Professor Munira Kadhim

The clinical outcome of radiation therapy is restricted due to acquired radio-resistance of a subpopulation of tumour cells that may cause tumour relapse and distant metastasis. While the effects of ionising radiation (IR) such as DNA damage and cell stress are well-documented, the potential role of IR in inducing invasive potential in cancer cells has not been broadly studied and aimed to be investigated in this study. MCF-7 cells irradiated with 0 Gy (control) or 2 Gy X-ray therapeutic dose of IR were assessed for cell viability, percentage of apoptotic cells, reactive oxygen species (ROS) levels, DNA fragmentation, Matrigel invasion, assessment of epithelial-mesenchymal transition (EMT) markers and Helix pomatia agglutinin (HPA) binding at 30 minutes, 4- or 24-hours post-IR. Reduction in cell viability, increase in apoptotic cells, ROS positive cells, and DNA fragmentation was observed, while functional invasiveness and EMT were exacerbated together with altered glycosylation in MCF-7 cells irradiated with 2 Gy X-ray compared to control cells. These findings indicate that despite the detrimental effects of 2 Gy X-ray IR on MCF-7 cells, a subpopulation of cells may have gained increased invasive potential. The exacerbated invasive potential may be attributed to enhanced EMT and altered glycosylation. Moreover, deregulation of Transforming growth factor beta (TGF-β) following IR may be one of the elements responsible for these changes, as it lies in the intersection of these invasion-promoting cell processes.The clinical outcome of radiation therapy is restricted due to acquired radio-resistance of a subpopulation of tumour cells that may cause tumour relapse and distant metastasis. While the effects of ionising radiation (IR) such as DNA damage and cell stress are well-documented, the potential role of IR in inducing invasive potential in cancer cells has not been broadly studied and aimed to be investigated in this study. MCF-7 cells irradiated with 0 Gy (control) or 2 Gy X-ray therapeutic dose of IR were assessed for cell viability, percentage of apoptotic cells, reactive oxygen species (ROS) levels, DNA fragmentation, Matrigel invasion, assessment of epithelial-mesenchymal transition (EMT) markers and Helix pomatia agglutinin (HPA) binding at 30 minutes, 4- or 24-hours post-IR. Reduction in cell viability, increase in apoptotic cells, ROS positive cells, and DNA fragmentation was observed, while functional invasiveness and EMT were exacerbated together with altered glycosylation in MCF-7 cells irradiated with 2 Gy X-ray compared to control cells. These findings indicate that despite the detrimental effects of 2 Gy X-ray IR on MCF-7 cells, a subpopulation of cells may have gained increased invasive potential. The exacerbated invasive potential may be attributed to enhanced EMT and altered glycosylation. Moreover, deregulation of Transforming growth factor beta (TGF-β) following IR may be one of the elements responsible for these changes, as it lies in the intersection of these invasion-promoting cell processes.

Along with the cells that are exposed to radiation, non-irradiated cells can unveil radiation effects as a result of intercellular communication, which are collectively defined as radiation induced bystander effects (RIBE). Exosome-mediated signalling is one of the core mechanisms responsible for multidirectional communication of tumor cells and their associated microenvironment, which may result in enhancement of malignant tumor phenotypes. Recent studies show that exosomes and exosome-mediated signalling also play a dynamic role in RIBE in cancer cell lines, many of which focused on altered exosome cargo or their effects on DNA damage. However, there is a lack of knowledge regarding how these changes in exosome cargo are reflected in other functional characteristics of cancer cells from the aspects of invasiveness and metastasis. Therefore, in the current study, we aimed to investigate exosome-mediated bystander effects of 2 Gy X-ray therapeutic dose of ionizing radiation on the invasive potential of MCF-7 breast cancer cells in vitro via assessing Matrigel invasion potential, EMT characteristics and the extent of glycosylation, as well as underlying plausible molecular mechanisms. The findings show that exosomes derived from irradiated MCF-7 cells enhance invasiveness of bystander MCF-7 cells, possibly through altered miRNA and protein content carried in exosomes.

