ORF27-05 - Transcriptomic Profiling of Human Adipose Depots Demonstrates Patient-Specific Coordination of Tissue Gene Expression

Human adipose tissues are solid organs distributed throughout the body in specific locations and are thought to have dichotomous functions: white adipose tissue (WAT) stores triglycerides as an energy reservoir, and brown adipose tissue (BAT) oxidizes metabolic fuels to generate heat. To define the transcriptional landscape of six principal adipose depots, we performed RNA sequencing on tissue from 11 autopsies (2F/9M, 16-71y) of patients with different underlying diseases. We examined whether depots exhibit distinct functional specialization and the extent to which transcriptional coordination occurs across the whole-body adipose system. Despite global similarities in terms of cell type proportions, each depot displayed a distinct transcriptomic signature and functional profile. Analysis of published mouse datasets revealed that tissue depots clustered more closely than individual mice (median Calinski-Harabasz Index [CH-I]: 3.999 vs 0.283, P = 0.0001), with depot identity accounting for 56.8% of gene expression variance versus 18.6% for individual mouse identity (P < 2.2e-16). Conversely, human donors clustered more closely than depots (median CH-I 3.329 vs 0.605, P = 0.007), with donor identity explaining 37.6% of variance compared to only 5.4% for depot identity (P < 2.2e-16). After incorporating covariates (age, sex, postmortem interval, BMI, RNA integrity), donor identity (15.6%) and BMI (15.1%) emerged as dominant variance contributors. Furthermore, we used the transcriptomic pattern of an accessible subcutaneous WAT depot to predict donor identity with 83% accuracy. To validate transcriptomic findings, we measured real-time metabolic activity using 18F-FDG PET/CT in 14 healthy women. Two distinct categories of human adipose tissues arose: depots with higher glucose uptake and depots with lower glucose uptake, corresponding to greater thermogenic (BAT) and energy storage capacities (WAT), respectively. These results aligned with immunofluorescence imaging and gene expression characterization of thermogenic markers, such as uncoupling protein 1 (UCP1), within donor tissue. When the dataset was partitioned for genes showing the greatest variability across depots, a set of 47 genes served as an exception, with greater intradepot similarity than intrasubject similarity, providing possible molecular signatures to define distinct adipose depots in humans. Our findings show that while each human adipose depot possesses a unique transcriptomic signature reflecting overlapping yet discrete functions, an individual’s adipose depots are more transcriptionally similar to each other than the same depot is across different individuals. This pattern has potential physiological implications, suggesting that individualized factors such as genetics, diet, and disease play impactful roles in coordinating the behavior of the whole-body adipose tissue system.