NRAC controls CD36-mediated fatty acid uptake in adipocytes and lipid clearance in vivo

Article NRAC controls CD36-mediated fatty acid uptake in adipocytes and lipid clearance in vivo Inderjeet Singh1,2,3, Yasuhiro Onogi 1,2,4, Filipe Menezes 5, Dina Khasanova5,LingruKang 1,2, Chenxi Wang1,2,JulioRuiz-Trave 2,6, Sapna Sharma 2,7,8,AhmedKhalil 1,2,6, Valentin K Reichenbach 1,2,YingziShi 1,2, Andrew Flatley9, Xiaocheng Yan1,2, Andreas Israel1,2, Nathalia R V Dragano 2,10, Juan Antonio Aguilar-Pimentel10, Anne Hoffmann11, Adhideb Ghosh 12, Falko Noé 12, Christian Wolfrum 12, Sebastian Cucuruz2,13, Ann-Christine König14, Ingo Burtscher 2,11, Stefanie M Hauck 14, Heiko Lickert2,13, Susanna M Hofmann 2,13,15, Regina Feederle 9,16, Sonja C Schriever 2,17, Rene Hernandez-Bautista 1,2,GencerSancar 2,18, Alberto Cebrian-Serrano2,6, Igor Tetko 5, Helmut Fuchs2,10, Valérie Gailus-Durner2,10, Matthias Blüher2,11,19,20, Martin Hraběde Angelis2,10,21 & Siegfried Ussar 1,2 ✉ Abstract Adipose tissue is a central organiser of systemic lipid homeostasis and a pharmacological target in obesity, orchestrating cellular responses to environmental cues. Nutritionally regulated adipose and cardiac enriched protein (NRAC) is a small adipocyte-specific transmembrane protein with unknown function. Here, we show that Nrac directly interacts with scavenger receptor CD36 via its first transmembrane domain. Forming a complex with CD36 and caveolin-1 under low extracellular fatty acid (FA) concentrations, NRAC modulates CD36-dependent fatty acid uptake in adipocytes. Upon increase in extracellular FA levels, NRAC is ubiquitinated and internalised, leading to CD36’s dissociation from caveolin-1 and clathrin-mediated endocytosis. This results in increased fatty acid uptake into fat cells, adipocyte hypertrophy, increased fat mass and elevated lipid clearance from the blood in chow-diet-fed mice. Finally, human NRAC expression and the intronic SNP rs12878589 are associated with body fat distribution and obesity. Together, these findings reveal a novel regulatory mechanism by which adipocytes sense and respond to extracellular fatty acid availability to fine-tune lipid uptake and storage at cellular and organismal level. Keywords CD36; Clathrin-mediated Endocytosis; Adipose Tissue; Hypertrophy; Fatty Acid Uptake Subject Categories Membranes & Trafficking; Metabolism https://doi.org/10.1038/s44318-025-00520-2 Received 4 November 2024; Revised 15 July 2025; Accepted 17 July 2025 Published online: 1 August 2025 Introduction Adipose tissue is centrally important for the regulation of glucose and lipid homeostasis. Too little and too much adipose mass result in the development of systemic insulin resistance and the metabolic syndrome (Bluher, 2020;Grundy,2015; Klein et al, 2022). Thus, adipocytes continuously sense pleiotropic nervous, endocrine and nutritional/environmental inputs to adjust their endocrine and lipid handling activity. Most of these inputs are processed at the adipocyte cell membrane through various classes of transmembrane receptors and transporters (Onogi et al, 2020). Fatty acid uptake and release are among the most important functions of white adipocytes. Fatty acid uptake is initiated by the release of free fatty 1Research Unit Adipocytes & Metabolism (ADM), Helmholtz Diabetes Center, Helmholtz Zentrum München, German Research Center for Environmental Health GmbH, Neuherberg 85764, Germany. 2German Center for Diabetes Research (DZD), Neuherberg 85764, Germany. 3Department of Medicine, Technische Universität München, Munich, Germany. 4Research Center for Pre-Disease Science (RCPDS), University of Toyama, Sugitani 2630, Toyama 9300194, Japan. 5Molecular Targets and Therapeutics Center, Institute of Structural Biology, Neuherberg 85764, Germany. 6Institute for Diabetes & Obesity, Helmholtz Diabetes Center, Helmholtz Center Munich, Helmholtz Zentrum München, German Research Center for Environmental Health GmbH, Ingolstaedter Landstrasse 1, Neuherberg 85764, Germany. 7Research Unit of Molecular Epidemiology, Helmholtz Zentrum München, Neuherberg 85764, Germany. 8Institute of Epidemiology, Helmholtz Zentrum München, Neuherberg 85764, Germany. 9Core Facility Monoclonal Antibodies (CF-MAB), Helmholtz Zentrum München, German Research Center for Environmental Health GmbH, Neuherberg 85764, Germany. 10Institute of Experimental Genetics, German Mouse Clinic, Helmholtz Zentrum München, German Research Center for Environmental Health GmbH, Neuherberg 85764, Germany. 11Helmholtz Institute for Metabolic, Obesity and Vascular Research (HI-MAG), Helmholtz Zentrum München at the University of Leipzig and University Hospital Leipzig, Leipzig, Germany. 12Institute of Food, Nutrition and Health, ETH Zurich, Schwerzenbach 8092, Switzerland. 13Institute of Diabetes and Regeneration Research (IDR), Helmholtz Diabetes Center, Helmholtz Zentrum Munich, Neuherberg 85764, Germany. 14Metabolomics and Proteomics Core (MPC), Helmholtz Center Munich, Neuherberg, Germany. 15Department of Medicine IV, University Hospital, Ludwig-Maximilians University, Munich, Germany. 16Munich Cluster for Systems Neurology (SyNergy), Feodor-Lynen-Straße 17, Munich D-81377, Germany. 17Research Unit NeuroBiology of Diabetes, Helmholtz Munich, Ingolstädter Landstraße 1, Neuherberg 85764, Germany. 18Department of Internal Medicine IV, Division of Diabetology, Endocrinology and Nephrology, University of Tübingen, Tübingen, Germany. 19Institute for Diabetes Research and Metabolic Diseases, Helmholtz Center Munich, University of Tübingen, Tübingen, Germany. 20Medical Department III –Endocrinology, Nephrology, Rheumatology, University of Leipzig Medical Center, Leipzig 04103, Germany. 21Chair of Experimental Genetics, TUM School of Life Sciences, Technische Universität München, Freising 85354, Germany. ✉E-mail: [email protected] 1234567890();,: © The Author(s) The EMBO Journal Volume 44 | Issue 18 | September 2025 | 5037 –5065 5037 Downloaded from https://www.embopress.org on November 3, 2025 from IP 2001:a62:331:8d01:b08d:a892:1074:db15. WT Nrac ko 0 2000 4000 6000 8000 10000 PGF Area (A.U.) <0.0001 WT Nracko 0 20000 40000 60000 SCF Area (A.U.) <0.0001 B C E WT Nracko PGF SCF BAT PGF SCF 0 5 10 15 Nrac expression mRNA Expression norm. Tbp WT Nracko <0.0001<0.0001<0.0001 WT Nracko 0 1 2 3 4 5 Fat mass Weight (g) 0.0358 WT Nracko 0 7 14 21 28 35 Lean mass Weight (g) 0.1506 D 810121416182022 0 7 14 21 28 35 Body weight Body weight (g) WT Nracko Age (in weeks) F WT Nrac ko 0 500 1000 1500 2000 2500 PGA FAU RFU 0.0004 0 2 4 6 8 10 LTT Time (h) Triglyceride (mM) WT Nracko p= 0.0263 01234 WT Nracko 0 500 1000 1500 2000 2500 SCA FAU RFU 0.0266 HG I WT 0 2 4 6 Serum TG Triglyceride (mM) 0.0259 Nracko WT 0 10 20 30 AUC Arbitrary units 0.0009 Nracko 01234 0 2 4 6 LTT females Time (h) Triglyceride (mM) WT Nrac ko ** WT Nrac KO 0 5 10 15 LTT AUC Arbitrary units 0.0059 WT Nrac KO 0 500 1000 1500 SCA FAU RFU 0.0084 WT Nrac KO 0 500 1000 1500 PGA FAU RFU 0.0609 LJK M OPQ WT (scr) 0 2000 4000 6000 8000 10000 NRAC OE FAU RFU 0.0393 0.0004 0.0003 WT (OE) Nracko (scr) Nracko (OE) N R WT Nracko 0.0 0.2 0.4 0.6 0.8 PGF weight Weight (g) 0.0888 WT Nracko Day 0 Day 8 WT Nracko 0.0 0.2 0.4 0.6 SCF weight Weight (g) 0.0292 The EMBO Journal Inderjeet Singh et al 5038 The EMBO Journal Volume 44 | Issue 18 | September 2025 | 5037 –5065 © The Author(s) Downloaded from https://www.embopress.org on November 3, 2025 from IP 2001:a62:331:8d01:b08d:a892:1074:db15. acids from triglyceride rich liporpoteins by endothelial lipoprotein lipase (LPL) (Boren et al, 2022;GonzalesandOrlando,2007)and subsequent uptake of fatty acids into adipocytes. About 50% of all fatty acids in adipocytes are taken up by the long-chain fatty acid translocase CD36 (Goldberg et al, 2009; Hao et al, 2020). Unlike classical transporters, CD36 transports fatty acids into the adipocytes via endocytosis. Previous studies showed that CD36 endocytosis takes place mainly via caveolin-mediated endocytosis (Hao et al, 2020), albeit clathrin mediated endocytosis