Ionising radiation (IR) is commonly used for cancer therapy; however, its potential influence on the metastatic ability of surviving cancer cells exposed directly or indirectly to IR remains controversial. Metastasis is a multistep process, by which the cancer cells dissociate from the initial site, invade, travel through the blood stream or lymphatic system, and colonise distant sites. This complex process has been reported to require cancer cells to undergo epithelial-mesenchymal transition (EMT), by which the cancer cells convert from an adhesive, epithelial to motile, mesenchymal form, and is also associated with changes in glycosylation of cell surface proteins, which may be functionally involved in metastasis. In this paper, we give an overview of metastatic mechanisms and of the fundamentals of cancer-associated glycosylation changes. While not attempting a comprehensive review of this wide and fast moving field, we highlight some of the accumulating evidence from in vitro and in vivo models for increased metastatic potential in cancer cells that survive IR, focusing on angiogenesis, cancer cell motility, invasion and EMT and glycosylation. We also explore the indirect effects in cellsexposed to exosomes released from irradiated cells. The results of such studies need to be interpreted with caution and there remains limited evidence that radiotherapy enhances metastatic capacity of cancers in a clinical setting, and undoubtedly has a very positive clinical benefit. However, there is potential that this therapeutic benefit may ultimately be enhanced through a better understanding the direct and indirect effects of IR on cancer cell behaviour.

Breast cancer is the most commonly occurring cancer in women worldwide and the second most common cancer overall. The development of new therapies to treat this devastating malignancy is needed urgently. Nanoparticles are one class of nanomaterial with multiple applications in medicine, ranging from their use as drug delivery systems and the promotion of changes in cell morphology to the control of gene transcription. Nanoparticles made of the natural polymer chitosan are easy to produce, have a very low immunogenic profile, and diffuse easily into cells. One hallmark feature of cancer, including breast tumours, is the genome instability caused by defects in the spindle-assembly checkpoint (SAC), the molecular signalling mechanism that ensures the timely and high-fidelity transmission of the genetic material to an offspring. In recent years, the use of nanoparticles to treat cancer cells has gained momentum. This is in part because nanoparticles made of different materials can sensitise cancer cells to chemotherapy and radiotherapy. These advances prompted us to study the potential sensitising effect of chitosan-based nanoparticles on breast cancer cells treated with reversine, which is a small molecule inhibitor of Mps1 and Aurora B that induces premature exit from mitosis, aneuploidy, and cell death, before and after exposure of the cancer cells to X-ray irradiation. Our measurements of metabolic activity as an indicator of cell viability, DNA damage by alkaline comet assay, and immunofluorescence using anti-P-H3 as a mitotic biomarker indicate that chitosan nanoparticles elicit cellular responses that affect mitosis and cell viability and can sensitise breast cancer cells to X-ray radiation (2Gy). We also show that such a sensitisation effect is not caused by direct damage to the DNA by the nanoparticles. Taken together, our data indicates that chitosan nanoparticles have potential application for the treatment of breast cancer as adjunct to radiotherapy.

Here we report an in vivo study investigating the role of EVs in mediating radiation effects. C57BL/6 mice were total-body irradiated with X-rays (0.1, 0.25, 2 Gy), 24 hours later EVs were isolated from the bone marrow and were intravenously injected into unirradiated (so-called bystander) animals. EV-induced systemic effects were compared to radiation effects in the directly irradiated animals. Similarly to direct radiation EVs from irradiated mice induced complex DNA damage in EV-recipient animals, manifested in an increased level of chromosomal aberrations and the activation of the DNA damage response. However, while DNA damage after direct irradiation increased with the dose, EV-induced effects peaked at lower doses. A significantly reduced hematopoietic stem cell pool in the BM as well as CD4+ and CD8+ lymphocyte pool in the spleen was detected in mice injected with EVs isolated from animals irradiated with 2 Gy. These EV-induced alterations were comparable to changes present in the directly irradiated mice. The pool of TLR4-expressing dendritic cells was different in the directly irradiated mice, where it increased after 2 Gy and in the EV-recipient animals, where it strongly decreased in a dose-independent manner. 
 

A panel of 8 differentially expressed miRNAs were identified in the EVs originating from both low and high dose-irradiated mice, with a predicted involvement in pathways related to DNA damage repair, hematopoietic and immune system regulation, suggesting a direct involvement of these pathways in mediating radiation-induced systemic effects. 
 

In conclusion, we proved the role of EVs in transmitting certain radiation effects, identified miRNAs carried by EVs potentially responsible for these effects and showed that the pattern of changes was often different in the directly irradiated and EV-recipient bystander mice, suggesting different mechanisms.

