If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The mammalian target of rapamycin (mTOR) and associated phosphatidylinositol 3-kinase/AKT/mTOR signaling pathway is commonly up-regulated in cancer, including bladder cancer. mTOR complex 2 (mTORC2) is a major regulator of bladder cancer cell migration and invasion, but the mechanisms by which mTORC2 regulates these processes are unclear. A discovery mass spectrometry and reverse-phase protein array–based proteomics dual approach was used to identify novel mTORC2 phosphoprotein targets in actively invading cancer cells. mTORC2 targets included focal adhesion kinase, proto-oncogene tyrosine-protein kinase Src, and caveolin-1 (Cav-1), among others. Functional testing shows that mTORC2 regulates Cav-1 localization and dynamic phosphorylation of Cav-1 on Y14. Regulation of Cav-1 activity by mTORC2 also alters the abundance of caveolae, which are specialized lipid raft invaginations of the plasma membrane associated with cell signaling and membrane compartmentalization. Our results demonstrate a unique role for mTORC2-mediated regulation of caveolae formation in actively migrating cancer cells.
Bladder cancer is the most common urinary tract malignancy, with an estimated 81,000 new bladder cancer cases and 17,000 bladder cancer–associated deaths this year alone in the United States.
Although most patients present with noninvasive disease, progression to invasive cancer occurs in approximately 40% to 60% of patients and results in an increased risk of metastasis and reduced disease-specific survival.
mTOR is an evolutionary conserved serine/threonine kinase that can integrate extracellular and environmental cues, such as growth factor signaling and nutrient status, to affect a diverse array of cell processes, such as cell proliferation, metabolism, and motility. Specifically, mTOR mediates these functions as an essential component of two multiprotein complexes, mTORC1 and mTORC2. These complexes can be distinguished by their subunit composition (raptor and PRAS40 for mTORC1 and rictor, stress-activated map kinase interacting protein 1 (mSin1), and protor for mTORC2) and their downstream functions.
as well as increased mTORC2 activity occurring in invasive, high-stage human bladder cancers. Given the important role of mTORC2 in regulating bladder cancer cell migration and invasion, a proteomics-based approach using mass spectrometry (MS) and reverse-phase protein array (RPPA) was used to discover and identify novel targets of mTORC2 signaling in motile and nonmotile conditions.
Phosphopeptide enrichment coupled to MS and RPPA represent two powerful complementary phosphoproteomic approaches that allow for discovery-based, quantifiable detection of proteins within biological samples. Although these technologies are able to effectively uncover important signaling events and map the activated signaling architecture of input samples, each have their own limitations.
Although an MS approach can provide a global unbiased view of phosphoprotein expression, it is also possible that important signaling-related phosphoproteins may be undetected because of low relative protein abundance within a sample, low stoichiometry of phosphorylated signaling proteins compared with phosphorylated high abundance proteins and/or rapid degradation or dephosphorylation of a protein before analysis. By contrast, RPPA has greater analytical sensitivity than MS because of its ability to measure the phosphorylation state of very low abundance signaling proteins from microscopic quantities of cells. However, the biggest limitation of RPPA is the availability and dependence of antibodies for detection of proteins. Furthermore, RPPA can be viewed as a somewhat biased approach because arrays of specific antibodies are selected to probe a biological sample(s) for changes in protein expression and/or activity. Nevertheless, both MS and RPPA serve as increasingly important and complementary technologies for protein signaling analysis in cancer and other diseases.
Here, a parallel MS- and RPPA-based proteomics discovery approach was used to elucidate novel downstream targets of mTORC2 signaling in motile bladder cancer cells, followed by validation and functional testing of a subset of protein signaling networks that appeared relevant for migration and invasion. Multiple classes of proteins regulated by mTORC2, including mediators of cell morphology, cell assembly and organization, cell adhesion, cytoskeletal rearrangement, and cell motility, were identified. A subset of putative mTORC2-mediated targets were validated, and the ability of mTORC2 to regulate dynamic phosphorylation of caveolin-1 (Cav-1) on Y14 and Cav-1 cellular localization, both of which represent novel downstream effects of mTORC2, was functionally tested. Furthermore, mTORC2 can also induce up-regulation of cavin-1, an essential component for caveolae formation and function. The results of this study have identified a novel role for mTORC2-regulation of caveolae formation during bladder cancer cell motility.
Materials and Methods
Human J82 and T24 bladder cancer cells were obtained and authenticated from the ATCC (Manassas, VA). Cells were grown in RPMI 1640 medium (Thermo Fisher Scientific, Grand Island, NY) supplemented with 10% fetal bovine serum (Gibco) and maintained at 37°C in a humidified chamber containing 5% CO2.
Gene Silencing and Serum Stimulation
SMARTpool siGENOME siRNA against RICTOR (siRictor; catalog number M-016984-02) and CAV1 (catalog number M-003467-01) were purchased from GE Dharmacon (Lafayette, CO) and transfected into cells using Lipofectamine RNAiMAX reagent (Invitrogen, Carlsbad, CA). Pooled siGENOME nontargeting control siRNA (siNTC) numbers 2-5 (catalog numbers D-001210-02 to -05) was used for controls. Briefly, cells were plated onto 100-mm dishes and incubated overnight. Each plate of cells was transfected in Opti-MEM reduced serum media (Gibco) containing siRNA (160 pmol) and maintained for 72 hours. For serum stimulation, media was aspirated from plates and cells were washed three times with phosphate-buffered saline (PBS) before the addition of RPMI 1640 medium supplemented with 10% fetal bovine serum during the last hour of gene silencing.
