SCN lesioning preserves BAT thermogenesis and disrupts fuel partitioning under TRF-STE

To examine the role of the SCN in thermogenic adaptation under combined nutritional and thermal stress, we implemented a TRF protocol in STE (TRF-STE), a model previously shown to induce hypothermia in wild-type (WT) mice [14]. Mice were restricted to a 4-hour feeding window (ZT16-ZT20) to minimize disruption of SCN-driven circadian rhythmicity [30]. Core body temperature (CBT) was recorded using implanted thermochips, and BAT surface temperature was monitored by infrared thermography. For reference, we reanalyzed previously published CBT data obtained from WT mice housed under conventional housing conditions (CC, 23°C−25°C) [14], where TRF induced only a modest CBT reduction when switched from ad libitum to TRF (Fig 1A, dash line). In contrast, under 21°C (STE) conditions, sham-operated mice (sham mice) displayed a much larger drop in CBT and BAT temperature upon food withdrawal at ZT20, consistent with a hypothermic adaptation to fasting (Fig 1A and 1B). By contrast, SCN-lesioned (SCNx) mice (S1A Fig, described in Materials and methods) showed attenuated hypothermia during the second TRF cycle, maintaining significantly sustained CBT and BAT temperatures than sham mice at 21°C or WT under CC (Fig 1A and 1B). To distinguish stress-induced hypothermia from the normal sleep-associated temperature dip, we analyzed the TRF at different circadian phases (ZT4-8, ZT10-14, and ZT22-2). Hypothermia coincided consistently with food deprivation across phases, independently of sleep onset (S1B Fig), indicating that SCN lesioning prevents adaptive hypothermia rather than abolishing the physiological circadian trough. Tail temperature, a proxy for peripheral vasodilation, remained unchanged between groups (Fig 1A).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. SCN lesioning preserves BAT thermogenesis and impairs fuel switching under TRF-STE conditions.

(A) Core body temperature (top), including WT mice under conventional conditions CC, 23°C–25°C; dashed line, data reanalyzed from Zhang and colleagues, 2020), interscapular BAT surface temperature (middle), and tail temperature (bottom) in sham or SCNx mice under Ad-STE and TRF-STE. The feeding window (ZT16-ZT20) is indicated by the green shaded area. WT: n = 3 (reanalyzed from Zhang and colleagues, 2020); sham and SCNx under Ad-STE: n = 10, sham and SCNx under TRF: n = 5. All mice were 8 weeks old at the start. After SCN lesioning, mice recovered for 1 week, underwent 1 week of wheel-running screening, and only arrhythmic mice without abrupt body weight gain were selected for 2 weeks of acclimation in 21°C chambers before TRF experiments. Data are shown as mean ± SEM. (B) Representative infrared thermographic images showing surface body temperature at ZT4 (rest phase) and ZT16 (active phase) under TRF-STE in sham and SCNx mice. (C–E) Measurements of total VO₂ (C), total VCO₂ (D), and total EE (E) on day 2 of TRF-STE (n = 10 per group). (F) Food intake analysis under Ad-STE and TRF-STE in sham (n = 6) and SCNx mice (n = 8). (G–J) Body composition analysis of fat mass and lean mass in sham (n = 10), SCNx (n = 12), and HFD-fed mice (n = 6). (K) GLM analysis of EE with group as the independent variable and body weight as the covariate under TRF-STE conditions (n = 10 per group). (L) CBT in sham (n = 5), HFD-fed (n = 4), and SCNx (n = 5) mice during TRF-STE. Data are mean ± SEM. (M) ΔRER at ZT16 following the transition from Ad-STE to TRF-STE in sham and SCNx mice (n = 10 per group). Unless otherwise indicated, data are presented as mean ± SD. NS: not significant, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. Tests used: unpaired two-tailed Student t test (C–J, and M), and two-way ANOVA with Sidak’S multiple comparisons test (A, L). The data underlying the graphs shown in the figure can be found in S1 Source Data.

https://doi.org/10.1371/journal.pbio.3003534.g001

To further examine whole-body energy metabolism, we assessed oxygen consumption (VO₂) and energy expenditure (EE) via indirect calorimetry. Sham and SCN-lesioned mice maintained comparable VO₂ and EE under both ad libitum (Ad-STE) (S1C–S1F Fig) and TRF-STE conditions (Figs 1C–1E and S1G), with no significant differences in food intake (Fig 1F). However, body composition analysis revealed an increase in fat mass and a concomitant reduction in lean mass in SCN-lesioned mice despite similar body weight (Fig 1G–1J). To determine whether SCN lesioning alters EE independently of body mass, we applied a General Linear Model (GLM) with body weight as a covariate. A significant group × mass interaction violated the assumptions required for analysis of covariance (ANCOVA), but GLM still identified a significant main effect of group on EE (Figs 1K and S1H). These results indicate that SCN lesioning reduces EE.

To test whether increased fat deposition could offset the elevated thermogenic demand imposed by TRF-STE, sham mice were fed a HFD prior to exposure. Despite marked increases in body weight and adiposity (Fig 1G–1J), HFD-fed mice still exhibited a significant drop in CBT during TRF-STE (Fig 1L). These results indicate that increased fat mass alone is insufficient to counteract hypothermia under combined nutritional and thermal stress, supporting a key role in SCN-mediated control of thermogenic adaptation.

SCN-lesioned mice displayed significantly lower locomotor activity under Ad-STE and TRF-STE (S1I and S1J Fig), which may partially account for their decreased EE (Figs 1K and S1H). However, their ability to maintain body temperatures despite reduced EE suggests the presence of a compensatory mechanism in energy metabolism. To investigate whether this compensation involved altered substrate utilization, we examined changes in respiratory exchange ratio (ΔRER) from Ad-STE to TRF-STE transition as an indicator of fuel preference. Under TRF-STE, sham mice displayed the expected shift from carbohydrate to lipid oxidation (Fig 1M), reflecting adaptive metabolic flexibility. In contrast, SCN-lesioned mice failed to make this transition, maintaining a glucose-dependent profile (Fig 1M). This impaired shift in fuel utilization indicates a loss of metabolic flexibility, likely attributable to impaired SCN function and decreased lipid mobilization.

SCN lesioning impairs lipid mobilization but enhances glucose-driven thermogenesis in BAT under TRF-STE

To investigate the mechanisms underlying preserving thermogenesis in SCN-lesioned mice under combined nutrient and thermal stress, we assessed tissue-specific glucose uptake and metabolic flexibility. Utilizing positron emission tomography/computed tomography (PET/CT) with ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) at ZT6 after a 10-hour fast, we observed a significant increase in glucose uptake in SCN-lesioned BAT compared to sham controls, a finding corroborated by gamma counting (Fig 2A–2C). In contrast, glucose uptake in WAT was significantly reduced, while hepatic uptake remained unchanged (Fig 2C), suggesting a selective redistribution of glucose toward BAT under SCN disruption and a metabolic shift favoring glucose as a thermogenic fuel.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. SCN lesioning impairs lipid mobilization and metabolic flexibility under TRE-STE.

(A–C) ¹⁸F-FDG PET-CT imaging and glucose uptake quantification. Representative PET-CT images 60 min after intravenous injection of ¹⁸F-FDG under TRF-STE (A). Quantification of ¹⁸F-FDG uptake across tissues by PET-CT (B). Radioactivity measurements in interscapular BAT, liver, and eWAT using gamma counting (C). n = 5 per group. (D and E) Glycolytic flux in interscapular BAT explants. Seahorse extracellular flux of ECAR in interscapular BAT explants from sham and SCNx mice under TRF-STE at ZT16 (D). Quantification of basal glycolysis, normalized to tissue weight (E). n = 11 per group. (F) QPCR analysis of glycolytic and carbohydrate metabolism genes (Pfkl, Pkm, and Lctl) in interscapular BAT from sham and SCNx mice under Ad-STE and TRF-STE (n = 6 per group). (G) Expression of adrenergic receptors (Adrb3, Adrb2) in eWAT from sham and SCNx mice under Ad-STE and TRF-STE (n = 6 per group). (H and I) Western blot analysis of HSL and phosphor-HSL (HSL Ser660) in eWAT from sham and SCNx mice under Ad-STE and TRF-STE (H), with densitometry analysis (I). n = 4 per group. (J) Expression of Adrb3 in interscapular BAT from sham and SCNx mice under Ad-STE and TRF-STE (n = 6 per group). (K and L) Western blot (K) and densitometry analysis (L) of ADRB3 levels in interscapular BAT from sham and SCNx mice under Ad-STE and TRF-STE (n = 8 per group). (M) NE levels in interscapular BAT measured by ELISA from sham and SCNx mice at ZT16 under Ad-STE and TRF-STE (n = 6 per group), normalized to total protein. (N and O) Western blot analysis of HSL and phosphor-HSL (HSL Ser660) in interscapular BAT from sham and SCNx mice under Ad-STE and TRF-STE (N), with densitometry analysis (O). n = 4 per group. All data are presented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 as determined by unpaired two-tailed Student t test (B, C, E-G, I, J, L, M, and O). The data underlying the graphs shown in the figure can be found in S1 Source Data. Raw blot images can be found in S1 Raw Images.