has been reports as well (Zeng et al, 2003). The regulation of CD36 endocytosis is important as CD36 localization in adipocytes is the major regulator of CD36-mediated fatty acid transport (Brailey et al, 2022;Daquinagetal,2021; Hao et al, 2020; Peche et al, 2023; Wang et al, 2024). However, compared to lipolysis (Ali et al, 2015), relatively little is known about the modulation of fatty acid uptake into adipocytes, or mechanisms distinguishing CD36-mediated fatty acid uptake in adipocytes compared to other cell types such as cardiac and skeletal muscle, liver and macrophages (Ali et al, 2015). Cell type-specific regulation of CD36-mediated fatty acid uptake could be mediated by modulation of transport activity by coreceptors or other modulatory proteins with a more restricted expression pattern. To date, very few adipocyte-selective cell surface proteins are known (Ussar et al, 2014). Among them, NRAC (Nutritionally regulated adipose and cardiac enriched protein; mouse homolog A530016L24Rik,humanhomolog-C14ORF180) shows some of the highest specificity to white and brown adipocytes, with additional low expression in heart (Zhang et al, 2012). NRAC is a two-pass membrane protein with both the Nand C-terminus located intracellularly and a small extracellular loop, with no additional identified protein domains. Blast searches do not reveal any homologous eukaryotic proteins that could indicate a potential function of NRAC. However, previous studies showed downregulation of Nrac gene expression in white adipose tissue of fasted and obese mice, with little impact of these nutritional changes in Nrac expression in brown adipose tissue (BAT) (Zhang et al, 2012). Knockdown studies in cultured in vitro differentiated adipocytes suggested a role of NRAC in adipocyte differentiation or lipid accumulation (Kerr et al, 2019). However, if and what role NRAC could play in adipocyte lipid storage in vivo remains unknown. Here we show that NRAC acts as a rheostat for CD36-mediated fatty acid uptake, dependent on the extracellular fatty acid concentration. We provide data that under low extracellular fatty acid concentrations NRAC forms a complex with CD36 and caveolin-1, limiting fatty acid uptake. Upon loss of NRAC or an increase in extracellular fatty acid levels, leading to a reduction in surface localization of NRAC via increased ubiquitination resulting in CD36 dissociation from caveolin-1 and is internalized via clathrin-mediated endocytosis. This results in increased fatty acid uptake into adipocytes and adipocyte hypertrophy, increased fat mass and elevated lipid clearance from the blood in chow diet fed mice. Results NRAC deletion leads to hypertrophic white adipocytes due to increased fatty acid uptake NRAC is predominantly expressed in white and brown adipocytes with little additional expression reported in heart (Zhang et al, 2012). To investigate the function of NRAC in vivo, we obtained whole-body Nrac knockout mice and verified complete loss of Nrac expression by qPCR (Fig. 1A). We also developed a rat monoclonal antibody targeting the intracellular N-terminus of murine NRAC, detecting murine NRAC in NRAC-GFP transfected cells (Fig. EV1A), as well as endogenous levels by western blotting (Fig. EV1B), immunostaining (Fig. EV1C), and immunoprecipitation (Fig. EV1D). Chow diet fed male Nrac knockout (Nracko) mice did not show differences in body weight (Fig. 1B).However,maleNrac ko mice had increased total fat mass (Fig. 1C), and subcutaneous (SCF) and perigonadal (PGF) adipose tissue mass (Fig. 1D,E), with no significant differences in lean mass (Fig. 1F) at 15 weeks of age compared to wild-type (wt) littermate controls. Histological analysis of SCF and PGF showed increased mean adipocyte size in both depots of Nracko mice (Fig. 1G,H), but no signs of increased immune cell infiltration (Fig. 1G) or elevated expression of Tnfa or Il6 (Fig. EV1E–H). Weights of liver, heart and pancreas showed no genotype-specific differences (Fig. EV1I). Elevated fatty acid uptake is a major driver of adipocyte hypertrophy (Gonzales and Orlando, 2007), which was elevated in primary subcutaneous and perigonadal adipocytes from Nracko mice compared to wt littermate controls (Fig. 1I,J). Interestingly, overexpression of NRAC in wt and Nracko in vitro differentiated adipocytes led to a reduction in fatty acid uptake compared to the controls (Fig. 2K). Supporting the physiological relevance of the increased adipocyte fatty acid uptake, we found reduced random fed serum triglyceride levels (Fig. 1L) and increased lipid clearance during a lipid tolerance test (Fig. 1M,N) in Nracko mice compared to wt littermate controls. Similar to male Nracko mice, Nracko females exhibited enhanced lipid clearance (Fig. 1O) and increased fatty acid uptake in subcutaneous adipocytes (Fig. 1P), with a similar trend in perigonadal adipocytes (Fig. 1Q). Figure 1. NRAC deletion promotes adipocyte fatty acid uptake in chow diet-fed male mice. (A) mRNA expression of Nrac in BAT, PGF and SCF tissues (n=7 wt, 5 ko). (B) Body weight (n=9 wt, 5 ko). (C) Fat mass (n=9 wt, 5 ko, Age 15 weeks). Tissue weight of (D) PGF and (E) SCF (n=6 wt, 8 ko, age 23 weeks). (F) Lean mass (n=9 wt, 5 ko, Age 15 weeks). (G) H&E staining of PGF and SCF (age 23 weeks) (scale bar 200 µm)(H) Quantification of adipocyte cell sizes. (I,J) Fatty acid uptake (FAU) in the primary isolated adipocytes from (I) SCF and (J) PGF of male mice (n=5 wt, 6 ko, age 12 weeks). (K) Fatty acid uptake in stably NRAC overexpressing (OE) wt and Nracko adipocytes. Scr plasmid was used as control. (L) Serum Triglyceride (TG) levels of 23 weeks old mice (n=8 wt, 7 ko). (M) Lipid tolerance test (LTT) (n=9 wt, 7 ko) (age 14 weeks) and, (N) baseline corrected area under the curve (AUC), (O) Lipid tolerance test in chow diet fed female mice (n=4 wt. 4ko). Fatty acid uptake in primary (P) subcutaneous adipocytes (SCA) and (Q) perigonadal adipocytes (PGA) isolated from female mice (n=4 wt, 4ko). (R) Oil red O staining on the primary differentiated subcutaneous adipocytes (scale bar 200 µm). Data are represented as mean ± SEM. *P< 0.05, **P< 0.01, ***P< 0.001 by Student’sttest or one or two-way ANOVA was used. Source data are available online for this figure. Inderjeet Singh et al The EMBO Journal © The Author(s) The EMBO Journal Volume 44 | Issue 18 | September 2025 | 5037 –5065 5039 Downloaded from https://www.embopress.org on November 3, 2025 from IP 2001:a62:331:8d01:b08d:a892:1074:db15. AB C WT Nrac ko 0 500 1000 1500 2000 2500 BA FAU RFU 0.0174 WT Nracko UCP1 α TUBULIN D EF WT Nracko 0.0 0.5 1.0 1.5 UCP1 Normalized protein expression 0.0135 WT Nracko G H -50 0 50 100 150 200 250 OCR FA OCR (pmol/min) >0.999 <0.0001 <0.0001 WT (BSA) Nracko (BSA) WT (Oleate) Nracko (Oleate) Nracko (Palmitate) WT (Palmitate) 0 50 100 150 200 250 OCR ISO OCR (pmol/min) 0.7516 <0.0001 <0.0001 WT (BSA) Nracko (BSA) WT (Oleate) Nracko (Oleate) Nracko (Palmitate) WT (Palmitate) WT Nracko 0 100 200 300 400 Isoproterenol OCR (pmol/min) 0.0173 WT Nracko 0 100 200 300 400 Proton Leak OCR (pmol/min) 0.0173 WT Nrac ko 0 200 400 600 800 Maximal Respiration OCR (pmol/min) 0.3832 WT Nracko 0 200 400 600 Spare Capacity OCR (pmol/min) 0.7922 I WT BSA 0 0 800 30 60 Iso 90 120 200 400 600 OCR Time (min) OCR (pmol/min) Nracko BSA WT Oleate Nracko Oleate WT Palmitate Nracko Palmitate FCCP AA/Rot BSA/FA JK MN 0 500 1000 1500 2000 FAU ISO RFU 0.0074 0.0004 WT (DMSO) WT (Iso) Nracko (DMSO) Nracko (Iso) WT Nracko 0 100 200 300 ATP-Linked Respiration OCR (pmol/min) 0.4286 L 0 100 200 300 400 500 OCR OCR (pmol/min) WT Nrac ko 030 60 90 120 Iso Oligo FCCP AA/Rot Time (min) WT Nrac ko 0 100 200 300 400 BAT Ucp1 0.0113 mRNA expression Norm Tbp The EMBO Journal Inderjeet Singh et al 5040 The EMBO Journal Volume 44 | Issue 18 | September 2025 | 5037 –5065 © The Author(s) Downloaded from https://www.embopress.org on November 3, 2025 from IP 2001:a62:331:8d01:b08d:a892:1074:db15. However, the observed adipose hypertrophy in chow diet fed male Nracko mice did not