(Abstracts provided by applicants)

Heparin Sulfate–A Novel Target for Cancer Treatment
Jeffrey Esko, Principal Investigator University California at San Diego, CA	U01 CA091290-01
Tumor growth depends on a variety of secreted growth and angiogenic factors. Cellular responses to many of these factors rely on the presence of cell surface heparan sulfate proteoglycans, which serve as co-receptors for several signaling pathways. Mounting evidence suggests that inhibition of heparan sulfate biosynthesis could bock the ability of tumor cells to respond growth and angiogenic stimuli. This hypothesis will be tested by altering heparan sulfate synthesis in tumor cells and model organisms. The novel molecular target in this proposal is the biosynthetic pathway for heparan sulfate. Blocking or reducing the expression of specific genes in the pathway by genetic strategies is needed to validate specific enzymatic targets for eventual pharmacologic intervention. Screening technologies are in place to identify compounds derived from natural products and synthetic libraries. To achieve the long-term goal of developing novel chemotherapeutic agents, we have the following specific aims: 1. Validate the heparan sulfate biosynthetic pathway as a target for cancer drug therapy. Previous studies of Chinese hamster ovary cells suggest that ablation of heparan sulfate formation prevents tumor formation in athymic mice. To extend these studies to more common tumor lines, heparan sulfate biosynthesis will be altered genetically using classic mutagenesis, anti-sense methods, new RNA interference procedures, and chimeraplasty (RNA/DNA hybrids). These studies will focus on EXT1 and NDST1 since these enzymes are responsible for the polymerization of the polysaccharide chain and initiation of all downstream modification reactions. 2. Identify new targets in the heparan sulfate biosynthetic pathway. Although many of the genes that encode the biosynthetic enzymes for heparan sulfate assembly have been identified, several critical components have not. To characterize the role of these other components and to study their role in tumor formation, new mutants of CHO cells will be isolated in the C5 epimerase and EXT2, a subunit of the co- polymerase complex. RNAi methods recently developed for Drosophila tissue culture cells will be employed to determine the function of D-Ext2, D-epimerase, and developed for Drosophila tissue culture cells will be employed to determine the function of D-Ext2, D-epimerase, and D-Ext1 in heparan sulfate biosynthesis. 3. Identify and characterize inhibitors of heparan sulfate biosynthesis. A high throughput screening method was been developed to identify compounds that inhibit heparan sulfate biosynthesis in cultured cells. To date, over 60,000 samples have been screened from the Natural Products Branch of the Developmental Therapeutics Program of the NCI and active extracts have been identified. Parallel screening of synthetic libraries is planned. Assays for screening drug candidates using sensitized genetic backgroups in Drosophila will be established and inhibition of PTEN-mediated overgrowth will be measured. The molecular target of active compounds will be examined by characterizing intermediates that accumulate in cells or tissues after drug treatment. Identification of active components coupled with large scale synthesis would allow us to test if the compounds have anti-tumor activity in mice.

Small Molecule Inhibitors of Bcl xL Survival Protein
David Hockenbery, Principal Investigator Fred Hutchinson Cancer Research Center, Seattle, WA	U01 CA91310-01
The Bcl-2-related survival proteins confer cellular resistance to a wide range of apoptosis-inducing agents. In work can-led out in our labs, a novel small molecular ligand to the Bcl-xL protein has been identified, which inhibits the molecular pore function of Bcl-xL and selectively kills Bcl-xL and, at higher doses, Bcl neg2-expressing cells. We made the initial observation that Bcl-xL-expressing hepatocyte cell lines are more sensitive than isogenic control cells to antimycin A (AA), a known inhibitor of mitochondrial electron transport. A 2-methoxy antimycin A analog lacking effects on mitochondrial respiration still exhibited selective toxicity for Bcl-xL plus cells and mitochondria. Computational molecular docking analysis predicted that antimycin A conforms to a conserved hydrophobic groove on the molecular surface of Bcl-xL. We confirmed this interaction by showing competitive binding of AA and its 2-methoxy derivative with a known hydrophobic groove ligand to recombinant Bcl-xL and Bcl-2 proteins, a BH3 domain peptide derived from the pro-apoptotic dimerization partner, Bak. Finally, we found that AA inhibits the pore-forming activity of Bcl-xL in synthetic Liposomes, demonstrating that this small ligand can directly inhibit the function of Bcl-2-related survival proteins. Two aims of this application investigate the structural determinants of AA binding to the Bcl-xL hydrophobic pocket and mechanism of pore inhibition. Initial screening of human hematopoietic cell lines for cytotoxic effects of antimycin A indicates myeloma cell lines, including multi-drug resistant sublines, are sensitive to AA and 2-methoxy AA. Several published studies have shown that myeloma cell survival is predominantly dependent on Bcl-xL, despite the expression of several related anti-apoptotic proteins. We propose, using pre-clinical models, to test whether multiple myeloma is particularly susceptible to Bcl-xL-targeted therapies, and validate Bcl-xL as the relevant target of 2-methoxy AA in myeloma cells. The questions to be addressed In three specific alms are as follows: Specific Aims: Evaluate efficacy and toxicology of 2-methoxy antimycin A3 in mouse myeloma and hepatoma tumor models. Characterize biochemical mechanism of antimycin A inhibition of Bcl-,xL pore-forming function, including analysis of mutations in Bcl-xL hydrophobic groove binding site; x-ray crystallography of antimycin complex with Bcl-xL; and membrane topology studies. Determine the role of endogenous pro-apoptotic dimer partners of Eel-if in the cytotoxic mechanism of antimycin A.