Recently we demonstrated radiation-induced non-targeted effects [NTE: genomic instability (GI) and bystander effects (BE)] are partially mediated by exosomes, particularly the RNA content. However the mechanistic role of exosomes in NTE is yet to be fully understood.

The present study used MCF7 cells to characterise the longevity of exosome-induced activity in the progeny of irradiated and unirradiated bystander cells. Exosomes extracted from conditioned media of irradiated and bystander progeny were added to unirradiated cells. Analysis was carried out at 1 and 20/24 population doublings following medium/exosome transfer for DNA/chromosomal damage. Results confirmed exosomes play a significant role in mediating NTE of ionising radiation (IR). This effect was remarkably persistent, observed >20 doublings post-irradiation in the progeny of bystander cells. Additionally, cell progeny undergoing a BE were themselves capable of inducing BE in other cells via exosomes they released. Furthermore we investigated the role of exosome cargo. Culture media from cells exposed to 2 Gy X-rays was subjected to ultracentrifugation and four inoculants prepared, (a) supernatants with exosomes removed, and pellets with (b) exosome proteins denatured, (c) RNA degraded, and (d) a combination of protein–RNA inactivation. These were added to separate populations of unirradiated cells. The BE was partially inhibited when either exosome protein or exosome RNA were inactivated separately, whilst combined RNA–protein inhibition significantly reduced or eliminated the BE. These results demonstrate that exosomes are associated with long-lived signalling of the NTE of IR. Both RNA and protein molecules of exosomes work in a synergistic manner to initiate NTE, spread these effects to naïve cells, and perpetuate GI in the affected cells.

Organisms are often exposed to environmental pressures that affect homeostasis, so it is important to understand the biological basis of stress-response. Various biological mechanisms have evolved to help cells cope with potentially cytotoxic changes in their environment. miRNAs are small non-coding RNAs which are able to regulate mRNA stability. It has been suggested that miRNAs may tip the balance between continued cytorepair and induction of apoptosis in response to stress. There is a wealth of data in the literature showing the effect of environmental stress on miRNAs, but it is scattered in a large number of disparate publications. Meta-analyses of this data would produce added insight into the molecular mechanisms of stress-response. To facilitate this we created and manually curated the miRStress database, which describes the changes in miRNA levels following an array of stress types in eukaryotic cells. Here we describe this database and validate the miRStress tool for analysing miRNAs that are regulated by stress. To validate the database we performed a cross-species analysis to identify miRNAs that respond to radiation. The analysis tool confirms miR-21 and miR-34a as frequently deregulated in response to radiation, but also identifies novel candidates as potentially important players in this stress response, including miR-15b, miR-19b, and miR-106a. Similarly, we used the miRStress tool to analyse hypoxia-responsive miRNAs. The most frequently deregulated miRNAs were miR-210 and miR-21, as expected. Several other miRNAs were also found to be associated with hypoxia, including miR-181b, miR-26a/b, miR-106a, miR-213 and miR-192. Therefore the miRStress tool has identified miRNAs with hitherto unknown or under-appreciated roles in the response to specific stress types. The miRStress tool, which can be used to uncover new insight into the biological roles of miRNAs, and also has the potential to unearth potential biomarkers for therapeutic response, is freely available at https://mudshark.brookes.ac.uk/MirStress.

Non-DNA targeted effects of ionising radiation, which include genomic instability, and a variety of bystander effects including abscopal effects and bystander mediated adaptive response, have raised concerns about the magnitude of low-dose radiation risk. Genomic instability, bystander effects and adaptive responses are powered by fundamental, but not clearly understood systems that maintain tissue homeostasis. Despite excellent research in this field by various groups, there are still gaps in our understanding of the likely mechanisms associated with non-DNA targeted effects, particularly with respect to systemic (human health) consequences at low and intermediate doses of ionising radiation. Other outstanding questions include links between the different non-targeted responses and the variations in response observed between individuals and cell lines, possibly a function of genetic background. Furthermore, it is still not known what the initial target and early interactions in cells are that give rise to non-targeted responses in neighbouring or descendant cells. This paper provides a commentary on the current state of the field as a result of the non-targeted effects of ionising radiation (NOTE) Integrated Project funded by the European Union. Here we critically examine the evidence for non-targeted effects, discuss apparently contradictory results and consider implications for low-dose radiation health effects.

