TRPC6 governs brown adipose thermogenesis via a BMPR2-p38 MAPK signaling axis

Abstract

Brown adipose tissue (BAT) thermogenesis combats obesity, but mechanisms linking calcium dynamics to thermogenic programming remain incompletely defined. Here, we identify the calcium channel TRPC6 as an essential BAT-intrinsic regulator of metabolic health. BAT-specific Trpc6 knockout (Trpc6^BTKO) mice exhibit spontaneous BAT whitening, mitochondrial dysfunction, and impaired cold tolerance. Upon high-fat diet (HFD) challenge, Trpc6^BTKO mice develop exacerbated obesity, hepatic steatosis, and insulin resistance. These phenotypes are driven by increased energy intake and reduced energy expenditure associated with impaired thermogenesis. TRPC6 deficiency suppresses mitochondrial biogenesis and thermogenesis. Mechanistically, TRPC6 mediates calcium influx and interacts directly with BMPR2, thereby selectively activating p38 MAPK signaling to drive thermogenic gene expression. Genetic disruption of the TRPC6-BMPR2 complex abolishes TRPC6-mediated thermogenesis. Thus, we define a non-redundant TRPC6-BMPR2-p38 MAPK signaling axis whose disruption underpins obesity and associated metabolic dysfunction, positioning it as a promising therapeutic target for metabolic disease.

Graphical abstract

Highlights

• •TRPC6 is a critical regulator of thermogenesis in brown adipose tissue.
• •Ablation of TRPC6 impairs thermogenesis and energy balance, resulting in cold intolerance.
• •BAT-specific TRPC6 deficiency exacerbates diet-induced obesity and systemic metabolic dysfunction.
• •TRPC6 regulates thermogenesis through BMPR2-dependent p38 MAPK activation.

1. Introduction

Obesity and its associated metabolic disorders represent a global health crisis [[1], [2], [3]], underscoring the critical need to define molecular regulators of systemic energy homeostasis. BAT is a key thermogenic organ, dissipating chemical energy as heat to counteract obesity and metabolic diseases [[4], [5], [6], [7], [8], [9]]. While the core thermogenic effectors like PGC-1α and UCP1 are well-established [[10], [11], [12], [13], [14], [15]], the upstream mechanisms integrating intracellular calcium (Ca²⁺) signals into the thermogenic program remain incompletely understood.

Calcium signaling is a fundamental regulator of metabolic processes, modulating mitochondrial function, lipid mobilization, and thermogenic gene expression [[16], [17], [18]]. The transient receptor potential (TRP) channels, which sense diverse stimuli (e.g., cold, thermal, mechanical, and chemical), are major mediators of calcium influx [19,20]. While specific TRP isoforms like TRPV2 and TRPM8 are known to influence BAT differentiation and UCP1-dependent thermogenesis [21,22], the role of TRPC6 in BAT thermogenesis and metabolic control remains unexplored.

TRPC6 is a widely expressed Ca²⁺-permeable channel implicated in cardiovascular, neuronal, immune, and renal physiology [[23], [24], [25], [26]]. We therefore hypothesized that TRPC6 acts as a BAT-autonomous regulator, non-redundantly controlling thermogenesis through a specific Ca²⁺-dependent signaling pathway, and that its impairment predisposes to metabolic disease. To test this, we employed BAT-specific Trpc6 knockout mice alongside complementary in vitro and mechanistic approaches. We demonstrate that TRPC6 ablation reduces thermogenic capacity and cold adaptation through the Ca²⁺-dependent BMPR2-p38 MAPK pathway. This study establishes the role of TRPC6 in BAT thermogenesis and elucidates the underlying mechanisms by which its disruption promotes obesity.

2. Results

2.1. Dynamic upregulation of TRPC6 in brown adipocytes during cold adaptation

To systematically identify cold-responsive regulators in BAT, we subjected wild-type mice to 4 °C cold exposure for 7 days (Fig. 1A). During thermogenic activation, BAT lipid droplets significantly diminished post-exposure (Fig. 1B–C). Subsequent RNA-seq analysis of BAT revealed profound cold-induced transcriptional reprogramming, with 1,168 upregulated and 879 downregulated genes (Fig. 1D). Further GO analysis displayed enrichment in fatty acid metabolism, lipid metabolic processes, and adaptive thermogenesis (Fig. 1E), consistent with thermogenic activation. As TRP channels are established regulators of adipose metabolism [[27], [28], [29], [30]], we noted that within the TRPC subfamily, TRPC6 was significantly induced by cold (Fig. 1F). Furthermore, a progressive upregulation of Trpc6 mRNA levels throughout the cold exposure period in BAT was observed (Fig. 1G), which was confirmed at protein levels by western blot (Fig. 1H–I) and immunostaining (Fig. 1J–K). Consistently, the protein levels of UCP1 and PGC-1α, key markers of thermogenic activation, were also elevated following cold stimulation (Fig. 1H–I).

Figure 1.

Expression profile of TRPC6 in cold stimulation and mature adipocytes of BAT. (A) Wildtype mice were transferred to 4 °C for up to 7 d. (B) Representative H&E staining of BAT at room temperature (RT) and cold exposure (Scale bar, 100 μm). (C) Lipid droplet area (n = 8). (D) Volcano plot of differential genes by RNA-seq. (E) GO biological process analysis. (F) Relative expression of TRPC family genes differentially expressed (RT and 4 °C, n = 3). (G) The mRNA levels of Trpc6 in BAT after 7 days cold exposure (n = 3). Statistical comparisons were made between D1 vs. D0, D4 vs. D0, and D7 vs. D0 using one-way ANOVA. (H) The protein levels and (I) quantification of protein levels of TRPC6, UCP1 and PGC-1α in BAT (n = 5). (J) IHC staining (TRPC6) of BAT (Scale bar, 100 μm). (K) Quantitative analysis of IHC staining (TRPC6) of BAT (n = 7). (L) Fluorescence staining of primary brown adipocytes (Scale bar, 50 μm). (M) The mRNA level of Trpc6 in adipocytes and SVFs of BAT (n = 3). (N) Fluorescence staining of C3H10T1/2-derived adipocytes (Scale bar, 100 μm). (O) The protein levels and (P) quantification of protein levels of TRPC6, UCP1 and PGC-1α in preadipocytes (D0, D1, and D3) and adipocytes (D1, D3, D5, and D7) (n = 3). (Q) Trpc6 mRNA level in adipocytes and preadipocytes of C3H10T1/2 cells (n = 3). Data are presented as mean ± SEM and ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001, ns, not significant by unpaired two-tailed Student's t tests.

