Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Nov 15.
Published in final edited form as: Clin Cancer Res. 2016 May 24;22(22):5605–5616. doi: 10.1158/1078-0432.CCR-15-1673

Tyrosine Kinase Signaling in Clear Cell and Papillary Renal Cell Carcinoma Revealed by Mass Spectrometry-Based Phosphotyrosine Proteomics

Scott M Haake 1, Jiannong Li 2, Yun Bai 2, Fumi Kinose 2, Bin Fang 3, Eric Welsh 4, Roy Zent 5, Jasreman Dhillon 6, Julio Pow-Sang 7, Yian Ann Chen 4, John Koomen 8,9, W Kimryn Rathmell 1, Mayer Fishman 7, Eric B Haura 2
PMCID: PMC5122466  NIHMSID: NIHMS791673  PMID: 27220961

Abstract

Purpose

Targeted therapies in renal cell carcinoma (RCC) are limited by acquired resistance. Novel therapeutic targets are needed to combat resistance and, ideally, target the unique biology of RCC subtypes.

Experimental Design

Tyrosine kinases provide critical oncogenic signaling and their inhibition has significantly impacted cancer care. In order to describe a landscape of tyrosine kinase activity in RCC that could inform novel therapeutic strategies, we performed a mass spectrometry-based system-wide survey of tyrosine phosphorylation in 10 RCC cell lines as well as 15 clear cell and 15 papillary RCC human tumors. To prioritize identified tyrosine kinases for further analysis, a 63 tyrosine kinase inhibitor (TKI) drug screen was performed.

Results

Among the cell lines, 28 unique tyrosine phosphosites were identified across 19 kinases and phosphatases including EGFR, MET, JAK2, and FAK in nearly all samples. Multiple FAK TKIs decreased cell viability by at least 50% and inhibited RCC cell line adhesion, invasion, and proliferation. Among the tumors, 49 unique tyrosine phosphosites were identified across 44 kinases and phosphatases. FAK pY576/7 was found in all tumors and many cell lines, while DDR1 pY792/6 was preferentially enriched in the papillary RCC tumors. Both tyrosine kinases are capable of transmitting signals from the extracellular matrix and emerged as novel RCC therapeutic targets.

Conclusions

Tyrosine kinase profiling informs novel therapeutic strategies in RCC and highlights the unique biology amongst kidney cancer subtypes.

Keywords: renal cell carcinoma, tyrosine kinase, mass spectrometry, FAK, DDR

Introduction

Renal cell carcinoma (RCC) is among the most common cancers diagnosed in the United States. Metastatic RCC is relatively insensitive to traditional therapies like chemotherapy and radiation and is generally incurable (1). However, the emergence of active agents targeting vascular epithelial growth factor receptor (VEGFR) and mammalian target of rapamycin (mTOR) heralded a new era in the treatment of RCC. However, patient outcomes remain poor despite these contemporary therapies (2).

There are multiple reasons that could explain the poor outcomes associated with the targeted treatment era. First, the responses to these targeted therapies are typically transient. The emergence of resistance is nearly inevitable resulting in poor 5-year survival (2). Second, the VEGFR-targeted therapies that have come to dominate the targeted treatment era in RCC were tailored for clear cell RCC, the most common histological subtype. These tumors are characterized by the near ubiquitous loss of function of the Von Hippel Lindau tumor suppressor, VHL (3). The result is inappropriate stabilization of HIF and a maladaptive hypoxic and angiogenic response, including markedly high levels of VEGF production (4). By targeting VEGF receptors, this class of tyrosine kinase inhibitors (TKIs) targets the core biology of clear cell RCC tumors. However, non-clear cell RCC tumors such as papillary RCC have functional VHL and thus classically do not exhibit the same angiogenic response. Predictably, their outcomes are inferior when treated with VEGFR therapies (5). Novel therapeutic targets could guide new drug development with a goal of delaying, treating, or preventing disease resistance. Furthermore, studying non-clear cell RCC offers the opportunity to tailor our treatment to their unique biology.

Tyrosine kinases provide signal transduction that is critical for the growth and survival of several cancers (6). While several tyrosine kinases are overexpressed in RCC, including EGFR (7) and MET (8), few approved therapies target these epithelial drivers in RCC (9–11). In order to describe a broad landscape of tyrosine kinase activity in RCC that could inform novel therapeutic strategies, we performed a mass spectrometry (MS)-based system-wide survey of tyrosine phosphorylation in RCC cell lines as well as clear cell and papillary RCC tumors (Fig. 1). In order to prioritize emerging tyrosine kinase targets based on functional data, we concurrently performed a large TKI screen across the ten RCC cell lines. One potential target to emerge from this approach was focal adhesion kinase (FAK). FAK was phosphorylated at an activating site in cell lines and both clear cell and papillary tumors. FAK TKIs were active in the screen and were shown to inhibit cell adhesion, proliferation, and invasion. Additionally, the receptor tyrosine kinase DDR1, for which the only known ligand is collagen, was highly phosphorylated at multiple activating sites in the papillary RCC tumors relative to the clear cell RCC tumors. This receptor tyrosine kinase (RTK) may represent a novel mediator of stromal signaling and a therapeutic target in papillary RCC. The combination of systems level MS techniques to study tyrosine phosphorylations with integrated functional studies identified novel therapeutic targets that warrant further investigation.

Figure 1.

Figure 1

Open in a new tab

Phosphotyrosine proteomics and functional interrogation of renal cell carcinoma (RCC). LC-MS/MS = liquid chromatography-tandem mass spectrometry.

Materials and Methods

Full descriptions of all materials and methods can be found under Supplementary Methods and Materials.