impair systemic glucose (Fig. EV1J) or insulin metabolism (Fig. EV1K), as glucose and insulin tolerance as well as percentage of glycated haemoglobin (%HbA1c) were not different between Nracko and wt littermate control mice (Fig. EV1L). Moreover, female mice did not show genotypespecific differences in body weight (Fig. EV1M), body composition (Fig. EV1N,O), glucose metabolism, (Fig. EV1P,Q), insulin sensitivity (Fig. EV1R) and adipose histology (Fig. EV1S,T). Assessment of NRAC expression between male and female mice across adipose depots revealed comparable expression between male and female adipose tissues (Fig. EV1U–W). Thus, the absence of adipose hypertrophy in female compared to male mice is most likely not the result of mechanistic differences of NRAC between male and female mice but rather a consequence of the increased whole-body energy expenditure in female mice (Mauvais-Jarvis et al, 2013). Previous studies showed that a knockdown of NRAC impairs adipogenic differentiation in vitro (Kerr et al, 2019; Zhang et al, 2012), which could contribute to compensatory adipocyte hypertrophy (Bakker et al, 2006). We did not observe any differences in adipogenesis of subcutaneous wt and Nracko preadipocytes differentiated in vitro, as shown by Oil Red O staining (Fig. 1R). Thus, loss of NRAC impacts on systemic lipid metabolism by redirecting fatty acids to white adipocytes for storage. NRAC deletion promotes thermogenesis in brown adipocytes without affecting systemic energy expenditure In addition to white adipocytes, NRAC is expressed in brown adipocytes. Fatty acids induce UCP-1 mediated mitochondrial uncoupling and thermogenesis (Bartelt et al, 2011). Therefore, we investigated the brown adipose tissue (BAT) phenotype of Nracko mice. Like white adipocytes, NRAC knockout increased fatty acid uptake of primary brown adipocytes (Fig. 2A). Moreover, stimulation of differentiated adipocytes with isoproterenol led to increased fatty acid uptake (Fig. 2B). In line with elevated fatty acid influx, we observed slightly increased Ucp1 mRNA (Fig. 2C) and protein levels in Nracko BAT (Fig. 2D). Histologically, we observed reduced lipid droplet size in BAT of Nracko mice (Fig. 2E). To assess the effects of NRAC deletion on brown adipocyte energy metabolism, we differentiated primary brown adipocytes in vitro and measured oxygen consumption using a Seahorse Extracellular Flux analyzer. Nracko brown adipocytes showed significantly increased oxygen consumption rates (OCR) upon isoproterenol stimulation, as well as increased proton leakage (Fig. 2F–H). No differences were observed upon FCCP induced maximal respiration (Fig. 2I), spare capacity (Fig. 2J) and ATP-linked respiration (Fig. 2K) indicating that Nracko brown adipocytes have higher thermogenic capacity, but require the induction of lipolysis. This is likely due to low extracellular fatty acid concentrations in the assay medium. Previous studies have demonstrated that fatty acid stimulation enhances the thermogenic capacity of brown adipocytes (Li et al, 2014), potentially through increased internalization of CD36. To investigate whether the elevated fatty acid uptake observed in Nracko adipocytes contributes to enhanced thermogenic function, we administered fatty acids (oleate and palmitate) via the Seahorse assay port. Notably, fatty acid stimulation led to a significant increase in OCR, with Nracko adipocytes exhibiting a markedly higher response compared to controls (Figs. 2L–NandEV2A–C). These findings suggest that loss of Nrac potentiates thermogenic activation through enhanced fatty acid-driven respiration. To explore the physiological regulation of NRAC, we examined its cellular localization under various conditions, including starvation, acute cold exposure, andtreatmentwiththeβ3-adrenergic agonist CL-316243. Among those, cold exposure and CL treatment resulted in a minor reduction in NRAC surface localization (Fig. EV2D). Measurement of whole-body energy expenditure and substrate utilization, did not reveal any differences between Nracko and wt littermate mice (Fig. EV2E–H). Moreover, acute treatment with CL-316243 (Fig. EV2I,J), or acute cold exposure (Fig. EV2K,L) also did not induce differences in systemic energy expenditure or substrate utilization. Thus, NRAC deletion increases brown adipocyte fatty acid influx and thermogenesis. However, the reduced serum triglyceride concentration appears to limit or eliminate the systemic consequences of this BAT phenotype. Loss of NRAC enhances adipocyte fatty acid uptake by modulating CD36 endocytic pathways To understand the mechanism underlying the increased adipocyte fatty acid uptake upon loss of NRAC, we identified interaction partners using proteomics following NRAC immunoprecipitation. NRAC and its interaction partners were immunoprecipitated using our monoclonal rat antibody from PGF of wt mice. PGF from Nracko mice was used as control. Using this approach, we identified 13 NRAC-interacting proteins (Fig. 3A). Among those, we detected proteins related to lipid uptake including CD36, caveolin-1, and cavin 1. Interaction of NRAC with CD36 and caveolin-1 was confirmed using immunoprecipitation from BAT, SCF (Fig. EV3A) and PGF (Fig. 3B) with subsequent western blotting. Figure 2. NRAC deletion increases UCP1 in BAT and oxygen consumption rates in brown adipocytes. (A) Fatty acid uptake (FAU) in primary mature brown adipocytes (BA) (n=5 wt, 6 ko, age 12 weeks). (B) FAU in differentiated brown adipocytes, treated with either DMSO or 5 µM isoproterenol for 20 min (n=5). (C)Ucp1 mRNA (n=6 wt, 8 ko) and (D) protein expression (n=4 wt and 3 ko) and quantification. (E) H&E staining of BAT (age 23 weeks) (scale bar 200 µm). (F) Oxygen consumption rate (OCR) in primary differentiated mature brown adipocytes (Iso Isoproterenol, Oligo Oligomycin, FCCP Carbonyl cyanide-4 (trifluoromethoxy), AA Antimycin A, Rot Rotenone), (G) OCR after isoproterenol injection (H) proton leak and (I) maximal respiratory capacity (J) spare capacity, (K) ATP-linked respiration (n=5 wt, 6 ko, age 12 weeks). (L) Oxygen consumption rate in the differentiated adipocytes from wt and Nracko cells. BSA conjugated oleate and palmitate (100 µM) were injected in the first port, Isoproterenol (Iso) in the second port, FCCP (Carbonyl cyanide-4 (trifluoromethoxy) in 3rd port and AA (Antimycin A) and Rot (Rotenone) in the 4th port. (M) OCR calculated after subtracting fatty acid (FA) stimulated values from basal and (N) isoproterenol (Iso) stimulated OCR from basal (n=10). Data are shown as mean ± SEM. *P< 0.05, **P< 0.01, ***P< 0.001 by Student’sttest in all figures except (E) where two-way ANOVA was used. Data on figure G-Kwere analyzed via Mann–Mann-Whitney Utest as data were not normally distributed. Source data are available online for this figure. Inderjeet Singh et al The EMBO Journal © The Author(s) The EMBO Journal Volume 44 | Issue 18 | September 2025 | 5037 –5065 5041 Downloaded from https://www.embopress.org on November 3, 2025 from IP 2001:a62:331:8d01:b08d:a892:1074:db15. A B D WT WT Nracko Nracko CD36 CAV1 GAPDH Input IP: CD36 C C WT WT Nracko Nracko Input IP: CD36 CD36 CAV1 ERK 1/2 PGF WT WT Nracko Nracko CD36 CAV1 GAPDH Input IP: CD36 SCF WT WT Nracko Nracko GAPDH CAV1 CD36 Input IP: CD36 BAT E F G CD36 CAVIN2 CAV1 2 3 0246 log (Fold Change) -log10 (pValue) WT Nracko Nracko WT VINCULIN NRAC CAV1 CD36 Input IP: NRAC Vehicle CPZ NRAC CAV1 CD36 Merge Nracko WT WT Nracko The EMBO Journal Inderjeet Singh et al 5042 The EMBO Journal Volume 44 | Issue 18 | September 2025 | 5037 –5065 © The Author(s) Downloaded from https://www.embopress.org on November 3, 2025 from IP 2001:a62:331:8d01:b08d:a892:1074:db15. CD36 facilitates long-chain fatty acid import mainly via caveolin-mediated endocytosis (Hao et al, 2020). Using primary in vitro differentiated subcutaneous adipocytes from wt and Nracko mice, we studied the spatial distribution of CD36 and caveolin-1. Immunostaining for CD36, caveolin-1 and NRAC in wild-type adipocytes showed that all proteins