Purinergic adipocyte-macrophage crosstalk promotes degeneration of thermogenic brown adipose tissue

Figures

• 

• 

• 

  Figure 1. Inflammatory BAT remodeling is triggered by imbalanced BAT activation. (A–E) Male Ucp1−/− mice and wild-type (WT) littermates were housed at 22 °C. (A) Representative BAT images of HE and Sirius Red stainings as well as immunostaining of MAC2. Scale bar, 50 µm. (B) Ucp1 expression in BAT of denervated (DNV) and sham-operated (sham) mice (n = 4). (*P = <0.0001, #P = <0.0001, *P = 0.0055; left to right). (C) Western Blot of BAT of DNV and sham mice. AKT served as loading control (n = 4). (D) Norepinephrine levels in BAT of DNV and sham Ucp1-/- and WT mice (n = 4). (*P = 0.0212, #P = 0.0303, #P = 0.0006; left to right). (E) Gene expression of inflammatory markers in BAT of DNV and sham mice (n = 4). (Tnf: *P = <0.0001, *P = 0.0007, *P = 0.0255; #P = 0.0012; Ccl2: *P = 0.0006, #P = 0.0029; Il1b: *P = 0.0369; Emr1: *P = <0.0001, #P = 0.0020; Cd4: *P = 0.0004, #P = 0.0015; Timp1: *P = 0.0071, #P = 0.0049; Col1a1: #P = 0.0445; left to right). (F–I) Wild-type mice housed at 22 °C were injected daily for three subsequent days with either vehicle (mock/mock), with the β3-adrenergic agonist CL316,243 alone (mock/CL), with the inhibitor of β-oxidation etomoxir alone (Eto/mock) or with the…

• 

  Figure EV1. Related to Fig. 1. Inflammatory BAT remodeling is triggered by imbalanced BAT activation. (A–G) Male Ucp1-/- mice and wild-type (WT) littermates were housed at 22 °C. (A) iBAT weight (n = 5). (*P = 0.0004). (B) BAT gene expression of Ucp1, fibrosis and inflammatory marker genes (n = 5). (Ucp1: *P = 2.0040E-05; Tnf: *P = 0.0054; Ccl2: *P = 0.0275; Il1b: *P = 0.0703; Ifng: *P = 0.0117; Emr1: *P = 0.0044; Cd4: *P = 0.0064; Cd8b1: *P = 0.0055; Timp1: *P = 0.0172; Col1a1: *P = 0.0091; left to right). (C) Body weight (n = 5). (D) BAT images of HE and Sirius Red stainings as well as immunostaining of MAC2. Scale bar, 50 µm. (n = 4). For better comparison, images of all individual mice are presented here, which includes the representative image shown in the main Fig. 1. (E) Quantification of lipid droplet area as shown in (D) (n = 4). (F) Quantification of MAC2 staining as shown in (D) (n = 4). (*P = 0.0038). (G) Quantification of Sirius red staining as shown in (D) (n = 4). (*P = 0.0077). (H) Male Ucp1-/- mice and wild-type (WT) littermates were housed at 22 °C or 6 °C. Western Blot of BAT (n = 5). (I) Quantification of Western blot shown in (H). (UCP1: *P = 0.0032; #P = 0.0016; *P =…

• 

  Figure EV2. Related to Fig. 1. Inflammatory BAT remodeling is triggered by imbalanced BAT activation. Wild-type mice housed at 22 °C were injected daily for three subsequent days with either vehicle (mock/mock), with the beta-3-adrenergic agonist CL316,243 alone (mock/CL), with the inhibitor of beta oxidation etomoxir alone (Eto/mock) or with the combination of both (Eto/CL). (A) Representative images of BAT of individual mice employing HE, Sirius Red and MAC2 (immune)-stainings, respectively. Scale bar, 50 µm. (n = 4). For better comparison, images of all individual mice are presented here, which includes the representative image shown in the main Fig. 1. (B) Quantification of lipid droplet area (n = 4, 3 images per section per mouse). (*P = 0.0073). (C) Quantification of MAC2 staining as shown in (A) (n = 4, 3 images per section per mouse). (*P = <0.0001). (D) Quantification of Sirius red staining as shown in (A) (n = 4, 3 images per section per mouse). (*P = 0.0004). (E) Liver gene expression of inflammatory marker genes (n = 4–5). (*P = 0.0246; *P = 0.0095; left to right). (F) Respiratory quotient (RQ) in mock/CL and Eto/CL mice. The first and second CL injection are indicated by arrows…

