Nicholas Denko - US grants

Evidence is gathered that the tumor microenvironment has a profound effect upon tumor cell development. Physiologic stresses of low oxygen (hypoxia), low glucose, and low growth factors are all found within a tumor and can exert selective pressures upon the tumor cell population. Specifically, hypoxia can induce genomic instability, cell cycle arrest, resistance to chemotherapy, apoptosis, and increased production of angiogenic factors; all of which can lead to a more aggressive tumor. In order to determine the mechanism by which hypoxia exerts these effects, novel hypoxia-induced messenger RNAs were identified. The goals of this study are 1) to identify the specific stresses that induce two related, hypoxia-induced messages, 2) to determine the mechanism for their accumulation, and 3) to investigate the role that the protein product plays in the hypoxic phenotype. The experimental design will utilize Northern blots, run-off transcript analysis, message stability studies, expression of mRNAs under heterologous promoters, expression of tagged protein products, and cell biologic investigation for hypoxia-like phenotypes. These studies may provide insight into novel targets for cancer therapy, or novel reagents for cancer-specific gene- therapy.

Bioinformatics and statistics have become increasingly important in recent years as the complexity of the pathways being studied has expanded. We have three project in this PPG that are using high- throughput, genome wide analyses to generate large data sets. It is one of the roles of this core to help manage the large volumes of data being produced and make that data available in a manageable form for the members of the Stanford community, and then the rest of the scientific community as well. We will establish algorithms that can be used by the different groups to cluster expression profiles using unsupervised as well as supervised machine learning. We will also establish web-based interfaces with the Stanford community, and yeast consortium to facilitate the analysis of the homozygote deletion pool. Finally, we will work with the projects on all experiments in order to help establish exacting criteria of statistical power and significance.

[unreadable] DESCRIPTION (provided by applicant): [unreadable] Hypoxia is a unique patho-physiologic stress of the solid tumor that has been shown both experimentally and clinically to influence tumor aggressiveness, metastatic potential and response to therapy. The mechanisms for these phenomena have not been determined, but are thought to be at least partially dependent upon gene and protein expression changes induced by hypoxia. Investigators have identified many components of the hypoxic stress response cascade starting from hypoxic sensing, to transcriptional changes, to protein expression changes and post-translational protein modifications. Tumor cells that are unable to start new hypoxia-responsive mRNA transcription grow poorly in model tumors (HIF knockouts). Likewise, cells that fail to express the hypoxic target protein vascular endothelial growth factor also grow poorly in model tumors (VEGF knockouts). The conclusion can therefore be made that studying other components of the stress response cascade is reasonable, because blocking them could have similarly profound impact on tumor growth. We propose to investigate if HIG2 can influence stress-dependent protein synthesis because we have evidence that HIG2 is associated with translational machinery during hypoxia. We have previously identified HIG2 as a novel 63 amino acid proteins that is robustly induced by hypoxia in a wide variety of normal cells and tumor cell lines. We now show that the HIG2 protein co-localizes with members of the cytoplasmic [unreadable] "Stress granule". Stress granules have been shown to regulate protein translation in response to heat shock, and we suggest that they serve a similar function during hypoxia. This grant proposal is based upon the hypothesis that HIG2 plays an important function in regulating protein synthesis during hypoxia through its association with components of the stress granule. To address this hypothesis, we propose to 1) Identify the mechanism by which HIG2 is targeted to the stress granule, 2) Determine binding partners for HIG2 during hypoxia and 3) Establish the functional significance of HIG2 within the stress granule during tumor formation. These experiments should determine if HIG2 is necessary for hypoxia-dependent translational control, and how important this control is to the survival of ceils within the hypoxic regions of human tumors. [unreadable] [unreadable]

