European Advances in Radiological Protection: EC-NRPB Agreement of Association Project Co-ordinator Reports for 1996-99, Contract Number F14P-CT95000

Synopsis

Summary:

A limited number of copies of the published report are available on request. There are also a limited number of copies available of the following reports published by NRPB for the European Commission.

• EUR 18456 European Advances in Radiological Protection. EC-NRPB Agreement of Association project coordinator mid-term review reports for 1996-1998. Contract number F14P-CT950008 (1998)
• EUR 17628 Radon Research in the European Union. K D Cliff and J C H Miles (eds). Contract number F13P-CT920064 (1997)
• EUR 16951 European Advances in Radiological Protection. CEC-NRPB Agreement of Association project coordinator reports for 1993-1995. Contract number F13P-CT920064 (1996)

The executive summary of the report is given below.

Executive summary

The EC-NRPB Association Agreement research contracts sought to clarify uncertainties in two distinct areas of radiological protection. The first concerned the fundamental processes of radiation-induced tumorigenesis and involved input from three groups of contractors. The second area concerned radiation effects on the developing brain before birth (in utero) and involved one group of contractors. As noted in all summaries of work, complex biological processes and radiation response parameters may be involved in tumour induction and in effects on the developing brain. A few of these processes are probably shared (principally DNA damage response) but most are fundamentally different. In both these areas our understanding of the risks to humans after radiation exposure is, however, limited by the power of epidemiological approaches to quantitatively resolve radiation effects at low doses. As a consequence, there continues to be a need to address the fundamental mechanisms that underlie these health effects in order to project epidemiological data obtained at relatively high doses to the low doses that are of prime importance in radiological protection. This brief overview considers some of the current uncertainties on radiation risk and the contribution to knowledge made by the research work of the contractors.

Radiation tumorigenesis

Empirical models that take little account of biological factors have principally been used to estimate tumour risk at low doses, to calculate life-time risk from epidemiological data based on incomplete follow-up, and to assess risks in different human populations. Biologically-based models are being developed and evaluated but, on many issues, the underlying concepts and parameters are not wholly secure. For example, questions remain on the specific tumour genes in cells that act as mutational targets for radiation, on the ways in which such gene mutations develop, and on the multistage sequence of mutations necessary for full radiation-induced malignancy to develop. Cellular factors in tumorigenesis such as cellular aging processes (senescence and immortalization), programmed cell death (apoptosis), and the onset of instability of the total DNA content of cells (the genome) also require clarification. The reports of the project leaders indicate that good progress was made in resolving uncertainties on some of these issues.

Mouse studies on the development of leukaemia and solid tumours provided good evidence in support of the views that:

• induced tumours are of monoclonal (single cell) origin.
• events associated with radiation damage tend to occur early in tumorigenesis.
• these induced events usually take the form of tumour-specific gene/chromosomal losses.
• radiation tumorigenesis tends to follow a multistage pathway without a characteristic radiation signature that can distinguish a given tumour.

These cytogenetic/molecular features apply most clearly to leukaemia and intestinal tumours but supporting data were also obtained from studies with lymphoma, osteosarcoma and radon-associated lung tumours in the rat.

Cytogenetic studies with radiotherapy-associated human solid tumours also supported a radiation mechanism driven by monoclonal genomic loss events. In the case of Chernobyl-associated human thyroid tumours the cytogenetic/molecular work confirmed monoclonality and the involvement of specific gene/chromosomal rearrangements of the same general form seen in spontaneous tumours of this type. One particular form of gene rearrangement in thyroid tumours ( ret/PTC3) appeared to be characteristic of tumour induction in early childhood. Other cellular, chromosomal and molecular studies in the area of thyroid carcinogenesis revealed evidence of multiple changes in gene expression, of patterns of chromosomal events and of the potential influence of the local structure of the chromosome on gene rearrangement.

Investigations on cellular senescence/immortalization processes in tumorigenesis succeeded in emphasizing the importance of specific controlling genes, of region-specific chromosomal instability/breakage and of the dynamic characteristics of chromosome termini (telomere) maintenance, chromosomal instability and clonal selection. The data generated gave clear evidence of human-rodent differences in cell senescence/immortalization processes which cautions against over-generalization of some concepts in radiation tumorigenesis. Some cellular studies also suggested that radiation acted to enhance senescence-associated chromosomal instability rather than inducing a novel process; in this context mouse studies also failed to establish a clear role for persistent chromosomal instability in radiation leukaemogenesis. Thus, uncertainties remain on the true status of induced genomic instability during post-irradiation tumorigenesis.

Other studies sought to clarify selected aspects of tumorigenesis associated with apoptosis and genomic stability. Using in vivo techniques it was shown that induced apoptosis and arrest of the cell division cycle in the mouse was dependent upon the p53 gene. Some p53 mutant cell lines did not recapitulate this result providing further evidence of p53-independent radiation response pathways; other in vitro studies revealed a close association between p53 mutation and chromosomal rearrangement/instability. A co-operative interaction between alpha-particle irradiation and retroviral infection was seen for in vitro transformation of murine bone forming cells and clonal chromosomal events were characterised in transformed glial cells cultured from rat brain.

