Abstract 3175: Methylmalonic acid in TME signaling

The importance of metabolic reprogramming in cancer has been long demonstrated by the association of systemic metabolic changes, such as aging, diet and exercise, with cancer outcomes. These systemic shifts, combined with local metabolic alterations within the tumor microenvironment (TME), can all cooperate to foster an environment conducive to cancer progression. In cells derived from primary and metastatic patient tumors, we found that mesenchymal-like cells displayed dysregulated propionate metabolism, leading to increased accumulation and secretion of methylmalonic acid (MMA), a novel aging-induced oncometabolite. This tumor cell-secreted MMA, in addition to increased MMA in the serum of elderly individuals, combine to form high local accumulation of MMA in the TME. We discovered that MMA acts on fibroblasts in the TME, activating them to cancer-associated fibroblasts (CAFs). MMA modifies the cargo of CAF-secreted extracellular vesicles (EVs), which function as a messenger to tumor cells, further promoting epithelial-to-mesenchymal transition, drug resistance, and increased metastasis in vivo. Here, we reveal a novel function of MMA in cancer, demonstrating for the first time that tumor-secreted MMA recruits the tumor microenvironment to drive cancer progression and metastatic reprogramming.

Citation Format: Vivien Low, Zhongchi Li, Ashley Laughney, Wenbing Jin, Noah Dephoure, Valbona Luga, Ethan Earlie, Bobak Parang, Chunjun Guo, Lewis Cantley, John Blenis. Methylmalonic acid in TME signaling [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 3175.

Coagulation factors directly cleave SARS-CoV-2 spike and enhance viral entry

Coagulopathy is a significant aspect of morbidity in COVID-19 patients. The clotting cascade is propagated by a series of proteases, including factor Xa and thrombin. While certain host proteases, including TMPRSS2 and furin, are known to be important for cleavage activation of SARS-CoV-2 spike to promote viral entry in the respiratory tract, other proteases may also contribute. Using biochemical and cell-based assays, we demonstrate that factor Xa and thrombin can also directly cleave SARS-CoV-2 spike, enhancing infection at the stage of viral entry. Coagulation factors increased SARS-CoV-2 infection in human lung organoids. A drug-repurposing screen identified a subset of protease inhibitors that promiscuously inhibited spike cleavage by both transmembrane serine proteases and coagulation factors. The mechanism of the protease inhibitors nafamostat and camostat may extend beyond inhibition of TMPRSS2 to coagulation-induced spike cleavage. Anticoagulation is critical in the management of COVID-19, and early intervention could provide collateral benefit by suppressing SARS-CoV-2 viral entry. We propose a model of positive feedback whereby infection-induced hypercoagulation exacerbates SARS-CoV-2 infectivity.

EXTH-12. INHIBITION OF EPIDERMAL GROWTH FACTOR RECEPTOR AND PLATELET-DERIVED GROWTH FACTOR RECEPTOR-ALPHA EXERTS SYNERGISTIC EFFICACY IN GLIOBLASTOMA

Epidermal growth factor receptor (EGFR) alterations, including amplification and activating mutations, occur in more than half of GBM cases. EGFR is located on Chr. 7, and Chr. 7 gain is one of the earliest events precipitating gliomagenesis. EGFR inhibitors, monoclonal antibodies, vaccines, and CAR-T cells have failed in GBM due to intrinsic heterogeneity and receptor tyrosine kinase (RTK) bypass pathways that mediate therapeutic resistance. New targeted therapeutic approaches to leverage synergistic combinations are desperately needed to improve GBM prognosis. Using the TCGA and other GBM databases, we previously demonstrated that PDGFRA amplification in patients with EGFR-amplified GBM carries significantly worse survival. EGFR and PDGFRA co-expression occur in more than one-third of GBM patients. The PDGFRA ligand PDGFA is also located on Chr. 7, and its expression is significantly increased with Chr. 7 gain and EGFR copy number increase. Therefore, Chr. 7 gain inherently leads to co-activation of both EGFR and PDGFRA signaling. We used patient-derived glioblastoma cells with Chr. 7 gain to test combined inhibition of EGFR and PDGFRA in vitro. We found that combined inhibition of both EGFR and PDGFRA using FDA-approved EGFR-targeted agents (Erlotinib, Gefitinib, Dacomitinib, Neratinib, and Osimertinib) and Crenolanib, respectively, leads to synergistic cytotoxicity in vitro. Inhibition of either EGFR or PDGFRA led to receptor cross-activation, and EGF and PDGF-AA-induced RTK activation was blocked by Neratinib and Crenolanib. Immunoprecipitation experiments and proximity ligation assays demonstrated that combined inhibition prevents EGFR and PDGFRA heterodimerization and pathways of therapeutic resistance. This combined inhibition led to decreased activation of downstream signaling pathways, including PI3K and MAPK. We show that combined inhibition of EGFR and PDGFRA exerts synergistic cytotoxicity in GBM and prevents resistance pathways that emerge during single-agent targeted therapy. These pathways are targetable with FDA-approved agents that could be used in patients with GBM with Chr. 7 gain.

P13.18 Inhibition of epidermal growth factor receptor and platelet-derived growth factor receptor-alpha exerts synergistic efficacy in glioblastoma

BACKGROUND

Despite our understanding of the genetic changes that precipitate gliomagenesis, targeted therapy has failed in glioblastoma (GBM) with median survival not significantly improved over the past two decades. Epidermal growth factor receptor (EGFR) alterations, including amplification and activating mutations, are among the most common genetic changes in GBM, occurring in more than half of cases. EGFR is located on Chr. 7, and Chr. 7 gain is one of the earliest events precipitating gliomagenesis. Various EGFR inhibitors, including tyrosine kinase inhibitors, monoclonal antibodies, vaccines, and CAR-T cells have failed in GBM due to intrinsic heterogeneity and receptor tyrosine kinase bypass pathways that mediate therapeutic resistance. New targeted therapeutic approaches to leverage synergistic combinations are desperately needed to improve GBM prognosis. Using the TCGA and other GBM databases, we have previously demonstrated that the presence of PDGFRAamplification in patients with EGFR-amplified GBM carries significantly worse survival. EGFR and PDGFRA co-expression occur in more than one-third of GBM patients. The PDGFRA ligand PDGFA is also located on Chr. 7, and its expression is significantly increased with Chr. 7 gain and EGFR copy number increase. Therefore, Chr. 7 gain inherently leads to co-activation of both EGFR and PDGFRA signaling pathways.

MATERIALS AND METHODS

We used models of patient-derived glioblastoma cells to test combined inhibition of epidermal growth factor receptor and platelet-derived growth factor receptor-alpha in vitro.

RESULTS

Using patient-derived GBM models with Chr. 7 gain, we found that combined inhibition of both EGFR and PDGFRA using a variety of FDA-approved EGFR-targeted agents (Erlotinib, Gefitinib, Dacomitinib, Neratinib, and Osimertinib) and Crenolanib, respectively, leads to synergistic cytotoxicity in vitro. We found that inhibition of either EGFR or PDGFRA alone led to receptor cross-activation, and EGF and PDGF-AA-induced receptor tyrosine kinase activation was blocked by Neratinib and Crenolanib. Immunoprecipitation experiments and proximity ligation assays demonstrated that combined inhibition prevents EGFR and PDGFRA heterodimerization and pathways of therapeutic resistance. This combined inhibition led to decreased activation of downstream signaling pathways, including phosphatidylinositol 3-kinase and mitogen-activated protein kinase.

CONCLUSIONS

We show that combined inhibition of EGFR and PDGFRA exerts synergistic cytotoxicity in GBM and prevents resistance pathways that emerge during single-agent targeted therapy against these receptor tyrosine kinases. These pathways are targetable with FDA-approved agents that could be used in patients with GBM with Chr. 7 gain.

P13.07 Cysteine induces glioblastoma cytotoxicity through mitochondrial reductive stress

BACKGROUND

Glioblastoma (GBM) remains a poorly treatable disease with high mortality. Tumor metabolism in GBM is a critical mechanism responsible for accelerated growth because of upregulation of glucose, amino acid, and fatty acid utilization. However, therapies targeting GBM metabolism, whether through the use of small-molecule compounds or dietary interventions to limit nutrient sources, have failed in clinical trials. Metabolic bypass is an important mechanism that is often overlooked in GBM trials, since many trials have focused instead on combining anti-metabolic therapy with cytotoxic treatments. The goal of this research is to use a multi-pronged treatment approach with targeted drug and dietary therapy to leverage metabolic susceptibilities in GBM.

MATERIALS AND METHODS

We first interrogated the TCGA database and a cancer metabolite database for alterations in glucose and amino acid signatures in GBM relative to other human cancers and relative to low-grade glioma. We identified the amino acid cysteine as contributing to a novel metabolic susceptibility pathway in GBM. To study the role of cysteine in GBM pathogenesis, we treated patient-derived GBM cells with a variety of FDA-approved cysteine-promoting compounds in vitro, including N-acetylcysteine (NAC). We measured cell proliferation, energy production, mitochondrial metabolism, and reactive oxygen species to study mechanisms of oxidoreductive stress. Results: From our TCGA and cancer metabolite database analyses, we found that GBM exhibits the highest levels of cysteine and methionine pathway gene expression of 32 human cancers and that GBM exhibits high levels of cysteine-related metabolites compared to low-grade gliomas. Cysteine compounds, including NAC, reduce growth of GBM cells, which is exacerbated by glucose deprivation. This growth inhibition is associated with reduced mitochondrial metabolism, manifest by reduction in ATP generation, NADPH/NADP+ ratio, mitochondrial membrane potential, and oxygen consumption rate. Through measurement of mitochondrial hydrogen peroxide, we found that NAC-treated cells exhibit a paradoxical increase in mitochondrial hydrogen peroxide levels, likely due to inhibition of thioreductase and glutathione reductase systems. Through genetic and pharmacological studies, we found that induction of thioredoxin-2 rescues NAC-mediated cytotoxicity and that inhibition of thioreductase and glutathione reductase exacerbates mitochondrial toxicity and reductive stress.

CONCLUSIONS

We show that cysteine compounds reduce cell growth and induce mitochondrial toxicity in GBM through reductive stress. This metabolic phenotype is exacerbated by glucose deprivation. This pathway is targetable with FDA-approved cysteine-promoting compounds and could synergize with glucose-lowering treatments, including the ketogenic diet, for GBM.

DDRE-07. FATTY ACID SYNTHESIS IS REQUIRED FOR BREAST CANCER BRAIN METASTASIS

Brain metastases are refractory to therapies that otherwise control systemic disease in patients with human epidermal growth factor receptor 2 (HER2+) breast cancer, and the unique brain microenvironment contributes to this therapy resistance. Nutrient availability can vary across tissues, therefore metabolic adaptations required for breast cancer growth in the brain microenvironment may also introduce liabilities that can be exploited for therapy. Here, we assessed how metabolism differs between breast tumors growing in the brain versus extracranial sites and found that fatty acid synthesis is elevated in breast tumors growing in the brain. We determine that this phenotype is an adaptation to decreased lipid availability in the brain relative to other tissues, which results in a site-specific dependency on fatty acid synthesis for breast tumors growing at this site. Genetic or pharmacological inhibition of fatty acid synthase (FASN) reduces HER2+ breast tumor growth in the brain, demonstrating that differences in nutrient availability across metastatic sites can result in targetable metabolic dependencies.

