Reduced expression of CXCL12 by impaired BM nestin⁺MSCs. (A) Expression of the CXCL12 mRNA in nestin⁺ MSCs analyzed by PCR in the ITP group (n = 30) and control group (n = 28). (B) Proportion of CXCL12⁺ nestin⁺ MSCs in nestin⁺ MSCs analyzed by FACS in the ITP group (n = 30) and control group (n = 28). (C) Proportion of nestin⁺ MSCs among CD45⁻ CD31⁻ CD235a⁻ stromal cells analyzed by FACS in the ITP group (n = 30) and control group (n = 28). (D) Proportion of CD45⁻ CD31⁻ CD235a⁻ stromal cells among mononuclear cells analyzed by FACS in the ITP group (n = 30) and control group (n = 28) (see supplemental Figure S3 for the gating strategy used). (E) Representative merge image of BM biopsy tissues immunofluorescently stained with CD34 (green), nestin (red), and DAPI (blue) antibodies in the ITP group (scale bar, 50 μm). (F) Representative image of BM biopsy tissue immunofluorescently stained with CD34 (green), nestin (red), and DAPI (blue) antibodies in the control group (scale bar, 50 μm). (G) Number of nestin⁺ MSCs per square millimeter of BM biopsy tissue analyzed by immunofluorescence in the ITP group (n = 30) and control group (n = 28). (H) Nestin mRNA in nestin⁺ MSCs analyzed by PCR in the ITP group (n = 30) and control group (n = 28). Data are pooled from 10 independent experiments with 4 to 8 samples per experiment (A-D,G,H). Data are shown as mean ± SD. Each point represents the mean adjusted value of 3 replicates for each individual patient. P values were calculated using the Mann-Whitney U test. *P < .001.