NCLX prevents cell death during adrenergic activation of the brown adipose tissue

Abstract

A sharp increase in mitochondrial Ca²⁺ marks the activation of brown adipose tissue (BAT) thermogenesis, yet the mechanisms preventing Ca²⁺ deleterious effects are poorly understood. Here, we show that adrenergic stimulation of BAT activates a PKA-dependent mitochondrial Ca²⁺ extrusion via the mitochondrial Na⁺/Ca²⁺ exchanger, NCLX. Adrenergic stimulation of NCLX-null brown adipocytes (BA) induces a profound mitochondrial Ca²⁺ overload and impaired uncoupled respiration. Core body temperature, PET imaging of glucose uptake and VO₂ measurements confirm a thermogenic defect in NCLX-null mice. We show that Ca²⁺ overload induced by adrenergic stimulation of NCLX-null BAT, triggers the mitochondrial permeability transition pore (mPTP) opening, leading to a remarkable mitochondrial swelling and cell death. Treatment with mPTP inhibitors rescue mitochondrial function and thermogenesis in NCLX-null BAT, while calcium overload persists. Our findings identify a key pathway through which BA evade apoptosis during adrenergic stimulation of uncoupling. NCLX deletion transforms the adrenergic pathway responsible for thermogenesis activation into a death pathway.

Fig. 1. Mitochondrial Ca²⁺ extrusion in BA is mediated by Na⁺/Ca²⁺ exchange and requires PKA.

a Representative traces of mitochondrial Ca²⁺ kinetics in primary brown adipocytes (BA) after application of NE (1.5 μM). Mitochondrial Ca²⁺ was monitored using the mitochondrial Ca²⁺ dye, Rhode2-AM. The dashed lines represent the linear fit used to calculate the Ca²⁺ influx and efflux rates. b Representative traces of mitochondrial Ca²⁺ transients upon application of NE to BA in the presence or absence of Na⁺ (NMDG⁺ iso-osmotically replacing Na⁺). c Quantification of mitochondrial Ca²⁺ efflux rates in the presence and absence of Na⁺. NMDG⁺ was used as a cationic replacement of Na⁺ (n = 4 independent experiments per condition). d Representative traces of mitochondrial Ca²⁺ transients upon application of NE (1.5 μM), ATP (100 μM), or ATP in cells pretreated with PKA activator, Forskolin (FSK (50 μM, 15 min)). e, f Quantification of mitochondrial Ca²⁺ fluxes in response to NE, ATP, or ATP+FSK stimulation. NE (n = 12), ATP (n = 4), ATP+FSK (n = 6). g Representative traces of mitochondrial Ca²⁺ transients in primary BA upon stimulation by NE in the presence and absence of PKA inhibitor, H-89 (5 μM, 1 h pre-incubation). h, i Quantification of mitochondrial Ca²⁺ traces under treatment of H-89 (n = 10 and 14 for control and H-89 in h, n = 8 and 13 for control and H-89 in i). Student’s t-test (c, h, i); one-way ANOVA with Tukey’s post hoc test (e, f). Data are expressed as means ± SEM. ^nsp > 0.05, *p < 0.05, **p < 0.001. Replicates are indicated by individual dots shown for each group in all graphs. Source data are available as a Source Data file.

Fig. 2. NCLX is required for mitochondrial uncoupled respiration in primary BA.

a Representative traces of mitochondrial Ca²⁺ fluxes in primary BA transfected either with siNCLX or siControl and monitored for mitochondrial Ca²⁺ transients evoked by NE. b, c Quantification of influx and efflux rates in primary BA transfected either with siNCLX or with siControl (n = 5 independent experiments per condition). d Western blot analysis of NCLX levels from NCLX KO and WT BAT. Stain-Free Gels (S.F. Gels) were used as a normalization method for total loaded protein. e, f Western blot analysis of MCU and quantification of its relative levels (n = 11 per group). g Representative mitochondrial Ca²⁺ transients of primary BA from NCLX KO and WT mice induced by NE. h, i Quantification of Ca²⁺ influx and efflux rates in NCLX KO and WT BA. Mitochondrial Ca²⁺ influx rates were unaffected while Ca²⁺ efflux rates were diminished in NCLX KO BA (n = 4 for WT and n = 11 for NCLX KO). j Representative traces of oxygen consumption rates (OCR) of cultured BA from NCLX KO and WT mice. Stimulation by NE was the first injection. OLIGOmycin was used to assess mitochondrial uncoupling efficiency. FCCP was then injected to assess maximal respiration and non-mitochondrial OCR was evaluated by Antimycin A injection (AA). k, l Quantification of NE response and maximal respiration after NE stimulation (N = 3 independent experiments with n = 18 for WT and n = 22 for NCLX KO). Student’s t-test (b, c, f, h, i, k, l); data are expressed as means ± SEM. ^nsp > 0.05, *p < 0.05, **p < 0.001, ***p < 0.0001. Replicates are indicated by individual dots shown for each group in all graphs. Source data are available as a Source Data file.

