Aberrant Lipid Metabolism in Macrophages Is Associated with Granuloma Formation in Sarcoidosis

Abstract

Rationale: Chronic sarcoidosis is a complex granulomatous disease with limited treatment options that can progress over time. Understanding the molecular pathways contributing to disease would aid in new therapeutic development. Objectives: To understand whether macrophages from patients with nonresolving chronic sarcoidosis are predisposed to macrophage aggregation and granuloma formation and whether modulation of the underlying molecular pathways influence sarcoidosis granuloma formation. Methods: Macrophages were cultivated in vitro from isolated peripheral blood CD14⁺ monocytes and evaluated for spontaneous aggregation. Transcriptomics analyses and phenotypic and drug inhibitory experiments were performed on these monocyte-derived macrophages. Human skin biopsies from patients with sarcoidosis and a myeloid Tsc2-specific sarcoidosis mouse model were analyzed for validatory experiments. Measurements and Main Results: Monocyte-derived macrophages from patients with chronic sarcoidosis spontaneously formed extensive granulomas in vitro compared with healthy control participants. Transcriptomic analyses separated healthy and sarcoidosis macrophages and identified an enrichment in lipid metabolic processes. In vitro patient granulomas, sarcoidosis mouse model granulomas, and those directly analyzed from lesional patient skin expressed an aberrant lipid metabolism profile and contained increased neutral lipids. Conversely, a combination of statins and cholesterol-reducing agents reduced granuloma formation both in vitro and in vivo in a sarcoidosis mouse model. Conclusions: Together, our findings show that altered lipid metabolism in sarcoidosis macrophages is associated with its predisposition to granuloma formation and suggest cholesterol-reducing therapies as a treatment option in patients.

Keywords: granuloma; lipid-laden; macrophages; non-Löfgren’s; sarcoidosis.

Fig. 1. Chronic sarcoidosis patient monocyte-derived macrophages supplemented with GM-CSF form in vitro granuloma aggregates

(A) Photomicrographs of day 6 monocyte-derived macrophages from chronic sarcoidosis patients and age-matched healthy controls (n=5) with corresponding quantification of macrophage aggregation according to cluster size. (B) Quantification of the average area and number of large and medium-sized macrophage aggregates/clusters that are bigger than 8x 10⁴ μm² taken from Fig. 1A. (C) Representative immunofluorescent staining of a macrophage aggregate for CD206 (pink) and CD68 (yellow) expression. Nuclei stained with DAPI (blue). Staining performed for 4 patient samples. (D) Quantification of F-actin/Phalloidin expression (mean fluorescent intensity, MFI) in aggregating macrophage clusters vs remaining cells (n=4 patient samples). (E) Immunofluorescent staining of a macrophage aggregate for F-actin/Phalloidin (orange) expression. Nuclei stained with DAPI (blue). (F) Quantification of p-S6 expression (mean fluorescent intensity, MFI) found in aggregating macrophage clusters vs remaining non-clustering cells and frequency of p-S6⁺ cells amongst total cells (n=4 patient samples). Representative image of sarcoidosis macrophage aggregate stained with phospho-S6 ribosomal protein (p-S6) antibody.

Fig. 2. Bulk RNA sequencing analyses of chronic sarcoidosis monocytes and macrophages

(A) Principal component analysis (PCA) plot of disease (sarcoidosis, n=5) and healthy control (n=4) monocytes and macrophages. (B) Bar plot of the number of enriched gene ontology (GO) sub-categories (x-axis) under each category (y-axis), based on genes differentially expressed in sarcoidosis monocytes versus control monocytes, and sarcoidosis macrophages versus control macrophages. (C) Bar plot of the pathways from MSigDB, Biocarta or Elsevier Pathway databases enriched in gene transcripts upregulated in sarcoidosis monocytes compared to healthy control monocytes (Enrichr pathway gene-set analyses, x-axis -log10 p-adjusted values). (D) Category-gene net plot of enriched pathways and genes from transcripts upregulated in sarcoidosis macrophages compared to healthy controls. Node size denote number of genes in pathway. Gene transcripts involved in individual pathways are represented extending from pathway categories. Color of gene transcript nodes represent its expression (log2 Fold change) relative to control. (E) Full STRING protein-protein interaction (PPI) analysis of corresponding proteins of genes upregulated in sarcoidosis macrophages. Nodes clustered using the Markov Cluster (MCL) Algorithm with inflation parameter =3. Cluster color labels clusters containing nodes with high interaction score. Inter-cluster edges are represented with dotted lines. Protein-protein enrichment value: 1.11e-16.