https://doi.org/10.1371/journal.pbio.3003534.g002

To functionally validate this substrate switch, we performed extracellular flux analysis on ex vivo BAT explants. Upon glucose stimulation (10 mM), BAT from SCN-lesioned mice exhibited a significantly increased extracellular acidification rate (ECAR), indicating enhanced glycolytic flux (Fig 2D and 2E). These results are consistent with the elevated RER observed during TRF-STE (Fig 1M), further supporting a glucose-driven thermogenic program in the absence of SCN regulation.

Consistent with these functional changes, SCN-lesioned BAT showed upregulated expression of glycolytic enzymes such as Pfkl and Pkm, indicating enhanced glucose catabolism (Fig 2F). Additionally, Lctl, a gene involved in alternative carbohydrate metabolism (e.g., lactose and galactose metabolism) [31], was also upregulated (Fig 2F), suggesting broader transcriptional reprogramming of carbohydrate metabolism in response to SCN disruption.

Given the crucial role of β3-adrenergic signaling in the regulation of both thermogenesis and lipolysis [15,32,33], we assessed sympathetic signaling in WAT. In epididymal WAT (eWAT), expressions of Adrb2 and Adrb3 were significantly reduced in SCN-lesioned mice under TRF-STE (Fig 2G), consistent with indicative of suppressed sympathetic tone. Functionally, this was reflected by a failure to induce phosphorylation of HSL at Ser660, a key activation site downstream of extracellular signal-regulated kinase (ERK), suggesting impaired lipolysis in SCN-lesioned eWAT in response to TRF-STE (Figs 2H, 2I, and S2A).

Conversely, in BAT, both Adrb3 mRNA and ADRB3 protein levels were elevated at ZT4 and remained sustained at ZT16 under TRF-STE in SCN-lesioned mice, in contrast to their downregulation in sham controls (Fig 2J–2L). Norepinephrine (NE) levels in BAT were also significantly higher in SCN-lesioned mice (Fig 2M), suggesting preserved or even enhanced β3-adrenergic signaling in BAT under TRF-STE. However, despite increased Adrb3 expression, HSL phosphorylation remained low in SCN-lesioned BAT under TRE-STE (Fig 2N and 2O), indicating a disconnect between receptor abundance and downstream lipolytic activation and accessibility of fatty acids to mitochondrial oxidation. Consistent with this, genes involved in fatty acid mobilization and oxidation, such as Cyp2e1, Irf4, and Cyp4a10, were significantly downregulated in SCN-lesioned BAT (S2B Fig). H&E staining and Oil Red O staining further revealed the presence of larger lipid droplets in SCN-lesioned BAT under both Ad-STE and TRF-STE, consistent with impaired fatty acid utilization (S2C and S2D Fig).

Together, these findings demonstrate that SCN lesioning drives glucose-driven thermogenesis in BAT, while simultaneously impairing lipid mobilization in eWAT and BAT, likely due to disrupted sympathetic regulation. The apparent dissociation between β3-adrenergic receptor expression and lipolytic activation highlights compensatory metabolic reprogramming in BAT under SCN disruption, whereby it switches from using lipid to carbohydrate substrates to sustain thermogenesis.

SCN modulates thermogenic capacity via sympathetic signaling

To elucidate the anatomical and functional connection between the SCN and BAT, we performed anterograde tracing using HSV-eGFP injected into the SCN, which identified GFP-positive axonal terminals in interscapular BAT. Complementary retrograde tracing with PRV-CAG-GFP revealed SCN-positive neurons projecting to interscapular BAT, establishing an anatomical link between the SCN and BAT (Figs 3A, 3B, and S3A) [34].

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. SCN regulates thermogenic capacity via sympathetic signaling.

(A) Schematic representation of anterograde tracing and an image showing HSV-eGFP-labeled axonal projections from the SCN terminating in the interscapular BAT region. (B) Schematic and image showing the retrograde tracing from interscapular BAT to SCN, visualizing tracing signals in the SCN. (C and D) 3D whole-mount TH immunostaining (C) and quantification of sympathetic nerve fiber density (D) in interscapular BAT from sham and SCNx mice under TRF-STE from 10 regions (n = 1 per group). (E) Body temperature response to 6-OHDA treatment in sham and SCNx mice from Ad-STE to TRF-STE. Data are mean ± SEM, n = 6 per group, from the same experiments as interscapular BAT and tail temperature as S3F and S3G Fig. (F) Thermal response to SR59230A treatment in sham (n = 6) and SCNx mice (n = 5) from Ad-STE to TRF-STE. Data are mean ± SEM. From the same experiments as interscapular BAT and tail temperature as S3H and S3I Fig. SR: SR59230A. (G) Relative mRNA expression of Pgc1α and Ucp1 at ZT4 and ZT16 under Ad-STE and TRF-STE (n = 6 per group). (H and I) Western blot analysis of PGC1α and UCP1 levels in interscapular BAT from sham and SCNx mice (H), with densitometry analysis (I). n = 4 per group. (J) UMAP plot showing snRNA-seq data and clustering into 8 major cell types. (K) Expression of Pgc1α and Ucp1 within adipocyte clusters at ZT4 and ZT16 in sham vs. SCNx mice; exact p-values reported. (L) Electron microscopy images showing mitochondria in interscapular BAT from sham and SCNx mice under TRF-STE conditions. The red square highlights a magnified view of the mitochondrial structure. (M) Quantification of mitochondrial DNA content in interscapular BAT from sham and SCNx mice under TRF-STE (n = 6 per group). (N and O) Western blot analysis of the OXPHOS in interscapular BAT from sham and SCNx mice under TRF-STE conditions (N), with densitometry analysis (O). n = 4 per group. (P) OCR was measured in mitochondria isolated from interscapular BAT of sham and SCNx mice using Seahorse XF Analyzer. Measurements were taken under basal conditions and after sequential injection of oligomycin (ATP synthase inhibitor), FCCP (mitochondrial uncoupler), and a combination of rotenone/antimycin A (complex I/III inhibitors of the electron transport chain). Data are mean ± SEM, n = 5 per group. (Q) Quantification of basal respiration, ATP-linked respiration, and maximal respiration. n = 5 per group. Unless otherwise indicated, data are presented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 as determined by unpaired two-tailed Student t test (D, G, I, M, O, and Q), two-way ANOVA with Sidak’S multiple comparisons test (E and F). Scale bars, 400 μm (C), 200 μm (A, middle), 100 μm (B), 50 μm (bottom of A), 10 μm (magnified view of A), 2 μm (L, upper), 500 nm (L, bottom). Schematic (A, B) created in BioRender.com. The data underlying the graphs shown in the figure can be found in S1 Source Data. Raw blot images can be found in S1 Raw Images.

https://doi.org/10.1371/journal.pbio.3003534.g003

To investigate whether the SCN-mediated regulation of BAT thermogenesis operates via sympathetic innervation, we first examined tyrosine hydroxylase (TH), a marker of catecholaminergic sympathetic activity. Although total TH protein levels in BAT were unchanged between groups (S3B and S3C Fig), Adipo-Clear 3D imaging revealed a marked increase in sympathetic fiber density in SCN-lesioned BAT compared to sham controls (Fig 3C and 3D). To functionally validate the necessity of local sympathetic input, we selectively ablated catecholaminergic fibers in BAT using 6-hydroxydopamine (6-OHDA). Effective denervation was confirmed by the significant reduction in TH signal (S3D and S3E Fig). Under TRF-STE conditions, 6-OHDA treatment resulted in thermogenic responses that were indistinguishable between sham and SCN-lesioned mice, indicating that intact sympathetic innervation is required for SCN-dependent BAT activation (Figs 3E, S3F, and S3G). To further assess the downstream effector mechanisms, we pharmacologically blocked β3-adrenergic signaling using SR59230A, a selective ADRB3 antagonist. Following SR59230A administration, sham and SCN-lesioned mice exhibited comparable core and BAT temperatures under TRF-STE (Figs 3F, S3H, and S3I), supporting the conclusion that β3-adrenergic signaling is a critical mediator of SCN-dependent thermogenic regulation.

