Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

snRNA-seq reveals a subpopulation of adipocytes that regulates thermogenesis

Abstract

Adipose tissue is usually classified on the basis of its function as white, brown or beige (brite)1. It is an important regulator of systemic metabolism, as shown by the fact that dysfunctional adipose tissue in obesity leads to a variety of secondary metabolic complications2,3. In addition, adipose tissue functions as a signalling hub that regulates systemic metabolism through paracrine and endocrine signals4. Here we use single-nucleus RNA-sequencing (snRNA-seq) analysis in mice and humans to characterize adipocyte heterogeneity. We identify a rare subpopulation of adipocytes in mice that increases in abundance at higher temperatures, and we show that this subpopulation regulates the activity of neighbouring adipocytes through acetate-mediated modulation of their thermogenic capacity. Human adipose tissue contains higher numbers of cells of this subpopulation, which could explain the lower thermogenic activity of human compared to mouse adipose tissue and suggests that targeting this pathway could be used to restore thermogenic activity.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.

from$8.99

All prices are NET prices.

Fig. 1: Identification of adipocyte populations in BAT.
Fig. 2: A distinct population of adipocytes is present in mouse iBAT and ingWAT and in human deep-neck BAT.
Fig. 3: Reduction of ALDH1A1 expression in mature adipocytes of iBAT promotes thermogenesis and whole-body energy expenditure in mice.
Fig. 4: ALDH1A1 regulates brown adipocyte thermogenic capacity through acetate.

Data availability

All RNA sequencing (RNA-seq) data that support this finding of this study have been deposited in ArrayExpress, with the accession codes E-MTAB-8561 for snRNA-seq of mouse interscapular brown adipocytes at room temperature by Smart-seq2; E-MTAB-8562 for snRNA-seq of mouse interscapular brown adipocytes in room-temperature, cold-exposure and thermoneutral conditions by 10X sequencing; E-MTAB-8564 for snRNA-seq of human BAT cells; E-MTAB-9199 for snRNA-seq of human subcutaneous WAT cells; and E-MTAB-9192 for bulk RNA-seq of mouse iBAT, subscWAT and ingWAT. The datasets can be explored interactively at https://batnetwork.org/. Please address correspondence and requests for materials to C.W. and requests for bioinformatic information to W.S. Source data are provided with this paper.

References

  1. 1.

    Cannon, B. & Nedergaard, J. Brown adipose tissue: function and physiological significance. Physiol. Rev. 84, 277–359 (2004).

    CAS  Google Scholar 

  2. 2.

    Lauby-Secretan, B. et al. Body fatness and cancer—viewpoint of the IARC working group. N. Engl. J. Med. 375, 794–798 (2016).

    Google Scholar 

  3. 3.

    Hajer, G. R., van Haeften, T. W. & Visseren, F. L. J. Adipose tissue dysfunction in obesity, diabetes, and vascular diseases. Eur. Heart J. 29, 2959–2971 (2008).

    CAS  Google Scholar 

  4. 4.

    Scheele, C. & Wolfrum, C. Brown adipose cross talk in tissue plasticity and human metabolism. Endocr. Rev. 41, 53–65 (2019).

    Google Scholar 

  5. 5.

    Roh, H. C. et al. Warming induces significant reprogramming of beige, but not brown, adipocyte cellular identity. Cell Metab. 27, 1121–1137 (2018).

    CAS  Google Scholar 

  6. 6.

    Rosenwald, M., Perdikari, A., Rülicke, T. & Wolfrum, C. Bi-directional interconversion of brite and white adipocytes. Nat. Cell Biol. 15, 659–667 (2013).

    CAS  Google Scholar 

  7. 7.

    Rosenwald, M. & Wolfrum, C. The origin and definition of brite versus white and classical brown adipocytes. Adipocyte 3, 4–9 (2014).

    CAS  Google Scholar 

  8. 8.

    Jung, S. M., Sanchez-Gurmaches, J. & Guertin, D. A. in Brown Adipose Tissue. Handbook of Experimental Pharmacology Vol. 251 3–36 (2018).

  9. 9.

    Min, S. Y. et al. Diverse repertoire of human adipocyte subtypes develops from transcriptionally distinct mesenchymal progenitor cells. Proc. Natl Acad. Sci. USA 116, 17970–17979 (2019).

    CAS  Google Scholar 

  10. 10.

    Song, A. et al. Low- and high-thermogenic brown adipocyte subpopulations coexist in murine adipose tissue. J. Clin. Invest. 130, 247–257 (2020).

    CAS  Google Scholar 

  11. 11.

    Rajbhandari, P. et al. Single cell analysis reveals immune cell-adipocyte crosstalk regulating the transcription of thermogenic adipocytes. eLife 8, e49501 (2019).

    Google Scholar 

  12. 12.