• 

  Figure 2. BAT degeneration is preceded by the infiltration of pro-inflammatory myeloid cells. (A–L) Wild-type mice housed at 22 °C were injected for indicated time spans or (M, N) daily for three consecutive days with either vehicle (mock/mock) or with the combination of the β3-adrenergic agonist CL316,243 and the inhibitor of β-oxidation etomoxir (Eto/CL). (A) Gene expression of Ucp1 and inflammatory marker genes in BAT; 4 h after Eto/CL injection (n = 4). (Ccl2: *P = 2.65E-05; Cd8b1: *P = 0.0103; left to right). (B) Gene expression of Ucp1 and inflammatory marker genes in BAT; 8 h after Eto/CL injection (n = 4). (Tnf: *P = 0.0025; Ccl2: *P = 0.000016; Il1b: *P = 0.0005; Emr1: *P = 0.0442; left to right). (C) Gene expression of Ucp1 and inflammatory marker genes in BAT; 24 h after Eto/CL injection (n = 4). (Ucp1: *P = 0.0199; Tnf: *P = 0.0410; Ccl2: *P = 0.0049; Il1b: *P = 0.0245; Emr1: *P = 0.0001; Cd8b1: *P = 0.0332; left to right). (D) Gene expression of Ucp1 and inflammatory marker genes in BAT; 48 h after Eto/CL injection (n = 4). (Tnf: *P = 2.14E-05; Ccl2: *P = 1.73E-06; Il1b: *P = 0.0008; Ifng: *P = 0.0001; Emr1: *P = 1.79E-06; Cd4: *P = 0.0003; Cd8b1: *P = 3.76E-05; left to…

• 

  Figure 3. Brown adipocyte dysfunction causes ATP secretion and is linked to purinergic receptor signaling in BAT. (A–D) Brown adipocytes differentiated from BAT stromal vascular fractions of WT mice were treated for 24 h with or without Eto (25 µM) in the presence or absence of CL (50 nM). (A) CCL2 concentration in cell culture supernatant (n = 9). (B) IL1β concentration in cell culture supernatant (n = 9). (C) Free fatty acid (FFA) concentration in cell culture supernatant (n = 9). (*P = <0.0001, *P = <0.0001; left to right). (D) ATP concentration in cell culture supernatant (n = 9). (*P = <0.0001). (E, F) Brown adipocytes differentiated from BAT stromal vascular fractions of Ucp1−/− mice were treated for 24 h with or without CL (50 nM). (E) FFA concentration in cell culture supernatant (n = 12). (*P = 7.38E-05). (F) ATP concentration in cell culture supernatant (n = 12). (*P = 0.0023). (G, H, J), Wild-type mice housed at 22 °C were injected daily for three subsequent days with either vehicle (mock/mock), with the β3-adrenergic agonist CL316,243 alone (mock/CL), with the inhibitor of β-oxidation etomoxir alone (Eto/mock) or with the combination of both (Eto/CL). (G) Gene expression of…

• 

  Figure EV3. Related to Fig. 3. Brown adipocyte dysfunction causes ATP secretion and is linked to purinergic receptor signaling in BAT. (A) Brown adipocytes differentiated from BAT stromal vascular fractions were treated for 24 h with or without Eto (25 µM) in the presence or absence of CL (50 nM). Gene expression of cytokines (n = 9). (*P = 0.0053; *P = 0.0054; *P = 0.0393, left to right). (B) Analysis of published murine snRNAseq data (Behrens et al, Mol Metab. 2025 Nov;101:102252. doi: 10.1016/j.molmet.2025.102252.). Presentation of P2rx4/P2rx7-double positive nuclei in a dataset of murine single nucleus RNA sequencing. Cell types with a nuclei count above 100 were re-analyzed. (C–F) Human BAT samples from four donors were sorted using low speed centrifugation to enrich mature adipocytes (mA), followed by MACS® to isolate CD31+ endothelial cells and CD11b+ myeloid cells. (C) Gene expression of the macrophage marker EMR1 in the three fractions (n = 4). (D) Gene expression of the endothelial cell marker GPIHBP1 in the three fractions (n = 4). (E) Gene expression of the adipocyte marker ADIPOQ in the three fractions (n = 4). (F) Gene expression of the thermogenic brown adipocyte marker UCP1…

• 

  Figure 4. Combined inhibition of P2X4 and P2X7 prevents pharmacologically induced BAT degeneration thereby preserving thermogenic capacity. (A–F, J–L), Wild-type (WT) and P2rx4−/− mice were pretreated with the P2RX7-inhibiting nanobody or vehicle (control). Then, mice were daily injected with Eto and CL on two consecutive days. (A) Representative BAT images of HE, Sirius Red and MAC2 (immune)-stainings. Scale bar, 50 µm. (B) BAT gene expression of inflammatory marker genes and Ucp1 (n = 6). (Tnf: *P = 0.0006; Ccl2: *P = 0.0003; Il1b: *P = 0.0016; Emr1: *P = 0.0073; Timp1: *P = 0.0004; Col1a1: *P = 0.0053; Ucp1: *P = <0.0001; left to right). (C) Western blot of BAT using γ-tubulin (γ-Tub) as loading control (n = 6). (D) Quantification of Western blot shown in (C). (UCP1: *P = <0.0001; MAC2: *P = <0.0001). (E) Representative FACS analysis of BAT-derived SVF showing vascular and infiltrated CD45+-CD11b+ cells. (F) Quantification of FACS analysis (n = 3, each pool contains BAT of two mice). (*P = 0.0108; *P = 0.0112, left to right). (G–I) Wild-type and P2rx4−/− mice were pretreated with the P2X7-inhibiting nanobody or vehicle (control). Then, the mice received Eto and CL for indicated time…

