The 786-O and Caki-2 RCC cell lines, which differ in HIF 2α, recapitulate the differences between VHL
-wt and VHL
-mut RCC. Furthermore, they have been extensively used for RCC studies 
. Before proteomic analysis, we confirmed the differential expression of HIF-2α between the RCC cell lines (Fig. S1
Our method of choice for this first study in RCC varying in HIF-2α expression was based on our earlier comparison of 2D-LC-MS/MS results with those obtained from shotgun LC-MS/MS analyses demonstrating that most molecular functions and biological processes were represented via gene ontology (GO) analysis using either methodology 
. The analytical replications were >95% across all the samples analyzed in both VHL
-wt and VHL
-mut cells. We employed rigorous criteria for the evaluation of significant protein expression differences between VHL
genotypes of RCC. Specifically, we required proteins to match Scaffold protein IDs, as well as have a minimum 2-fold change and an ANOVA P
-value of <0.05 as determined by Progenesis LC-MS analyses (MS).
The introduction of Progenesis LC-MS into our workflow of label-free proteomics we employed earlier has resulted in a substantial improvement over previous software which was implemented for quantitative expression profiling based on spectral counting 
. We have been able to verify a higher proportion of differentially regulated proteins obtained by integration of extracted peptide intensities (via Progenesis LC-MS software), which would have been excluded due to criteria applied to the evaluation of spectral counts (e.g., requiring minimum spectral counts and using stringent G
-tests). Apparently, statistics applied to extracted peptide ion intensities did not compromise evaluation of lower-abundance proteins for differential expression, which would support arguments to phase out spectral counting methods in future quantitative studies 
. lists representative VHL
genotype- regulated RCC proteins. Additional proteins and peptides identified from Mascot database searching can be found in the Supporting Information (Table S1
Representative differentially expressed proteins of VHL-wt (Caki-2) and VHL-mut (786-O) RCCs identified by database search (Mascot) and quantified by extracted peptide intensity (MS) features generated with Progenesis LC-MS.
IPA generated four networks from differentially expressed proteins. Proteins with significant differences in expression were involved in regulation of cell-to-cell signaling and interaction, tissue development, and cancer (Network-1; ). In this network, the well-known tumor suppressor gene TP53 takes an apparent central role. TP53 mutations are the most frequent genetic alterations found in human cancer 
, and renal pathology and DNA adduct analysis have identified the gene as one of the most important pathways modulated in the kidney by human carcinogens 
. In addition, a network was identified for nucleic acid metabolism, small molecule biochemistry, cellular assembly and organization (Network-2, Fig. S2A
) with presumed involvement of the Myc oncogene 
under the direct control of the VHL
-mut downregulated 40S ribosomal protein S20 (RSP20) and myosin-9 (MYH9), as well as through interaction with the VHL
-mut upregulated heterogeneous nuclear ribonucleoprotein K (HNRNPK) according to our proteomic analysis of 786-O and Caki-2 RCC cell lines (Table S1
). Networks associated with cellular development, cellular growth and proliferation, hematological system development and function (Network-3, Fig. S2B
), hereditary disorder, metabolic disorder, and carbohydrate metabolism (Network-4, Fig. S2C
) were also revealed by IPA. The complex network that could be built from combining Networks 1–4 was given in the Supporting Information (Fig. S2D
). The quantitative difference in the expression of HSPD1, a protein found to be an important node in two networks (1 and 3, and ) was confirmed by Western blot analyses (Fig. S3
Ingenuity Pathway Analyses (IPA).
Assessment of cell-survival by MTT assay.
Differential Regulation of Cellular Energy Pathways in VHL-mut and VHL-wt RCC Cells
IPA revealed signaling pathways that regulate tumor energetics such as protein ubiquitination, methylglyoxal degradation III, glycolysis, gluconeogenesis, and pentose phosphate pathways in RCC cells according to VHL
-genotype (). Differentially expressed proteins ( and Table S1
), along with their impact on the cellular energy metabolism as implicated by the IPA analysis (), also indicated significant differences in metabolic preferences of VHL
-mut versus VHL
-wt RCC. Metabolism in cancer cells is mainly driven by the flux of metabolites through three processes: glycolysis, pentose phosphate pathway and glutaminolysis 
. Rapidly proliferating cancer cells have enhanced aerobic glycolysis compared to non-malignant cells, which is commonly known as Warburg effect 
. Previously published proteomic studies have strongly indicated the up-regulation of glycolysis in RCC cells 
. However, pathways that maintain cellular energy status specifically in HIF2α over-expressing VHL
-mut background, which corresponds to an advanced stage of RCC, have not been reported. The tumor cells with impaired glycolysis maintain the downstream metabolite flux by deriving the intermediate metabolites, fructose 6-phosphate and dihydroxyacetone phosphate, from the pentose phosphate pathway 
. To our surprise, glucose-6-phosphate isomerase (GPI) of the glycolytic pathway ( and ) and UDP-glucose 6-dehydrogenase (UGDH) of the pentose phosphate pathway (Fig. S2A
and Table S1
) were downregulated. However, a significant upregulation of mitochondrial malate dehydrogenase 2 (MDH2) was observed in VHL
-mut RCC cells (Fig. S2B
and Table S1
). The differential expression of cytosolic GPI of the glycolytic pathway and mitochondrial MDH2 of the glutaminolytic pathway were strongly supported by Progenesis LC-MS analysis based on extracted peptide intensities (Fig. S4
and Fig. S5
). This novel finding strongly indicated a possible metabolic shift towards glutaminolysis in advanced stages of RCC that overexpress HIF2α. Glutamine is converted to glutamate and shunted into the tricarboxylic acid (TCA) cycle as α-ketoglutarate that ultimately yields malate. MDH2 converts malate to oxaloacetate to afford NADH that is further shuttled into mitochondrial respiratory chain to yield ATP by the action of ATP synthase 
. In addition to mitochondrial MDH2, ATP synthase α and β (ATP5A1 and ATP5B; and Fig. S2A
) also were overexpressed in VHL
-mut RCC. Thus, the upregulation of mitochondrial MDH2, ATP5A1 and ATP5B in VHL
-mut cells, in conjunction with downregulation of cytosolic GPI, strongly indicated a shift towards alternative glutaminolytic metabolic pathway in RCC associated with an advanced stage of malignancy.