Importantly, Trpc6 expression was enriched in mature brown adipocytes, as evidenced by significantly higher mRNA levels in isolated mature adipocytes compared to the stromal vascular fractions (SVFs) (Fig. 1L-M). Notably, this specificity appeared unique to BAT, with differential expression in white adipose depots (Fig. S5G). Moreover, consistent with a role in adipocyte maturation, in vitro differentiation of C3H10T1/2 preadipocytes showed progressive upregulation of TRPC6 at both mRNA and protein levels (Fig. 1N-Q), which was synchronized with the upregulation of the key thermogenic markers UCP1 and PGC-1α (Fig. 1O-P), thereby correlating with the process of adipogenic maturation and thermogenesis. Collectively, our findings identify TRPC6 as a cold-responsive TRP channel selectively enriched in functional brown adipocytes.

2.2. BAT-specific TRPC6 deletion exacerbates HFD-induced obesity and metabolic dysfunction

To specifically investigate TRPC6's role in BAT-mediated energy metabolism, we generated brown adipocyte-specific Trpc6 knockout mice (Trpc6^BTKO) by crossing Trpc6^fl/fl mice with UCP1-Cre transgenic mice (Fig. S1A–D). Under a chow diet, Trpc6^BTKO mice and Trpc6^fl/fl mice exhibited comparable body weight, food intake, and adipose depot masses (Fig. 2A–E; Fig. S2A–B). However, BAT from Trpc6^BTKO mice displayed a pale appearance of whitening, characterized by enlarged lipid droplets and unilocular morphology (Fig. 2D; Fig. 2F–H). Importantly, this adipocyte hypertrophy was BAT-specific, with no detectable differences observed in inguinal white adipose tissue (iWAT) or epididymal WAT (eWAT) (Fig. S2C–H).

Figure 2.

TRPC6 deficiency in BAT aggravates HFD-induced Obesity. (A–H) Trpc6^fl/fl and Trpc6^BTKO mice were fed normal chow diet (NCD) for 9 weeks. (A) Representative images of mice. (B) Body weight (n = 6). (C) Averaged weekly food intake (n = 9 weeks/cage). (D) Representative images of BAT. (E) Relative BAT weight (n = 6). (F) Representative H&E stains of BAT (Scale bar, 100 μm). (G) Lipid droplet area distribution and (H) means in BAT (n = 7–8). (I–V) Trpc6^fl/fl and Trpc6^BTKO mice were fed high-fat diet (HFD) for 14 weeks. (I) Representative appearance of mice. (J) Body weight (n = 6). (K) Body composition of mice (n = 6). (L) Representative images of BAT. (M) Relative weights of adipose tissue and liver (n = 7–10). (N) Representative H&E stains of BAT (Scale bar, 100 μm). (O) Lipid droplet area distribution and (P) means in BAT (n = 7–8). (Q) Representative H&E stains of liver (Scale bar, 200 μm). (R, T) Glucose tolerance test (GTT) and insulin tolerance test (ITT) (n = 6). (S, U) Analysis of the GTT and ITT data using an area of the curve (n = 6). Data are presented as mean ± SEM and ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001, ns, not significant.

Upon HFD challenge, Trpc6^BTKO mice developed heightened weight gain and significantly higher fat mass compared to Trpc6^fl/fl mice (Fig. 2I–K; Fig. S3A). The depot weight significantly increases not only in BAT, but also in iWAT and eWAT (Fig. 2L-M; Fig. S3B). Additionally, BAT showed severely reduced multilocularity and pronounced whitening (Fig. 2N-P), while iWAT and eWAT displayed similar phenotypes (Fig. S3C–H). Furthermore, Trpc6^BTKO mice showed significantly increased liver weight (Fig. 2M) and exhibited significantly elevated lipid droplets (Fig. 2Q), alongside impaired glucose tolerance and insulin resistance (Fig. 2R-U). Collectively, these findings demonstrate that TRPC6 ablation in BAT predisposes to HFD-induced obesity and its associated metabolic dysfunctions.

2.3. TRPC6 deficiency impairs energy balance and cold tolerance

To further elucidate the underlying cause of weight gain in Trpc6^BTKO mice, we assessed their energy balance. Trpc6^BTKO mice exhibited significantly increased food intake (Fig. 3A) and reduced energy expenditure (Fig. 3B; Fig. S3I and S3L). This decreased expenditure was evidenced by significantly lower O₂ consumption and CO₂ production (Fig. 3C–D; Fig. S3J–K). Respiratory exchange ratio (RER) (Fig. 3E) and physical activity (Fig. S3M) did not differ between groups. These findings indicate that TRPC6 deficiency impacts energy balance by simultaneously promoting energy intake and suppressing energy expenditure.

Figure 3.

TRPC6 deficiency impairs energy balance and cold tolerance. (A) Averaged weekly food intake of Trpc6^fl/fl and Trpc6^BTKO mice fed HFD (n = 13 weeks/cage). (B) Energy expenditure (n = 4). (C) Oxygen consumption (n = 4). (D) Carbon dioxide production (n = 4). (E) Respiratory exchange ratio (n = 4). (F) Trpc6^fl/fl and Trpc6^BTKO mice were transferred to 4 °C for up to 7 d. (G) Core body temperatures upon acute 4 °C exposure (n = 6). (H) Representative forward-looking infra-red (FLIR) image and (I) analysis of capturing surface body temperature at 2 h acute cold exposure (n = 6). (J) Representative images of BAT after 7 days cold exposure. (K) Relative BAT weight (n = 6). (L) Representative H&E stains of BAT (Scale bar, 100 μm). (M) Lipid droplet area distribution and (N) means in BAT (n = 6–7). Data are presented as mean ± SEM and ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001, ns, not significant by unpaired two-tailed Student's t tests.

Trpc6^BTKO mice manifested as an impairment in thermogenic capacity and cold tolerance. Upon acute 4 °C exposure, Trpc6^BTKO mice exhibited significantly accelerated declines in core body temperature and significantly reduced BAT surface temperature (Fig. 3F–I). Although their BAT appeared paler, no significant differences in weight or size were detected after 7 days of cold exposure (Fig. 3J–K). BAT exhibited severe whitening, characterized by adipocytes with enlarged lipid droplets and a loss of multilocular morphology (Fig. 3L-N). In contrast, iWAT morphology remained unchanged under cold stress (Fig. S4A–E), and no alterations in adipose depots were observed (Fig. S4A–E). Collectively, these results demonstrate that BAT-intrinsic TRPC6 is essential for thermogenic capacity and cold defense.

