Targeting cannabinoid receptor 1 for antagonism in pro-fibrotic alveolar macrophages mitigates pulmonary fibrosis

Abstract

Pulmonary fibrosis (PF) is a life-threatening disease that requires effective and well-tolerated therapeutic modalities. Previously, the distinct pathogenic roles of cannabinoid receptor 1 (CB1R) and inducible nitric oxide synthase (iNOS) in the lungs and their joint therapeutic targeting were highlighted in PF. However, the cell-specific role of CB1R in PF has not been explored. Here, we demonstrate that CB1R in alveolar macrophages (AMs) mediates the release of anandamide into the alveoli, which promotes PF by inducing pro-fibrotic macrophages that are accessible to locally delivered antifibrotic therapy. A multitargeted therapy may improve therapeutic efficacy in PF. Pulmonary delivery of 0.5 mg/kg/d MRI-1867 (zevaquenabant), a peripherally acting hybrid CB1R/iNOS inhibitor, was as effective as systemic delivery of 10 mg/kg/d and also matched the efficacy of nintedanib in mitigating bleomycin-induced PF. A systems pharmacology approach revealed that zevaquenabant and nintedanib treatments reversed pathologic changes in both distinct and shared PF-related pathways, which are conserved in human and mouse. Moreover, zevaquenabant treatment also attenuated fibrosis and pro-fibrotic mediators in human precision-cut lung slices. These findings establish CB1R-expressing AMs as a therapeutic target and support local delivery of dual CB1R/iNOS inhibitor zevaquenabant by inhalation as an effective, well-tolerated, and safe strategy for PF.

Keywords: Drug therapy; Fibrosis; Immunology; Inflammation; Macrophages; Pulmonology.

Figure 1. Deletion of CB₁R in myeloid cells, but not in AT2 cells, prevents PF.

(A) Change in percentage in body weight after bleomycin administration in different groups showing prevention of body weight loss in CB₁R-KO and myeloid CB₁R-KO mice but not in AT2 CB₁R-KO mice in 28-day bleomycin-induced PF model (n = 7 per group). O.P., oropharyngeal. (B) Significant reduction of bleomycin-induced hydroxyproline content found in the CB₁R-KO and myeloid CB₁R-KO mice but not in AT2 CB₁R-KO mice at 28 days postbleomycin (2-way ANOVA, ****P < 0.0001, ***P < 0.001, **P < 0.01, n = 6–7 per group). (C) Representative images of Masson’s trichrome–stained lung sections showing reduced collagen deposition and better architecture in CB₁R-KO and myeloid CB₁R-KO mice but not in AT2 CB₁R-KO mice. (D–F) Retention of various pulmonary functions (FEV_0.1, FVC, and Crs) at the normal level was found in CB₁R-KO and myeloid CB₁R-KO mice but not in AT2 CB₁R-KO mice after being challenged with bleomycin (2-way ANOVA, ****P < 0.0001, ***P < 0.001, n = 6–7 per group). (G and H) Levels of AEA in BALF and lung were reduced in bleomycin-challenged CB₁R-KO and myeloid CB₁R-KO mice but not in AT2 CB₁R-KO mice (2-way ANOVA, ***P < 0.001, **P < 0.01, *P < 0.05, n = 3–10 per group).

Figure 2. Deletion of CB₁R in myeloid cells attenuates pro-fibrotic macrophages and microenvironment.

