Renaissance Computing Institute

CI in Support of Research

Cyberinfrastructure in Support of Research Melanoma Project

Overview
Systemic biological models are essential to biological researchers who seek to assess the interactions and complex interplay between the environment and basic biological processes. Additionally, the growing number of public databases filled with genomic and biological function information means that researchers have access to an unprecedented amount of data useful for scientific discovery. The Cyberinfrastructure in Support of Research project aims to provide the technological infrastructure to support this kind of 21st century research. Its target audience is researchers who study the response of human DNA to exposure to genotoxins, such as radiation from prolonged exposure to the sun. The aim of these researchers is to create a biologically faithful model of the human systems of response to DNA damage that accounts for all the complex interactions among cell cycle checkpoints following exposure to carcinogenic agents.

A full understanding of these processes and interactions could lead to new strategies for protecting humans from genotoxins that trigger a DNA damage response in cells. The work could impact our understanding of cancer chemotherapy, human aging and birth defects, which are all influenced by DNA damage response. As the complex organization and structure of the systems of response to DNA damage are established, more effective therapies and intervention strategies can be designed.

The cyberinfrastructure challenge in interdisciplinary computational biology is to integrate tools and techniques now used in the physical sciences and engineering with biological and biomedical studies and models. Through the combined efforts of researchers in biology, biomedicine and biochemistry and experts in applied mathematics, computational science, and advanced modeling and data analysis, this project aims to create state-of-the-art simulations and models of the human system of response to DNA damage on NCSA supercomputing clusters.

The RENCI Contribution
RENCI has collaborated with Prof. William Kaufmann and his lab in the Linberger Cancer Center at the University of North Carolina at Chapel Hill to develop tools and software to integrate massive amounts of protein-protein interaction data found in biological databases (i.e., BIND, REACTOME, HPRD, CPATH) with gene expression data generated for this project and visualization tools. These tools will create a method for studying system-level response to DNA damage that is based on integrated genomic data. RENCI and the Kaufmann lab have also collaborated to devise computer models of cell division and DNA damage repair mechanisms.

Research Findings
To maintain genomic stability in normal human cells, cell cycle checkpoints are activated when DNA damage occurs due to internal or external causes. Mutations in DNA damage responsive genes may reduce the effectiveness of checkpoint functions designed keep cells functioning normally and this can lead to the development of cancer. To elucidate the patterns of gene regulation associated with cell cycle checkpoint functions, changes in global gene expression were determined using Agilent Human 22k arrays after treating normal human diploid fibroblast lines—cells that help to maintain normal structural framework—with four DNA-damaging agents: ionizing radiation (IR), 254 nm ultraviolet radiation (UV), doxorubicin (Dox) and cadmium chloride (Cd).

To determine the roles of ATM and p53 in gene regulation in response to DNA damage, similar studies were performed with ATM- or p53-deficient fibroblast lines after radiation treatment. Quantitative flow cytometric analyses of cell cycle compartments (G1, S, G2 and M) showed that IR, UV, Dox and Cd, at doses inhibiting clonal expansion by 40 percent, caused genotoxin-specific changes in cell cycle progression and checkpoint function. Growth arrest in G1 and G2 was associated with repression of large numbers of cell cycle-regulated and growth-associated genes as determined by gene ontology analysis. The genotoxins also produced unique time-dependent gene expression. For example, UV induced unique time-dependent patterns of change in a set of 44 genes at 6 h post irradiation that was not induced by IR. Gene ontology analysis of this UV-specific pattern identified RNA transcription as the biological process controlled by the set of co-regulated genes. As expected, treatment with IR caused significantly less change in gene expression in ATM- and p53-deficient cell lines. Less induction of DNA-damage responsive genes soon after IR treatment and less repression of cell cycle-regulated genes at a later time were associated with the defective cell cycle checkpoint functions in these cell lines.

Funding
National Science Foundation under Grant No. CA SCI 0525308 and CSA SCI 0525308 through the National Center for Super Computing Applications, University of Illinois-Urbana/Champaign.

Project Leaders

RENCI Team

Partners
Lineberger Comprehensive Cancer Center, UNC Chapel Hill
Center for Environmental Health and Susceptibility, UNC Chapel Hill
RENCI
National Center for Supercomputing Applications

Links
Melanoma Project Website
RENCI teams with Carolina Medical Researchers to Develop Better Bioinformatics Tools