Whole cell extracts were prepared using radioimmunoprecipitation assay buffer (25 mmol/L Tris hydrochloride, pH 7.6, 150 mmol/L NaCl, 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS) containing PhosSTOP phosphatase and cOmplete Mini EDTA-free protease inhibitor cocktails (Roche, Mannheim, Germany). Cell lysates were cleared by centrifugation at 16,000 × g for 15 minutes at 4°C, and protein concentration was determined by the bicinchoninic acid method. Equal amounts (10 μg) of protein were separated by SDS-PAGE using 4% to 15% Tris-Glycine Gradient gels (Bio-Rad, Hercules, PA) and immobilized onto polyvinylidene fluoride membranes. Membranes were blocked in 5% (w/v) nonfat dry milk in 50 mmol/L Tris hydrochloride, pH 7.4, 150 mmol/L NaCl (TBS) plus 0.1% Tween-20 (TBS-T) and then incubated at 4°C overnight with the appropriate primary antibody diluted in 5% (w/v) nonfat dry milk or bovine serum albumin. The next day, membranes were washed with TBS-T (three times for 5 minutes each) then incubated with the appropriate horseradish peroxidase–conjugated secondary antibody for 1 hour at room temperature. Membranes were washed again in TBS-T (three times for 5 minutes each) and proteins detected using Clarity Western ECL blotting substrate (Bio-Rad). Primary antibodies for 4EBP1 (1:1000; catalog number 9644), pan-AKT (1:1000; catalog number 4685), autocrine motility factor receptor (1:1000; catalog number 9590), Cav-1 (1:1000; catalog number 3267), cofilin (1:1000; catalog number 5175), epidermal growth factor receptor (EGFR) (1:1000; catalog number 4267), focal adhesion kinase (FAK; 1:1000; catalog number 13009), human epidermal growth factor receptor 2 HER2; 1:1000; catalog number 4290), mTOR (1:1000; catalog number 2983), N-cadherin (1:1000; catalog number 13116), phosphorylated (p-) 4EBP1 T37/46 (1:2000; catalog number 2855), p-AKT S473 (1:2000; catalog number 4060), p-AKT T308 (1:1000; catalog number 4056), p-CAV-1 Y14 (1:1000; catalog number 3251), p-cofilin S3 (1:1000; catalog number 3313), p-EGFR Y1173 (1:1000; catalog number 4407), p-ezrin T567/radixin T564/moesin T558 (1:1000; catalog number 3141), p-FAK Y397 (1:1000; catalog number 8556), p-FAK Y567/577 (1:1000; catalog number 3281), p-FAK Y925 (1:1000; catalog number 3284), p-HER2 Y1248 (1:1000; catalog number 2247), p-HER2 Y877 (1:1000; catalog number 2241), p-mTOR S2448 (1:1000; catalog number 5536), p-p70-S6K T389 (1:1000; catalog number 9234), p–platelet-derived growth factor receptor (p-PDGFR)-α Y754 (1:1000; catalog number 2992), p-PDGFR-β Y751 (1:1000; catalog number 4549), p-S6 S235/236 (1:2000; catalog number 4858), p-S6 S240/244 (1:2000; catalog number 5364), p-SRC Y527 (1:1000; catalog number 2105), p-SRC Y416 (1:1000; catalog number 6943), p-Talin S425 (1:1000; catalog number 5426), p–vascular endothelial growth factor receptor 2 Y1175 (1:1000; catalog number 3770), p70-S6K (1:1000; catalog number 2708), rictor (1:1000; catalog number 2140), S6 (1:2000; catalog number 2217), snail (1:1000; catalog number 3879), SRC (1:1000; catalog number 2123), zonula occludens-1 (ZO-1; 1:1000; catalog number 8193), and β-catenin (1:1000; catalog number 8480) were purchased from Cell Signaling (Danvers, MA). Primary antibodies for liprin α-1 (1:1000; catalog number ab26192), liprin β-1 (1:1000; catalog number ab104117), and cavin-1 (1:1000; catalog number ab48824) were purchased from AbCam (Cambridge, UK). Primary antibodies for p-Adducin S662 (1:1000; catalog number 06-820) and mSin1 (1:1000; catalog number NBP1-89569) were purchased from EMD Millipore (Burlington, MA) and Novus (Centennial, CO), respectively. Actin antibody (1:5000; Sigma-Aldrich, St. Louis, MO) was routinely used as a loading control. Densitometry analysis was performed using ImageJ software version 1.48 (NIH, Bethesda, MD; http://imagej.nih.gov/ij).
Immunoprecipitation of Rictor
Immunoprecipitation of rictor-containing mTOR complexes was performed as previously described with minor modification.
T24 and J82 cells were lyzed for 30 minutes on ice with 0.3% CHAPS lysis buffer (40 mmol/L HEPES, pH 7.5, 120 mmol/L NaCl, and 1 mmol/L EDTA) containing protease and phosphatase inhibitors. After centrifugation at 16,000 × g for 15 minutes, fresh whole cell lysates (1 mg) were incubated with 3 μg of rictor (catalog number 5379; Cell Signaling) or 3 μg of rabbit IgG isotype control Sepharose bead conjugates (catalog number 3423; Cell Signaling) for 2 hours with rotation at 4°C. Beads were then collected by centrifugation at 2000 × g for 5 minutes and washed three times with lysis buffer. The captured immunoprecipitates were eluted by heating in 35 μL 2× Laemmli sample buffer containing 50 mmol/L 2-mercaptoethanol for 5 minutes at 95°C and then analyzed by immunoblot analysis. Co-immunoprecipitation of endogenous Cav-1 was detected with a monoclonal mouse anti–Cav-1 antibody (catalog number MAB5736; R&D Systems, Minneapolis, MN).
Cell Migration Assay
Modified scratch-wound migration assays were performed as previously described.
Cell lysates were prepared by resuspension for 1 hour in lysis buffer consisting of Tris hydrochloride (50 mmol/L, pH 7.4), NaCl (150 mmol/L), Triton X-100 (0.5% w/v), NP-40 (0.5% w/v), 80 mmol/L dithiothreitol, 10 μL/mL of protease inhibitor cocktails (Sigma-Aldrich), 1 mmol/L phenylmethylsulfonyl fluoride, 1 mmol/L Na3VO4, and PhosSTOP phosphatase inhibitor cocktail (Roche, Mannheim, Germany), sonicated for 30 seconds, and centrifuged at 16,000 × g for 10 minutes. The supernatants were precipitated with 4 volumes of acetone (Sigma-Aldrich) overnight at −20°C and centrifuged at 9000 × g for 5 minutes. The pellets were dried by lyophylization (Heto, Dry Winner) for 2 hours.
