A PI3K- and GTPase-independent Rac1-mTOR mechanism mediates MET-driven anchorage-independent cell growth but not migration

Abstract

Receptor tyrosine kinases (RTKs) are often overexpressed or mutated in cancers and drive tumor growth and metastasis. In the current model of RTK signaling, including that of MET, downstream phosphatidylinositol 3-kinase (PI3K) mediates both cell proliferation and cell migration, whereas the small guanosine triphosphatase (GTPase) Rac1 mediates cell migration. However, in cultured NIH3T3 and glioblastoma cells, we found that class I PI3K mediated oncogenic MET-induced cell migration but not anchorage-independent growth. In contrast, Rac1 regulated both processes in distinct ways. Downstream of PI3K, Rac1 mediated cell migration through its GTPase activity, whereas independently of PI3K, Rac1 mediated anchorage-independent growth in a GTPase-independent manner through an adaptor function. Through its RKR motif, Rac1 formed a complex with the kinase mTOR to promote its translocation to the plasma membrane, where its activity promoted anchorage-independent growth of the cell cultures. Inhibiting mTOR with rapamycin suppressed the growth of subcutaneous MET-mutant cell grafts in mice, including that of MET inhibitor-resistant cells. These findings reveal a GTPase-independent role for Rac1 in mediating a PI3K-independent MET-to-mTOR pathway and suggest alternative or combined strategies that might overcome resistance to RTK inhibitors in patients with cancer.

Fig. 1. Class I PI3K inhibition reduces MET-dependent cell migration but not anchorage-independent growth.

(A) Transwell migration assay performed with wild-type or M1268T MET-expressing NIH3T3 cells treated with DMSO, LY294002 (LY, 10 μM) or PHA-665752 (PHA, 100 nM). Data are mean ± SEM from N = 5 (DMSO), 4 (LY), or 3 (PHA) independent biological replicates. Student’s t test: NS, non-significant; ***P<0.005. (B and C) Colony formation assays in the cells described and treated as in (A) and grown in soft agar. Representative images are shown (B; scale bar: 5 mm), and the colony area was calculated by ImageJ (C). Data are mean ± SEM from N=3 independent biological replicates. Student’s t test: NS, non-significant; ***P<0.005. (D and E) Western blotting for phosphorylated (“P-AKT”, Ser⁴⁷³) and total AKT, with tubulin as loading control, in lysates from cells described and treated as in (A) or with dynasore (E; 80 μM). Below, mean densitometry of phosphorylated AKT relative to total ± SEM from N=3 independent biological replicates. Student’s t test (unless indicated, compared to control): NS, non-significant; ***P<0.005. (F and G) As in (A) and (C), respectively, in cells treated with DMSO or GDC0941 (F, 100 nM; G, as indicated). Data are mean ± SEM from at least 3 independent biological replicates. Student’s t test: NS, non-significant; ***P<0.005. (H) Relative colony formation by wild-type or D1246N MET-expressing cells grown in soft agar and treated with DMSO, GDC0941 (100 nM) or PHA-665752 (PHA, 100 nM). Data are mean ± SEM from N=3 independent biological replicates. One-way ANOVA test followed by Tukey’s multiple comparisons test: NS, non-significant; ***P<0.005. (I) Western blotting for phosphorylated (Tyr^1234/1235) and total MET and AKT in lysates from U87MG cells treated with DMSO or PF-02341066 (100 nM). p170 and p145: the precursor and the mature forms of the beta chain of MET. Blots were analyzed as in (D and E). (J) Relative colony area of U87MG cells grown in soft agar and treated with DMSO, PF-02341066 (100 nM) or GDC0941 (100 nM). Data are means ± SEM from at least 3 independent biological replicates. Student’s t test (compared to control): NS, non-significant; *P<0.05.

Fig. 2. Class I PI3K promotes MET-dependent cell migration through Rac1 activation.