FSMP-10. CYSTEINE INDUCES CYTOTOXICITY IN GLIOBLASTOMA THROUGH MITOCHONDRIAL HYDROGEN PEROXIDE PRODUCTION

Glioblastoma (GBM) is a poorly treatable disease with high mortality. Tumor metabolism in GBM is a critical mechanism responsible for growth because of upregulation of glucose, amino acid, and fatty acid utilization. However, little is known about the specific metabolic alterations in GBM that are targetable with FDA-approved compounds. To investigate metabolic signatures unique to GBM, we interrogated the TCGA and a cancer metabolite database for alterations in glucose and amino acid signatures in GBM relative to other human cancers and relative to low-grade glioma. From these analyses, we found that GBM exhibits the highest levels of cysteine and methionine pathway gene expression of 32 human cancers and that GBM exhibits high levels of cysteine metabolites compared to low-grade gliomas. To study the role of cysteine in GBM pathogenesis, we treated patient-derived GBM cells with FDA-approved cyst(e)ine-promoting compounds in vitro, including N-acetylcysteine (NAC) and the cephalosporin antibiotic, Ceftriaxone (CTX), which induces cystine import through system Xc transporter upregulation. Cysteine-promoting compounds, including NAC and CTX, inhibit growth of GBM cells, which is exacerbated by glucose deprivation. This growth inhibition is associated with reduced mitochondrial metabolism, manifest by reduction in ATP, NADPH/NADP+ ratio, mitochondrial membrane potential, and oxygen consumption rate. Mechanistic experiments revealed that cysteine compounds induce a rapid increase in the rate of H2O2 production in isolated GBM mitochondria, an effect blocked by the H2O2 scavenger, catalase. Such findings are consistent with reductive stress, a ROS-producing process whereby excess mitochondrial reducing equivalents prevent electron transfer to oxidized electron acceptors, inducing O2 reduction to H2O2. We show that cysteine-promoting compounds reduce cell growth and induce rapid mitochondrial toxicity in GBM, which may be due to reductive stress. This pathway is targetable with FDA-approved cysteine-promoting compounds and could synergize with glucose-lowering treatments, including the ketogenic diet, for GBM.

DDRE-22. TARGETING SERINE SYNTHESIS IN BRAIN METASTASIS

The brain environment is low in amino acids, including serine and glycine, both of which are important for tumor growth as they are precursors of proteins and nucleotide bases. How tumor cells overcome these conditions to proliferate and survive in the brain is incompletely understood. Here, we show that 3-phosphoglycerate dehydrogenase (PHGDH), which catalyzes the first and rate-limiting step of glucose-derived serine synthesis, enables brain metastasis in multiple human types and in preclinical models. Genetic suppression and small molecule inhibition of PHGDH attenuated brain metastasis, but not extra cranial tumors, and improved the overall survival of mice bearing brain metastasis. These results demonstrate that the tumor nutrient microenvironment determines tumor cell sensitivity to loss of serine synthesis pathway activity and raise the possibility that serine synthesis inhibitors may be useful in the treatment of brain metastases.

TAMI-38. CYSTEINE-PROMOTING COMPOUNDS INDUCE MITOCHONDRIAL TOXICITY IN GLIOBLASTOMA THROUGH ALTERED PYRUVATE AND SERINE METABOLISM

Glioblastoma (GBM) remains a poorly treatable disease with high mortality. Tumor metabolism in GBM is a critical mechanism responsible for accelerated growth because of upregulation of glucose, amino acid, and fatty acid utilization. However, little is known about the metabolic alterations that are specific to GBM and that are targetable with FDA-approved compounds. To investigate tumor metabolism signatures unique to GBM, we interrogated the TCGA and a cancer metabolite database for alterations in glucose and amino acid signatures in GBM relative to other human cancers and relative to low-grade glioma. From these analyses, we found that GBM exhibits the highest levels of cysteine and methionine pathway gene expression of 32 human cancers and that GBM exhibits high levels of cysteine-related metabolites compared to low-grade gliomas. To study the role of cysteine in GBM pathogenesis, we treated patient-derived GBM cells with a variety of FDA-approved cyst(e)ine-promoting compounds in vitro, including N-acetylcysteine (NAC) and the cephalosporin antibiotic, Ceftriaxone (CTX), which induces cystine import through System Xc transporter upregulation. Cysteine-promoting compounds, including NAC and CTX, inhibit growth of GBM cells, which is exacerbated by glucose deprivation. This growth inhibition is associated with reduced mitochondrial metabolism, manifest by reduction in ATP, NADPH/NADP+ ratio, mitochondrial membrane potential, and oxygen consumption rate. Metabolic tracing experiments with 13C6-glucose demonstrate that L-serine is rapidly depleted in GBM cells upon treatment with NAC and CTX, and exogenous serine rescues NAC- and CTX-mediated cell growth inhibition. In addition, these compounds reduce GBM mitochondrial pyruvate transport. We show that cysteine-promoting compounds reduce cell growth and induce mitochondrial toxicity in GBM, which may be due to rapid serine depletion and reduced mitochondrial pyruvate transport. This metabolic phenotype is exacerbated by glucose deprivation. This pathway is targetable with FDA-approved cysteine-promoting compounds and could synergize with glucose-lowering treatments, including the ketogenic diet, for GBM.

Abstract B24: Phase II trial of AKT inhibitor MK-2206 in patients with advanced breast cancer who have tumors with PIK3CA or AKT mutations, and/or PTEN loss/PTEN mutation

Background: The PI3K/AKT pathway is activated through PIK3CA or AKT1 mutations and PTEN loss in breast cancer. We conducted a phase II trial with an allosteric AKT Inhibitor MK-2206 in patients with advanced breast cancer who had tumors with PIK3CA/AKT1 mutations and/or PTEN loss/mutation.

Methods: The primary endpoint was objective response rate (ORR). Secondary endpoints were 6-month progression-free survival (6m PFS), predictive and pharmacodynamic markers, safety and tolerability. Patients had pretreatment and on-treatment biopsies as well as collection of peripheral blood mononuclear cells (PBMC) and platelet-rich plasma (PRP). Next-generation sequencing, immunohistochemistry and reverse phase protein arrays (RPPA) were performed.

Results: Twenty-seven patients received MK-2206. Eighteen patients were enrolled into the PIK3CA/AKT1 mutation arm (cohort A): 13 had PIK3CA mutations, four had AKT1 mutations and one had a PIK3CA mutation as well as PTEN loss. ORR and 6m PFS were both 5.6% (1/18), with one patient with HR+ breast cancer and a PIK3CA E542K mutation experiencing a partial response (on treatment for 36 weeks). Nine patients were enrolled on the PTEN loss/mutation arm (cohort B). ORR was 0% and 6m PFS was 11% (1/9), observed in a patient with triple-negative breast cancer and PTEN loss. The study was stopped early due to futility. The most common adverse events were fatigue (48%) and rash (44%). On pretreatment biopsy, PIK3CA and AKT1 mutation status was concordant with archival tissue testing. However, two patients with PTEN loss based on archival testing had PTEN expression on the pretreatment biopsy. MK-2206 treatment was associated with a significant decline in pAKT S473 and pAKT T308 and PI3K activation score in PBMC and PRPs, but not in tumor biopsies. By IHC, there was no significant decrease in median pAKT S473 or Ki-67 staining, but a drop was observed in both responders.

Conclusions: MK-2206 monotherapy had limited clinical activity in advanced breast cancer patients selected for PIK3CA/AKT1 or PTEN mutations or PTEN loss. This may, in part, be due to inadequate target inhibition at tolerable doses in heavily pretreated patients with pathway activation, as well as tumor heterogeneity and evolution in markers such as PTEN conferring challenges in patient selection.

Citation Format: Yan Xing, Nancy U. Lin, Mathew Mauer, Huiqin Chen, Armeen Mahvash, Aysegul Sahin, Argun Akcakanat, Yisheng Li, Vandana Abraham, Jennifer Litton, Mariana Chavez-McGregor, Vicente Valero, Sarina A. Piha-Paul, David Hong, Kim-Anh Do, Emily Tarco, Dianna Riall, Karina Agda Eterovic, Lewis Cantley, Gordon B. Mills, L. Austin Doyle, Eric Winer, Gabriel Hortobagyi, Ana Maria Gonzalez-Angulo, Funda Meric-Bernstam. Phase II trial of AKT inhibitor MK-2206 in patients with advanced breast cancer who have tumors with PIK3CA or AKT mutations, and/or PTEN loss/PTEN mutation [abstract]. In: Proceedings of the AACR Special Conference on Targeting PI3K/mTOR Signaling; 2018 Nov 30-Dec 8; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Res 2020;18(10_Suppl):Abstract nr B24.

Abstract GS6-06: A neoadjuvant trial with letrozole identifies PRR11 in the 17q23 amplicon as a mechanism of resistance to endocrine therapy in ER-positive breast cancer

Although the 17q23 amplicon has been associated with luminal B breast cancer (BC) and high risk of recurrence, a specific gene or genes in this region that would be causal to endocrine resistance have not yet been uncovered. We performed whole transcriptome analysis on RNA extracted from 58 estrogen receptor (ER)+ BCs treated with neoadjuvant letrozole for median 7.2 months. PRR11 (Proline rich 11), located in 17q23, was upregulated in non-responding tumors as defined by relapse after a median follow up of 5 years and/or a preoperative endocrine prognostic index (PEPI) ≥4. Differential gene expression analysis between tumors expressing low vs high PRR11 mRNA showed that BC signatures associated with proliferation, IGF-1 and PI3K signaling were enriched in tumors with high PRR11 expression.

Rate of PRR11 amplification is 15.2% in the Metastatic Breast Cancer project, but 9.5% and 9.4% in METABRIC and The Cancer Genome Atlas (TCGA), respectively. Gene Set Enrichment Analysis revealed an enrichment of hallmark gene sets associated with proliferation in PRR11-amplified ER+ BCs in METABRIC and TCGA. Integrated analysis of gene expression with on-treatment Ki67 levels from three independent studies with operable ER+ BCs treated with neoadjuvant aromatase inhibitor (ACOSOG-Z1031, NCT00651976, Llombart-Cussac et al.) showed that PRR11 was the only gene in 17q23 with a significant correlation with a high Ki67 levels across all studies.