The cell pellets were resuspended in 200 μL of 8 mol/L urea, and protein concentrations were measured by Bradford Assay (Bio-Rad). Protein samples were then reduced with 10 mmol/L dithiothreitol for 30 minutes at 37°C and then alkylated by 50 mmol/L iodoacetamide for 20 minutes at room temperature. The concentrated urea in the sample was diluted to a final concentration of 2 mol/L, and the proteins were digested by trypsin at 37°C for 6 hours in a buffer containing ammonium bicarbonate (50 mmol/L, pH 9). The digestion mixture was then acidified by adding glacial acetic acid to a final concentration of 2% and desalted by SepPak C18 column (Waters Corp., Milford, MA).
Phosphopeptides were enriched from the desalted 1-mg tryptic peptides using TiO2 column (200 μm × 2 cm) packed in-house.
A total of 100 fmol of standard phosphopeptide angiotensin II phosphate was added to the SepPak-cleaned sample. The sample was then mixed with an equal volume of loading buffer (200 mg/mL DHB, 5% trifluoroacetic acid, 80% acetonitrile), and loaded into the TiO2 column using the Pressure Cell (Brechbühler Inc., Schlieren, Switzerland) with flow rate of 3 μL/minute. The column was washed by 200 μL of Wash Buffer 1 (40 mg/mL dihydroxybenzoic acid, 2% trifluoroacetic acid, 80% acetonitrile) and 2 × 200 μL of a second wash buffer 2 (2% trifluoroacetic acid, 50% acetonitrile) to remove nonphosphopeptides. Phosphopeptides were eluted from the column with the elution buffer (5% ammonia solution). Ammonia in the eluate was removed by lyophilization (approximately 3 minutes), and the sample was acidified by adding glacial acetic acid to a final concentration of 2% and desalted by ZipTip (EMD Millipore).
The purified phosphopeptides were analyzed by high-sensitive reversed-phase liquid chromatography coupled nanospray tandem MS using an LTQ-Orbitrap mass spectrometer (Thermo Fisher Scientific). LTQ-Orbitrap provides high-accuracy mass measurement that is essential for the validation of modified peptide identification and the reduction of false-positive identification. The reversed-phase LC column was slurry-packed in house with 5 μm of 200 Å pore size C18 resin (Michrom BioResources, Inc., Auburn, CA) in a 100 μm i.d. × 10-cm-long piece of fused silica capillary (Polymicro Technologies, Phoenix, AZ) with a laser-pulled tip. After packing, the new column, the high-performance liquid chromatography system (Surveyor MS Pump Plus from Thermo Fisher Scientific), and the LTQ-Orbitrap were tested by analyzing 100 fmol of Yeast Enolase Standard & Tryptic Digestion (catalog number PTD/00001/46; Michrom Bioresources, Inc.) to ensure that stable electrospray ionization, desired mass accuracy, peak resolution, peak intensity, and retention time could be obtained. Additional iteration was performed to ensure reproducibility. A total of 100 fmol of standard peptide angiotensin I was spiked into the sample as an internal standard. After sample injection, the column was washed for 5 minutes with mobile phase A (0.1% formic acid), and peptides were eluted using a linear gradient of 0% mobile phase B (0.1% formic acid, 80% acetonitrile) to 40% B in 120 minutes at 200 nL/minute then to 100% B in an additional 10 minutes. The high-performance liquid chromatography gradient was shallower than that of general proteomic analysis because phosphopeptides are relatively hydrophilic. Before and after analyzing one sample, the column was washed with high-performance liquid chromatography mobile phase B for 30 minutes, then mobile phase A for 20 minutes at a high flow rate (1 μL/minute) to reduce potential carryover. The LTQ-Orbitrap mass spectrometer was operated in a data-dependent mode in which each full MS scan (60,000 resolving power) was followed by eight MS/MS scans where the eight most abundant molecular ions were dynamically selected and fragmented in by collision-induced dissociation using a normalized collision energy of 35%. The fragmented ions were detected by LTQ. The dynamic exclusion time was 30 seconds, and the dynamic exclusion size was 200. The FT master scan preview mode, charge state screening, monoisotopic precursor selection, and charge state rejection were enabled so that only the 1+, 2+, and 3+ ions were selected and fragmented by collision-induced dissociation.
Tandem mass spectra collected by Xcalibur version 2.0.2 were searched against the National Center for Biotechnology Information human protein database (released in September 2009 with 37,391 entries) using Proteome Discoverer software version 2.1 (Thermo Fisher Scientific) with full tryptic cleavage constraints, static cysteine alkylation by iodoacetamide, variable methionine oxidation, and variable phosphorylation of Ser/Thr/Tyr. Mass tolerance for precursor ions was 5 ppm, and mass tolerance for fragment ions was 0.25 Da. The Proteome Discoverer search results were filtered by criteria Xcorr versus charge 1.5, 1.8, 2.5 for 1+, 2+, 3+ ions; ranked top #1; probability of randomized identification of peptide <0.1. Confident peptide identification were determined using these stringent filter criteria for database match scoring followed by manual evaluation of the results. The false discovery rate was estimated by searching a combined forward-reversed database as described by Elias and Gygi.
Cellular lysates were printed in triplicate onto nitrocellulose-coated slides (Grace Bio-labs, Bend, OR) using an Aushon 2470 arrayer (Aushon BioSystems, Billerica, MA). Before proceeding with immunostaining, each array was treated with Reblot antibody stripping solution (Chemicon, Temecula, CA) for 15 minutes and blocked in I-block solution (Tropix, Bedford, MA) for 1 hour to reduce nonspecific binding. Each array was probed with one primary antibody on an automatic Autostainer (Dako Cytomation, Carpinteria, CA) using the Catalyzed Signal Amplification System kit (Dako Cytomation). Antibody specificity was tested for single-band specificity and ligand induction via Western blot analysis. Fluorescent detection was achieved using the streptavidin-conjugated IRDye680 (LI-COR Biosciences, Lincoln, NE) according to the manufacturer's instructions. The total amount of protein contained in each sample was measured by Sypro Ruby Protein Blot Stain (Molecular Probes, Eugene, OR) as previously described.
