Tumor suppressors TSC1 and TSC2 differentially modulate actin cytoskeleton and motility of mouse embryonic fibroblasts

Abstract

TSC1 and TSC2 mutations cause neoplasms in rare disease pulmonary LAM and neuronal pathfinding in hamartoma syndrome TSC. The specific roles of TSC1 and TSC2 in actin remodeling and the modulation of cell motility, however, are not well understood. Previously, we demonstrated that TSC1 and TSC2 regulate the activity of small GTPases RhoA and Rac1, stress fiber formation and cell adhesion in a reciprocal manner. Here, we show that Tsc1(-/-) MEFs have decreased migration compared to littermate-derived Tsc1(+/+) MEFs. Migration of Tsc1(-/-) MEFs with re-expressed TSC1 was comparable to Tsc1(+/+) MEF migration. In contrast, Tsc2(-/-) MEFs showed an increased migration compared to Tsc2(+/+) MEFs that were abrogated by TSC2 re-expression. Depletion of TSC1 and TSC2 using specific siRNAs in wild type MEFs and NIH 3T3 fibroblasts also showed that TSC1 loss attenuates cell migration while TSC2 loss promotes cell migration. Morphological and immunochemical analysis demonstrated that Tsc1(-/-) MEFs have a thin protracted shape with a few stress fibers; in contrast, Tsc2(-/-) MEFs showed a rounded morphology and abundant stress fibers. Expression of TSC1 in either Tsc1(-/-) or Tsc2(-/-) MEFs promoted stress fiber formation, while TSC2 re-expression induced stress fiber disassembly and the formation of cortical actin. To assess the mechanism(s) by which TSC2 loss promotes actin re-arrangement and cell migration, we explored the role of known downstream effectors of TSC2, mTORC1 and mTORC2. Increased migration of Tsc2(-/-) MEFs is inhibited by siRNA mTOR and siRNA Rictor, but not siRNA Raptor. siRNA mTOR or siRNA Rictor promoted stress fiber disassembly in TSC2-null cells, while siRNA Raptor had little effect. Overexpression of kinase-dead mTOR induced actin stress fiber disassembly and suppressed TSC2-deficient cell migration. Our data demonstrate that TSC1 and TSC2 differentially regulate actin stress fiber formation and cell migration, and that only TSC2 loss promotes mTOR- and mTORC2-dependent pro-migratory cell phenotype.

Figure 1. Effects of TSC1 and TSC2 on actin cytoskeleton.

A: F-actin staining of serum-deprived NIH 3T3 fibroblasts and matched Tsc1^+/+, Tsc1^−/− and Tsc2^+/+, and Tsc2^−/− MEFs. B: F-actin staining (red) of Tsc1^−/− and Tsc2^−/− MEFs transfected with either GFP-TSC1, GFP-TSC2, or GFP. Representative images of two separate experiments were taken using a Nikon Eclipse TE-2000E microscope at 200x magnification. Scale bar, 20 µm.

Figure 2. Effects of expression and siRNA-induced down-regulation of TSC1 and TSC2 in NIH 3T3 fibroblasts on actin cytoskeleton.

Serum-deprived NIH 3T3 fibroblasts were transfected with GFP-TSC1, GFP-TSC2, or control GFP or microinjected with siRNA TSC1 or siRNA TSC2 and GFP followed by rhodamine phalloidin staining to detect F-actin (red) and immunostaining with anti-GFP antibody (green) to identify transfected or injected cells. Scale bar, 20 µm.

Figure 3. Morphology and dynamics of Tsc1^−/− and Tsc2^−/− MEFs during wound closure.

Phase-contrast micrographs of cell motility during wound closure at 4 h after wound scraping. Arrows indicate direction of cell movement. Images were taken using a Nikon Eclipse TE2000-E microscope at 100X magnification in the phase contrast channel. Images are representative from three independent experiments. Scale bar, 100 µm.

Figure 4. TSC1 and TSC2 differentially regulate migration and invasiveness.

A: Representative membrane showing migration of serum-deprived Tsc2^+/+, Tsc2^−/−, Tsc1^+/+, and Tsc1^−/− MEFs. Serum-deprived cells were placed on collagen-saturated membranes in serum-free medium, and allowed to migrate in the Boyden chamber for 4 h in the absence of any stimuli. Then membranes were fixed, stained with Hemacolor stain set, and analyzed using Gel Pro software. B: Statistical analysis of migration experiments. Data represent mean values ± SE from measurements performed in triplicate from six separate experiments by ANOVA (Bonferroni-Dunn test). Basal migration of Tsc2^+/+ cells was taken as 1 fold. C: Tsc2^−/− MEFs have increased invasiveness. Invasiveness of serum-deprived Tsc2^+/+, Tsc2^−/−, Tsc1^+/+, and Tsc1^−/− MEFs was analyzed using the Cultrex 96 Well BME Cell Invasion Assay kit according to manufacturer’s protocol. Data represent the percentage of invaded cells per total number of cells taken as 100%. Data represent mean values ± SE from two independent experiments by ANOVA (Bonferroni-Dunn test).

