Abstracts

Genome Wide Evaluation of Normal Human Tissue in Response to Controlled, In vivo Low-Dose Low LET Ionizing Radiation Exposure: Pathways and Mechanisms P.I.: Z. Goldberg

Abstract

The true risks of low dose ionizing radiation (LDIR) are fundamentally undefined in the human. It remains unproven that the molecular level information gathered through in vitro experiments can be used to assess human tissue risks. Similarly, more complex tissue-like laboratory models must also be validated against actual human responses. Thus, human translational data must be obtained against which to correlate in vitro findings to evaluate their "real-life" applicability to allow the development of scientifically sound public policy on safe exposure limits to LDIR.

This project, started in September 2001, has begun to address this information gap by developing a benchmark data set of normal human tissue response to low linear energy transfer (LET) low dose ionizing radiation when exposure occurs in vivo, and thus represents the first direct testing of the linear no-threshold model of radiation effects. This project will continue and expand the patient-based genomic studies from the first granting period producing a more detailed temporal dose response pattern, with larger biopsy sizes allowing more complete evaluation of "best fit modeling" of radiation effects in the low dose range (i.e., possible non-linearity). Further, we will evaluate a laboratory 3-D skin model to provide crucial validation of the model. Finally, we will complete mechanistic studies in the 3D skin model and in human keratinocytes and fibroblasts, focusing on the pro-inflammatory genes identified in the genomic studies in the first granting period to test the following hypotheses:

Hypotheses

The complexity of human tissue low dose radiation response has a definable biosignature despite inter-person variability, and this signature is not duplicated in in vitro model systems. However, we hypothesize that the strongest signals from the human skin genome wide analysis will be seen in the models, and the relationship from the actual human tissue to the in vitro model can be determined quantitatively for key radiation responsive gene groups, thus validating the in vitro model for mechanistic studies. Further, we hypothesize that tissue level expression of radiation response is highly regulated by the epithelial-mesenchymal interaction, as a particular type of cell-cell dialogue/bystander effect.

We will test this through 3 specific aims: (1) To define characteristic gene expression profile changes in non-cancerous human cells following LDIR exposure in the <1-10 cGy range by comparing comprehensive gene expression in human skin samples before and after LDIR exposure obtained during therapeutic irradiation. In this specific aim we will build upon our existing database of single fraction exposures and develop a more comprehensive time-course dataset, collecting samples between 12-24 hours post exposure; (2) To validate the 3D skin model (EpiDermFT, MatTek) by characterizing its genomic responses to low dose IR exposures against the known responses of actual human skin irradiated in vivo; and (3) To investigate the role that the stromal/fibroblast layer may play in dampening the IR induced inflammatory responses in the epithelial component of skin (ie, the cross-tissue bystander effect). These studies will examine the epithelial and stromal tissue components separately, and their interactions, using the EpiDermFT skin model to elucidate molecular signaling/intercellular cross talk that protect tissues from expression of IR damage, with a focus on inflammatory/cox-2 signaling. Analysis of the role of specific transcripts with media transfer/modulation experiments with selective gene silencing in primary human keratinocytes and fibroblasts will also be completed.

These specific aims will extend the landmark dataset generated for OBER/DOE in the first granting period and begin to translate those findings into pathways of mechanistic evaluation in a laboratory model validated against actual human response.

» Return to the top