Images were acquired using the PowerScanner (TECAN, Mönnedorf, Switzerland), and spot intensity values were quantified using MicroVigene software version 18.104.22.168 (VigeneTech, Carlisle, MA) as previously described.
Immunofluorescence staining of bladder cancer cells grown on poly-l-lysine coated glass coverslips was performed using standard procedures. Cells were fixed with ice-cold methanol for 10 minutes and blocked with 10% normal goat serum (Gibco) in PBS with 0.1% Triton X-100 for 30 minutes at room temperature. Cav-1 was stained using an Alexa Fluor 488–conjugated mouse anti–Cav-1 monoclonal antibody (1:10; catalog number IC5736G; R&D Systems) for 1 hour at room temperature. Coverslips were mounted onto glass slides using ProLong Gold Antifade mountant with DAPI for nuclear counterstaining (Molecular Probes). Images were obtained on a Leica DM IRE2 inverted fluorescent microscope (Buffalo Grove, IL) with a Hamamatsu ORCA-100 digital camera (Sewickley, PA) using SimplePCI software version 6 (Hamamatsu Corp.).
Preparation of cell lysates for transmission electron microscopy was performed using standard procedures. J82 cells were collected by trypsinization and fixed with Karnovsky's fixative (2.5% glutaraldehyde and 4% paraformaldehyde in 0.1 mol/L PBS, pH 7.4) at 4°C for a minimum of 3 hours. After rinsing with PBS, cells were postfixed in 1.5% osmium tetroxide in PBS for 45 minutes. Cells were then washed with deionized water and dehydrated in ascending concentrations (35% to 100%) of ethanol for 15 minutes each. Samples were then immersed in propylene oxide, 1:1 propylene oxide, and epoxy resin mixture (Quetol 651 embedding kit; Electron Microscopy Sciences, Hatfield, PA), 2:3 propylene oxide and epoxy resin mixture, and then overnight at 75°C with pure epoxy resin mixture under constant agitation and vacuum to optimize specimen infiltration. Thin sections (60 nm) were then cut using an RMC MTX-L ultra-microtome (Boeckeler Instruments, Inc., Tucson, AZ) and mounted on thin-bar 200-mesh copper grids. The grids were then stained with methanolic uranyl acetate saturated solution followed by bismuth subnitrate staining for contrast enhancement. Samples were examined with a Zeiss EM 10 C transmission electron microscope (Oberkochen, Germany) and images captured using a GATAN model 785 digital camera (Pleasanton, CA).
Formalin-fixed, paraffin-embedded tumor blocks from 53 patients were curatively obtained following University of California, San Diego Institutional Review Board approval. Invasive disease (pT2+) was enriched in this patient population. Clinicopathologic information for this patient cohort is provided in Table 1.
Briefly, tissue sections from paraffin-embedded blocks were cut onto precoated slides, followed by deparaffinization, rehydration, and heat-induced antigen retrieval using sodium citrate buffer. Sections were then blocked with 10% normal goat serum in TBS for 1 hour at room temperature and then incubated overnight at 4°C with primary antibody against Cav-1 diluted in blocking buffer. After four washes in TBS, blocking of endogenous peroxidase was performed by incubation with 3% H2O2 in TBS for 10 minutes. For enzymatic detection, tissue was counterstained with species-appropriate prediluted horseradish peroxidase polymer secondary antibody and chromagen developed with diaminobenzidine using a Rabbit/Mouse specific horseradish peroxidase/diaminobenzidine (ABC) Detection IHC Kit according to the manufacturer's instructions (AbCam).
Ingenuity Pathway Analysis Analysis
The Ingenuity Pathway Analysis program (Qiagen, Venlo, Netherlands) was used to analyze phosphoproteins that were elevated in the control or RICTOR knockdown group. P < 0.05 was used to select differentially expressed phosphoproteins. The shortlisted phosphoprotein names were then mapped to corresponding Entrez gene identification numbers. Both pathway and gene ontology enrichment analyses were conducted to identify candidates involved in some type of molecular and cellular component function as classified by the gene ontology nomenclature.
The cBio Portal for Cancer Genomics web tool was used to analyze the provisional RNA sequencing data from The Cancer Genome Atlas (TCGA; https://portal.gdc.cancer.gov/projects/TCGA-BLCA, last accessed January 25, 2019) for bladder urothelial carcinoma. This study includes 408 tumor samples. All searches were performed according to the cBioPortal instructions.
Tests for statistical significance were determined using the t-test with P < 0.05. Comparison between Kaplan-Meier survival curves was performed using the Mantel-Cox log rank test. All statistical analyses were performed using GraphPad Prism software version 6 (GraphPad Software, La Jolla, CA).
MS-Based Analytics to Identify mTORC2 Targets
Pharmacologic inhibitors of mTOR, such as PP242, KU-0063795, and Torin-2, have served as valuable tools to study various biological processes associated with phosphatidylinositol 3-kinase (PI3K)/AKT/mTOR signaling.
However, these inhibitors affect both mTOR-containing complexes, making it difficult to distinguish mTORC1- versus mTORC2-mediated effects. Accordingly, targeted gene silencing of RICTOR was performed to selectively ablate mTORC2 activity, which was confirmed by reduced AKT S473 phosphorylation. This approach does not alter mTORC1-dependent phosphorylation of downstream signaling effectors.
A migration model was implemented whereupon serum stimulation after serum starvation was used to induce cell migration, and this was compared with nonmotile, serum-starved cells transfected with siNTC.