    Habib, N. et al. Massively parallel single-nucleus RNA-seq with DroNc-seq. Nat. Methods 14, 955–958 (2017).

    CAS  Google Scholar 

  13. 13.

    Habib, N. et al. Div-Seq: single-nucleus RNA-seq reveals dynamics of rare adult newborn neurons. Science 353, 925–928 (2016).

    ADS  CAS  Google Scholar 

  14. 14.

    Picelli, S. et al. Full-length RNA-seq from single cells using Smart-seq2. Nat. Protocols 9, 171–181 (2014).

    CAS  Google Scholar 

  15. 15.

    Shinoda, K. et al. Genetic and functional characterization of clonally derived adult human brown adipocytes. Nat. Med. 21, 389–394 (2015).

    CAS  Google Scholar 

  16. 16.

    Schwalie, P. C. et al. A stromal cell population that inhibits adipogenesis in mammalian fat depots. Nature 559, 103–108 (2018).

    ADS  CAS  Google Scholar 

  17. 17.

    Merrick, D. et al. Identification of a mesenchymal progenitor cell hierarchy in adipose tissue. Science 364, eaav2501 (2019).

    CAS  Google Scholar 

  18. 18.

    Gaublomme, J. T. et al. Nuclei multiplexing with barcoded antibodies for single-nucleus genomics. Nat. Commun. 10, 2907 (2019).

    ADS  Google Scholar 

  19. 19.

    Aran, D. et al. Reference-based analysis of lung single-cell sequencing reveals a transitional profibrotic macrophage. Nat. Immunol. 20, 163–172 (2019).

    CAS  Google Scholar 

  20. 20.

    Sebastian, B. M. et al. Identification of a cytochrome P4502E1/Bid/C1q-dependent axis mediating inflammation in adipose tissue after chronic ethanol feeding to mice. J. Biol. Chem. 286, 35989–35997 (2011).

    CAS  Google Scholar 

  21. 21.

    Lee, Y.-H., Petkova, A. P., Konkar, A. A. & Granneman, J. G. Cellular origins of cold-induced brown adipocytes in adult mice. FASEB J. 29, 286–299 (2015).

    CAS  Google Scholar 

  22. 22.

    Cinti, S. Obesity, Type 2 Diabetes and the Adipose Organ (Springer, 2018).

  23. 23.

    Keijer, J., Li, M. & Speakman, J. R. What is the best housing temperature to translate mouse experiments to humans? Mol. Metab. 25, 168–176 (2019).

    CAS  Google Scholar 

  24. 24.

    Perdikari, A. et al. BATLAS: deconvoluting brown adipose tissue. Cell Rep. 25, 784–797 (2018).

    CAS  Google Scholar 

  25. 25.

    Kiefer, F. W. et al. Retinaldehyde dehydrogenase 1 regulates a thermogenic program in white adipose tissue. Nat. Med. 18, 918–925 (2012).

    CAS  Google Scholar 

  26. 26.

    Eguchi, J. et al. Transcriptional control of adipose lipid handling by IRF4. Cell Metab. 13, 249–259 (2011).

    CAS  Google Scholar 

  27. 27.

    Ziouzenkova, O. et al. Retinaldehyde represses adipogenesis and diet-induced obesity. Nat. Med. 13, 695–702 (2007).

    CAS  Google Scholar 

  28. 28.

    Wu, J. et al. Beige adipocytes are a distinct type of thermogenic fat cell in mouse and human. Cell 150, 366–376 (2012).

    CAS  Google Scholar 

  29. 29.

    Yang, K. et al. Aldehyde dehydrogenase 1 a1 regulates energy metabolism in adipocytes from different species. Xenotransplantation 24, e12318 (2017).

    Google Scholar 

  30. 30.

    Yasmeen, R. et al. Autocrine function of aldehyde dehydrogenase 1 as a determinant of diet- and sex-specific differences in visceral adiposity. Diabetes 62, 124–136 (2013).

    CAS  Google Scholar 

  31. 31.

    Alvarez, R. et al. A novel regulatory pathway of brown fat thermogenesis. Retinoic acid is a transcriptional activator of the mitochondrial uncoupling protein gene. J. Biol. Chem. 270, 5666–5673 (1995).

    CAS  Google Scholar 

  32. 32.

    Matthias, A. et al. Thermogenic responses in brown fat cells are fully UCP1-dependent. J. Biol. Chem. 275, 25073–25081 (2000).

    CAS  Google Scholar 

  33. 33.

    Tourniaire, F. et al. All-trans retinoic acid induces oxidative phosphorylation and mitochondria biogenesis in adipocytes. J. Lipid Res. 56, 1100–1109 (2015).

    CAS  Google Scholar 

  34. 34.