Targeted Intervention Against EphA2 on Cancer Cells
Deborah Knapp, Principal Investigator Michael Kinch, (former)Principal Investigator Purdue University, West Lafayette, IN	U01 CA91318-01
Elevated levels of tyrosine kinase activity are necessary the growth and dissemination of malignant carcinoma cells. Our laboratory recently applied new technologies in monoclonal antibody production and identified dramatic changes in the expression and function of the EphA2 tyrosine kinase in malignant carcinomas. EphA2 is overexpressed in cancer cells (by 50-500 fold), with the highest levels of EphA2 consistently found on cells with metastatic potential. We have also shown that EphA2 overexpression confers metastatic potential but that the growth and invasiveness of metastatic cells can be reversed by activating EphA2 at the cell surface. Based on these results, we hypothesize that EphA2 provides unextraordinary opportunity for antibody-based targeting of metastatic carcinoma. To test this, we recently used innovative strategies to generate antibodies against epitopes on the extracellular domain of EphA2. Here, we propose to develop new assay systems to test EphA2 as a target for monoclonal antibody treatment of metastatic cancer cells. Our Specific Aims are: i) to identify EphA2 antibodies that can target tumor cells, ii) to test if these antibodies block metastatic cell growth or invasiveness, and iii) to determine the mechanisms of antibody action. Upon conclusion of our studies, we will have validated EphA2 as a molecular target for cancer treatment and will have generated the reagents and expertise to exploit a critical vulnerability on malignant cells.

Selective Mannosidase Inhibitors as Cancer Therapeutics
Kelley Moreman, Principal Investigator University of Georgia, Athens, GA	U01 CA91295-01
In response to the RFA, we have focused on identifying a selective inhibitor of the target enzyme, Golgi alpha-mannosidase II. Many lines of evidence of demonstrate that inhibiting this s enzyme target can retard tumor progression in vivo. All of the inhibitors of this enzyme, however, have an unacceptable, serious side effect that precludes their further study as therapeutic drugs because they are active against another alpha-mannosidase located in the lysosome, as well as a-mannosidase II. Inhibition of the lysosomal enzyme causes a phenocopy of the deadly lysosomal storage disease, Golgi alpha-mannosidosis. Significant differences in the substrate binding specificities of these two mannosidases can now be exploited to develop an inhibitor that is selective for Golgi alpha-mannosidase II. The focus of this proposal is to develop a selective inhibitor of this enzyme that is active in vivo, thereby validating Golgi alpha-mannosidase II as a novel target for a potential anti-tumor therapeutic. Several key scientific advances make this proposal timely, and we have assembled a team of investigators at the Complex Carbohydrate Research Center, Univ. of Georgia, to take advantage of these advances. First, the human Golgi alpha-mannosidase II has been cloned and expressed by a member of our team, in order to be used in inhibitor screening experiments. The lysosomal enzyme has been expressed similarly, and recent data on the structure of the active sites of the enzymes will aid in inhibitor design. High through-put inhibitor assays have been developed to screen thousands of potential inhibitors. A novel class of inhibitor compounds, the sulfonium salts, have been developed by a third team member. These compounds, as well as the inhibitor mannostatin, will be modified by a directed combinatorial synthetic strategy by a fourth member to generate libraries of compounds that will be screened to find identify potent selective inhibitors of the Golgi a-mannosidase II in vitro. Lead compounds will be used again in modified combinatorial syntheses to refine inhibitory properties. All of these leads will be tested for in vivo inhibition using a cell-based rapid screening assay for selected inhibition of N-glycan processing. Compounds that show in vivo selective inhibition will then be tested for their abilities to inhibit the Golgi mannosidase in mouse tissues when delivered orally or by s.c. osmotic pumps. The experiments described in this proposal, therefore, will allow us to isolate lead compounds that selectively inhibit N-linked glycosylation events known to promote tumor progression and demonstrate the validity of the Golgi alpha-mannosidase II as a prime target for tumor therapeutic development.