localized at the plasma membrane. However, while caveolin-1 was retained at the cell surface in Nracko adipocytes, CD36 was no longer at the plasma membrane, but appeared in clusters accumulating around lipid droplets in primary (Fig. 3C) as well as in immortalized differentiated white adipocytes (Fig. EV3B). Immunoprecipitation of CD36 confirmed a strongly reduced interaction between CD36 and caveolin-1 in Nracko adipocytes (Fig. 3D). Moreover, the interaction of CD36 and caveolin-1 was strongly reduced in PGF, SCF and BAT of Nracko mice compared to wt littermate controls (Fig. 3E–G), verifying the phenotype in vivo. CD36 transports fatty acids via endocytosis of the fatty acid/ CD36 complex. Loss of NRAC resulted in a dissociation between CD36 and caveolin-1, indicating that upon loss of NRAC, CD36 is utilizing an alternative endocytic pathway. Indeed, inhibition of clathrin-mediated endocytosis using chlorpromazine for one hour, restored CD36 cell surface localization (Figs. 3CandEV3B).Thus, loss of NRAC results in a switch from caveolin to clathrin-mediated CD36 endocytosis, increasing fatty acid uptake. NRAC interacts with CD36 via its first transmembrane domain Mouse NRAC is a 165-amino-acid, two-pass membrane protein composed of a large N-terminal intracellular domain (id1, residues 1–112), two transmembrane domains (tm1, residues 113–135; tm2, residues 145–164), and an extracellular loop (el, residues 136–144). DeepLoc 2.0 (Thumuluri et al, 2022) predictions indicate that the signal peptide overlaps with the tm1 region (residues 110–130). To investigate how NRAC interacts with CD36 and to identify the domain responsible for this interaction, we generated several Nrac mutant constructs. One construct lacked tm1 (NRAC 1–112 +SP), where the signal peptide and transmembrane domain of CD36 (SP = GLIAGAVIGAVLAVFGGILMPV) was used. Additional deletion constructs included NRAC 1–135 (containing id1 and tm1) and NRAC 1–144 (containing id1, tm1, and el) (Fig. 4A). These constructs were co-transfected with CD36 in HEK293T cells, which lack endogenous NRAC and CD36, and immunoprecipitation (IP) of NRAC was performed. The results showed that removal of tm1 abolished the interaction between NRAC and CD36, while deletion of other regions did not impair the interaction (Fig. 4B), indicating that tm1 is essential for NRAC–CD36 binding. Given that the NRAC–CD36 interaction plays a key role in inhibiting CD36-mediated fatty acid uptake, we next tested whether disrupting this interaction would impair NRAC function. HEK293T cells were co-transfected with CD36 and either fulllength NRAC (NRAC 1–165) or the truncated construct lacking tm1 (NRAC 1–112 +SP). While overexpression of full-length NRAC significantly reduced fatty acid uptake, the truncated NRAC constructfailedtodoso(Fig.4C). These findings confirm that NRAC’s interaction with CD36 is required for its ability to regulate CD36-mediated fatty acid uptake. Molecular dynamics (MD) simulations are a powerful tool for predicting and analyzing protein–protein interactions at atomic resolution. To explore whether human NRAC and CD36 exhibit similar interaction dynamics as observed in mice, we performed MD simulations of human NRAC (C14ORF180) and CD36. Over the course of 500 ns, the simulations revealed a stable interaction between human NRAC and CD36 (Fig. EV4A–C).Notably,thekey interacting residues were predominantly located within the first transmembrane domain (tm1) of human NRAC (Fig. 4D), supporting the experimental findings and highlighting a conserved mechanism of interaction. High fat diet feeding resembles NRAC deficiency, redirecting CD36 to clathrin-mediated endocytosis in adipocytes Based on these data, we hypothesized that high fat diet (HFD) feeding would result in accelerated development of insulin resistance and hyperglycemia in Nracko mice due to preexisting adipocyte hypertrophy and increased fatty acid uptake. Surprisingly, high-fat diet feeding for 10 weeks did not accelerate the development of glucose intolerance (Fig. EV5A,B) systemic insulin resistance (Fig. EV5C), hepatosteatosis (Fig. EV5D–F), or serum markers of adipose inflammation (Fig. EV5G–I) or hepatic function (Fig. EV5J). Conversely, HFD feeding abolished all adipose tissue phenotypes observed in CD fed mice. HFD fed Nracko mice had comparable lipid clearance (Fig. 5A) and serum triglyceride levels (Fig. 5B), with slightly increased HDL levels (Fig. EV5K). HFD fed Nracko mice also showed no differences in body weight (Fig. EV5L), body composition (Fig. 5C,D) and adipocyte size (Fig. EV5M,N). In line with this, fatty acid uptake in primary mature perigonadal (Fig. 5E) and subcutaneous (Fig. 5F) adipocytes from HFD fed mice was not different between wt and Nracko adipocytes. To understand the reason for the diminished phenotype in HFD fed Nracko mice, we treated mature in vitro differentiated subcutaneous white adipocytes with oleate to mimic the increased fatty acid environment in HFD fed mice. Previous data showed that long-chain fatty acid stimulation leads to CD36 internalization in adipocytes (Hao et al, 2020). Indeed, we observed that acute (Fig. EV6A,B) and chronic (Fig. 5G) fatty acid stimulation leads to increased CD36 internalization in a time and dose-dependent manner. Figure 3. NRAC regulates CD36/ caveolin-1 interaction and CD36 internalization. (A) Mass spectrometry analysis of immunoprecipitated samples from PGF of wt and Nracko mice (n=4). Proteins present in at least two wt samples, a wt to ko ratio >2 and an adjusted Pvalue < 0.05 were blotted. The significance threshold was set at log2 fold change above 1 and -log 10 (Pvalue) above 2. Significantly up-regulated proteins are shown in red. (B) Co-immunoprecipitation of NRAC with CD36 and CAVEOLIN-1 (CAV1). (C) Immunostaining of primary differentiated mature adipocytes either vehicle (0.05% ethanol) or Chlorpromazine (CPZ) (10 μM) treated for 1 h (scale 25 µm). (D) Co-immunoprecipitation of CD36 and CAV-1 from differentiated adipocytes. Co-immunoprecipitation of CD36 and CAV-1 from PGF (E), SCF (F) and BAT (G) of male mice. Source data are available online for this figure. Inderjeet Singh et al The EMBO Journal © The Author(s) The EMBO Journal Volume 44 | Issue 18 | September 2025 | 5037 –5065 5043 Downloaded from https://www.embopress.org on November 3, 2025 from IP 2001:a62:331:8d01:b08d:a892:1074:db15. In addition, oleate treatment also induced internalization of caveolin-1, in both wt and Nracko adipocytes. Inhibition of clathrinmediated endocytosis using chlorpromazine restored cell surface localization of CD36 in both wt and Nracko adipocytes but had no effectsoncaveolin-1localization (Figs. 5G and EV6C). Importantly, we observed a strongly reduced surface localization of NRAC upon oleate treatment in wt adipocytes (Figs. 5G and EV6A–C). In line with these in vitro data, immunostaining of NRAC in chow diet and HFD fed mice confirmed that NRAC surface localization is reduced upon HFD feeding in adipose tissue in vivo (Fig. 5H), albeit mRNA expression was increased in both SCF and PGF (Fig. EV6D,E). Female HFD fed mice showed a NRAC (1-165) NRAC (1-112+ SP) NRAC(1-135) NRAC (1-144) id1: intracellular domain 1 (region 1-112) tm1: transmembrane domain 1 (region 113-135) el: extracellular loop ( region136-144) tm2: transmembrane domain 2 (region 145-164) id2: intracellular domain 2 (residue 165) SP: CD36 signal peptide id1 id2 tm1 tm2 el A CD36 CD36+NRAC (1-165) CD36+ NRAC (1-112+SP) 0 50000 100000 150000 200000 250000 FAU NRAC Mutants RFU 0.0100 0.3110 B C VINCULIN NRAC CD36-mCherry IP: NRAC CD36-mCherry CD36mCherry NRAC (1-165) NRAC (1-112+SP) NRAC (1-135) NRAC (1-144) D R98 S103 V110 L118 R124 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 RPHGGSLLLQLCVCVLLVLALGLYCGR R98 G102 V110 L118 G123 NRAC-CD36 Average distance (Å) Figure 4. First transmembrane domain of NRAC is required for its association with CD36. (A) Cartoon of full-length NRAC and various constructs from mouse NRAC. (B) Co-immunoprecipitation of NRAC and