Radiation-induced adaptive response (AR) is a phenomenon of increased radioresistance mediated by a low priming dose of ionizing radiation (IR) applied prior to a higher challenging dose. We have previously shown that in mouse-embryo fibroblasts (MEF) and human A549 cells, AR is associated with enhanced repair of DNA double-strand breaks (DSB) by the DNA-PK-dependent pathway of non-homologous end-joining (D-NHEJ). Importantly, AR was 'transmitted' to non-irradiated bystander cells through transfer of medium from cells that had received a priming dose of IR. Here, we examine the influence of the genetic background in these responses.

MATERIALS AND METHODS:

Two plasmid-based assays specifically designed to measure the efficiency of NHEJ and HRR (homologous recombination repair) were deployed. MEF and the primary human fibroblast cell lines HF12 and HF19 were exposed to 10 mGy to 5 Gy X-rays. Bystander effects were investigated using the medium-transfer technique.

RESULTS:

In contrast to MEF, which induce robust AR to NHEJ, even as a bystander response, human fibroblasts fail to develop such phenomena.

CONCLUSIONS:

The development of AR is cell-type-specific. The same holds true for the development of AR as a bystander effect. Better understanding of the underlying mechanisms will help to understand the molecular basis of these differences in response.

Communication between irradiated and un-irradiated (bystander) cells can cause damage in cells that are not directly targeted by ionizing radiation, a process known as the bystander effect. Bystander effects can also lead to chromosomal/genomic instability within the progeny of bystander cells, similar to the progeny of directly irradiated cells. The factors that mediate this cellular communication can be transferred between cells via gap junctions or released into the extracellular media following irradiation, but their nature has not been fully characterized. In this study we tested the hypothesis that the bystander effect mediator contains an RNA molecule that may be carried by exosomes. MCF7 cells were irradiated with 2 Gy of X rays and the extracellular media was harvested. RNase treatment abrogated the ability of the media to induce early and late chromosomal damage in bystander cells. Furthermore, treatment of bystander cells with exosomes isolated from this media increased the levels of genomic damage. These results suggest that the bystander effect, and genomic instability, are at least in part mediated by exosomes and implicate a role for RNA.

PURPOSE:

This work investigates the hypothesis that genetic background plays a significant role in the signalling mechanisms underlying induction and perpetuation of genomic instability following radiation exposure.

MATERIALS AND METHODS:

Bone marrow from two strains of mice (CBA and C57) were exposed to a range of X-ray doses (0, 0.01, 0.1, 1 and 3 Gy). Different cellular signalling endpoints: Apoptosis, cytokine levels and calcium flux, were evaluated at 2 h, 24 h and 7 d post-irradiation to assess immediate and delayed effects.

RESULTS:

In CBA (radiosensitive) elevated apoptosis levels were observed at 24 h post X-irradiation, and transforming growth factor-β (TGF-β) levels which increased with time and dose. C57 showed a higher background level of apoptosis, and sustained apoptotic levels 7 days after radiation exposure. Levels of tumor necrosis factor-α (TNF-α were increased in C57 at day 7 for higher X-ray doses. TGF-β levels were higher in CBA, whilst C57 exhibited a greater TNF-α response. Calcium flux was induced in reporter cells on exposure to conditioned media from both strains.

CONCLUSIONS:

These results show genetic and dose specific differences in radiation-induced signalling in the initiation and perpetuation of the instability process, which have potential implications on evaluation of non-targeted effects in radiation risk assessment.