We next examined the expression of canonical thermogenic markers, Ppargc1a (Pgc1α) and Ucp1 in BAT. In sham mice, TRF-STE intensified the diurnal variation in these markers, with notably reduced expression at ZT16. In contrast, SCN-lesioned mice maintained expression of both markers at ZT16, with PGC1α showing a more robust and sustained response than UCP1 (Fig 3G–3I), suggesting a failure to suppress thermogenesis in the absence of SCN regulation.

In order to analyze the cellular heterogeneity and the dynamic expression characteristics of thermogenic genes (Pgc1α, Ucp1) in specific cell populations in BAT, we performed single-cell nuclear RNA sequencing (snRNA-seq) of BAT from sham and SCN-lesioned mice at ZT4 and ZT16 under TRF-STE conditions. Clustering analysis identified eight major clusters, ranging from 100 to 9,500 cells per cluster, with each expressing Pgc1α and Ucp1 (Figs 3J and S3J–S3L, S1, and S2 Tables). In brown adipocytes of SCN-lesioned mice, expression of both genes was slightly lower at ZT4 but significantly elevated at ZT16 compared to sham controls (Fig 3K). These findings confirm that SCN lesioning prevents the physiological TRF-STE-mediated suppression of Pgc1α and Ucp1 expression, so that defects in SCN results in sustained thermogenic activation in brown adipocytes.

SCN-lesioned BAT displayed increased mitochondrial density, enhanced lamellar cristae structure (Fig 3L), and a 1.5-fold increase in mitochondrial DNA (mtDNA) content (Fig 3M), along with upregulation of oxidative phosphorylation complex proteins (OXPHOS) (Fig 3N and 3O). To determine functional mitochondrial activity, we measured oxygen consumption rate (OCR) using Seahorse metabolic flux analysis. BAT mitochondria from SCN-lesioned mice showed elevated basal respiration, increased OCR in response to oligomycin and FCCP (Fig 3P and 3Q), indicating enhanced mitochondrial capacity and uncoupling potential under SCN disruption. To test whether this effect depended on sympathetic input, BAT was treated with 6-OHDA prior to respiration assays. Following treatment with 6-OHDA, OCR remained at comparable levels in both sham and SCN-lesioned mice (S3M and S3N Fig).

Collectively, these findings demonstrate that the SCN exerts critical control over BAT thermogenic capacity by modulating sympathetic outflow. Disrupted SCN results in activation β3-adrenergic signaling, and promotes mitochondrial biogenesis and function to prevent the normal adaptive decreased thermogenic response under TRF-STE conditions.

SCN lesioning reverses the TRF-STE induced BAT senescence and promotes cell proliferation via S100B

To investigate the molecular mechanism underlying SCN-regulated BAT thermogenesis, we performed RNA sequencing (RNA-seq) on interscapular BAT from sham and SCN-lesioned mice at ZT4 and ZT16 under Ad-STE and TRF-STE. Among the 25,240 genes analyzed, SCN-lesioned mice showed 77 upregulated and 24 downregulated genes at ZT4 and 316 upregulated and 274 downregulated genes at ZT16 compared to sham controls under TRF-STE (fold change > 2, p < 0.01) (Fig 4A). SCN lesioning significantly downregulated genes associated with fatty acid oxidation (Irf4, Cyp2e1, Cyp4a10, Cyp2b10, and Cyp2f2) and lipid transport (Mfsd2a, Slc7a8, and Slc43a1), while upregulating genes involved in lipogenesis and lipid storage (Amd1, Mest, Gdf15, Fgf18, Ptgds, and Hpgds) as well as carbohydrate metabolism (Pfkl, Pkm, and Lctl) (S4A Fig). These transcriptomic changes suggest impaired lipid mobilization alongside enhanced lipid accumulation on one side and increased glycolytic activity in BAT from SCN-lesioned mice, consistent with our earlier observations of decreased lipolysis and increased glucose utilization (Fig 2).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. SCN lesioning counteracts TRF-STE-induced BAT senescence via S100B.

(A) Volcano plot showing differentially expressed genes (DEGs) in interscapular BAT from sham and SCNx mice under TRF-STE. DEGs were defined by fold change > 2, p < 0.01. Commonly upregulated genes are highlighted in red and downregulated genes in green at both ZT4 and ZT16. n = 3 per group. (B) IPA of DEGs at ZT4 and ZT16. Pathways with significant enrichment (p < 0.01) are shown, with activation z-scores indicated for each pathway. (C) DAVID pathway analysis of S100 family-associated DEGs. Enriched KEGG pathways are shown in yellow, and Gene Ontology biological processes (BP) are shown in gray. (D and E) Correlation analysis of RNA-seq-derived S100b expression with cell proliferation markers (D) and senescence-related genes (E) in sham and SCNx mice. Pearson correlation coefficients and p-values are shown. (F and G) Cell proliferation analysis via EdU incorporation. Representative images showing EdU incorporation in interscapular BAT from sham and SCNx mice at ZT4 and ZT16 under Ad-STE and TRE-STE with 4°C and 30°C as control (F). Quantification of EdU⁺ nuclei as a percentage of total (Hoechst-stained) nuclei (G). n = 5 per group. (H–K) Assessment of S100B-induced proliferation in PDGFRα⁺ preadipocytes. PDGFRα-positive stromal vascular fraction (SVF) cells (H), representative images of EdU-positive cells treated with recombinant S100B (I), statistical analysis of EdU⁺ nuclei (J), and cell growth curves of control vs. S100B-treated preadipocytes (K). (L) Expression of differentiation-related genes during SVF from interscapular BAT induced to differentiate into mature adipocytes. n = 6 per group. (M) Representative β-galactosidase staining images showing senescent cell abundance in interscapular BAT from sham and SCNx mice under TRF-STE. Data are presented as mean ± SD. Statistical significance was determined using unpaired two-tailed Student t test (G, J, and L), Pearson correlation analysis (D and E), and two-way ANOVA with Sidak’S multiple comparisons test (K). *p < 0.05, **p < 0.01 and ****p < 0.0001. Scale bars, 100 μm (F and M), 50 μm (I), 20 μm (H and magnified view of I). The data underlying the graphs shown in the figure can be found in S1 Source Data.

https://doi.org/10.1371/journal.pbio.3003534.g004

Notably, SCN lesioning reversed the TRF-STE-mediated suppression of proliferation-related genes such as S100b, Ccnd1, Prelp, and Top2a, which were physiologically downregulated in sham mice at both time points (ZT4 and ZT16) (Fig 4A). Moreover, genes involved in methylation and detoxification (Ahcy, Gnmt, Pxmp2, and Cyp2e1) were consistently downregulated in SCN-lesioned mice (Figs 4A and S4B), suggesting a possible metabolic reallocation away from auxiliary pathways toward inappropriately activated thermogenic and proliferative programs.

Ingenuity Pathway Analysis (IPA) revealed significant enrichment of circadian rhythm, erythropoietin signaling, iron homeostasis, and S100 family signaling pathways, the latter of which showed significant activation (z-scores: −2.111 at ZT4, −2.143 at ZT16) in SCN-lesioned BAT (Fig 4B, S3 Table; IPA: p < 0.01). Complementary KEGG and Gene Ontology analyses showed that the S100 signaling intersects with GPCR signaling, angiogenesis, and cAMP/adrenergic pathways (Fig 4C), indicating a contributing role in adaptive thermogenic responses.