MS was used as a phosphoproteomic discovery approach to capture global changes of phosphorylation events in these cell populations that were increased or decreased by the presence or absence of mTORC2. After downstream workflows, stringent filter criteria from MS analysis identified 212 and 199 unique peptides in control and rictor-silenced cells, respectively. Ten unique phosphoproteins in control siRNA-transfected cells and 12 unique phosphoproteins in rictor-silenced were identified by MS analysis in at least two of the three biological replicates (Table 2).
Table 2Phosphoproteins Identified by Mass Spectrometry Analysis Putatively Regulated by mTORC2
Increased target protein phosphorylation by mTORC2
Ingenuity Pathway Analysis was used to perform functional enrichment of these proteins to determine the biological and/or signaling pathways affected by the presence or absence of mTORC2 activity. Ingenuity Pathway Analysis analysis revealed 26 and 22 significantly altered (P < 0.05) cellular functions up-regulated or down-regulated by mTORC2 activity, respectively (Figure 1, A and B). Among these, the top five categories up-regulated by mTORC2 activity were cellular assembly and organization, cell death and survival, cellular growth and proliferation, cellular function and maintenance, and gene expression. By contrast, the top five most overrepresented categories down-regulated by mTORC2 activity included cell death and survival, cellular growth and proliferation, cellular development, cell morphology, and cellular assembly and organization.
Given the role of mTORC2 in regulating cytoskeletal remodeling, cell adhesion, and cell motility, the investigation was focused on the novel targets of mTORC2 for phosphoproteins involved in cellular assembly and organization, cell morphology, and cellular movement. Among these functions, mTORC2 increased phosphorylation of eight proteins and decreased phosphorylation of 32 proteins (Figure 2A and Table 3). One protein, nonmuscle myosin heavy chain 10, was identified in both categories. The Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) database was used to assemble a protein interaction network of the 39 commonly represented phosphoproteins.
Thirty-six phosphoproteins formed one large distinct network with significantly enriched interactions (Figure 2B). Two of the top five gene ontology biological processes that were significantly enriched within this network included movement of cell or subcellular components (n = 19/35; P = 1.26 × 10−9) and regulation of cell motility (n = 14/35; P = 7.76 × 10−9).
Table 3Common Phosphoproteins Represented Among Cellular Assembly and Organization, Cell Morphology, and Cellular Movement as Determined by IPA Analysis
Toward the goal of identifying novel signaling effectors of mTORC2 activity, the differential protein and phosphoprotein changes that occur within the bladder cancer cell migration model were examined using RPPA analysis. This method serves as a complementary approach to MS and the global analysis of mTORC2 signaling perturbations in the rictor-silenced cell migration model. Lysates were probed with ≥150 antibodies that represented a diverse set of signaling pathways, biological functions, and protein classes, including PI3K/AKT/mTOR signaling, mitogen-activated protein kinase (MAPK)/extracellular-signal-regulated kinase (ERK) signaling, Janus kinase (JAK)/STAT signaling, AMP-activated protein kinase (AMPK) signaling and energy sensing, PKC signaling, cell stress and immune response, focal adhesion and cell motility, cell cycle regulation, cell death and apoptosis, translation and transcription, tumor suppressors and oncogenes, and receptor tyrosine kinases (RTKs) (Supplemental Figure S1). Rictor silencing and consequent loss of mTORC2 activity decreased AKT S473 phosphorylation, whereas downstream mTORC1 targets (p-AKT T308, p-PRAS40 T246, p-S6 S235/236 and S240/244, p-p70-S6K S371, T389 and T412, and p-4EBP1 S65 and T70) were unaffected in serum-stimulated cells (Figure 3).
RPPA analysis also identified that mTORC2 can regulate additional phosphoprotein targets across multiple signaling pathways. Loss of mTORC2 activity resulted in activation of cell death and apoptosis-signaling effectors, including caspases 3, 6, 7, and 9, poly (ADP-ribose) polymerase, and increased phosphorylation of several members of the Bcl-2 family of apoptosis proteins, which confirmed previous reports linking mTORC2 with apoptosis in leukemic, breast, and non–small cell lung carcinoma cell line models.
By contrast, reduced pathway activation of numerous signaling effectors associated with the MAPK/ERK, JAK/STAT, AMPK, and PKC pathways, as well as proteins involved in cell motility and focal adhesion (p-adducin S662, p-cofilin S3, p-CrkII Y221, and p-SRC Y416), cell cycle regulation (cyclins A and D1, p-Chk-1 S345, and p-Rb S780), and transcription and translation (p-eIF2a S51, p-eIF4E S209, and p-FKHR S256 and T24) was observed.
Validation of mTORC2 Effects on Discovery Phosphoprotein Targets
To validate the findings from parallel MS and RPPA analysis, a large subset of proteins and phosphoproteins was selected to confirm the directional change in expression or phosphorylation status between siNTC or siRictor-transfected bladder cancer cells through immunoblot analysis under nonmotile (serum starved) and motile (serum stimulated) conditions in J82 and T24 bladder cancer cells.
The effects of mTORC2 signaling were confirmed on known targets to verify the specificity of rictor silencing in selectively reducing mTORC2 activity. Rictor silencing ablated mTORC2 activity as evidenced by reduced phosphorylation of AKT S473 (Figure 3A). In addition, mTORC1 phosphorylation targets were unaffected by rictor silencing, with no phosphorylation alterations evident in AKT T308, S6 235/236 and S240/244, p70-S6K T389, and 4EBP1 T37/46, consistent with data from the RPPA analysis and prior studies.
Given the previously described role of mTORC2 in regulating bladder cancer cell migration and invasion, immunoblot validation was performed to confirm the effects of mTORC2-driven expression and phosphorylation changes of proteins known or proposed to be associated with these cellular processes.