DESCRIPTION (provided by applicant): Tumor hypoxia has been shown in several clinical studies to be a prognostic indicator of poor treatment outcome. These findings have been independent of therapeutic modality, with hypoxie tumors that are treated surgically doing equally as poorly as those treated with radiation therapy. Genetic models in mice, and patients with VHL-deficient tumors, have suggested that the ability to induce genes in response to hypoxia with the transcription factor HIF-1 may play a stimulatory role in tumorigenesis. Hypoxia, however, is also a potent repressor of gene expression, but few studies have investigated the mechanism of hypoxia-dependent gene downregulation. We present the hypothesis that gene repression by hypoxia can contribute to tmnorigenesis, and the NC2 (negative co-factor 2) molecules are responsible for at least some of this repression. NC2alpha/beta (Drl/DRAP) was originally characterized as a biochemical activity that was able to repress transcription in vitro. We have preliminary evidence that repressive activity is induced in hypoxic cells, and extracts from hypoxic cells fail to transcribe (at least) some templates in vitro. Biochemical characterization of this activity suggests that NC2 may be a general repressor of transcription in hypoxia. NC2 induction by hypoxia could then contribute to tumor formation through the downregulation of such hypoxia-repressed target genes as tumor suppressors thrombospondin 1 and 2 (TSP1/TSP2), apoptotic inhibitors stathmin and survivin, and tumor marker alpha fetal protein (AFP). We therefore propose to establish a genetic system for the analysis of NC2's role in tumor development. We will address this goal with three specific aims, first establish that NC2 is responsible for hypoxic repression of specific target promoters in vivo, second perform structure-function relationship of NC2alpha and NC2beta to identify domains that are critical for hypoxic gene repression, and third construct tumor cell lines with compromised NC2 activity that test proof in principle that hypoxic gene repression is important in model tumor formation. These studies will determine if hypoxic gene repression through NC2 activity could be a potential therapeutic target for future molecular therapeutics in the treatment of human cancers.

DESCRIPTION (provided by applicant): Tumor hypoxia has been shown in several clinical studies to be a prognostic indicator of poor treatment outcome. These findings have been independent of therapeutic modality, with hypoxie tumors that are treated surgically doing equally as poorly as those treated with radiation therapy. Genetic models in mice, and patients with VHL-deficient tumors, have suggested that the ability to induce genes in response to hypoxia with the transcription factor HIF-1 may play a stimulatory role in tumorigenesis. Hypoxia, however, is also a potent repressor of gene expression, but few studies have investigated the mechanism of hypoxia-dependent gene downregulation. We present the hypothesis that gene repression by hypoxia can contribute to tmnorigenesis, and the NC2 (negative co-factor 2) molecules are responsible for at least some of this repression. NC2alpha/beta (Drl/DRAP) was originally characterized as a biochemical activity that was able to repress transcription in vitro. We have preliminary evidence that repressive activity is induced in hypoxic cells, and extracts from hypoxic cells fail to transcribe (at least) some templates in vitro. Biochemical characterization of this activity suggests that NC2 may be a general repressor of transcription in hypoxia. NC2 induction by hypoxia could then contribute to tumor formation through the downregulation of such hypoxia-repressed target genes as tumor suppressors thrombospondin 1 and 2 (TSP1/TSP2), apoptotic inhibitors stathmin and survivin, and tumor marker alpha fetal protein (AFP). We therefore propose to establish a genetic system for the analysis of NC2's role in tumor development. We will address this goal with three specific aims, first establish that NC2 is responsible for hypoxic repression of specific target promoters in vivo, second perform structure-function relationship of NC2alpha and NC2beta to identify domains that are critical for hypoxic gene repression, and third construct tumor cell lines with compromised NC2 activity that test proof in principle that hypoxic gene repression is important in model tumor formation. These studies will determine if hypoxic gene repression through NC2 activity could be a potential therapeutic target for future molecular therapeutics in the treatment of human cancers.

