|Home | About | Journals | Submit | Contact Us | Français|
Renal cell carcinoma (RCC), the most lethal type of genitourinary cancer, is generally resistant to chemotherapy and radiation therapy. Surgical excision of the tumor at a localized stage remains the mainstay for curative therapy. A number of drugs developed in recent years have shown limited to significant efficacy in treating RCC. These drugs act by blocking critical signaling pathways associated with RCC tumor growth and survival and angiogenesis. Beyond well-validated signaling targets such as VHL, VEGFR and mTOr, additional pathways including HGF/c-MET and wnt/β-catenin have emerged as important to RCC pathogenesis. Mutations in one or more components of these signaling networks may affect tumor response to therapy. This review summarizes the state of knowledge about signaling pathways in RCC and discusses the known genetic and epigenetic alterations that underlie dysregulation of these pathways.
Renal cell carcinoma (RCC) is the most common type (>80%) of kidney cancer. Claiming more than 100,000 lives per year worldwide, RCC accounts for about 3% of all adult cancers and its incidence is rising. With newer therapies, the median survival period of patients with advanced RCC is about 26 months.1 Men are at a greater risk for developing RCC than women. Hereditary factors, tobacco smoking, obesity, hypertension and related medication and chronic renal failure are some of the common risk factors for RCC.2
RCC originates from the tubular structures of the kidney and is classified into four major histological cell types. Clear cell (cc) is the most common type, accounting for about 75–80% of all cases of RCC. Other types are papillary (10–15%), chromophobe (5%) and collecting duct (1%) RCC. Around 4% of RCCs are hereditary and 96% sporadic.3,4 Virtually all familial cases of ccRCC are due to inherited mutation in the von Hippel-Lindau (VHL) tumor suppressor gene. In addition, in at least two-thirds of sporadic cases of ccRCC, the VHL gene is inactivated either by point mutation, deletion or promoter hypermethylation.5–7
An activating point mutation in the tyrosine kinase domain of the c-MET proto-oncogene is responsible for one form of hereditary papillary RCC. Point mutation of MET is found in only 5–13% of sporadic papillary RCCs.8 A second and rare form of hereditary papillary RCC arises from inactivating mutations of the Krebs cycle enzyme fumarate hydratase (FH) tumor suppressor gene.9,10 There is no evidence of mutation of FH in sporadic papillary RCC.11 While chromophobe RCC is seen in patients with the hereditary Birt-Hogg-Dube (BHD) syndrome, the BHD tumor suppressor gene is rarely mutated in sporadic chromophobe RCC.12 The rare collecting duct RCC arises from the distal portion of the nephron and is usually of high grade and very aggressive. No genes have been shown to be mutated in collecting duct RCC but deletions of chromosomes 1q, 8p and 13q have been reported.13 Because the inherited RCC genes VHL, MET, FH (fumarate hydratase), FLCN (folliculin), SDH (succinate dehydrogenase) and also the TSC1 and 2 (tuberous sclerosis complex) genes are all involved in metabolic pathways related to oxygen, iron, energy and nutrient sensing, Linehan et al. have described RCC as a metabolic disease.14
Cancer is invariably accompanied by changes in signal transduction. Alterations in proto-oncogenes and tumor suppressor genes leads to dysregulated signal transduction that underlies the abnormal growth and proliferation of cancer cells. In some cases, tumor-associated mutations specifically target genes coding for critical signaling proteins. Alternatively, signaling proteins that are centrally located in important cancer-associated signaling networks can serve as therapeutic targets, even though their function is not specifically altered as part of the disease etiology.15,16 This review will summarize the major signaling pathways known to be involved in RCC, the genetic and epigenetic alterations that impact these pathways and the biologic agents targeted to these pathways in RCC.
