RAGE is essential for oncogenic KRAS-mediated hypoxic signaling in pancreatic cancer

A hypoxic tumor microenvironment is characteristic of many cancer types, including one of the most lethal, pancreatic cancer. We recently demonstrated that the receptor for advanced glycation end products (RAGE) has an important role in promoting the development of pancreatic cancer and attenuating the response to chemotherapy. We now demonstrate that binding of RAGE to oncogenic KRAS facilitates hypoxia-inducible factor 1 (HIF1)α activation and promotes pancreatic tumor growth under hypoxic conditions. Hypoxia induces NF-κB-dependent and HIF1α-independent RAGE expression in pancreatic tumor cells. Moreover, the interaction between RAGE and mutant KRAS increases under hypoxia, which in turn sustains KRAS signaling pathways (RAF-MEK-ERK and PI3K-AKT), facilitating stabilization and transcriptional activity of HIF1α. Knock down of RAGE in vitro inhibits KRAS signaling, promotes HIF1α degradation, and increases hypoxia-induced pancreatic tumor cell death. RAGE-deficient mice have impaired oncogenic KRAS-driven pancreatic tumor growth with significant downregulation of the HIF1α signaling pathway. Our results provide a novel mechanistic link between NF-κB, KRAS, and HIF1α, three potent molecular pathways in the cellular response to hypoxia during pancreatic tumor development and suggest alternatives for preventive and therapeutic strategies.

Pancreatic cancer is the fourth leading cause of cancer mortality in the USA. 1 The high mortality rate is due to the high incidence of metastatic disease at initial diagnosis, aggressive clinical course, and failure of systemic therapies. Over the past decade, numerous trials have been conducted to improve the outcomes in patients with metastatic disease by combination therapies employing gemcitabine as a foundation. 2 Despite a modest improvement in progression-free survival in combination with other drugs, the longer disease-free survival has not translated into any advantage in the overall survival. 3 New therapeutic strategies that specifically target cancer-specific pathways are required for the successful treatment and prevention of this disease. The small guanosine triphosphatase KRAS is frequently mutated in 98% of pancreatic cancers, making it an ideal target for therapeutic intervention. [4][5][6] Attempts to target this pathway have largely been unsuccessful. A better understanding of the molecular pathways involved in KRAS-driven development and progression of pancreatic cancer could potentially lead to improved targeted therapies. 7 Evidence from experimental and clinical studies increasingly points to an important role for hypoxia in the pathogenesis of solid cancers, 8 including pancreatic cancer. [9][10][11][12][13][14] Hypoxia-inducible factor 1 (HIF1) is a master transcription factor that regulates the oxygen supply balance and demand in response to intratumoral hypoxia and aberrantly, with oncogenic mutations. 15 HIF1 is a heterodimer composed of an α-subunit (HIF1α) and a β-subunit (HIF1β). Its activity primarily depends on the stability and modification of HIF1α. 15 Under normoxic conditions, HIF1α is unstable and undergoes immediate proteasomal degradation. Under hypoxic conditions, HIF1α is stabilized and induces the transcription of several genes involved in the development of inflammation, anaerobic metabolism, metastasis, and drug resistance. 8,15 The receptor for advanced glycation end products (RAGE, also termed AGER) is a transmembrane receptor of the immunoglobulin super family encoded within the Class III region of the major histocompatibility complex. RAGE activation is responsible for enhancing inflammation and has been implicated in several chronic diseases including diabetes, neurodegenerative disorders, and cancer. 16,17 RAGE is expressed by malignant cells as well as other cells within the tumor microenvironment including leukocytes, endothelial cells, and fibroblasts. We recently demonstrated that RAGE has a unique role in pancreatic tumorigenesis and drug resistance. [18][19][20][21][22] Knock down or knock out of RAGE in vitro or in vivo attenuates oncogenic KRAS-driven pancreatic tumor progression 19 and reverses the resistance to multiple drugs. 20,21 We demonstrated that the mechanism by which this occurs is partly involved in inflammatory response-associated metabolic changes 19,22,23 and cell death-associated autophagy. 20,21,24,25 Thus, targeting RAGE represents a novel approach for pancreatic cancer therapy.
Using murine studies as well as molecular, and cellular studies in vitro, we demonstrate here that binding of RAGE to mutant KRAS is a direct, positive regulator of HIF1α-dependent hypoxia signaling in pancreatic tumor development. These findings further refine our understanding of how RAGE promotes the development of pancreatic cancer and offer a novel mechanism by which an inflammatory gene can promote cell survival through hypoxic pathways.