    Li, H., Toth, E. & Cherrington, N. J. Alcohol metabolism in the progression of human nonalcoholic steatohepatitis. Toxicol. Sci. 164, 428–438 (2018).

    CAS  Google Scholar 

  35. 35.

    Mews, P. et al. Alcohol metabolism contributes to brain histone acetylation. Nature 574, 717–721 (2019).

    ADS  CAS  Google Scholar 

  36. 36.

    Liu, X. et al. Acetate production from glucose and coupling to mitochondrial metabolism in mammals. Cell 175, 502–513 (2018).

    CAS  Google Scholar 

  37. 37.

    Balaz, M. et al. Inhibition of mevalonate pathway prevents adipocyte browning in mice and men by affecting protein prenylation. Cell Metab. 29, 901–916 (2019).

    CAS  Google Scholar 

  38. 38.

    Husted, A. S., Trauelsen, M., Rudenko, O., Hjorth, S. A. & Schwartz, T. W. GPCR-mediated signaling of metabolites. Cell Metab. 25, 777–796 (2017).

    CAS  Google Scholar 

  39. 39.

    Hernández, M. A. G., Canfora, E. E., Jocken, J. W. E. & Blaak, E. E. The short-chain fatty acid acetate in body weight control and insulin sensitivity. Nutrients 11, 1943 (2019).

    Google Scholar 

  40. 40.

    Müller, M. et al. Circulating but not faecal short-chain fatty acids are related to insulin sensitivity, lipolysis and GLP-1 concentrations in humans. Sci. Rep. 9, 12515 (2019).

    ADS  Google Scholar 

  41. 41.

    Drokhlyansky, E. et al. The human and mouse enteric nervous system at single-cell resolution. Science 182, 1606–1622.e23 (2020).

    CAS  Google Scholar 

  42. 42.

    Chen, W., Gardeux, V., Meireles-Filho, A. & Deplancke, B. Profiling of single-cell transcriptomes. Curr. Protoc. Mouse Biol. 7, 145–175 (2017).

    CAS  Google Scholar 

  43. 43.

    Satija, R., Farrell, J. A., Gennert, D., Schier, A. F. & Regev, A. Spatial reconstruction of single-cell gene expression data. Nat. Biotechnol. 33, 495–502 (2015).

    CAS  Google Scholar 

  44. 44.

    Qiu, X. et al. Reversed graph embedding resolves complex single-cell trajectories. Nat. Methods 14, 979–982 (2017).

    CAS  Google Scholar 

  45. 45.

    Ihaka, R. & Gentleman, R. R: a language for data analysis and graphics. J. Comput. Graph. Stat. 5, 299–314 (1996).

    Google Scholar 

  46. 46.

    Fleming, S. J., Marioni, J. C. & Babadi, M. CellBender remove-background: a deep generative model for unsupervised removal of background noise from scRNA-seq datasets. Preprint at https://doi.org/10.1101/791699 (2019).

  47. 47.

    Wolock, S. L., Lopez, R. & Klein, A. M. Scrublet: computational identification of cell doublets in single-cell transcriptomic data. Cell Syst. 8, 281–291 (2019).

    CAS  Google Scholar 

  48. 48.

    McGinnis, C. S., Murrow, L. M. & Gartner, Z. J. DoubletFinder: doublet detection in single-cell RNA sequencing data using artificial nearest neighbors. Cell Syst. 8, 329–337 (2019).

    CAS  Google Scholar 

  49. 49.

    Stuart, T. et al. Comprehensive integration of single-cell data. Cell 177, 1888–1902 (2019).

    CAS  Google Scholar 

  50. 50.

    Modica, S. et al. BMP4 promotes a brown to white-like adipocyte shift. Cell Rep. 16, 2243–2258 (2016).

    CAS  Google Scholar 

  51. 51.

    Rosenwald, M., Efthymiou, V., Opitz, L. & Wolfrum, C. SRF and MKL1 independently inhibit brown adipogenesis. PLoS One 12, e0170643 (2017).

    Google Scholar 

  52. 52.

    Kane, M. A. & Napoli, J. L. Quantification of endogenous retinoids. Methods Mol. Biol. 652, 1–54 (2010).

    CAS  Google Scholar 

  53. 53.

    Sun, W. et al. Cold-induced epigenetic programming of the sperm enhances brown adipose tissue activity in the offspring. Nat. Med. 24, 1372–1383 (2018).

    CAS  Google Scholar 

  54. 54.

    Cinti, S., Cigolini, M., Morroni, M. & Zingaretti, M. C. S-100 protein in white preadipocytes: an immunoelectronmicroscopic study. Anat. Rec. 224, 466–472 (1989).

    CAS  Google Scholar 

  55. 55.

