University of North Carolina at Chapel Hill

Profiles of Susceptibility to Toxicant Stress

William Kaufmann, Ph.D.
bill_kaufmann@med.unc.edu

Project Description

The toxicogenomics research program at UNC-CH is focused on genetic determinants of susceptibility to environmental toxicants and chemotherapeutic drugs. Human and murine model systems will be used to determine the patterns of alterations in gene expression when cells and animals are exposed to several classes of toxins. Toxin classes under study include agents that induce oxidative stress, DNA double strand breaks, alkylation of DNA bases, as well as nuclear receptor agonists. cDNA microarrays containing >10,000 expressed genes will be used to determine for human cells and mice the dose- and time-kinetics of toxicant-induced changes in gene expression. In two interactive research projects diploid human mammary epithelial cells, fibroblasts and lymphoblasts, will be treated with doses of toxicants ranging about the mean lethal dose (D0 for inactivation of colony formation). Cell lines with heterozygous and homozygous mutations in the tumor suppressor genes p53, ATM and BRCA1, or with genetic inactivation of suppressor function, will be compared to wildtype lines to establish profiles of genetic susceptibility to toxicant stress. Similarly, a third research project will determine patterns of response to the toxicants in liver, mammary gland and colon of ten strains of mice with varying susceptibilities to carcinogenesis. These studies will determine whether murine strain-dependent susceptibility to carcinogenesis resembles human genetic susceptibility due to mutations in tumor suppressors. A Toxicology Resource Core proposes a demonstration project to determine the profiles of altered gene expression in livers of mice exposed to non-genotoxic chemicals which activate the nuclear receptors AhR, CAR and PPARa. The three research projects and the Toxicology Resource Core will be served by an Administrative Core, a Microarray Facility Core, and an Informatics Facility Core. The UNC-C toxicogenomics research program has the scientific expertise and organizational infrastructure to provide significant and substantial benefits to the Toxicogenomics Research Consortium.

 Back to top

Project 1: Gene Expression Signatures of Toxicant-exposed Human Mammary Epithelial Cells

Principal Investigator: Chuck Perou, Ph.D.
cperou@med.unc.edu

A mechanistic understanding of how toxicants act at the molecular level is essential to knowing how these agents cause damage and how cells respond to this damage. In addition, many commonly prescribed chemotherapeutics have specific mechanisms of action, but many also have pleiotropic toxicological effects that can influence how patients and tumors respond to these agents. A well known example is doxorubicin/adriamycin, which is often defined as a topoisomerase II inhibitor; however, doxorubicin is also an oxidative stress-inducing compound that can damage DMA independent of topoisomerase inhibition. We and others have hypothesized that the mechanism of action of a given compound might result in specific gene expression changes that will reflect a compendium of that compound's effects, and that by studying these patterns we might be better able to understand how a compound works. In addition, by assaying many different compounds with known and unknown targets, we can build a library of toxicant-elicited gene expression profiles that would serve as a molecular "signature" for each compound, or for a class of compounds.

In this application, we propose to use two different epithelial cell line models representing both breast basal and luminal epithelial cell tumors (Perou et al. 2000). More specifically, we will determine the gene expression patterns elicited by the addition of toxicants that damage DNA and other toxicant/chemotherapeutics, on our panel of six cell lines with three goals in mind. First, we will begin to build a library of toxicant-elicited gene expression "signatures" that may help us define and better understand the effects that these toxicants have. Second, we will determine if specific genetic alterations, specifically p53 alterations, influence the observed toxicant-elicited gene expression patterns and/or the overall sensitivity of these cell lines to these compounds. Finally, we will compare our in vitro library of toxicant and chemotherapeutic-induced gen expression patterns to our already existing in vivo library of over 100 individual breast tumor expression profiles to determine if any of the patterns elicited by specific toxicants in vitro are also elicited in vivo in a patient exposed to the same toxicant/chemotherapeutic.

