Radiation Leukaemogenesis at Low Doses DE-FG02-05 ER 63947 Final Technical Report 15 May 2005 - 14 May 2010.

摘要

This report provides a complete summary of the work undertaken and results obtained under US Department of Energy grant DF-FG02-05 ER 63947, Radiation leukaemogenesis at low doses. There is ample epidemiological evidence indicating that ionizing radiation is carcinogenic in the higher dose range. This evidence, however, weakens and carries increasing uncertainties at doses below 100-200 mSv. At these low dose levels the form of the dose-response curve for radiation-induced cancer cannot be determined reliably or directly from studies of human populations. Therefore animal, cellular and other experimental systems must be employed to provide supporting evidence on which to base judgements of risk at low doses. Currently in radiological protection a linear non-threshold (LNT) extrapolation of risk estimates derived from human epidemiological studies is used to estimate risks in the dose range of interest for protection purposes. Myeloid leukaemias feature prominently among the cancers associated with human exposures to ionising radiation (eg UNSCEAR 2006; IARC 2000). Good animal models of radiation-induced acute myeloid leukaemia (AML) are available including strains such as CBA, RFM and SJL (eg Major and Mole 1978; Ullrich et al 1976; Resnitzky et al 1985). Early mechanistic studies using cytogenetic methods in these mouse models established that the majority of radiation-induced AMLs carried substantial interstitial deletions in one copy of chromosome (chr) 2 (eg Hayata et al 1983; Trakhtenbrot et al 1988; Breckon et al 1991; Rithidech et al 1993; Bouffler et al 1996). Chr2 aberrations are known to occur in bone marrow cells as early as 24 hours after in vivo irradiation (Bouffler et al 1997). Subsequent molecular mapping studies defined a distinct region of chr2 that is commonly lost in AMLs (Clark et al 1996; Silver et al 1999). Further, more detailed, analysis identified point mutations at a specific region of the Sfpi1/PU.1 haemopoietic transcription factor gene which lies in the commonly deleted region of chr2 (Cook et al 2004; Suraweera et al 2005). These lines of evidence strongly implicate the Sfpi1/PU.1 gene as a tumour suppressor gene, dysregulation of which leads to myeloid leukaemia. The main focus of this project was to utilize the CBA mouse model of radiation leukaemogenesis to explore mechanisms of low dose and low dose-rate leukaemogenesis. A series of mechanistic investigations were undertaken, the central aim of which was to identify the events that convert normal cells into myeloid leukaemia cells and explore the dose-response relationships for these. Much of the work centred on the Sfpi1/PU.1 gene and its role in leukaemogenesis. Specific studies considered the dose-response and time-course relationships for loss of the gene, the functional consequences of Sfpi1/PU.1 loss and mutation on transcriptional programmes and developing an in vivo reporter gene system for radiation-induced alterations to PU.1 expression. Additional work sought further genetic changes associated with radiation-induced AMLs and a better characterization of the cell of origin or 'target cell' for radiation-induced AML. All the information gathered is of potential use in developing biologically realistic mathematical models for low dose cancer risk projection.