    Pliner, H. A., Shendure, J. & Trapnell, C. Supervised classification enables rapid annotation of cell atlases. Nat. Methods 16, 983–986 (2019).

    CAS  Google Scholar 

  56. 56.

    Yu, G., Wang, L.-G., Han, Y. & He, Q.-Y. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS 16, 284–287 (2012).

    CAS  Google Scholar 

  57. 57.

    Enomoto, M., Bunge, M. B. & Tsoulfas, P. A multifunctional neurotrophin with reduced affinity to p75NTR enhances transplanted Schwann cell survival and axon growth after spinal cord injury. Exp. Neurol. 248, 170–182 (2013).

    CAS  Google Scholar 

Download references

Acknowledgements

We are grateful to R. Freimann for assistance with nuclei FACS; E. Yángüez for assistance with 10X and Smart-seq2 experiments; E. D. Rosen for discussions and suggestions; D. Peleg for discussions and suggestions on HPLC analysis; O. Ashenberg, Y. Shen, Y. He, U. Ghoshdastider, G. Tan, B. Deplancke, P. Rainer and T. Wang for comments on the bioinformatics analyses; and W. Koppenol and E. Kiehlmann for histology tissue sections. Data produced and analysed in this paper were generated in collaboration with the Functional Genomics Center Zurich, the Cytometry Facility of University of Zurich and the Scientific Center for Optical and Electron Microscopy of ETH. The work was supported by the Swiss National Science Foundation (SNSF 185011 to C.W.).

Author information

Affiliations

Authors

Contributions

W.S. conceived the study; W.S. and C.W. designed the study; W.S. and H.D. performed all of the experimental work except for that described below. W.S. analysed the transcriptome data with input from A.R.; W.S., H.D., M.B., Z.K. and J.U. collected BAT from patients; P.S. performed surgery for collection of human BAT; W.S., H.D., M.S. and E.D. developed the nuclei acquisition methods; C.W. and W.S. acquired retinaldehyde quantification data; L.B. acquired human MADS data; T.W.S. and A.S.H. advised and oversaw Gpr43 work; G.C., A.G. and S.C. acquired immunoelectron microscope pictures; and G.R. acquired Optifast clinical data. C.W., W.S. and A.R. wrote the manuscript. H.D., M.B., S.C., L.D. and A.S.H. helped with editing of the manuscript.

Corresponding authors

Correspondence to Wenfei Sun or Christian Wolfrum.

Ethics declarations

Competing interests

A.R. is a co-founder and equity holder of Celsius Therapeutics, an equity holder in Immunitas Therapeutics and a scientific advisory board member of Thermo Fisher Scientific, Syros Pharmaceuticals, Asimov and Neogene Therapeutics. As of 1 August 2020, A.R. is an employee of Genentech.

Additional information

Peer review information Nature thanks Patrick Seale and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data figures and tables

Extended Data Fig. 1 Identification of subpopulations of mouse brown adipocytes.

a, Schematic illustration of adipocyte snRNA-seq and nuclei FACS plot of AdipoCre−/−NucRed and AdipoCre−/+NucRed. bf, snRNA-seq of 377 adipocyte nuclei from iBAT at room temperature, by Smart-seq2 protocol, yielding 1,999 genes (median) detected. b, Unsupervised clustering shown as UMAP. c, Feature plots for Auts2, Chst1, Cyp2e1, Atp2b4, Adipoq, Lipe, Cidec, Plin1, Ucp1, Cidea, Ppargc1a, Syne2, Ly6a and Cd34. d, e, Trajectory analysis of brown adipocyte nuclei by RNA velocity (d) and by Monocle (e). f, Heat map of signature genes for each population of brown adipocyte nuclei. gk, snRNA-seq of 8,827 adipocyte nuclei from iBAT at room temperature, by 10X protocol. g, Unsupervised clustering shown as UMAP. h, Violin plot of Pde3a, Cish, Atp5e, Cyp2e1 i, Feature plots for Cyp2e1, Atp2b4, Aldh1a1, Auts2, Ucp1, Cidea, Ppargc1a, Syne2, Adipoq, Lipe, Cidec and Plin1. j, Violin plot of Cd34, Ly6a and Pdgfra. k, Heat map of signature genes for each population. lo, snRNA-seq of 6,560 adipocytes in BAT at thermoneutrality, by 10X protocol. l, Unsupervised clustering shown as UMAP. m, Violin plot of Kng2, Plcb1, Atp5e, Ryr1 and Dcn. n, Feature plots of Adipoq, Lipe, Cidec, Plin1, Cyp2e1, Auts2, Aldh1a1, Atp2b4, Ucp1, Cidea, Ppargc1a and Syne2. o, Heat map of signature genes for each population. p, q, snRNA-seq of 11,074 adipocytes from mouse iBAT in cold-exposure conditions. p, Unsupervised clustering shown as UMAP. q, Heat map of signature genes for each population. This figure is related to Fig. 1.