NF-κB-mediated RAGE expression increases stabilization and transactivation of HIF1α under hypoxic conditions in pancreatic tumor cells. Given that hypoxia increases RAGE expression independent of HIF1α we next asked whether RAGE has a role in the regulation of HIF1α's stabilization and transactivation. Such results suggest that NF-κB-mediated RAGE expression is a positive regulator of the HIF1α signaling pathway, regulating its stabilization and transactivation.
RAGE-mediated KRAS pathway activation contributes to hypoxia-induced HIF1α activity in pancreatic tumor cells. We next evaluated potential mechanisms by which RAGE could enhance HIF1α-signaling under hypoxia. Our previous work demonstrated that RAGE has a critical role in the regulation of KRAS-driven pancreatic tumorigenesis. 19 We now examined whether RAGE and KRAS directly interacted with each other. Immunoprecipitation Hypoxia-induced autophagy via the RAGE-KRAS-HIF1α pathway is a survival mechanism in pancreatic tumor cells. Although HIF1α has a major role in the cell survival response to hypoxia, 30,31 it also is associated with cell death in some instances. 32 Given that the RAGE-mediated KRAS pathway is an important regulator of HIF1α signaling ( Figures  2 and 3), we therefore determined whether the knock down of RAGE regulates cell survival and cell death under hypoxic conditions. Similar to our previous study, 21 knockdown of RAGE and HIF1α decreased cell viability as assessed in a CCK8 assay (Figure 4a) and increased apoptosis demonstrated by enhanced caspase-3 activity (Figure 4b) following hypoxia. The RAF inhibitor (e.g., RAF265), MEK inhibitor (e.g., U0126), PI3K inhibitor (e.g., LY294002), and AKT inhibitor (e.g., 1,3-Dihydro-1-(1-((4-(6-phenyl-1H-imidazo (4,5-g)quinoxalin-7-yl)phenyl)methyl)-4-piperidinyl)-2H-benzimidazol-2-one trifluoroacetate salt hydrate) all increased hypoxia-induced cell death (Figure 4c). These findings suggest that activation of the RAGE-KRAS-HIF1α pathway under hypoxia promotes the survival of pancreatic tumor cells. Previous studies suggested that autophagy, a lysosome-mediated degradation pathway, is a major prosurvival mechanism of various tumor cells in response to hypoxia. 33,34 Similarly, the knock down of key autophagy regulators (ATG5 and Beclin1) by shRNA inhibits their protein expression (data not shown) and promotes pancreatic tumor cell death under hypoxic conditions (Figure 4d), supporting a pro-survival role of autophagy in the hypoxic tumor microenvironment. 35  (e and f) Indicated Panc02 cells were treated with 1% O 2 for 24 h with or without MEK inhibitor (e.g., U0126, 10 μM) and PI3K inhibitor (e.g., LY294002, 10 μM). Protein levels were assayed by western blot is induced, whereas it accumulates when autophagy is impaired. 36 Consistently, the knock down of RAGE and HIF1α in Panc02 cells prevents the expected decrease in p62 levels following induction of hypoxia ( Figure 4e). Moreover, MEK inhibition (e.g., U0126) and PI3K inhibition (e.g., LY294002) also inhibit autophagy (LC3-II/LC3-I ratio) and increase apoptosis (cleaved PARP) in response to hypoxia (Figure 4f). Taken together, our findings suggest that autophagic flux is enhanced under hypoxic conditions, dependent on cell survival signals generated by activation of the RAGE-KRAS-HIF1α pathway.

Genetic ablation of RAGE impairs mutated KRAS and
HIF1α signaling in vivo. Studies of human pancreatic ductal adenocarcinoma have been greatly facilitated by the development of a genetically engineered mouse model that expresses oncogenic KRAS under a pancreatic promoter Pdx1-Cre:Kras G12D/+ (KC mice). 37 Our published findings demonstrate that RAGE promotes KRAS G12D -driven pancreatic tumorigenesis. 19 Knock out of RAGE delayed KRASdriven pancreatic intraepithelial neoplasia progression and prolonged murine survival (KCR mice). 19 Mutant KRAS enhances HIF1α expression and activity in the development of several tumors. 38 To address the role of RAGE in the regulation of KRAS-mediated HIF1α hypoxia signaling, we stained HIF1α and its target proteins in KC and KCR mice. Compared with KC mice, KCR mice exhibited significantly downregulated expression of HIF1α ( Figure 5a) and its targets genes involved in angiogenesis (VEGFA), anaerobic metabolism (GLUT1, LDHA, and PDPK1), and metastasis (CXCR4) (Figure 5b). Indeed, our previous studies demonstrated that depletion of RAGE in mice decreases the KRASmediated autophagy response and bioenergetics in vivo. 19,22 Similar to our observations in vitro (Figures 3c and 4e), knock out of RAGE inhibits KRAS-mediated p-ERK1/2, p-AKT, and LC3-II ( Figure 5a). Collectively, these findings suggest that RAGE expression is required for KRAS-mediated activation of HIF1α signaling and pancreatic cancer bioenergetics and growth in vivo.