The potential for heritable variation in human tumorigenic response is recognized in radiological protection but, at present, there is insufficient knowledge to take such genetic factors into specific account. Accordingly a major topic in two contracts was investigation of these genetic factors using mouse models of different tumour types. The p53-deficiency and Apc-deficiency models noted in the summaries are genetic homologues of strongly expressing human cancer-prone disorders. Work with these models yielded evidence that such mutational types (genotypes) express substantially elevated risk of lymphoma/sarcoma ( p53-deficiency) and intestinal tumours ( Apc-deficiency) after exposure to radiation; also that this risk may be modified by variation in the overall genetic constitution of the host mice. The presence of such natural genetic variation in tumour susceptibility in the mouse was revealed in other studies on bone tumours, skin tumours, leukaemia and lymphoma with good progress being made in locating the variant genes responsible for inter- strain differences in susceptibility. Together, these studies imply that variant genes with effects on radiation response are not uncommon and are usually specific to tumour-type.

Overall, the work of the three groups of contractors considering radiation tumorigenesis has made a significant contribution to knowledge. In respect of tumour induction mechanisms, the work tends to favour monoclonal multistage models with radiation acting principally to induce gene/chromosomal deletions at the earliest point in tumour development (initiation). The finding of characteristic chromosomal exchanges in radiation-associated human thyroid tumours suggests that a deletion-type mechanism for radiation tumorigenesis may not always apply. However the true status of these gene-specific exchanges has yet to be established. Other mechanistic features (apoptosis, senescence/immortalization and chromosomal stability) were clarified and none was suggestive of unexpected contributions from radiation damage. In general, the data obtained provide no good reason to question the view that, at low doses, tumorigenic risk will, in most circumstances, rise as a simple function of dose.

The genetic studies with mouse models succeeded in highlighting the potential importance for radiation risk of naturally variant genes that may be relatively common in mammalian populations. Although such work remains at an early stage of development, there are clear indications that the approaches developed by the contractors will be of substantial value in future studies. In particular, the work performed provides a good foundation for progress on gene identification. With continuing developments in mammalian genetics it will become possible to extend such genetic analysis to humans.

Brain development

Brain development is a protracted multistage process involving cell proliferation, migration, and specialization (differentiation); these sequential processes of development include selective apoptosis and growth together with cellular connection and sheathing (axonal growth, synaptogenesis and myelination). Most of the critical phases in this complex process occur in utero. Maximum sensitivity to ionising radiation occurs during cell division and migration within epithelial cell structures in the brain (neuroepithelium), between the 8th and 15th week of gestation in humans. During this time, the developing brain is composed of heterogeneous populations of cells with different developmental histories and fates, implying that radiation-induced damage even to a few cells lying within a very small region can affect a number of different areas in the mature brain.

Survivors of the atomic bombing in Japan who were exposed in utero during this sensitive period show a linear increase in the frequency of mental retardation with radiation dose (40% per Gy). For the purposes of radiological protection, the validity of a linear dose-response relationship without a threshold for radiation-induced damage to the brain is doubted but cannot be excluded. It seems more likely that the underlying mechanism for radiation-induced effects on brain development demand inactivation or malfunction of a critical number of cells (a deterministic-type effect). The participants in this contract investigated the effects of acute and protracted low dose irradiation on DNA damage and repair, apoptosis, gene expression, cell formation and migration, synaptogenesis and behaviour, focusing particularly on possible thresholds for radiation effects in animal models.

The work showed that doses as low as 10-25 cGy can induce a variety of potentially relevant cellular effects in the brain. Apoptosis was induced by irradiation both in vivo and in vitro, p53 and various genes involved in the control of cell growth and differentiation such as nerve growth factor were also induced by exposure. Altered patterns of cell proliferation and migration in the neuroepithelium led to long-term changes in the brain and its complex constituent structures (reduced brain weight, atrophy of the cingulum, decrease of cortical and commisural diameters, diminution of the hippocampus and misalignment of cortical Va neurons). In general, protracted exposures were less damaging than acute exposure, and the earlier the in utero exposure, the greater the effect.

However, long-term effects on complex, multifactorially determined processes such as animal behaviour are more difficult to define. Effects of prenatal irradiation on adult behaviour were seen at higher doses (around 1 Gy, although similar deficits have been reported following 35 cGy). Spatial learning, which is dependent upon the integrity of one particular brain structure (the hippocampus), was particularly sensitive. In addition, exposure later in gestation was more effective in inducing behavioural deficits, possibly the result of small but specific changes in the organisation of connections in brain areas involved in the control of a particular behaviour.

Overall, with regard to radiological protection, the results obtained through the experimental work carried out in this project show that during the period of development of cortical structures the brain is highly sensitive to radiation. Deterministic effects prevail during the initial phase of damage which may subsequently be modified by compensation within the brain. In these experimental studies the thresholds for radiation effects were in the range 10 to 25 cGy, approaching the lowest range of doses associated with mental retardation in the Japanese atomic bomb survivors. In general, these thresholds were shifted to higher values following exposure during later stages of gestation. Thus the fundamental research conducted in this area has made a valuable contribution to radiological protection by reinforcing the view that functionally significant radiation effects on the developing brain are most unlikely to occur at the low doses that apply to the vast majority of human in utero exposures.

Last reviewed: 30 October 2009