One of the key issues of current radiation research is the biological effect of low doses. Unfortunately, low dose science is hampered by the unavailability of easily performable, reliable and sensitive quantitative biomarkers suitable detecting low frequency alterations in irradiated cells. We applied a quantitative real time polymerase chain reaction (qRT-PCR) based protocol detecting common deletions (CD) in the mitochondrial genome to assess direct and non-targeted effects of radiation in human fibroblasts. In directly irradiated (IR) cells CD increased with dose and was higher in radiosensitive cells. Investigating conditioned mediummediated bystander effects we demonstrated that low and high (0.1, 2 Gy) doses induced similar levels of bystander responses and found individual differences in human fibroblasts. The bystander response was not related to the radiosensitivity of the cells. The importance of signal sending donor and signal receiving target cells was investigated by placing conditioned medium from a bystander response positive cell line (F11-hTERT) to bystander negative cells (S1-hTERT) and vice versa. The data indicated that signal sending cells are more important in the medium-mediated bystander effect than recipients. Finally, we followed long term effects in immortalized radiation sensitive (S1-hTERT) and normal (F11-hTERT) fibroblasts up to 63 days after IR. In F11-hTERT cells CD level was increased until 35 days after IR then reduced back to control level by day 49. In S1-hTERT cells the increased CD level was also normalized by day 42, however a second wave of increased CD incidence appeared by day 49 which was maintained up to day 63 after IR. This second CD wave might be the indication of radiation-induced instability in the mitochondrial genome of S1-hTERT cells. The data demonstrated that measuring CD in mtDNA by qRT-PCR is a reliable and sensitive biomarker to estimate radiation-induced direct and non-targeted effects.

Increasing evidence indicates that radiationinduced genomic instability plays an important role in the development of cancer. However, radiation quality and genetic background can influence the outcome. The goal of this study was to quantify radiation-induced changes in lymphocyte populations in mouse strains known to differ in susceptibility to genomic instability (C57BL/6, resistant; CBA/Ca, susceptible). The effects of whole-body exposure to ?-rays and protons, with and without aluminum shielding, were compared. Total radiation doses of 0, 0.1, 0.5, and 2.0 Gy were delivered and subsets of mice from each group were euthanized on days 1 and 30 after exposure for spleen and bone marrow analyses. In the spleen on day 1, lymphocyte counts were decreased (p

Adaptive response (AR) is a term describing resistance to ionizing radiation-induced killing or formation of aberrant chromosomes that is mediated by pre-exposure to low ionizing radiation doses. The mechanism of AR remains elusive. Because cell killing and chromosome aberration formation derive from erroneous processing of DNA double-strand breaks (DSB), AR may reflect a modulation of DSB processing by nonhomologous end joining (NHEJ) or homologous recombination repair. Here, we use plasmid end-joining assays to quantify modulations induced by low ionizing radiation doses to NHEJ, the dominant pathway of DSB repair in higher eukaryotes, and investigate propagation of this response through medium transfer to nonirradiated bystander cells. Mouse embryo fibroblasts were conditioned with 10 to 1000 mGy and NHEJ quantified at different times thereafter by challenging with reporter plasmids containing a DSB. We show robust increases in NHEJ efficiency in mouse embryo fibroblasts exposed to ionizing radiation > 100 mGy, irrespective of reporter plasmid used. Human tumor cells also show AR of similar magnitude that is compromised by caffeine, an inhibitor of DNA damage signaling acting by inhibiting ATM, ATR, and DNA-PKcs. Growth medium from pre-irradiated cells induces a caffeine-sensitive AR in nonirradiated cells, similar in magnitude to that seen in irradiated cells. In bystander cells, gamma H2AX foci are specifically detected in late S-G(2) phase and are associated with Rad51 foci that signify the function of homologous recombination repair, possibly on DNA replication-mediated DSBs. The results point to enhanced NHEJ as a mechanism of AR and suggest that AR may be transmitted to bystander cells through factors generating replication-mediated DSBs. Cancer Res; 70(21); 8498-506. (C)2010 AACR.

Environmental (222) radon exposure is a human health concern, and many studies demonstrate that very low doses of high LET alpha-particle irradiation initiate deleterious genetic consequences in both radiated and non-irradiated bystander cells. One consequence, radiation-induced genomic instability (RIGI), is a hallmark of tumorigenesis and is often assessed by measuring delayed chromosomal aberrations We utilised a technique that facilitates transient immobilization of primary lymphocytes for targeted microbeam irradiation and have reported that environmentally relevant doses, e.g. a single (3)He(2+) particle traversal to a single cell, are sufficient to Induce RIGI Herein we sought to determine differences in radiation response in lymphocytes isolated from five healthy male donors Primary lymphocytes were irradiated with a single particle per cell nucleus. We found evidence for inter-individual variation in radiation response (Rid, measured as delayed chromosome aberrations) Although this was not highly significant, it was possibly masked by high levels of intra-individual variation While there are many studies showing a link between genetic predisposition and RIGI, there are few studies linking genetic background with bystander effects in normal human lymphocytes In an attempt to investigate inter-individual variation in the induction of bystander effects, primary lymphocytes were irradiated with a single particle under conditions where fractions of the population were traversed We showed a marked genotype-dependent bystander response in one donor after exposure to 15% of the population The findings may also be regarded as a radiation-induced genotype-dependent bystander effect triggering an instability phenotype