Cell lines

Cell lines A704 (HTB-45), A498 (HTB-44), ACHN (CRL-1611), Caki-1 (HTB-46), and Caki-2 (HTB-47) were purchased from American Type Culture Collection (ATCC). 786-O (CRL-1932), RXF393, UO31, SN12C, and TK10 (CRL-2396) were a gift from Dr. Javier Torres-Roca (Moffitt Cancer Center, Tampa, FL). All cell lines were cultured in RPMI1640 with 10% fetal bovine serum, maintained in a central repository at MCC, routinely tested for mycoplasma contamination, and have been authenticated with short-tandem repeat (STR) analysis (ATCC).

Human tumor tissues

Tumor tissues were collected as part of the Total Cancer Care protocol (12) and approved by the University of South Florida Institutional Review Board (Tampa, FL). Patients gave informed consent before enrollment in the Total Cancer Care protocol. Tumor tissues were snap frozen following nephrectomy. All tissues contained more than 90% tumor cells when examined by light microscopy. Centralized pathology review was performed at time of TCC enrollment. So that phosphotyrosine (pY) peptide patterns could be compared between tumor types, clear cell and papillary RCC tumors were balanced for patient characteristics including age, gender, race, tumor size, stage, and nuclear grade (Supplementary Table 1). Type 1, type 2, and undefined papillary RCC were included in the analysis.

Phosphopeptide immunoprecipitation, analysis, and data processing

Immunoprecipitation and purification of pY peptides was performed using PhosphoScan pTyr100 (Cell Signaling Technology) according to the manufacturer’s recommendations and as described previously (13). Briefly, whole cell extracts were prepared from either 1) 2 × 108 cells of each cell line or 2) 100 mg of tumor tissue using denaturing lysis buffer containing 9 M urea and 20 mM HEPES (pH 8.0) supplemented with phosphatase inhibitors (1 mM sodium orthovanadate, 2.5 mM sodium pyrophosphate, 1 mM β-glycerophosphate) followed by sonication on ice. Extracted proteins (50 mg for each cell line) were then reduced with 4.5 mM DTT and alkylated with 10 mM iodoacetamide. Trypsin digestion was carried out at room temperature overnight and resulting tryptic peptides were then acidified with 1% trifluoroacetic acid and desalted with C18 Sep-Pak cartridges (Waters) according to the manufacturer’s procedure. Peptides were lyophilized and then dissolved in immuno-affinity purification (IAP) buffer containing 50 mM MOPS/NaOH (pH 7.2), 10 mM sodium phosphate, and 50 mM sodium chloride. The pY peptides were immunoprecipitated with immobilized pTyr100 antibody overnight at 4°C followed by three washes with IAP buffer and two washes with H2O. The pY peptides were eluted from beads twice with 0.15% trifluoroacetic acid and the volume was reduced to a final concentration of 20 μL via vacuum centrifugation. Before MS tumor analysis, 100 fmol of a multi-peptide standard (Pierce 88320) was spiked into each sample. Samples were analyzed by nano-liquid chromatography/tandem mass spectrometry (LC-MS/MS). Two technical replicates were performed for each sample. Aliquots of unused tumor pY immunoprecipitate was analyzed in a targeted quantification experiment (liquid chromatography-multiple reaction monitoring or LC-MRM) to validate the label free relative quantification data for select phosphopeptides. Further LC-MS/MS and LC-MRM details are described in Supplementary Materials and Methods.

Generation of protein-protein interaction network for pY proteins

The pY proteins corresponding to the clear cell RCC cell lines and the human tumors were input into Cytoscape (version 2.8.3; http://www.cytoscape.org; (14)). Protein-protein interactions (PPI) between nodes were imported using the PSICQUIC plug-in to search multiple databases (15). Network visualization was performed with Cytoscape.

Drug screening and in vitro assays

Viability assays were performed in black-wall 384-well microtiter plates for the 10 RCC cell lines. Cells were seeded at a density of 1,000 cells/well. Drugs or DMSO were added after 24 hours and cells were incubated for another 72 hours. 63 tyrosine kinase inhibitors (Fig. 4a, Supplementary Fig. 1–2) were screened for each cell line at 0.5 μM and 2.5 μM (each in duplicate and averaged), a concentration that approximates or is slightly greater than predicted serum levels in humans and thus assures minimum false negatives. Viability was evaluated using the CellTiter-Glo assay (Promega) with addition of reagent after 72 hour drug incubation and luminescence was read on a SpectraMax M5 plate reader (Molecular Devices). Matrigel adhesion and invasion assays (16) as well as 3H-thymidine proliferation assays (17) were performed as described previously.

Figure 4.

Figure 4

Open in a new tab

Tyrosine kinase inhibitors provide functional correlation. a) A total of 63 tyrosine kinase inhibitors (TKIs) were screened. Data for the most frequently observed tyrosine kinases are shown. The FAK TKIs CEP-37440 (CEP) and defactinib (DEF) were observed to decrease b) cell adhesion and c) proliferation in a dose-dependent manner for multiple RCC cell lines. d) In addition, the FAK TKIs decreased RCC cell line proliferation ([drug] = 1.0 µM). DMSO = dimethyl sulfoxide; *p<0.05; **p<0.01

Western blotting

Western blotting was performed as described in our previous studies (13, 18). Primary antibodies used for our study were purchased from Cell Signaling Technology (EGFR pY1068, EGFR pY1197, MET pY1234/5, total MET), Santa Cruz (total EGFR, total FAK), BD Sciences (FAK pY397) and Sigma-Aldrich (β-actin).