Fig. 3. Thermogenesis is impaired in NCLX-null mice.

a Survival curves of 8–10-week-old NCLX KO and WT male mice cold stressed at 4 °C. Animals reaching temperatures 28 °C and below were returned to room temperature for recovery and counted as drop-outs. Survival rates depict the rate at which animals had to be removed from 4 °C to room temperature to avoid mortality (n = 4 mice for WT and n = 6 mice for NCLX KO). CBT, core body temperature. b Core body temperature traces of the animals from experiment (a) during cold exposure (4 °C). The dashed red line indicates animal removal. c VO₂ of NCLX KO and WT mice subjected to cold (4 °C) at t = 0 and t = 7 h (n = 5 mice per group). d–f Non‐shivering thermogenesis of NCLX KO and WT mice assessed by measuring O₂ consumption at baseline and followed by β-3 adrenergic stimulation (CL-316,243 injection (1 mg/kg)) in anesthetized mice at 30 °C. d Traces of VO₂ of NCLX KO and WT mice. e VO₂ at baseline and under CL stimulation. f Fold increase of O₂ consumption after the CL stimulation (n = 5 mice for WT and n = 6 mice for NCLX KO). g Survival curves of WT and NCLX heterozygous (+/−) littermates cold stressed at 4 °C; mice reaching 28 °C or lower were returned to room temperature for recovery and counted as drop-outs (n = 4 mice for WT and n = 9 mice for NCLX+/−). h Core body temperature of the animals shown in g. i VO₂ of WT and NCLX+/− mice at basal and after CL-316,243 injection (1 mg/kg), under anesthesia at 30 °C. Traces of VO₂ of NCLX+/− and WT mice. j VO₂ at baseline and under CL stimulation (n = 4 mice per group). Student’s t-test (c, e, f, j); two-way ANOVA (d, i). Data are expressed as means ± SEM. ^nsp > 0.05, *p < 0.05, **p < 0.001, ***p < 0.0001. Replicates are indicated by individual dots shown for each group in all graphs. Source data are available as a Source Data file.

Fig. 4. Respiratory dysfunction mediated by NCLX loss in BA is acutely induced following adrenergic stimulation.

a, b Western blot analysis of NCLX-null and WT BA for UCP1 expression levels S.F. Gel was used as a normalization reference for total protein loading (n = 11 per group). c, d Western blot analysis of NCLX-null and WT BA for TOM20 expression levels (n = 7 per genotype). e Mitotracker Green (MTG) fluorescence intensity in NCLX-null and WT primary BA, measured by Operetta high-content imaging system (n = 16 per condition from total N = 3 imaging experiments). f–j Membrane potential and morphometric analyses of mitochondrial morphology in NCLX-null and WT BA (N = 4 independent experiments and n = at least 41 cells per condition). f Representative images of NCLX-null and WT BA stained with TMRE and MTG. Scale bar, 10 μm. g Quantification of mitochondrial membrane potential of NCLX-null and WT BA. TMRE signal is normalized to MTG signal to generate a ratio that avoids focal plane artefacts. h Quantification of mitochondrial aspect ratio (AR), i circularity and j roundness in NCLX-null and WT BA using the MTG channel. k, l Western blot analysis of primary NCLX-null and WT BA for OXPHOS complex subunits II–V (CII–CV). Cells were treated with or without NE for 6 h (n = 3–7 per group). m Western blot of super complex (SC) I+III in BAT homogenates from cold-exposed and non-exposed NCLX-null and WT mice. n Quantification of the fraction of complex III that is assembled into SC(I+III) relative to total complex III from cold-exposed and non-exposed NCLX-null and WT mice (N = 3–4 mice per condition). o Assessment of mitochondrial complexes activity in cold-exposed and non-exposed NCLX-null and WT BAT: CI (n = 11–18), CII (n = 9–20), and CIV (n = 5–8). p Quantification of basal respiration in intact non-stimulated BA (N = 6–8 independent experiments with n = 32 for WT and n = 21 for NCLX KO). q Quantification of maximal spare respiratory capacity in NE-stimulated and non-stimulated NCLX-null and WT BA. Maximal respiration was induced by FCCP (N = 3–8 independent experiments with n = 18–32 per condition). Student’s t-test (b, d, e, g, h, i, j, p); one-way ANOVA with Tukey’s post hoc test (l, n, o, q). Data are expressed as means ± SEM. ^nsp > 0.05, **p < 0.001. Replicates are indicated by individual dots shown for each group in all graphs. Source data are available as a Source Data file.