CD36. NRAC constructs and CD36-mCherry were cotransfected and IP was performed using our NRAC antibody. (C) Fatty acid uptake assay in HEK293T cells transfected with either CD36-mCherry alone or cotransfected with NRAC (1–165) and NRAC (1–112 +SP) (n=6). (D) The average minimum distance between NRAC residues and CD36 residue was calculated over 500 ns of molecular dynamics simulation. A structural representation of the interaction surface is shown on the top. The first NRAC residue observed to interact with CD36 is R98, located within the intrinsically disordered region (IDR). The first helical interaction begins at G102 and extends to G123, with R124 exhibiting borderline interaction behavior. Data in (C) are shown as mean ± SEM. *P< 0.05, **P< 0.01, ***P< 0.001 by one-way ANOVA. Source data are available online for this figure. The EMBO Journal Inderjeet Singh et al 5044 The EMBO Journal Volume 44 | Issue 18 | September 2025 | 5037 –5065 © The Author(s) Downloaded from https://www.embopress.org on November 3, 2025 from IP 2001:a62:331:8d01:b08d:a892:1074:db15. AG 01234 0 20 40 60 LTT HFD Time (h) Serum TG levels (mM) WT Nracko WT Nracko 0 50 100 150 200 AUC AUC 0.8367 C WT Nracko 0 5 10 15 20 Serum TG levels TG levels (mM) 0.0949 B WT (BSA) WT (Oleate) 0 1000 2000 3000 4000 5000 PGA FAU RFU 0.4261 0.2224 Nracko (BSA) Nracko (Oleate) WT (BSA) WT (Oleate) 0 1000 2000 3000 4000 SCA FAU RFU 0.9772 0.9130 Nracko (Oleate) Nracko (BSA) E Oleate (24h) +CPZ (1h) Oleate (24h) NRAC CAV1 CD36 Merge WTNracko WTNracko D F H NRAC + Ub-HA NRAC NRAC NRAC NRAC BSA+MG132 Oleate+MG132 Input Input IP: HA IP: HA IB: UBIQUITIN NRAC VINCULIN NRAC + Ub-HA NRAC + Ub-HA NRAC + Ub-HA I WT Nracko 0 10 20 30 Lean Mass Weight (g) 0.2467 WT Nracko 0 5 10 15 Fat Mass Weight (g) 0.5194 NRAC PLIN1 Merge Nracko Nracko WT WT Chow Diet High fat diet Inderjeet Singh et al The EMBO Journal © The Author(s) The EMBO Journal Volume 44 | Issue 18 | September 2025 | 5037 –5065 5045 Downloaded from https://www.embopress.org on November 3, 2025 from IP 2001:a62:331:8d01:b08d:a892:1074:db15. Human data The human data utilized in this study were obtained from the cross-sectional cohort (CSC) of the Leipzig Obesity Biobank (LOBB, https://www.helmholtz-munich.de/en/hi-mag/cohort/ leipzig-obesity-bio-bank-lobb). This cohort includes paired samples of abdominal subcutaneous and omental visceral adipose tissue from a total of 1,479 participants. These individuals are characterized as either having normal/overweight (N= 31; 52% women; average age: 55.8 ± 13.4 years; average BMI: 25.7 ± 2.7 kg/ m²) and obese (N= 1448; 71% women; average age: 46.9 ± 11.7 years; average BMI: 49.2 ± 8.3 kg/m²). Adipose tissue samples were collected during elective laparoscopic abdominal surgeries, adhering to established protocols (Langhardt et al, 2018;Mardinoglu et al, 2015). Body composition and metabolic parameters were evaluated using standardized methods as previously detailed (Bluher, 2020;Klotingetal,2010). Exclusion criteria for participation included being under 18 years of age, having a history of chronic substance or alcohol abuse, smoking within the last 12 months prior to surgery, suffering from acute inflammatory diseases, using glitazones concurrently, having end-stage malignancies, experiencing weight loss greater than 3% in the three months before surgery, uncontrolled thyroid disorders, and Cushing’s disease. The study received approval from the Ethics Committee of the University of Leipzig (approval number: 159-1221052012) and was conducted in accordance with the principles set forth in the Declaration of Helsinki. All participants provided written informed consent prior to their inclusion in the study. We generated ribosomal RNA-depleted RNA sequencing data using the SMARTseq protocol (Picelli et al, 2014;Songetal,2018). The libraries were sequenced as single-end reads on a Novaseq 6000 (Illumina, San Diego, CA, USA) at the Functional Genomics Center in Zurich, Switzerland. The preprocessing steps were performed as previously outlined (Hagemann et al, 2023). In summary, adapter and quality-trimmed reads were aligned to the human reference genome (assembly GRCh38.p13, GENCODE release 32), and gene-level expression quantification was conducted using Kallisto (Bray et al, 2016) version 0.48. For samples with read counts exceeding 20 million, we downsampled them to 20 million reads using the R package ezRun version 3.14.1 (https:// github.com/uzh/ezRun, accessed on April 27, 2023). Data normalization was performed using a weighted trimmed mean (TMM) of the log expression ratios, incorporating adjustments for age, transcript integrity numbers (TINs), and gender, except in cases where male and female samples were analyzed independently. All analyses were performed in R version 4.3.1 (www.R-project.org). MD simulations Initial structure generation NRAC lacks a resolved crystal structure, but several UniProt entries exist for different species (Consortium, 2024). CD36, in complex with the CIDRαdomain from MCvar1 PfEMP1 (PDB-ID: 5LGD), has been crystallized. However, the specific region of interest— where NRAC is likely to bind is unresolved so needed computational modeling (Hsieh et al, 2016). The human sequences of each protein were used inAlphaFold2 predictions (Jumper et al, 2021). Experimental evidence suggests that binding occurs within the membrane, involving residues in NRAC’sTM1regionanda transmembrane helix of CD36. The transmembrane topology was predicted using DeepTMHMM (Hallgren et al, 2022), and the membrane orientation was determined with the PPM server (Lomize et al, 2012). Two potential interaction sites were identified: NRAC residues 105–121 and CD36 residues 8–30 or 440–466. Full-length 3D structures of both proteins were generated using the LIT-AlphaFold pipeline with AlphaFold2 (Mirdita et al, 2022) (Urvas et al, 2024)(Yuetal,2023). From a visual inspection of PPM’s output, the most promising interacting residues were selected, i.e., residues 105–121 for NRAC and 14–30 or 440–456 for CD36. These regions were docked using HADDOCK2.4 (Honorato et al, 2024), and the top clusters were selected for molecular dynamics simulations. Molecular dynamics simulations Molecular dynamics simulations were conducted to assess the stability of the protein–protein complexes. Simulations were performed using GROMACS (Abraham et al, 2015)withthe CHARMM36 force field (Huang and MacKerell Jr, 2013). As an initial step, explicit water simulations were run to evaluate complex stability in aqueous solution, given the largely lipophilic nature of the interaction surface. System preparation and execution followed standard protocols (Lemkul, 2018). Subsequently, the complexes were embedded in a lipid bilayer using CHARMM-GUI (Lee et al, 2016) for all-atom membrane simulations. Multiple membrane compositions were tested; the final system comprised 40 cholesterol and 110 POPC (1-palmitoyl2-oleoyl-sn-glycero-3-phosphocholine) molecules per leaflet. All simulations were run for 500 ns. Analysis was performed using custom Python scripts, and visualizations were generated with ChimeraX (Meng et al, 2023). Statistical analysis All statistics were calculated using GraphPad Prism 10. Data are presented as mean ± standard error of the mean (SEM) unless stated differently in the figure legend. Statistical significance was determined by unpaired Student’sttest or, using oneor two-way analysis of variance (ANOVA), followed by Tukey’s multiple comparison test if not otherwise stated in figure legends. Differences reached statistical significance with P< 0.05. The Spearman correlation coefficient was employed to evaluate the relationship between human C14ORF180 (human NRAC) expression and metabolic parameters, with adjustments made for multiple testing using the false discovery rate. Data availability The data supporting the findings of this study are available within the paper and its supplementary information files. The human RNA-seq data from the LOBB have not been deposited in a public repository