Statistical Methods & Data Analysis

Tumor samples were randomized prior to MS analysis to minimize potential batch effects. Internal controls within each sample (Thermo peptide standard) and external controls (MS analysis of control cell line before analysis of RCC tumors, after every 20 MS runs, and at conclusion of MS analysis of RCC tumors) were included to monitor potential batch effects (Supplementary Fig. 3 and 4). In order to maximize protein coverage, Mascot, Sequest, and Andromeda search engines were performed to identify phosphorylated proteins (see supplementary methods for details). To facilitate and streamline comparisons among samples, only proteins that were identified as phosphorylated in at least 5 samples (cell line or tumor) were considered for further analysis (Fig. 2 and 5). The data regarding individual phosphosites came from only the Andromeda search results (Fig. 3 and 6). Label free quantitative data was calculated using MaxQuant (version 1.2.2.5) using the Andromeda search results (Fig. 6) (19). Peptide ion signal intensities were log2 transformed to allow for the use of parametric statistical tests. A two-sample, two-tailed t-test was used to contrast phosphopeptide intensity in clear cell versus papillary tumors (Fig. 6a), phosphopeptide intensity in tumor samples calculated from extracted ion chromatograms (Fig. 6b), and output from multiple reaction monitoring experiments (Fig. 6c). To account for multiple hypothesis testing, a false discovery rate was estimated for phosphopeptide intensity in tumor samples and reported as a q-value with a cut-off of 0.05 (Fig. 6a). Protein and phosphosite functional annotation was performed with reference to databases at www.phosphosite.org and www.uniprot.org.

Figure 2.

Figure 2

Open in a new tab

Identification of tyrosine phosphorylated proteins in RCC cell lines and tumors. a) A total of 198 tyrosine phosphorylated proteins from several different categories were identified among the renal cell carcinoma cell lines and tumors (limited to those phosphorylated proteins identified in at least 5 samples). b) The rate of phosphorylation was compared among cell lines and tumors for the proteins most frequently identified in our analysis. The proteins more commonly phosphorylated in c) tumors, d) tumors and cell lines, and e) cell lines are shown. Node size is proportional to number of tumors in which a tyrosine phosphorylated protein was identified (label = HGNC gene name). Color is proportional to number of cell lines in which a tyrosine phosphorylated protein was identified (dark blue = more cell lines). Edges represent protein-protein interactions curated from the literature.

Figure 5.

Figure 5

Open in a new tab

Identification of phosphotyrosine peptides in clear cell and papillary renal cell carcinoma (RCC) tumors. The rate of phosphorylation was compared among clear cell and papillary RCC tumors for the proteins most frequently identified in our analysis. The proteins more commonly phosphorylated in a) clear cell RCC tumors, b) clear cell and papillary tumors, and c) papillary RCC tumors are shown. Node size is proportional to number of papillary tumors in which a tyrosine phosphorylated protein was identified (label = HGNC gene name). Color is proportional to number of clear cell RCC tumors in which a tyrosine phosphorylated protein was identified (dark blue = more clear cell RCC tumors). Edges represent protein-protein interactions curated from the literature.

Figure 3.

Figure 3

Open in a new tab

Tyrosine kinase and phosphatase phosphotyrosine peptide detection across RCC cell lines. a) Individual phosphopeptides were identified across clear cell renal cell carcinoma (RCC) cell lines (grey = detected). b) Western blots for select phosphorylated and total tyrosine kinases across cell lines along with ratio of phosphorylated-to-total protein band intensity.

Figure 6.

Figure 6

Open in a new tab

Kinase and phosphatase phosphotyrosine peptide intensity across clear cell and papillary RCC tumors. a) Individual phosphotyrosine peptides were identified across clear cell and papillary renal cell carcinoma (RCC) tumors. Relative phosphopeptide quantitation is represented by color (white = phosphopeptide not detected, blue = lower phosphopeptide intensity, red = greater phosphopeptide intensity). Grey cells = q<0.05. b) Phosphotyrosine peptide quantitation was validated by analysis of extracted ion chromatograms (EICs). EICs were manually checked for accurate peak integration. c) The relative phosphotyrosine peptide intensity of EGFR and DDR1 were validated in a separate, quantitative, and targeted proteomics experiment (LC-MRM).

When contrasting the patient characteristics corresponding to human tumors (Supplementary Table 1), all continuous variables were described with the median and range values. Analyses of categorical data utilized a two-tail Fisher’s exact test. Difference of medians was evaluated with Mann-Whitney U test.

Results

RCC cell lines and tumors display FAK phosphorylation but different patterns of receptor tyrosine kinase phosphorylation

In order to study RCC in an unbiased and systems level manner, we performed immunoprecipitation followed by LC-MS/MS to catalog pY peptides from 10 RCC cell lines and 30 tumors (Fig. 1). Overall, 198 proteins were identified and their function annotated (Fig. 2a, Supplementary Table 2). The proteins corresponded to several different classes including receptor tyrosine kinases (EGFR, MET), non-receptor tyrosine kinases (JAK2, FAK or FAK1 or PTK2), serine/threonine kinases (CDK1, MK01 or MAPK1), phosphatases (SHIP2 or INPPL1), cell adhesion proteins (ITB1 or ITGB1), cytoskeletal proteins (ABLM1, MYO1E), GTP-associated proteins (ARHG5 or ARHGEF5), and others. The cell lines and tumors had similarities with 102 shared tyrosine phosphorylated proteins (Fig. 2b). However, marked distinctions were also evident as 53 pY proteins were unique to cell lines and 22 unique to tumors. To further explore these patterns, the ~50 most commonly identified phosphorylated proteins were analyzed further. If the rate of phosphorylation for a protein was at least 50% higher for cell lines relative to tumors (or vice versa), it was considered “more common” in that subset (Fig. 2c and 2e). Proteins that were commonly phosphorylated but did not meet this cutoff were considered “common in tumors and cell lines” (Fig. 2d). The MAP kinase pathway was tyrosine phosphorylated in both tumors and cell lines as was PTK2 protein FAK1 or FAK), supporting their universal importance in RCC biology. However, multiple members of the PI3K pathway as well as EGFR were more common in the cell lines suggesting distinct signal transduction relative to tumors. By contrast, the collagen receptor DDR1 was phosphorylated in tumor samples only. Thus, while MAPK and FAK were phosphorylated in both cell lines and tumors, distinct patterns of activation emerged among some RTKs.