Fig. 5. Adrenergic stimulation in NCLX-null BA leads to mitochondrial swelling and cytochrome c release.

a–c Super-resolution confocal images of fixed WT or NCLX-null BA, with or without adrenergic stimulation for 6 h. a NCLX KO and WT BA treated with or without NE and immunostained for cytochrome c (red). Samples were co-stained for TOM20 to label mitochondrial localization. Scale bar is 10 μm, for the zoom images scale bar is 2 μm. b Quantification of cytochrome c release. Cytochrome c in the cytosol was calculated by measuring its total integrated fluorescence intensity (FI) outside the mitochondria per cell, normalized to total cellular area (n = 27–46 cells for each of the four conditions). c Quantification of mitochondrial swelling. Zero indicates bars where quantification yielded zero events of swelling (n = 28–37 cells for each of the four conditions). d Time-lapse of swelling formation following NE stimulation in NCLX-null and WT BA. Cells were labeled with mCherry-Fis1 probe marking their outer mitochondrial membrane. Swelling spatially starts at one pole of the cell and gradually propagated to the center of the cell. Scale bar, 10 and 2 μm for the zoom images. e–h Electron microscopy of NE-stimulated NCLX-null and WT BA. e Representative images of NE-stimulated NCLX-null and WT mitochondria of primary BA. Note that while the WT mitochondria are intact, adrenergic stimulation leads to cristae disruption, mitochondrial swelling, and rupture in the NCLX KO mitochondria. Scale bar, 1 and 0.1 μm for the zoom images. f Quantification of cristae density. g Circumference and h mitochondrial swelling assessed by cross-sectional area of individual mitochondria (n = 185 mitochondria at least were analyzed per condition). Student’s t-test (f, g, h); one-way ANOVA with Tukey’s post hoc test (b, c). Data are expressed as means ± SEM. ^nsp > 0.05, *p < 0.05, **p < 0.01, ***p < 0.0001. Replicates are indicated by individual dots shown for each group in all graphs. Source data are available as a Source Data file.

Fig. 6. Adrenergic stimulation in NCLX-null BA leads to Ca²⁺-mediated cell death in vitro and in vivo.

a–c Ca²⁺ chelation and inhibition of Ca²⁺ uptake prevent mitochondrial swelling in NCLX KO BA. a A scheme depicting points of pharmacological intervention targeting Ca²⁺ entry and accumulation. EGTA (3 mM) was used to buffer extracellular Ca²⁺, BAPTA-AM (25 μM) was used to chelate intracellular Ca²⁺, and Ru360 (10 μM) was used to inhibit MCU and Ca²⁺ entry to the mitochondria. b Quantification of mitochondrial swelling in NE-stimulated NCLX KO and WT BA treated with Ca²⁺ chelators or Ru360 assessed by super-resolution imaging of TOM20 staining for mitochondrial outer membrane (n = 18–33). c Representative images of NE-stimulated NCLX KO and WT BA treated with Ca²⁺ chelators or Ru360. Note that chelating Ca²⁺ or blocking mitochondrial Ca²⁺ entry was sufficient to inhibit mitochondrial swelling. Scale bar, 10 and 1 μm for the zoom images. Swollen mitochondria are indicated by the white arrows. d–f Cell death measured by TUNEL staining. Mice were submitted to intermittent cold-stress protocol for 5 days. BAT was harvested, fixed, and stained with TUNEL. Hematoxylin was used as a counterstaining for all nuclei (n = 3–6 mice per genotype). d TUNEL staining of BAT lobes from cold-stressed NCLX KO and WT mice. TUNEL-positive cells (brown) are indicated with red arrows. Scale bar, 500 and 50 μm for the zoom images. e Quantification of the average number of positive cells per cross-sectional area in each genotype. f Quantification of the TUNEL-positive cells as a percentage of total cells. Student’s t-test (e, f); one-way ANOVA with Tukey’s post hoc test (b). Data are expressed as means ± SEM. ^nsp > 0.05, *p < 0.05, **p < 0.01, ***p < 0.0001. Replicates are indicated by individual dots shown for each group in all graphs. Source data are available as a Source Data file.