due to restrictions imposed by patient consent but can be obtained from Matthias Blüher upon request. LIT-AlphaFold can be found at https://github.com/LIT-CCM-lab/LIT-AlphaFold. The code of AlphaFold, as well as the related models’weights, genetic databases, and template database can be found at https:// github.com/google-deepmind/alphafold. The mass spectrometry data are deposited to the ProteomeXchange Consortium (http:// The EMBO Journal Inderjeet Singh et al 5052 The EMBO Journal Volume 44 | Issue 18 | September 2025 | 5037 –5065 © The Author(s) Downloaded from https://www.embopress.org on November 3, 2025 from IP 2001:a62:331:8d01:b08d:a892:1074:db15. proteomecentral.proteomexchange.org) via the PRIDE partner repository with the dataset identifier PXD PXD054842. The source data of this paper are collected in the following database record: biostudies:S-SCDT-10_1038-S44318-025-00520-2. Expanded view data, supplementary information, appendices are available for this paper at https://doi.org/10.1038/s44318-025-00520-2. Peer review information A peer review file is available at https://doi.org/10.1038/s44318-025-00520-2 References Abraham MJ, Murtola T, Schulz R, Páll S, Smith JC, Hess B, Lindahl E (2015) GROMACS: high performance molecular simulations through multi-level parallelism from laptops to supercomputers. SoftwareX 1-2:19–25 Ali AH, Mundi M, Koutsari C, Bernlohr DA, Jensen MD (2015) Adipose tissue free fatty acid storage in vivo: effects of insulin versus niacin as a control for suppression of lipolysis. Diabetes 64:2828–2835 Bakker AH, Nijhuis J, Buurman WA, van Dielen FM, Greve JW (2006) Low number of omental preadipocytes with high leptin and low adiponectin secretion is associated with high fasting plasma glucose levels in obese subjects. Diab Obes Metab 8:585–588 Bartelt A, Bruns OT, Reimer R, Hohenberg H, Ittrich H, Peldschus K, Kaul MG, Tromsdorf UI, Weller H, Waurisch C et al (2011) Brown adipose tissue activity controls triglyceride clearance. Nat Med 17:200–205 Bluher M (2020) Metabolically healthy obesity. Endocr Rev 41:bnaa004 Boren J, Taskinen MR, Bjornson E, Packard CJ (2022) Metabolism of triglyceriderich lipoproteins in health and dyslipidaemia. Nat Rev Cardiol 19:577–592 Brailey PM, Evans L, Lopez-Rodriguez JC, Sinadinos A, Tyrrel V, Kelly G, O’Donnell V, Ghazal P, John S, Barral P (2022) CD1d-dependent rewiring of lipid metabolism in macrophages regulates innate immune responses. Nat Commun 13:6723 Bray NL, Pimentel H, Melsted P, Pachter L (2016) Near-optimal probabilistic RNA-seq quantification. Nat Biotechnol 34:525–527 Consortium TU (2024) UniProt: the Universal Protein Knowledgebase in 2025. Nucleic Acids Res 53:D609–D617 Daquinag AC, Gao Z, Fussell C, Immaraj L, Pasqualini R, Arap W, Akimzhanov AM, Febbraio M, Kolonin MG (2021) Fatty acid mobilization from adipose tissue is mediated by CD36 posttranslational modifications and intracellular trafficking. JCI Insight 6:e147057 Goldberg IJ, Eckel RH, Abumrad NA (2009) Regulation of fatty acid uptake into tissues: lipoprotein lipaseand CD36-mediated pathways. J Lipid Res 50(Suppl):S86–90 Gonzales AM, Orlando RA (2007) Role of adipocyte-derived lipoprotein lipase in adipocyte hypertrophy. Nutr Metab 4:22 Grundy SM (2015) Adipose tissue and metabolic syndrome: too much, too little or neither. Eur J Clin Invest 45:1209–1217 Hagemann T, Czechowski P, Ghosh A, Sun W, Dong H, Noe F, Wolfrum C, Bluher M, Hoffmann A (2023) Laminin alpha4 expression in human adipose tissue depots and its association with obesity and obesity related traits. Biomedicines 11:2806 Hallgren J, Tsirigos KD, Pedersen MD, Almagro Armenteros JJ, Marcatili P, Nielsen H, Krogh A, Winther O (2022) DeepTMHMM predicts alpha and beta transmembrane proteins using deep neural networks. Preprint at https:// www.biorxiv.org/content/10.1101/2022.04.08.487609v1 Hao JW, Wang J, Guo H, Zhao YY, Sun HH, Li YF, Lai XY, Zhao N, Wang X, Xie C et al (2020) CD36 facilitates fatty acid uptake by dynamic palmitoylationregulated endocytosis. Nat Commun 11:4765 Hofmann SM, Perez-Tilve D, Greer TM, Coburn BA, Grant E, Basford JE, Tschop MH, Hui DY (2008) Defective lipid delivery modulates glucose tolerance and metabolic response to diet in apolipoprotein E-deficient mice. Diabetes 57:5–12 Honorato RV, Trellet ME, Jimenez-Garcia B, Schaarschmidt JJ, Giulini M, Reys V, Koukos PI, Rodrigues J, Karaca E, van Zundert GCP et al (2024) The HADDOCK2.4 web server for integrative modeling of biomolecular complexes. Nat Protoc 19:3219–3241 Hsieh FL, Turner L, Bolla JR, Robinson CV, Lavstsen T, Higgins MK (2016) The structural basis for CD36 binding by the malaria parasite. Nat Commun 7:12837 Huang J, MacKerell AD Jr (2013) CHARMM36 all-atom additive protein force field: validation based on comparison to NMR data. J Comput Chem 34:2135–2145 Jeong DW, Park JW, Kim KS, Kim J, Huh J, Seo J, Kim YL, Cho JY, Lee KW, Fukuda J, Chun YS (2023) Palmitoylation-driven PHF2 ubiquitination remodels lipid metabolism through the SREBP1c axis in hepatocellular carcinoma. Nat Commun 14:6370 Jumper J, Evans R, Pritzel A, Green T, Figurnov M, Ronneberger O, Tunyasuvunakool K, Bates R, Zidek A, Potapenko A et al (2021) Highly accurate protein structure prediction with AlphaFold. Nature 596:583–589 Kerr AG, Sinha I, Dadvar S, Arner P, Dahlman I (2019) Epigenetic regulation of diabetogenic adipose morphology. Mol Metab 25:159–167 Klein S, Gastaldelli A, Yki-Jarvinen H, Scherer PE (2022) Why does obesity cause diabetes? Cell Metab 34:11–20 Kloting N, Fasshauer M, Dietrich A, Kovacs P, Schon MR, Kern M, Stumvoll M, Bluher M (2010) Insulin-sensitive obesity. Am J Physiol Endocrinol Metab 299:E506–515 Kohler G, Milstein C (1975) Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 256:495–497 Kurki MI, Karjalainen J, Palta P, Sipila TP, Kristiansson K, Donner KM, Reeve MP, Laivuori H, Aavikko M, Kaunisto MA et al (2023) FinnGen provides genetic insights from a well-phenotyped isolated population. Nature 613:508–518 Langhardt J, Flehmig G, Kloting N, Lehmann S, Ebert T, Kern M, Schon MR, Gartner D, Lohmann T, Dressler M et al (2018) Effects of weight loss on glutathione peroxidase 3 serum concentrations and adipose tissue expression in human obesity. Obes Facts 11:475–490 Lee J, Cheng X, Swails JM, Yeom MS, Eastman PK, Lemkul JA, Wei S, Buckner J, Jeong JC, Qi Y et al (2016) CHARMM-GUI input generator for NAMD, GROMACS, AMBER, OpenMM, and CHARMM/OpenMM simulations using the CHARMM36 additive force field. J Chem Theory Comput 12:405–413 Lemkul JA (2018) From proteins to perturbed hamiltonians: a suite of tutorials for the GROMACS-2018 molecular simulation package [Article v1.0]. Living J Comput Mol Sci 1:5068 Li Y, Fromme T, Schweizer S, Schottl T, Klingenspor M (2014) Taking control over intracellular fatty acid levels is essential for the analysis of thermogenic function in cultured primary brown and brite/beige adipocytes. EMBO Rep 15:1069–1076 Lomize MA, Pogozheva ID, Joo H, Mosberg HI, Lomize AL (2012) OPM database and PPM web server: resources for positioning of proteins in membranes. Nucleic Acids Res 40:D370–376 Mardinoglu A, Heiker JT, Gartner D, Bjornson E, Schon MR, Flehmig G, Kloting N, Krohn K, Fasshauer M, Stumvoll M et al (2015) Extensive weight loss reveals distinct gene expression changes in human subcutaneous and visceral adipose tissue. Sci Rep 5:14841 Inderjeet Singh et al The EMBO Journal © The Author(s) The EMBO Journal Volume 44 | Issue 18 | September 2025 | 5037 –5065 5053 Downloaded from https://www.embopress.org on November 3, 2025 from IP 2001:a62:331:8d01:b08d:a892:1074:db15. Mauvais-Jarvis F, Clegg DJ, Hevener AL (2013) The role of estrogens in control of energy balance and glucose homeostasis. Endocr Rev 34:309–338 Meng EC, Goddard TD, Pettersen EF, Couch GS, Pearson ZJ, Morris JH, Ferrin TE (2023) UCSF ChimeraX: tools for structure building and analysis. Protein Sci 32:e4792 Mirdita M, Schutze K, Moriwaki Y, Heo L, Ovchinnikov S, Steinegger M (2022) ColabFold: making protein folding accessible to