Accumulating evidence suggests that subcutaneous and visceral adipose tissues are differentially associated with metabolic disorders. In obesity, subcutaneous adipose tissue is beneficial for metabolic homeostasis because of repressed inflammation. However, the underlying mechanism remains unclear. Here, we demonstrate that γ-aminobutyric acid (GABA) sensitivity is crucial in determining fat depot-selective adipose tissue macrophage (ATM) infiltration in obesity. In diet-induced obesity, GABA reduced monocyte migration in subcutaneous inguinal adipose tissue (IAT), but not in visceral epididymal adipose tissue (EAT). Pharmacological modulation of the GABA _B receptor affected the levels of ATM infiltration and adipose tissue inflammation in IAT, but not in EAT, and GABA administration ameliorated systemic insulin resistance and enhanced insulin-dependent glucose uptake in IAT, accompanied by lower inflammatory responses. Intriguingly, compared with adipose-derived stem cells (ADSCs) from EAT, IAT-ADSCs played key roles in mediating GABA responses that repressed ATM infiltration in high-fat diet-fed mice. These data suggest that selective GABA responses in IAT contribute to fat depot-selective suppression of inflammatory responses and protection from insulin resistance in obesity.

Some of the recent studies on drug sensitivity prediction have applied graph neural networks to leverage prior knowledge on the drug structure or gene network, and other studies have focused on the interpretability of the model to delineate the mechanism governing the drug response. However, it is crucial to make a prediction model that is both knowledge-guided and interpretable, so that the prediction accuracy is improved and practical use of the model can be enhanced. We propose an interpretable model called DRPreter (drug response predictor and interpreter) that predicts the anticancer drug response. DRPreter learns cell line and drug information with graph neural networks; the cell-line graph is further divided into multiple subgraphs with domain knowledge on biological pathways. A type-aware transformer in DRPreter helps detect relationships between pathways and a drug, highlighting important pathways that are involved in the drug response. Extensive experiments on the GDSC (Genomics of Drug Sensitivity and Cancer) dataset demonstrate that the proposed method outperforms state-of-the-art graph-based models for drug response prediction. In addition, DRPreter detected putative key genes and pathways for specific drug–cell-line pairs with supporting evidence in the literature, implying that our model can help interpret the mechanism of action of the drug.

Genomic profiles of cancer patients such as gene expression have become a major source to predict responses to drugs in the era of personalized medicine. As large-scale drug screening data with cancer cell lines are available, a number of computational methods have been developed for drug response prediction. However, few methods incorporate both gene expression data and the biological network, which can harbor essential information about the underlying process of the drug response. We proposed an analysis framework called DrugGCN for prediction of Drug response using a Graph Convolutional Network (GCN). DrugGCN first generates a gene graph by combining a Protein-Protein Interaction (PPI) network and gene expression data with feature selection of drug-related genes, and the GCN model detects the local features such as subnetworks of genes that contribute to the drug response by localized filtering. We demonstrated the effectiveness of DrugGCN using biological data showing its high prediction accuracy among the competing methods.

Adipocytes have unique morphological traits in insulin sensitivity control. However, how the appearance of adipocytes can determine insulin sensitivity has not been understood. Here, we demonstrate that actin cytoskeleton reorganization upon lipid droplet (LD) configurations in adipocytes plays important roles in insulin-dependent glucose uptake by regulating GLUT4 trafficking. Compared to white adipocytes, brown/beige adipocytes with multilocular LDs exhibited well-developed filamentous actin (F-actin) structure and potentiated GLUT4 translocation to the plasma membrane in the presence of insulin. In contrast, LD enlargement and unilocularization in adipocytes downregulated cortical F-actin formation, eventually leading to decreased F-actin-to-globular actin (G-actin) ratio and suppression of insulin-dependent GLUT4 trafficking. Pharmacological inhibition of actin polymerization accompanied with impaired F/G-actin dynamics reduced glucose uptake in adipose tissue and conferred systemic insulin resistance in mice. Thus, our study reveals that adipocyte remodeling with different LD configurations could be an important factor to determine insulin sensitivity by modulating F/G-actin dynamics.

Thermogenic adipocytes generate heat to maintain body temperature against hypothermia in response to cold. Although tight regulation of thermogenesis is required to prevent energy sources depletion, the molecular details that tune thermogenesis are not thoroughly understood. Here, we demonstrate that adipocyte hypoxia-inducible factor α (HIFα) plays a key role in calibrating thermogenic function upon cold and re-warming. In beige adipocytes, HIFα attenuates protein kinase A (PKA) activity, leading to suppression of thermogenic activity. Mechanistically, HIF2α suppresses PKA activity by inducing miR-3085-3p expression to downregulate PKA catalytic subunit α (PKA Cα). Ablation of adipocyte HIF2α stimulates retention of beige adipocytes, accompanied by increased PKA Cα during re-warming after cold stimuli. Moreover, administration of miR-3085-3p promotes beige-to-white transition via downregulation of PKA Cα and mitochondrial abundance in adipocyte HIF2α deficient mice. Collectively, these findings suggest that HIF2α-dependent PKA regulation plays an important role as a thermostat through dynamic remodeling of beige adipocytes.