Phenotypic and functional effects of CD9 downregulation. REH cells were transduced with 2 different shRNAs targeting CD9 messenger RNA (mRNA). (A) Membrane expression of CD9 and CXCR4 measured by flow cytometry and represented by gray histograms. Fluorescence with the control isotype is indicated by the black line on the histograms. (B) Adhesion assay. REH cells were used to seed superfibronectin-coated 96-well plates. The adherent cells were counted in an MTS (3-(4,5 dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl-2-(4-sulfophenyl)-2H-tetrazolium) proliferation assay. The graph represents the optic density (OD) values, with the bars indicating the median values. **P < .01, ****P < .0001 in Wilcoxon test. (C) Migration assay. Cell migration in response to a gradient of CXCL12 (100 ng/mL) was measured in a Boyden chamber. Results are presented as migration rates using a scatter dot plot representation. The histograms indicate the means of 8 independent experiments. *P < .05 in Wilcoxon test. (Di) The effects of CD9 depletion on REH cell morphology were observed by scanning electron microscopy. (Dii) The number of villi per cell and the area of the membrane villi were determined with ImageJ software. The bars indicate the median values. n is the number of cells analyzed: REH CD9^high, n = 21; REH CD9^medium, n = 41; REH CD9^low, n = 21; **P < .01, ***P < .001 in Mann-Whitney tests. (E) REH cells stably transduced with shRNA targeting CD9 were cultured and injected IV (10⁵ cells) into 4-week-old NSG mice. The general condition of the mice was monitored daily until their death. Kaplan-Meier survival curves were plotted. n is the number of mice used: REH CD9^high, n = 7; REH CD9^medium, n = 14; REH CD9^low, n = 8;*** P < .001, ****P < .0001 in log-rank (Mantel-Cox) tests.