miR-21 targets Ski in primary hematopoietic cells and early-stage MDS. (A) Sequence alignment between miR-21 and a region of the SKI 3′ UTR that is conserved between humans and mice. (B) miR target validation: luciferase sensor assay for the empty vector (EV), wild-type (WT), or mutated SKI 3′ UTR in MCF7 cells (mean ± standard error of the mean [SEM] for 3 independent experiments). (C) miR-21 forced expression: representative immunoblot of retroviral transduced lineage^– bone marrow (from 3 independent experiments). Nearby protein marker sizes (in kDa) are indicated. (D) Primary marrow cell context for miR-21 activity: representative immunoblot of freshly isolated lineage^– bone marrow from mice with the indicated genotypes (above the immunoblot) (from 3 independent experiments). Nearby protein marker sizes (in kDa) are indicated. (E) Steady-state Ski mRNA levels in freshly isolated lineage^– bone marrow cells (from 3 biological replicates). (F) Validation in miR-21–expressing MDS: confocal microscopy images of immunofluorescent anti-SKI staining of human bone marrow sections with the indicated diagnoses. The scale bars represent 100 μm for 20× images and 25 μm for 60× images. Red text indicates samples with MIR21 quantitation. (G) Quantitation of immunofluorescence staining in panel F. Red dots indicate samples with MIR21 quantitation. (H) Steady-state MIR21 mRNA levels in bone marrow samples. (I) Comparison between the fraction of SKI-expressing cells and MIR21 mRNA levels in MDS patients. Data in (B,E,G-I) are displayed as mean ± SEM. **P < .01; ***P < .001. Ab, antibody; DAPI, 4′,6-diamidino-2-phenylindole; RCMD, refractory cytopenia with multilineage dysplasia.