Extended Data Fig. 2 Dynamics of subpopulations of mouse brown adipocytes under thermoneutral, room-temperature and cold-exposure conditions.

a, Mouse iBAT in the cold-exposure condition; feature plots of Ucp1, Cidea, Ppargc1a, Syne2, Adipoq, Lipe, Cidec, Plin1, Cyp2e1, Auts2, Aldh1a1 and Atp2b4. bl, Integrated analysis of mouse iBAT 10x snRNA-seq data. b, UMAP plots of integrated mouse adipocyte nuclei from iBAT grouped by different conditions. c, Pseudotime plot of integrated mouse adipocyte nuclei from iBAT by Monocle. d, e, Feature plots for brown markers Ucp1, Ppargc1a (d) and P4 markers Cyp2e1, Aldh1a1 (e), split by different conditions. f, P-RT-4, P-CE-4 and P-TN-4 cells shown in the integrated UMAP plot, split by different conditions. g, h, Feature plots for brown/adipocyte markers Ucp1, Ppargc1a, Adipoq and Pparg (g) and P4 markers Cyp2e1, Gnai1, Aldh1a1 and Trhde (h). il, P4_score calculates the summation of Aldh1a1, Cyp2e1, Auts2, Trhde, Ptger3, Gnai1, Asph and Cachd1 for each cell. Feature plots for P4_score (i), split by conditions (k); violin plots for P4_score (j), split by different conditions (l). m, Feature plots for CCDC80, PDGFRA, PDGFRB, PPARG, CD3D, CD3E, C1QB, VWF, ALDH1A1, PTGER3, GNAI1 and TRHDE of snRNA-seq for human deep-neck BAT. This figure is related to Fig. 1.

Extended Data Fig. 3 Identification of subpopulations of human brown and white adipocytes.

ad, Feature or violin plots for P4_score (a), UMI and gene per nucleus (b) or annotation of adipocytes by Garnett classifier55 (c) and heat map (d) of snRNA-seq for human deep-neck BAT. P4_score calculates the sum of ALDH1A1, CYP2E1, AUTS2, TRHDE, PTGER3, GNAI1, ASPH and CACHD1 for each cell. eg, Feature or violin plots for CIDEA, PPARGC1A, PPARGC1B, GK, ALDH1A1, PTGER3, GNAI1 and TRHDE expression (e) or P4_score (f) and heat map (g) of human brown adipocyte populations. h, UCP1 mRNA level of deep-neck supraclavicular BAT and subcutaneous WAT for each individual. n = 16 individuals, t = 2.2. i, Unsupervised clustering of 2,438 nuclei from human subcutaneous WAT, yielding 608 (median) genes; annotation by SingleR. jl, Feature or violin plots for P4_score (j) or PPARGC1A, ADIPOQ, PPARG, PDGFRA, PECAM1, VWF, PTPRC, IL7R, CFD, DCLK1, CNTNAP2 and COL19A1 expression (k) and heat map (l) of snRNA-seq for human subcutaneous WAT. Statistical significance was calculated using two-tailed paired t-test (h). This figure is related to Fig. 1.

Extended Data Fig. 4 CYP2E1 marks a specific subpopulation of adipocytes in mouse and human adipose tissue.

a, CYP2E1 immunohistochemical staining in iBAT at room temperature. Red arrows indicate CYP2E1-positive staining. b, Quantification of CYP2E1+ cells with multilocular and unilocular morphology in the interscapular adipose tissue. n = 5 independent experiments, t = 8.0. ce, Bulk RNA-seq of subscWAT, iBAT and ingWAT. n = 3 mice. c, Biological pathways enriched in iBAT over subscWAT, analysed by clusterProfiler56, which calculates a P value using the hypergeometric distribution adjusted for multiple comparison; d, Heat map for subscWAT, iBAT and ingWAT of the top-40 P4 markers. e, RNA expression level of P4 and adipocyte markers in bulk RNA-seq of iBAT and subscWAT. f, mRNA expression levels of Cyp2e1 and Ucp1 in mouse iBAT kept in room-temperature, cold-exposure and thermoneutral conditions. n = 8 mice, FCyp2e1 = 18.4, FUcp1 = 47.6. g, Quantification of CYP2E1+ cells in mouse iBAT in room-temperature, cold-exposure or thermoneutral conditions. n = 6 independent experiments, F = 112.2. h, CYP2E1 immunofluorescence staining in mouse iBAT after cold exposure. i, CYP2E1 immunofluorescence staining in mouse iBAT at thermoneutrality. j, CYP2E1 immunofluorescence staining in mouse ingWAT at room temperature. k, CYP2E1 immunofluorescence staining in mouse ingWAT after cold exposure. l, CYP2E1 immunofluorescence staining in mouse visWAT at room temperature. m, Quantification of CYP2E1+ cells in mouse ingWAT and visWAT at room temperature. n = 5 independent experiments. n, H&E staining of human deep-neck BAT. Two representative images from six individuals. o, Immunofluorescence staining of CYP2E1 and UCP1 in human deep-neck BAT.p, Immunohistochemical staining of CYP2E1 in human deep-neck BAT for individuals 1, 3, 9, 11, 12 and 14. Red arrows indicate CYP2E1-positive staining. Data are mean average ± s.e.m. Statistical significance was calculated using a two-tailed unpaired Student’s t-test (b) or one-way ANOVA (f, g). Scale bars, 50 μm. This figure is related to Fig. 2

