Genetic Toxicology Association Spring Meeting
Thursday, April 26, 2007
Clayton Hall Conference Center, University of Delaware
Identification of Novel Therapeutic and Diagnostic Targets for Lung Cancer through Proteomics
Elizabeth G. Joseloff
Lung cancer is the most commonly diagnosed cancer worldwide and a leading cause of cancer-related deaths. Limited response to existing therapies contributes to the 15% 5-year survival rate indicating a need for both improved lung cancer diagnostic biomarkers as well as therapeutics. To address this, we have performed mass spectrometry (MS) to analyze the cell surface proteome of lung cancer cell lines and tumor tissues to identify novel targets for monoclonal antibody therapeutics. We have also used our MS analysis in combination with drug sensitivity data to identify novel lung diagnostic biomarkers to predict and monitor patient response to treatment.
Back to Top ↑
Biomarkers of Key Events for Use in the EPA’s Cancer Risk Assessment Process
R. Julian Preston (NHEERL, EPA)
The revised EPA Guidelines for Carcinogen Risk Assessment have identified a framework for incorporating mechanistic data into the approach, thereby placing less reliance on uncertainty (default) factors. The initial phase is to describe a particular mode-of-action in terms of a set of key events that are necessary for converting a normal cell to a transformed cell and ultimately to a metastatic tumor. These key events are by definition measurable and can be quantitated. Preston and Williams (Critical Reviews in Toxicology 35:673-683, 2005) described these key events for DNA-reactive chemicals. The mode-of-action/key event approach developed for rodent tumors for a particular chemical can be utilized to establish if the same mode-of-action/key events are plausible or are operating in humans using a Human Relevance Framework (Meek, et al., Critical Reviews in Toxicology 33:591-653, 2003). This process can identify human relevance for a particular mode-of-action and the need to conduct a cancer risk assessment, but also can indicate data gaps that if filled can reduce uncertainty in risk assessments. A significant advantage to the use of mechanistic data in the identification of key events is that biological indicators of these key events can be used to estimate tumor outcomes at doses below those at which tumors themselves (rodent or human) can be measured. These indicators can be used in a qualitative sense for better defining the shapes of a tumor dose-response at low exposure levels. They are in fact occasionally used in a quantitative way to estimate tumor frequencies at low exposure levels.
The ability using current molecular biology techniques to assess whole genome responses at the DNA, mRNA or protein levels allows for the selection of informative bioindicators of response (tumors or preneoplastic lesions, for example). In addition, using the sophisticated computational models that are being developed for defining systems at the cellular, tissue, organ or organism level, it should be possible to predict even more specifically the outcomes of exposures to environmental chemicals in humans. The ability to extrapolate from high to low dose, across species, and from acute to chronic exposure should be greatly enhanced.
Back to Top ↑
Validation of Pathological and Genomic Biomarkers for Chemically-Induced Liver Tumors and their Contribution to Mechanism Based Risk Assessment
Mark Fielden, Ph.D.
Alternatives to the two year rodent bioassay for identifying carcinogenic chemicals would facilitate increased cost savings for development of therapeutics and identification of potentially carcinogenic chemicals, however, current short term tests are unsuitable for regulatory purposes. Alternatively, identifying non-genotoxic carcinogens and their mode of action early in product development would facilitate a pro-active risk assessment that would not normally occur until after a definitive two year carcinogenicity study. To this end, we developed a 5 day repeat dose in vivo test in rats that utilizes an hepatic gene expression-based biomarker signature to predict and understand the mode of action for non-genotoxic hepatocarcinogens. The biomarker was validated on 47 independent chemicals and found to predict hepatocarcinogenicity with greater accuracy compared to increased liver weight, hepatocellular hypertrophy, hepatic injury and a number of other putative gene-based biomarkers. In contrast to these and other alternative methods, an understanding of potential modes of action can be derived through comparison of the biomarker profile of test chemicals to hepatocarcinogens of known mechanism, thus facilitating simultaneous prediction and mechanism-based assessment of human cancer risk.
Back to Top ↑
Overview of DNA Adducts as Biomarkers of Human Cancer Risk
Miriam C. Poirier (Carcinogen-DNA Interactions Section, NCI, NIH)
A critical first step in the carcinogenic process, for cancers induced by chemicals, is DNA damage. This can include formation of covalently-bound carcinogen-DNA adducts, other chemical modifications of DNA bases (oxidative damage), and alterations in DNA ultra-structure (DNA strand cross-links, DNA strand breaks, chromosomal aberrations, and chromosomal loss). Whereas repair mechanisms exist for many types of DNA damage, residual damage may lead to the insertion of an incorrect base during DNA replication, transcription and translation on mutated templates, and synthesis of an altered protein. The result may be a clone of cells with a proliferative growth advantage. Whereas the etiology of chemically-induced cancers in humans requires a multi-step process that occurs over many years, and often involves a chronic chemical exposure, many studies have measured DNA adducts in human tissues years before the appearance of tumors.