2.4. TRPC6 ablation disrupts mitochondrial biogenesis and thermogenesis in mice

We further investigate the role of TRPC6 in BAT thermogenesis. BAT from Trpc6^BTKO mice exhibited significantly reduced protein levels of key thermogenic regulators UCP1, PGC-1α as well as components of the mitochondrial OXPHOS complexes (NDUFB8, SDHB, UQCRC2, MTCO1) (Fig. 4A–B). Correspondingly, mRNA levels of thermogenic markers, including Dio2, Cidea, and Prdm16, were also significantly decreased (Fig. 4C). Immunohistochemistry further confirmed attenuated UCP1 and PGC-1α protein levels in the multilocular adipocytes (Fig. 4D–E). Importantly, these changes were BAT-specific, with no alterations in thermogenic gene expression in iWAT (Fig. S4F).

Figure 4.

TRPC6 deficiency suppresses the thermogenic program via PGC-1α/UCP1 signaling in BAT. (A–G) Trpc6^fl/fl and Trpc6^BTKO mice were transferred to 4 °C for up to 7 d. (A) The protein levels and (B) quantification of protein levels of TRPC6, UCP1, PGC-1α, β-Actin, and OXPHOS complex proteins in BAT (n = 6). (C) The mRNA levels of thermogenic genes in BAT (n = 6). (D) IHC staining (UCP1 and PGC-1α) of BAT (Scale bar, 100 μm). (E) Quantitative analysis of IHC staining (UCP1 and PGC-1α) of BAT (n = 7). (F) Relative mitochondrial DNA (mtDNA) levels (n = 8). (G) Representative TEM images of mitochondria in BAT (Top, Scale bar, 1 μm; Bottom, Scale bar, 1 μm). (H–N) Trpc6^fl/fl and Trpc6^BTKO mice were fed HFD for 14 weeks. (H) The protein levels and (I) quantification of protein levels of TRPC6, UCP1, PGC-1α, β-Actin, and OXPHOS complex proteins in BAT (n = 6). (J) The mRNA levels of thermogenic genes in BAT (n = 7–9). (K) IHC staining (UCP1 and PGC-1α) of BAT (Scale bar, 100 μm). (L) Quantitative analysis of IHC staining (UCP1 and PGC-1α) of BAT (n = 8). (M) Relative mtDNA levels (n = 8). (N) Representative TEM images of mitochondria in BAT (Top, Scale bar, 1 μm; Bottom, Scale bar, 1 μm). Data are presented as mean ± SEM and ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001, ns, not significant by unpaired two-tailed Student's t tests.

Quantitative analysis revealed significant impairment in mitochondrial function. Specifically, mitochondrial DNA (mtDNA) copy number was significantly reduced after cold exposure (Fig. 4F). Electron microscopy also revealed reduced mitochondrial number and disrupted cristae structure in brown adipocytes (Fig. 4G), indicating compromised mitochondrial biogenesis. Furthermore, HFD-fed Trpc6^BTKO mice also displayed suppressed UCP1 and PGC-1α protein levels (Fig. 4H–I), and downregulated OXPHOS components in BAT (Fig. 4H–I). Similarly, the mRNA expression of Ucp1, Ppargc1a and other thermogenic markers (Cox7a1, Cox8b, Dio2, Cidea) was also significantly downregulated (Fig. 4J). Immunohistochemical staining reinforced this suppression, with significantly reduced UCP1 and PGC-1α in BAT (Fig. 4K–L). Moreover, mitochondrial analysis revealed reduced mtDNA content (Fig. 4M) and decreased mitochondrial number and cristae (Fig. 4N). Notably, Trpc6 knockout also impaired the induction of thermogenic programming in iWAT, evidenced by attenuated browning and downregulated mRNA expression of Cox7a1 and Cox8b (Fig. S3N). Collectively, our findings demonstrate that TRPC6 underpins BAT thermogenesis through PGC-1α and UCP1, and its ablation leads to compromised mitochondrial biogenesis and function and adaptive thermogenesis.

2.5. TRPC6 controls thermogenesis in brown adipocytes

To investigate the impact of TRPC6 on adipocytes, we isolated SVFs from BAT of Trpc6^+/+ and Trpc6^−/− mice and differentiated them into primary brown adipocytes (Fig. S1F–I). Strikingly, Trpc6 ablation in these primary adipocytes resulted in significantly larger lipid droplets (Fig. 5A–C) and concomitantly attenuated Ca²⁺ influx (Fig. 5D–E). This was further accompanied by reduced protein levels of UCP1, PGC-1α and the components of OXPHOS complexes (Fig. 5F–G), alongside significant transcriptional suppression of key thermogenic genes, including Ucp1, Ppargc1a, Cox7a1, Cox8b, Dio2, and Cidea (Fig. 5H). Critically, these functional and molecular impairments were specific to brown adipocytes, as white adipocytes showed no significant alterations (Fig. S5A–F).

Figure 5.

Impact of TRPC6 on adipocytes in vitro. (A) Representative fluorescence staining of primary brown adipocytes from Trpc6^+/+ and Trpc6^−/− (Scale bar, 50 μm). (B–C) Lipid droplets area (n = 11). (D) Ca²⁺ influx was detected by using the Ca²⁺ indicator Fluo-4 AM and (E) Ca²⁺ relative fluorescence intensity (n = 6). (F) The protein levels and (G) quantification of protein levels of TRPC6, UCP1, PGC-1α, β-Actin, and OXPHOS complex proteins (n = 3). (H) The mRNA levels of thermogenic genes (n = 4). (I) Oxygen consumption rate of Mito-stress test and (J) assay parameters of Mito-stress test (n = 9). (K) Representative Oil Red O-stained (top; Scale bar, 200 μm) and phase contrast images (bottom; Scale bar, 100 μm), and (L) quantitative analysis of lipid droplet area of C3H10T1/2-derived adipocytes infected with sh-Trpc6 or vector (Ctrl) (n = 10–18). (M) The protein levels and (N) quantification of protein levels of TRPC6, UCP1, PGC-1α, β-Actin, and OXPHOS complex proteins (n = 3). (O) The mRNA levels of thermogenic genes (n = 3). (P) Representative fluorescence staining (Scale bar, 100 μm) and (Q) lipid droplets area analysis of C3H10T1/2-derived adipocytes transduced with Trpc6^OE or vector (Ctrl) (n = 15). (R) The protein levels and (S) quantification of protein levels of TRPC6, UCP1, PGC-1α, β-Actin and OXPHOS complex proteins (n = 3). (T) The mRNA levels of thermogenic genes (n = 3). (U) Representative fluorescence images of mitochondrial morphology (Scale bar, 10 μm). (V) Quantitative analysis of mitochondrial density (n = 5). Data are presented as mean ± SEM and ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001, ns, not significant by unpaired two-tailed Student's t tests.