Several kinases, including EGFR and MET, are phosphorylated in RCC cell lines

Rather than only examine whether a protein had any pY (Fig. 2), we also evaluated specific phosphosites in cell lines (Supplementary Table 3). Furthermore, our focus was narrowed to kinases and phosphatases given their key roles in regulating cancer signaling (Fig. 3a). EGFR pY1110, pY1172 and pY1197 was observed in multiple cell lines, consistent with receptor activation (20–22). Similarly, the presence of the MET activation site pY1234/5 in all cell lines implies the presence of ubiquitous MET activation. FAK pY576/7 was detected in nearly all the cell lines (Fig. 3a). This FAK phosphosite within the kinase domain of the enzyme is required for maximal enzymatic activity and thus suggests FAK activation (23). Thus, the identification of specific phosphosites was able to provide insight into the functional relevance of pY events.

Western blotting was performed to further explore the pY patterns of the cell lines for EGFR, MET, and FAK (PTK2) (Fig. 3b). EGFR is known to be activated and phosphorylated in RCC, possibly related to VHL inactivation (24). EGFR pY1197 was weakly positive by western blot in several cell lines despite being identified in all cell lines via MS analysis (Fig. 3a and 3b). To further validate EGFR activation, western blot for EGFR pY1068 was performed. EGFR pY1068 is well known to correlate with EGFR activation (20, 21) and has an excellent performing antibody though, given the large size and double phosphorylation status of the corresponding tryptic peptide can be difficult to detect in MS analysis. The positivity of EGFR pY1068 on western blot supports the MS data that suggests EGFR activation. Thus, the weak EGFR pY1197 signal on western blot yet strong signal via MS may simply be due to increased sensitivity for MS techniques. MET is also known to be activated and phosphorylated in RCC and interest is increasing given the approval of cabozantinib, a TKI whose targets include MET (8, 11, 25). Western blot for MET pY1234/5 confirms that this phosphorylation event is common among the cell lines. Again, its detection in some cell lines via MS but not western blot suggests increased sensitivity of MS techniques. FAK (PTK2) was evaluated further as it was seen to be frequently phosphorylated in cell lines and tumors (Fig. 2d) and its function in RCC had not been well characterized in the literature. The performance of commercially available FAK pY576/7 antibodies was suboptimal and lacked reproducibility in our hands. However, the MS data (Fig. 2d and 3a) showed near-universal FAK activation in RCC and we verified its activation by testing phosphorylation of the canonical autophosphorylation and activation site, FAK pY397 (23). Overall, western blots validate the MS data that EGFR, MET, and FAK are phosphorylated at activating sites in RCC cell lines.

TKI screen highlights activity of FAK tyrosine kinase inhibitors

A functional screen utilizing 63 tyrosine kinase inhibitors (TKIs) was employed to assist in nominating functionally relevant signaling networks for further analysis (Supplementary Fig. 1–2). We focused our analysis on TKIs targeting tyrosine kinases observed to be phosphorylated in most cell lines, specifically EGFR, FAK, JAK2, and MET (Fig. 4a). Among the most active class of TKIs were those targeting FAK. Specifically, CEP-37440 at 2.5 µM reduced cell viability to <50% in 6 of the 10 cell lines tested. Given that FAK was phosphorylated in both tumors and cell lines via MS and western blot (Fig. 2d, 3a, and 3b) and the FAK TKIs were active in our drug screen (Fig. 4a), we decided to further investigate the impact of FAK TKI treatment on RCC cell lines. For this analysis, two FAK TKIs with activity across multiple cell lines (CEP-37440 and defactinib) and two cell lines sensitive to multiple FAK TKIs (ACHN and UO31) were chosen. Consistent with the canonical role of FAK in mediating cellular adhesion, the FAK TKIs demonstrated dose-dependent inhibition of RCC cellular adhesion to Matrigel (Fig. 4b). Similarly, the FAK TKIs potently inhibited cellular proliferation in a dose-dependent manner (Fig. 4c) and produced a dramatic effect on the ability of the RCC cells to invade through a Matrigel membrane (Fig. 4d). This accumulation of evidence regarding likely FAK activation in RCC cell lines and tumors and the activity of FAK TKIs suggests that further investigation is warranted to explore FAK TKIs as novel RCC therapies.

FAK is phosphorylated in both tumor types while other subtype-specific pathways emerge

Next, we focused on the 145 pY proteins identified in the tumors (Fig. 2b, Supplementary Table 2). Corresponding proteins were diverse and included receptor tyrosine kinases (EGFR, DDR1), non-receptor tyrosine kinases (FAK or PTK2SRC), serine/threonine kinases (PRPF4B), phosphatases (TENC1 or TNS2), cell adhesion proteins (PKP4), cytoskeletal proteins (PALLD, PAXI or PXN), and others. Focusing on the ~50 most frequently identified pY proteins, the rate of tyrosine phosphorylation for each protein was compared among clear cell and papillary RCC tumors. If the rate of tyrosine phosphorylation was ≥50% higher for either clear cell or papillary RCC, it was considered “more common” in that subset (Fig. 5a and 5c). Common proteins that did not meet this cutoff were considered “common in clear cell and papillary RCC tumors” (Fig. 5b). EGFR was commonly phosphorylated in clear cell RCC tumors, as predicted since loss of VHL function can lead to EGFR activation (Fig. 5a) (24). Similar to the cell lines, MAPK phosphorylation was common in both tumor subtypes (Fig. 5b). FAK (PTK2) was phosphorylated across the RCC spectrum, including cell lines (Fig. 2d) and both tumor types (Fig. 5b). Meanwhile, DDR1 was previously shown to be enriched in tumors (Fig. 2c) and now appears to be preferentially phosphorylated in the papillary RCC tumors (Fig. 5c).