all. Nat Methods 19:679–682 Morigny P, Boucher J, Arner P, Langin D (2021) Lipid and glucose metabolism in white adipocytes: pathways, dysfunction and therapeutics. Nat Rev Endocrinol 17:276–295 Onogi Y, Khalil A, Ussar S (2020) Identification and characterization of adipose surface epitopes. Biochem J 477:2509–2541 Peche VS, Pietka TA, Jacome-Sosa M, Samovski D, Palacios H, Chatterjee-Basu G, Dudley AC, Beatty W, Meyer GA, Goldberg IJ, Abumrad NA (2023) Endothelial cell CD36 regulates membrane ceramide formation, exosome fatty acid transfer and circulating fatty acid levels. Nat Commun 14:4029 Peterson JM, Seldin MM, Tan SY, Wong GW (2014) CTRP2 overexpression improves insulin and lipid tolerance in diet-induced obese mice. PLoS ONE 9:e88535 Picelli S, Faridani OR, Bjorklund AK, Winberg G, Sagasser S, Sandberg R (2014) Full-length RNA-seq from single cells using Smart-seq2. Nat Protoc 9:171–181 Reilly SM, Saltiel AR (2017) Adapting to obesity with adipose tissue inflammation. Nat Rev Endocrinol 13:633–643 Song Y, Milon B, Ott S, Zhao X, Sadzewicz L, Shetty A, Boger ET, Tallon LJ, Morell RJ, Mahurkar A, Hertzano R (2018) A comparative analysis of library prep approaches for sequencing low input translatome samples. BMC Genomics 19:696 Thumuluri V, Almagro Armenteros JJ, Johansen AR, Nielsen H, Winther O (2022) DeepLoc 2.0: multi-label subcellular localization prediction using protein language models. Nucleic Acids Res 50:W228–W234 Urvas L, Chiesa L, Bret G, Jacquemard C, Kellenberger E (2024) Benchmarking AlphaFold-generated structures of chemokine-chemokine receptor complexes. J Chem Inf Model 64:4587–4600 Ussar S, Lee KY, Dankel SN, Boucher J, Haering MF, Kleinridders A, Thomou T, Xue R, Macotela Y, Cypess AM et al (2014) ASC-1, PAT2, and P2RX5 are cell surface markers for white, beige, and brown adipocytes. Sci Transl Med 6:247ra103 Villanueva-Carmona T, Cedo L, Nunez-Roa C, Maymo-Masip E, Vendrell J, Fernandez-Veledo S (2023) Protocol for the in vitro isolation and culture of mature adipocytes and white adipose tissue explants from humans and mice. STAR Protoc 4:102693 Wang J, Li DL, Zheng LF, Ren S, Huang ZQ, Tao Y, Liu Z, Shang Y, Pang D, Guo H et al (2024) Dynamic palmitoylation of STX11 controls injury-induced fatty acid uptake to promote muscle regeneration. Dev Cell 59:384–399.e385 Yu D, Chojnowski G, Rosenthal M, Kosinski J (2023) AlphaPulldown—aPython package for protein-protein interaction screens using AlphaFold-Multimer. Bioinformatics 39:btac749 Zeng Y, Tao N, Chung KN, Heuser JE, Lublin DM (2003) Endocytosis of oxidized low density lipoprotein through scavenger receptor CD36 utilizes a lipid raft pathway that does not require caveolin-1. J Biol Chem 278:45931–45936 Zhang R, Yao F, Gao F, Abou-Samra AB (2012) Nrac, a novel nutritionallyregulated adipose and cardiac-enriched gene. PLoS ONE 7:e46254 Acknowledgements We thank Vignesh Karthikaisamy and Mauricio Berriel Diaz for providing the pRK5-HA-Ubiquitin-WT plasmid. IS was supported by the National Overseas Scholarship from the Government of India. YO received support through the Post-doctoral Fellowship program from The Uehara Memorial Foundation, Japan (201830047) and the Alexander von Humboldt-Stiftung, Germany. LK, CW and XY were supported by China Scholarship Council. DK and IT got funding from the Horizon Europe funding programme under the Marie Skłodowska-Curie Actions Doctoral Networks grant agreement “Explainable AI for Molecules - AiChemist”, no. 101120466. German Mouse Clinic received grant from German Federal Ministry of Education and Research (Infrafrontier grant 01KX1012 to MHdA) and German Center for Diabetes Research (DZD) (MHdA). MB received funding from grants from the DFG Projektnummer 209933838—SFB 1052 (project B1) and by Deutsches Zentrum für Diabetesforschung (DZD, Grant: 82DZD00601). We thank Wenfei Sun and Hua Dong for supporting human adipose tissue RNA sequencing. Author contributions Inderjeet Singh: Conceptualization; Data curation; Formal analysis; Validation; Investigation; Visualization; Methodology; Writing—original draft; Project administration; Writing—review and editing. Yasuhiro Onogi: Conceptualization; Supervision; Investigation; Writing—review and editing. Filipe Menezes: Formal analysis; Investigation; Methodology; Writing—review and editing. Dina Khasanova: Investigation; Methodology; Writing—review and editing. Lingru Kang: Investigation; Methodology; Writing—review and editing. Chenxi Wang: Investigation; Methodology; Writing—review and editing. Julio Ruiz-Trave: Formal analysis; Investigation; Methodology; Writing—review and editing. Sapna Sharma: Formal analysis; Investigation; Methodology; Writing— original draft; Writing—review and editing. Ahmed Khalil: Investigation; Methodology; Writing—review and editing. Valentin K Reichenbach: Formal analysis; Investigation; Visualization; Methodology; Writing—review and editing. Yingzi Shi: Investigation; Methodology; Writing—review and editing. Andrew Flatley: Formal analysis; Validation; Investigation; Methodology; Writing—review and editing. Xiaocheng Yan: Investigation; Methodology; Writing—review and editing. Andreas Israel: Investigation; Methodology; Project administration; Writing—review and editing. Nathalia RV Dragano: Data curation; Formal analysis; Investigation; Methodology; Writing—review and editing. Juan Antonio Aguilar-Pimentel: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Writing—review and editing. Anne Hoffmann: Data curation; Formal analysis; Investigation; Methodology; Writing—review and editing. Adhideb Ghosh: Formal analysis; Investigation; Methodology; Writing—review and editing. Falko Noé: Formal analysis; Investigation; Methodology; Writing—review and editing. Christian Wolfrum: Supervision; Methodology; Writing—review and editing. Sebastian Cucuruz: Formal analysis; Investigation; Methodology; Writing—review and editing. Ann-Christine König: Formal analysis; Investigation; Methodology; Writing— review and editing. Ingo Burtscher: Investigation; Methodology; Writing— review and editing. Stefanie M Hauck: Supervision; Writing—review and editing. Heiko Lickert: Supervision; Writing—review and editing. Susanna M Hofmann: Supervision; Methodology; Writing—review and editing. Regina Feederle: Supervision; Methodology; Writing—review and editing. Sonja C Schriever: Investigation; Writing—review and editing. Rene HernandezBautista: Investigation; Methodology; Writing—review and editing. Gencer Sancar: Funding acquisition; Investigation; Methodology; Writing—review and editing. Alberto Cebrian-Serrano: Supervision; Investigation; Methodology; Writing—review and editing. Igor Tetko: Supervision; Writing—review and editing. Helmut Fuchs: Supervision; Methodology; Writing—review and editing. Valérie Gailus-Durner: Supervision; Methodology; Writing—review and editing. Matthias Blüher: Data curation; Formal analysis; Supervision; Methodology; Writing—review and editing. Martin Hrabé de Angelis: Supervision; Writing— review and editing. Siegfried Ussar: Conceptualization; Formal analysis; Supervision; Funding acquisition; Validation; Visualization; Writing—original draft; Project administration; Writing—review and editing. Source data underlying figure panels in this paper may have individual authorship assigned. Where available, figure panel/source data authorship is The EMBO Journal Inderjeet Singh et al 5054 The EMBO Journal Volume 44 | Issue 18 | September 2025 | 5037 –5065 © The Author(s) Downloaded from https://www.embopress.org on November 3, 2025 from IP 2001:a62:331:8d01:b08d:a892:1074:db15. listed in the following database record: biostudies:S-SCDT-10_1038-S44318025-00520-2. Funding Open Access funding enabled and organized by Projekt DEAL. Disclosure and competing interests statement MB received honoraria as a consultant and speaker from Amgen, AstraZeneca, Bayer, Boehringer-Ingelheim, Lilly, Novo Nordisk, Novartis, and Sanofi. All other authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. Creative Commons Public Domain