PRR11 knockdown inhibited E2-independent growth of HCC1428 LTED (long-term estrogen deprived) and MCF7 LTED cells in culture and MCF7 xenografts. PRR11 siRNA also inhibited growth of fulvestrant-resistant and tamoxifen-resistant MCF7 cells. Conversely, PRR11 transduction induced MDA-MB-134VI cell growth under estrogen-depleted conditions. Using a PCR array with 84-cell cycle genes, we identified SKP2, CDKN1A, CCNB2, CCNA2, CKS2 and CCNB1 as genes downregulated by PRR11 knockdown. Except for SKP2 and CDKN1A, expression of all those genes was elevated in PRR11-amplifiedER+ BCs in TCGA and METABRIC.

Suggesting a link to activation of PI3K signaling, we found the proline-rich motif of PRR11 associates with the SH3 domain of the p85 regulatory subunit of PI3K. We hypothesized that this association suppresses p85 homodimer formation, thus facilitating binding of PI3Kα (p110α)-p85 dimers to IRS1, retention of p110α at the plasma membrane and, hence, activation of PI3K/AKT. To test this, we co-transfected HEK293T cells with HA-p85 and FLAG-p85. Forced expression of PRR11 reduced HA-p85 and FLAG-p85 homodimers as shown by HA and FLAG pulldowns followed by FLAG and HA immunoblots, respectively. PRR11 overexpression enhanced insulin-stimulated association of IRS1 to p110α and activation of AKT. PRR11 knockdown reduced insulin/IGF-1/2-stimulated p-AKT. In METABRIC and TCGA, PRR11 amplification and PIK3CA mutations are exclusive of each other, suggesting these alterations would be functionally linked with the same pathway.

Connectivity map analysis with the list of genes significantly overexpressed in ER+/PRR11-amplified BCs predicted PI3K inhibitors as perturbations that suppress such gene list. In the MGH/Sanger dataset, PRR11-amplified BC cell lines displayed significantly higher sensitivity to the pan-PI3K inhibitor pictilisib compared to cell lines without PRR11 amplification. Finally, inhibition of PI3Kα by siRNA or alpelisib abrogated E2-independent growth and insulin-stimulated growth of PRR11-transduced MDA-MB-134VI and MCF10A cells, respectively, suggesting p110α is required for the growth promoting effects of PRR11.

These data suggest that 1) PRR11 is a mediator of resistance to antiestrogens via amplification of PI3K/AKT signaling, and 2) PI3Kα is a potential therapeutic target in ER+ BCs harboring PRR11 amplification.

Citation Format: Kyung-min Lee, Angel Guerrero-Zotano, Ariella Hanker, Alberto Servetto, Dhivya Sudhan, Luigi Formisano, Valerie Jansen, Paula González-Ericsson, Melinda Sanders, Thomas Stricker, Lewis Cantley, Carlos Arteaga. A neoadjuvant trial with letrozole identifies PRR11 in the 17q23 amplicon as a mechanism of resistance to endocrine therapy in ER-positive breast cancer [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr GS6-06.

EXTH-36. PI3K INHIBITION IN CONJUNCTION WITH THE KETOGENIC DIET REDUCES GROWTH AND NEUROINFLAMMATION IN PEDIATRIC HIGH-GRADE GLIOMA

Pediatric high-grade glioma remains a poorly treatable disease with high mortality. Targeted therapies have gained interest in this disease, but efficacy is limited by therapeutic resistance, often because of tumor heterogeneity. Phosphoinositide 3-kinase (PI3K) inhibitors represent a strong drug class for pediatric glioma, but their use is associated with insulin feedback that reactivates the PI3K pathway and drives therapeutic resistance. Here, we target insulin feedback that is the primary mechanism of PI3K inhibitor-related therapeutic resistance in pediatric high-grade glioma using the ketogenic diet. We treated patient-derived pediatric high-grade glioma stem cells with vehicle or the pan-PI3K inhibitor, BKM-120, in conjunction with phenformin to decrease glucose utilization. These cells exhibited 65% less proliferation when exposed to BKM-120 and phenformin. We treated NOD scid gamma (NSG) mice containing patient-derived pediatric high-grade glioma xenografts with vehicle or BKM-120 on a regular or ketogenic diet to determine whether reducing insulin feedback increases BKM-120 efficacy. Mice with intracranial glioma xenografts survived longer when treated with BKM-120 on the ketogenic diet than with BKM-120 or the ketogenic diet alone. We measured pro-inflammatory cytokines in glioma cells treated with BKM-120 and phenformin in comparison to vehicle-treated cells to determine their effect on neuro-inflammation. We also applied conditioned medium from glioma cells treated with BKM-120 and phenformin to cortical neurons to measure oxidative stress. We found that phenformin reduced the production of pro-inflammatory cytokines, including TNF-alpha, IFN-gamma, IL-1beta, and IL-6, by BKM-120-treated glioma cells. Cortical neurons treated with conditioned medium from BKM-120- and phenformin-treated glioma cells exhibited less oxidative stress than those treated with BKM-120 alone. Our results demonstrate that lowering glucose utilization and insulin feedback increases efficacy of PI3K inhibition and decreases neuro-inflammation. By using the ketogenic diet to reduce systemic glucose levels, this strategy may enhance efficacy and reduce morbidity of PI3K inhibitors in this population.

P11.63 PI3K inhibition in conjunction with the ketogenic diet reduces growth and neuroinflammation in pediatric high-grade glioma

Background: Pediatric high-grade glioma remains a poorly treatable disease with high mortality.Therapeutic advances have lagged behind that of adult glioblastoma, due to small patient numbers, inappropriate generalization from adult tumor types, and unique biology. The use of targeted therapy has recently gained interest in this disease, but efficacy is limited by therapeutic resistance, often as a result of tumor heterogeneity. In the case of phosphatidylinositol 3-kinase (PI3K) inhibition, clinically relevant PI3K inhibitors represent a strong class of drugs for pediatric high-grade glioma, but their use is associated with insulin feedback that reactivates the PI3K pathway and drives therapeutic resistance. Here, we target insulin feedback that is the primary mechanism of PI3K inhibitor-related therapeutic resistance in pediatric high-grade glioma using the ketogenic diet.

Materials and Methods: Patient-derived pediatric high-grade glioma stem cells were treated with vehicle or the pan-PI3K inhibitor, BKM-120, with glucose deprivation or phenformin to decrease glucose utilization. NSG mice containing patient-derived pediatric high-grade glioma xenografts were treated with vehicle or BKM-120 on a regular or ketogenic diet to determine whether reducing insulin feedback increases BKM-120 efficacy. To determine the effect of glioma cells on neuro-inflammation, we measured pro-inflammatory cytokines in glioma cells treated with BKM-120 and phenformin in comparison to vehicle-treated cells. We then applied conditioned medium from glioma cells treated with BKM-120 and phenformin to cortical neuronal cultures to measure oxidative stress and neuro-inflammation. Results: Pediatric high-grade glioma cells exhibited increased toxicity when exposed to BKM-120 with glucose deprivation or phenformin. Furthermore, mice with intracranial high-grade glioma xenografts survived longer when treated with BKM-120 on the ketogenic diet than with BKM-120 or the ketogenic diet alone. Phenformin reduced the production of pro-inflammatory cytokines by BKM-120-treated glioma cells. Cortical neurons treated with conditioned medium from BKM-120- and phenformin-treated glioma cells exhibited less oxidative stress than those treated with BKM-120 alone. Our results demonstrate that strategies to lower glucose utilization and insulin feedback increase efficacy of BKM-120 and decrease neuro-inflammation.

Conclusions: We show that strategies to lower glucose utilization and insulin feedback increase efficacy of BKM-120. Furthermore, reducing insulin feedback decreases the production of pro-inflammatory cytokines in tumor cells and reduces oxidative stress in neurons treated with conditioned medium from BKM-120-treated glioma cells. By using the ketogenic diet to reduce glucose levels, this strategy may enhance efficacy of PI3K inhibitors in this patient population.

Abstract 2062: Evaluating the effect of tumor necrosis on glycolytic flux and cell survival in glioblastoma

Glioblastoma continues to rank as one of the most lethal primary human tumors. The degree of necrosis in glioblastoma remains one of the most powerful predictors of poor prognosis, but the relationship between necrosis and poor prognosis is not known. It is unclear whether endogenous tumor necrosis is merely an indicator of an aggressive phenotype or whether tumor necrosis itself propagates the aggressive phenotype. Determining the role of necrosis in mediating the malignant potential of glioblastoma therefore represents an important aspect of therapy. We describe a model of patient-derived glioblastoma spheroids expressing a genetically encoded fluorescent NADH/NAD+ biosensor. This biosensor reports glycolytic flux and can be imaged with fluorescent microscopy. We imaged spheroids using multiphoton microscopy and differentiated cells by high versus low NADH/NAD+ levels. FACS analysis of dissociated spheroids followed by RT-PCR analysis demonstrated increased expression of genes known to be upregulated in peri-necrotic regions, including procollagen-lysine,2-oxoglutarate 5-dioxygenase 2 (PLOD2), activating transcription factor 5 (ATF5), and protein O-fucosyltransferase 2 (POFUT2). These cells also had increased expression of hypoxia-responsive genes, including vascular endothelial growth factor (VEGF) and jumonji domain-containing protein 6 (JMJD6). We created an intracranial glioblastoma patient-derived xenograft model, with cells expressing this NADH/NAD+ biosensor. When imaging these tumors in vivo, we were able to demonstrate that cells with high NADH/NAD+ levels are more likely to survive within the peri-necrotic environment. These findings indicate that increased glycolytic flux facilitates survival from expanding regions of necrosis. Future experiments will focus on biopsy of cellular populations with high versus low NADH/NAD+ levels within the peri-necrotic niche. We will then perform RNAseq and metabolomics analysis to determine the genetic and metabolic adaptations that facilitate survival within this microenvironment. This model to study cell survival within the peri-necrotic region will improve awareness of necrosis-mediated growth proliferation and may lead to treatments targeted to tumor necrosis that could improve patient prognosis.

Citation Format: Evan Noch, Isaiah Yim, Lewis Cantley. Evaluating the effect of tumor necrosis on glycolytic flux and cell survival in glioblastoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 2062.