Seahorse metabolic flux analysis revealed significant mitochondrial dysfunction in Trpc6^−/− brown adipocytes. Specifically, basal respiration, maximal respiration, ATP production, and spare capacity were all significantly reduced, while proton leak remained unaffected (Fig. 5I–J), indicating a specific defect in coupled respiration. Recapitulating these findings in an established brown adipocyte model, Trpc6 knockdown in C3H10T1/2-derived adipocytes similarly increased lipid accumulation (Fig. 5K–L; Fig. S6A–B) and reduced both thermogenic protein (Fig. 5M−N) and mRNA levels (Fig. 5O). Conversely, supporting a direct regulator role, Trpc6 overexpression in C3H10T1/2-derived adipocytes promoted a thermogenic phenotype, characterized by smaller lipid droplets (Fig. 5P–Q; Fig. S6C–D) and the upregulation of thermogenic genes at the protein and mRNA levels (Fig. 5R–T), and an increased mitochondrial number (Fig. 5U–V). Together, our findings demonstrate that TRPC6 functions as an essential modulator of the thermogenic program in brown adipocytes.

2.6. TRPC6 regulates thermogenesis through selective p38 MAPK activation

Transcriptomic analysis of cold-exposed Trpc6^BTKO BAT revealed significant enrichment of MAPK signaling pathways (Fig. 6A). Given the established role of p38 MAPK in UCP1 activation and mitochondrial biogenesis [[31], [32], [33]], we specifically assessed its activation status. Across all models, Trpc6 ablation led to suppressed phosphorylation of p38 MAPK (p-p38), while the phosphorylation of both ERK and JNK remained unaffected (Fig. 6B–E) in both HFD and cold-exposure conditions. This specific suppression of p38 activation was recapitulated in Trpc6^−/− brown adipocytes (Fig. 6F–G) and in C3H10T1/2-derived adipocytes subjected to Trpc6 knockdown (Fig. 6H–I). Conversely, Trpc6 overexpression in C3H10T1/2-derived adipocytes significantly increased p-p38 levels (Fig. 6J–K).

Figure 6.

Knockout of TRPC6 in BAT impairs thermogenesis by inhibiting the p38-MAPK pathway. (A) KEGG enrichment process analysis of RNA-seq of Trpc6^fl/fl and Trpc6^BTKO mice after cold exposure (n = 3). (B) The protein levels and (C) quantification of protein levels of TRPC6, β-Actin, and p38 MAPK in BAT after cold exposure (n = 6). (D) The protein levels and (E) quantification of protein levels of TRPC6, β-Actin, and p38 MAPK in BAT after HFD (n = 6). (F) The protein levels and (G) quantification of protein levels of TRPC6, β-Actin, and MAPK in primary brown adipocytes (n = 3). (H) The protein levels and (I) quantification of protein levels of TRPC6, β-Actin, and MAPK in C3H10T1/2-derived adipocytes infected with sh-Trpc6 or vector (n = 3). (J) The protein levels and (K) quantification of protein levels of TRPC6, β-Actin, and MAPK in C3H10T1/2-derived adipocytes infected with Trpc6^OE or vector (n = 3). (L–M) Primary brown adipocytes loaded with Fluo-4 AM. Baseline [Ca²⁺]_i was acquired. Primary brown adipocytes, incubated in Ca²⁺-containing or Ca²⁺-free bath solutions, were exposed to HPF (5 μM) at 150 s. (L) A representative experiment was described where fluorescence was represented as F/F₀ as a function of time. (M) The first peak (n = 3). (N–O) C3H10T1/2-derived adipocytes loaded with Fluo-4 AM. (N) A representative experiment was described where fluorescence was represented as F/F₀ as a function of time. (O) The first peak (n = 3). (P–Q) Primary brown adipocytes were treated with vehicle (Ctrl) or HPF for 48 h. (P) The protein levels and (Q) quantification of protein levels of UCP1, PGC-1α, β-Actin, and p38 MAPK (n = 3). (R–S) Trpc6^−/− adipocytes were treated with vehicle or HPF for 48 h. (R) The protein levels and (S) quantification of protein levels of UCP1, PGC-1α, β-Actin, and p38 MAPK (n = 3). Unpaired two-tailed Student's t tests and one-way ANOVA were used appropriately. Data are presented as mean ± SEM and ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001, ns, not significant.

To pharmacologically interrogate this pathway dependence, we employed hyperforin (HPF) [34], a TRPC6 agonist. Adipocytes were treated with HPF in either Ca²⁺ or Ca²⁺-free buffer. HPF significantly increased intracellular calcium ([Ca²⁺]_i) in primary brown adipocytes under Ca²⁺ conditions, but this effect was completely abolished in Ca²⁺-free conditions (Fig. 6L–M; Fig. S7A). Similar results were also observed in C3H10T1/2-derived adipocytes, demonstrating the calcium dependence of HPF's action (Fig. 6N–O; Fig. S7B). Strikingly, hyperforin treatment rescued p38 phosphorylation in wild-type (Fig. 6P–Q) but failed in Trpc6^−/− brown adipocytes (Fig. 6R–S), demonstrating that TRPC6 is required for p38 MAPK activation. Therefore, our integrated genetic, cellular, and pharmacological evidence establishes that TRPC6 specifically gates BAT thermogenesis by selectively activating the p38 MAPK signaling pathway.

2.7. TRPC6 controls thermogenesis through BMPR2-dependent p38 MAPK activation

Protein interaction network analysis identified a direct interaction between TRPC6 and BMPR2 (Fig. 7A), a receptor of the bone morphogenetic protein (BMP) family. BMP signaling is an established regulator of adipocyte differentiation [[35], [36], [37], [38]], and more evidence also highlights its crucial role in modulating adipocyte thermogenesis [[39], [40], [41]]. This suggests a functional link between TRPC6 and BMP-mediated regulation of adipocyte thermogenesis.

Figure 7.