Dedication waiver http://creativecommons.org/publicdomain/zero/1.0/ applies to the data associated with this article, unless otherwise stated in a credit line to the data, but does not extend to the graphical or creative elements of illustrations, charts, or figures. This waiver removes legal barriers to the re-use and mining of research data. According to standard scholarly practice, it is recommended to provide appropriate citation and attribution whenever technically possible. © The Author(s) 2025 Inderjeet Singh et al The EMBO Journal © The Author(s) The EMBO Journal Volume 44 | Issue 18 | September 2025 | 5037 –5065 5055 Downloaded from https://www.embopress.org on November 3, 2025 from IP 2001:a62:331:8d01:b08d:a892:1074:db15. Expanded View Figures Figure EV1. NRAC deletion does not affect glucose metabolism or insulin sensitivity in chow diet fed mice. Validation of NRAC antibody by (A) western blotting of HEK293T cells transfected with either NRAC-GFP or GFP plasmids. Untransfected cells were used as control. (B) Western blotting to detect endogenous NRAC in BAT, SCF and PGF and (C) immunostaining of subcutaneous differentiated mature adipocytes to detect endogenous NRAC (scale bar 10 µm), and (D) Immunoprecipitation of NRAC from perigonadal fat (PGF). (E–H) mRNA expression of Tnfa and Il6 in PGF and SCF of male mice (n=6 wt, 5ko), (I) weights of liver, heart and pancreas (n=9 wt, 5ko). (J) Glucose tolerance test and area under the curve (age 8 weeks). (K) Insulin tolerance test (ITT) and area under the curve (AUC) in 9 weeks old male mice. (L) Percentage of HbA1c in 12-week-old male mice. (M) Body weight development of chow diet fed female wt and Nracko mice. (N) Fat and (O) lean mass of 15-week-old chow diet fed female mice. (P) Glucose tolerance test and area under the curve of 8 weeks old female mice. (Q) %HbA1c in 12-week-old female mice. (R) Insulin tolerance test (ITT) and area under the curve (AUC) in 9-week-old female mice. (S) H&E staining of SCF and PGF tissues of female mice (Age 23 weeks), and (T) quantification of lipid droplets. n=9wt and 7ko for males and 5wt and 7ko for females. (U) mRNA expression of Nrac from BAT, SCF and PGF of male and female mice (n=4). Data are shown as mean ± SEM. *P< 0.05, **P< 0.01, ***P< 0.001 by Student’sttest or one or two-way ANOVA. Source data are available online for this figure. The EMBO Journal Inderjeet Singh et al 5056 The EMBO Journal Volume 44 | Issue 18 | September 2025 | 5037 –5065 © The Author(s) Downloaded from https://www.embopress.org on November 3, 2025 from IP 2001:a62:331:8d01:b08d:a892:1074:db15. AC PGF SCF 0 5 10 15 20 25 Body weight females Body weight (g) WT Nrac KO Age (in weeks) 810 16 12 14 18 20 22 0 100 200 300 400 GTT females Time (min) Glucose (mg/dl) WT Nracko 030 60 90 120 WT Nracko 0 200 400 600 800 AUC Arbitrary units 0.7045 WT Nracko 0 1 2 3 4 5 Fat mass Weight (g) 0.2297 WT Nracko 0 5 10 15 20 25 Lean mass Weight (g) 0.5698 WT Nracko 0 1 2 3 4 5 HbA1c HbA1c levels (%) 0.1701 WT Nracko 0 20000 40000 60000 PGF 0.4414 0 50 100 150 ITT females CD Time (min) Glucose (mg/dl) 030 60 90 120 0306090120 0 100 200 300 400 GTT Time (min) Glucose (mg/dl) J K Q N T R S LM Control GFP NRAC-GFP Vinuculin NRAC-GFP GFP NRAC-GFP IB: NRAC IB: GFP WT Nracko 0 100 200 300 400 AUC Arbitrary units 0.7911 EF GH WT Nracko WT Nracko 0 5000 10000 15000 20000 AUC Arbitrary units 0.0581 D B NRAC βACTIN BAT SCF PGF WT Nracko WT WT Nracko Nracko WT Nracko 0.00 0.02 0.04 0.06 0.08 0.10 PGF Tnfa mRNA expression Norm Tbp 0.2593 WT Nrac ko 0.00 0.01 0.02 0.03 PGF Il6 mRNA expression Norm Tbp 0.1503 WT Nrac ko 0.0 0.2 0.4 0.6 SCF Tnfa mRNA expression Norm Tbp 0.3843 WT Nrac ko 0.00 0.01 0.02 0.03 SCF Il6 mRNA expression Norm Tbp 0.9539 OP WT Nracko 0.0 0.5 1.0 1.5 2.0 Liver Weight (g) 0.4241 WT Nracko 0.00 0.05 0.10 0.15 0.20 0.25 Heart Weight (g) 0.6220 WT Nracko 0.0 0.1 0.2 0.3 0.4 0.5 Pancreas Weight (g) 0.9514 I 0 50 100 150 ITT Glucose(mg/dl) 30 60 90 120 0Time (min) 0 10 20 30 Nrac mRNA Expression norm. Tbp 0.9974 0.7304 0.5834 BAT SCF PGF Males Females U WT WT Nracko Nracko NRAC GAPDH Input IP: NRAC WTNrac ko 0 2 4 6 HbA1c HbA1c levels (%) 0.3259 WT Nrac ko 0 100 200 300 400 AUC Arbitrary units 0.4228 WT Nracko WT Nracko 0 2000 4000 6000 8000 10000 SCF Arbitrary units 0.0685 Arbitrary units Inderjeet Singh et al The EMBO Journal © The Author(s) The EMBO Journal Volume 44 | Issue 18 | September 2025 | 5037 –5065 5057 Downloaded from https://www.embopress.org on November 3, 2025 from IP 2001:a62:331:8d01:b08d:a892:1074:db15. (NRAC/DAPI) The EMBO Journal Inderjeet Singh et al 5058 The EMBO Journal Volume 44 | Issue 18 | September 2025 | 5037 –5065 © The Author(s) Downloaded from https://www.embopress.org on November 3, 2025 from IP 2001:a62:331:8d01:b08d:a892:1074:db15. Figure EV2. NRAC deletion does not affect energy homeostasis, food intake and substrate utilization in mice under chow diet. (A) Oxygen consumption rates (OCR) in wt and Nracko differentiated adipocytes. Injection strategy included BSA or BSA conjugated oleate and palmitate (100 µM each) (port 1), oligomycin (port 2), FCCP (port 3) and Antimycin A and rotenone in port 4. (B) OCR calculated by subtracting OCR after fatty acid (FA) stimulation from basal respiration. (C) Proton leak calculated by subtracting oligomycin OCR from non-mitochondrial OCR (n=9). (D) Immunostaining of BAT from wt mice either overnight starved (16 h), random fed, acute cold, (4 h) and CL-316243 injected mice (n=3), age 12 weeks (scale 100 µm). (E) Basal metabolic rate (BMR). (F) Energy expenditure and respiratory exchange ratio (RER), (G,H) in saline treated mice, (I,J) in CL316243 treated mice and (K,L) upon acute cold exposure (n=6 WT and 8KO, Age 16 weeks). Data are shown as mean ± SEM. *P< 0.05, **P< 0.01, ***P< 0.001 by Student’sttest or one or two way ANOVA. In figures (E–G,I,K) the data are plotted as linear regression. Source data are available online for this figure. Inderjeet Singh et al The EMBO Journal © The Author(s) The EMBO Journal Volume 44 | Issue 18 | September 2025 | 5037 –5065 5059 Downloaded from https://www.embopress.org on November 3, 2025 from IP 2001:a62:331:8d01:b08d:a892:1074:db15. A CPZ Vehicle WT WT Nracko Nracko NRAC CAV1 CD36 Merge B VINCULIN CAV1 CD36 NRAC BAT SCF Input IP: NRAC IP: NRACInput WT Nracko Nracko WT WT WT Nracko Nracko Figure EV3. NRAC regulates CD36/Cav1 interaction in adipocytes. (A) Immunoprecipitation of NRAC from BAT and SCF. (B) Immunostaining of differentiated immortalized preadipocytes treated with vehicle (0.05% ethanol) or Chlorpromazine (CPZ) (10 μM) for 1 h (scale 25 µm). Source data are available online for this figure. The EMBO Journal Inderjeet Singh et al 5060 The EMBO Journal Volume 44 | Issue 18 | September 2025 | 5037 –5065 © The Author(s) Downloaded from https://www.embopress.org on November 3, 2025 from IP 2001:a62:331:8d01:b08d:a892:1074:db15. Figure EV4. First transmembrane domain of NRAC is required for NRAC-CD36 interaction. (A) MD-simulations of human NRAC and CD36. The full simulation block is shown, with the central grey stripe representing the lipid bilayer membrane. Water molecules are depicted as red and white particles, while Na⁺and Cl−ions appear as pink spheres at a physiological concentration of 0.15 M. The bright green structures correspond to the proteins CD36 and NRAC, with CD36 embedded in the upper membrane leaflet and NRAC in the lower leaflet, and the zoomed in area of the same region. (B) CD36–NRAC complex is embedded in the membrane, with CD36 located on top and NRAC below. The membrane is shown in grey. (C) The CD36–NRAC complex is embedded in the membrane, with CD36 on top and NRAC below. Phospholipids in the membrane are represented as grey circles. Source data are available online for this figure. Inderjeet Singh et al The EMBO Journal © The Author(s) The EMBO Journal Volume 44 | Issue 18 | September 2025 | 5037 –5065 5061 Downloaded from https://www.embopress.org on November 3, 2025 from IP 2001:a62:331:8d01:b08d:a892:1074:db15.