SAT-LB057 High-Fructose Corn Syrup Enhances Intestinal Tumor Growth in Mice

An increasing number of studies have suggested a potential link between sugar consumption and cancer. However, a direct relationship between the two remains controversial. Here, we treated APC-mutant mice daily with a modest amount of high-fructose corn syrup (HFCS) via oral gavage to mimic sugar-sweetened beverage (SSB) consumption and evaluated the effects of SSBs on intestinal tumorigenesis. We observed a dramatic increase in tumor size and tumor grade in HFCS-treated mice despite no signs of obesity or metabolic syndrome. A bolus of HFCS spikes the level of fructose and glucose in the intestinal lumen and serum, respectively, allowing tumors to take up both sugars efficiently. Ketohexokinase (KHK) in tumors rapidly converts fructose to fructose-1-phosphate (F1P), lowering cytosolic ATP in the process. Consequently, this leads to the activation of glycolysis as ATP depletion relieves the allosteric inhibition of phosphofructokinase (PFK), the most critical regulatory enzyme in glycolysis. This accelerated glycolysis contributes to the increased de novo lipogenesis pathway that enhances tumor growth in HFCS-treated mice. Importantly, the deletion of KHK or fatty acid synthase (FASN) in APC-mutant mice abolishes the HFCS-induced metabolic changes and tumor growth. Overall, our results provide a mechanistic rationale for restricting SSB consumption to decrease the rate of growth of early stage colorectal cancers. Unless otherwise noted, all abstracts presented at ENDO are embargoed until the date and time of presentation. For oral presentations, the abstracts are embargoed until the session begins. Abstracts presented at a news conference are embargoed until the date and time of the news conference. The Endocrine Society reserves the right to lift the embargo on specific abstracts that are selected for promotion prior to or during ENDO.

Abstract TS2-1: The complexity of targeting PI3K for cancer therapy

Phosphoinositide 3-Kinase (PI3K) is activated by insulin and other growth factors to mediate cell growth. The PI3K enzyme encoded by the PIK3CA gene is one of the most frequently mutated oncogenes in human cancer. This same enzyme mediates insulin responses in liver, muscle, fat and other tissues. The most common mutations in this gene enhance the ability of PI3K to be activated by insulin. Since breast cancers typically express relatively high levels of the insulin receptor, high serum insulin levels are likely to enhance activation of PI3K in breast cancers. PI3K inhibitors that target PIK3CA have the expected physiological effect of raising serum glucose and insulin levels and the elevated insulin is likely to compromise the effectiveness of therapy. Data will be presented showing that in mouse models of human breast cancers and other cancers, dietary interventions and pharmaceutical interventions that limit elevation of serum insulin during PI3K therapy improve responses to PI3K inhibitors.

Citation Format: Cantley L. The complexity of targeting PI3K for cancer therapy [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr TS2-1.

Abstract A078: Towards understanding noncanonical phosphatidylinositol kinases in the maintenance of prostate metabolism

An estimated 1 in 7 men will develop prostate cancer (PCa) with many progressing to advanced castrate-resistant disease. Unlike other tissue types, normal prostate cell growth and development is heavily dependent on the androgen receptor (AR) signaling pathway. While the introduction of novel AR antagonists for clinical treatment has improved outcomes, most castration-resistant prostate cancer (CRPC) patients ultimately develop resistance to these therapies. A need exists to better understand the mechanisms that control the transition of prostate cells from a hormone-dependent to castrate-resistant state.

Androgens strongly influence the metabolic state of PCa cells to favor sustained cellular growth. We hypothesize there are effectors working in conjunction with AR to coordinate alterations to androgen-dependent metabolism that are linchpins in the orchestration of the transition to CRPC. Leading candidates are members of phosphoinositol (PI) pathways, which have a high frequency of alteration in PCa (i.e phosphoinositide 3-kinase (PI3K)). Herein we explore a family of poorly understood lipid kinases called the type II phosphatidylinositol-5-phosphate 4-kinases (PI5P4Ks) and predict them to be critical regulators of cancer cell survival. PI5P4Ks are druggable targets that act by phosphorylating the lipid phosphatidylinositol-5-phosphate (PI 5-P) at the 4 position of the inositol ring to generate phosphatidylinositol-4,5-bisphosphate (PI-4,5-P2; PIP2). We implicate the three PI5P4K isoforms (PI5P4Kα, PI5P4Kβ, and PI5P4Kγ) encoded by the genes PIP4K2A, B, and C, to be important regulators of cancer metabolism that play a role in the maintenance of prostate biology and oncogenesis.

Analysis of transcript data revealed expression of PIP4K2A, B, and C in primary PCa patient samples, which was correlated with an AR activation gene signature and hotspot tumor suppressor deletion. As well, isoform expression was assessed for differential expression in relation to an integrated neuroendocrine prostate cancer mRNA score (TCGA; n=333). PI5P4Kα and PI5P4Kβ protein was detected in primary and advanced prostate cancer using optimized antibodies of patient tissue TMAs (n= 72). Using in vitro LNCaP cell models, siRNA knockdown systems were tested to evaluate the molecular consequence of targeting PIP4K2A and PIP4K2B in androgen-dependent systems. Stable knockdown using fluorescently labeled lentiviral shRNA constructs significantly reduced proliferation of shPIP4K2 treated cells. As well, we have produced a prostate-specific PI5P4K knockout mouse model by expressing probasin-driven Cre in a homozygous 129/SvEv Pip4k2aflx/flx murine strain. Finally, implementation of a discovery-based metabolomic platform (Metabolon HD4) was used to profile the overall shift in metabolite species that results from downregulating the expression of PIP4K2A in androgen-dependent cell models.

In summary, we have developed novel insights into the role of a family of noncanonical PI kinases in prostate biology. There are a growing number of PI3K/AKT inhibitors being tested in combination with androgen deprivation therapy in clinical trials, but there is still almost nothing known about the potential crosstalk of the greater PI kinase network. These data convincingly implicate a fundamental role for PI5P4Ks in PCa androgen signaling and metabolism, as well as lay the foundation of phenotypic understanding of what PI5P4K is responsible for in the prostate.

Citation Format: Joanna Triscott, Matteo Benelli, Verena Sailer, Davide Prandi, Brooke Emerling, Francesca Demichelis, Lewis Cantley, Mark A. Rubin. Towards understanding noncanonical phosphatidylinositol kinases in the maintenance of prostate metabolism [abstract]. In: Proceedings of the AACR Special Conference: Prostate Cancer: Advances in Basic, Translational, and Clinical Research; 2017 Dec 2-5; Orlando, Florida. Philadelphia (PA): AACR; Cancer Res 2018;78(16 Suppl):Abstract nr A078.

Abstract IA08: Reprogramming human cancer cells in CDK4/6 inhibitor therapy

CDK4 and CDK6 drive cell cycle progression through early G1 and are frequently deregulated in human cancer. Selective inhibition of CDK4/6 with palbociclib (PD 0332991) has demonstrated exciting clinical efficacy in diverse human cancers. However, the mechanism that discriminates sensitivity from resistance to targeting CDK4/6 remains obscure.

Mantle cell lymphoma (MCL) is an incurable non-Hodgkin's lymphoma where deregulated CDK4 activity and cyclin D1 expression underlies unrestrained proliferation and disease progression. In a phase I clinical trial in recurrent MCL, we demonstrated that palbociclib alone produced a durable clinical response with a favorable toxicity profile. To investigate the mechanism for targeting CDK4/6, we have now shown in primary cancer cells that 1) inhibition of CDK4/6 leads to early G1 arrest that requires Rb, the CDK4/6 substrate; 2) prolonged early G1 arrest (pG1) sensitizes Rb-proficient cancer cells to killing by diverse clinically-relevant agents; and 3) pG1 sensitization stems from continuous expression of genes scheduled for early G1 only. This leads to an imbalance in genes expression, which is exacerbated in synchronous progression to S phase (pG1-S) after palbociclib withdrawal, due to incomplete restoration of cell cycle-coupled gene expression.

In a hypothesis-driven phase Ib clinical trial, targeting CDK4 with palbociclib in sequential combination with the proteasome inhibitor bortezoimb was well tolerated (n=16). It exhibited a durable palbociclib dose-dependent clinical activity, including one complete remission for over 3 years with only one progression at the optimal dose combination (n=6). Longitudinal integrative whole transcriptome and whole exome sequencing of tumor cells isolated from serial lymph node biopsies demonstrated that palbociclib initially induced pG1 in MCL cells of all patients, regardless of copy number variation or mutation (ATM, p53). Cell cycle control by palbociclib is thus initially intact in MCL, but is insufficient to predict the clinical response.

As predicted, only genes programmed for early G1 were expressed pG1, concurrent PI3K inactivation in primary MCL cells. However, <1% of the 1400 genes suppressed (not programmed) in pG1 in clinically-responding patients were activated in non-responding patients. These genes were critical for redox homeostasis, suggesting that PI3K inactivation and redox stress mediates pG1 sensitization.

This study represents the first investigation of genes that discriminate sensitivity from resistance in targeting CDK4/CDK6 in human cancer, through integrative longitudinal analysis of whole exome and whole transcriptome sequencing in concert with protein expression analysis and functional studies. Selective inhibition of CDK4 induces pG1 in all MCL patients, which apparently reprograms MCL for clinical response to bortezomib through PI3K inactivation and suppression of genes for redox homeostasis.

Citation Format: Selina Chen-Kiang, Maurizio Di Liberto, Priyanka Vijay, David Chiron, Xiangao Huang, Scott Ely, Olivier Elemento, Christopher Mason, Lewis Cantley, John P. Leonard, Peter Martin. Reprogramming human cancer cells in CDK4/6 inhibitor therapy. [abstract]. In: Proceedings of the AACR Precision Medicine Series: Cancer Cell Cycle - Tumor Progression and Therapeutic Response; Feb 28-Mar 2, 2016; Orlando, FL. Philadelphia (PA): AACR; Mol Cancer Res 2016;14(11_Suppl):Abstract nr IA08.

Abstract 1625: CUB domain-containing protein 1 (CDCP1) promotes metastatic phenotype in pancreatic ductal adenocarcinoma

CUB domain-containing protein 1 (CDCP1) is a transmembrane glycoprotein that is overexpressed and highly tyrosine phosphorylated in multiple cancer types including pancreatic ductal adenocarcinoma (PDAC), a highly lethal disease. CDCP1 has been implicated in the promotion of metastases, namely cancer cell migration, invasion, and resistance to anoikis. In addition, prior studies in other cancer cell lines have shown that hypoxia inducible factors (HIF-1α and HIF-2α), which are key regulators of metastasis, induce both expression of CDCP1 and phosphorylation at tyrosine 734. The role of CDCP1 and the mechanisms that regulate CDCP1 still need further elucidation in PDAC.

In this study, we demonstrated that CDCP1 is highly expressed and tyrosine phosphorylated in PDAC cell lines MiaPaCa2, PaTu89902, and Panc-1 cell lines compared to the non-tumorigenic pancreatic ductal cell line HPNE. shRNA knockdown of CDCP1 showed a reduction in the phosphorylation of Src and PKC-delta. To determine whether HIF regulates CDCP1 in PDAC, we used PDAC cell lines that stably express shRNAs targeting HIF-1α and HIF-2α. Hypoxia induced an increase in CDCP1 expression in a HIF-1α-dependent fashion. Additionally, CDCP1 knockdown in PaTu8902, a highly metastatic PDAC line, impaired cancer cell invasion by transwell invasion assay but did not show resistance to anoikis by soft agar assay. To validate our data in vivo, we performed intra-pancreatic injection of shCDCP1 PaTu8902 cells and scramble control transfected with GFP/luciferase into nude mice. Live imaging demonstrated that 2 weeks after initial injection resulted in decreased tumor establishment in mice injected with the shCDCP1 cell line compared to control. We propose that CDCP1 expression and tyrosine phosphorylation in PDAC is mediated by hypoxia inducible factors and that CDCP1 confers invasive properties in vitro and may play a role in initial tumor establishment in vivo. Based on these results, CDCP1 is a promising novel target for PDAC.

