Brown-fat-mediated tumour suppression by cold-altered global metabolism

The study investigates whether activation of brown adipose tissue (BAT) thermogenesis alters systemic metabolism to suppress tumour growth. Cancer cells commonly exhibit elevated aerobic glycolysis (the Warburg effect), characterized by high glucose uptake and lactate production, which supports proliferation and survival. Cold exposure activates BAT-mediated non-shivering thermogenesis (NST), which consumes glucose. The central hypothesis is that BAT activation under cold conditions competes with tumours for circulating glucose, thereby limiting tumour glycolysis and growth. The work aims to define the impact of cold-induced BAT activation on diverse tumour types, delineate underlying mechanisms, and assess translational relevance in humans.

Prior literature establishes glycolysis as a hallmark of cancer metabolism, with many tumours upregulating GLUT1 and glycolytic enzymes, especially under hypoxia via HIF-1α. The Warburg effect provides intermediates for biosynthesis but is ATP-inefficient, necessitating high glucose flux. BAT is specialized for energy expenditure via UCP1-dependent NST, activated by cold, diet, or β-adrenergic stimulation; adult humans retain metabolically active BAT. Glucose contributes substantially to BAT thermogenesis, and impairing GLUTs or hexokinase reduces adipose thermogenic metabolism. Browning of white adipose tissue (WAT) also occurs with cold, potentially affecting systemic metabolism. However, the consequences of BAT activation for tumour growth and metabolism had not been defined. This study builds on these foundations to test whether BAT activation can systemically deprive tumours of glucose and suppress glycolysis-driven growth.

Experimental design encompassed multiple murine cancer models and human pilot studies:

• Mouse models: Subcutaneous xenografts of colorectal cancer (CRC) in immunocompetent C57BL/6 mice; additional xenografts of fibrosarcoma, breast cancer, melanoma, and pancreatic ductal adenocarcinoma; spontaneous genetic models MMTV-PyMT (breast cancer) and ApcMin (intestinal adenoma). Intra-organ (liver) implantation of mouse CRC and human CRC in immunodeficient mice tested tissue/organ independence.
• Environmental conditions: Thermoneutrality (30 °C), room temperature (22 °C), and cold exposure (4 °C). Duration adjusted to achieve comparable tumour sizes where indicated. Body, subcutaneous, and tumour temperature measurements excluded direct chilling effects.
• BAT/WAT assessment: Histology and immunofluorescence of BAT and subcutaneous WAT (sWAT) for UCP1, perilipin, COX4, and CD31; morphological assessment for multilocular adipocytes and mitochondrial content; evaluation of sWAT browning.
• Tumour microenvironment (TME) and proliferation: Immunostaining of tumour sections for CA9 (hypoxia), Ki-67 (proliferation), CD31 (angiogenesis), cleaved caspase-3 (apoptosis), CD45 (myeloid cells), IBA1 (macrophages), and FSP1 (cancer-associated fibroblasts) with quantitative image analysis.
• Metabolic readouts: 18F-FDG PET-CT to quantify glucose uptake (SUV normalized to body weight) in BAT and tumours; indirect calorimetry for thermogenic metabolism; fasting blood glucose measurements; insulin tolerance tests (ITT) and glucose tolerance tests (GTT).
• Interventions: Surgical interscapular BAT removal (iBATectomy) to test BAT dependence; dietary high glucose (15% in drinking water) to elevate circulating glucose and assess rescue of tumour growth under cold; pharmacologic β3-adrenergic activation with CL-316,243.
• Genetics: Ucp1 knockout (Ucp1−/−) mice to test requirement for UCP1-mediated NST.
• Molecular and biochemical analyses: RNA-seq with gene set enrichment analysis (GSEA) to assess pathways (carbohydrate and fatty acid metabolic processes); targeted metabolomics profiling of glycolytic intermediates in BAT and tumours; LDH activity assays; qPCR for Glut genes and glycolysis-related genes; immunoblotting of total and phosphorylated PI3K, AKT, mTOR, and GLUT1.
• Human studies: Healthy volunteers (3 males, 3 females, 22–25 years) underwent mild cold exposure (16 °C, 2–6 h/day for 14 consecutive days) with PET imaging to assess BAT activation. A pilot in an 18-year-old patient with Hodgkin’s lymphoma involved 7 days mild cold (22 °C) vs warm (28 °C) exposure with PET-CT assessment of BAT and tumour 18F-FDG uptake.
• Statistics: Two-sided unpaired t-tests, one-way ANOVA with Tukey post hoc tests, and Wald tests, as appropriate.