Chemokines as Molecular Targets in Cancer Drug Discovery
Joost Oppenheim, Principal Investigator NCI-Frederick, Frederick, MD	U01 Z01BC 9289-01
At times the mobilization of cells by chemoattractants, such as the small chemoattractant cytokines known as chemokines, can have deleterious consequences such as contributing to tumor growth and bone marrow graft rejection. Recent reports have demonstrated that higher levels of chemokine production correlate with progression of breast tumors and melanomas. Mechanistic studies have demonstrated that in addition to inducing leukocyte migration, chemokines induce microvascular endothelial cells to migrate contributing to neovascularization. Therefore, by participating in angiogenesis, chemokines may support solid tumor growth by effecting the tumor microenvironment. Additionally, there is evidence that select chemokines protect lymphoma and melanoma cells from steroid or cytokine-induced cell death. Consequently, chemokines produced by tumor cells may act as survival factors thereby protecting the tumor from cytokine-induced cell death. Thus, there are two possible mechanisms by which chemokines contribute to tumor biology (1) an autocrine growth, survival factor mechanism and a (2) paracrine mechanism of angiogenesis. Based on the above in vivo correlative data this study intends to demonstrate that chemokines and their receptors are likely anti-tumor targets and will identify drugs with anti-angiogenic and anti-tumor cell survival activities.

Contrasting Properties of Integrin Cytoplasmic Domains
Renata Pasqualini, Principal Investigator University of Texas MD Anderson Cancer Center Houston, TX	U01-CA91134-01
Integrins that bind to vitronectin are highly expressed in neovasculature and play an important regulatory role in angiogenesis. At least two cytokine-induced pathways lead to angiogenesis in vivo, and evidence indicates that these pathways can be distinguished by their dependency on specific av integrins. Our objective is to define the molecules involved in alpha v beta3- alpha v beta5-selective angiogenic signaling. We hypothesize that (i) different molecules associate with each of these integrins after angiogenesis is triggered by a defined cytokine and (ii) the assembly of specific molecules associating with the beta3 or the beta5 cytoplasmic domain results in selective signaling. Two independent, yet complementary strategies will be used to approach these questions: isolation of antibodies against specific elements present in alpha v beta3 and alpha v beta5 focal adhesion preparations and panning of phage display peptide libraries on beta3 and beta 5 cytoplasmic domains. These studies will shed light into the molecular basis of selective signal transduction pathways initiated by alpha v beta3 and alpha v beta5. A better understanding of angiogenesis and new ways of manipulating such process in physiological and pathological situations may result from this work.