Citation Format: Sunnie Kim, Brooke Emerling, Benjamin Hopkins, Ryan Loughran, Salva Naranjos, Lewis Cantley. CUB domain-containing protein 1 (CDCP1) promotes metastatic phenotype in pancreatic ductal adenocarcinoma. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 1625.

Abstract C112: Cabozantinib eradicates advanced murine prostate cancer by activating anti-tumor innate immunity

Several kinase inhibitors targeting aberrant signaling pathways in tumor cells have been deployed in cancer therapy. However, their impact on the tumor immune microenvironment remains poorly understood. The tyrosine kinase inhibitor cabozantinib showed striking responses in early phase clinical trials, particularly in cancer patients with bone metastases. Here we show that cabozantinib rapidly eradicates invasive, poorly-differentiated PTEN/p53 deficient murine prostate cancer. This was associated with increased neutrophil chemotactic factor expression, including CXCL12 and HMGB1 production by tumor cells, and robust infiltration of neutrophils into the tumor. Critically, cabozantinib-induced tumor clearance in mice was abolished by antibody-mediated granulocyte depletion or HMGB1 neutralization or blockade of neutrophil chemotaxis with the CXCR4 inhibitor, plerixafor. Collectively, these results demonstrate that cabozantinib triggers neutrophil-mediated anti-tumor innate immune response that results in tumor clearance.

(Manuscript submitted to Science)

Citation Format: Akash Patnaik, Kenneth Swanson, Sabina Signoretti, Huihui Ye, Eva Csizmadia, Jesse Novak, Marina Gehring, Katja Helenius, Athalia Pyzer, Laleh Montaser, Lily Wang, Olivier Elemento, Elena Levantini, John Clohessy, John Asara, Kathleen Kelly, Pier Paolo Pandolfi, Jacalyn Rosenblatt, David Avigan, Steven Balk, Lewis Cantley. Cabozantinib eradicates advanced murine prostate cancer by activating anti-tumor innate immunity. [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2015 Nov 5-9; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2015;14(12 Suppl 2):Abstract nr C112.

Abstract CT232: SU2C Phase Ib study of the PI3K-alpha inhibitor BYL719 (alpelisib) with letrozole in ER+/HER2-metastatic breast cancer

Background: Mutations in PIK3CA, the gene encoding the p110α subunit of PI3K, have been associated with antiestrogen resistance in ER+ BC. In general, antiestrogen-resistant cancers retain ER and responsiveness to estradiol. This suggests that treatment of ER+ BC with PI3K inhibitors should also include antiestrogens.

Methods: We conducted a phase Ib 3+3 dose escalation trial of letrozole (2.5 mg/day) with the PI3Kα inhibitor BYL719 in post-menopausal patients (pts) with ER+/HER2- metastatic breast cancer (MBC) refractory to previous endocrine therapy. BYL719 doses ranged from 300-400 mg/daily. Treatment continued until unacceptable toxicity or disease progression. Disease was assessed every 2 months. PIK3CA mutation status was determined by SNaPshot or targeted next generation sequencing (NGS) in all patients’ tumors.

Results: Twenty-six pts were accrued; all patients were refractory to endocrine therapies in the metastatic setting, 18 had prior aromatase inhibitor (AI) therapy in their first or second-line MBC treatment. Median age was 53 years (31-72); 76% of pts had bone and 61% had visceral metastases; 50% of pts had a somatic PIK3CA hotspot mutation. Toxicities are summarized in Table 1. Rash was the dose limiting toxicity* at 350 mg/day. Of the 5 pts with a partial response, 3 (60%) had a PI3KCA mutation, 6/9 pts (67%) that were on treatment for ≥6 months; 3/6 pts (50%) that were on treatment for ≥12 months and 4/6 pts (67%) that are still on study (range 11 - 19 months) had a PI3KCA mutation.

Discussion: The combination of letrozole/ BYL719 is safe and tolerable in pts with AI-refractory ER+/HER2-negative MBC. The MTD and recommended dose for phase II trials of BYL719 in combination with letrozole was 300 mg/day. Rash and hyperglycemia were observed at this dose, suggesting drug-induced inhibition of PI3K. Pts with PIK3CA-mutant tumors appeared to derive better clinical benefit, although activity of the combination was also seen in patients with PIK3CA-wild type cancers. Additional NGS of >300 cancer genes and their correlation with clinical response will be updated at the meeting. Clinical trial information: NCT01791478.

Adverse Events (% by Grade)Toxicity300 mg (N = 20)300 mg (N = 20)350 mg (N = 6)350 mg (N = 6)% by Grade% by Grade% Total% by Grade% by Grade% Total2323Hyperglycemia181256202070Diarrhea12128110050Fatigue2504330080Rash180431020*40Nausea0062101060Vomiting00250030Anorexia602510040Dysgeusia00180050

Citation Format: Ingrid A. Mayer, Vandana Abramson, Justin Balko, Melinda Sanders, Dejan Juric, David Solit, Yisheng Li, Lewis Cantley, Eric Winer, Carlos Arteaga. SU2C Phase Ib study of the PI3K-alpha inhibitor BYL719 (alpelisib) with letrozole in ER+/HER2-metastatic breast cancer. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr CT232. doi:10.1158/1538-7445.AM2015-CT232

Abstract 5497: Cabozantinib eradicates de novo castrate-resistant PTEN/p53 deficient murine prostate cancer via activation of neutrophil-mediated anti-tumor innate immunity

The majority of kinase inhibitors in cancer clinical trials have been evaluated in the context of cell autonomous induction of apoptosis. However, apoptosis can result in protumorigenic immunosuppression, which limits the durability of an anti-tumor response. Recent clinical trial data shows striking clinical and radiographic responses with kinase inhibitor cabozantinib (XL-184) in metastatic solid tumors, particularly in the context of bone metastases. However, the mechanism responsible for this robust antitumor response remain unknown. Here we show that cabozantinib eradicates poorly differentiated invasive cancer that develop in the context of prostate-specific PTEN and p53 loss, respectively, within 48 hours of cabozantinib treatment. This rapid tumor clearance was temporally associated with infiltration of mature polymorphonuclear leukocytes into the tumor bed. These anti-tumor effects of cabozantinib appear to be MET-independent, since the MET inhibitor PF-04217903 did not phenocopy the effects of cabozantinib in the prostate-specific PTEN/p53 deficient mouse model in vivo. Unexpectedly, in vitro treatment of PTEN/p53 deficient cell lines derived from murine tumors showed insignificant apoptosis, but robust extracellular release of HMGB1, a neutrophil chemoattractant, and marker of immunogenic cell death. To elucidate the relevance of an immunogenic anti-tumor mechanism in vivo, we performed RNA-seq profiling and quantitative RT-PCR analysis, which showed an acute increase in anti-tumor inflammatory cytokine gene expression signature and neutrophil activation/chemotaxis markers following acute cabozantinib treatment. Critically, blockade of neutrophil chemotaxis/trafficking with dexamethasone or depletion with anti-Ly6G antibody, reversed the effects of cabozantinib towards eradication of advanced PTEN/p53 deficient tumors. Finally, cytokine array profiling of supernatant from bone marrow metastases from castrate-resistant prostate cancer patients showed increased neutrophil markers and decreased IL-6 levels within the bone microenvironment following 6 weeks of cabozantinib treatment., thus polarizing the neutrophils into” N1” effector cells. Collectively, these results shed light on a novel anti-tumor immunogenic mechanism for cabozantinib within the tumor microenvironment of PTEN/p53 deficient tumors, which unleashes a profound innate anti-tumor immune response. These findings suggest the possibility of combination therapies of cabozantinib with T-cell checkpoint blockade or vaccine-based approaches, to augment immunologic responses in advanced cancers.

Citation Format: Akash Patnaik, Kenneth Swanson, Katja Helenius, Thornley Thomas, Athalia Pyzer, Vilmosne Csizmadia, Sabina Signoretti, Todd Morgan, Yugang Wang, Olivier Elemento, Lily Wang, Elena Levantini, John Clohessy, John Asara, David Smith, Jacalyn Rosenblatt, David Avigan, Steven Balk, Lewis Cantley. Cabozantinib eradicates de novo castrate-resistant PTEN/p53 deficient murine prostate cancer via activation of neutrophil-mediated anti-tumor innate immunity. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 5497. doi:10.1158/1538-7445.AM2015-5497

Abstract PR13: Pancreatic tumor stem cells resistant to inhibition of oncogenic signaling are dependent on mitochondrial function

Pancreatic ductal adenocarcinoma (PDAC) continues to have a poor prognosis despite extensive characterization of underlying genetic lesions. Although KRAS mutation is a signature driver event required for tumor development and maintenance, all the attempts made to target oncogene-driven signaling pathways have been met with minimal impact on survival and tumors eventually relapse, suggesting that ablation of oncogenic pathways fails to eradicate tumor cells. We used a genetically engineered mouse model of PDAC in which we could control the expression of mutant KRAS in a time- and tissue-specific manner to study the effects of ablating oncogenic signaling on PDAC cells. As previously described (Ying et al. 2012), this model showed rapid tumor growth when KRAS was expressed, followed by robust tumor regression upon oncogene ablation. However a small population of tumor cells survive genetic and pharmacological inactivation of KRAS. These surviving cells are tumor stem cells able to remain in a quiescent state for months before repopulating the original tumor upon reactivation of KRAS signaling. In depth transcriptomic and metabolomic analyses revealed surviving cells have different metabolic features with respect to the bulk of tumor cells, showing a prominent expression of genes governing mitochondrial function, β-oxidation and autophagy, as well as strong reliance on mitochondrial respiration, yet these cells also displayed impaired glycolytic capacity. In accord with their decreased dependence on glycolysis for cellular energetics and their complete absence of glycolytic reserve, these PDAC surviving tumor stem cells show high sensitivity to oxidative phosphorylation inhibitors, which prevented tumor recurrence. Our integrated analysis paves the way for new therapeutic strategies combining KRAS signaling and oxidative phosphorylation inhibition to completely eradicate pancreatic ductal adenocarcinoma.

This abstract is also presented as Poster A87.

Citation Format: Andrea Viale, Piergiorgio Pettazzoni, Costas Lyssiotis, Haoqiang Ying, Nora Sánchez, Matteo Marchesini, Alessandro Carugo, Tessa Green, Sahil Seth, Virginia Giuliani, Timothy Heffernan, Alec Kimmelman, Huamin Wang, Jason Fleming, Lewis Cantley, Ronald DePinho, Giulio Draetta. Pancreatic tumor stem cells resistant to inhibition of oncogenic signaling are dependent on mitochondrial function. [abstract]. In: Proceedings of the AACR Special Conference on Pancreatic Cancer: Innovations in Research and Treatment; May 18-21, 2014; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2015;75(13 Suppl):Abstract nr PR13.