Under TRF-STE, sham mice displayed upregulation of senescence-associated genes such as Cdkn1a (p21) and Gadd45γ, especially at ZT16 following prolonged fasting (S4C Fig), suggesting an energy-conserving senescence response. In contrast, SCN lesioning suppressed this physiological senescence response and maintained elevated expression of S100b and Ccnd1, which were repressed in sham controls under TRF-STE (S4D Fig).

S100B, a calcium-binding protein with known roles in cell signaling, proliferation, and stress adaptation [28], was positively correlated with proliferation markers (Trp53, Mki67, Ccnd1, Top2a, Cd34, and Prelp) and inversely correlated with senescence markers (Cdkn1a and Gadd45γ) based on RNA-seq data (Fig 4D and 4E). In SCN-lesioned mice under TRF-STE, the increase in S100B was associated with enhanced expression of proliferation markers and suppression of senescence-associated genes (S4E and S4F Fig), suggesting that SCN activity modulates BAT cellular plasticity via S100B-dependent mechanisms.

To directly assess proliferation, 5-ethynyl-2′-deoxyuridine (EdU) labeling revealed significantly increased EdU⁺ nuclei in BAT from SCN-lesioned mice under both Ad-STE and TRF-STE conditions compared to sham controls (Fig 4F and 4G). Consistently, anti-PDGFRα immunostaining of BAT sections showed an increased number of PDGFRα⁺ preadipocytes in SCN-lesioned mice relative to sham mice under both Ad-STE and TRF-STE conditions (S4G and S4H Fig). In vitro, recombinant S100B promoted proliferation of PDGFRα⁺ preadipocytes (Fig 4H–4K) and accelerated early expression of brown adipocyte differentiation markers (Fig 4L), indicating dual roles in precursor expansion and thermogenic commitment. In parallel, senescence-associated β-galactosidase (SA-β-gal) staining showed a marked increase in senescent cells in sham BAT under TRF-STE, which was significantly reduced in SCN-lesioned BAT (Fig 4M). Together, these findings support a model in which SCN-regulated S100B expression protects BAT from stress-induced senescence and enhances its regenerative and thermogenic capacity under nutrient and thermal stress.

S100B functions as a nutrient-sensitive effector that amplifies β3-adrenergic signaling and thermogenic capacity in BAT

S100B emerged from our transcriptomic analyses as a potential effector of thermogenic plasticity under conditions of SCN inactivation and metabolic stress. To assess its responsiveness to energetic states, we measured S100b expression in interscapular BAT under diverse nutritional and thermal conditions. A 48-hour fast completely suppressed S100b, whereas HFD feeding significantly increased its expression (Fig 5A). Thermoneutral acclimation (30°C) markedly suppressed S100b, while cold exposure (4°C) failed to induce S100b, despite robust Ucp1 upregulation (Fig 5B). These findings demonstrate that S100b is regulated by nutrient availability rather than cold stress. Furthermore, serum concentrations of S100B, triglycerides (TG), and non-esterified fatty acids (NEFA) were unchanged between SCN-lesioned and sham mice (Fig 5C–5E), indicating that S100b regulation occurs in a tissue-autonomous manner within BAT, independent of systemic metabolic cues. This finding contrasts with prior reports associating circulating S100B with obesity and neuroinflammation [22] and underscores the localized nature of SCN-mediated control of S100B in BAT.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. S100B senses nutrients and amplifies β3-adrenergic signaling.

(A) S100b expression levels under various nutritional states (fasting, HFD, and TRF) and environmental temperatures (4°C, 21°C, and 30°C). n = 4 per group. (B) S100b and Ucp1 expression levels in indicated temperatures. n = 4 per group. (C) Serum S100B levels were measured by ELISA in sham and SCNx mice under Ad-STE and TRF-STE at ZT4 and ZT16. n = 4 per group. (D and E) Serum biochemical assays showing levels of TG (D) and NEFA (E) in sham and SCNx mice under Ad-STE (n = 6 per group). (F) Representative immunohistochemistry images of S100B protein in interscapular BAT from sham and SCNx mice under Ad-STE and TRF-STE conditions. Scale bars, 50μm. (G and H) Western blot analysis (G) and densitometric quantification (H) of S100B, p21, and CCND1 protein levels in interscapular BAT from sham and SCNx mice at ZT4 and ZT16. n = 4 per group. Red arrows indicate quantified bands. (I) Relative mRNA levels of indicated genes in interscapular BAT from sham and SCNx BAT following ADRB3 antagonist with SR59230A. n = 6 per group. SR: SR59230A. (J and K) Structural prediction of the human ADRB3-S100B complex using AlphaFold2/ColabFold. The human ADRB3/S100B complex model (J) was predicted by ColabFold and colored in blue, cyan, yellow, and orange according to different prediction confidence (pLDDT, predicted local distance difference test). Surface electrostatic representation showing ADRB3 (deep teal) and S100B (orange) with positive and negative charges indicated in blue and red, respectively (K). (L) coIP of ADRB3 and S100B in HEK293T cells. (M and N) Western blot (M) and densitometry analysis (N) of ADRB3 protein levels in interscapular BAT following S100b knockdown or overexpression in vivo. n = 4 per group. Red arrows indicate quantified bands. (O and P) ECAR analysis (O) and quantification of basal glycolysis and glycolytic capacity (P) in primary preadipocytes treated with S100B and/or β3-agonist CL-316243. Glucose, oligomycin, and 2-DG were sequentially injected. Data are presented as mean ± SEM. Control, CL, 0.1 µM S100B and 0.2 µM S100B: n = 6; 0.1 µM S100B + CL and 0.2 µM S100B+CL: n = 3. (Q) Schematic model illustrating the proposed ADRB3-S100B signaling axis. SCN lesioning enhances SNS activity, thereby promoting ADRB3 signaling and upregulating S100B expression. S100B, in turn, increases ADRB3 sensitivity, establishing a positive feedback loop that sustains thermogenesis and stimulates preadipocyte proliferation. Created in BioRender.com. Unless otherwise indicated, data are presented as mean ± SD. NS: not significant, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 by unpaired two-tailed Student t test (A–E, H, I, N, and P). The data underlying the graphs shown in the figure can be found in S1 Source Data. Raw blot images can be found in S1 Raw Images.

https://doi.org/10.1371/journal.pbio.3003534.g005

Histological analysis confirmed reduced S100B expression in BAT of sham mice under TRF-STE, whereas SCN-lesioned mice maintained high levels (Fig 5F). snRNA-seq confirmed this result, revealing elevated S100b expression specifically in brown adipocytes of SCN-lesioned mice at both ZT4 and ZT16, with no significant changes detected in other cell types (S5A Fig and S5B, S4 Table). Western blot analyses confirmed corresponding increases in S100B and CCND1 proteins, alongside decreased levels of the senescence marker p21 in SCN-lesioned BAT (Fig 5G and 5H).

To determine whether β3-adrenergic signaling regulates S100B expression and function, we treated mice with SR59230A. This intervention eliminated the expression differences in S100b, Ccnd1, and Cdkn1a between SCN-lesioned and sham mice (Fig 5I), indicating that ADRB3 signaling modulates the S100B-mediated balance between proliferation and senescence. These results are consistent with our earlier observations showing that SR59230A inhibits SCN-lesion-induced BAT thermogenesis (Fig 3F).

To evaluate the possibility of a direct molecular interaction between S100B and ADRB3, we employed ColabFold protein modeling, which predicted a high-confidence interaction localized to the intracellular domain of ADRB3 (Fig 5J and 5K). This interaction was experimentally validated by co-immunoprecipitation (coIP), confirming physical association between the two proteins (Fig 5L). Modulation of S100b expression in BAT altered ADRB3 protein abundance and electrophoretic mobility (Fig 5M and 5N), though this was not attributable to differences in protein degradation (S5C Fig), suggesting that S100B enhances receptor sensitivity rather than stability, as previously proposed [35,36].

To assess the functional consequences of this interaction, we treated primary brown adipocytes (PPDIVs) with recombinant S100B. This significantly increased ECAR, indicative of enhanced glycolytic metabolism (Fig 5O and 5P). Co-treatment with the β3-agonist CL-316243 synergistically amplified ECAR, indicating that S100B potentiates β3-adrenergic signaling to promote glucose-dependent thermogenesis (Fig 5O and 5P).

