BATLAS: Deconvoluting brown adipose tissue. BATLAS: Deconvoluting brown adipose tissue

Recruitment and activation of classical brown and inducible brite/beige adipocytes has received increasing attention in recent years as a strategy to improve systemic metabolic control.Nevertheless, the origin and expression signatures of brown and brite/beige adipocytes are still under debate, mainly due to the complexity of tissue biopsies. To study different adipocyte types in detail, we generated pure samples of brown, brite/beige, and white mature adipocytes by fluorescence activated cell sorting. Using this resource, we could demonstrate that on a transcriptional level, brown and brite/beige mature adipocytes are identical. Employing a machine learning approach for paired analysis of transcriptional data of pure mouse brown,brite/beige and white adipocytes and human brown and white whole adipose tissue obtained by PET-CT-guided biopsies, we were concomitantly able to identify a gene signature that can classify brown and white adipose tissue depots. Using the newly developed algorithm, we were able to predict the brown adipocyte content in a mixed population of adipocytes from different human biopsies that can be used for in-depth characterization of complex tissue samples from adipose tissue and might therefore support the development strategies to increase brown adipocyte formation in humans.