In vivo evolution of tumour cells after the generation of double-strand DNA breaks

Cell death following exposure to toxic agents can occur through several routes, including apoptosis, extended or permanent cell cycle arrest, necrosis and ‘mitotic catastrophe’. These various types of cell death display different morphological changes and biochemical characteristics. Such variations can result from differences between cell types and the mechanisms by which cell death is triggered. At the cellular level, necrosis is characterised by an increased plasma membrane permeability, cell swelling, a decline in protein synthesis and autolysis (Wyllie et al, 1980; Schwartz and Osborne, 1993). It can be observed in solid tumours of mice treated with arsenic trioxide (Lew et al, 1999). Apoptotic cell death plays a major role in the regulation of cell growth in multicellular organisms. It constitutes a systematic means of cell suicide during normal embryonic development and morphogenesis (von Wangenheim and Peterson, 1998), aging (Monti et al, 1992) or in response to pathogenic infections and other irreparable cell damage. Induction of apoptosis by anticancer drugs has been demonstrated using, for example, paclitaxel (Milross et al, 1996) or cisplatin (Trimmer and Essigmann, 1999). Recently, after cell treatment with ionising radiation, a ‘premitotic’ and a ‘postmitotic’ apoptosis could be distinguished as a function of the morphological features of the treated cells (Shinomiya et al, 2000). Apoptosis requires the activation of proteases and endonucleases (Wyllie et al, 1984; Anderson et al, 1997). However, Tounekti et al (1993) showed that, in vitro, cell uptake of large quantities of BLM, a nonpermeant drug that generates double-strand DNA breaks (DSB), results in the very rapid generation of the apoptotic morphological changes, as well as in a DNA degradation similar to that observed in usual apoptosis. This evolution was termed pseudoapoptosis (Tounekti et al, 1993) because it is caused by the DSB generated by the bleomycin (BLM) and does not require induction of proteases nor endonucleases involved in typical apoptosis: BLM itself acts as a microendonuclease (Tounekti et al, 1995). In vitro, mitotic cell death is a slow process in which cell metabolism remains functional for a period equivalent to more than two cell cycles, the cell death being noticeable only after the passage of the cell through one or more consecutive mitoses (Chang and Little, 1991; Radford, 1991). These mitoses are, nevertheless, aberrant and do not result in the formation of two daughter cells (Dewey et al, 1970; Joshi et al, 1982). This results in the generation of small abortive colonies by otherwise clonogenic cells. Mitotic cell death can be initiated by the generation of a few unrepaired DSB like those generated by small doses of ionising radiation or the internalisation of low amounts of radiomimetic drugs like BLM (Tounekti et al, 1993, 2001; Yanagihara et al, 1995). Thus, in vitro, the number of DSB seems to determine the cell death pathway. Single-strand DNA breaks (SSB) in very large numbers also result in cell death (Yoshida et al, 1993; Cohen-Jonathan et al, 1999). We also reported that not only the absolute numbers of DSB or SSB but also their ratio determine cell death pathway: true apoptosis, pseudoapoptosis or mitotic cell death (Tounekti et al, 2001). On these basis, we decided to perform a systematic analysis of the consequences of the generation of different numbers of DSB in tumour cells in vivo. Control of DSB number, as well as synchronisation of the DSB generation was achieved using different doses of BLM and cell permeabilisation by means of locally delivered electric pulses (EPs). Indeed, (a) external electric fields generate changes in the transmembrane voltage and, under appropriate conditions, provoke the transient and reversible permeabilisation of the cells, in vitro as well as in vivo and (b) BLM molecules do not diffuse through the plasma membrane and they enter only at the time of the cell electropermeabilisation, in amounts proportional to the external concentration at the moment of EP. Moreover, permeabilising EP can be applied locally in vivo on tumour nodules: the huge increase in the antitumour effectiveness of BLM achieved by the increase in BLM uptake is the basis of electrochemotherapy (ECT), a new way to treat solid tumours that is already under clinical evaluation (Domenge et al, 1996; Mir et al, 1998; Gehl and Geertsen, 2000; Rodriguez-Cuevas et al, 2001; Rols et al, 2002). Therefore, it seemed possible to study in vivo the cell death pathway caused by different amounts of DSB using these experimental conditions. Indeed, in ECT, as performed in our preclinical and clinical studies, EPs alone are not toxic and the toxicity observed is actually the toxicity caused by the facilitated uptake of BLM.