TRPC6 knockdown suppresses thermogenic program through BMPR2. (A) Protein–protein interaction (PPI) network of TRPC6 predicted by STRING database. (B–C) Surface electrostatic potential of TRPC6 and BMPR2 (Blue: positive charge, Red: negative charge). (D) Molecular docking analysis indicated possible binding interactions between TRPC6 and BMPR2 proteins (Purple: TRPC6, Turquoise: BMPR2). (E) Immunoprecipitation confirmed the interaction between TRPC6 and BMPR2. (F) Representative fluorescence staining and (G) gray value of TRPC6 and BMPR2 in BAT (Red: TRPC6, Green: BMPR2, Blue: DAPI; Scale bar, 10 μm). (H) The protein levels and (I) quantification of protein levels of TRPC6, BMPR2, and β-Actin in primary brown adipocytes from Trpc6^+/+ and Trpc6^−/− (n = 3). (J) The protein levels and (K) quantification of protein levels of TRPC6, BMPR2, and β-Actin in C3H10T1/2-derived adipocytes infected with sh-Trpc6 or vector (n = 3). (L) The protein levels and (M) quantification of protein levels of TRPC6, BMPR2, and β-Actin in C3H10T1/2-derived adipocytes infected with Trpc6^OE or vector (n = 3). (N) Representative fluorescence staining (Scale bar, 100 μm) and (O) lipid droplets area analysis of C3H10T1/2-derived adipocytes transfected with Bmpr2 siRNA (si-Bmpr2) or vector (Ctrl) (n = 11–12). (P) The protein levels and (Q) quantification of protein levels of TRPC6, UCP1, PGC-1α, β-Actin, and OXPHOS complex proteins (n = 3). (R) The protein levels and (S) quantification of protein levels of BMPR2, β-Actin, and p38 MAPK (n = 3). (T–U) C3H10T1/2-derived adipocytes of Trpc6^OE transfected with Bmpr2 siRNA or vector. (T) The protein levels and (U) quantification of protein levels of TRPC6, BMPR2, UCP1, PGC-1α, β-Actin, and p38 MAPK (n = 3). (V) The protein levels and (W) quantification of protein levels of UCP1, PGC-1α, β-Actin, and p38 MAPK in primary brown adipocytes following 48 h treatment with vehicle (Ctrl) or DHC (1 μM) (n = 3). (X–Y) Trpc6^−/− adipocytes were treated with vehicle or DHC for 48 h. (X) The protein levels and (Y) quantification of protein levels of UCP1, PGC-1α, β-Actin, and p38 MAPK (n = 3). Unpaired two-tailed Student's t tests and one-way ANOVA were used appropriately. Data are presented as mean ± SEM and ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001, ns, not significant.

Structural validation via molecular docking revealed a high–affinity complex stabilized by hydrogen bonds with BMPR2 residues Thr559, Asp560, Ser561, and Glu188 (Fig. 7B–D). Critically, co-immunoprecipitation experiments confirmed their physical interaction (Fig. 7E), while immunofluorescence further demonstrated membrane co-localization in BAT (Fig. 7F–G). Functionally, we established that TRPC6 regulates BMPR2 expression. Trpc6 knockdown significantly reduced BMPR2 protein levels in C3H10T1/2-derived adipocytes and primary adipocytes (Fig. 7H–K), whereas Trpc6 overexpression conversely increased BMPR2 expression (Fig. 7L–M). Strikingly, Bmpr2 knockdown phenocopied Trpc6 ablation, inducing lipid droplet enlargement (Fig. 7N–O; Fig. S6E–F), suppressing thermogenesis (Fig. 7P–Q; Fig. S6G), and impairing p38 phosphorylation (Fig. 7R–S). Most importantly, Bmpr2 silencing abolished the thermogenic enhancement mediated by Trpc6 overexpression (Fig. 7T–U), demonstrating the absolute requirement for BMPR2 in this pathway. Additionally, we found that a p38 agonist, dehydrocorydaline chloride (DHC) [42,43], promotes thermogenesis in adipocytes and rescues the thermogenic defect caused by TRPC6 deficiency (Fig. 7V–Y; Fig. S5H–I). Collectively, these results define a mechanism wherein TRPC6 governs BAT thermogenesis by forming a signaling complex with BMPR2 to activate p38 MAPK.

3. Discussion

This study elucidates a fundamental role for the calcium-permeable channel TRPC6 in governing BAT thermogenesis and systemic metabolic health. Through integrated analyses of BAT-specific Trpc6 knockout mice and complementary in vitro and mechanistic investigations, we demonstrate that TRPC6 is indispensable for maintaining BAT identity, mitochondrial function, and adaptive thermogenesis. Critically, we uncover the TRPC6-BMPR2-p38 MAPK signaling axis as the essential pathway coupling calcium influx to the transcriptional activation of the PGC-1α/UCP1 program, thereby gating energy dissipation.

Although TRPC6 has been implicated in nephrotic syndrome, neurodegeneration, and ischemia-reperfusion injury [[44], [45], [46], [47]], it has not previously been linked to BAT thermogenesis. This gap is notable given the metabolic functions of other TRP channels: TRPV2 modulates BAT thermogenesis and differentiation [48], TRPM8/TRPV1 stimulates UCP1-dependent thermogenesis [49,50], and TRPV1/TRPC5 promotes hyperphagia [[51], [52], [53]]. Our work now establishes TRPC6 as a critical, BAT-intrinsic thermoregulator operating via a mechanism distinct from its relatives.

Previous global Trpc6 KO caused increased body weight and food intake, which was linked to impaired hypothalamic ARC POMC neuron axonal projections that disrupt feeding regulation [54]. Our findings now reveal a pronounced peripheral metabolic function: BAT-intrinsic TRPC6 ablation drives obesity susceptibility through a dual mechanism—suppressing thermogenic energy expenditure while simultaneously increasing energy intake. The spontaneous BAT whitening and mitochondrial dysfunction observed under basal chow diet conditions demonstrate that TRPC6 is constitutively required for maintaining brown adipocyte identity. Under metabolic stress, BAT-specific TRPC6 loss accelerated obesity, systemic adiposity, hepatic steatosis, and insulin resistance. Importantly, cold intolerance and impaired lipid utilization confirm a thermogenic defect independent of central appetite circuits, positioning TRPC6 as a vital regulator of adaptive thermogenesis. Moreover, beige adipose tissue also plays a critical role in adaptive thermogenesis [7]. However, it remains challenging to precisely distinguish and quantify the respective contributions of brown versus beige adipocytes to the overall thermogenic phenotype in vivo.

Cold-induced TRPC6 upregulation facilitates calcium influx, which is intrinsically coupled to thermogenic activation. In this study, we reveal a direct interaction between TRPC6 and BMPR2, forming a signaling complex essential for selective p38 MAPK activation. This partnership is functionally obligate: BMPR2 knockdown not only phenocopied the lipid accumulation and thermogenic suppression seen in TRPC6-deficient adipocytes but also abolished TRPC6-mediated thermogenic enhancement. The pathway specificity is underscored by our finding that TRPC6 ablation impaired cold-induced p38 phosphorylation without affecting ERK/JNK activity. This TRPC6-BMPR2 axis thus bridges calcium flux to transcriptional control via p38-mediated PGC-1α/UCP1 induction, a previously unknown linkage. It is therapeutically significant that both pharmacological TRPC6 activation and genetic overexpression rescued p38 signaling and thermogenic output, validating the pathway's causal and druggable nature. These findings are consistent with the fact that BMP signaling mediated by BMPR2, has been well established as a key regulator of adipocyte differentiation and thermogenic programming [[38], [39], [40], [41],[55], [56], [57]].