Abstract A87: Pancreatic tumor stem cells resistant to inhibition of oncogenic signaling are dependent on mitochondrial function

This abstract is also being presented as a short talk in Session 7: New Therapies. A full abstract is printed in the Proffered Abstracts section (PR13) of the Conference Proceedings.

Citation Format: Andrea Viale, Piergiorgio Pettazzoni, Costas Lyssiotis, Haoqiang Ying, Nora Sánchez, Matteo Marchesini, Alessandro Carugo, Tessa Green, Sahil Seth, Virginia Giuliani, Timothy Heffernan, Alec Kimmelman, Huamin Wang, Jason Fleming, Lewis Cantley, Ronald DePinho, Giulio Draetta. Pancreatic tumor stem cells resistant to inhibition of oncogenic signaling are dependent on mitochondrial function. [abstract]. In: Proceedings of the AACR Special Conference on Pancreatic Cancer: Innovations in Research and Treatment; May 18-21, 2014; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2015;75(13 Suppl):Abstract nr A87.

Active Pin1 is a key target of all-trans retinoic acid in acute promyelocytic leukemia and breast cancer

A common key regulator of oncogenic signaling pathways in multiple tumor types is the unique isomerase Pin1. However, available Pin1 inhibitors lack the required specificity and potency for inhibiting Pin1 function in vivo. By using mechanism-based screening, here we find that all-trans retinoic acid (ATRA)--a therapy for acute promyelocytic leukemia (APL) that is considered the first example of targeted therapy in cancer, but whose drug target remains elusive--inhibits and degrades active Pin1 selectively in cancer cells by directly binding to the substrate phosphate- and proline-binding pockets in the Pin1 active site. ATRA-induced Pin1 ablation degrades the protein encoded by the fusion oncogene PML-RARA and treats APL in APL cell and animal models as well as in human patients. ATRA-induced Pin1 ablation also potently inhibits triple-negative breast cancer cell growth in human cells and in animal models by acting on many Pin1 substrate oncogenes and tumor suppressors. Thus, ATRA simultaneously blocks multiple Pin1-regulated cancer-driving pathways, an attractive property for treating aggressive and drug-resistant tumors.

Abstract 2951: Vitamin C is selectively toxic to cancer cells harboring KRAS or BRAF mutations by targeting GAPDH

Although more than half of colorectal cancer patients have either KRAS or BRAF mutations, targeted therapies for these subgroups of cancer patients are still lacking. Here, we report that KRAS or BRAF mutant cells exhibit a significant increase in the uptake of oxidized vitamin C, dehydroascorbate (DHA), via the GLUT1 glucose transporter. Increased uptake of DHA in the mutant cell lines causes oxidative stress as intracellular DHA is immediately reduced back to vitamin C by reduced glutathione. Elevated reactive oxygen species inactivate GAPDH and thus divert glycolytic flux into the pentose phosphate pathway. Because KRAS or BRAF mutant cells are highly dependent on glycolysis, this altered glucose metabolism induced by vitamin C ultimately leads to an energy deficit in KRAS or BRAF mutant cells and thereby cell death. There are currently a growing number of clinical trials in phases I/II examining the effect of high dose vitamin C as a treatment for cancers, including colorectal cancers. Our results provide a rationale for the pharmacological use of vitamin C as a therapeutic agent to treat colorectal cancer patients with oncogenic KRAS or BRAF mutations.

Citation Format: Jihye Yun, Adam Kavalier, Edouard Mullarky, Kaitlyn Bosch, Jatin Roper, Carlo Rago, Jihye Paik, John Asara, Steven Gross, Bert Vogelstein, Nickolas Papadopoulos, Lewis Cantley. Vitamin C is selectively toxic to cancer cells harboring KRAS or BRAF mutations by targeting GAPDH. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 2951. doi:10.1158/1538-7445.AM2014-2951

Abstract CT418: A Phase Ib study of BKM120 combined with abiraterone acetate for castrate-resistant, metastatic prostate cancer

Introduction: There is significant cross-talk between PI3-kinase (PI3K) and androgen receptor (AR) signaling pathways, respectively, which are both critical for cell survival in castrate-resistant prostate cancer (CRPC). The primary study objective (NCT01741753) is to determine the safety profile and maximum tolerated dose (MTD) of BKM120 (pan-PI3K inhibitor) in combination with abiraterone/prednisone (A/P) in CRPC patients. The secondary study objectives are to assess the impact of PTEN status on duration of response/time to progression in the expansion cohort, and to evaluate the impact of BKM120 on a PI3-kinase activation fingerprint in metastatic bone or lymph node tissue samples. An exploratory objective of the study is to assess the effect of BKM120 on transcription of a set of AR-regulated genes in metastatic bone biopsy samples.

Methods: The trial design involves a 14 day lead-in phase with BKM120 alone, to assess single-agent toxicity and perform correlative studies. A/P is combined with BKM120 at the end of 14 days using the standard 3+3 dose-escalation design with 3 dose levels of BKM120 (80 mg, 100 mg, 120 mg, respectively), and participants are assessed for safety and MTD on the combination therapy. To determine pharmacodynamic impact of single agent BKM120 on the PI3K activation signature at a metastatic site, a mandatory CT-guided bone or lymph node biopsy is performed prior to BKM120 initiation and at the end of 2 weeks on BKM120 single-agent therapy. Immunohistochemical (IHC) stains for three markers (p-AKT, p-S6 and PTEN) are used to obtain a semi-quantitative PI3K activation score, based on the quartile levels of continuous staining scores of each marker.

Results: The clinical, PSA biochemical responses, immunohistochemical and molecular RT-PCR data for three enrolled patients are summarized here. The first patient had symptomatic bone pain with limited functional mobility at baseline, and was treated with the first BKM120 dose level of 80 mg daily. He had a decline in PSA from 156.5 to 129.6 during the lead-in phase with single-agent BKM120 treatment with partial symptomatic pain relief. Interestingly, comparison of pre- and post-treatment metastatic bone tumor IHC staining on single-agent BKM120 showed inhibition of phospho-S6 staining. Upon adding A/P to BKM120 treatment, he had a striking decline in his PSA to 23.3 within 1 week, which reached a nadir of 9.2 within 4 weeks of combination therapy treatment. This striking biochemical improvement corresponded to a marked decline in narcotic pain requirements, increased mobility and quality-of-life. However, he developed clinical, biochemical and radiographic disease progression after 4 cycles of combination therapy treatment, so had to withdraw from study. The second and third patient had a >90% biochemical improvement within 4 weeks of combination therapy treatment and remain on study to date. To evaluate cross-talk between AR and PI3K signaling in the bone microenvironment, RT-PCR analyses on pre- and BKM120-treated bone biopsy samples were performed. The results showed that the feedback relationship between PI3K and AR signaling are heterogeneous, and dependent on the PTEN status of the tumor.

Conclusion: The pan-PI3K inhibitor/abiraterone combination is well-tolerated, and demonstrates promising anti-tumor activity in CRPC patients.

Citation Format: Akash Patnaik, Massimo Loda, Justin Kung, Jim Wu, MaryEllen Taplin, Philip Kantoff, Lewis Cantley, Steven Balk, Glenn Bubley. A Phase Ib study of BKM120 combined with abiraterone acetate for castrate-resistant, metastatic prostate cancer. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr CT418. doi:10.1158/1538-7445.AM2014-CT418

Abstract OT1-4-02: Phase I study of the combination of BKM120 and olaparib for the treatment of high grade serous ovarian cancer (HGSC) and triple negative breast cancer (TNBC)

Background: In vivo synergy of the PI3-kinase inhibitor BKM120 with the PARP inhibitor olaparib has been observed using a mouse model of BRCA1-related breast cancer and sporadic TNBC (Juvekar et al and Ibrahim et al, Cancer Discovery 2012). In addition, olaparib has single agent activity in both HGSC and BRCA-associated breast cancer. The PI3kinase pathway is activated in both TNBC and HGSC (www.cancergenome.nih.gov). These preclinical and clinical data have served as the rationale for this phase I, multi-center study (NCT01623349) combining the oral PARP inhibitor olaparib with the oral PI3-kinase inhibitor BKM120 in patients with recurrent HGSC or recurrent TNBC. This study is being conducted through the Stand Up to Cancer (SU2C)'s Targeting PI3-kinase in Women's Cancers Dream Team.

Study Design: This study has a 3 + 3 design, escalating if 0/3 or 1/6 participants have a dose limiting toxicity (DLT) during the first cycle of therapy (first 28 days). The study objectives are to determine the recommended phase II dose (RP2D) of daily continuous oral olaparib (using the tablet formulation) and BKM120, assess toxicities, safety, and preliminary activity of this combination, and determine pharmacokinetic profiles of both agents. In addition, there are several translational endpoints including elucidation of downstream signaling effects of the PI3-kinase pathway, examination of BRCA1 immunostaining, and assessment of BRCA1 promoter hypermethylation and somatic mutations in BRCA1 and BRCA2 using archived formalin fixed paraffin embedded (FFPE) tissue. Serial IL-8 and circulating DNA levels are also being monitored as well. Eligibility includes a diagnosis of recurrent TNBC or HGSC, PS 0 or 1, measurable or evaluable cancer, and normal lab values and organ function. Prior PARP inhibitor exposure is allowed. In addition, breast cancer or ovarian cancer patients with any histologic subtype are eligible if they have a known germline BRCA1 or BRCA2 mutation. At the RP2D, 10 pts each with a diagnosis of TNBC or HGSC will be enrolled to further determine safety and efficacy profiles in addition to more thoroughly studying translational endpoints. As of June 7, 2013, 16 patients have been enrolled into this study with a planned accrual of approximately 50 patients which may change based on number of dose levels tested during dose escalation. In addition, an amendment is pending that will add a second cohort studying the combination of olaparib and BYL719 based on robust pre-clinical activity observed in murine models which will increase our total accrual. Once this new cohort is open, both arms will enroll simultaneously.

For further information, contact Ursula Matulonis at: umatulonis@partners.org.

Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr OT1-4-02.