Over the past two decades, our understanding of radiation biology has undergone a fundamental shift in paradigms away from deterministic ‘hit-effect’ relationships and towards complex ongoing ‘cellular responses’. These responses include now familiar, but still poorly understood, phenomena associated with radiation exposure such as genomic instability and bystander effects. Although these responses share some common features (e.g. they occur at high frequency following very low doses, are heterogeneous in their induction and are observed at time points far removed from the initial radiation exposure), the precise relationship between genomic instability and bystander effects remains to be elucidated. This review will provide a synthesis of the known, and proposed, interrelationships among irradiated and bystander cellular responses to radiation. It also discusses our current experimental approach for gaining a clearer understanding of the relationship between damage induction and long-term effects in both irradiated and bystander cells.

Genomic instability is effectively induced by ionizing radiation. Recently, evidence has accumulated supporting a relationship between genetic background and the radiation-induced genomic instability phenotype. This is possibly due to alterations in proteins responsible for maintenance of genomic integrity or altered oxidative metabolism. Studies in human cell lines, human primary cells, and mouse models have been performed predominantly using high linear energy transfer (LET) radiation, or high doses of low LET radiation. The interplay between genetics, radiation response, and genomic instability has not been fully determined at low doses of low LET radiation. However, recent studies using low doses of low LET radiation suggest that the relationship between genetic background and radiation-induced genomic instability may be more complicated than these same relationships at high LET or high doses of low LET radiation. The complexity of this relationship at low doses of low LET radiation suggests that more of the population may be at risk than previously recognized and may have implications for radiation risk assessment.

We have demonstrated chromosomal instability in the clonal descendants of hemopoietic stem cells after irradiating murine bone marrow with alpha-particles. However, because cells that are irradiated by alpha-particles are defined by a Poisson distribution of individual particle traversals, there is an inevitable proportion of unirradiated cells in the surviving population. The calculated expected proportions of irradiated and nonirradiated cells indicate that the number of clonogenic cells transmitting chromosomal instability is greater than the number expected to be hit and survive. To investigate further this discrepancy, we studied the effects of interposing a grid between the cells and the alpha-particle source so that the surviving population consists predominantly of untraversed stem cells. Comparison with the same irradiation conditions without the grid reveals that the same level of instability is induced. The data confirm that alpha-particles induce chromosomal instability but instability is demonstrated in the progeny of nonirradiated stem cells and must be due to unexpected interactions between irradiated and nonirradiated cells. This untargeted effect has important implications for mechanistic studies of radiation action and for assessment of radiation risk.

alpha-particles, which are ionising radiation of high linear-energy-transfer emitted, for example, from radon or plutonium, pass through tissue as highly structured tracks. Single target cells in the path of the tracks might be damaged by even low-dose alpha-irradiation. We found non-clonal cytogenetic aberrations, characterised by a high frequency of chromatid aberrations with chromosome aberrations, in clonal descendants of haemopoietic stem cells after exposure to alpha-particles of bone marrow cells from two of four haematologically normal individuals (up to 25% abnormal metaphases). The data are consistent with a transmissible genetic instability induced in a stem cell resulting in a diversity of aberrations in its clonal progeny many cell divisions later.

When investigating the biological effects of ionizing radiation on the haemopoietic system, a confounding problem lies in possible differences between the biological effects of sparsely ionizing, low linear energy transfer radiation such as X-, beta- or gamma-rays, and densely ionizing, high linear energy transfer radiation such as alpha-particles. To address this problem we have developed novel techniques for studying haemopoietic cells irradiated with environmentally relevant doses of alpha-particles from a plutonium-238 source. Using a clonogenic culture system, cytogenetic aberrations in individual colonies of haemopoietic cells derived from irradiated stem cells have been studied. Exposure to alpha-particles (but not X-rays) produced a high frequency of non-clonal aberrations in the clonal descendants, compatible with alpha-emitters inducing lesions in stem cells that result in the transmission of chromosomal instability to their progeny. Such unexpected instability may have important implications for radiation leukaemogenesis.