Together, these results position S100B as a nutrient-sensitive, SCN-regulated effector of BAT thermogenic plasticity. Acting through the ADRB3-S100B axis, SCN signaling integrates sympathetic tone with transcriptional and metabolic programs that enhance glycolysis, stimulate preadipocyte proliferation, and repress senescence. This positive feedback loop supports sustained BAT thermogenesis under combined nutritional and thermal stress (TRF-STE), and highlights S100B as a critical molecular node in SCN-mediated metabolic adaptation (Fig 5Q).

S100B is both necessary and sufficient to mediate SCN-dependent thermogenic resilience in BAT

To determine whether S100B is functionally required for SCN-mediated thermogenesis, we selectively knocked down S100b in the BAT of SCN-lesioned mice using AAV-U6-shRNA(S100b)-EGFP-P2A, with a scrambled shRNA vector as control (Fig 6A). Knockdown of S100b significantly increased expression of the senescence marker Cdkn1a (p21) and decreased expression of the proliferation-associated gene Ccnd1 (Fig 6B–6D). Moreover, EdU incorporation assays showed significantly reduced BAT cell proliferation upon S100b knockdown (Fig 6E and 6F), confirming S100B’s essential role in supporting BAT cellular turnover. Functionally, this was accompanied by a significant drop in body temperature during TRF-STE, reversing the thermoprotective phenotype conferred by SCN lesioning (Figs 6G and S6A).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. SCN-mediated BAT thermogenesis via ADRB-S100B axis in response to TRF-STE.

(A) Schematic of AAV-mediated S100b knockdown or scramble control delivery into interscapular BAT of SCNx mice. (B) Relative mRNA expression of the indicated genes in interscapular BAT from SCNx mice under TRF-STE with scramble or S100b targeting shRNA. n = 6 per group. (C and D) Western blot analysis of S100B and p21 (C) and corresponding densitometry quantification (D) in interscapular BAT from SCNx mice under TRF-STE (n = 4 per group). (E and F) EdU incorporation analysis in interscapular BAT from SCNx mice following S100b knockdown or scramble control. Representative images (E) and quantification of EdU-positive nuclei relative to total Hoechst-positive nuclei (F). (G) Body temperature profiles of SCNx mice with S100b knockdown or scramble control under Ad-STE and TRF-STE (scramble: n = 6, S100b knockdown: n = 5). Data presented as mean ± SEM. Corresponding interscapular BAT and tail temperature data are shown in S6A Fig. (H) Schematic illustrating AAV-mediated overexpression of S100b (AAV-DIO-S100b + AAV-CAG-Cre) or EGFP control in interscapular BAT of WT mice. (I) Relative mRNA levels of indicated genes in interscapular BAT from EGFP or S100b overexpressing mice under TRF-STE (n = 6 per group). (J and K) Western blot analysis of S100B and p21 proteins (J) and corresponding densitometry (K) from interscapular BAT of EGFP or S100b-overexpressing mice (n = 4 per group). (L and M) EdU incorporation analysis in interscapular BAT from EGFP and S100b-overexpressing mice. Representative images (L) and quantification of EdU⁺ cells (M). n = 6 per group. (N) Body temperature responses under Ad-STE and TRF-STE in WT mice with EGFP or S100b overexpression in interscapular BAT. Data presented as mean ± SEM. n = 6 per group. See S6B Fig for interscapular BAT and tail temperature profiles. Unless otherwise indicated, data are presented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001. Significance determined by unpaired two-tailed Student t test (B, D, F, I, K, and M) or two-way ANOVA with Sidak’S multiple comparisons test (G, N). Scale bars, 50 μm (E, L), 20 μm (magnified view of E, L). Schematic (A, H) created in BioRender.com. The data underlying the graphs shown in the figure can be found in S1 Source Data. Raw blot images can be found in S1 Raw Images.

https://doi.org/10.1371/journal.pbio.3003534.g006

To assess sufficiency, we overexpressed S100b in BAT via Cre-dependent AAV-CMV-DIO-S100b co-injected with AAV-CAG-Cre-EGFP (Fig 6H). S100b overexpression downregulated p21 and Gadd45γ, reduced p21 protein levels, and upregulated Ccnd1 (Fig 6I–6K). These molecular changes translated into significantly increased BAT cell proliferation (Fig 6L and 6M) and resistance to TRF-STE-induced hypothermia (Figs 6N and S6B). Together, these data demonstrate that S100B is sufficient to drive preadipocyte proliferation and sustain thermogenesis under stress conditions.

As S100b expression in BAT is regulated by SCN activity, we next tested whether diverse modes of SCN disruption converge on similar outcomes. First, constant light exposure (LL), a noninvasive method known to induce behavioral arrhythmia and circadian disruption [37], was used as a physiological model of SCN impairment (Fig 7A). LL-exposed mice showed disrupted locomotor activity (S7A Fig) and a BAT molecular phenotype resembling SCN-lesioned mice, including elevated S100b and Ccnd1, and reduced p21 (Fig 7B–7D), enhanced EdU incorporation (Fig 7E and 7F), and attenuated TRF-STE-induced hypothermia (Figs 7G and S7B).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 7. Multiple complementary models of SCN dysfunction in response to TRF-STE.

(A) LL exposure paradigm is used to disrupt SCN rhythmicity. Arrhythmic mice were selected for subsequent analysis (S7A Fig). (B) Relative mRNA expression of indicated genes in interscapular BAT from WT mice under light/dark cycle (LD) or LL during TRF-STE (n = 4 per group). (C and D) Western blot analysis (C) and densitometry (D) of S100B and p21 proteins in interscapular BAT from mice under LD or LL conditions (n = 4 per group). (E and F) EdU staining in interscapular BAT from mice under LD or LL. Representative images (E) and quantification of EdU⁺ cells (F) normalized to Hoechst staining. (G) Body temperature profiles of WT mice under LD or LL conditions during TRF-STE. Data presented as mean ± SEM. n = 4 per group. See S7B Fig for additional temperature metrics. (H) Working model illustrating the SCN-ADRB3-S100B axis in BAT during TRF conducted at ZT16-ZT20 in a subthermoneutral environment, the SCN regulates BAT thermogenic plasticity via SNS output. SCN lesioning enhances ADRB3 signaling and S100B expression, which together form a positive feedback loop that amplifies β3-adrenergic sensitivity, promotes preadipocyte proliferation, suppresses senescence, and sustains glucose-driven thermogenesis. Unless otherwise indicated, data are presented as mean ± SD. *p < 0.05, **p < 0.01 and ****p < 0.0001. Significance determined by unpaired two-tailed Student t test (B, D, and F) or two-way ANOVA with Sidak’S multiple comparisons test (G). Scale bars, 50 μm (E), 20 μm (magnified view of E). Schematic (A) and working model (H) created in BioRender.com. The data underlying the graphs shown in the figure can be found in S1 Source Data. Raw blot images can be found in S1 Raw Images.

https://doi.org/10.1371/journal.pbio.3003534.g007

To further confirm SCN regulation of this axis, we used two SCN-targeted lesion models: [1] NMDA-induced excitotoxicity, which causes focal damage to SCN neurons through glutamatergic overactivation (S7C Fig) [38]; and GABAergic neuronal ablation via AAV-DIO-Caspase3 delivered with AAV-VGAT1-Cre (S7D Fig). In both models, BAT exhibited elevated S100b and Ccnd1 and suppressed p21 expression, though to a lesser extent than in SCN lesioning, reinforcing that S100B activation is a conserved feature across diverse modes of SCN disruption (S7E and S7F Fig).

In summary, these findings establish that the SCN regulates BAT thermogenesis via a β3-adrenergic-S100B signaling axis. Across multiple independent models, SCN lesioning, light-induced arrhythmia, excitotoxicity, or targeted neuronal apoptosis, converge on S100B activation, promoting preadipocyte proliferation, suppressing senescence, and sustaining thermogenesis under thermal and nutritional stress. These results identify S100B as a critical downstream effector of SCN-driven sympathetic output, linking circadian control with metabolic resilience in adipose tissue.

Animal and ethics statement

All animal experiments were conducted in accordance with the Regulation for the Administration of Affairs Concerning Experimental Animals of the People’s Republic of China and the national standard Laboratory Animals: Guideline for Ethical Review of Animal Welfare (GB/T 35892-2018). All procedures were reviewed and approved by the Animal Care and Use Committee of Soochow University under protocol numbers YX-19-3 and YX-24-02.