Hypoxia causes gene expression changes largely through the induction of the HIF1 transcription factor, and many of these changes are thought to help adapt to the adverse environment where oxygen is limiting. One class of hypoxia-induced genes that have been extensively studied is the glycolytic enzymes. These molecules are thought to be necessary to maintain energy production when reduced oxygen will not support oxidative phosphorylation within the mitochondria. While glycolysis is important to cellular growth in hypoxia, we find that the mitochondrion does not just passively stop functioning. HIF-proficient cells actively reduce oxygen consumption in hypoxia while HIF-deficient cells do not. We therefore used expression profiling and data mining during the past funding cycle to identify hypoxia-induced proteins that are targeted to the mitochondria. These putative HIF-1 regulated mitochondrial proteins do not cause apoptosis, but our functional data supports the novel concept that they actively regulate mitochondrial activity in response to hypoxia. We therefore propose to address the following four questions in this application. In specific aim 1, we will determine if HIF-dependent gene expression changes result in altered oxygen consumption in the mitochondria through the induction of target genes BNip3/L, and/or pyruvate dehydrogenase kinase 1 (PDK1) and/or hypoxia-induced gene 1 (HIG1). In specific aim 2 we will test the hypothesis that pharmacologic reversal of these HIF-1 dependent changes will increase oxygen consumption, diminish intracellular oxygen concentrations, and result in sensitivity to oxygen-dependent therapies such as the hypoxic cytotoxins tirapazamine (TPZ).or dinitobenzamide mustard Pr-104. In specific aim 3 we will ask if this pharmacologic treatment that makes tumors more hypoxic also makes them more aggressive and likely to metastasize. Lastly in specific aim 4 we hypothesize that we can identify additional novel regulators of hypoxic mitochondrial function through a screen of the yeast deletion library using growth in hypoxia on non- fermentable carbon source media. The proposed experiments will allow us to determine how these molecules contribute to hypoxic regulation of mitochondrial function, what they contribute to the growth of model tumors, and what impact they have on the tumor's response to oxygen-dependent therapy.

DESCRIPTION (provided by applicant): The alterations in tumor cell metabolism that result in a glycolytic phenotype have long been recognized, and are exploited as the basis of the FDG-PET scan. This aberrant metabolism is also thought to give tumor cells a growth advantage. Our recent work has identified the hypoxia-inducible gene pyruvate dehydroggenase kinase 1 (PDK1) as one factor that contributes to this metabolic shift by downregulating mitochondrial function. We propose to test the hypothesis that we can revert tumor cell metabolism to a more oxidative state by inhibition of PDK1 with the small molecule dichloroacetate (DCA). DCA has been shown to have anti- cancer effects in model tumors, and has been safely given to patients with metabolic disorders for over 30 years. We feel that this justifies a clinical trial in patients with advanced Head and Neck cancer whose treatment options have been exhausted. One innovative aspect of this trial will be the evaluation of the patients with functional PET imaging to determine if there is a change in the tumor metabolism in response to DCA. We will use standard FDG-PET to evaluate glucose uptake in the tumor and a novel hypoxic radiotracer EF5 to determine oxygen tension within the tumor. As DCA shifts metabolism to oxidative rather than glycolytic, FDG signal should decrease and EF5 signal should increase. Changes in tracer uptake will be correlated to molecular findings in vitro. We will obtain fine needle aspirate biopsies from the patient's pre and post DCA and determine the level of phosphorylation of the E1a subunit of the pyruvate dehydrogenase complex (the target of PDK1). We will also collect patient serum pre and post DCA and determine if there is any increase in the level of secreted gene products of hypoxia-responsive genes such as osteopontin or galectin-1. This trial will lay the groundwork for a phase 2 trial evaluating the clinical efficacy of DCA as an anti-cancer drug. PUBLIC HEALTH RELEVANCE: Novel anti-cancer treatments based on tumor cell metabolism represent an exciting new avenue of study. The proposed trial evaluates novel functional imaging modalities with a novel metabolic modulating anti-cancer agent. The combination of these technologies will give the most rigorous investigation of the potential for this class of drugs. )

DESCRIPTION (provided by applicant): Tumor hypoxia has been recognized as a hindrance to successful radiation therapy for over 50 years. Attempts to overcome this obstacle by delivering more oxygen to the tumor, however, have been clinically disappointing, largely due to the functional limitations of the tumor vasculature. Instead of reducing hypoxia by increased delivery of oxygen, this application proposes to limit hypoxia by reducing oxygen consumption within the tumor. If the supply of oxygen delivered to the tumor is constant, then transient reduction in demand will increase overall functional oxygenation. Commonly prescribed anti-diabetic biguanidedrugs (metformin, phenformin) have been shown to reduce mitochondrial function in vitro at least in part through inhibition of electron transport chain (ETC) complex 1. We propose to test the hypothesis that pharmacologic downregulation of mitochondrial metabolism will reduce cellular demand for oxygen and result in decreased tumor hypoxia and specific radiosensitization of model tumors. This approach will be especially effective when using hypofractionated radiation protocols where oxygen enhancement can have a profound effect on overall tumor cell killing. We have organized this proposal into the following four specific aims. 1) Determine the role of tumor suppressor LKB1 in mediating the effect of biguanides on mitochondrial metabolism. 2) Establish the relative importance of glucose versus glutamine as a mitochondrial fuel in regulating mitochondrial response to intervention with biguanides. 3) Quantitate the biochemical effect of biguanides on mitochondrial function, tumor hypoxia, and glucose consumption in vivo. And 4) Establish the optimal level of radiosensitization in both subcutaneous and orthotopic model tumors treated with biguanides and radiation. It is important to note that because normal tissue is typically well oxygenated, thi systemic approach will specifically radiosensitize tumors, without causing enhanced normal tissue toxicity.