Discussion
Pancreatic cancer, is characterized by near-universal mutations in the KRAS oncogene and a hypoxic and proinflammatory tumor microenvironment. In this study, we demonstrate that NF-κB-driven expression of RAGE sustains KRAS and HIF1α activation in response to hypoxia. This facilitates pancreatic tumor growth and proliferation. These findings provide a novel mechanistic link between NF-κB, KRAS, and HIF1α, three potent oncogenes in the cellular response to hypoxia and pancreatic tumor development.
We have demonstrated that constitutive RAGE activation is a characteristic signature in pancreatic cancer cells. 19 However, how RAGE is activated in the hypoxic tumor microenvironment is not fully understood. Many stressful stimuli such as inflammation, hypoxia, and oxidative stress can activate the NF-κB pathway and increase RAGE expression in immune and tumor cells. Our previous study demonstrated a positive feedback loop between RAGE and NF-κB during oxidative stress in pancreatic tumor cells exposed to chemotherapy. 20 NF-κB activation within the tumor microenvironment regulates a number of genes important in cell death, metabolism, immunity, and inflammatory responses, contributing to oncogenesis. 39 HIF1α is a master regulator of hypoxia 15 and its expression and activity is also controlled by NF-κB. 40,41 Here, we demonstrated that genetic and pharmacologic inhibition of NF-κB, but not HIF1α, significantly inhibits hypoxia-induced RAGE expression. These findings suggest that NF-κB is the major transcription factor regulating the expression of RAGE in response to various stressors, including hypoxia, in pancreatic tumor cells.
Although our knowledge of the molecular function of RAGE during neoplastic transformation and malignant progression is limited, recent experimental data support a direct link between RAGE activation and proliferation, survival, migration, and invasion of tumor cells. 42,43 Blockade of the RAGE signaling pathway suppresses tumor growth, angiogenesis, and metastases. 16,17 We and others have demonstrated that RAGE-deficient mice are resistant to chemically-induced skin carcinogenesis, colitis-associated cancer induction, as well as pancreatic tumorigenesis. 19,[43][44][45] However, RAGE has shown tumor suppressor activity in several cancer types, including esophageal squamous cell 46 and oral squamous cell carcinomas, 47 lung carcinoma, 48 chondrosarcoma, 49 and rhabdomyosarcoma. 50 Thus, understanding the expression and function of RAGE in the setting of cancer is important in developing novel targeted therapies.
Our findings indicate that RAGE preferentially binds to mutant KRAS to activate the RAF-MEK-ERK and PI3K-AKT pathways, which contributes to HIF1α activation in vitro and in vivo. Oncogenic KRAS mutation is the key regulator in pancreatic cancer maintenance. 4,6 Many receptors for growth factors and cytokines are able to activate KRAS either by direct or indirect protein interaction. 4,6 Similarly, we demonstrated that hypoxia increases the interaction between RAGE and KRAS in pancreatic tumor cells. Active KRAS modulates a number of signaling pathways. The two best-studied pathways that are activated by KRAS are the RAF-MEK-ERK pathway and the PI3K-AKT pathway. The RAF-MEK-ERK pathway ultimately promotes cell proliferation and migration, whereas the PI3K-AKT pathway leads to enhanced protein synthesis and limits apoptosis. The oncogenic KRAS-mediated signaling pathway is required for HIF1α stabilization and activation in colon cancers. 38,51 Our findings show that knockdown of RAGE in pancreatic tumor cells inhibits mutant KRAS activation and subsequent HIF1α stabilization and transactivation. Moreover, knock out of RAGE in mice prevents oncogenic KRAS-driven HIF1α signaling activation.