Translational Apparatus as a Target for Drug Discovery
Vitaly Polunovsky, Principal Investigator University of Minnesota, Minneapolis, MN	U01 CA91220-01
A major limiting factor in anti-neoplastic therapy is the failure of some tumor types to respond to anticancer treatments, and the appearance of resistant cell populations in originally responsive malignancies upon relapse. There is a large body of evidence that genetic changes leading to increased cap-dependent translation are accompanied by cancer cell chemoresistance and are associated with a poor prognosis. We have recently discovered that inhibiting assembly of the cap-dependent translation initiation apparatus eIF4F, (a trimolecular complex of eIF4G, eIF4A and eIF4E) in Ras Vl2 transformed cells by transfer of the gene encoding translational repressor 4E-BPl (which sequesters eIF4E, the 5? mRNA cap binding protein), sensitizes these cells to non-genotoxic and genotoxic cytostatic drugs in vitro, and dramatically reduces their tumorigenicity. Most importantly from a therapeutic point of view, we have found that disruption of eIF4F assembly specifically activates apoptosis in cancer cells, but not in non-transformed cells, identifying eIF4F as a potential novel molecular target for anticancer drug discovery . Thus, here we propose to test the hypothesis that phosphoramidate nucleotide derivatives that repress cap-dependent translation initiation by interfering with the interaction between the 5? rnRNA cap and eIF4E will activate apoptosis in a wide spectrum of cancer cells with an upregulated translation apparatus; and sensitize them to safe doses of chemotherapeutic agents without harming desirable bystander cells. Our Aims include: Aim 1. Synthesize a library of nucleotides predicted to inhibit binding of eIF4E to the 5? rnRNA cap; Aim 2. Test candidate compounds in an ordered series of high and medium throughput in vitro assay systems; Aim 3. Utilize preclinical models of breast and lung cancer to test the most promising compound (based on Aim 2 results) for its ability to collaborate with well-tolerated doses of available cancer therapeutics to inhibit xenograft growth in athymic mice. To achieve these aims, we have assembled an experienced investigative team committed to anticancer drug discovery.

TGF122Beta-receptor antagonists as anticancer agents
Michael Reiss, Principal Investigator UNIVof MED/DENT NJ-R W JOHNSON MED SCHOOL Piscataway, NJ	U01-CA94431-01
Transforming Growth Factor-betas are polypeptides that are constitutively secreted and activated by many carcinomas. They contribute to the tumor's ability to invade and metastasize, to induce angiogenesis and to escape from immune destruction. By the same token, cancer cells themselves are generally refractory to TGFbeta-mediated growth arrest. This particular set of circumstances raises the question whether blocking the effects of tumor-derived TGFbeta on normal tissue (stromal cells, microvascular endothelial cells and immune cells) might constitute a novel approach to cancer treatment. In the past, several different strategies have been employed to counteract the biological effects of TGFbeta in cancer and other diseases. These have included the use of TGFbeta neutralizing antibodies, of TGFbeta- binding proteins, such as decorin, and of TGFbeta1 antisense RNA oligonucleotides. Animal experiments and small-scale clinical studies using each of these approaches have provided proof of concept that inactivation of TGFbeta has the predicted anti-tumor effect. However, larger scale testing and further clinical development of any of these compounds has been marred by technical difficulties and limited availability. We now propose an alternative strategy to blocking TGFbeta action by targeting the key molecule in TGFbeta signaling, i.e. the type I TGFbeta receptor (TbetaR-I) serine-threonine kinase. Small molecular selective TbetaR-I antagonists are likely to be more effective than the approaches mentioned above, and should not be subject to the same limitations in terms of production and bioavailability. Our collaborators at SCIOS, Inc. have identified several promising lead compounds that inhibit TbetaR-I kinase activity in cell-free as well as in cellular systems in vitro. In addition, we have developed the capability to measure effects of TbetaR-I kinase inhibitors in vivo, using a proprietary highly sensitive antibody that selectively detects phosphorylated Smad2. We intend to examine the effects of lead compounds against normal cells in vitro using a number of different assays for TGFbeta's biological effects. The best TbetaR-I antagonists will then be tested for their antitumor activity against transplantable tumors in mice, with particular attention to their effects on metastasis, angiogenesis and anti-tumor immunity. Finally, optimization of the compounds in terms of potency, selectivity and bioavailability will be carried out to derive analogs with a favorable toxicity profile that can be developed further for clinical use.

