics by activating mTOR signaling. We investigated transcriptional alterations in genes associated with lipid metabolism. We found that the expression of acetyl-coenzyme A carboxylase alpha (ACACA), which serves as the rate-limiting enzyme in de novo fatty acid synthesis, increased following PPD stimulation (Figure 3B). This upregulation was effectively inhibited by treatment with CSL311. In contrast, the expression of acyl-coenzyme A dehydrogenase (ACADM), an enzyme essential for the catabolism of medium-chain fatty acids, remained unaffected by PPD stimulation (Figure 3C). Further assessment of lipid droplet synthesis in CD68⁺ macrophages within granulomas was conducted using BODIPY staining techniques (Figure 3D). The results showed that treatment with CSL311 significantly reduced the proportion of BODIPY⁺ macrophages in the granulomas (Figure 3E). Additionally, both the average size of lipid droplets (Figure 3F) and their quantity within CD68⁺ macrophages (Figure 3G) were diminished post-treatment with CSL311, suggesting a significant impact on lipid metabolism associated with βc receptor antagonism.

We subsequently conducted a pre-clinical model of lung sarcoidosis utilizing humanized hβcTg mice. These humanized transgenic (Tg) mice are characterized by the expression of the human (h) βc receptor (hβcTg) and the absence of murine βc/βIL-3 receptors, as previously described (16). The administration protocol, as previously illustrated, involved the subcutaneous injection of 50 μg of free vimentin, followed by an intravenous infusion of histidine-tagged vimentin bound to Ni-NTA agarose beads (His-Vim). To establish a proof-of-concept for βc receptor antagonism as a potential therapeutic strategy, we aimed to inhibit βc signaling by administering CSL311 to vimentin-treated mice (Figure 4A). It is important to highlight that this study did not include a group that received only Ni-NTA beads. Additionally, prior research showed that vimentin-beads alone elicited only a minimal level of granulomatous inflammation, and the addition of vimentin immunization markedly potentiated granuloma formation, supporting an antigen-specific effect from vimentin (22). Histopathological analysis conducted on day 15 revealed no granuloma formation in the saline control group (data not shown). In contrast, the vimentin-treated groups exhibited distinct lung granulomas (red arrowhead) along with the presence of multinucleated giant cells (MGCs) (Figure 4B. white arrowhead). Notably, while treatment with CSL311 (VIM-CSL311) did not result in a reduction in the total number of granulomas, it significantly decreased the average size of the granulomas compared to the isotype treatment group (VIM-ISO, Figures 4C, D). Furthermore, in the analysis of bronchoalveolar lavage (BAL) cellular components, CSL311 significantly reduced vimentin-induced total BAL cell counts, driven by reductions in BAL macrophages, neutrophils, and lymphocytes (Figure 4E).

To gain deeper insights into how βc receptor antagonism reduced granuloma burden at a molecular level, we conducted RNA-seq analysis of lung samples collected from the vimentin model. A heatmap analysis of vimentin-induced differentially expressed genes (DEGs) revealed distinct transcriptional profiles between saline controls (SAL group) and VIM-ISO group. Remarkably, VIM-CSL311-treated mice clustered closely with the SAL group, indicating that CSL311 treatment reversed vimentin-induced changes in gene expression (Figure 5A). Volcano plots highlighted key significantly upregulated and downregulated DEGs in VIM-ISO group compared to SAL group (Figure 5B) and VIM-CSL311 group compared to VIM-ISO group (Figure 5C). After vimentin-treatment (VIM-ISO vs SAL), genes associated with lung remodeling (Spp1, Npy, Saa1 and Saa3), macrophage activation (Mmp12, Mmp13, Arg1 and Trem2), inflammatory chemokines and cytokines (Ccl24, Cxcl2, Cxcl3 and Il1b), and MGC formation (Dcstamp) were among the most significantly upregulated DEGs. Conversely, genes linked to tissue integrity and adhesion (Pgm5, Pcdhb2) were among the most downregulated DEGs (Figure 5B). Following CSL311 treatment (VIM-CSL311 vs VIM-ISO), genes associated with myeloid cell response (Ccl6, Marco, Cd300lf, Tfec, Cd69, Clec4n and Clec4a2), T cell response (Ccl17 and Sh2d1b1), mast cell and eosinophil cell response (Mcpt8, Mcemp1, Fcer1a, Prss34, Ear1, and Rnase2b), and lipid metabolism (Fabp1, Cpne5, Vism2a) were significantly downregulated (Figure 5C). CSL311 treatment also affected the expression levels of several key genes associated with granuloma inflammation, immune cell infiltration into granulomas, fibrosis related to sarcoidosis, and lipid metabolism in foamy macrophages (Supplementary Figure S5). Next, a comprehensive pathway analysis was conducted utilizing the MSigDB database. In mice treated with vimentin, a significant upregulation of energy metabolic pathways was observed, including mitochondrial respiration, the tricarboxylic acid (TCA) cycle, and oxidative phosphorylation. Promisingly, these alterations were effectively reversed following treatment with CSL311. Additionally, the analysis highlighted the activation of inflammatory pathways associated with chemokine and cytokine signaling, particularly IL-4 and IL-13 signaling, in the disease state. The study also offered valuable insights into the chemotaxis of myeloid cells—specifically neutrophils, eosinophils, macrophages, and dendritic cells—as well as lymphocyte migration. Treatment with CSL311 demonstrated a significant capacity to suppress key inflammatory pathways (Figure 5D). Additional findings indicated that pathways related to tissue remodeling, particularly those implicated in lung fibrosis, matrisome-associated genes, and secreted factors, were significantly upregulated in the disease condition. These pathways were subsequently attenuated through treatment with CSL311 (Figure 5D).

Since sarcoidosis is characterized as a granulomatous disease, we analyzed the pulmonary sarcoidosis patient dataset, utilizing the Granulomatosis HP gene set from the Human Phenotype Ontology (HPO) database. This gene set offers a comprehensive framework for understanding the various pathological processes involved in granuloma formation. Our analysis revealed that genes related to human granulomatosis were markedly upregulated in progressive sarcoidosis tissue compared to both normal tissue (NES=1.868) and tissue from patients with self-limiting disease (NES=1.927) (Supplementary Figures S6A, B). Following this, we applied the mouse homologs of the Granulomatosis gene set to analyze our mouse RNA-seq data using GSEA. As depicted in Figure 6A, we found that genes associated with this pathway were significantly upregulated in vimentin-induced sarcoidosis compared to saline controls (VIM-ISO vs SAL, NES=1.92). Moreover, our RNA-seq analysis in mice further demonstrated that treatment with CSL311 effectively reduced the expression of genes associated with the Granulomatosis gene set (VIM-CSL311 vs VIM-ISO, NES=-1.942) (Figures 6B, C). Our analysis also revealed valuable insights, indicating that pathways related to lipid metabolism and atherosclerosis were activated in the vimentin model (Figure 6D), aligning well with findings from human sarcoidosis patient datasets (Supplementary Figures S6C, D). Importantly, we observed that the lipid metabolism pathway was downregulated following treatment with CSL311 (Figures 6E, F). This change is consistent with the results from the in vitro model, where we observed a significant reduction in lipid droplet formation in granuloma macrophages following CSL311 treatment. These findings suggest potential therapeutic avenues that warrant further exploration.