Abstract C60: Vitamin C inhibits the survival and growth of colorectal cancer cells with KRAS or BRAF mutations

Although more than half of colorectal cancer patients have either KRAS or BRAF mutations, targeted therapies for these subgroups of cancer patients are lacking. We previously showed that isogenic human colorectal cancer cell (CRC) lines harboring a KRAS or BRAF mutation up-regulate the expression of glucose transporter 1 (GLUT1), and increase glucose uptake and glycolysis. In this study, we go on to show that KRAS or BRAF mutant cells exhibit a significant increase in the uptake of vitamin C, mainly in its oxidized form, dehydroascorbic acid (DHA), relative to their wild-type counterparts through GLUT1. Notably, CRCs did not uptake vitamin C in its reduced form, possibly due to the fact that its known transporters, sodium vitamin C cotransporters (SVCTs) were generally not present in CRCs. Deletion of GLUT1 in mutant cells, overexpression of GLUT1 in wild-type cells and competition assays between glucose and DHA confirmed that GLUT1 is both necessary and sufficient for the uptake of vitamin C in CRCs. Both mutant and wild-type cells were able to grow in low glucose (1-2 mM) at a similar rate. In the same condition, we found vitamin C to be selectively toxic to KRAS and BRAF mutant cells, causing them to undergo apoptosis. In contrast, in high glucose levels (>10 mM), a similar concentration of vitamin C had no effect on the survival and growth of mutant cells. In addition, vitamin C significantly inhibited the growth of cancer cells with KRAS or BRAF mutations in vivo xenografts models. To find the mechanism by which vitamin C induces cell death in KRAS or BRAF mutant cells, we performed targeted metabolomics using LC/MS. Both unlabeled total metabolite analysis and flux analysis with C13-Glucose isotope strongly suggested that vitamin C re-routes glucose usage from aerobic glycolysis and non-oxidative pentose phosphate pathway (PPP) to the oxidative PPP to compensate for the lower levels of NADPH which was rapidly consumed to reduce GSSH to GSH. Having taken up exclusively DHA, the oxidized state of vitamin C, the mutant cells immediately use GSH to reduce Vitamin C, resulting in the accumulation of GSSH. Shunting glucose to the oxidative PPP reduces aerobic glycolysis, lowering ATP levels, and decreases glucose anabolism, disfavoring the synthesis of nucleic acids, lipid and amino acids, all needed for the survival and growth of mutant cells. Surprisingly the level of reactive oxidative stress (ROS) was largely unchanged following vitamin C treatment possibly because increased levels of vitamin C served as an anti-oxidant in mutant cells. There are currently more than three clinical trials phase I/II examining the effect of high dose vitamin C via intravenous administration in patients against lymphoma or solid cancers including CRCs. Our findings will provide proof of concept that KRAS or BRAF mutations in CRC can be predictive biomarkers for vitamin C therapy.

Citation Format: Jihye Yun, Adam Kavalier, Jatin Roper, Steve Gross, Carlo Rago, Nickolas Papadopoulos, Bert Vogelstein, Lewis Cantley. Vitamin C inhibits the survival and growth of colorectal cancer cells with KRAS or BRAF mutations. [abstract]. In: Proceedings of the Third AACR International Conference on Frontiers in Basic Cancer Research; Sep 18-22, 2013; National Harbor, MD. Philadelphia (PA): AACR; Cancer Res 2013;73(19 Suppl):Abstract nr C60.

PTEN loss as a context-dependent determinant of patient outcomes in obese and non-obese endometrioid endometrial cancer patients

5521

Background: Aberrations in the PI3K pathway, the central relay pathway of insulin signals, occur in the majority of endometrioid endometrial cancers. We explored the prognostic utility of PIK3CA, PIK3R1, and PTEN mutations, as well as PTEN protein loss, in the context of patient weight. Methods: Patients (pts) treated for endometrial cancer at a single institution between 2000 and 2009 were identified. Tumor DNA was extracted and exome sequencing performed using a 454 platform with confirmation of hot spot mutations by Sequenom. PTEN protein expression was determined by immunohistochemistry and reverse phase protein array (RPPA). RPPA for 135 relevant proteins was performed using a GeneTAC arrayer to create spot arrays. Slides were scanned, analyzed, and quantitated using Microvigene software. Results: One hundred eighty seven endometrioid endometrial cancer specimens were included. Median age was 61 yrs and median body mass index (BMI) was 33.5 kg/m2. The majority of pts had early stage (I/II) disease (74%) and grade 2 tumors (66%). There were no statistically significant associations between progression free survival (PFS) and PIK3CA, PIK3R1, PTEN mutation or loss. However, when stratified by BMI, PTEN loss was associated with a significantly improved PFS (p< 0.006) in obese (BMI > 30 kg/m2) pts. In contrast, PTEN loss was associated with a worse PFS (p<0.06) in non-obese (BMI < 30 kg/m2) pts. Further, PTEN loss in obese and non-obese pts resulted in distinct protein changes by RPPA, with canonical PI3K pathway activation observed only in the non-obese PTEN loss cohort. PTEN loss in obese pts was associated with decreased expression of CATENIN and phosphorylated FOXO3A. Conclusions: These data suggest the impact of PTEN loss on tumor biology and clinical outcomes must be interpreted in the context of BMI and provide potential explanation for prior discrepant findings on effect of PTEN status on prognosis in endometrial cancer. These data describe a clinically important interaction between metabolic state and tumor genetics that could potentially unveil the biologic underpinning of obesity-related cancers and may be relevant to ongoing clinical trials with PI3K pathway inhibitors.

Phase II, two-stage, two-arm, PIK3CA mutation stratified trial of MK-2206 in recurrent endometrial cancer (EC)

5524

Background: EC has high rates of PI3K pathway alteration including PTEN mutation (50%) or IHC loss (>50%), PIK3CA mutation (25-40%) and PIK3R1 (20%) mutation. MK-2206 is an allosteric inhibitor of AKT, an effector kinase of PI3K signals. We hypothesized that pts whose tumors harbored PIK3CA mutations would be more likely to benefit from MK-2206 than those without PIK3CA mutation. Methods: Pts had recurrent or advanced EC; all histologies except MMMT were eligible. Up to 2 prior chemo lines were permitted; excluding prior treatment with PI3K/MTOR inhibitors. The first 19 pts were treated with MK-2206 200mg QW; due to initial skin toxicity rates, the starting dose was amended to 135mg QW. Co-primary endpoints were objective response and 6 mo PFS. The first 37 pts were stratified retrospectively. PIK3CA MT included R88Q, K111N, E110K, E418K, C420R, E453K, E542K/V,E545K, Q546R, H701P, M1043V, H1047R/L/Y changes. Independent Simon 2-stage tests were planned within PIK3CA MT and WT stratum: for MT, n1=15 and n2=10 pts would allow discrimination of RR<5% and 6moPFS<10% versus RR>25% or 6moPFS>35%; for WT, n1=31 and n2=24 pts would discriminate RR<5% and 6moPFS<10% versus RR>20% or 6moPFS>25%. Results: 37 pts were enrolled (1 ineligible) before accrual was stopped as timely CLIA-compliant prospective mutation analysis was not feasible. By PIK3CA mutation analysis, 9 pts were MT: 1 pt had both PR and 6moPFS. 27 pts were WT: 1 pt had PR and 3 pts had 6moPFS. 2 pts with 6moPFS were treated at 200mg; 2 pts with 6moPFS were treated at 135mg. Each group had 1 PR. The most common toxicity was grade 3 rash (19%). Grade 3 and 4 toxicities occurred in 50% and 8% of pts. Exploratory analysis of histology found all pts with 6moPFS were classified as serous (4 of 8) as compared to all other histologies (0 of 28, p=.001). Targeted exome sequencing and copy number analysis of the PI3K pathway and PTEN IHC are underway. Associative studies will be reported. Conclusions: There is limited single agent activity of MK-2206 in both PIK3CA MT and PIK3CA WT EC populations. Activity was detected in pts with serous histology tumors and warrants further study. Clinical trial information: NCT01307631. [Table: see text]

Abstract 4588: Identification of CDCP1 as a HIF-2α target gene involved in the regulation of cancer cell migration and metastasis

CUB domain-containing protein 1 (CDCP1) is a transmembrane protein that is highly expressed in stem cells and frequently overexpressed and tyrosine phosphorylated in cancer. CDCP1 promotes cancer cell metastasis. However, the mechanisms that regulate CDCP1 are not well defined. Studies from our laboratory revealed a biochemical pathway by which CDCP1 participates in the activation of Src-family kinase (SFK) members and the coupling of SFK-activation to the phosphorylation and regulation of protein kinase C-delta (PKC-δ). Here we show that hypoxia induces CDCP1 expression and tyrosine phosphorylation in a HIF-2α, but not HIF-1α, dependent fashion. shRNA knockdown of CDCP1 impairs cancer cell migration under hypoxic conditions, while overexpression of HIF-2α promotes the growth of tumor xenografts in association with enhanced CDCP1 expression and tyrosine phosphorylation, as well as, significantly promotes lung metastases in NOD/SCID mice. To investigate the relationship between HIF-2α and CDCP1 expression, we performed a correlation analysis in the largest up-to-date collection (Sanger Cell Line Project) of cancer cell line microarray data (n=732). We found a dramatic concordance in the expression of HIF-2α and CDCP1 (Pearson's correlation, P &lt;1x10-20), indicating that cancers with high HIF-2α expression tend to have high levels of CDCP1 expression. We next asked whether other known HIF-2α target genes also correlate in this expression analysis. Remarkably, MET and EGFR, which are hypoxia regulated and known HIF-2α target genes, also displayed a strong correlation with HIF-2α and CDCP1 expression. Immunohistochemistry analysis of tissue microarray samples from tumors of patients with clear cell renal cell carcinoma (ccRCC) shows that increased CDCP1 expression correlates with decreased overall survival. Interestingly, high-grade ccRCCs (G3, G4) expressed significantly higher (P = 0.03, t-test) levels of CDCP1 protein compared to lower grade tumors (G1, G2), suggesting that CDCP1 expression increases progressively with higher ccRCC tumor grade. Furthermore, hypoxia activates Src signaling and the Src inhibitor (Dasatinib) prevents the hypoxia-induced phosphorylation of CDCP1. Thereby, reinforcing that CDCP1 is an SFK-associated receptor, which promotes migration and metastasis and suggests that hypoxia-induced CDCP1 signaling may further stimulate a more aggressive cancer phenotype. Together, these data support a critical role for CDCP1 as a unique HIF-2α target gene involved in the regulation of cancer metastasis, and suggest that therapeutic approaches targeting CDCP1, such as monoclonal antibodies, could be beneficial in the treatment of metastatic cancers.

Supported by NIH grant 5R01GM056203-15 to L.C.C and Dana Farber/Harvard Cancer Center Career Development Award to B.M.E.