Fig. 3. Lipid metabolism pathways play a role in the formation of sarcoidosis in vitro granulomas

(A) BODIPY 493/503 neutral lipid expression in sarcoidosis macrophage clusters (n= 4 patient samples) => 30 microns wide, and => 70 microns μm wide versus all cells (identified by DAPI nuclei staining) in sarcoidosis monocyte-derived macrophage culture. (B) Representative images of sarcoidosis macrophages at day 6 of culture with GM-CSF after 3-day treatment with standard media with 10% fetal calf serum and control DMSO diluent (SDM) or 10% lipoprotein-deficient serum and 5 μM Lovastatin. Green spots: BODIPY 493/503 neutral lipids, DAPI (blue). (C) Number of large macrophage aggregates (>=70μm width) after 3 days of treatment with standard media with 10% fetal calf serum and control DMSO diluent (SDM), or 10% Lipoprotein deficient media alone (LPD), or 5μM Lovastatin or 12μM Lovastatin and 10% lipoprotein-deficient serum. Two-tailed paired T-tests carried out. (D) Quantification of in vitro granuloma size (area in m²) after 3 days of treatment with standard media with 10% fetal calf serum and control DMSO diluent (SDM), or 10% Lipoprotein deficient media alone (LPD), or 5μM Lovastatin or 12μM Lovastatin and 10% lipoprotein-deficient serum. Two-tailed paired T-tests carried out. Significant Fixed effect (type III) p < 0.01 (mixed effect analysis). (E) Treated and control-treated sarcoidosis macrophages (treatment as described in Fig. 3B) were stained with Zombie Red (Biolegend) dye to assess viability by flow cytometry (n= 6 patient samples). (F) Total intensity (integral density) of BODIPY 493/503 neutral lipid expression in sarcoidosis macrophages treated with standard media with 10% fetal calf serum and control DMSO diluent (SDM), or 10% Lipoprotein deficient media alone (LPD), or 5μM Lovastatin or 12 μM Lovastatin and 10% lipoprotein-deficient serum was measured and values normalized to the SDM macrophage condition and expressed as a percentage (%). Two-tailed paired T-tests carried out. Significant Fixed effect (type III) p < 0.01 (mixed effect analysis). (G) Total intensity (integral intensity/density) of SREBF1 in sarcoidosis macrophages treated with standard media with 10% fetal calf serum and control DMSO diluent (SDM), or 10% Lipoprotein deficient media alone (LPD), or 5μM Lovastatin or 12μM Lovastatin and 10% lipoprotein-deficient serum was measured and values normalized to the SDM macrophage condition and expressed as a percentage (%). Two-tailed paired T-tests carried out. Significant Fixed effect (type III) p < 0.05 (mixed effect analysis). (H) Phospho-S6 (pS6) expression in sarcoidosis macrophages treated with 10% lipoprotein-deficient serum and control DMSO or 5μM Lovastatin and 10% lipoprotein-deficient serum determined by flow cytometry (n=6 patient samples). (I) Ki-67 expression in sarcoidosis macrophages treated with 10% lipoprotein-deficient serum and control DMSO or 5μM Lovastatin and 10% lipoprotein-deficient serum determined by flow cytometry (n= 6 patient samples). (J) Frequency of ki-67+ stained cells determined by flow cytometry after treatment with control DMSO diluent in standard media containing 10% fetal calf serum or 5μM Lovastatin and 10% lipoprotein-deficient serum (n= 6 patient samples).

Fig. 4. Lesional skin granulomas from sarcoidosis patients contain increased neutral lipids and an aberrant lipid metabolism profile