Like many solid tumors, kidney tumors are frequently characterized by hypoxic conditions due to local imbalance between oxygen (O2) supply and consumption.17 Indeed, hypoxia and compensatory hyperactivation of angiogenesis are thought to be particularly important in RCC compared to other tumor types, given the highly vascularized nature of kidney tumors and the specific association of mutation in VHL, a critical regulator of the hypoxic response, with onset of RCC.18 VHL, in association with elongins B and C and Cullins (Cul2), forms a ubiquitin ligase complex that mediates degradation of hypoxia-inducible target proteins.19
Hypoxic signaling is mediated by a small group of transcription factors called the hypoxia-inducible factors (HIFs), which in turn regulate the expression of over 200 genes involved in crucial pathways related to tumorigenesis including angiogenesis, invasion and mitogenesis. Target proteins regulated by HIFs include VEGF, PDGF, EGFR TGFα, HEF-1, GLUT1 and MUC1 (for a more complete list see Semenza 2010).20 HIFs function as heterodimers consisting of an α-subunit (either HIF-1α or HIF-2α) and a β-subunit (HIFβ). The rate-limiting step in their function is the level of the HIF-α subunit, which varies depending on oxygen availability. Under hypoxic conditions, HIF-1α translocates to the nucleus, where it binds to HIF-β and the transcriptional coactivators p300/CBP to induce the expression of genes such as VEGF, PDGF-β and TGFα, that contain hypoxia-response elements (HRE) in their promoters (Fig. 1).21 In normal cells, this activation is transient. Under normoxic (normal O2 supply) conditions, HIF-1α is hydroxylated by proline hydroxylase and asparagine hydroxylase,22,23 which allow it to be bound by VHL and targeted for proteasomal degradation, limiting the transcription of HIF-dependent gene targets necessary for angiogenesis.24 In hypoxic RCC tumors, in the absence of VHL, HIFα proteins remain constitutively expressed thereby inducing VEGF and other HIF targets. Recent studies on HIFα subunits suggest that based on endogenous HIF expression and VHL status, ccRCC tumors can be classified as three subtypes (i) VHL wild-type, (ii) VHL-/-, HIF-1α and HIF-2α (H1H2) and (iii) VHL-/-, HIF-2α (H2) tumors.25 A related study found that HIF-2α expression, but not HIF-1α, assists tumor cell survival.26
Increased expression of many of the HIF target genes is implicated in promoting cancer,27,28 inducing both changes within the tumor (cell-intrinsic) and changes in the growth of adjacent endothelial cells to promote blood vessel growth. The best-studied HIF targets, VEGF and PDGF, are potent endothelial cell mitogens. The expression level of VEGF in RCC is known to strongly correlate with microvessel density, a measure of the degree of angiogenesis.29 A key step in angiogenesis is the upregulation of growth factor receptors on endothelial cells such as VEGFR and PDGFR. Other HIF targets include genes involved in glucose metabolism (Endoglin and others),20 cell proliferation and survival (TGFα and EGFR) and metastasis (MUC1). The HIF-1α target gene carbonic anhydrase IX (CAIX) has been extensively studied as a prognostic marker for RCC.30 Expression of the microRNA miR-210 is upregulated in ccRCC compared to normal renal tissue.31,32 MiR-210 is upregulated in hypoxia and its expression can be induced by both HIF-1α and HIF-2α.33
Protein kinase B (Akt) and mammalian target of rapamycin (mTOR) are hubs for key oncogenic processes including cell proliferation, survival and angiogenesis. mTOR inhibitors have showed promise for RCC in phase I clinical trials. Autocrine binding of VEGF and PDGF to their receptor tyrosine kinases (VEGFR, PDGFR, KIT) on RCC tumor cells activates PI3K, which in turn promotes the generation of phosphatidylinositol-3,4,5-triphosphate (PIP3). PIP3 recruits the cytoplasmic kinase AKT to the cell membrane, where it is activated by phosphorylation at two independent sites mediated by PDK1 and mTOR (TORC2) respectively.14 AKT activation inhibits apoptosis by phosphorylating and inactivating proapoptotic proteins such as procaspase 9, the bcl2 family member BAD, and apoptosis signal regulating kinase 1 (ASK1).34–36 It also inactivates GSK-3β, which would normally phosphorylate and induce the degradation of cell cycle promoting proteins such as cyclin D1,37 and proliferation-promoting transcription factors such as c-myc, β-catenin, c-Jun and Notch. Signaling from VEGF and PDGF through AKT also activates mTOR. This protein functions as a component of two distinct complexes: TORC1 (sensitive to rapamycin), which positively regulates protein synthesis and cell cycle and TORC2 (insensitive to rapamycin), which regulates cell polarity and spatial growth by remodeling actin cytoskeleton.38,39 As mentioned above, the TORC2 complex is also thought to phosphorylate AKT and thus mTOR is both upstream and downstream of AKT activation.