(A) Bleomycin-induced PF model. (B) A significant increase in the CB₁R protein expression in the lung immune cells was observed in the fibrotic mice (1-way ANOVA, *P < 0.05, n = 6 per group). (C) During lung fibrosis, CB₁R expression increased only in the macrophages among all myeloid cells in the lungs (2-way ANOVA, ***P < 0.001, **P < 0.01, n = 6 per group). (D) Among different subsets of macrophages, CB₁R expression only increased in tissue-resident alveolar macrophages (Tr-AMs) and monocyte-derived alveolar macrophages (Mo-AMs), indicating the significance of CB₁R in the fibrotic alveolar microenvironment (2-way ANOVA, ****P < 0.0001, **P < 0.01, n = 6 per group). (E) CB₁R-KO and myeloid CB₁R-KO mice were challenged with bleomycin-induced PF model. (F) A reduction in the total macrophage population was found in both CB₁R-KO and myeloid CB₁R-KO mice (1-way ANOVA, ****P < 0.0001, n = 9–13 per group). (G–I) Phenotypic alterations in different subpopulations of macrophages: Mo-AMs, Tr-AMs, and IMs. Infiltration of Mo-AMs was found in the fibrotic lungs, which was reduced in both CB₁R-KO and myeloid CB₁R-KO mice (1-way ANOVA, ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05, n = 9–13 per group). (J–M) The total CD206⁺ macrophages and subpopulations of CD206⁺ macrophages, Tr-AMs, Mo-AMs, and IMs, in different groups. CB₁R deletion significantly attenuated the total CD206⁺ macrophages as well as CD206⁺ MO-AMs and IMs (1-way ANOVA, ****P < 0.0001, ***P < 0.001, n = 9–13 per group). (N) A multiplex Luminex assay was used to measure secreted cytokines in BALF. (1-way ANOVA, *P < 0.05 indicates significant difference compared with control. ^#P < 0.05 indicates a significant difference compared with WT bleomycin group. n = 4–11 per group.)

Figure 3. CB₁R activation in AMs could regulate the activation of pro-fibrotic macrophages.

(A–D) Primary alveolar macrophages (pAMs) were isolated from healthy mice (control) and fibrotic mice 14 days after exposure to saline and bleomycin, respectively. Gene expression of Cnr1, Tgfβ, and Il-6 in pAMs ex vivo at 24 hours in the absence or presence of 1 μM of rimonabant (1-way ANOVA, **P < 0.01, *P < 0.05, n = 4 per group). (E–H) Gene expression of Cnr1, Tgfβ, and Il-6 in MH-S cell line (mouse alveolar macrophage cell line) at 48 and 72 hours after exposure to bleomycin (1 mU/mL) in the absence or presence of 1 μM of rimonabant. Control group was treated with saline as a vehicle for bleomycin. (1-way ANOVA, ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05, n = 6 per group.) (I–K) Cell surface protein expression of CB₁R and CD206 and coexpression in MH-S cell line at 48 and 72 hours after exposure to bleomycin (1 mU/mL) in the absence or presence of 1 μM of rimonabant determined by flow cytometry analysis. Control group was treated with saline as a vehicle of bleomycin. (1-way ANOVA, ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05, n = 6 per group.)

Figure 4. Spatial transcriptomics analysis of the human IPF lungs reveals the importance of targeting CB₁R (CNR1) and iNOS (NOS2).

(A) Comparative expression of pro-fibrotic macrophage population denoted by FABP4⁺ and SPP1⁺ macrophages between healthy human and IPF human lungs (25). (B) Expression of CNR1 was increased in both FABP4⁺ and SPP1⁺ macrophages in the human IPF lungs. (C) Expression pattern of fibrogenic and/or scar-associated macrophage markers in both FABP4⁺ and SPP1⁺ macrophages in the human IPF lungs reflecting their association with CNR1 expression. (D) Spatial plots visualize the niche annotation in human IPF lung samples. Spatial patterns of FABP4, SPP1, CNR1, and NOS2 in human IPF lung samples. Airway: basal, ciliated; Alveolar: AT0, AT1, AT2; Fibroblast: alveolar, peribronchial, pericyte; Fibrotic: aberrant basaloid, myofibroblast, IL-1B⁺ macrophages; Immune: dendritic, B/plasma, T cell; SMCs_Adv_Meso: smooth muscle, adventitial fibroblast, mesothelial cells. (E–G) Correlation analysis between the expression of CNR1 and key pro-fibrotic macrophage markers, FABP4, SPP1, and MRC1, respectively, in the IPF lungs. (H) Comparative expression of NOS2 in different lung population niche between healthy human and IPF human lungs.

Figure 5. Pulmonary delivery of MRI-1867 at a reduced dose reveals antifibrotic efficacy.