Source data.

Extended Data Fig. 5 ALDH1A1 and CYP2E1 are specific markers for the P4 adipocyte subpopulation.

a, Immunofluorescence co-staining of CYP2E1 and ALDH1A1 in mouse iBAT at room temperature. b, ALDH1A1 protein levels in mature adipocytes and SVF of three adipose tissue depots. n = 7 mice, F = 78.06, df = 41. c, Immunofluorescence staining of ALDH1A1 in the interscapular adipose tissue of Ucp1-GFP transgenic mice at room temperature. d, Aldh1a1 mRNA expression levels of mouse iBAT in room-temperature, cold-exposure or thermoneutral conditions. n = 8 mice, F = 25.1. e, ALDH1A1 protein expression levels of mouse iBAT in room-temperature, cold-exposure or thermoneutral conditions. n = 4 mice (except n = 5 for the room-temperature cohort), F = 75.6. f, Schematic map of the AAV knockdown constructs. g, Confocal images of mouse iBAT from AAV-injected mice. h, Immunoblot of ALDH1A1 in ingWAT (n = 7 mice, t = 2.23) and visWAT (n = 7 mice, t = 0.53) after injection of shRNA AAV into iBAT. Data are mean average ± s.e.m. Statistical significance was calculated using a one-way ANOVA (b, d, e) or two-tailed unpaired Student’s t-test (h). Scale bars, 50 μm. See gel source data in Supplementary Fig. 1a, d, e. This figure is related to Fig. 3

Source data.

Extended Data Fig. 6 Reduction of ALDH1A1 expression in BAT regulates UCP1 expression in BAT.

a, Immunofluorescence staining of ALDH1A1 in iBAT of AAV-injected mice after cold exposure. Representative image from four independent experiments. b, Immunohistochemical staining of UCP1 in subscWAT from shRNA-AAV-injected mice after cold exposure. Representative image from four independent experiments. Scale bars, 50 μm. This figure is related to Fig. 3.

Extended Data Fig. 7 Reduction of ALDH1A1 expression in BAT induces BAT thermogenic capacity.

a, Surface temperature of iBAT shRNA-AAV-injected mice at thermoneutrality. n = 6 mice, t = 4.17, df = 10. b, Immunohistochemical staining of UCP1 in iBAT of iBAT shRNA-AAV-injected mice at thermoneutrality. c, Protein levels of UCP1 in iBAT of iBAT shRNA-AAV-injected mice at thermoneutrality. n = 8 mice, t = 8.58, df = 14. d, Immunohistochemical staining of UCP1 in ingWAT of shRNA-AAV-injected mice after cold exposure. e, UCP1 protein levels in ingWAT from shRNA-AAV-injected mice after cold exposure. nshAldh1a1 = 6 mice, nscramble = 7 mice. f, CYP2E1 protein levels in iBAT from shRNA-AAV-injected mice at room temperature. nshAldh1a1 = 8 mice, nscramble = 10 mice. g, Glucose uptake of ingWAT, visWAT, brain and blood in mice injected with iBAT shRNA AAV. n = 4 mice. h, Time-resolved oxygen consumption of AAV-injected mice after cold exposure. n = 5 mice, df = 8, tRT = 2.57, tCE = 2.33. i, Schematic illustration of the overexpression AAV construct with mini Ucp1 promoter. j, ALDH1A1 and UCP1 protein levels of iBAT from iBAT AAV-GFP- or AAV-Aldh1a1-injected mice. n = 7 mice, df = 12, tALDH1A1 = 6.8 tUCP1 = 6.6. k, Time-resolved oxygen consumption of iBAT AAV-GFP- or AAV-Aldh1a1-injected mice after cold exposure. n = 5 mice. l, Surface temperature of iBAT AAV-GFP- or AAV-Aldh1a1-injected mice at room temperature. n = 6 mice, t = 2.8, df = 10. Data are mean average ± s.e.m. Statistical significance was calculated using a two-tailed unpaired Student’s t-test. Scale bars, 50 μm. See gel source data in Supplementary Figs. 1f, g, 2a, b. This figure is related to Fig. 3

Source data.