(A) Representative immunofluorescence images of lesional skin stained with CD68, BODIPY 493/503 or DAPI from a chronic sarcoidosis patient (n=4 patient samples). (B) Quantification of BODIPY 493/503 neutral lipid expression in lesional skin granuloma vs remaining tissue in lesional skin vs non-lesional skin (n=4 in duplicates). (C) Frequency of BODIPY 493/503 neutral lipid⁺ CD68⁺ macrophages in sarcoidosis lesional skin granuloma vs remaining tissue in lesional skin vs non-lesional skin. (D) Representative images of sarcoidosis lesional skin stained with CD68, SREBF1 and DAPI (n=4 patient samples). (E) SREBF1 expression in CD68⁺ macrophages found in sarcoidosis lesional skin, remaining non-granulomatous tissue in lesional skin (n=4) and non-lesional skin (n=3). (F) Frequency of SREBF1⁺ CD68⁺ macrophages in sarcoidosis lesional skin, remaining tissue in lesional skin (n=4) and non-lesional skin (n=3). (G) Frequency of MARCO⁺ CD68⁺ macrophages in sarcoidosis lesional skin, lesional skin remaining tissue (n=4) and non-lesional skin (n=3). (H) Violin plots showing expression of MARCO in granuloma-associated (GA) macrophages and homeostatic macrophages (Cluster 0 and Cluster 1 from Krausgruber et. al., 2023). (I) Violin plots showing expression of lipid metabolism pathways genes (ETHE1, EBP, MSMO1, IDI1, SQLE, LDLR and SCD) in granuloma-associated (GA) macrophages and homeostatic macrophages (Cluster 0 and Cluster 1 from Krausgruber et. al., 2023).

Fig. 5. Macrophages from skin of sarcoidosis model mice display an aberrant lipid metabolism profile

(A) UMAP of scRNA-seq transcriptome profiles from swollen paw and tail skin of female age-matched Tsc2^{floxed/floxed} CD11c-Cre (Tsc2^KO) sarcoidosis model mice and Tsc2^{floxed/floxed} control (Tsc2^WT) littermates between ages 36-41 weeks old (n=3 each). (B) UMAP from (Fig. 5A) annotated with expression of macrophage markers Cd68 and Adgre1 (F4/80). (C) UMAP from (Fig. 5A) annotated with expression of macrophage markers Mertk and Fcgr1 (CD64). (D) UMAP from (Fig. 5A) annotated with expression of lipid metabolism cluster genes – Srebf1, and Lpin1. (E) Violin plots showing expression of lipid metabolism associated genes (Ethe1, Ebp, Msmo1) in Tsc2^KO sarcoidosis mouse skin macrophages compared to Tsc2^WT control macrophages. Student’s t-test was performed. p= <0.001 for all three plots. (F) Frequency of BODIPY 493/503⁺ macrophages (gated on CD11b, activated macrophage marker Mac-2, CD64 and F4/80) amongst all live 7AAD-cells in the skin of sarcoidosis model mice (Tsc2^KO) vs control Tsc2^WT mice (33-36 weeks old male mice, n=4 each) and representative dot plot showing population of F4/80, BODIPY 493/503 expressing cells amongst the gated macrophages. (G) Expression of BODIPY 493/503 neutral lipids in skin macrophages of sarcoidosis model mice (Tsc2^KO) vs control Tsc2^WT mice (33-36 weeks old male mice, n=4 each).

Fig. 6. Reduction of disease severity after mice with severe sarcoidosis were treated with statin and cholesterol-deficient diet

(A) Hind paw thickness of TSC2^KO sarcoidosis mice after treatment with cholesterol-deficient diet (CFT) and Atorvastatin/ Lipitor or control diet (n=9) and control DMSO diluent (n=7). (B) Representative image of sarcoidosis mice after treatment with cholesterol-deficient diet (CFT) and Atorvastatin/ Lipitor or control diet and control DMSO diluent. (C) Measurement of sarcoidosis mouse spleen length after treatment with cholesterol-deficient diet (CFT) and Atorvastatin/ Lipitor or control diet and control DMSO diluent. (D) Lung weight to body weight ratio index of sarcoidosis mice after treatment with cholesterol-deficient diet (CFT) and Atorvastatin/ Lipitor or control diet and control DMSO diluent. (E) Kaplan-Meier survival curve of severe sarcoidosis mice (weeks old) after 28 days of treatment with cholesterol-deficient diet (CFT) and Atorvastatin/ Lipitor (n=9) and or control diet and control DMSO diluent (n=9). (F) Number of BODIPY 493/503 positive F4/80⁺ Mac-2⁺ macrophages in lungs of sarcoidosis mice treated with cholesterol-deficient diet (CFT) and Atorvastatin/ Lipitor treated mice or control diet, and control DMSO diluent. (G) Photomicrographs of Mac-2-stained pulmonary granulomas of sarcoidosis mice treated with cholesterol-deficient diet (CFT) and Atorvastatin/ Lipitor treated mice or control diet, control DMSO diluent. (H) Number of Mac-2⁺ granulomas/ mm² of lung in lungs of sarcoidosis mice treated with cholesterol-deficient diet (CFT) and Atorvastatin/ Lipitor treated mice or control diet, control DMSO diluent.