(A and B) Pharmacokinetic profile and area under the curve of MRI-1867 at 0.5 mg/kg b.w. O.P. and 10 mg/kg b.w. I.P. doses showing similar concentrations in the lungs over 24 hours (unpaired t test, P > 0.05, n = 3 per group). (C and D) Comparative exposure of MRI-1867 in the serum and brain between O.P. (0.5 mg/kg b.w.) and I.P. (10 mg/kg b.w.) doses, indicating reduced exposure via pulmonary delivery (n = 3 per group). (E) Upper gastrointestinal motility assay showing in vivo systemic peripheral CB₁R antagonism potency of MRI-1867 (1-way ANOVA, ****P < 0.0001, **P < 0.01, n = 5–8 per group). (F) Improvement in body weight loss was found after treatment with MRI-1867 (n = 8–10 per group). (G) Attenuation of hydroxyproline content by MRI-1867 treatment showing antifibrotic efficacy (1-way ANOVA, ****P < 0.0001, **P < 0.01, n = 8–22 per group). (H) Masson’s trichrome staining of the mice lung showed a reduction in the collagen deposition and alveolar space constriction by the treatment with MRI-1867. (I and J) MRI-1867 via both O.P. and I.P. dosage showed comparable improvement in pulmonary functions as depicted by FEV and FVC (1-way ANOVA, ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05, n = 4–14 per group). (K and L) MRI-1867 at both dosages reduced anandamide levels in the BALF and lungs (1-way ANOVA, ****P < 0.0001, **P < 0.01, *P < 0.05, n = 4–7 per group). (M and N) MRI-1867 at 0.5 mg/kg b.w. O.P. dose reduced the gene expression of Cnr1 and Nos2 at a similar degree as compared with the I.P. dose of 10 mg/kg b.w. (1-way ANOVA, ***P < 0.001, **P < 0.01, *P < 0.05, n = 4–14 per group.) GI, gastrointestinal.

Figure 6. Modulation of the fibrotic microenvironment by MRI-1867.

(A) Immunophenotyping by flow cytometry revealed a reduction in the total macrophage population by both dosages of MRI-1867. (B–D) Phenotypic alterations in different subpopulations of macrophages: tissue-resident alveolar macrophages (Tr-AMs), monocyte-derived alveolar macrophages (Mo-AMs), and interstitial macrophages (IMs). Infiltration of Mo-AMs was found in the fibrotic lungs and MRI-1867 treatment reduced this population (1-way ANOVA, ****P < 0.0001, ***P < 0.001, **P < 0.01, n = 5–7 per group). (E) MRI-1867 predominantly targets the CD206⁺ macrophages by reducing the population increase due to fibrosis (1-way ANOVA, ****P < 0.0001, ***P < 0.001, n = 5–7 per group). (F–H) The subpopulations of CD206⁺ macrophages, Tr-AMs, Mo-AMs, and IMs, in different groups. MRI-1867 significantly attenuated the CD206⁺ MO-AMs and IMs (1-way ANOVA, ****P < 0.0001, ***P < 0.001, **P < 0.01, n = 5–7 per group). (I) MRI-1867 ameliorates the fibrotic microenvironment by altering cytokines and chemokines in the BALF from mice. Heatmap showing modifications in the stimulatory, pro-fibrotic, and chemotactic cytokines by MRI-1867 (independent t test, *P < 0.05 vs. Ctrl, ^#P < 0.05 vs. bleomycin-O.P. Veh, n = 4–10 per group). (J) Attenuation of fibrogenic macrophage signature by pulmonary delivery of MRI-1867 in the lungs. Heatmap is generated from the z-score–normalized expression (mRNA) of the macrophages isolated from the lungs of different groups (independent t test, P < 0.05, n = 3–5 per group). (K) Inhalational MRI-1867 treatment via the O.P route at 0.5 mg/kg reversed the majority of bleomycin-induced fibroproliferative pathways in the fibrotic lungs. MRI-1867 treatment regulates the pathways involved in different phases of fibrosis, thereby exerting its antifibrotic efficacy (independent t test, P < 0.05, n = 3 per group).

Figure 7. Dual inhibition of CB₁R and iNOS by MRI-1867 provides superior antifibrotic efficacy in aged mice.