While the above trends are interesting, the comparisons were only semi-quantitative and not amenable to statistical testing. In addition, the above trends are not phosphosite-specific. Thus, we next evaluated specific tyrosine phosphosites in the tumors (Supplementary Table 4). Unlike the cell lines, where the MS experiment was designed as a survey of tyrosine phosphosites, the tumor experiments were designed to be quantitative, thus facilitating direct comparison of pY intensities among clear cell and papillary tumors.

When filtered for peptides corresponding to kinases and phosphatases that were identified in at least 5 tumors, 40 unique pY peptides corresponding to 37 proteins were identified, of which seven pY peptides demonstrated significant differential intensity between clear cell and papillary RCC tumors (Fig. 6a). The extracted ion chromatograms (EIC) for these pY peptides with differential intensity were manually verified for accurate peak integration and the relative quantification was successfully validated for six of the pY peptides (Fig. 6b). Based on this EIC quantitation, EGFR pY1197 (activating site, fold change-7.05, p<0.001), MK01 (ERK2) pY187 (activating site, fold-2.94, p=0.003) (26), MK03 (ERK1) pY204 (activating site, fold-1.95, p=0.036), and TENC1 pY483 (phosphosite function unknown, fold-2.11, p=0.046) were consistently higher in clear cell RCC. Conversely, DDR1 pY792/6 (activating site, fold-11.48, p=0.012) and PRP4B pY849 (phosphosite function unknown, fold-2.66, p=0.004) were higher in papillary RCC. A separate targeted proteomics experiment using liquid chromatography-multiple reaction monitoring (LC-MRM) was conducted utilizing unused aliquots of the phosphopeptide immunoprecipitates in order to further validate the differential intensity of pY peptides correlating to EGFR and DDR1 from the LC-MS/MS experiment. The MRM results validated the increase in DDR1 phosphorylation in papillary tumors for pY513 (activating site, fold-4.01, p=0.006) and pY796 (activating site, fold-9.53, p=0.002) as well as the increase in EGFR pY1197 in clear cell RCC tumors (activating site, fold-2.43, p=0.013) (Fig. 6c). Furthermore, immunohistochemistry was performed on external samples of human FFPE clear cell and papillary RCC tumors in order to further characterize DDR1 and FAK phosphorylation amongst the tumor subtypes (Supplementary Figure 5). These experiments demonstrate robust signal correlating with DDR1 pY513 in the papillary RCC relative to clear cell RCC, providing further validation of our mass spectrometry results. Meanwhile, FAK pY397 continues to be seen in both tumor types.

This quantitative, phosphosite-specific tumor data reveals some novel insights. First, while the MAPK pathway and FAK (PTK2) was observed as phosphorylated in both tumor types in prior analyses (Fig. 5b), a quantitative analysis reveals that there is relatively higher phosphorylation of multiple MAPK members in the clear cell RCC tumors (Fig. 6a-b). One possible explanation may be that the increased EGFR phosphorylation/activation observed for the clear cell RCC tumors is contributing to this higher MAPK activation. Meanwhile, FAK (PTK2) demonstrated similar levels of pY576/7 across both RCC tumor subtypes. Finally, DDR1 consistently emerges as preferentially phosphorylated in papillary tumors. Thus, some pathways continue to emerge as universally activated in RCC (e.g. FAK) while others appear to be subtype specific (e.g. EGFR, DDR1).

Discussion

We utilized the immunoprecipitation and purification of pY peptides coupled with MS-based peptide identification and quantitation of RCC cell lines and tumors as a method to 1) seek out activated signaling pathways as therapeutic targets and 2) further resolve the divergent biology of these kidney cancer subtypes. As phosphorylations often regulate the function of target proteins and kinases, their study offers the potential to study cancer systems with some degree of functional annotation. Furthermore, MS allows for a relatively unbiased analysis that is not dependent upon specific phosphoantibodies. To provide further resolution of the functional relevance of identified pathways, we coupled this pY peptide survey with a functional screen utilizing TKIs. The resultant data are distinct and complementary to the genomic data that has been featured so prominently in the molecular study of these tumors.

This analysis of tyrosine phosphorylation of RCC cell lines and tumors offers a unique opportunity to contrast their respective biology. Interestingly, ~25% of all tyrosine phosphorylated proteins were unique to the cell lines (i.e. not detected in tumors) while several proteins were exclusive to the tumors (Fig. 2b). Thus, while cell lines represent valuable resources given their ease of manipulation, human tumors are essential to guide investigation. For example, the collagen RTK DDR1 emerged early in our analysis as commonly phosphorylated in tumors yet undetected in the cell lines (Fig. 2c). Multiple reasons exist that could explain this discrepancy. First, as the cell lines were grown in plastic without collagen, DDR1 may lack sufficient ligand to activate/phosphorylate to the level that we can detect. However, the human tumors with their associated extracellular matrix display frequent DDR1 phosphorylation. Second, DDR1 was preferentially phosphorylated in the papillary RCC tumors. Since cell lines for rarer, non-clear cell RCC subtypes are not widely available, the cell lines likely do not replicate the biological diversity found within the RCC spectrum. As DDR1 is emerging as an important and druggable driver of tumorigenesis in cancer (27–29), its discovery in papillary RCC is of great interest and would have potentially been missed without inclusion of the human studies. New tumor models that better recapitulate the tumor-associated stroma may be needed to continue preclinical studies of DDR1 in papillary RCC.

