Mammalian target of rapamycin complex 2 (mTORC2) is a critical determinant of bladder cancer invasion

Abstract

Bladder cancer is the fourth most common cause of cancer in males in the United States. Invasive behavior is a major determinant of prognosis. In this study, we identified mammalian target of rapamycin complex 2 (mTORC2) as a central regulator of bladder cancer cell migration and invasion. mTORC2 activity was assessed by the extent of phosphorylation of Ser473 in AKT and determined to be approximately 5-fold higher in specimens of invasive human bladder cancer as opposed to non-invasive human bladder cancer. The immortalized malignant bladder cell lines, UMUC-3, J82 and T24 demonstrated higher baseline mTORC2 activity relative to the benign bladder papilloma-derived cell line RT4 and the normal urothelial cell line HU1. The malignant bladder cancer cells also demonstrated increased migration in transwell and denudation assays, increased invasion of matrigel, and increased capacity to invade human bladder specimens. Gene silencing of rictor, a critical component of mTORC2, substantially inhibited bladder cancer cell migration and invasion. This was accompanied by a significant decrease in Rac1 activation and paxillin phosphorylation. These studies identify mTORC2 as a major target for neutralizing bladder cancer invasion.

Figure 1. Increased mTORC2 activity occurs in invasive bladder cancer.

We compared A, non-invasive low-grade papillary UCa (non-inv LG; scale bar 80 microns); B, non-invasive high-grade papillary UCa (non-inv HG; scale bar 80 microns); and C, invasive high-grade UCa to determine mTORC2 activity (scale bar 100 microns). D, immunoblotting for p-Ser473 as a marker of mTORC2 activity was performed and shows higher mTORC2 activity in both non-invasive HG and invasive lesions. E, densitometry was used to quantify p-Ser473/ total AKT signal intensities; averages and standard error of the mean (SEM) were normalized to the average signal intensity for the non-invasive LG samples.

Figure 2. Increased mTORC1 and mTORC2 activity is observed in malignant bladder cancer cells and can be induced with serum.

A, immunoblotting shows higher mTORC1 (p-S6) and mTORC2 (p-Ser473) activity in malignant UMUC-3, T24 and J82 bladder cancer cells, compared with normal urothelial HU1 cells and benign RT4 cells. B, serum stimulation induced mTORC2 activity in malignant bladder cancer cells. mTORC1 signaling was responsive to rapamycin in this model; mTORC2 was not.

Figure 3. Temporal profiling of mTORC2 activity in malignant bladder cancer cells shows an early response to serum stimulation.

A, serum stimulation in J82 cells shows high levels of mTORC2 (p-Ser473) activity by 5 min that persists through 6 h. B, rictor gene-silencing using pooled siRNA dramatically reduces detectable rictor protein and ablates p-Ser473 relative to that observed in cells transfected with NTC siRNA. No effect on mTORC1 signaling (p-S6) was detected.

Figure 4. mTORC2 is critical in bladder cancer cell migration and transwell invasion.

A, serum starvation blocked J82 bladder cancer cell motility. Addition of serum induced robust migration of J82 cells that were transfected with NTC siRNA, as determined using a modified scratch assay. Migration was inhibited by rictor gene-silencing. Scale bar 100 microns. B, quantification shows a reduction in the distance of migration of cells transfected with rictor-specific siRNA (□) compared with cells transfected with NTC siRNA (▲). Also shown are cells that were no serum-stimulated (♦). A total of 6 frames were analyzed for each time point per condition. C, transwell invasion on Matrigel-coated membranes shows higher numbers of malignant bladder cancer cells that invade. D, rictor gene-silencing significantly reduces cell invasion of all three malignant bladder cell lines in transwell assay. (*) statistically significant comparison using student t-test, p<0.05).

Figure 5. mTORC2 regulates invasion of bladder cancer cells into the human bladder wall.

A, schematic showing orientation of human bladder wall invasion assay that includes lamina propria (LP), muscularis propria (MP) and perivesical fat (PV fat). B, H&E section showing invading J82 cells in the lamina propria after 72 h (scale bar 200 microns) C, cells highlighted in previous panel showed positive immunostaining for pancytokeratin (scale bar 100 microns). D, immunofluorescent stain shows invasion of malignant J82 cells (red), at 72 hr (scale bar 50 microns). E, quantification of tissue invasion by RT4 and J82 cells (microns). F, rictor gene-silencing significantly inhibits invasion of bladder all by J82 cells. (*) statistically significant comparison using student t-test, p<0.05).

Figure 6. mTORC2 regulates cell spreading and Rac1 activity in bladder cancer cells.

J82 cells were either transfected with rictor-specific siRNA or NTC siRNA 72 h prior to the start of the experiment. A, immunofluorescent staining shows broad cytoplasmic processes and peripheral phospho-paxillin in cells that are transfected with NTC siRNA. This is reduced in cells with rictor gene-silencing. Scale bar 30 microns. B, cell-surface areas were determined for J82 cells transfected with NTC (♦) or rictor-specific siRNA (▲) and allowed to spread for the indicated times (mean±SEM, n>15 for each time point; (*) p<0.05). C, immunoblotting for phospho-paxillin in J82 cells that were transfected with rictor-specific or NTC siRNA and allowed to adhere for 30 min. D, GTP-Rac1 was determined in J82 cells transfected with NTC or Rictor-specific siRNA. Rictor knockdown reduced Rac1 levels to 60.2% of control. Cells with GTPγS were analyzed as a positive control and GDP as negative control. E, GTP-loaded RhoA in J82 cells transfected with rictor-specific or NTC siRNA. Rictor knockdown reduced RhoA levels to 83.3% of control.