However, to date only a single publication reports that components of this pathway, specifically Akt, mTOR and p70S6K are constitutively activated in RCC compared to normal renal tissue.40 No point mutations of PIK3CA have been found in RCC. Negative regulation by the PTEN tumor suppressor gene may be the underlying mechanism for Akt activation in cancer41–43 including RCC.44 Inactivation of PTEN, generally by homozygous deletion as point mutations are rare, has been identified in a subset of RCC.45
Wnts are a family of secreted glycoproteins that regulate cell proliferation, differentiation and cell migration.46 The ultimate effector of canonical Wnt signaling is the transcriptional coactivator β-catenin, which is emerging as a key molecule in the pathogenesis of renal cancer. In normal quiescent cells, β-catenin is trapped in a highly processive enzyme complex containing casein kinase 1 (CK1), glycogen synthase kinase 3β (GSK3β), adenomatosis polyposis coli protein (APC) and axin. β-catenin is phosphorylated at serine and threonine residues by this complex and targeted for proteosomal degradation.47 Wnt positively regulates β-catenin, inhibiting its phosphorylation, ubiquitination and degradation. Stabilized β-catenin enters the nucleus and, together with a member of the LEF-TCF (lymphoid enhancer-binding factor 1-T cell specific transcription factor 7) family of transcription factors, activates target genes such as the MYC oncogene.48 MYC also shows copy number amplification in a subset of primary ccRCC49 and papillary RCC.50 Wnt is also thought to mediate its effect on cell growth and tumor promotion by activating the mTOR pathway.51 TSC2 is sequentially phosphorylated by AMPK and GSK3 for its activation and subsequent inhibition of mTOR. Wnt activates the mTOR pathway by inhibiting GSK3.51
There are several lines of evidence for the involvement of the Wnt signaling pathway in RCC. Though β-catenin-activating point mutations are rare in RCC,52 elevation of β-catenin levels by induced overexpression induces renal tumors in mice.53–55 In a subset of RCC, the APC gene promoter is aberrantly hypermethylated,56 providing one means to liberate nuclear β-catenin. Peruzzi et al. discovered β-catenin is degraded by the E3-ubiquitin ligase activity of VHL and loss of VHL enables HGF-driven oncogenic β-catenin signaling as a novel target for VHL thus implicating Wnt signaling in the pathogenesis of renal cancer.57 Further evidence for the activation of Wnt signaling pathway in RCC comes from a recent article by Kojima et al.58 which describes the homozygous deletion of CXXC4, a gene coding for Idax (an inhibitor of Wnt signaling pathway) in aggressive RCC. The secreted-Frizzled receptor proteins (sFRPs), Dickkopf 2 (DKK2) and Wnt inhibitory factor 1 (WIF-1) are Wnt antagonists and expression of these genes is also silenced by aberrant hypermethylation in RCC.59–63
Thus, Wnt has a dual role in pathogenesis of RCC. It not only induces transcription through activation of β-catenin, but also stimulates translation and cell growth through activation of the mTOR pathway. Linehan et al.64 suggest that loss of VHL could lead to combined de-repression of HIFs and β-catenin, which in turn might contribute to malignancy in ccRCC. Recently, a VHL-interacting protein Jade-1 (gene for apoptosis and differentiation in epithelia) has been shown to be a novel E3 ubiquitin ligase that ubiquitinates β-catenin leading to its degradation. Jade-1 is positively regulated by VHL and is thought to function as a renal tumor suppressor.65,66 Loss of VHL results in reduced levels of Jade-1 and consequent increased levels of β-catenin, providing yet another mechanism by which VHL loss promotes renal tumorigenesis.
In addition to a crucial role in the initiation of a tumor, hypoxia is also involved in tumor metastasis. Yang et al. described a link between hypoxia, HIF-1α and the transcription factor TWIST,67 which shows that upregulation of TWIST results in induction of (EMT). The kidney is mesenchymal in origin and develops through MET (mesenchymal to epithelial transition) to form epithelial structures, which further differentiate to form mature nephrons.68 In ccRCC this transition is reversed and results in EMT and dedifferentiation. EMT is an essential process before metastasis can occur. EMT requires co-activation of several important signaling receptors such as FGFR, EGFR, HGF and other proteins that leads to the activation of transcriptional regulators such as Snail, Slug, ZEB1 and SIP1 which in turn regulate changes in gene expression patterns that underlie EMT.69 The main target for these regulators is E-cadherin, which is crucial for maintaining an epithelial phenotype. Loss of E-cadherin results in the dissociation of intercellular epithelial junctional complexes. The E-cadherin gene is hypermethylated in 11% of primary RCC70 but it is not known if this subset is associated with metastatic potential.