In some cases, discrepancies between RPPA and representative immunoblot signals may be the result of using different antibodies with different epitope specificities and/or interassay variation. Loss of rictor affected the expression and/or phosphorylation of proteins involved in cell motility and focal adhesion, many of which have not been previously associated with mTORC2 activity (Figure 3B). In rictor-ablated cells, increased expression or phosphorylation of the cell adhesion factors FAK Y397, Y576/577 and Y925, total FAK, N-cadherin, β-catenin, p-ezrin T567, total talin-1, and snail was observed in both serum-starved and serum-stimulated conditions when compared with NTC control cells. By contrast, expression of liprin β-1 and TJP1/ZO-1 and phosphorylation of cofilin S3 were increased in rictor-silenced cells in the absence of serum, suggesting differential effects of mTORC2 activity under motile versus nonmotile conditions. Loss of mTORC2 activity resulted in reduction of talin-1 S425 and adducin S662 phosphorylation and protein expression of liprin α-1 and autocrine motility factor receptor in both serum-starved and stimulated conditions. These immunoblot findings are consistent with results obtained by RPPA analysis and represent validation of novel signaling effectors and downstream mediators of mTORC2. A subset of these proteins (N-cadherin, β-catenin, snail, and TJP1/ZO-1) have been implicated in epithelial-mesenchymal transition and with increased metastatic potential in a number of tumor systems, suggesting programmatic changes in cell function may also be regulated by mTORC2.
Phosphorylation of Cav-1 and Modification of Caveolae-Associated Receptor Tyrosine Kinases Are Regulated by mTORC2
A major motivation of this study was to identify novel targets of mTORC2 activity in the context of active bladder cancer cell migration. The unbiased global view of phosphoprotein expression changes due to mTORC2 silencing afforded by MS complemented by the high analytical sensitivity of RPPA to measure phosphorylation statuses of proteins relevant to cell motility and focal adhesion would allow for discovery of proteins using our bladder cancer cell migration model. To this end, MS analysis identified Cav-1 (Table 3) as a candidate downstream target of mTORC2 and is functionally associated with pathways relevant to cell migration, including cellular assembly and organization, cell morphology, and cellular movement. Cav-1 belongs to a family of integral membrane proteins and is an essential component of small (50 to 100 nm) flask-shaped invaginations of the plasma membrane called caveolae.
Aside from roles in caveolae formation and stability, Cav-1 acts as a scaffolding protein within caveolae and noncaveolar subdomains of lipid rafts to cluster and regulate numerous signaling molecules, such as G proteins, integrins, SRC family tyrosine kinases, PI3Ks, and RTKs.
On the basis of identification of an abundance of phosphorylated Cav-1 in rictor-expressing control cells using MS, up-regulation of Cav-1 activity by mTORC2 was anticipated. In particular, two well-characterized phosphorylation sites, Y14 and S80, were considered as possible residues that are potentially phosphorylated after serum stimulation of siNTC-transfected control cells. Residue Y14 phosphorylation is essential for Cav-1 association with SH2-domain–containing adaptor proteins, such as Grb7, at focal adhesion sites that promote integrin-mediated cell migration.
Cellular stress induces the tyrosine phosphorylation of caveolin-1 (Tyr(14)) via activation of p38 mitogen-activated protein kinase and c-Src kinase. Evidence for caveolae, the actin cytoskeleton, and focal adhesions as mechanical sensors of osmotic stress.
Phosphorylation of this residue is also required for caveolae-mediated endocytosis of cell surface signaling proteins.
To test whether Cav-1 Y14 is differentially phosphorylated in rictor-expressing versus rictor-silenced cells, immunoblot analysis of Y14 phosphorylation was performed in response to serum stimulation. A rapid and robust up-regulation of Cav-1 Y14 phosphorylation was observed after the addition of serum in rictor-expressing J82 cells as evidenced by immunoblot and corresponding densitometry analyses (Figure 4, A and B). By contrast, mTORC2-silenced cells showed constitutive phosphorylation of Y14 in both the presence and absence of serum, suggesting that rictor knockdown results in loss of dynamic regulation of Cav-1. Given the effect of mTORC2 silencing on Cav-1 phosphorylation, the expression of cavin-1, a necessary component for caveolae formation,
was subsequently investigated and its expression found to be markedly up-regulated on rictor silencing and serum stimulation.
To verify the potential interaction of Cav-1 with mTORC2, mTORC2-associated proteins were immunoprecipitated from T24 and J82 bladder cancer cell lysates using a rictor antibody–Sepharose bead conjugate. Intact mTORC2 was confirmed based on co-immunoprecipitation of enriched amounts of mTOR and the mTORC2-specific component, mSIN1 (Figure 4C). In both cell lines, co-immunoprecipitation of Cav-1 was observed with rictor antibody, confirming an interaction between the mTORC2 complex and Cav-1.
Growth factor RTKs coordinate numerous biological processes and are subjected to multiple levels of regulation, including attenuation of receptor autophosphorylation by protein tyrosine phosphatases, receptor sequestration and degradation, and up-regulation of inhibitory proteins that counteract downstream signaling effectors.
Both mTORC1 and mTORC2 signaling complexes coordinate negative feedback signals to some growth factor RTKs, such as insulin and insulin-like growth factor 1 receptors, PDGFR-β, and HER kinase receptors.
Consistent with these reports, reduced phosphorylation status of Y1135/36 on insulin-like growth factor 1 receptor, Y751 and Y716 on PDGFR-β, and Y1248 on HER2 was detected in rictor-silenced J82 cells by RPPA and immunoblot analysis (Figure 4D and Supplemental Figure S1). Because Cav-1 functions as a scaffolding protein to regulate signal transduction and rictor silencing eliminates dynamic Cav-1 phosphorylation, we hypothesized that mTORC2 feedback may regulate RTKs through modulation of Cav-1 activity. The effect of Cav-1 silencing was therefore evaluated on the phosphorylation status of EGFR and HER2. No apparent differences in EGFR Y1173, HER2 Y877, and HER2 Y1248 phosphorylation and EGFR and HER2 expression were observed in Cav-1–silenced J82 cells compared with rictor silencing, suggesting that loss of mTORC2 activity may directly affect regulation of EGFR and HER2 through Cav-1 (Figure 4E).