• Cold exposure at 4 °C markedly suppressed growth of multiple solid tumours (CRC, fibrosarcoma, breast cancer, melanoma, pancreatic ductal adenocarcinoma). In CRC xenografts, ~80% growth inhibition by day 20 vs 30 °C; room temperature (22 °C) did not suppress tumours.
• Overall survival of CRC-bearing mice was nearly doubled under cold acclimatization vs thermoneutrality.
• TME changes with cold: decreased tumour hypoxia (CA9), reduced CD31+ microvessel density, lower tumour cell proliferation (Ki-67), reduced CD45+ myeloid cells; apoptosis (cleaved caspase-3) unchanged. Similar findings in spontaneous MMTV-PyMT breast tumours and ApcMin intestinal adenomas.
• Tumour suppression occurred in intrahepatic models, excluding local cooling as a cause; core, subcutaneous, and intratumour temperatures did not decrease under cold.
• Cold activated BAT and promoted sWAT browning in tumour-bearing mice, with increased UCP1, COX4, CD31, and multilocular adipocyte morphology.
• 18F-FDG PET-CT revealed redistribution of glucose uptake from tumours to interscapular BAT under cold; tumour 18F-FDG signals were markedly reduced despite similar sizes.
• Systemic metabolism: cold decreased fasting blood glucose and improved insulin sensitivity and glucose clearance (ITT/GTT) in xenograft and spontaneous models.
• BAT dependence: Surgical BAT removal increased blood glucose under cold and abolished cold-induced tumour suppression and TME improvements; in MMTV-PyMT mice, BAT removal eliminated the cold-induced reduction in tumour 18F-FDG uptake.
• Metabolic reprogramming: RNA-seq GSEA showed attenuation of glycolysis and lipid metabolism in cold-exposed tumours. Metabolomics showed increased glycolytic intermediates in BAT and decreased levels (e.g., G6P, F1,6BP, GAP, 3-PG, 2-PG, PEP, pyruvate, lactate) in tumours under cold; LDH activity unchanged.
• GLUTs and signalling: Tumour Glut1, Glut4, and Glut7 mRNA levels decreased under cold; BAT Glut4 and glycolysis-related genes increased. PI3K/AKT/mTOR activation in tumours was reduced by cold.
• High-glucose feeding (15%): at thermoneutrality accelerated tumour growth; under cold abolished tumour suppression across CRC, melanoma, and pancreatic models; restored tumour glycolysis, 18F-FDG uptake, GLUT1 expression, and PI3K/AKT activation.
• UCP1 requirement: Ucp1−/− mice lost cold-induced tumour suppression; BAT 18F-FDG uptake under cold was negligible; tumour glucose uptake and glycolysis were similar to thermoneutral levels; tumour proliferation and hypoxia increased to control levels.
• Human relevance: Mild cold activated substantial BAT in healthy adults (supraclavicular, cervical, parasternal regions). In a lymphoma patient, mild cold increased BAT 18F-FDG uptake and concomitantly reduced tumour 18F-FDG uptake; warm exposure reversed these effects.

The findings support a mechanism in which cold-activated BAT competes with tumours for circulating glucose, limiting tumour glycolysis and growth. Extensive controls excluded direct local cooling effects: internal organ tumour models, temperature measurements, and spontaneous genetic models all showed suppression under cold. BAT activation, not merely WAT browning, appears central to glucose sequestration, as shown by PET-CT redistribution of 18F-FDG and the loss of effect after BAT removal or Ucp1 deletion. Suppression of tumour PI3K/AKT/mTOR signalling and downregulation of GLUTs further reflect impaired glycolytic metabolism in tumours. High-glucose feeding rescued tumour growth under cold by restoring GLUT1 expression and PI3K/AKT activation, indicating that glucose availability and GLUT1-mediated uptake are critical nodes. While WAT may contribute, data point to BAT as the primary glucose sink under these conditions. The human pilot results, though preliminary, align with the preclinical mechanism: mild cold activates BAT and associates with reduced tumour glucose uptake. This BAT-centric, metabolism-altering approach could complement existing therapies and may be broadly applicable across tumour types that rely on glycolysis.

Activation of BAT thermogenesis via cold exposure potently suppresses the growth of diverse solid tumours by lowering systemic glucose availability and impairing tumour glycolysis. The effect depends on BAT and UCP1, is reversible by BAT removal or high-glucose supplementation, and is associated with downregulated GLUT expression and PI3K/AKT/mTOR signalling in tumours. PET-CT demonstrates glucose redistribution from tumours to BAT, and a pilot human study confirms BAT activation by mild cold and reduced tumour glucose uptake. This work introduces a simple, feasible therapeutic paradigm leveraging global metabolic reprogramming to inhibit tumour growth. Future research should include controlled clinical trials to evaluate efficacy and safety in patients, mechanistic dissection of glucose–GLUT1–PI3K/AKT signalling under cold, exploration of WAT and other nutrient contributions, optimization of cold or pharmacologic BAT-activation protocols, and combination strategies with standard anticancer therapies.

Human data are preliminary (small healthy cohort and a single cancer patient); rigorous clinical trials are needed. Room temperature (22 °C) did not activate BAT sufficiently to suppress tumours, indicating sensitivity to ambient temperature. Ucp1 deletion can cause broader mitochondrial defects, potentially confounding interpretation of UCP1 dependence. Contribution of WAT browning cannot be completely excluded. The high-glucose rescue may involve additional pathways (e.g., insulin/IGF signalling) beyond GLUT1, warranting deeper mechanistic studies. Findings are in murine models, and translatability across tumour types and patient populations requires validation.