Two independent microarray studies reported miR-141 and miR-200c to be significantly downregulated in ccRCC compared to normal renal tissue.32,71 MiR-141 and miR-200c inhibit EMT by directly targeting ZEB1 and SIP1, which are repressors of E-cadherin.72,73 In metastatic cancer cells, these miRNAs are downregulated thereby allowing ZEB1 and SIP1 to repress E-cadherin. SIP1 can also directly activate vimentin expression, a marker for ccRCC.74 Another important function of E-cadherin is to sequester β-catenin in the cytoplasm. Repression of E-cadherin by Snail and Slug also releases β-catenin, which relocates to the nucleus and activates transcription of mesenchymal markers such as vimentin, fibroblast-specific protein 1 (FSP1), Snail, slug, etc. Recently, it has been shown that deregulation of HIF-1α in VHL-negative cells is linked to downregulation of E-cadherin and induction of the EMT. An interesting multifunctional cytokine implicated in the EMT is TGFβ. It has differential effects on the EMT in normal and tumor cells. In normal and pre-malignant cells, TGFβ acts as a tumor suppressor. Malignant cells often become resistant to the growth inhibitory effects of TGFβ and overexpress this cytokine. A different tumor microenvironment can induce TGFβ to function as a tumor promoter.75,76
Kidney tissue is an abundant source of hepatocyte growth factor (HGF) and its activator, urokinase.77 Changes in expression and activity of HGF and its receptor c-MET, have been associated predominantly with papillary RCC because germline oncogenic mutations in the gene coding for c-MET are responsible for one form of hereditary papillary RCC.78,79 However, activating point mutations in MET, located on human chromosome 7, are found in only 5–13% of sporadic papillary RCC8,80 and most papillary RCCs show trisomy 7 without mutation of MET. Met inhibitors are in clinical trials for treatment of RCC and other cancers.81 Some of the therapeutic strategies to inhibit this pathway are: inhibition of autophosphorylation of c-MET; blocking interaction between HGF and c-MET; and suppression of the downstream signaling cascade of activated c-MET.82
HGF binding to MET leads to phosphorylation of two tyrosine residues at the C-terminus of MET, which in turn leads to the recruitment of adapter proteins such as Gab1, Grb2, SHC, STAT3 and PI3K and activation of the Ras/MAPK and PI3K/AKT effector pathways to promote RCC growth and metastasis.83 MET phosphorylation also induces tyrosine phosphorylation of β-catenin (different from the serine/threonine phosphorylation by GSK3/APC complex described above in the Wnt pathway) causing β-catenin dissociation from E-cadherin followed by nuclear translocation of β-catenin and transcription activation.84 Peruzzi et al. showed that VHL expression in RCC cells suppressed HGF-stimulated β-catenin signaling and therefore loss of VHL in RCC could enable HGF-driven oncogenic β-catenin signaling.57
In addition to the gene or miRNA alterations relevant to the pathways discussed above, inactivating point mutations in the histone modifying genes SETD2, JARID1C, UTX and MLL2 were reported in 12–17% of ccRCC.85 Mutation of the p600 retinoblastoma associated protein ZUBR1 gene (10%), the NF2 gene (7%) and the WRN and NBN DNA double strand repair genes (2–3%) were also found in ccRCC.85 Tumor suppressor genes play a vital role in regulating signaling pathways. Inactivation by point mutation, deletion or promoter hypermethylation of these tumor suppressor genes enables tumor cells to evade this mode of control. Biallelic mutations of the tumor suppressor genes p53 or Rb are found in less than 2% of RCC. The p16 and p14 tumor suppressor genes are inactivated by homozygous deletion,86 or more rarely promoter hypermethylation, in around 10–20% of primary RCC. In addition to classical tumor suppressor genes, promoter hypermethylation of other cancer genes is evident in RCC notably of the RASSF1A gene a mediator of pro-apoptotic effects of K-Ras.87,88
Oncogenic RAS or BRAF mutations leading to activation of downstream signaling pathway are found in only 1% of kidney tumors.85,89,90 No point mutations of the EGFR or HER2/neu proto-oncogenes have been reported in RCC.85,90
