Figure 4

Reducing CD44 is sufficient to alter normal phenotype in breast cells. MCF-10a cells were transfected for 48 hours with full-length CD44 (CD44WT), and single-site (C286A, C286S) or double-site (SA, AA) palmitoylation-impaired mutants. (A) Triton X-100-insoluble and Triton X-100-soluble fractions were isolated, and were confirmed to be enriched in respectively lipid raft (Flotillin-1 (Flot-1)) and nonraft (transferrin receptor (TfR)) marker proteins. Overexpression of mutant CD44 reduced CD44 recovery from raft-containing fractions compared with that in untransfected controls and CD44WT-expressing cells. (B) Calculated ratios of CD44 affiliation with detergent-insoluble fractions reflected significant reductions in CD44 raft affiliation in mutant cells compared with untransfected control cells, and furthermore in CD44 double-site mutants compared with CD44WT. (C) CD44 palmitoylation was assessed by 1-biotinamido-4-(4′-(maleimidomethyl)cyclohexanecarboxamido)butane (BMCC) assay in CD44 immunoprecipitates of control, CD44WT-expressing and CD44 mutant-expressing cells. Palmitoylated CD44 was detected with streptavidin (CD44-Palm), and total CD44 detected using CD44 antibody (CD44-Total). No CD44 was immunoprecipitated in the isotype-matched IgG control lanes, and no palmitoylated CD44 was detected in the hydroxylamine (HAM)-negative control conditions. In the HAM-positive condition, palmitoylated CD44 was reduced in mutant-expressing cells. (D) Quantification of palmitoylated CD44 (as a ratio of total CD44) revealed significant reductions in cells overexpressing a double-site CD44 palmitoylation-impaired mutant (AA) relative to control or CD44WT-expressing cells (*P < 0.05; **P < 0.01, Student’s t test). (E) Phase-contrast micrographs 48 hours post transfection revealed a change in the morphology of control MCF-10a colonies (Ctrl) following CD44 overexpression, with a protrusive, disseminated epithelial-to-mesenchymal transition-like appearance in all of the CD44-overexpressing cells (arrows), which was particularly pronounced in those expressing the CD44 palmitoylation-impaired mutants. (F) Migration was measured via scratch-wound assays, in which wound widths at each time point were expressed relative to their cognate measurement at time = 0 for each condition. These assays revealed significant enhancements in the migratory capacity of MCF-10a cells expressing CD44 palmitoylation-impaired mutants relative to control cells (*P < 0.05, two-way analysis of variance ). Error bars, standard error of the mean; n = 3 experiments. WT, wild-type.