(A) Improvement in body weight loss was found after treatment with MRI-1867 via the O.P. route in 52-week-old (aged) mice. All the treatments were started on day 10, then continued up to day 28, and mice were euthanized on day 28 after bleomycin (1 U/kg b.w., O.P., n = 8–16 per group). (B) Bleomycin-induced elevated hydroxyproline levels were only significantly reduced by dual targeting of CB₁R and iNOS with MRI-1867 and nintedanib in the 28-day bleomycin-induced PF model in 52-week-old mice (1-way ANOVA, ****P < 0.0001, **P < 0.01, n = 8–13 per group). (C) Masson’s trichrome staining of the mouse lungs showed a reduction in the collagen deposition and alveolar space constriction by the treatment with MRI-1867, rimonabant, 1400W, and nintedanib but better architecture found in MRI-1867 treatment compared with rimonabant, 1400W, and nintedanib. (D–F) Pulmonary delivery of MRI-1867 showed significant improvement in pulmonary functions as depicted by FEV_0.1, FVC, and Crs (1-way ANOVA, ****P < 0.0001, ***P < 0.001, **P < 0.01, n = 8–14 per group). Nintedanib could not improve the pulmonary functions in these mice.

Figure 8. Transcriptomics analysis reveals the shared and unique effects of MRI-1867 and nintedanib treatments on attenuating multiple fibrosis-related genes and pathways in bleomycin-induced PF.

(A) Venn diagram demonstrates numbers of DEGs that were reversed by either MRI-1867 (0.5 mg/kg, O.P.) or nintedanib (60 mg/kg, P.O.) treatment compared with the vehicle-treated group in bleomycin-induced PF in mice. (B) Significantly reversed fibrosis-related pathways (FDR < 0.05) based on 2,208 DEGs by either MRI-1867 or nintedanib treatment. (C) Significantly reversed fibrosis-related pathways (FDR < 0.05) based on 828 DEGs by only MRI-1867 but not nintedanib treatment. (D) Significantly reversed fibrosis-related pathways (FDR < 0.05) based on 1,003 DEGs by only nintedanib but not MRI-1867 treatment.

Figure 9. Translationally relevant genes and pathways reversed by either MRI-1867 or nintedanib treatments.

(A) Venn diagram demonstrates numbers of fibrosis-related DEGs in lung transcriptome that were reversed by either MRI-1867 (0.5 mg/kg, O.P.) or nintedanib (60 mg/kg, P.O.) treatment compared with the vehicle-treated group in bleomycin-induced PF in mice and those shared in IPF patients’ lung transcriptome (NCBI GEO GSE92592, 39 samples) (65). (B) Significantly reversed fibrosis-related pathways shared in human and mouse lung transcriptome (FDR < 0.05) based on 709 DEGs by either MRI-1867 or nintedanib treatment. (C) Circos plot exhibiting the uniquely reversed effects of bleomycin by pulmonary delivery of MRI-1867 treatment but not by nintedanib treatment compared with human late-stage IPF patients’ lung transcriptomics data (GSE92592, 39 samples) (65) and the reversed PF-dependent biological pathways. (DESeq2, FDR < 0.05.)

Figure 10. MRI-1867 attenuates fibrosis in hPCLS.

(A) Experimental plan with hPCLS, an ex vivo model exhibiting the characteristics of IPF. Schematic of treatment with a control cocktail or fibrotic cocktail containing TGF-β. Vehicle (DMSO) and MRI-1867 (10 μM) were treated 2 days after the fibrosis induction. TGF-β, transforming growth factor-β; TNF-α, tumor necrosis factor-α; PDGF-AB, platelet-derived growth factor-AB; LPA, lysophosphatidic acid. (B) Reduction of collagen deposition by MRI-1867 as measured by hydroxyproline content in hPCLS (1-way ANOVA, *P < 0.05, n = 6 per group). (C) Masson’s trichrome staining shows characteristics of alveolar septal thickening and collagen fiber formation, which was markedly reduced by MRI-1867 treatment. (D–G) MRI-1867 treatment significantly reduced the gene expression of COL1A1, FN1, ACTA2, and IRF5 (1-way ANOVA, ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05, n = 4 per group). (H) Regulation of pro-fibrotic cytokines by MRI-1867 in fibrotic hPCLS culture supernatant (independent t test, FDR < 0.1. n = 6 per group).