 Back to top

Project 2: Profiles of Genetic Susceptibility to Oxidative Stress

Principal Investigator: Bill Kaufmann, Ph.D.
bill_kaufmann@med.unc.edu

Several tumor suppressor genes, notably p53, ATM and BRCA1, serve a caretaker function by enforcing cell cycle checkpoint responses to DNA damage and enhancing repair of premutagenic and preclastogenic DNA damage. The proposed experiments emerge from our earlier findings that normal cell lines derived from different tissue types show a differential response to DNA damaging agents. This project will test the hypothesis that each of these genes can mitigate the genotoxic consequences of oxidative DNA damage in karyotypically stable, diploid human cells. If our hypothesis is correct, then we will be able to define sets of genes that are up- and down-regulated in human cells in response to oxidative DNA damage. Our strategy will be to use human fibroblast and mammary epithelial lines, immortalized by transduction of hTERT to induce expression of telomerase, and human lymphoblasts and treat these cells with three agents that damage cells through generation of oxygen-derived free radicals: ionizing radiation, hydrogen peroxide, and doxorubicin (also known as adriamycin). Thus, the profiles of response to DNA damage will be obtained in diploid lines with stable genomes. Our approach will be to measure the induction of two types of oxidized bases, thymine glycol and 8-hydroxyguanine, in DNA to establish the molecular dose of oxidative DNA damage produced by each agent in each cell type and the rates of repair these lesions. Topoisomerase-ll associated DNA strand breaks will also be quantified as doxorubicin is known to inhibit the strand passing reaction of topoisomerase-ll. Measurements of DNA repair, apoptosis and cell cycle checkpoint response will be correlated with changes in gene expression using >10K gene human cDNA microarrays. We will determine whether the three cell types display patterns of response to DNA that reflect their known differences in checkpoint and apoptotic responses, and whether inactivation of p53, ATM and BRCA1 influences these responses. This effort will begin to address the mechanisms of genetic and cell-typespecific variation in susceptibility to toxicant stress.

 Back to top

Project 3: Mouse Strain-Specific Molecular Profiles in Response to Toxicants

Principal Investigator: David Threadgill, Ph.D.
david_threadgill@med.unc.edu

The mouse has become the stalwart of molecular dissection in the post-genome era. In support of broadening the value of the mouse as an experimental model for all facets of biology, including toxicological research, a multi-national effort is underway to develop a comprehensive comparative database on the physiology of inbred strains representing a broad diversity. To bolster the linkages between the genome and phenome projects and the response to toxicants, we will perform an extensive comparative gene expression profile in the mouse as a first step toward genetically dissecting unique genotype-specific toxicant-induced gene expression profiles.

A detailed comparative analysis will be targeted to three tissues, colon, mammary gland, and liver, in a set of inbred strains selected to provide the strongest link to the developing mouse phenome project. The first aim will be to develop and statistically test approaches to minimize individual and experimental variables. The optimal approaches will be applied in the second aim to a series of ten inbred mouse strains to produce a reference database on baseline gene expression profiles in the target tissues. The third aim is to perform expression profiling on the reference inbred strains in response to an alkylating toxicant that is commonly used as a model for sporadic colorectal cancer. Results from this aim will be compared to known colorectal cancer response rates in these inbred strains. The fourth aim will be to develop strain-specific dose and time course expression profiles in response to toxicants widely used as chemotherapeutics. Samples from the three target tissues will also be evaluated for DNA damage.

Data collected in pursuit of the four specific aims will be databased and mined for unique signatures indicative of common responses to different classes of toxicants and for specific genetic background dependent responses. Furthermore, data from this project will be analyzed in conjunction with other projects from the toxicogenomics consortium to identify common species-dependent and-independent responses to particular chemotherapeutic toxicants. These results will provide valuable insights into individual response to toxicants and potential animal models for specific patterns of toxicant responses observed in humans.

 Back to to