The identified TRPC6-BMPR2-p38 axis may represent a potential therapeutic target for obesity. Human transcriptomic data from GeneCards, Human Protein Atlas and GTEx show that TRPC6 expression is comparatively high in adipose tissue and increases as brown adipocytes differentiation (Fig. S8). While this suggests a potential physiological role, direct functional evidence in humans remains to be established. Moreover, a few questions are required to be addressed. Firstly, there is lacking direct evidence to confirm causality of TRPC6 in human obesity. Secondly, due to TRPC6's vital roles in other organs such as kidney and heart, achieving adipose-selective activation will be crucial to avoid off-target effects. Future efforts are required to develop tissue-specifical activators or modulator.

In summary, TRPC6 functions as a non-redundant molecular switch for BAT thermogenesis via its dual coupling of calcium handling to BMPR2-dependent p38 MAPK activation. Disruption of this axis compromises mitochondrial integrity, drives BAT whitening, and precipitates obesity-related metabolic syndrome. Beyond its established roles in other tissues, we established TRPC6 as a central BAT-intrinsic rheostat, with its functional partnership with BMPR2 defining a previously unrecognized membrane signaling complex in thermogenic regulation. The TRPC6-BMPR2-p38 MAPK-UCP1 axis presents an actionable node for therapeutic intervention, highlighting TRPC6 activators and the TRPC6-BMPR2 interface as promising avenues for combating obesity and metabolic disease.

4. Materials and methods

4.1. Animal experiments

UCP1-Cre mice were purchased from the Model Animal Research Center of Nanjing University (Nanjing, China). Trpc6^−/− and Trpc6^fl/fl mice were obtained from GemPharmatech Co., Ltd (Nanjing, China). Trpc6^fl/fl mice on a C57BL/6J background were intercrossed with UCP1-Cre mice on a C57BL/6J background to generate Trpc6^BTKO mice (brown adipose tissue-specific knockout). All animals were maintained at 22 °C with a humidity of 50 ± 5% in a 12 h/12 h light/dark cycle and were fed normal chow diet and water. To induce obesity, male mice aged 6–8 weeks were fed a diet containing 60% fat (Xietong Biotech, XTHF60) for 12–16 weeks to induce obesity. For cold exposure in middle-aged mice, Trpc6^fl/fl and Trpc6^BTKO mice were housed separately at 4 °C with adequate food and water, maintaining the same light/dark cycle. Experimental animals were grouped in a randomized manner, and investigators were to the allocation of mice groups when conducting experimental protocols. At the conclusion of the experiments, mice were anesthetized with 2.5% isoflurane, followed by exsanguination and tissue collection. The animal protocol was reviewed and approved by the Animal Experiment Center of Zhujiang Hospital of Southern Medical University (approval number LAEC-2022-088).

4.2. Body composition analysis

The body composition of mice was measured using a small animal body composition analyzer (Niumag, QMR20-060H–I) as previously described [58], which provided measurements of fat mass, lean mass, and free fluid. Prior to the measurements, mice were fasted overnight and anesthetized with tribromoethanol injection. The machine was preheated, standard sample calibration was performed, and mice were placed in the detection tube in a prone position. Measurements were taken three consecutive times with the average value recorded.

4.3. High-definition multiplexed respirometry system

As previously described [59], mice were individually housed in the metabolic cage system (Sable Systems, PRO-MRM-4) under strictly controlled environmental conditions (temperature: 22 °C, humidity: 50 ± 5%, 12 h/12 h light/dark cycle). Before collecting data, mice were acclimated to the metabolic cages for 24 h to minimize stress-related artifacts. Over a 48-hour period, the system tracked oxygen consumption (VO₂), carbon dioxide production (VCO₂), and respiratory exchange ratio (RER) at 5-minute intervals. Oxygen consumption, carbon dioxide production, and energy expenditure were normalized to body weight as previously described [60].

4.4. Infrared thermography

According to previous publications [60], the BAT temperature was measured using a thermal imager (Fotric, 323Pro) at room temperature or 4 °C. Images were obtained from the backs of the mice and presented using a rainbow high-contrast color palette, with temperatures linearly represented between 25 °C and 38 °C. The emissivity setting was fixed at 0.95.

4.5. Glucose and insulin tolerance tests

Trpc6^fl/fl and Trpc6^BTKO mice were fed HFD for 12 weeks. For the insulin tolerance test (ITT): after fasting for 6 h, mice were intraperitoneally injected with insulin (0.75 IU/kg). Blood glucose levels were promptly measured using a glucose meter by tail venipuncture immediately before injection (0 min), and after injection (30, 60, 90, 120 min). For the glucose tolerance tests (GTT): mice were fasted for 16 h before intraperitoneal injection of 10% glucose solution (0.01 ml/g). Blood glucose measurements for the GTT followed the same protocol as for the ITT.

4.6. Real-time PCR

Total RNA was extracted from the adipose tissues of mice in each group and cultured cells using TRIzol reagent (Accurate Biology, AG21024). cDNA was then synthesized using the Evo M-MLV Reverse Transcription Kit (Accurate Biology, AG11706) from extracted RNA. PCR amplification was performed using the AG qPCR Kit (Accurate Biology, AG11701) with a 10 μl reaction system: initial denaturation at 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s, 60 °C for 60 s. The melting curve ranged from 60 to 97 °C. Data analysis was performed using the 2^−ΔΔCT relative quantification method, employing β-actin and RPS18 as the normalization controls. Primer sequences were designed based on the National Center for Biotechnology Information (NCBI) Primer BLAST (Supplementary Table 1 of sequences) and synthesized by IGE Biotechnology Co., Ltd (Guangzhou, China).

4.7. Mitochondria/nuclear DNA quantification

The protocol was modified from previous research [61]. DNA was isolated using Universal Genomic DNA Kit (Servicebio, G3633). qPCR was run using the following primers: ND1 and RPS18 primers (sequences provided in Supplement Table 1).

