Figure 1Exogenous arginine increases bladder cancer migration and invasion. A: Representative images of modified scratch migration assay to assess two-dimensional cell migration in response to arginine in 0.1% serum-containing media. The first dotted line indicates the cell migration at 0 hour, and the second dotted line indicates the cell migration at 24 hours. B: Quantitation shows a dose-dependent increase in T24 migration after arginine addition. C: Dose-dependent increase in transwell invasion in T24 cells after arginine addition. D: Representative images of DAPI-stained cells invading through Matrigel-coated invasion chambers. E: Immunoblot assay of endothelial nitric oxide synthase (eNOS) and inducible nitric oxide synthase (iNOS) in a panel of bladder cell lines. F: Quantitation shows increased expression of eNOS and iNOS in invasive cell lines compared with noninvasive cell lines. Data are expressed as means ± SEM (B, C, and F). n = 3. ∗∗∗P < 0.001 (analysis of variance). Original magnification, ×100 (A, D, and E).

Figure 2Endothelial nitric oxide synthase (eNOS) and inducible nitric oxide synthase (iNOS) regulate motility of bladder cancer cells. A: Immunoblots of T24, J82, and RT112 cells transfected with nontargeting control siRNA (si-NTC), eNOS siRNA (si-eNOS), and iNOS siRNA (si-iNOS)–show target-specific effects. B: Representative images of invading T24 cells after transfection with si-NTC, si-eNOS, or si-iNOS. C and D: Quantification of T24 cell invasion (C) and J82 cell invasion (D) after siRNA transfection. E: Representative images of cell migration assay performed with si-RNA transfected T24 cells. The first dotted line indicates the cell migration at 0 hour, and the second dotted line indicates cell migration at 24 hours. F–H: Quantification of migration data after siRNA transfection in T24, J82, and RT112 cells. I: Application of NOS inhibitors L-Name and 1400 W and the soluble guanylyl cyclase inhibitor ODQ reduce bladder cancer cell migration in T24 cells. Inhibition effects can be overcome by co-application of NO or cGMP analogues. Data are expressed as means ± SEM. n = 3. ∗P < 0.05, ∗∗∗P < 0.001, and ∗∗∗∗P < 0.0001 (t-test). Original magnification, ×100. Deta, Deta-NONOate.

Figure 3Inducible nitric oxygen synthase (iNOS) and endothelial nitric oxide synthase (eNOS) expression shows a progressive increase in expression with advancing human bladder cancer stage. A: Normal urothelium and noninvasive and invasive urothelial carcinoma were assessed for eNOS and iNOS expression by immunohistochemistry (IHC). B: Evaluation of intensity of IHC shows an increase in eNOS expression in urothelial tumor relative to normal urothelium, with an additional increase in expression within paired lymph node metastasis. C: iNOS expression is primarily increased in urothelial cancer relative to normal urothelium, without an additional increase in metastatic deposits. D and E: eNOS expression but not iNOS expression in primary urothelial carcinoma tumors is associated with a greater risk of metastasis. Data are expressed as means ± SEM (B–E). n = 3. ∗P < 0.05, ∗∗∗∗P < 0.0001. Original magnification, ×100 (A).

Figure 4Nitric oxide (NO) scavenger and donor application confirmed that NO is involved in motility effects. A and B: Representative images of T24 cells with increasing concentrations of the NO scavenger cobinamide during a migration assay (the first dotted line indicates the cell migration at 0 hour, and the second dotted line indicates cell migration at 24 hours) (A) and quantitation of these effects (B). C and D: Representative images of T24 invading cell number with increasing cobinamide concentration (C) and quantitation of these effects (D). E: Migration distance increases with increasing concentration of the NO donor Deta-NONOate. F: Similar effects were with Deta-NONOate application in the invasion assay. Data are expressed as means ± SEM (B and D–F). n = 3. ∗∗∗P < 0.001 (analysis of variance). Original magnification, ×100 (A and C).