Lastly, the role of Cav-1 was evaluated on bladder cancer cell migration using a 24-hour modified scratch-wound migration assay. Cav-1 silencing had a small but significant (8%; P = 0.01, two-tailed t-test) inhibitory effect on J82 cell migration (Figure 4F) in contrast to a modest and significant decrease (35%; P = 0.0002, two-tailed t-test) in T24 cells (Figure 4G). These results are consistent with previous reports establishing a promigratory role of Cav-1 in bladder cancer cells.
mTORC2 Can Induce Cav-1 Redistribution and Modify Caveolae Abundance
Whereas tyrosine-14 phosphorylation of Cav-1 is linked to the regulation of cell signaling and caveolae-mediated endocytosis, serine-80 phosphorylation localizes Cav-1 to endoplasmic reticulum membranes.
Because commercial antibodies are not currently available for immunoblot detection of p-Cav-1 S80 that would allow directly determining whether this residue is activated by mTORC2, immunofluorescence microscopy was used to determine the localization pattern of Cav-1 in response to mTORC2 ablation and serum stimulation. In both control and rictor-silenced cells, a punctate staining pattern of Alexa Fluor 488–labeled Cav-1 was present throughout the cell (Figure 5A). However, in only control cells, Cav-1 was also localized to the edge and rear of cells in a manner consistent with polarization of Cav-1 during cell migration.
Together, these results suggest that the Y14 and not the S80 residue of Cav-1 may be modified by mTORC2 signaling because a reticular staining pattern consistent with endoplasmic reticulum localization was not observed in either experimental condition.
Because the data showed increased cavin-1 expression, modification of Y14 Cav-1 phosphorylation, and redistribution of Cav-1 with rictor silencing, we hypothesized that loss of mTORC2 activity might also alter the abundance of caveolae. Transmission electron microscopy was used to evaluate changes in caveolae abundance in mTORC2-ablated J82 cells. In both NTC and rictor siRNA–transfected cells, membrane-dense flask-shaped invaginations were observed on the plasma membrane and vesicular and fully-invaginated caveolae within the cell cytosol (Figure 5, B and C). The size of these vesicles (approximately 100 nm) was consistent with caveolae, as previously described.
However, compared with rictor-expressing control cells (Figures 5, D and F), an increased abundance of caveolae was observed in rictor-silenced cells (Figures 5, E and G), suggesting that mTORC2 activity affects caveolae formation through regulation of cavin-1 expression and/or Cav-1 Y14 phosphorylation.
Association of Caveolin-1 with Overall Survival in Human Bladder Cancer
Having identified Cav-1 as a potential target of mTORC2, which can promote bladder cancer cell invasion,
Cav-1 expression was analyzed in a cohort of patients with urothelial carcinoma to determine its association with outcomes. Previous studies have found that increased Cav-1 expression correlates with increasing bladder tumor grade and stage,
although there are limited reports of Cav-1 association with outcomes. Cav-1 expression was assessed using immunohistochemistry in 53 primary high-grade bladder cancers that were associated with metastasis in a subset of cases. Cav-1 was expressed in the cell cytoplasm and/or cell membrane in 49 of 53 cases (92.5%) (Figure 6A). The remaining four patients (7.5%) lacked detectable Cav-1 expression. Representative images with matching hematoxylin and eosin stains of Cav-1–negative and –positive specimens are shown in Figure 6, B and C.
It was next tested whether Cav-1 expression can predict patient outcomes. High Cav-1 expression was associated with a 5-year survival rate of 18.5% compared with a 46.1% 5-year survival rate associated with low Cav-1 expression. The publicly available transcriptomics bladder cancer data from the provisional TCGA RNA sequencing data set were next examined. High and low CAV1 expression (n = 101 each) was defined as the top and bottom quartiles, respectively, among the patient cohort (n = 408). A significant reduction in overall survival (P = 0.006) was observed in patients with high CAV1 expression compared with low CAV1 expression, with median survival rates of 2.0 and 7.2 years, respectively (Figure 6D).
mTOR plays a central role in cellular growth and proliferation, metabolism, and cell motility through two distinct multiprotein complexes, mTORC1 and mTORC2. Given its significance in maintaining cellular homeostasis, it is fitting that dysregulation of mTOR activity is associated with myriad malignancies, including bladder cancer. Although much is known about the regulation and functions of mTORC1, studies that define mTORC2 activity are more limited. Our laboratory has previously described a critical role for mTORC2 in promoting bladder cancer migration and invasion, with downstream effects on RhoA and Rac1 likely responsible in part for mediating this response.
To identify additional potential targets of mTORC2 signaling during bladder cancer motility, MS- and RPPA-based proteomic analysis was used concomitantly. Diverse categories of phosphoproteins pivotal in a wide range of cellular processes were identified, including those involved in cell motility, such as autocrine motility factor receptor; cytoskeletal arrangement, such as adducin and cofilin; and cell and focal adhesion, including FAK, SRC, N-cadherin, β-catenin, ezrin, talin-1, liprin α-1, liprin β-1, TJP1/ZO-1, and snail. Cav-1 was also identified as a novel downstream target of mTORC2, with mTORC2 regulation of Cav-1 phosphorylation and localization contributing to altered expression of known cancer-associated RTKs, such as EGFR and focal adhesion complexes, which were also identified through our screen.
Cav-1 has diverse functions that can positively or negatively regulate cell signaling pathways. For example, both EGFR and the insulin receptor can be negatively and positively activated by Cav-1 expression.
The differential regulatory roles of Cav-1 on cell surface receptors underscore the complexity of Cav-1 involvement in disease and biological processes as additionally evidenced by the variable effects on bladder cancer cell migration observed when Cav-1 is silenced. Cav-1 was first described as a direct substrate of SRC kinase that localizes to caveolae.
Subsequent studies also revealed the localization of Cav-1 to noncaveolar lipid rafts at or near focal adhesion sites where Y14-phosphorylated Cav-1 regulates focal adhesion complex dynamics to control directional migration.