C57BL/6J mice were purchased from GemPharmatech Co. (Nanjing, China) and maintained in a pathogen-free (SPF) animal facilities under a 12-hour light/12-hour dark cycle at controlled temperature and humidity. Animals had ad libitum access to food and water unless subjected to TRF protocols as indicated. To induce obesity, 8-week-old C57BL/6J male mice were fed a HFD containing 60% fat, 20% carbohydrate, and 20% protein (#XTHF60, Xietong, China) for 7 weeks. Body weight and body fat composition were monitored and recorded.

Cell line

HEK293T cells are cultured in a high-glucose DMEM medium containing 10% fetal bovine serum and penicillin-streptomycin. Depending on the experimental requirements, cells are cultured in 10 cm, 35 mm, or 4-chamber coverslips. HEK293T cells are transfected with plasmids using LipoMax Transfection (#32012, SUDGEN, China) according to the manufacturer’s instructions. Observe fluorescence and capture under a confocal fluorescence microscope (TCS SP8, Leica) 24 h after transfection. Alternatively, cells can be collected for coIP or protein stability assay.

SCN-lesioned mice

Eight-week-old male C57BL/6J mice were anesthetized with isoflurane and maintained on a heating pad to preserve CBT throughout the procedure. After fixation in a stereotaxic apparatus, a 0.5 mm diameter burr hole was drilled into the skull, and a 0.15 mm diameter platinum-iridium electrode (custom-fabricated by Suzhou Kedo Brain Technology Co.) was stereotactically lowered into the SCN. The electrode was polyimide-coated except for the tip. The stereotaxic coordinates used to target the SCN were: anteroposterior (AP): −0.05 mm, mediolateral (ML): ± 0.15 mm, dorsoventral (DV): −5.85 mm from the bregma. A lesion was induced by passing a 0.4 mA direct current for 30 s using a Ugo Basile lesioning device (Model 53500). Sham-operated mice underwent identical procedures except that electrodes were inserted only to 5.4 mm DV depth, and no current was applied. After 1 week of postoperative recovery, all mice were placed in wheel-running cages under a 12-hour light/12-hour dark (LD) cycle for 1 week to assess circadian locomotor activity. Periodogram analysis showed that all sham mice retained robust circadian rhythms, whereas approximately 30% of SCNx mice became arrhythmic (S1A Fig). For downstream physiological and molecular experiments, only SCN-lesioned mice that completely lost circadian rhythmicity and had comparable average body weight and size to sham controls were included. At the conclusion of the experiments, histological analysis was performed to verify the accuracy and extent of SCN lesions [14]. For excitotoxic lesion experiments, 30 nl of NMDA (50 mM) was bilaterally injected into the SCN using the same stereotaxic coordinates. Two weeks later, wheel-running assays were performed to confirm circadian disruption, followed by GFAP immunostaining to assess gliosis and lesion specificity. For the viral-mediated lesioning, AAV2/9-DIO-Casp3 (5.49 × 10¹² vg/ml) and AAV2/9-VGAT1-Cre (3.2 × 10¹² vg/ml) were mixed in a ratio of 2:1, and a total of 150 nl was bilaterally injected into the SCN. The control group received AAV2/9-DIO-EGFP (5.21 × 10¹² vg/ml) co-injected with AAV2/9-VGAT1-Cre. Behavioral circadian rhythm was assessed via wheel-running activity 2 weeks postinjection.

Mice locomotor analyses

After surgery, the mice were housed individually in cages with running wheels, had free access to food and water, and were maintained under a 12-hour light-dark cycle. The locomotor rhythms were recorded using Clocklab (V6.1.02) with continuous recording for 1 week. Arrhythmic mice were selected for subsequent manipulations.

Constant light treatment mice

Eight-week-old male mice were individually housed in boxes that could regulate the light environment. The experimental group of mice was exposed to constant light, while the control group was maintained on a 12-hour light-dark cycle, with a light intensity of 200 lux for both groups. This treatment lasted for at least 2 months, and the mice’s free wheel-running activity was recorded.

Different temperatures and nutritional conditions

Eight-week-old male mice were individually housed in an incubator (PGX-350B, Ningbo Saifu) with adjustable temperatures. After a one-week acclimation period at room temperature, they were placed under different temperature conditions or given various dietary patterns (chow diet, ZT16-20 TRF, fasting, or HFD). After a 48-hour treatment, the mice were euthanized, and their interscapular BAT was rapidly collected.

TRF-STE

SCN-lesioned mice and sham mice were individually housed in cages placed in an incubator with adjustable temperature and light. The temperature control of the incubator is based on a high-precision temperature control module, ensuring variations of less than 1°C. Each layer of the incubator is checked with a mercury-in-glass thermometer to ensure that the temperature of each layer is uniform. The light schedule is 12 h of light (light intensity is approximately 200 lux) followed by 12 h of darkness. The mice were first given ad libitum at 21°C for 2 weeks to acclimate to the incubator. Then they were then placed in new, clean cages, and the food was removed at ZT20. Each cage was covered with wood shavings to provide a small nest. The mice were given food on the following day (ZT16), which was the first day of the TRF-STE protocol. Calculate and record the amount of food eaten by the mice each day.

Infrared thermograph and core body temperature recording

The backs of the mice were shaved 2 days before recording, and a Thermochip (SureFlap Ltd) was injected into the abdominal cavity of the mice to detect CBT. Changes in CBT of mice were detected using a thermochip monitoring device (GPR+, Destron Fearing). An infrared imager (Flir E8) was used to detect the surface temperature of the BAT of the mice from a top view. The camera was kept at the same height for each detection, and the images were automatically synthesized by the device and processed using Flir Tools software.

Indirect calorimetry

Mice were individually acclimated for 3 days in metabolic cages of a comprehensive animal monitoring system (Oxymax, Columbus Instruments) and provided with food and water ad libitum. The room lighting followed a 12-hour light/12-hour dark cycle, and the temperature was freely adjusted to meet the different experimental needs. For TRF mice, food was provided 4 h after the lights were turned off (ZT16), and food was removed 8 h after the lights were turned off (ZT20). O₂, CO₂, RER (CO₂/O₂), EE, and locomotor activity were recorded throughout the experiment.

Interscapular BAT adeno-associated virus (AAV) injection

After being anesthetized with isoflurane, the fur over the scapular region of SCN-lesioned or 8-week-old male C57BL/6J mice was shaved, and the skin was disinfected with 75% alcohol. An incision was made to expose the interscapular BAT pads bilaterally, and a total of 1 × 10¹² vg/ml AAV was injected into multiple sites of each pad using a microsyringe. The mice were then allowed to recover for 1 week before entering the TRF-STE protocol. The following viral vectors were used: AAV2/8-U6-shS100b-EGFP (5.08 × 10¹² vg/ml) and AAV2/8-U6-scramble RNA-EGFP (5.08 × 10¹² vg/ml), which were purchased from BrainVTA (China); and AAV2/8-CMV-DIO-S100b-EGFP (4.3 × 10¹² vg/ml) and AAV2/8-CMV-DIO-EGFP (4.5 × 10¹² vg/ml), which were obtained from Obiosh (China).

Anterograde, retrograde tracing, and virus

The retrograde tracing viral tracer used in this study was the pseudorabies virus (PRV-CAG-EGFP). After the 8-week-old male C57BL/6J mice were anesthetized with isoflurane, an incision was made in the skin at the site of the interscapular BAT to expose the fat pads on both sides. A total of 3 μl of the virus was injected into both sides of the interscapular BAT at multiple points using a microsyringe (Size: 2 μl, RWD). The Brain slices were removed for observation on days 3, 4, and 5.

The anterograde virus used in this study was the herpes simplex virus (HSV-EGFP). After the mice were anesthetized, 200 nl of HSV was administered to the unilateral SCN. After seven days, the interscapular BAT was removed, fixed, sliced, and observed.

Mouse PET scans and gamma counting

PET/CT (SNPC-304, Pingseng Technology) scans, and gamma counting (WIZARD 2,480, PerkinElmer) mouse scans were performed at the First Affiliated Hospital of Soochow University. On the second day of TRF-STE, the mice were transported using an onboard refrigerator to facilitate temperature control of the environment. After the mice were anesthetized with isoflurane, ¹⁸F-FDG was injected into the tail vein. Sixty minutes later, the mice were scanned with PET/CT, and the mice were always anesthetized. At the end of the scan, the mice were killed by cervical dislocation, and organs were rapidly harvested and weighed for the gamma counter to quantify radioactivity, which was normalized based on organ weight and dose at the time of the scan. The percentage of the injected dose per gram was calculated for the tissues.