While DDR1 may only be phosphorylated in papillary RCC tumors, FAK was phosphorylated in essentially all RCC samples. FAK is an important component of focal adhesions, which are multi-protein complexes that link the cytoskeleton to components of the extracellular matrix via integrins (30). Focal adhesions regulate cellular attachment and are often involved in the transition to a motile phenotype that promotes cancer invasion and metastasis (30). As discussed above, cell lines have important limitations and do not always accurately recapitulate the biology of the disease they were developed to model. However, in the current MS data set, FAK was phosphorylated at pY576/7 in both the cell lines and tumors. Thus, studying cell lines to determine the functional impact of FAK inhibition in RCC may be an example of their utility. While further investigations are necessary to fully understand the function of FAK in RCC, the current studies provide important insights. First, multiple FAK TKIs were capable of causing significantly decreased cell viability in multiple RCC cell lines during the drug screen (Fig. 4a). The decrease in cell viability caused by the FAK TKIs is at least partially due to decreased cell proliferation (Fig. 4c). The mechanism of decreased cell viability and proliferation remains to be completely resolved, but given the decrease in cell adhesion due to FAK TKI treatment (Fig. 4b), anoikis is one possibility (31). Furthermore, the ability of the FAK TKIs to potently inhibit migration of RCC cells through Matrigel, a model of basement membrane, is highly relevant given the need for cancer cells to invade during tumorigenesis and eventually metastasis. Together, these data support further study of FAK inhibition as a novel therapy in RCC.

MET is classically associated with type 1 papillary RCC, where it can be activated through a variety of mechanisms including gain-of-function mutations (32). In addition, it has also been shown to be activated in clear cell RCC, particularly in advanced (8) and VEGFR-resistant disease (33). The interest in MET has intensified recently given the published activity of cabozantinib, a TKI whose targets include VEGFR and MET, in clear cell RCC patients who previously progressed on VEGFR TKI (11). However, while MET pY1234/5 was identified in most of the cell lines, it was not identified in any of the clear cell RCC human tumors and only four of the papillary RCC tumors (Supplementary Table 2). A few possible explanations exist for why MET was not detected as phosphorylated more frequently in the tumor data. First, as few tumors in our analysis were from metastatic patients, a cohort with more advanced disease may have higher levels of MET phosphorylation. In addition, given the data suggesting MET may be especially relevant in the context of VEGFR TKI resistance, it may be less activated in our treatment-naïve samples (33). Also interesting is that fact that, of the 7 MET TKIs tested in our drug screen (which included cabozantinib), only tivantinib decreased cell viability to <50% (Fig. 4a). It should be noted that tivantinib was recently shown to have significant off-target effects and thus may not be acting through MET (34). This data suggests that MET TKIs may not be directly cytotoxic to the tumor cell. Indeed, the clinical activity observed for cabozantinib may stem from endothelial cell MET inhibition (i.e. anti-angiogenesis properties) or inhibition of a different cabozantinib target. Further investigation is needed to determine the role of MET in different RCC subtypes.

One of the goals of the current study was a comparison of the tyrosine phosphorylation across the proteomes of clear cell and papillary RCC. Accordingly, we took efforts to include similar tumor sizes and to roughly match for tumor stage and grade (Supplementary Table 1). Papillary tumors tend to be small with 75% of such tumors representing T1 disease (35). Therefore, matching the size of papillary and clear cell RCC tumors contributed to the portfolio of primary tumors studied. This study, while limited in size and representation of advanced disease, provides a discovery cohort for uncovering comparative differences in the proteome of important histologic subtypes in RCC. Future studies to validate the extent of tumors demonstrating these signals, and particularly to examine the phosphoproteome in metastatic tumors, would be valuable, and may reveal additional activated signaling pathways.

The proteome represents an important focus of biology whose study can complement our knowledge derived from the genome. For example, a recent proteogenomics study of human colorectal tumors suggested that mRNA levels correlated with protein levels in a significant manner only one-third of the time (36). Furthermore, the study of post-translational modifications (PTMs) of the proteome can add functional context and provide unique insights. While the study of tyrosine phosphorylations was our current focus, other protein PTMs are known to be critical modulators of cellular function and are thus worthy of future study, including serine/threonine phosphorylations, ubiquitination, and acetylation. The discovery of these multiple PTMs may be best facilitated by broad-based MS-based techniques, reserving the development of specific antibodies to these distinct PTMs for validation and mechanistic analytics. Indeed, such complementary analyses offer an exciting opportunity to study the integration of several factors influencing tumor biology including chromatin modifications, somatic mutations, the tumor microenvironment, and signal transduction.

Supplementary Material

1
2
3
4
5

Translational Relevance.

Novel therapeutic targets are needed to combat the acquired resistance that plagues metastatic renal cell carcinoma (RCC) patients treated with targeted agents. Given the importance of tyrosine kinases in cancer signaling, we performed a mass spectrometry-based survey of tyrosine kinases to propose novel therapeutic targets in clear cell and papillary RCC. These data highlighted tyrosine kinases capable of mediating interactions between kidney cancer and the extracellular matrix (ECM). Specifically, FAK was phosphorylated in nearly all samples and multiple FAK tyrosine kinase inhibitors demonstrated significant activity. DDR1, a collagen receptor tyrosine kinase capable of providing oncogenic signaling, demonstrated higher phosphorylation in papillary RCC tumors. These studies highlight the divergent biology of kidney cancer subtypes and reveal novel therapeutic targets.

Acknowledgments

We would like to thank the patients and their families for their participation in the Moffitt Cancer Center Total Cancer Care (TCC) initiative.

Research support: This work has been supported by a Conquer Cancer Foundation Young Investigator Award (SMH), the Moffitt Cancer Center Department of Genitourinary Oncology (SMH), as well as RO1-DK083187, RO1-DK075594, R01-DK069221 and VA Merit Award 1I01BX002196 (RZ). The Total Cancer Care study was enabled, in part, by the generous support of the DeBartolo Family, and the study also received valuable assistance from the Proteomics and Bioinformatics Core Facility at the H. Lee Moffitt Cancer Center & Research Institute, an NCI-designated Comprehensive Cancer Center (P30-CA76292).