IPA Analysis: Cellular Pathways Significantly Represented by HIF2α Regulation in RCC Cell Lines.
In the context of downregulation of key glycolytic and pentose phosphate pathway enzymes, especially GPI and UGDH, revealed by proteomic analysis of VHL-mut RCC, we further investigated the importance of glutaminolysis. We first analyzed the degree of cell survival using MTT assays at 24, 48 and 72 h during incubation with aminooxyacetate (AOA), an inhibitor of glutaminolysis (). The VHL-mut RCC cells were more susceptible to inhibition of glutaminolysis relative to VHL-wt RCC cells at all concentrations of AOA and at all time-points tested. In addition to using glutaminolytic inhibitor, we also studied the extent of cell survival upon culturing in medium lacking glutamine. Again, VHL-mut RCC cells were more susceptible to glutamine depletion than VHL-wt RCC cells ().
By comparing the results of cell survival assays after glutaminolytic pathway inhibition by using AOA and glutamine depletion by using medium lacking glutamine, we observed that VHL-mut cells cultured in glutamine depleted medium had moderately higher survival when compared to survival upon treatment with the glutaminolytic inhibitor. This could be due to the presence of glutamate in the cell culture medium, which could compensate for glutamine deprivation when cultured without glutamine. By acting downstream of glutamate, the increased inhibitory effect of AOA is due to its ability to prevent conversion of glutamine to α-ketoglutarate, which would explain the low survival of VHL-mut RCC cells in the presence of AOA when compared to lack of glutamine substitution in the medium alone. Collectively, when compared with VHL-wt RCC, the cell survival assay both in the presence of AOA and upon cell cultures in glutamine-depleted medium confirmed the predominant dependence of VHL-mut RCC on active glutaminolytic pathway for survival.
Regulation of Complex Cellular Energetics, Proliferative and Metastatic Signaling Networks in VHL-mut RCC
The differential regulation of various cellular processes, systemic protein networks, and toxicological pathways are increasingly becoming a point of reference in both enhancing the understanding about critical players in tumor biology as well as in designing prudent monotherapies and multi-drug combinations for aggressive cancers. In this regard, the IPA analyses revealed the pathways, potential interactions, and networks of differentially regulated proteins in advanced VHL
-mut RCC, which can guide further interventional studies in RCC. The top metabolic cellular pathways that were differentially regulated included glycolysis, gluconeogenesis, protein ubiquitination pathway, and pentose phosphate pathway (). These dominant differences in energy metabolism pathways provide additional corroborative evidence for the significance of decreased glycolysis and increased glutaminolytic preferences of VHL
-mut RCC versus VHL
-wt RCC, which was also confirmed by additional molecular assays in this study. Given the increasing interventional focus on targeting specific metabolic transformation and dependencies in cancers, these differential metabolic pathways revealed by IPA could guide such integrated metabolic targeted therapies in RCC 
The top differentially regulated toxicological pathways were related to hypoxia-inducible factor signaling and oxidative stress (). Moderate levels of oxidative stress stimulate cancer cell proliferation, whereas high levels of ROS are toxic even to cancer cells 
. The current drug of choice in RCC sorafenib not only inhibits tyrosine kinases, but also induces cell death through induction of mitochondrial oxidative stress 
. Thus, the differential regulation of key networks of energy metabolism and oxidative stress pathways assumes significance in further directing the investigations into molecular mechanisms that regulate differential drug sensitivity for RCC drugs like sorafenib. In this context, the regulation of HSPD1 by the VHL
-genotype of RCC may deserve attention. The heat shock family of proteins is one of the major regulators of the cellular response to oxidative stress and, hence, regulates the response to chemotherapeutic interventions 
. HSPD1 is a predominant mitochondrial protein involved in maintaining the integrity of mitochondrial proteins and its expression is increased in response to ROS generation induced due to hypoxia, high temperatures and upon exposure to toxic chemicals 
IPA Analysis: Toxicological Cellular Pathways Significantly Represented by HIF2α Regulation in RCC.
Finally, the top differentially regulated molecular and cellular functions between VHL genotypes of RCC were cell death and survival, cancer, nucleic acid metabolism, cellular growth and proliferation, and energy production (). Other significant differentially regulated cellular functions included RNA post-transcriptional modification and cellular compromise which are of fundamental relevance to cancer cell survival and proliferation.
Summary of Ingenuity Pathways Analyses (IPA) revealing differentially regulated molecular and cellular functions in VHL-mut RCC compared to VHL-wt RCC.