(A) Transwell migration assays performed with wild-type or M1268T MET-expressing cells transfected with a negative control or siRNA targeting p110α, p110β, or both, and treated with DMSO (control), A66 (500 nM) or TGX221 (TGX, 40 nM) either alone or combined. Data are means ± SEM, N=3 or 4 independent biological replicates. Student’s t test (compared to control, or as indicated): NS, non-significant; **P<0.01, ***P<0.005. (B) Transwell migration by U87MG cells treated with DMSO (N=8), PF-02341066 (100 nM, N=3), A66 (500 nM, N=3), TGX221 (40 nM, N=3) or A66 and TGX221 combined (A66+TGX, N=4). Data are means ± SEM from independent biological replicates. Student’s t test (compared to control, or as indicated): NS, non-significant; *P<0.05, **P<0.01, ***P<0.005.One-way ANOVA test followed by Tukey’s multiple comparisons test (compared to control): NS, non-significant; *P<0.05, ***P<0.005. (C and D) Wild-type and M1268T MET-expressing cells were transfected with negative control (RNAi control) or p110α and p110β combined (RNAI p110 α+ β) siRNAs. The percentage of cells with Rac1 at the plasma membrane (C) or lacking stress fibres (D) was counted, following immunostaining with an antibody against Rac1 or rhodamine-phalloidin, respectively. Data are means ± SEM from N=3 independent biological replicates. Student’s t test: NS, non-significant; *P<0.05, **P<0.01, ***P<0.005.

Fig. 3. mTORC1 promotes oncogenic Met-dependent anchorage-independent and tumor growth.

(A) Relative colony area of wild-type, M1268T (N=6 independent biological replicates) and D1246N MET-expressing cells (N=3 independent biological replicates) grown in soft agar and treated with DMSO or wortmannin (100 nM). (B) Relative colony area of wild-type, M1268T and D1246N MET-expressing cells (N=3 independent biological replicates) and U87MG cells (N=4 independent biological replicates) grown in soft agar and treated with DMSO or rapamycin (2 nM). (C and D) Western blots for phosphorylated p70-S6K and mTOR (Ser²⁴⁸¹) in lysates from wild-type, M1268T and D1246N MET-expressing cells treated with DMSO, LY294002 (LY, 10 μM) or PHA-665752 (PHA, 100 nM). N= 3 or 4 independent biological replicates. (E and F) Tumour volume of graft of 5x10⁵ cells expressing wild-type, M1268T or D1246N MET injected subcutaneously into nude mice. Tumors were measured daily and, at 30-50 mm3, treated topically with DMSO or rapamycin (2 nM), then assessed at days 4 (E) and 6 (F) thereafter. N=5 mice per group. Data (A to F) are mean values ± SEM. Student’s t test (compared to control or as indicated): NS, non-significant; *P<0.05, **P<0.01, ***P<0.005.

Fig. 4. Oncogenic Met promotes anchorage-independent growth via a Rac1-mTOR pathway.

(A and B) Western blots for phosphorylated p70-S6K (P-p70-S6K), p70-S6K (A and B), phosphorylated ERK1/2 (P-ERK1/2), ERK1/2 and HSC70 (B) were performed on M1268T MET-expressing cells treated with DMSO, dynasore (80 μM) or UO126 (10 μM), as indicated. Data below blots are mean ± SEM, phospho:total ratio, obtained by densitometry of N=3 independent biological replicates. (C) Western blots for phosphorylated p70-S6K (P-p70-S6K), p70-S6K, HSC70 and Rac1 were performed on M1268T MET-expressing cells transfected with negative control or Rac1 siRNA (“RNAi”). Below are mean levels of indicated phosphorylated protein over total protein ± SEM, obtained by densitometry of Western blots. N=3 independent biological replicates. (D and E) Relative colony area of (D) wild-type and M1268T MET-expressing cells and (E) U87MG cells transfected with negative control or Rac1 RNAi and grown in soft agar. N=3 independent biological replicates. Data (A to E) are mean values ± SEM. P values were obtained with the Student’s t test. NS: non-significant, *P<0.05, ***P<0.005.

