Fig. 3: CD24high osteolineage cells represent inflammatory senescent cells in old mice targeted by genetic senolytic clearance.
a t-SNE visualization and FlowSOM clustering of n = 80,000 CD45-Lin- bone and marrow cells (n = 40 INK-ATTAC mice [n = 15 young, n = 12 old + vehicle, n = 13 old + AP] – 2,000 cells sampled per mouse) analyzed by CyTOF. Cells are colored by clustered population (see Table 1 for defining markers); b Heatmap representation of the 11 cell clusters and protein expression of identification markers; c Log2 fold-Change mean expression of p16 or BCL-2 in each cluster with age. d Log2 fold-change of %p16KB cells in each cluster across aging; e Schematic of p16+ senescent cell clearance in 24-month-old AP-treated INK-ATTAC mice; f t-SNE plots of FlowSOM clusters of bone/bone marrow cells from old vehicle- or AP-treated INK-ATTAC mice. Black arrows indicate cleared clusters while green arrows indicate increased cluster abundance; g Quantification of cluster abundance changes between vehicle- and AP-treated mice (Log2-fold change to vehicle); h CD24 expression feature plots in samples from vehicle- or AP-treated mice; i Feature plots of p16 and FLAG (INK-ATTAC transgene) protein expression; j Heatmap and k, l volcano plots for statistical comparisons of senescence marker expression between cleared clusters in vehicle-treated old mice. Schematic in (e) was generated using BioRender. Box plots show median and interquartile ranges with error bars representing minimum and maximum values. (c, d, g) Two-sided Mann–Whitney test or unpaired t test, as appropriate. k, l Multiple two-sided t tests with Holm-Sidak Correction. Source data are provided as a Source Data file.