4.8. Protein extraction and western blot

Proteins were initially extracted from adipose tissues or cultured cells using RIPA lysis buffer (Cwbio, CW2333S) containing protease inhibitors and phosphatase inhibitors (Cwbio, CW2200S/CW2383S). The protein concentration of the samples was determined using the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific, 23227). The proteins were then denatured through heat treatment after the addition of loading buffer (Fude, FD002). Equal amounts of total protein were separated using 10% or 12.5% SDS-PAGE and then transferred to PVDF membranes (Millipore, IPFL00005). Following blocking with 5% skimmed milk (Epizyme, PS112) for 1.5 h at room temperature, the membranes were incubated overnight at 4 °C with various primary antibodies. Afterward, the membranes were incubated with horseradish peroxidase-conjugated anti-rabbit or anti-mouse secondary antibodies for 1 h at room temperature. Protein bands were visualized using an ECL kit (Millipore, WBULS0100), and the band intensities were quantified using ImageJ software. β-Actin served as an internal control for assessing protein expression levels. The antibodies used in this study are summarized in Supplementary Table 2.

4.9. Molecular docking analysis

To investigate the interaction mechanism between the target protein and its binding partner, protein–protein docking was performed using the GRAMM Docking Web Server. After docking, the results were evaluated using PDBePISA's scoring system. Finally, the protein–protein interactions were predicted, and a visualization of the complex was created using PyMOL. In the figure, the TRPC6 protein is depicted as a purple-colored cartoon, BMPR2 as a turquoise-colored cartoon, with their respective binding sites represented as colored stick models. When focusing on the binding region, the corresponding protein is highlighted to show the binding site specific to it.

4.10. Co-immunoprecipitation

Cells were lysed using IP lysis buffer (Beyotime, P0013) supplemented with protease and phosphatase inhibitor cocktails (Cwbio, CW2200S/CW2383S). After centrifuging the lysates at 12,000 rpm for 15 min at 4 °C, the supernatants were incubated with 2 μg of anti-FLAG (Proteintech, 66008-4-Ig) under gentle rotation at 4 °C overnight. Subsequently, Protein A/G Magnetic Beads (MCE, HY-K0202) were added, followed by incubation with rotation at room temperature for 1 h. The bound proteins were eluted through incubation with 1 × loading buffer (Fude, FD002) and rotation for 15 min, and finally, the eluates were denatured by boiling for 10 min.

4.11. Immunofluorescence staining

After embedding mouse adipose tissue with OCT embedding medium (Sakura, 4583), 6-μm thick sections were cut using a cryostat (Leica, CM1950). The sections were then fixed with 4 % paraformaldehyde (Beyotime, P0099) for 15 min, rinsed with PBS (Boster, AR0030), and permeabilized with Triton X-100 (Solarbio, IT9100) for 15 min. Subsequently, the sections were blocked with goat serum (Boster, AR0009) for 30 min. Specific primary antibodies were diluted according to the manufacturer's instructions and incubated overnight at 4 °C. The following day, the sections were incubated for 1 h with secondary antibodies conjugated to AlexaFluor488 and AlexaFluor568 (Invitrogen, A-11008/A-11004), followed by DAPI (Solarbio, S2110) mounting. After mounting, images were captured using fluorescence microscope (Nikon, Ti2-E) and analyzed statistically using ImageJ software.

4.12. Histopathology

Fresh adipose and liver tissue specimens from mice were fixed overnight at 4 °C in a 4% paraformaldehyde solution (Beyotime, P0099). The tissues were then dehydrated using gradient ethanol, cleaned with xylene, and embedded in paraffin. These paraffin-embedded tissues were sliced into 4 μm-thick sections and stained with hematoxylin and eosin (H&E). Images were captured using a digital pathological slide scanner (3D HISTECH). Additionally, the adipose area in the H&E-stained sections was quantitatively analyzed using ImageJ software.

4.13. Immunohistochemical staining

Paraffin sections (4 μm) of adipose tissue were deparaffinized in xylene and hydrated in alcohol series, and then the antigens on the sections were repaired using sodium citrate buffer at 100 °C before cooling to room temperature. Sections were incubated at room temperature with 3% H₂O₂ for 15 min to eliminate endogenous peroxidase activity and wash with ddH₂O three times. They were then blocked with goat serum for 30 min. Later, followed by the primary antibody of UCP1 (Abcam, ab234430, 1:500) and PGC-1α (Abclonal, A20995, 1:300) incubation at 4 °C overnight. After washing with PBS, the sections were then incubated with HRP-conjugated secondary antibody (Epizyme, LF102, 1:500) at 37 °C for 1 h. The sections were incubated with the DAB kit (Zsgb Biotech, ZLI-9017) for 3–5 min. The sections were then rinsed, stained with hematoxylin, dehydrated, cleaned, and sealed. Images of the sections were captured and scanned using a digital pathological slide scanner as previously described.

4.14. Cell culture and treatment

The C3H10T1/2 cells (Lot No. CL-0325) were obtained from Procell Life Science & Technology Co., Ltd (Wuhan, China). The cell line (RRID: CVCL_0190) was tested and confirmed to be negative for bacterial, fungal, and mycoplasma contamination. All experiments were performed using cells between passages 6 and 15. The cells were cultured in high-glucose (4.5 g/L) DMEM (Gibco, C11995500BT) supplemented with 10% fetal bovine serum (FBS; ExCell, FSP500) and 0.5% penicillin/streptomycin (Gibco, 15140122) at 37 °C in a 5% CO₂ incubator. The differentiation protocol was performed as previously described [62]. For brown adipocytes differentiation, confluent cells were treated with induction medium consisting of high-glucose DMEM containing 10% FBS, 2 μg/ml dexamethasone (Sigma, D1756), 0.5 mM 3-isobutyl-1-methylxanthine (IBMX; TargetMol, 28822-58-4), 62.5 μM indomethacin (Macklin, 53-86-1), 1 μM insulin (Solarbio, I8830), 1 μM rosiglitazone (Macklin, R832516), and 1 nM triiodothyronine (T3; Sigma, T2877). After 3 days of induction, the medium was substituted with differentiation medium composed of high-glucose DMEM, 10% FBS, 1 μM insulin, 1 μM rosiglitazone, and 1 nM T3 for 7 days culture. Before cell harvesting, Bodipy staining and Hoechst staining were performed using a 2.5 μM lipid probe (MCE, HY-W090090) and Hoechst 33342 reagent (Solarbio, C0031), respectively. Images were captured using fluorescence microscope. The lipid droplets were stained with the Oil Red O staining kit (Beyotime, C0158S), and images were then acquired under a microscope. The area occupied by these lipid droplets was subsequently analyzed using ImageJ software.