Figure 5. TSC1 and TSC2 re-expression or siRNA-induced knock-out validate their differential role in regulating cell migration.

A: Re-expression of TSC1 rescues Tsc1^−/− MEFs migration. MEFs were transiently transfected with TSC1 for 48 h, serum deprived followed by migration assay. B: Re-expression of TSC2 inhibits Tsc2^−/− MEF migration. Tsc2^+/+ and Tsc2^−/− MEFs were transiently transfected with TSC2, serum deprived followed by migration assays. TSC1 and TSC2 re-expression were confirmed by immunoblot analysis of equalized in protein content samples. Migration of Tsc1^−/− (A) and Tsc2^−/− MEFs (B) transfected with control plasmid was taken as 1 fold. Data represent mean values ± SE from two different experiments with three replicates for each condition by ANOVA (Bonferroni-Dunn test). C: Downregulation of TSC2, but not TSC1, promotes migration of wild type MEFs. Tsc2^+/+ and Tsc1^+/+ MEFs were transfected with siRNA TSC1, siRNA TSC2, and control siRNA. 48 h post-transfection, migration assays were performed. Protein levels were detected by immunoblot analysis with specific anti-TSC1 and anti-TSC2 antibodies under the same experimental conditions. Migration of wild type Tsc1^+/+ (right) or Tsc2^+/+ MEFs (left) transfected with siGLO RISC-Free siRNA was taken as 1 fold. Data represent mean values ± SE from measurements performed in triplicate from two separate experiments.

Figure 6. siRNATSC1 and siRNATSC2 induce opposite effects on NIH 3T3 fibroblast migration.

Cells were transfected with siRNA TSC1 (A), siRNA TSC2 (B), or control siRNA. 48 h post-transfection, protein levels were detected by immunoblot analysis with anti-TSC1 or anti-TSC2 antibodies. C, Upper panel: Representative image of hemacolor-stained membrane with migrated NIH 3T3 fibroblasts for 4 h. 3T3 fibroblasts were transfected with siRNA TSC1, siRNA TSC2, and siGLO RISC-Free siRNA as control cells, serum-deprived followed by migration assay in the presence or absence of 10 ng/ml PDGF performed in triplicate for each experimental condition. C, Lower panel: Statistical analysis of NIH 3T3 cell migration. Data represent mean values ± SE from two independent experiments, six repetitions in each experiment. *P<0.01 for siRNA TSC1 vs. control siRNA, **P<0.001 for siRNA TSC2 vs. control siRNA by ANOVA (Bonferroni-Dunn).

Figure 7. mTOR and Rictor, but not Raptor, modulate stress fiber formation in TSC2-null ELT3 cells.

A: Cells were co-microinjected with siRNA mTOR, siRNA Rictor, siRNA Raptor, or control siRNA, and GFP to detect microinjected cells, serum-deprived followed by F-actin staining. Images were taken using a Nikon Eclipse TE-2000-E Microscope at 400X magnification. Scale bar, 20 µm. B: Statistical analysis. Data represent the percentage of cells without stress fibers per total number of microinjected cells taken as 100%. Data represent mean values ± SE by ANOVA (Bonferroni Dunn). C: mTOR activity is required for stress fiber formation in TSC2-null cells. Cells transfected with HA-tagged mTOR-KD were serum-deprived followed by staining with rhodamine phalloidin and immunostaining with anti-HA antibody to detect transfected cells. Data represent the percentage of cells without stress fibers per total number of transfected cells taken as 100%. Data represent mean values ± SE by ANOVA (Bonferroni Dunn). Scale bar, 20 µm.

Figure 8. mTOR and Rictor mediate cell migration.

A: siRNA-induced mTOR depletion inhibits Tsc2^−/− MEF migration. Tsc2^−/− MEFs were transfected with siRNA mTOR or control siRNA followed by migration assay. Migration of control siRNA-transfected cells was taken as 100%. B: Rictor is required for serum-induced MEF migration. Tsc2^−/− MEFs were transfected with siRNA Rictor,siRNA Raptor, or control siRNA, and migration assays were subsequently performed. Migration of Tsc2^−/− MEFs transfected with control siRNA was taken as 1 fold. C: Rictor is required for cell migration under nutrient-replete conditions but is not sufficient to modulate all migration. Migration assays were performed with Rictor^+/+ and Rictor^−/− MEFs under basal (unstimulated) and serum-stimulated (10% FBS) conditions. Migration of Rictor^+/+ MEFs under basal conditions was taken as 100%.