Figure 5Inducible nitric oxide synthase (iNOS) and mammalian target of rapamycin complex 2 (mTORC2) activity is enriched in invadopodia and may mediate local effects. A: Actin (green) shows rearrangement after cobinamide or Deta-NONOate treatment in T24 cells. The arrows indicated the aggregates developed by actin cytoskeleton rearrangement. Cell nuclei are shown in blue (DAPI stain). Serum-free and serum-containing media were used as controls. B: Schematic of invadopodia assay. C: Immunoblot of rictor (mTORC2 component), iNOS, endothelial nitric oxide synthase (eNOS), phospho-AKT (pAKT) (Ser473) (mTORC2 target), total AKT, phospho-S6 (pS6) Ser235/236 (mTORC1 target), total S6, Rac1 (invadopodia marker), and histone H3 (cell body marker) in cell body (CB) and invadopodia (IN) fractions of untreated T24 cells. D: Quantitation shows a dose-dependent reduction in invadopodia formation with cobinamide treatment of T24 cells. E: Images showing cobinamide effect in one of three independent experiments. F: By contrast, quantitation shows a dose-dependent increase in invadopodia formation as evident with Deta-NONOate application to T24 cells. G: Deta-NONOate images from one representative experiment of three. Data are expressed as means ± SEM (B–D and F). n = 3. ∗P < 0.05, ∗∗∗P < 0.001 (t-test). Original magnification, ×100 (A, E, and G).

Figure 6Mammalian target of rapamycin complex 2 (mTORC2) activity regulates endothelial nitric oxide synthase (eNOS) and inducible nitric oxide synthase (iNOS) expression and cGMP formation. A: Immunoblot of rictor, iNOS, eNOS, phospho-AKT (pAKT0 Ser473, total AKT, phospho-S6 (p-S6) Ser235/236, and total S6 in T24 cells stably transfected with nontargeting control (NTC), eNOS, or iNOS shRNA. Rac1 was used as an invadopodia control and histone H3 as a cell body control. B: Immunoblot comparison of siRNA NTC (si-NTC) and siRNA rictor (si-rictor) transfected T24 cells, with (plus sign) and without (minus sign) serum stimulation, shows that rictor ablation reduces iNOS and eNOS protein expression. Although elimination of mTORC2 activity through lack of pAKT Ser473 signal is evident with rictor silencing, mTORC1 activity, by induction of p-S6 with serum stimulation, remains intact. C and D: Stable transfection of T24 cells with shRNA NTC (sh-NTC) or shRNA rictor (sh-rictor) confirms ablation of mTORC2 signaling effects, with a modest reduction in iNOS and eNOS proteins (C), as confirmed by NOS/actin ratios analyzed by densitometry (D). E: T24 cells show changes in cGMP levels treated with 50 μmol/L cobinamide or Deta-NONOate. F: Rictor and NOS silencing reduce cGMP levels in T24 cells. Data are expressed as means ± SEM. n = 3. ∗P < 0.05 versus untreated (t-test); †P < 0.05 versus si-NTC (t-test).

Figure 7Endothelial nitric oxide synthase (eNOS), inducible nitric oxide synthase (iNOS), and rictor are important in the establishment of zebrafish tail vein metastasis. A: Representative images show that fluorescently labeled, highly invasive J82, T24, and UM-UC-3 bladder cancer cells demonstrate a higher number of zebrafish tail vein metastasis compared with the less invasive RT4 and RT112 cells. The metastasized cancer cells were indicated by arrows. Embryos were used at 48 hours post fertilization, and injections were into the perivitelline cavity. B: UM-UC-3 cells transfected with either shRNA rictor (sh-rictor) or shRNA iNOS (sh-iNOS) (red) were co-injected with control shRNA NTC (sh-NTC) cells (green) into zebrafish to compare effects of rictor and iNOS on frequency of metastasis. C and D: Quantitation of metastasis by manual counting. E: cGMP levels were also decreased with either sh-rictor or sh-iNOS transfection of UM-UC-3 cells. Data are expressed as means ± SEM (C–E). n = 3. ∗P < 0.05 for target shRNA-treated cells versus control sh-NTC cells (t-test). Original magnification, ×100 (A and B).