4.15. Lentivirus-mediated silencing/overexpression of TRPC6

To knockdown Trpc6, sh-Trpc6 was generated by GeneChem Co., Ltd (Shanghai, China). For Trpc6 overexpression, we constructed two viral vectors: an untagged version (GenePharma Co., Ltd.) and a dual-tagged version containing both Flag and eGFP tags (IGE Biotechnology Co., Ltd.). Transduction efficiency was assessed by fluorescence microscopy. The specific shRNA target sequence for mouse Trpc6 was as follows: 5′-CCGGGCTGCACATTGCCAGGAATATCTCGAGATATTCCTGGCAATGTGCAGCTTTTTG-3′.

4.16. siRNA transfection

Bmpr2 siRNA and scrambled siRNA control were synthesized by GenePharma Co., Ltd (Suzhou, China), with the following sequences: si-Bmpr2: Sense: 5′-GCAAUCUCCCACCGAGAUUTT-3′; Antisense: 5′-AAUCUCGGUGGGAGAUUGCTT-3′. Transfection was performed using GenePharma siRNA-mate plus transfection reagent according to the manufacturer's instructions. The transfection complexes were incubated in complete medium for 24 h. Subsequent experiments were conducted 48–72 h post-transfection.

4.17. SVF isolation and primary cell culture

Cells were extracted from the stromal vascular fractions (SVF) of male Trpc6^+/+ and Trpc6^−/− mice aged 6–8 weeks. Brown adipose tissue was finely minced and digested in 2.4 μg/ml dispase II (Sigma, D4693), 10 mg/ml collagenase type II (Sigma, 9001-12-1) and 10 mM CaCl₂ (Macklin, 10043-52-4) in PBS at 37 °C for 30 min. The digestion process was stopped by complete DMEM, and the digested tissue was filtered through a 40 μm strainer. The cell suspensions were centrifuged at 4 °C for 10 min, 1,000 rpm, which were then suspended and plated onto plates. SVF cells were cultured in high-glucose (4.5 g/ml) DMEM with 10% FBS and 0.5% pen/strep. The differentiation protocol was performed as previously described [62]. To investigate the effects of the compounds, differentiated mature adipocytes were treated with vehicle, 5 μM hyperforin (HPF; MCE, HY-116330A) or 1 μM dehydrocorydaline chloride (DHC; MCE, HY-N0674A) for 48 h.

4.18. Oxygen consumption rate assays

OCR was quantified using an Agilent Seahorse XF96 instrument (Agilent Technologies, 103015-100). After 3 days of induction, brown preadipocyte cells from Trpc6^+/+ and Trpc6^−/− BAT were digested with 0.25% trypsin (Gibco, 25200072) and seeded into seahorse 96-well plates (2000 cells/well). Oxygen consumption rate assays were conducted on the seventh day of primary adipocytes differentiation. Prior to the assay, the instrument was preheated, and the probe plate was hydrated overnight. Cell culture plates were rinsed three times with a working assay solution (pH 7.4) consisting of 1 mM glucose, 2 mM glutamine, and 10 mM pyruvate. Subsequently, 180 μl of the working solution was dispensed into each well. Cells were incubated in a CO₂-free environment at 37 °C for 50 min. The hydrated probe plate was added with drugs (oligomycin 1.5 μM, carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone 3 μM, antimycin/rotenone 1 μM) and then put into the cell plate to run the XF Cell Mito Stress Test program. Cell counts for each well were determined using Cytation 5, and the data collected in Seahorse Software Wave Desktop (v2.6) were adjusted according to the corresponding cell counts.

4.19. Calcium imaging

Cells were washed with PBS. Each tube was then dispensed with 0.5 ml of 2 μM Fluo-4 AM (Beyotime, S1061S) working solution and incubated in darkness at 37 °C for 30 min. Subsequently, cells were washed with PBS and promptly analyzed using flow cytometry (excitation 488 nm, emission collected at 512/520 nm). Furthermore, intracellular Ca²⁺ concentration ([Ca²⁺]_i) was evaluated by fluorescence microscopy as described previously [63]. Cells were loaded with 2 μM Fluo-4 AM at 37 °C for 30 min, and [Ca²⁺]_i was measured using time-lapse imaging (3 s intervals) on a fluorescence microscope. Baseline [Ca²⁺]_i was acquired for 120 s. Then, cells were incubated in Ca²⁺-containing (containing 1.8 mM Ca²⁺) or Ca²⁺-free bath solutions (containing 100 μM EGTA, Beyotime, ST068) in the presence of 5 μM HPF. [Ca²⁺]_i was described as a fluorescence intensity ratio of F/F₀, where F₀ indicated the resting fluorescence intensity, and F was the fluorescence intensity of cells treated with HPF stimulation. We quantify the data by calculating the F/F₀ increment of the first peak.

4.20. Transmission electron microscopy

BAT was minced into 1 mm × 1 mm and primarily fixed with Karnovsky's fixative. After postfixing with 1% osmium tetroxide for 1 h followed by staining with 2% uranylacetate in maleate buffer, pH 5.2 for 1 h and dehydrating with ethanol gradient. They were embedded into resin and baked overnight at 70 °C. Blocks were sectioned at 60 nm using an ultramicrotome (Leica, EM UC7) and transferred onto copper grids. Images were captured using transmission electron microscope (FEI, Talos L120C).

4.21. RNA-seq analysis

Total RNA was extracted from the tissue samples. The RNA libraries were sequenced on the illumina NovaseqTM 6000 platform by LC Biotechnology Co., Ltd (Hangzhou, China). Differentially expressed genes (DEGs) were identified using DESeq2 with the following criteria: |log2(fold change) | > 1 and adjusted P-value <0.05. Functional enrichment analysis, including GO terms and KEGG pathways, was carried out with ClusterProfiler. Heatmaps were generated using pheatmap, and principal component analysis (PCA) was conducted to evaluate sample clustering. All experiments included at least three biological replicates.

4.22. Statistical analysis

The statistical analyses in validation experiments were conducted using IBM SPSS Statistics (SPSS, v26.0), and the graphs were generated using GraphPad Prism 10.0 software. The data are presented as mean ± SEM. Unpaired Student's t-test and one-way ANOVA were used to compare means of numerical variables where appropriate. P < 0.05 was considered statistically significant, while ns signifies non-significance (P > 0.05).

CRediT authorship contribution statement

Danyingzhu Xie: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Ying Peng: Visualization, Methodology. Susu Zhang: Visualization, Resources, Methodology. Xudong Mai: Visualization, Methodology. Weiheng Wen: Visualization, Methodology. Ming Wang: Supervision, Resources, Funding acquisition. Qiao-Ping Wang: Writing – review & editing, Writing – original draft, Visualization, Supervision. Jia Sun: Writing – review & editing, Writing – original draft, Visualization, Supervision, Resources, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article.

Data availability

Data will be made available on request.