A large body of animal model literature has shown that DNA damage is necessary but not specific for carcinogenesis involving chemical exposures. For genotoxic chemical carcinogens, dose response curves for chemical exposure, DNA adduct formation, and mutagenesis, often parallel cancer induction in tissues that develop tumors. Furthermore, metabolic modulations that reduce the formation of DNA adducts in target tissues of rodents will also reduce tumor risk. DNA adducts may form in organs that do not develop tumors, indicating that species- and tissue-specific risk factors (for example cell proliferation) also contribute to tumor induction. However, tumors do not appear in the absence of DNA damage.
It would be useful to elucidate the role played by DNA adduct formation in human cancer etiology. The human situation is confounded by multiple simultaneous exposures, but epidemiologic approaches (e.g. the prospective nested case-control study design) may be used to reveal an association between DNA adduct formation and human cancer risk. If such an association can be validated it may be possible to devise strategies for cancer prevention. Individuals might be targeted for chemoprevention efforts designed to inhibit the formation of DNA damage, even though overall exposure cannot be altered.
The current talk will examine the chemical nature of DNA adducts, discuss measurement and extent of DNA adduct formation in humans, provide evidence that indicates a role for DNA adduct formation in human cancer risk and discuss strategies for chemoprevention.
Back to Top ↑
An Integrative, Systems Approach to Drug and Biomarker Discovery: Genomics, Proteomics, and Bioinformatics
John N. Weinstein, M.D., Ph.D. (Genomics & Bioinformatics Group, Lab of Molecular Pharmacology, Center for Cancer Research, National Cancer Institute)
In November 2006, we and our collaborators launched the "Spotlight on Molecular Profiling" series in the journal Molecular Cancer Therapeutics. The series highlights 'integromic' molecular profiling studies, focused initially on the NCI-60 cell line panel – with implications for carcinogenesis as well as for toxicology and therapeutics. The NCI-60 panel, used by the NCI Developmental Therapeutics Program to screen >100,000 compounds and natural products for anticancer activity, is the most comprehensively characterized set of cells anywhere. Our molecular profilings completed or in progress have included transcript expression using cDNA arrays, Affymetrix arrays (HU6800, U95, and U133), spotted oligo arrays, and real-time RT-PCR; microRNA expression by oligo array; protein expression by 2-D gels and reverse-phase lysate microarrays; BAC, Agilent, and NimbleGen array CGH; spectral karyotyping; DNA methylation by bisulfite sequencing and HELP assay; mutations by resequencing and Affymetrix SNP chip. The overall profiling enterprise can be considered as a forerunner to The Cancer Genome Atlas project -- with more diverse molecular characterizations but in the easier context of cultured cells. The various types of molecular information are presented in CellMiner (http://discover.nci.nih.gov/cellminer (Shankavaram, in preparation), a user-friendly relational database of molecular data and metadata on the NCI-60. Proofs of principle for 'integromic' analysis include: (i) identification of candidate biomarkers for differential diagnosis of ovarian and colon metastases; (ii) analyses that led to clinical development of oxaliplatin, now standard-of-care for colon cancer; (iii) discovery of "MDR1-inverse" compounds, which, paradoxically, are more potent in cells that express MDR1; (iv) development of the "Permissive Apoptosis-Resistance" (PAR) model for acquired resistance to cancer drugs and xenobiotic toxicants; (v) identification of asparagine synthetase (ASNS) as a potential 'causal' biomarker for treatment of ovarian cancer with L-asparaginase (L-ASP). With thanks to our many other collaborators in the overall enterprise. http://discover.nci.nih.gov. References: (1) Weinstein, Spotlight on molecular profiling: 'integromic' analysis of the NCI-60 cancer cell lines. Mol. Cancer Ther. 2006; 5: 2601. (2) Lorenzi, et al., ibid 2006; 5: 2613. (3) Ikediobi, et al., ibid 2006; 5: 2606. (4) Ludwig and Weinstein, Nature Rev. Cancer 2005; 5: 845.
Back to Top ↑
FDA Perspective on Toxicogenomic Biomarkers
Federico Goodsaid, Ph.D. (Senior Staff Scientist in Genomics at OCPB/CDER/FDA)
Back to Top ↑