Multi-Level Kinetic and Dynamic Biologically-Based Risk Models for Low Dose Radiation

William Griffith
University of Washington
Seattle, WA 98105
E-mail: Griffith@u.washington.edu
Website: http://depts.washington.edu/chc/riskchar.html

Why the Project?

Major revisions of our understanding of the mechanisms of how ionizing radiation causes effects have occurred in the last decade. This is due to accumulating evidence that radiation effects, called bystander effects, can occur in cells distant from those being irradiated. Observations of mutations and other changes in cells not irradiated have been made in many types of cell systems and suggest that bystander effects are due to secreted factors, although specific factors in general have not been identified. The secreted factors have been shown to be transmissible through both gap junctions and extra cellular media.

Project Goals:

In this work we integrate many of these observations in quantitative models of the secreted factors to better understand how the secreted factors may operate and how to better extrapolate in vitro observations to in vivo systems. We use a value of information approach to identify how the models may be improved by the collection of additional types of observations and data sets.

Experimental Approach:

Our basic model treats each cell as a separate compartment with the gap junctions described as first order rates in a two dimensional geometry, and the soluble factor is modeled as diffusing through the media from the target cell where it is absorbed by the bystander cells from the media. Biological effects are represented as probabilities where the probability depends upon the concentration of the secreted factor in the cell. The utility of this type of model is to compare similar types of bystander effects across cell systems and to characterize ranges of parameter values. These types of models may be useful as a component in an integration of multiple biological models to provide estimates for risk assessment. The major limitation in developing these types of models is that bystander effects studies for most biological endpoints do not support estimation of even a minimal number of parameters needed for a model of secreted factors. This indicates that the value of other types of studies would be to characterize the parameters needed for models of secreted factors in bystander effects studies.

Expected Outcome:

This model is of utility for risk assessment as it allows for the evaluation of alterations in specific developmental dynamic processes across times and doses. It also emphasizes the importance of evaluating the long-term potential for radiation effects at low doses.