SnRNA-seq

SnRNA-seq and analysis of interscapular BAT were performed by Shanghai OB Biotech Co., at ZT16 on the second day of TRF-STE. Mice were euthanized, and interscapular BAT was rapidly dissected. Each group of samples consists of interscapular BAT from five mice. These tissues were processed into a cell suspension. The cell suspension was filtered twice through a 40 µm cell strainer (#352340, BD), centrifuged at 300 rpm for 5 min at 4°C, and the pellet was resuspended after the addition of red blood cell lysis buffer (#130-094-183, MACS). After standing for 10 min, centrifuge at 300 rpm for another 5 min, remove the supernatant, resuspend the cell pellet, and count the cells (Luna, LUNA-II). The cell concentration was then adjusted to 1,000 cells/µl.

Single-cell libraries were prepared for all samples using 10× Genomics Chromium Next GEM Single Cell 3′ Reagent Kits V3.1 (1000268, Chromium). The well-constructed library was subjected to high-throughput sequencing on the Illumina NovaSeq 6000 PE150 platform. Based on a preliminary quality control using Cell Ranger (V7.1.0), further quality control of experimental data was performed to exclude data from multiples, doublets, or unbound cells. This is followed by downstream analysis. In this project, we used PCA (Principal Components Analysis) and UMAP (Uniform Manifold Approximation and Projection) algorithms for dimensionality reduction. Based on the results of PCA, UMAP was used to visualize the clustering of single-cell populations. The clustering algorithm used was SNN (Shared Nearest Neighbors), which ultimately resulted in the optimal cell grouping. The Presto test method is used to perform differential expression analysis between the specified cell population and all other cell populations. Filtering criteria include a log fold change > 0 and a minimum percentage (min.pct) of gene expression across all cells > 0.25. This process helps to identify marker genes for each cell population. The top 10 marker genes are ranked from highest to lowest based on Gene Diff values. Pct1 represents the proportion of cells expressing the marker gene within the current group of cells. Pct2 represents the proportion of cells expressing the marker gene in the remaining groups. The ratio of these two percentages is the value of gene-diff (S1 Table).

RNA-seq

At ZT4 and ZT16 on the second day of TRF-STE, mice were euthanized by cervical dislocation, and interscapular BAT was collected and stored in liquid nitrogen. Total RNA from interscapular BAT was extracted using Trizol reagent (Ambion) and chloroform, and then the concentration was detected using NanoDrop 2000 (Thermo Fisher). RNA-seq was performed by Genewiz (Suzhou, China). First, the total RNA was quality-controlled using Agilent 2200 tapstation, then RNA-seq was performed on an Illumina NovaSeq 6000 platform with PE 150-bp reads, and sequencing was performed using Illumina bcl2fastq. Raw binary base calling data were converted to sequence data. Sequencing data were compared to the mm10 reference genome at UCSC using Bowtie (V1.1.2). Finally, FPKM values of the genes in each sample were calculated using RSEM (V1.2.8). Screening for differentially expressed genes was based on fold change > 2, p < 0.01.

Ingenuity Pathway Analysis (IPA)

Pathway enrichment for the differentially expressed genes from RNA-seq was analyzed by IPA (PMID: 24336805) (p < 0.01). The highest-ranking canonical pathways were selected, focusing on those with significant activation z-scores at both ZT4 and ZT16 (absolute [Z score] < −2). The data is presented in S3 Table.

The Alphafold2 model of the ADRB3/S100B complex

Five models of the ADRB3/S100B complex were generated by ColabFold 1.5.5 in the local computer (https://github.com/YoshitakaMo/localcolabfold) using human ADRB3 (Uniprot ID P13945) and S100B (Uniprot ID P04271) sequences, and both pdb templates and amber refinement were used during modeling. The top-one-ranked model was chosen for the structure analysis using PyMOL (https://pymol.org/).

EDU labeling in vivo

On the day before TRF-STE, mice were injected intraperitoneally with EdU (25 μg/g body weight, #900584, Sigma), with injections every 8 h for a continuous period of 3 days and a final injection 2 h before euthanasia. The mouse interscapular BAT was collected and preserved in 4% PFA. This was then cut into 7 μm sections (RM2106, Leica) and stained using the Click-iT Edu Imaging Kit (C10340-red or C10337-green, Invitrogen). After washing with PBS, the sections were immersed in a DAPI staining solution (1:5000 in PBS) for 10 min at room temperature. After the slides were sealed with 50% glycerol (Sangon, China), they were observed and photographed using an inverted fluorescence microscope (IX53, OLYMPUS) or a confocal microscope (TCS SP8, Leica), and the EdU-positive cells were counted within the same field of view, and the ratio to the total number of cells was calculated.

Protein stability assay

Transfect HEK293T cells with the expression plasmid. After 24 h of transfection, treat the cells with the protein synthesis inhibitor cycloheximide (CHX, MCE) at a concentration of 20 µg/ml. The cell suspension is then collected at the appropriate time and subjected to western blot analysis.

Stromal vascular fraction (SVF) cell culture

Brown adipose tissue SVF was isolated from 6-week-old C57BL/6J male mice. Mice were killed by cervical dislocation, and the interscapular BAT was dissected out and rinsed three times with pre-cooled PBS at 4°C. After mincing, the tissue was treated with type I collagenase (SCR103, Sigma) at 37°C for 40 min. After centrifugation at 1,200g for 2 s, the upper layer of the cell suspension was collected, followed by centrifugation at 1,700g for 5 min to collect the lower layer of pelleted cells. The cells are then resuspended in a high-glucose DMEM medium (containing 20% fetal bovine serum and penicillin-streptomycin), plated into 35 mm dishes, and cultured at 37°C with 5% CO₂.

For EdU-labeled cell proliferation of SVF, 0.1 µM recombinant S100B protein (TMPJ-00990, Targetmol USA) or saline and 10 µM EdU were added to the medium. After 48 h, cells were washed three times with PBS, fixed with 4%PFA for 10 min, and visualized using the Click-iT Edu Imaging Kit. The nucleus was labeled with DAPI for 10 min at room temperature.

For SVF immunofluorescence, after the cells were fixed with 4%PFA, they were blocked with 5% normal donkey serum for 1 h at room temperature. Anti-PDGFRα antibody (1:500, Cell Signaling Technology, 3174S) was then added and incubated overnight at 4°C. The next day, after washing three times with PBS, they were incubated with the secondary antibody (goat anti-rabbit, Alexa Fluor 647) for 1 h at room temperature in the dark. After nuclear labeling with DAPI, the cells were imaged by confocal fluorescence microscopy.

For the analysis of cell growth curves, after the addition of 0.1 µM S100B recombinant protein, the cell culture plate was placed in a fully automated live cell dynamic imaging system (Incucyte SX5, Sartorius), and the cell count was recorded every hour for a continuous period of 72 h. Seahorse extracellular flux analysis for glycolytic stress test.

SVF was isolated from the interscapular BAT of 6-week-old C57BL/6J male mice. SVF was added to Seahorse XF24 cell culture microplates to induce differentiation. Change the growth medium every other day until cells reach 80%−100% confluence. The growth medium was replaced with an induction medium containing 0.64 μM dexamethasone, 0.625 mM isobutylmethylxanthine, 8.4 μM rosiglitazone, and 1.6 μg/ml insulin. The induction medium was changed every other day and replaced with a differentiation medium (composition is the same as the induction medium except for the addition of 1.63 μM 3, 3′, 5-triodo-L-thyronine sodium salt). The differentiation medium was changed every other day, and the glycolytic stress test was performed. Glycolysis stress test: The Agilent Seahorse XF Glycolysis Stress Test Kit was used according to the standard Agilent Seahorse protocol. Prior to the glycolytic stress test, the SVF medium was pretreated with 0.1 μM/0.2 μM recombinant S100B protein and/or 0.1 μM ADRB3 agonist CL316243 for 18 h. After six consecutive measurements of basal ECAR, 10 mM glucose, 2 μM oligomycin, and 50 mM 2-DG were sequentially injected for three ECAR measurements. Wave (Agilent Technologies V2.6.4) software was used to analyze the parameters.