Extended Data Fig. 8 Reduction of ALDH1A1 during brown adipocyte formation induces thermogenic capacity.

a, Immunofluorescence staining and quantification of CYP2E1 and LD540 in ex-vivo-differentiated cells from iBAT. b, Immunofluorescence staining of ALDH1A1 and CYP2E1 in ex-vivo-differentiated cells from ingWAT. c, mRNA level of Adipoq, Ucp1, Cyp2e1 and Aldh1a1 during brown adipocyte differentiation ex vivo. n = 6 independent experiments. d, Protein levels of CYP2E1 and ALDH1A1 during brown adipocyte differentiation ex vivo. e, Immunofluorescence staining and quantification of ALDH1A1 and UCP1 in ex-vivo-differentiated adipocytes from iBAT. Brown area in the quantification bar plot denotes the overlap of UCP1+ALDH1A1+ double-positive population. n = 10 independent experiments. f, Immunofluorescence staining of ALDH1A1 and TOMM20 in ex-vivo-differentiated cells from iBAT. Brown area in the quantification bar plot denotes the overlap of TOMM20+ALDH1A1+ double-positive population. n = 10 independent experiments. g, Protein levels of ALDH1A1 in ex-vivo-differentiated cells from iBAT after Aldh1a1-siRNA-mediated knockdown. n = 4 independent experiments, t = 6.97, df = 6. h, Immunofluorescence staining of UCP1 and LD540 in ex-vivo-differentiated cells from iBAT after Aldh1a1-siRNA-mediated knockdown. n = 5 independent experiments, tUCP1+% = 5.73, df = 8. i, mRNA levels of Ucp1, Ppargc1a, Cidea, Dio2, Adipoq, Cebpa and Pparg in ex-vivo-differentiated cells from iBAT after Aldh1a1- and scramble-siRNA knockdown. n = 6 independent experiments, df = 10, tUcp1 = 3.46, tPpargc1a = 2.59, tCidea = 7.30, tDio2 = 2.49. j, Cellular respiration (OCR) after siRNA-mediated knockdown in brown adipocytes. n = 4 independent experiments, df = 6, tBasal = 3.82, tAtp = 2.14, tIso = 12.2, tMax = 8.39, tUncoupling = 5.28. k, Time-resolved OCR of ex-vivo-differentiated cells from iBAT. Cells transfected with Aldh1a1 or scramble siRNA were mixed in different ratios (related to Fig. 4d). l, Schematic illustration of the co-culture experiment. Cells transfected with Aldh1a1 or scrambled siRNA were cultured in the bottom or top chamber as indicated. m, UCP1 protein level of co-cultured cells in the bottom well. n = 4 independent experiments, F = 5.64. n, Cellular respiration of co-cultured cells in the bottom well. n = 5 independent experiments. Data are mean average ± s.e.m. Statistical significance was calculated using a two-tailed unpaired Student’s t-test (gi) or ordinary one-way ANOVA (m). Scale bars, 50 μm. See gel source data in Supplementary Fig. 2c, d. This figure is related to Fig. 4

Source data.

Extended Data Fig. 9 Overexpression of ALDH1A1 during brown adipocyte formation reduces thermogenic capacity.