Autophagy has an important role in both physiological and pathological conditions. However, the role of autophagy in tumorigenesis depends on tumor type and stage. 52,53 An early study indicated that autophagy promotes pancreatic cancer growth. 54 Inhibition of autophagy by chloroquine limits pancreatic cancer growth in in vivo xenograft models. 54 More recently, deletion of certain genes (e.g., ATG5 and ATG7) has been found to be involved in the core mechanism of autophagy, leading to the blocking of the progression of KRAS-driven tumor from precursor lesions into full pancreatic ductal adenocarcinoma in mice. 55,56 This halt in progression is due to the activation of p53, a transcription factor that regulates a large number of genes and guards against genomic instability. 55 However, a recent study shows that ATG5 deletion impairs the progression of pre-malignant PanIN to invasive pancreatic ductal adenocarcinoma in the setting of p53 loss and that p53 status does not impact response to autophagy inhibition. 56 These results highlight the complex relationship between autophagy and p53 in the pancreatic cancer. Our previous study demonstrated that RAGEmediated autophagy is required for STAT3 signaling activation during pancreatic tumorigenesis. 19 Consistent with our results in mice, hypoxia-induced autophagy via the RAGE-KRAS-HIF1α pathway is a survival mechanism in pancreatic tumor cells.
In summary, this study demonstrates that RAGE expression is regulated by a NF-κB-dependent pathway and the upregulated binding of RAGE to mutant KRAS facilitates HIF1α activation in a range of pancreatic tumor cells, as well as a spontaneous pancreatic tumor model in transgenic mice ( Figure 6). Our study extends our previous work by showing that expression of RAGE is involved in inflammation, metabolism, and autophagy changes in pancreatic cancer growth and progression. [18][19][20][21][22] Preferential binding of RAGE to mutant KRAS also suggest that acquisition of this mutation   , and p62 (1 : 500)) in 5% milk. After incubation with peroxidase-conjugated secondary antibodies in 5% milk for 1-2 h at room temperature, protein bands were visualized by enhanced chemiluminescence (Pierce) and then quantified using the Gel-pro Analyzer software (Media Cybernetics, Bethesda, MD, USA). The diagram shows quantitation relative to actin.
Immunofluorescence analysis. Cells were cultured on glass cover-slips and fixed in 4% formaldehyde for 20 min at room temperature before detergent extraction with 0.1% Triton X-100 for 10 min at room temperature. Cover slips were saturated with 2% bovine serum albumin (BSA) in phosphate buffered saline (PBS) for 1 h at room temperature and processed for immunofluorescence with primary antibodies followed by secondary Alexa Fluor 488 or Cy3-conjugated IgG (Invitrogen, San Diego, CA, USA), respectively. Nuclear morphology was visualized with the use of Hoechst 33342 (Invitrogen) stain.
Immunohistochemistry analysis. Paraffin embedded specimens were deparaffinized in xylene, subjected to heat mediated antigen-retrieval in 10 mM sodium citrate (pH 6.0), permeabilized in 0.2% Triton-100 (Sigma) and treated with 100-400 μl blocking solution (Cell Signal Technology). After removal of blocking solution, tissue sections were incubated with 100-400 μl primary antibody as indicated in Figure 5 overnight at 4°C followed by horseradish peroxidaseconjugated secondary antibody (Cell Signal Technology). The signal was amplified with SignalStain Boost Detection Reagent (Cell Signal Technology) and detected using SignalStain DAB reagent (Cell Signal Technology).
Immunoprecipitation analysis. Cells were lysed at 4°C in ice-cold modified radioimmunoprecipitation (RIPA) lysis buffer (Millipore, Billerica, MA, USA) and cell lysates were cleared by centrifugation (12 000 g, 10 min). Concentrations of proteins in the supernatant were determined by bicinchoninic acid assay. Before immunoprecipitation, samples containing equal amount of proteins were pre-cleared  Figure 6 Hypoxia promotes pancreatic cancer initiation, progression, and metastasis through the activation of NF-κB-RAGE-KRAS-HIF1α pathway with protein A or G agarose/sepharose (4°C, 3 h) and subsequently incubated with various irrelevant IgG or specific antibodies in the presence of protein A or G agarose/sepharose beads for 2 h or overnight at 4°C with gentle shaking. Following incubation, agarose/sepharose beads were washed extensively with PBS and the proteins were eluted by boiling in 2 × sodium dodecyl sulfates (SDS) sample buffer before SDS-PAGE electrophoresis.
Caspase-3 activity assay. Caspase-3 activity assay was performed using Caspase-3 Colorimetric Assay Kit from Calbiochem (San Diego, CA, USA) according to the manufacturer's instructions.
Statistical analysis. Data are expressed as mean ± S.D. of three independent experiments. One-way ANOVA was used for comparison among the different groups. When the ANOVA was significant, post hoc testing of differences between groups was performed using an LSD test. A P-valueo0.05 was considered significant.
Cell Death and Disease is an open-access journal published by Nature Publishing Group. This work is licensed under a Creative Commons Attribution 4.0 International Licence. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons licence, users will need to obtain permission from the licence holder to reproduce the material. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0