SVF induction differentiation

SVF was isolated from the interscapular BAT of 6-week-old C57BL/6J male mice. First, SVF was cultured using a growth medium. After the cells became fully confluent (24 h later), the growth medium was replaced with an induction medium. The experimental group was additionally supplemented with 0.1 µM recombinant S100B protein, while the control group received only fresh medium. The differentiation medium was refreshed every 2 days. Cells were collected on days 0, 3, and 5 of induction. RNA was extracted from the collected cells, reverse transcribed, and used for qPCR to detect the expression of adipocyte differentiation-related genes.

Measurement of mitochondrial oxidation

The OCR was measured using a microfluorimetric Seahorse XF24 Analyzer (Agilent Technologies). Mice that had undergone 2 days of TRF were sacrificed, and their interscapular BAT was quickly removed. Mitochondria were isolated from the tissue using MIB (250 mM sucrose, 5 mM HEPES, 2 mM EGTA, and 1% BAS, PH = 7.2) as follows: the tissue was homogenized in MIB, then centrifuged at 800g to remove debris, and finally centrifuged at 10,000g to collect the mitochondria. The mitochondrial pellet was resuspended in MAS (155 mM KCl, 10 mM KH₂PO₄, 2 mM MgCl₂, 5 mM HEPES, 1 mM EGTA, plus 0.1% BSA). The assay solution also contained 10 mM pyruvate, 5 mM malate, and 1 mM GDP, with the pH adjusted to 7.2. The assay protocol was as follows: Port A: 5 mM ADP; Port B: 4 µM oligomycin; Port C: 5 µM FCCP; Port D: 5 µM antimycin A and 5 µM rotenone. Wave (Agilent Technologies V2.6.4) software was used to analyze the parameters.

Seahorse analysis of ex vivo BAT

On day 2 of TRF-STE, at ZT16, mice were euthanized, and the interscapular BAT was quickly removed, rinsed with PBS, weighed at 5 mg, and placed into the Seahorse XFe24 Islet Capture FluxPak, and then minced. Each well was filled with 500 μl of XF DMEM Medium (with 2 mM glutamine) and incubated at 37°C for 50 min. The tissue’s ECAR was detected using the microfluorimetric Seahorse XF24 Analyzer, with 10 mm glucose added during the detection. Wave (Agilent Technologies V2.6.4) software was used to analyze the parameters.

Whole-mount TH immunostaining and 3D imaging

On the second day of TRF-STE at ZT16, mice were anesthetized with pentobarbital sodium and slowly perfused with PBS, followed by 4% PFA. Interscapular BAT was excised, fixed in 4% PFA at 4°C overnight, and washed twice with PBS the next day. The fixed tissue was cleared using the NSⅢ Tissue Clearing Kit (#240920-12, Nuohai Life Science, Shanghai) according to the manufacturer’s instructions. Cleared samples were incubated in primary antibody solution (Abcam, AB152) at 37°C with gentle shaking for 15 days, washed with PBS, and then immersed in secondary antibody solution (Thermo Fisher Scientific, A32733) at 37°C for 10 days. After refractive-index matching, 3D fluorescence imaging was performed on a Nuohai LS-18 tiling light-sheet microscope, and images were rendered in Amira (Thermo Fisher Scientific). Nerve-fiber signals were segmented and traced using the segmentation and filament modules of Amira, and final data were visualized with Amira’s rendering tools.

Immunohistochemistry for interscapular BAT sections

After the mice were anaesthetized by intraperitoneal injection of sodium pentobarbital (4% in saline), the mice were perfused with PBS and 4% PFA. Mouse interscapular BAT was collected and processed into 7 μm paraffin sections. The sections were then deparaffinized and antigen retrieved with a sodium citrate antigen retrieval solution (pH 6.0). After washing with PBS, endogenous peroxidase was removed with 3% H₂O₂, the sections were incubated with 0.3% TritonX-100 (Sigma) for 20 min at room temperature, washed in PBS and incubated with 10% (in PBS) normal goat serum for 1 h at room temperature, and the primary antibody rabbit anti-S100B (1:200,380829, ZEN-BIO) or PDGFRα (1:500, Cell Signaling Technology, 3174S) was diluted in blocking solution and incubated overnight at 4°C. The next day, the primary antibody was washed off with PBS and incubated with horseradish peroxidase-conjugated secondary antibody for 1 h. The final sections were stained with a DAB staining solution mixture (KIT-5920, MXB biotechnologies China) for visualization, and hematoxylin was used to label the nucleus.

Western blotting and quantification

Protease and phosphatase inhibitors (Roche) were added to the RIPA lysis buffer. The mouse interscapular BAT was homogenized thoroughly in the RIPA buffer, then centrifuged at 12,000g for 15 min at 4°C, and the supernatant was collected. The protein was quantified using the bicinchoninic acid assay (P0011, Beyotime China), and the total protein was diluted to a concentration of 1 μg/μl. SDS-PAGE gels of different concentrations were used to quantify the corresponding proteins. The dilution ratios for the primary antibodies were as follows: HSP90 (1:1000, Proteintech, 13171-1-AP), UCP1 (1:1000, Proteintech, 23673-1-AP), HSL (1:1000, Cell Signaling Technology, 4107T), p-HSL (Ser660) (1:1000, Cell Signaling Technology, 45804S), Total OXPHOS complex (6 μg/ml, Abcam, ab110413), PGC1α (1:1000, Santa Cruz, sc-13067), S100B (1:500, 380829, ZEN-BIO), p21 (1:1000, Proteintech, 28248-1-AP), ADRB3 (1:1000, Bioss, bs-1063R), CCND1 (1:5000, Proteintech, 26939-1-AP). Immerse the PVDF membrane (Millipore) in the diluted primary antibodies and incubate overnight. After three washes in TBST, the membrane was incubated with peroxidase-conjugated secondary antibodies (anti-rabbit/mouse IgG H + L HRP-linked, 7074s/7076s, Cell Signaling Technology) for 1 h at room temperature. Finally, the membranes were imaged using an automatic chemiluminescence imaging system (ChemiScope S6, Clinx Shanghai) and quantified using ImageJ (National Institutes of Health).

Quantitative RT-PCR

The total RNA was measured using the NanoDrop 2000 (Thermo Fisher). The PrimeScript RT Reagent Kit (RR037A, Takara) was used for genomic DNA isolation and reverse transcription to obtain cDNA. To quantify mDNA, a DNA Isolation Kit (DC102, Vazyme) was used to extract the genomic DNA from interscapular BAT according to the manufacturer’s instructions, and quantitative PCR was performed using SYBR Green (RR420, Takara) with the ViiA7 (Life Technologies). The relative expression levels of genes were calculated using the 2^ΔΔCT (cycle threshold) method, and the final results were normalized to the expression level of actin.

ELISA

Mice were killed by cervical dislocation, fresh blood was collected, centrifuged at 3000g for 10 min at room temperature, and the serum was collected and stored in an ultra-low temperature freezer. The mouse S100B ELISA kit (#ED-20090, Lunch Chang Shuo Biotech, China) or NE ELISA kit (#E-EL-0047, Elabscience) was used according to the manufacturer’s instructions to detect the level of S100B or NE in each serum sample.

Denervation of the sympathetic nervous system

6-OHDA (10 mg/ml, #H4381, Sigma) was dissolved in 0.15 mol/l saline containing 1% ascorbic acid. A midline skin incision was then made along the upper dorsal surface to expose both interscapular BAT pads, and 10 μg of 6-OHDA was injected into each pad using a 10 μl microsyringe. Perform TRF-STE one week after injection.

Quantification and statistical analysis

All experimental data were first tested for normal distribution using the Shapiro–Wilk test in GraphPad Prism V8.0.2 (GraphPad, USA). Subsequently, Student t tests or two-way ANOVAs were performed based on experimental needs, while GLM for mouse EE was calculated using SPSS Statistics V25.0 (IBM, USA), with mouse body weight as the covariate and EE as the dependent variable. Statistical details of the experiments were provided in the legends and corresponding methods sections. P-values are indicated in the Figs. Statistical significance is indicated as p < 0.05 (*p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001), NS: not significant.