• 

  Figure EV4. Related to Fig. 4. Combined inhibition of P2X4 and P2X7 prevents pharmacologically induced BAT degeneration thereby preserving thermogenic capacity. (A–C) Wild-type (WT), P2rx4-/- and P2rx7-/- mice were daily injected with Eto and CL on two consecutive days. (A) BAT gene expression of Ucp1, purinergic receptors and inflammatory markers (n = 4–5). (*P = <0.0001). (B) Representative BAT images of HE, Sirius Red and MAC2 (immune)-stainings. Scale bar, 50 µm. (C) Quantification of macrophages and Sirius red staining (n = 4, 3 images per section per mouse). (*P = 0.0284). (D–I) Wild-type (WT) and P2rx4-/- mice were pretreated with the P2X7-inhibiting nanobody or vehicle (control). Then, mice were daily injected with Eto and CL on two consecutive days. (D) BAT images of HE staining of individual mice. Scale bar, 50 µm. For better comparison, images of all individual mice are presented here, which includes the representative image shown in the main Fig. 4. (E) Quantification of lipid droplet area (n = 6, 3 images per section per mouse). (F) BAT images of MAC2 immunestaining of individual mice. Scale bar, 50 µm. For better comparison, images of all individual mice are presented here,…

• 

  Figure 5. Myeloid P2X4 expression determines BAT degeneration. (A–E) Flow cytometry-based validation of the newly generated P2rx4fl/fl-LysMCre mouse model. (A) Gating of macrophages and mast cells in BAT. (B) Flow cytometry of P2X4 in peritoneal macrophages after lavage of P2rx4fl/fl-LysMCre- (red line) and P2rx4fl/fl-LysMCre+ (grey line) mice. (C) Quantification of P2X4+ macrophages of P2rx4fl/fl-LysMCre- and P2rx4fl/fl-LysMCre+ mice shown in (B). (*P = <0.0001). (D) Flow cytometry of P2X4 in peritoneal mast cells after lavage of P2rx4fl/fl-LysMCre- (red line) and P2rx4fl/fl-LysMCre+ (grey line) mice. (E) Quantification of P2X4+ mast cells of P2rx4fl/fl-LysMCre- and P2rx4fl/fl-LysMCre+ mice shown in (D). (F–J) P2rx4fl/fl-LysMCre- and P2rx4fl/fl-LysMCre+ mice were pretreated with the P2X7-inhibiting nanobody. Then, mice were daily injected with Eto and CL on two consecutive days. (F) Representative BAT images of HE, Sirius Red, UCP1 and MAC2 (immune)-stainings. Scale bar, 50 µm. (G) Western blot of BAT using γ-tubulin (γ-Tub) as loading control (n = 5). (H) Quantification of Western blot shown in (G). (UCP1: *P = 6.0905E-05; MAC2: *P = 0.0469). (I) BAT gene expression of inflammatory and…

• 

  Figure EV5. Related to Fig. 5. Myeloid P2X4 expression determines BAT degeneration. (A–G) P2rx4fl/fl-LysMCre- and P2rx4fl/fl-LysMCre+ mice were pretreated with the P2X7-inhibiting nanobody. Then, mice were daily injected with Eto and CL on two consecutive days. (A) BAT images of HE staining of individual mice. Scale bar, 50 µm. For better comparison, images of all individual mice are presented here, which includes the representative image shown in the main Fig. 5. (B) Quantification of lipid droplet area (n = 5, 3 images per section per mouse). (*P = 0.0008). (C) BAT images of MAC2 immunestaining of individual mice. Scale bar, 50 µm. For better comparison, images of all individual mice are presented here, which includes the representative image shown in the main Fig. 5. (D) Quantification of macrophages (n = 5, 3 images per section per mouse). (*P = 0.0239). (E) BAT images of Sirius Red staining of individual mice. Scale bar, 50 µm. For better comparison, images of all individual mice are presented here, which includes the representative image shown in the main Fig. 5. (F) Quantification of Sirius Red staining (n = 5, 3 images per section per mouse). (*P = 0.0358). (G) BAT images of UCP1…