Fig. 5. Rac1 promotes anchorage-independent growth of MET-mutant cells independently of its GTPase activity.

(A) Western blots for phosphorylated mTOR on Ser²⁴⁸¹ (P-mTOR), mTOR, HSC70 and Rac1 of M1268T Met-expressing cells transfected with negative control (RNAi control) or Rac1 (RNAi Rac1) siRNAs. Below, the mean levels of indicated phosphorylated protein over total protein ± SEM obtained by densitometry of Western blots. N=3 independent biological replicates. (B) Colony area of M1268T Met-expressing cells grown in soft agar and treated with DMSO, Ehop-016 (4 μM) or NSC23766 (NSC) (100 μM). N=3 independent biological replicates. (C) Colony area of M1268T MET-expressing cells transfected with control, Vav2 or Tiam1 siRNAs grown in soft agar. N=4 independent biological replicates. (D) Colony area of M1268T-MET-expressing cells transiently transfected with GFP-Rac1-T17N dominant-negative construct. After flow cytometry separation, the cells expressing GFP-Rac1-T17N or the GFP negative cells (no GFP) were grown in soft agar and were treated with DMSO or PHA-665752 (PHA) (100 nM), N=4 independent biological replicates. Data (A to E) are mean values +/- SEM. P values were obtained with the Student’s t test. NS: non-significant, **P<0.01, ***P<0.005.

Fig. 6. Rac1 is required for mutant MET-dependent localization of Rac1 and mTOR at the plasma membrane.

(A) Confocal sections of wild-type and M1268T MET-expressing cells. Scale bar: 10μm. Cells were fixed and stained with DAPI (blue) and immunostained for Rac1 (green) and mTOR (red). (B and C) Percentage of cells with mTOR at the plasma membrane when cells were (B) treated with DMSO or PHA-665752 (PHA, 100 nM) or (C) transfected with negative control or Rac1 siRNA. N=3 independent biological replicates. Data (B and C) are mean values ± SEM. P values were obtained with the Student’s t test. NS: non-significant, *P<0.05.

Fig. 7. Rac1 associates with and promotes plasma membrane localization of mTOR and anchorage-independent growth of MET-mutant cells through its C-terminal RKR motif.

(A) Wild-type and M1268T MET-expressing cells were transiently transfected with GFP-Rac1-T17N or GFP-Rac1-T17N-AAA constructs. GFP-Rac1 was immunoprecipipated (IP) with GFP-Trap beads (GFP). Pull-down with control beads (Ctl) was performed. Western blots were performed to detect the GFP immunoprecipitates (IP) and the co-immunoprecipitated (co-IP) proteins mTOR and Tiam1. Lysates for each condition were also blotted and are indicated with “+”. Numbers below were obtained following densitometry of the blots. Thy represent the level of mTOR co-immunoprecipitated with GFP in M1268T MET-expressing cells. Thus, for each construct, mTOR level was normalised to GFP level. The obtained value was normalised to mTOR levels in lysate. The value was set as 1 for GFP-Rac1-T17N construct. Data are mean +/- SEM, N=3 independent biological replicates. (B and C) M1268T MET-expressing cells were transiently transfected with GFP-Rac1-T17N (B) or GFP-Rac1-T17N-AAA (B and C) constructs. After flow cytometry separation, the cells expressing GFP or the GFP negative cells (No GFP) were (B) immunostained for mTOR and the percentage of cells with mTOR at the plasma membrane was evaluated, and (C) grown in soft agar and treated with DMSO or PHA-665752 (PHA, 100 nM). Data are mean +/- SEM, N=3 independent biological replicates. P values were obtained with the Student’s t test. NS: non-significant, **P<0.01, ***P<0.005. (D) Model suggested for the role of Rac1 downstream of oncogenic MET to induce cell migration, and anchorage-independent growth.