Footnotes

There are no conflicts of interest to disclose by the authors.

References

  • 1.Motzer RJ, Bander NH, Nanus DM. Renal-cell carcinoma. The New England journal of medicine. 1996;335:865–875. doi: 10.1056/NEJM199609193351207. [DOI] [PubMed] [Google Scholar]
  • 2.Heng DY, Xie W, Regan MM, Warren MA, Golshayan AR, Sahi C, et al. Prognostic factors for overall survival in patients with metastatic renal cell carcinoma treated with vascular endothelial growth factor-targeted agents: results from a large, multicenter study. J Clin Oncol. 2009;27:5794–5799. doi: 10.1200/JCO.2008.21.4809. [DOI] [PubMed] [Google Scholar]
  • 3.Cancer Genome Atlas Research N. Comprehensive molecular characterization of clear cell renal cell carcinoma. Nature. 2013;499:43–49. doi: 10.1038/nature12222. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Gnarra JR, Zhou S, Merrill MJ, Wagner JR, Krumm A, Papavassiliou E, et al. Post-transcriptional regulation of vascular endothelial growth factor mRNA by the product of the VHL tumor suppressor gene. Proc Natl Acad Sci U S A. 1996;93:10589–10594. doi: 10.1073/pnas.93.20.10589. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Ravaud A, Oudard S, De Fromont M, Chevreau C, Gravis G, Zanetta S, et al. First-line treatment with sunitinib for type 1 and type 2 locally advanced or metastatic papillary renal cell carcinoma: a phase II study (SUPAP) by the French Genitourinary Group (GETUG)dagger. Annals of oncology : official journal of the European Society for Medical Oncology / ESMO. 2015 doi: 10.1093/annonc/mdv149. [DOI] [PubMed] [Google Scholar]
  • 6.Blume-Jensen P, Hunter T. Oncogenic kinase signalling. Nature. 2001;411:355–365. doi: 10.1038/35077225. [DOI] [PubMed] [Google Scholar]
  • 7.Minner S, Rump D, Tennstedt P, Simon R, Burandt E, Terracciano L, et al. Epidermal growth factor receptor protein expression and genomic alterations in renal cell carcinoma. Cancer. 2012;118:1268–1275. doi: 10.1002/cncr.26436. [DOI] [PubMed] [Google Scholar]
  • 8.Gibney GT, Aziz SA, Camp RL, Conrad P, Schwartz BE, Chen CR, et al. c-Met is a prognostic marker and potential therapeutic target in clear cell renal cell carcinoma. Annals of oncology : official journal of the European Society for Medical Oncology / ESMO. 2013;24:343–349. doi: 10.1093/annonc/mds463. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Ravaud A, Hawkins R, Gardner JP, von der Maase H, Zantl N, Harper P, et al. Lapatinib versus hormone therapy in patients with advanced renal cell carcinoma: a randomized phase III clinical trial. J Clin Oncol. 2008;26:2285–2291. doi: 10.1200/JCO.2007.14.5029. [DOI] [PubMed] [Google Scholar]
  • 10.Choueiri TK, Vaishampayan U, Rosenberg JE, Logan TF, Harzstark AL, Bukowski RM, et al. Phase II and biomarker study of the dual MET/VEGFR2 inhibitor foretinib in patients with papillary renal cell carcinoma. J Clin Oncol. 2013;31:181–186. doi: 10.1200/JCO.2012.43.3383. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Choueiri TK, Escudier B, Powles T, Mainwaring PN, Rini BI, Donskov F, et al. Cabozantinib versus Everolimus in Advanced Renal-Cell Carcinoma. The New England journal of medicine. 2015;373:1814–1823. doi: 10.1056/NEJMoa1510016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Fenstermacher DA, Wenham RM, Rollison DE, Dalton WS. Implementing personalized medicine in a cancer center. Cancer journal (Sudbury, Mass) 2011;17:528–536. doi: 10.1097/PPO.0b013e318238216e. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Bai Y, Li J, Fang B, Edwards A, Zhang G, Bui M, et al. Phosphoproteomics identifies driver tyrosine kinases in sarcoma cell lines and tumors. Cancer Res. 2012 doi: 10.1158/0008-5472.CAN-11-3015. Epub ahead of print. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Cline MS, Smoot M, Cerami E, Kuchinsky A, Landys N, Workman C, et al. Integration of biological networks and gene expression data using Cytoscape. Nature protocols. 2007;2:2366–2382. doi: 10.1038/nprot.2007.324. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Aranda B, Blankenburg H, Kerrien S, Brinkman FS, Ceol A, Chautard E, et al. PSICQUIC and PSISCORE: accessing and scoring molecular interactions. Nature methods. 2011;8:528–529. doi: 10.1038/nmeth.1637. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Chen D, Roberts R, Pohl M, Nigam S, Kreidberg J, Wang Z, et al. Differential expression of collagen- and laminin-binding integrins mediates ureteric bud and inner medullary collecting duct cell tubulogenesis. American journal of physiology Renal physiology. 2004;287:F602–F611. doi: 10.1152/ajprenal.00015.2004. [DOI] [PubMed] [Google Scholar]
  • 17.Bulus N, Feral C, Pozzi A, Zent R. CD98 increases renal epithelial cell proliferation by activating MAPKs. PLoS One. 2012;7:e40026. doi: 10.1371/journal.pone.0040026. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Li J, Rix U, Fang B, Bai Y, Edwards A, Colinge J, et al. A chemical and phosphoproteomic characterization of dasatinib action in lung cancer. Nat Chem Biol. 2010;6:291–299. doi: 10.1038/nchembio.332. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Cox J, Mann M. MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification. Nat Biotechnol. 2008;26:1367–1372. doi: 10.1038/nbt.1511. [DOI] [PubMed] [Google Scholar]