Citation Format: Brooke M. Emerling, Cyril Benes, Eric Bell, George Poulogiannis, Kevin Courtney, Hui Lui, Rayman Choo-Wing, Gary Bellinger, Stephen Soltoff, Lewis Cantley. Identification of CDCP1 as a HIF-2α target gene involved in the regulation of cancer cell migration and metastasis. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 4588. doi:10.1158/1538-7445.AM2013-4588

Abstract MS1-1: Targeting PI3K

Phosphoinositide 3-Kinase (PI3K) is a central enzyme in a signaling pathway that mediates cellular responses to growth factors. The signaling pathway downstream of PI3K is highly conserved from worms and flies to humans and genetic analysis of the pathway has revealed a conserved role in regulating glucose metabolism and cell growth. Mutational events that lead to hyperactivation of the PI3K pathway result in hamartoma syndromes and cancers. Activating mutations in PIK3CA, encoding the p110alpha catalytic subunit of PI3K (PIK3CA) or inactivating mutations in PTEN, a phosphoinositide 3-phosphatases that reverses the effects of PI3K, are among the most common events in solid tumors, especially breast cancers. Drugs that target PI3K are in clinical trials for a variety of cancers. It is likely that PI3K pathway inhibitors will need to be combined with other drugs to be broadly effective. We have employed genetically engineered mouse models that develop cancers due to mutations in genes in the PI3K pathway and are using these models to explore the efficacy of PI3K pathway inhibitors as single agents or in combination with other drugs. We are also interrogating the role of PI3K in the growth and survival of Brca1/p53 mutant breast cancers. The role of PI3K inhibitors for treating cancers in these mouse models and in human trials will be discussed.

Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr MS1-1.

Abstract LB-365: Effective use of PI3K/MTOR and MEK inhibitors prior to hormone ablative therapy in PTEN-loss driven murine prostate cancer

The PTEN and p53 tumor suppressors are the most commonly altered genes in human cancer, including prostate cancer (PCa). Loss of PTEN is associated with increased Gleason score and clinical recurrence, and the majority of human metastatic PCas have PTEN loss via multiple mechanisms. Mice with prostate-specific homozygous deletion of PTEN develop invasive PCa albeit with prolonged latency of 6-8 months. Combined PTEN/p53 inactivation in mouse prostate elicits invasive cancer by 9 weeks of age and invariable lethality by 6 months of age. Since PTEN loss results in PI3K/mTOR pathway activation, we evaluated the impact of GSK458 (PI3K/mTOR inhibitor) and GSK418 (PI3K beta/delta isoform-specific inhibitor), singly and in combination with GSK212 (MEK inhibitor), in uncastrated, prostate-specific PTEN/p53 double knockout mice (4-6 months) and PTEN mice (11-14 months), respectively, harboring advanced PCa. The drugs were administered by daily oral gavage for 3 weeks with serial 18FDG-PET/MRI imaging at baseline, 2 days, 1 week, 2 weeks and 3 weeks post-treatment respectively. GSK458 treatment of PTEN/p53 and PTEN mice results in reduction in FDG-PET uptake as early as 24 h post-treatment, with 40% tumor shrinkage by 1 week post-treatment, but rapid regrowth of 18FDG-avid tumor by 2-3 weeks post-treatment. This acquired resistance was found to be mediated in part through upregulation of multiple receptor tyrosine kinases. In contrast, PTEN/p53 and PTEN mice did not respond to GSK418, by either FDG-PET or MRI analysis. We observed increased pERK/total ERK ratio by Western blot analysis and increased p-ERK staining by immunohistochemistry in GSK458 and GSK418-treated PTEN/53 mice, respectively. Treatment of PTEN/p53 and PTEN mice with GSK212 resulted in an approx. 40% reduction in tumor volume over 3 weeks. Treatment of PTEN/p53 mice with a combination of GSK458 plus GSK212 resulted in approx. 60% reduction of tumor volume at 3 week post-treatment. Strikingly, treatment of prostate-specific PTEN only mice with GSK458 plus GSK212 combination resulted in a &gt;90% tumor regression at 3 weeks post-treatment. These results demonstrate the potential utlitity of PI3K/MEK-directed combination therapies in the neoadjuvant setting for locally advanced disease or hormone-sensitive phase in metastatic disease, thus delaying the need for hormone ablative therapy and its associated morbidity in advanced PCa. The data underscore the value of genetically engineered mouse models to co-clinically evaluate biomarkers of response and resistance to targeted therapies, and elucidate mechanisms of acquired resistance early in clinical development. The design of “personalized” combination therapies to overcome resistance to PI3K/MEK-directed therapies in the hormone-naive and castration-resistant contexts are currently underway in multiple GEMMs and Phase Ib co-clinical trials in advanced PCa.

Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr LB-365. doi:1538-7445.AM2012-LB-365

Abstract LB-418: Targeting the PI3K/mTOR pathway in genetically engineered mouse models of prostate cancer

The PTEN and p53 tumor suppressors are among the most commonly inactivated or mutated genes in human cancer, including prostate cancer. Loss of PTEN is associated with increased pathologic Gleason score and risk of clinical recurrence, and 20-60% of human metastatic prostate cancers have loss of heterozygosity at the PTEN locus, resulting in hyperactivation of the PI3K/mTOR pathway. Mice with germline heterozygosity for PTEN have been shown to develop prostate intraepithelial neoplasia (PIN) at a high rate (&gt;60%) and mice with prostate specific homozygous deletion of PTEN develop invasive prostate cancer, albeit with prolonged latency of approx. 6 months. Combined inactivation of PTEN and p53 in mouse prostate elicits invasive cancer by 9 weeks of age and invariable lethality by 7 months of age.

There are several PI3K pathway-directed therapies currently in Phase I clinical trials, but the underlying tumor genetic signature of patients most likely to respond to these therapies is largely unknown. To understand the significance of targeting the PI3K/mTOR pathway in advanced prostate cancer driven by PTEN +/− p53 loss, we evaluated the impact of GSK458 (a dual PI3K/mTOR inhibitor) in prostate-specific PTEN/p53 double knockout mice and prostate-specific PTEN-knockout mice. The mice were imaged by synchronized 18FDG-PET and T2-weighted MRI, respectively, for baseline tumor metabolic and volumetric assessment prior to drug administration. GSK458 was administered at 3 mg/kg by daily oral gavage for 3 weeks with serial 18FDG-PET and T2-weighted MRI imaging at 2 days, 1 week, 2 weeks and 3 weeks after initiation of treatment, followed by sacrifice, prostate harvest and standard hisopathologic and immunohistochemical staining.

GSK458 treatment of PTEN/p53-deficient and PTEN-deficient mice results in target inhibition, based on pharmacodynamic assessment by 18FDG-PET uptake. MRI and histopathologic analysis demonstrate that there is a significant reduction, but not complete regression of tumor burden in both intraepithelial and poorly differentiated atypical components within stroma and partial stromal collapse following 3 weeks of GSK458 treatment. These data highlight the feasibility of monitoring dual pharmacodynamics/antitumor effects of PI3K-directed therapies using 18FDG-PET/MRI imaging and underscore the utility of genetically engineered mouse models to predict response to targeted therapies in genetically stratified human clinical trials. The evaluation of PI3K-isoform specific inhibitors and the design of rational combinations to overcome de novo and acquired resistance mechanisms to PI3K-directed therapies are currently being explored in multiple genetically engineered mouse model systems.

Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr LB-418. doi:10.1158/1538-7445.AM2011-LB-418

Abstract LB-255: A mass spectrometry platform to quantitatively profile cancer cell metabolism from cells, tumors, and fixed tissue

The metabolic requirements of cancer and proliferating cells are different from that of normal differential tissue and may have diverse applications in the treatment of cancers. However, many of the molecular mechanisms that reorganize metabolism to support cell proliferation are unknown. To study cancer cell metabolism, we implemented a mass spectrometry based platform to quantitatively profile endogenous metabolites from proliferating cell lines, tumor tissues and formalin fixed paraffin embedded (FFPE) tissue. Cell lines are derived from several cancers including lung, multiple myeloma, prostate, and gliobastoma (GBM). In some cases, these were compared to a drosophila reference cell line. Patients were also profiled for disease classification from their cerebrospinal fluid (CSF). We also were successful in extracting polar metabolites from FFPE tissue more than five years old stored at room temperature. For FFPE tissue samples, we show that we can observe differences in disease states involving PI3K-TSC-TOR pathway and we compared different extraction methods for acquiring metabolomics data from FFPE blocks. We show that we can cluster GBM versus normal patients from analyzing their CSF. We target more than 255 unique metabolites using selected reaction monitoring (SRM) based analyses with an AB/Sciex 5500 QTRAP mass spectrometer coupled to a Shimadzu UFLC using normal phase chromatography. For a single 18 min run, our platform allows for unprecedented sensitivity, quantitation and coverage of metabolites that comprise of diverse metabolic pathways from as little as a single 6 cm tissue culture dish of cells or approximately 2–3 million cells from tissue samples. We find that amide XBridge columns (Waters) at 275 uL/min perform well in both negative and positive ion switching mode and that the sampling rate of the instrument is sufficiently fast (cycle time of 1.6 sec with 3 msec dwell times) to effectively capture up to 300 metabolite targets without scheduled SRM runs. Peak areas of metabolites are integrated using MultiQuant 1.1 software (Applied Biosystems). Peak areas from triplicate runs are hierarchically clustered and statistical analyses are applied to generate P values for metabolite changes over different biological conditions. We also probed metabolic flux in pathways by targeting a set of 13C glucose labeled metabolites. In addition, we have also analyzed a model cell line after stimulation with Insulin and EGF to examine if growth factor induced metabolic changes are evolutionarily conserved. Using metabolic inhibitors, such as Iodoacetic acid, KCN, etc., we have been able to characterize the consequences of inhibiting glycolysis and oxidative phosphorylation, respectively. Finally, we considered kinase inhibitors and measured their effects on metabolism in proliferating cancer cell lines.

Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr LB-255. doi:10.1158/1538-7445.AM2011-LB-255

AMP-activated protein kinase regulates neuronal polarization by interfering with PI 3-kinase localization

Axon-dendrite polarization is crucial for neural network wiring and information processing in the brain. Polarization begins with the transformation of a single neurite into an axon and its subsequent rapid extension, which requires coordination of cellular energy status to allow for transport of building materials to support axon growth. We found that activation of the energy-sensing adenosine 5'-monophosphate (AMP)-activated protein kinase (AMPK) pathway suppressed axon initiation and neuronal polarization. Phosphorylation of the kinesin light chain of the Kif5 motor protein by AMPK disrupted the association of the motor with phosphatidylinositol 3-kinase (PI3K), preventing PI3K targeting to the axonal tip and inhibiting polarization and axon growth.

Extended work hours and risk of acute occupational injury: A case-crossover study of workers in manufacturing

OBJECTIVE

This study was designed to determine whether injury risk among manufacturing workers was related to hours worked during the previous week.

METHODS

A case-crossover design was utilized to contrast hours worked prior to an injury shift with those worked prior to a non-injury shift for hourly workers. Paired t-tests were used to determine significance of the difference. Conditional logistic regression was used to assess dose-response.

RESULTS

Hours worked prior to injury significantly exceeded hours during the control week. Workers who worked more than 64 hr in the week before the shift had an 88% excess risk compared to those who worked 40 hr or fewer, P < 0.05.

CONCLUSION

The study provides evidence that injury risk is related to time worked during the previous week. Control of overtime in manufacturing may reduce risk of worker injury.