RCC confined to the kidney can be successfully treated by surgical resection alone. Cytokine or the newer targeted therapies are used to treat patients with locally advanced or metastatic disease.91 Cytokine immunotherapies such as high-dose interleukin 2 (IL-2) or interferon α (IFNα) work effectively only in a minority of ccRCC patients.92 Among the drugs available to treat metastatic RCC are the kinase inhibitors sunitinib, sorafenib and pazopanib. Sunitinib has been evaluated for its inhibitory activity against more than 80 different kinases, some of which are implicated in tumor growth, angiogenesis and metastasis.93,94 Sunitinib inhibits a broad range of receptor tyrosine kinases including VEGF receptors (VEGFR1, VEGFR2 and VEGFR3), PDGF receptors (PDGFRα and PDGFRβ), fms-like tyrosine kinase 3 (FLT3) and stem cell factor receptor (KIT). Sorafenib targets an even broader spectrum of kinases, but its inhibiting activity is weaker compared to sunitinib. Pazopanib is another angiogenic inhibitor that is recently approved for treatment of advanced RCC.95
Temsirolimus, a powerful inhibitor of the PI3K/Akt/mTOR pathway is often the front-line therapy of choice in patients with metastatic RCC who have poor prognosis.96 It is a rapamycin analog and binds to FK506-binding protein 12 (FKBP12). The temsirolimus-FKBP12 complex binds to mTOR and inhibits its kinase activity by an allosteric mechanism. Temsirolimus has been shown to decrease the levels of HIF-1α and VEGF in vitro.97 Another rapamycin analog, everolimus, was recently approved for treatment of RCC.98 Metastatic RCC patients who are refractory to sunitinib and sorafenib are treated with everolimus as a second-line therapy. Another mTOR inhibitor, deferolimus, is yet another promising rapamycin analog in RCC treatment but not yet approved. Since VEGF, an important angiogenic factor, is overexpressed in most or all ccRCC, bevacizumab, a humanized VEGF-neutralizing antibody has shown promising results in inhibiting tumor angiogenesis. Importantly, all the VEGF-targeted therapies are believed to predominantly act on the normal endothelial cells surrounding the tumor rather than the tumor per se.99–101 These therapies mainly inhibit angiogenesis and exhibit anitumor activity by blocking the supply of oxygen and nutrients to the tumor cells. Unlike VEGF-targeted therapies, mTOR inhibitors are thought to act directly upon the tumor cells.102,103 In general, non-ccRCC show a lower response rate to the kinase inhibitors and mTOR inhibitors. MET-inhibitors are in trial and have demonstrated anitumor activity for papillary RCC. Both sunitinib and sorafenib inhibit the KIT proto-oncogene, which shows overexpression in chromophobe RCC.104
Current chemotherapeutics can only increase the overall survival of patients from weeks to months and cannot cure RCC. Despite major advances in VEGF-targeted therapy, nearly all tumors are either inherently resistant or develop acquired resistance to these agents. Since targeting angiogenesis alone may not be sufficient for the eradication of RCC, there is a compelling need to identify both mechanisms of resistance and novel drugs to treat the disease. Genetic and epigenetic markers are emerging as promising biomarkers to predict response to therapy in solid tumors. A better understanding of chemoresistance and biomarkers for stratification of patients who will benefit or not from existing therapy, e.g., the 30% that show no response to sunitinib, is needed. As yet, there are no such chemoresponse markers available for RCC. Future studies, in particular the Cancer Genome Atlas consortium, will further elucidate the genome, mRNA and miRNA transcriptome and methylome of RCC revealing the pathways and networks perturbed in RCC and thus likely identify new biomarkers and therapeutic targets.
The authors would like to thank and acknowledge Erica Golemis and Gary R. Hudes at FCCC for their reading and input to this review. This publication was supported in part by grant number P30 CA006927 from the National Cancer Institute. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Cancer Institute or the National Institutes of Health. Additional funds were provided by Fox Chase Cancer Center via institutional support of the Kidney Keystone Program.
Previously published online: www.landesbioscience.com/journals/cbt/article/13247