Back

Office of Biological and Environmental Research

DOE Lowdose Radiation Program Workshop IV

Abstract

_____________________________________________________________________

Title: BIOLOGICALLY BASED MULTISTAGE MODELING OF RADIATION EFFECTS

Author: William Hazelton, PhD

Institution: Fred Hutchinson Cancer Research Center

Risks associated with exposure to low doses of ionizing radiation are highly uncertain. Recently,
considerable research effort has been devoted to improve our understanding of biological processes that mediate the low dose radiation response. For example, microbeam experiments have looked at the effects of a controlled number of alpha particles passing through targeted cells. This has led to a number of unexpected findings, including bystander effects with cell killing, mutation, and genetic instability occurring not only in cells with irradiated nuclei, but in cells where only the cytoplasm was traversed and also in neighbors of irradiated cells. Genetic instability leads to progressive chromosomal aberrations or nucleotide mutations. Adaptive response reduces the response to a high dose of radiation following a low dose within 24 hours or so. At very low doses, some cells appear to be hypersensitive to radiation.

Multistage clonal expansion models

We intend to incorporate these phenomena into stochastic models of carcinogenesis that include multiple events leading to initiation of stem or progenitor cells, followed by clonal expansion of initiated cells, and malignant conversion and progression to detection of lesions. These multistage clonal expansion (MSCE) models can be readily extended to incorporate genetic instability, bystander effects, adaptive response and low dose hypersensitivity. We anticipate that collaboration with experimentalists and other modelers in the workshops will be crucial in focusing on aspects of these phenomena to incorporate into the models that best capture the salient features while avoiding unnecessary detail.

Genetic instability

Genetic instability is characterized by progressively increasing rates for mutation and
chromosomal abnormalities that may occur over the span of many cell divisions. This may
increase rates for the mutations required for cancer, but may also lead to altered growth and
increasing cell death rates.

The MSCE model can be generalized to include genomic instability by allowing more than one
pathway along which normal cells undergo initiation, clonal expansion, and malignant
conversion. We model two types of mutations – those that cause progress on the pathway to
cancer, and also destabilizing mutations that link parallel cancer pathways. A transition to a new
cancer pathway leads to higher mutation rates, and possibly different cell division and death
rates. Both types of mutation are assumed to occur in the course of asymmetric cell divisions.

Bystander effect

The bystander effect describes the fact that non-irradiated cells that are neighbors of irradiated
cells often show altered mutation and cell killing rates in comparison to distant non-irradiated or
sham irradiated cells, and also in comparison to irradiated cells. These effects can be
incorporated in the MSCE modeling framework by allowing alternate parallel pathways for
irradiated, neighbors, and non-irradiated cells. The probabilities for taking a direct hit or being a
neighbor to a cell that is hit depend on the tissue structure, shielding, and other factors. Rates for
each pathway will be optimized to best represent the available data for cell death, division,
initiation, and oncogenic transformation that become available. Prior to irradiation, all cells are
in the non-irradiated pathway. During radiation exposure, transition probabilities would shift
weight into the other two pathways based on an appropriate physical model of the process.

Adaptive response and low dose hypersensitivity

Adaptive response occurs when a small radiation exposure (typically between 10 mSv and about
100 mSv) reduces cell killing and mutation in response to a subsequent high dose radiation
exposure. Adaption typically lasts for around 24 hours. We propose to represent low-dose
hypersensitivity and adaptive response in the MSCE modeling framework by incorporating
specific features in the dose response functions. Low dose hypersensitivity will be represented
as a modification to the dose response curves. For adaptive response, the shape of the dose
response functions affecting initiation, promotion, and malignant conversion will depend on
whether or not a dose occurred previously in the proper dose range and time window to act as a
priming dose. Thus the response will reflect the previous dose and the time since previous dose.

Combined cell cycle and multistage clonal expansion model

The strength of the adaptive response depends on the state of the cell cycle. We propose to study
this in more detail using an extension of the MSCE model that includes transitions between cell
cycle states in both normal and initiated stem cells. We plan to combine a stochastic version of
the cell cycle with the MSCE model to better represent the variable effects of radiation on
mutation, apoptosis and cell cycle arrest, as influenced by the cell cycle state.

Biological archetypes

We anticipate that a hierarchy of biological archetypes will be useful in determining parameters
of these detailed models, including cell cultures, 3-dimensional human cell clusters, cultured
human tissues, and epidemiological cohorts. Mechanistic models of these phenomena will be
optimized using data from in vitro experiments. These sub models will be incorporated as
components of in vivo models, allowing coefficients multiplying these effects to be optimized to
test for the significance of the phenomena when fitting epidemiological cohorts. Likelihood
methods will be used for model selection and validation at each level.