The identification of an enrichment of phosphorylated Cav-1 in the rictor-expressing control bladder cancer cell model by MS suggested that mTORC2 could regulate Cav-1 activity through phosphorylation of the Y14 and/or S80 residues. Co-immunoprecipitation of Cav-1 was observed with mTORC2, and immunoblot analysis showed that Cav-1 phosphorylation on Y14 was robustly up-regulated after serum-stimulation in rictor-expressing control cells and could be ablated when rictor was silenced. However, this does not exclude the possibility of an indirect interaction between mTORC2 and Cav-1.
To further support the hypothesis that mTORC2 regulates Cav-1, the subcellular distribution pattern of Cav-1 was analyzed in cells with or without rictor silencing after serum stimulation to promote cell motility. In serum-stimulated control cells, Cav-1 was frequently polarized to the membrane and trailing edge of cells. By contrast, in rictor-silenced cells, Cav-1 staining appeared punctate within the cell cytosol with no apparent evidence of endoplasmic reticulum localization indicative of S80 phosphorylation.
Therefore, the observed difference in Cav-1 polarization is likely the result of Y14 rather than S80 phosphorylation. However, this assessment of mTORC2 regulation of S80 phosphorylation is limited because of the lack of a commercial antibody to this site. Given the observed localization of Cav-1 to the plasma membrane under motile conditions in control cells and the mTORC2-driven cavin-1 expression change identified with rictor silencing, whether mTORC2 could regulate formation or abundance of caveolae in this model system was tested. Rictor silencing resulted in increased caveolae formation seen by transmission electron microscopy. On the basis of these data, mTORC2 appears to be an upstream mediator of Cav-1 Y14 phosphorylation and may be responsible for regulating caveolae formation and abundance. Because caveolae can cluster and enhance the signaling of numerous RTKs within the cell, mTORC2 may also have a role in mediating bladder cancer caveolar-associated RTK protein expression and/or stability, including EGFR and HER2.
Cell migration is a complex process that requires spatial and temporal regulation of numerous signaling components at integrin-mediated adhesion sites. These adhesions dynamically form and turn over at the leading edge of a cell to generate traction forces while they disassemble at the rear to produce rear-end retraction and detachment for cell body translocation. This highly regulated process is understood to be primarily mediated by FAK and SRC tyrosine kinases that signal to numerous downstream effectors, including Cav-1 and the Rho family GTPases Rac1 and RhoA, the latter of which was previously found to be regulated by mTORC2 in bladder cancer cells.
Under serum-free conditions in which cells do not migrate, Y14 of Cav-1 was not phosphorylated in control cells but was markedly phosphorylated on serum stimulation. Because mTORC2-related expression changes in focal adhesion complex proteins were seen, whether mTORC2 could spatially regulate Cav-1 localization to cell adhesions to prevent focal adhesion turnover, which thereby maintains a resting (nonmotile) state, was examined. This theory is supported by findings reported by Beardsley et al
that suggest that loss of Cav-1 polarity in stationary cells and targeted silencing of Cav-1 impede directional movement. The authors also found that immunofluorescence staining for Cav-1 in stationary cells appeared punctate throughout the cell, a similar result observed in rictor-silenced, serum-stimulated bladder cancer cells.
Prior studies have failed to demonstrate an association between Cav-1 expression and overall survival for patients with bladder cancer,
although the patient numbers were limited. Public gene expression data from the provisional TCGA bladder cancer study that includes a large patient data set were therefore examined. A highly significant association was found between high CAV1 expression and reduced median and overall survival.
In summary, two parallel phosphoproteomics approaches were used to identify protein and phosphoprotein targets of mTORC2, including regulators of cell motility, cell morphology, cellular assembly and organization, and several other functions that have not been previously linked with mTORC2 activity. mTORC2 was found to dynamically regulate Cav-1 Y14 phosphorylation, alter Cav-1 localization, and mediate caveolae formation. Although known to regulate some components of focal adhesion complexes, additional cell adhesion and cell motility proteins were also identified as potential targets of mTORC2, and these proteins may be regulated in conjunction with Cav-1. Specifically, the effects on the activities of FAK and SRC on rictor silencing would suggest a functional association among these kinases and mTORC2. Indeed, a link between mTORC2 activation and Fyn, a SRC family kinase, in cooperation with FAK was previously described during focal adhesion signaling that defines marrow-derived mesenchymal stem cell fate and cytoskeletal structure emanating from mechanical stimuli, such as cell stress or strain.
However, it is unclear whether the same association exists during growth factor–induced activation of mTORC2. Future studies that dissect the role of mTORC2 in the regulation of the caveolar microenvironment are important to determine effects on RTK signaling in the context of cell motility and stasis and the regulation of additional proteins associated with focal adhesions.
We thank Dr. Henry C. Powell (University of California, San Diego) for his assistance and expert advice for the electron microscopy experiments, Drs. David M. Sabatini (Massachusetts Institute of Technology) and Dudley W. Lamming (University of Wisconsin) for technical protocols for rictor co-immunoprecipitation, and Dr. Kun-Liang Guan (University of California, San Diego) for critically reviewing the manuscript.
A.M.H., S.G., S.M., E.F.P., and D.E.H. designed research; A.M.H., S.G., M.Z.L., K.N., J.M., W.Z., A.H., and J.W. performed research; B.C., K.B., and S.R. analyzed data; A.M.H, K.B., S.R, S.M., E.F.P., and D.E.H. reviewed and/or wrote the article; and D.E.H. is the guarantor of this work and, as such, had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Cellular stress induces the tyrosine phosphorylation of caveolin-1 (Tyr(14)) via activation of p38 mitogen-activated protein kinase and c-Src kinase. Evidence for caveolae, the actin cytoskeleton, and focal adhesions as mechanical sensors of osmotic stress.
Supported by the Department of Pathology, University of California, San Diego .
Disclosures: E.F.P. has US government– and George Mason University–assigned patents concerning the reverse-phase protein array technology and receives licensing and royalty distribution from these patents. E.F.P. is a consultant to ADVX Investors Group, LLC, which has licenses in reverse-phase protein array–based immunoprecipitation from George Mason University and US government–owned patents.