  • 20.Zhang G, Fang B, Liu RZ, Lin H, Kinose F, Bai Y, et al. Mass spectrometry mapping of epidermal growth factor receptor phosphorylation related to oncogenic mutations and tyrosine kinase inhibitor sensitivity. Journal of proteome research. 2011;10:305–319. doi: 10.1021/pr1006203. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Sorkin A, Helin K, Waters CM, Carpenter G, Beguinot L. Multiple autophosphorylation sites of the epidermal growth factor receptor are essential for receptor kinase activity and internalization. Contrasting significance of tyrosine 992 in the native and truncated receptors. The Journal of biological chemistry. 1992;267:8672–8678. [PubMed] [Google Scholar]
  • 22.Smith MA, Hall R, Fisher K, Haake SM, Khalil F, Schabath MB, et al. Annotation of human cancers with EGFR signaling-associated protein complexes using proximity ligation assays. Science signaling. 2015;8:ra4. doi: 10.1126/scisignal.2005906. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Calalb MB, Polte TR, Hanks SK. Tyrosine phosphorylation of focal adhesion kinase at sites in the catalytic domain regulates kinase activity: a role for Src family kinases. Molecular and cellular biology. 1995;15:954–963. doi: 10.1128/mcb.15.2.954. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Kaelin WG., Jr The von Hippel-Lindau tumour suppressor protein: O2 sensing and cancer. Nat Rev Cancer. 2008;8:865–873. doi: 10.1038/nrc2502. [DOI] [PubMed] [Google Scholar]
  • 25.Linehan WM, Spellman PT, Ricketts CJ, Creighton CJ, Fei SS, Davis C, et al. Comprehensive Molecular Characterization of Papillary Renal-Cell Carcinoma. The New England journal of medicine. 2015 doi: 10.1056/NEJMoa1505917. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Ferrell JE, Jr, Bhatt RR. Mechanistic studies of the dual phosphorylation of mitogen-activated protein kinase. The Journal of biological chemistry. 1997;272:19008–19016. doi: 10.1074/jbc.272.30.19008. [DOI] [PubMed] [Google Scholar]
  • 27.Ambrogio C, Gomez-Lopez G, Falcone M, Vidal A, Nadal E, Crosetto N, et al. Combined inhibition of DDR1 and Notch signaling is a therapeutic strategy for KRAS-driven lung adenocarcinoma. Nature medicine. 2016 doi: 10.1038/nm.4041. [DOI] [PubMed] [Google Scholar]
  • 28.Valencia K, Ormazabal C, Zandueta C, Luis-Ravelo D, Anton I, Pajares MJ, et al. Inhibition of collagen receptor discoidin domain receptor-1 (DDR1) reduces cell survival, homing, and colonization in lung cancer bone metastasis. Clin Cancer Res. 2012;18:969–980. doi: 10.1158/1078-0432.CCR-11-1686. [DOI] [PubMed] [Google Scholar]
  • 29.Ali-Rahmani F, FitzGerald DJ, Martin S, Patel P, Prunotto M, Ormanoglu P, et al. Anticancer Effects of Mesothelin-targeted Immunotoxin Therapy are Regulated by Tyrosine Kinase DDR1. Cancer Res. 2015 doi: 10.1158/0008-5472.CAN-15-2401. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Sulzmaier FJ, Jean C, Schlaepfer DD. FAK in cancer: mechanistic findings and clinical applications. Nat Rev Cancer. 2014;14:598–610. doi: 10.1038/nrc3792. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Buchheit CL, Weigel KJ, Schafer ZT. Cancer cell survival during detachment from the ECM: multiple barriers to tumour progression. Nat Rev Cancer. 2014;14:632–641. doi: 10.1038/nrc3789. [DOI] [PubMed] [Google Scholar]
  • 32.Linehan WM, Spellman PT, Ricketts CJ, Creighton CJ, Fei SS, Davis C, et al. Comprehensive Molecular Characterization of Papillary Renal-Cell Carcinoma. The New England journal of medicine. 2016;374:135–145. doi: 10.1056/NEJMoa1505917. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Ciamporcero E, Miles KM, Adelaiye R, Ramakrishnan S, Shen L, Ku S, et al. Combination strategy targeting VEGF and HGF/c-met in human renal cell carcinoma models. Molecular cancer therapeutics. 2015;14:101–110. doi: 10.1158/1535-7163.MCT-14-0094. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Remsing Rix LL, Kuenzi BM, Luo Y, Remily-Wood E, Kinose F, Wright G, et al. GSK3 alpha and beta are new functionally relevant targets of tivantinib in lung cancer cells. ACS chemical biology. 2014;9:353–358. doi: 10.1021/cb400660a. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Beck SD, Patel MI, Snyder ME, Kattan MW, Motzer RJ, Reuter VE, et al. Effect of papillary and chromophobe cell type on disease-free survival after nephrectomy for renal cell carcinoma. Annals of surgical oncology. 2004;11:71–77. doi: 10.1007/BF02524349. [DOI] [PubMed] [Google Scholar]
  • 36.Zhang B, Wang J, Wang X, Zhu J, Liu Q, Shi Z, et al. Proteogenomic characterization of human colon and rectal cancer. Nature. 2014;513:382–387. doi: 10.1038/nature13438. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS791673-supplement-1.xlsx (257.9KB, xlsx)
2
NIHMS791673-supplement-2.docx (219.5KB, docx)
3
NIHMS791673-supplement-3.tif (4.6MB, tif)
4
NIHMS791673-supplement-4.docx (12.6KB, docx)
5
NIHMS791673-supplement-5.docx (22.1KB, docx)

RESOURCES