Hsp90s as targets in the development of anticancer drugs
Neal Rosen, Principal Investigator Sloan-Kettering Institute for Cancer Research New York, NY	U01 CA91178-01
The hsp90 family of proteins consists of Hsp90 alpha and beta, Grp94 and Trap-1. They are abundant chaperone proteins that play roles in protein refolding and processing and the conformational maturation of several key signaling proteins, including steroid receptors, Raf kinase and several transmembrane tyrosine kinases. Hsp90 is overexpressed in tumors and may play a role in maintaining the transformed phenotype by stabilizing signaling proteins and mutated protooncogenes and by mediating tumor cell survival in hypoxic, harsh environments. Ansamycin antibiotics and radicicol are natural products that bind to a conserved pocket in the hsp90 family proteins. Occupancy of this pocket by drug alters their function and causes the degradation of the signaling proteins that require Hsp90. Addition of ansamycins to cancer cells causes RB-dependent G1 arrest, differentiation and apoptosis. Cells with defective RB function arrest in prometaphase and undergo apoptosis. One ansamycin, 17 - allylaminogeeldanamycin (17- AAG) is currently in clinical trial. The main goal of this research is to identify compounds that bind lead compounds that bind to the hsp90 pocket and have selective activities that confer novel indecence properties. A lead compound, a geldanamycin (GM) dimer has been identified with selectivity for HER-family tyrosine kinases. We have also begun to synthesize compounds designed to bind to the hsp90-family pocket and to develop chemical libraries of molecules that bind to the different hsp90 family members. These molecules will be screened for their ability to bind selectively to the chaperones, degrade specific targets, and inhibit tumor cells with particular molecular lesions. Our goal is to derive selective compounds and use them to determine the biologic effects of inhibiting each of the hsp90 family members and to identify new, target-directed drugs with anticancer activity.

FLT3 as a Target for Drug Discovery to Treat Leukemia
Donald Small, Principal Investigator Johns Hopkins University, Baltimore, MD	U01 CA91177-01
Our long-term goals are to increase the cure rate and decrease chemotherapy-relate toxicity for patients with leukemia. In particular, we are interested in patients whose leukemias express an activated FLT3 receptor, either as a result of mutation or ligand co-expression. 20% of AML patients have mutations of the FLT3 receptor. These patients have a particularly poor prognosis with almost no one being cured in either the pediatric or adult populations. The mutation of the FLT3 receptor consists of small (18-105 bp) internal tandem duplications (ITDs) which are unique for each patient but all map to the juxtamembrane region of the receptor. These mutations constitutively activate the tyrosine kinase domain of FLT3 receptor which is required for signaling. This makes the development of novel strategies for these patients imperative if we are to effect improvements in their outcome. If FLT3 constitutive activation is contributing to leukemogenesis in these patients, one strategy would be to develop small molecule inhibitors of the kinase domain of the receptor. We therefore propose two major specific aims in this proposal with FLT3 as the novel molecular target for drug discovery. The first id to validate FLT3 as a molecular target by developing several animal models that mimic different ways that the FLT3 can become constitutively activated. We will also introduce second "hits" that often occur in leukemia to study the spectrum of the hematopoietic disease to which an activated FLT3 can contribute. The second aim is to develop assays utilizing cell lines and proteins that can be utilized for high-throughout screening of FLT3 inhibitors. This will enable the discovery of lead compounds that are unable to inhibit FLT3 in a highly potent and specific fashion. We believe that this work will lead to the identification of compounds able to inhibit FLT3 which will ultimately be used as novel therapeutics against leukemia.

Novel Agents that Activate Apoptosis in Tumor Cells
Allan Weissman, Principal Investigator Karen Vousden, (former)Principal Investigator NCI-Frederick, Frederick, MD	U01 Z01BC 10327-01
The development of cancer depends on the accumulation of specific genetic alterations that allow aberrant proliferation of the tumor cell. Protection from such aberrant growth is provided by several failsafe mechanisms that work by inducing apoptotic cell death in cells undergoing oncogenic changes. Therefore, for a tumor cell to survive, it must acquire genetic alterations that perturb the link between abnormal growth and cell death. One very common mechanism to develop resistance to apoptosis is the inability to release cytochrome c from mitochondria, a key event in the induction of apoptosis by oncogenic changes. The p53 tumor suppressor protein plays a pivotal role in this pathway, and loss of the ability to induce p53, either by mutation of p53 itself or by alterations in the pathways leading to p53 activation, occur in most cancers. These alterations will be exploited by designing assays to identify drugs to reactivate the apoptotic pathways in tumors scells. Emphasis will be placed on the development of assays to identify inhibitors of the ubiquitin ligase activities of MDM2 and E6/E6AP, the principal regulators of p53, for use in tumors that retain wild type p53. Screens to identify cytochrome c mimetics will also be developed. Such compounds would repair apoptotic defects downstream of p53, and so be effective in tumors expressing mutant p53. Unlike normal cells, cancer cells are continuously exposed to strong apoptotic stimuli, and survive only by virtue of genetic defects that compromise the ability of the cell to respond to these signals. Reactivating the apoptotic pathways would, in principal, efficiently and selectively kill cancer cells.