a, b, FACS gating of adipocytes, CYP2E1+ and GFP+ cells from ex-vivo-differentiated adipocytes (a), and FACS control for the CYP2E1 antibody (left) and GFP (right) (b). c, mRNA expression levels of Cyp2e1, GFP, Pparg, Aldh1a1 and Ucp1 in FACS-sorted cells. n = 5 independent experiments. d, mRNA expression levels of Ucp1 and Pparg in FACS and co-cultured cells. n = 5 independent experiments, FUcp1 = 134.7, FPparg = 1.2. e, Time-resolved cellular respiration of FACS selected and co-cultured cells (related to Fig. 4e). f, Schematic illustration of the overexpression lentiviral construct with mini CMV promoter, based on pLenti-MP257. g, Protein levels of ALDH1A1 in ex-vivo-differentiated cells from iBAT infected with AAV-GFP or AAV-Aldh1a1. nGFP = 8 independent experiments, nAldh1a1 = 6 independent experiments, t = 3.62. h, Immunofluorescence staining of ALDH1A1 and UCP1 in ex vivo differentiated cells from iBAT infected with AAV-GFP or AAV-Aldh1a1, nGFP = 11 independent experiments, nAldh1a1 = 10 independent experiments, tUCP1+ = 21.4, tALDH1A1+ = 30.1. (i) mRNA expression levels of Ucp1, Pparg, Cebpa, Adipoq and Cyp2e1 in ex-vivo-differentiated cells from iBAT infected with AAV-GFP or AAV-Aldh1a1. n = 6 independent experiments, tUcp1 = 13.1. j, Cellular respiration of Aldh1a1- or GFP-overexpressing cells. n = 6 independent experiments. k, Retinaldehyde quantification in the supernatant of ex-vivo-differentiated cells from iBAT, treated with the indicated siRNA and/or acetate. n = 4 independent experiments, F = 3.99, P = 0.058. l, Retinaldehyde quantification of intracellular abundance of retinaldehyde in ex-vivo-differentiated cells from iBAT, treated with the indicated siRNA and/or acetate. n = 4 independent experiments, F = 4.348, P = 0.048. m, Cellular respiration of ex-vivo-differentiated cells from iBAT, treated with the indicated retinoids. n = 5 independent experiments. n, Cellular respiration of ex-vivo-differentiated cells from iBAT, treated with all-trans retinaldehyde at the indicated concentrations. n = 5 independent experiments. o, Acetate levels in iBAT of mice kept in cold-exposure, room-temperature and thermoneutral conditions. n = 6 independent experiments, F = 97.79. p, UCP1 and LD540 staining in ex-vivo-differentiated cells from iBAT, treated with the indicated levels of acetate during differentiation from day 4 to 8. n = 7 independent experiments, t = 32.4. Data are mean average ± s.e.m. Statistical significance was calculated using a one-way ANOVA (d, k, l,o, p) or two-tailed unpaired Student’s t-test (gi). Scale bars,50 μm. See gel source data in Supplementary Fig. 2f. This figure is related to Fig. 4

Source data.

Extended Data Fig. 10 ALDH1 regulates brown adipocyte thermogenic capacity through acetate signalling.

a, Ucp1 and Pparg mRNA expression levels in ex-vivo-differentiated cells from iBAT, treated with various levels of acetate during differentiation from day 4 to 8. n = 5 independent experiments, t = 50.7. b, Time-resolved cellular respiration in ex-vivo-differentiated cells from iBAT treated with the indicated levels of acetate (related to Fig. 4g). c, Cellular respiration in differentiated human MADS treated with the indicated levels of acetate during differentiation from day 16 to 17. n = 4 independent experiments (except n = 5 for the acetate 1 mM cohort), FMax = 16.2, FcAMP = 32.95. d, Cellular respiration in ex-vivo-differentiated cells from iBAT treated with the indicated levels of acetate in the presence or absence of Aldh1a1 knockdown. n = 5 independent experiments, F = 22.7. e, Cellular respiration in ex-vivo-differentiated cells from iBAT, treated with acetate (1 mM), AR-C155858 (1 μM) or SR13800 (1 μM). n = 4 independent experiments, FMax = 49.6. f, Cellular respiration in ex-vivo-differentiated cells from iBAT, treated with acetate (1 mM), CCCP (1 μM), α-cyano-4-hydroxycinnamic acid (1 mM), monensin (10 μM) or mercury chloride (5 μM). n = 4 independent experiments, FMax = 39.1. g, Gpr43, Ucp1, Pparg, Cebpa and Adipoq mRNA expression levels in ex-vivo-differentiated cells from iBAT with Gpr43 knockdown. n = 6 independent experiments, tGpr43 = 9.5, tUcp1 = 7.2. h, Time-resolved cellular respiration in brown adipocytes treated with the indicated levels of acetate with or without Gpr43 knockdown (related to Fig. 4h). i, Cellular respiration in ex-vivo-differentiated cells from iBAT, treated with the indicated levels of acetate and the GPR43 agonist CFMB. n = 5 independent experiments (except n = 6 for the CFMB 1 μM cohort), F = 23.9. j, Cellular respiration in ex-vivo-differentiated cells from ingWAT, treated with the indicated levels of acetate and GPR43 agonist. n = 5 independent experiments (except n = 4 for DMSO; acetate 0.1 mM and DMSO; acetate 1 mM cohorts), FMax = 25.8. k, Acetate levels in the culture medium of in ex-vivo-differentiated cells from iBAT with Aldh1a1 knockdown, treated with the indicated levels of acetate. n = 5 independent experiments, F = 652.7. Data are mean average ± s.e.m. Statistical significance was calculated using a one-way ANOVA (a, cf, ik) or two-tailed unpaired Student’s t-test (g). This figure is related to Fig. 4

Source data.

Supplementary information

Supplementary Information

This file contains Supplementary Figures 1 and 2 and Supplementary Tables 1-9.

Reporting Summary

Source data

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Sun, W., Dong, H., Balaz, M. et al. snRNA-seq reveals a subpopulation of adipocytes that regulates thermogenesis. Nature 587, 98–102 (2020). https://doi.org/10.1038/s41586-020-2856-x

Download citation

Further reading

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Search

Quick links