Fig. 6: β-cell diabetes dysfunction involves different molecular patterns that are unique or shared with other conditions, including different diabetes models and aging.
From: Delineating mouse β-cell identity during lifetime and in diabetes with a single cell atlas
a, The activity of β-cell diabetes-trajectories (NOD and db/db + mSTZ) DEG groups across fine β-cell states (red rectangles mark examples highlighted in text). Cell groups representative of healthy and diabetic states in DGE analysis are marked with blue and orange rectangles, respectively. b, Overlap of DEGs across diabetes models (T1D NOD, T2D db/db + mSTZ) and endocrine cell types. c, Expression of DEG groups between aged males and females across coarse β-cell states, split by sex. Marked are cell groups highlighted in the text and groups representative of healthy and diabetes model cells from DGE analysis. d, Gene expression of diabetes markers that were validated on protein level; shown for diabetes models and associated controls. e, Validation of selected diabetes model β-cell DEGs on protein level with immunohistochemistry. The images are representative examples of three independent animals. Scale bars, 50 μM. For every antibody pair, the left plot shows an overlay of channels and the right shows individual channels. f, PAGA graph showing connectivity (lines) between fine β-cell states (dots) imposed on β-cell UMAP. The connections between healthy, intermediate and diabetes model states are marked in solid lines. g, Expression of DEGs with the same direction in NOD and db/db + mSTZ trajectories in healthy, intermediate and diseased states per dataset (dataset 8–16wNOD is abbreviated as NOD). Expression is normalized per gene and dataset. imm., immature; M, male; F, female; NOD-D, NOD diabetic; D.-inter, diabetic intermediate. In a, c, d and g relative expression is computed as the average of cell groups normalized to [0,1] for each gene feature.