Tumor hypoxia has been recognized as a hindrance to successful radiation therapy for over 50 years. The search for small molecules to act as oxygen mimetics and selectively sensitize hypoxic tumors to radiotherapy has had many worthy candidates in the past, but none have been successful in Phase III clinical trials. In addition, attempts to overcome hypoxia by delivering more oxygen to the tumor, however, have been clinically disappointing, largely due to the functional limitations of the tumor vasculature. Increasing systemic oxygen delivery does not result in increased tumor oxygenation. Instead of reducing hypoxia by increased delivery of oxygen, this application proposes to limit hypoxia by reducing oxygen consumption within the tumor. If the supply of oxygen delivered to the tumor is constant, then transient reduction in demand will result in increased tumor oxygenation. We have identified several commonly prescribed drugs (papaverine is the lead compound) that dramatically reduce mitochondrial function in vitro. We propose to test the hvpothesis that pharmacologic down regulation of mitochondrial metabolism will reduce cellular demand for oxygen and result in decreased tumor hypoxia and radiosensitization of model tumors. This approach will be especially effective when added to hypofractionated radiation protocols where oxygen enhancement can result in a profound increase on overall tumor cell killing. We have organized this proposal into the following four specific aims. 1) Determine the effective concentration of two mechanistically different, FDA approved drugs to inhibit the mitochondrial function of a panel of cancer cell lines in vitro. 2) Quantify the biochemical effect of drug treatment on mitochondrial function, tumor hypoxia, and glucose consumption in model tumors with Core B. 3) Establish the optimal level of radiosensitization in both subcutaneous and orthotopic model tumors treated with metabolic modifiers in collaboration with Project 1 and 4, and 4) Initiate a clinical trial that directly measures tumor oxygenation changes in patients with advanced Head and Neck Squamous Cell Carcinomas (HNSCC) with Project 2. It is important to note that because normal tissue is typically well oxygenated, systemic delivery of well tolerated metabolic modifiers will specifically radiosensitize tumors, without causing enhanced normal tissue toxicity.

? DESCRIPTION (provided by applicant): This shared instrument application is to support the purchase of a small animal radiation research platform (SARRP). The technology available from Xstrahl is unique and combines a cone beam CT with a computer controlled rotating treatment x-ray tube. In combination with the treatment planning software, this hardware allows for 3-dimensional radiotherapy treatment planning and delivery to model organisms that mimics the volumetric approach used to treat human tumors. This next-generation system leverages the recent advances in small animal CT to generate true image guided radiotherapy, or IGRT. The device incorporates a 360o rotating gantry with a stage that rotates and translates in 3 dimensions to allow for multiple x-ray beamlets to be delivered from any angle and converge upon the tumor. This system will replace the current technology at Ohio State University that is based on 50 year old design for cabinet X-ray machines. The cabinet x-ray device at Ohio State University is an RS2000, a point source X-ray generator fixed to the top of a sealed cabinet. In order to treat animals, one must use manual placement of lead shielding in such a way as to shield a portion of the mouse and allow x-ray delivery only to the model tumor. The SARRP will enhance the research of 6 major users with 9 currently funded NIH projects that span the departments of Radiation Oncology, Neurosurgery, and Molecular Virology, Immunology and Medical Genetics at OSUCCC and the Experimental Therapeutics program at Nationwide Children's Research Institute. In addition, the SARRP device will be incorporated into the training for the graduate program in medical physics run in the department of Radiation Oncology. Finally, the SARRP will be combined into the services that are offered at the OSU Neuroscience Center (P30) core for orthotopic brain tumor implantation. This core is designed to support the development of new technologies for the treatment of brain tumors. Current and future investigators using this core will now be able to incorporate radiotherapy with their novel modalities in the hope of generating innovative, synergistic therapeutics to treat brain tumors.