Fig. 7. Inhibition of mitochondrial permeability transition in NCLX-null mice restores thermogenic function.

a, b OCR of NCLX-null and WT primary BA that were pretreated either with the mPTP inhibitor, NIM811, or with vehicle control, DMSO. a Representative OCR traces in BA from NCLX KO and WT pretreated with NIM811 (500 nM) or vehicle control. b Quantification of NE response in NCLX KO and WT BA pretreated with NIM811 or vehicle control (n = 28–36 per condition from N = 8–9 independent experiments; See Supplementary Fig. 10 for WT treatment traces). c Representative mitochondrial Ca²⁺ transients of NCLX KO and WT BA pretreated with NIM811 or vehicle control and stimulated by NE. d Quantification of Ca²⁺ amplitude, influx, and efflux rates (n = 4–11 independent experiments per condition). e Representative images of cells stained for TOM20 to label mitochondrial localization and assess mitochondrial swelling. Scale bar, 10 and 1 μm for the zoom images. f Quantifications of mitochondrial swelling in NE activated WT and NCLX KO BA pretreated with NIM811 or vehicle control (n = 68–118 cells). g Cell death in WT and NCLX KO BA pretreated with NIM811 or vehicle control after 72 h of NE stimulation. Cell death was measured in vitro by TUNEL assay (n = 77–83 per condition from N = 5–6 independent experiments). h Representative images of WT and NCLX KO BA after 72 h of NE stimulation. Cells were pretreated with NIM811 or vehicle control. TOM20 was used to label mitochondrial localization while DAPI was used to label the nuclei. Scale bar, 10 μm. i Schematic graph for in vivo NIM811 experimental procedure. Mice were pretreated for 5 days with a 50 mg/kg dose NIM811 or a vehicle introduced subcutaneously. On the day of the experiment, mice were acutely cold stressed (6–8 h) and then immediately PET-imaged with the glucose analog ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) to measure BAT activity under 4 °C. j Survival curves of 8–10-week-old male NCLX KO and WT mice pretreated with or without NIM811 and cold stressed at 4 °C. Mice reaching 28 °C or lower were returned to room temperature for recovery (n = 5–10 mice per group). k Representative PET-CT images of ¹⁸F-FDG uptake in cold-stressed mice. Each image is a composition of a CT image superimposed on a PET image. In the CT image, radio-opacity is reflected in grayscale. In the PET image, ¹⁸F-FDG intensity is reflected in a hot-metal scale. BAT is marked with blue arrows. ¹⁸F-FDG uptake is presented as percent injected dose per gram (%ID/g). l Quantification of ¹⁸F-FDG uptake in BAT and normalized to liver uptake as a reference tissue (n = 3–11 mice per group). One-way ANOVA with Tukey’s post hoc test (b, d, f, g, l); log-rank (j). Data are expressed as means ± SEM. ^nsp > 0.05, *p < 0.05, **p < 0.01, ***p < 0.0001. Replicates are indicated by individual dots shown for each group in all graphs. Source data are available as a Source Data file.

Fig. 8. Mechanism by which brown adipocytes avoid cell death during thermogenesis.

Adrenergic stimulation of brown adipocytes leads to intracellular increase in Ca²⁺ which can induce mitochondrial Ca²⁺ overload and permeability transition. While in other cell types, such scenario will progress to apoptosis, in BA, adrenergic stimulation leads to activation of Ca²⁺ extrusion through NCLX, thereby preventing permeability transition and cell death. Inhibition of permeability transition using the Cyclophilin D inhibitor, NIM811, can prevent cell death and restore function even in the absence of NCLX, and while Ca²⁺ overload persists. _iCa²⁺, intracellular ${\mathrm{Ca}}^{\mathrm{2+}}$; _m${\mathrm{Ca}}^{2+}$, mitochondrial ${\mathrm{Ca}}^{2+}$.