ABSTRACT Tumor hypoxia reduces the effectiveness of radiotherapy by reducing the effective dose delivered to the tumor. Cells that are severely hypoxic require 2.8-fold greater dose to achieve the same cell kill as those that are fully oxygenated. For this reason, many groups have tried to deliver more oxygen to tumors as a therapy to reduce hypoxia and increase radiosensitivity. These strategies did effectively radiosensitize model tumors in rodents, but did not prove successful in human trials. We have looked at tumor oxygenation differently from the biological perspective and the evaluation of success perspective. In terms of biology, if we could clinically reduce oxygen demand rather than increase its supply, we could effectively reduce hypoxia and produce tumor radiosensitization. We have identified papaverine as an FDA-approved molecule with the ability to inhibit mitochondrial function at clinical doses. Studies in mouse tumors support the idea that papaverine can radiosensitize through regulation of oxygen consumption, producing ?Metabolic Radiosensitization?. In terms of the evaluation of papaverine, we propose to test its potential as a clinical radiosensitizer in a heterogeneous population of spontaneous canine soft tissue sarcomas being treated at OSU Veterinary Medical School. The heterogeneous host- and tumor genetics generate a clinical trial that we hypothesize will be much more predictive of human success than a rodent study. Papaverine is not targeted to a specific cancer mutation, and should be effective as a radiosensitizer in any solid tumor where hypoxia exists. The proposed study will be a phase 0 pharmacodynamics study that will determine if papaverine can increase oxygenation in canine soft tissue sarcoma as measured by near infrared spectroscopy (FD-NIRS) technology in real time. We propose to test 10 animals with 1 mg/kg and 10 animals with 2 mg/kg papaverine. Correlative biological studies will determine if baseline tumor hypoxia or other biological variables can predict response to papaverine. These studies will determine if canine soft tissue sarcoma is a feasible system to test new hypoxic tumor radiosensitizers, and if papaverine should be considered as a potential radiosensitizer in future interventional trials.

ABSTRACT Tumor hypoxia reduces the effectiveness of anti-cancer treatment with radiotherapy, some chemotherapy and immune checkpoint blockade therapy. For radiotherapy, biophysical measures show that hypoxic cells require 2.8-fold greater dose to achieve the same cell kill as those that are fully oxygenated. For immunotherapy, hypoxia has been shown to contribute to immune evasion and even accelerate T cell exhaustion. For these reasons, many groups have tried to deliver more oxygen to tumors as an adjuvant to increase tumor sensitivity. Unfortunately, this approach has met with disappointing clinical results. We have looked at tumor oxygenation differently, as a supply and demand mismatch, with the supply being inadequate to meet the demand of the growing tumor mass. Therefore, if we could reduce oxygen demand rather than increase supply, we could effectively reduce hypoxia and sensitize tumors. Because mitochondria are the major sink for oxygen within a cell, we propose that novel mitochondria inhibitors would reduce oxygen demand to match the limited supply. We have identified papaverine (PPV) as an FDA-approved molecule with the ability to inhibit mitochondrial function at clinical doses. Published studies from our group showed that in mouse tumors that papaverine can radiosensitize through inhibition of mitochondrial function, producing ?Metabolic Radiosensitization?. Papaverine was originally isolated from the poppy and used as a smooth muscle vasorelaxant presumably due to inhibition of phosphodiesterase 10A. This activity makes it an effective drug for cerebral vasospasm, but causes a systemic drop in blood pressure and potential adverse cardiovascular effects. We therefore propose in this application to synthesize and evaluate new small molecule derivatives of papverine that we have designed to remove its activity as a phosphodiesterase inhibitor, but retain its activity as a mitochondrial complex 1 inhibitor. Using these PPV derivatives, and sophisticated mouse models of cancer, we intend to prove that inhibition of mitochondrial function is an effective strategy for removing hypoxia in solid tumors without affecting well oxygenated normal tissue. Preliminary data supports the overall theory that mitochondrial inhibitors increase tumor oxygenation and sensitivity to radiotherapy and immune checkpoint blockade therapy.