Haematopoiesis involves commitment, proliferation and differentiation of a common progenitor into several specialised blood cell types that participate in a wide range of functions such as host defence against infection and haemostasis. Up-regulation of the haematopoietic system plays a critical role in establishing the proper response against invading pathogens, in removing cancerous cells or in tissue remodelling during normal development. Furthermore, deregulations of the haematopoietic differentiation program are at the origin of numerous diseases including leukaemia and lymphoma. Unravelling the genetic circuitry and the molecular mechanisms that control normal and pathological blood cell formation and differentiation is a major challenge. Recent evidence has shown that several features of haematopoietic development are conserved from Drosophila to vertebrates. Notably, transcription factors of the RUNX and GATA families, which play key roles in haematopoiesis in vertebrates, also control blood cell development in Drosophila. Taking advantage of this phylogenetic conservation and of the powerful genetic tools available in Drosophila, we propose to use Drosophila haematopoiesis as a paradigm to study the function and molecular mechanisms of action of the GATA and RUNX transcription factors and to characterise new genes controlling the conserved aspects of haematopoiesis. First, we will use a recently developed genome-wide collection of UAS-dsRNA transgenic flies, from Dr R. Ueda to identify, in a large-scale genetic screen, the genes that control the different steps of Drosophila larval blood cell development. Furthermore, we have shown that the Drosophila orthologues of GATA1 and RUNX1, Serpent and Lozenge, interact physically and functionally to induce the differentiation of a particular blood cell type in vivo. Interestingly, this interaction is conserved in mammals in which it may also regulate blood cell development. To further characterize the molecular mechanisms of action and of cooperation between GATA and RUNX factors, we will perform a genome wide RNAi screen for modulators of Srp/Lz-induced transactivation in Drosophila S2 cells. Translocations or mutations affecting Runx1/AML1 are associated with the development of a large proportion of leukaemia in human. For instance the RUNX1-ETO translocation product is present in more than 10% of acute myeloid leukaemia, and the TEL-RUNX1 translocation is associated to 25% of acute lymphoblastic leukaemia. Another aspect of our project will be to use Drosophila as a model system to analyse the mode of action of these two oncogenic RUNX derivatives. In particular, we intend to perform both a molecular and a genetic screen to identify genes that control the activity of RUNX1-ETO and/or TEL-RUNX1. Similarly to the screen for modulators of Srp/Lz-induced transactivation, we will look for modulators of RUNX1-ETO or TEL-RUNX1-controlled transcription by RNAi on S2 cells. Furthermore, we will perform an in vivo screen using UAS-dsRNA transgenic flies to identify suppressors of TEL-RUNX1 and RUNX1-ETO-induced phenotype in vivo. The different genes identified in these screens will then be studied in depth in vitro and in vivo. In particular, we will characterise the function of these genes in Drosophila haematopoiesis and immune response. We will also elicit the analysis of their human orthologues during normal and leukaemic blood cell development. The completion of our project will allow the identification and the functional characterisation of several new factors that control the activity of GATA and RUNX transcription factors and/or haematopoiesis. Our results should shed new lights on the molecular mode of action of these transcription factors not only during Drosophila haematopoiesis but also in human haematopoietic development as well as in disease such as leukaemia. Hopefully, our results could also help to elaborate new therapeutic strategies.

DNA Double Strand Breaks (DSBs) are highly harmful lesions since they can lead to mutations and chromosome rearrangements, themselves promoting tumour progression. Eucaryotic cells have thus developed efficient DNA repair processes, mainly homologous recombination (HR), and Non Homologous End Joining (NHEJ), to face this danger. However DNA packaging into chromatin creates a significant barrier to DSBs detection and repair, and therefore, major chromatin remodelling events accompany DNA repair. Due to the lack of convenient tools, the chromatin changes induced by and around DSBs, together with their functions, still remains largely unknown, especially in mammalians cells. We have developed a new experimental system, based on the use of a restriction enzyme fused to the ligand binding domain of the oestrogen receptor (AsiSI-ER), which permits a tightly controlled induction of sequence-specific DSBs (at known loci) throughout the genome, in human cells. This new DSB-inducible system, combined with high throughput genome wide technologies (such as ChIP-chip, or GCC), provides now the unique opportunity to study specific chromatin changes during the repair of DSBs, since it enables high resolution profiling of any DSB induced chromatin modification and DNA repair proteins around breaks. Using this system, we have already established the first high resolution map of a DSB-induced chromatin modification (gammaH2AX), and have investigated its spreading properties (manuscript submitted). With such a system in hand, we propose to proceed in an investigation of the relationships between chromatin structure and DSB repair. First of all, we plan to map, by ChIP-chip, various chromatin modifications and DNA repair proteins, around DSBs. Such a thorough description of the chromatin landscape(s) before and after DSBs induction should give insights concerning chromatin modification events that occur during DNA repair. In addition, by monitoring the state of chromatin before DNA damage, we should be able to characterize the influence of chromatin structure, on DNA repair efficiency (specifically concerning the choice between NHEJ and HR). To further gain insights in the influence of chromatin structure on DNA repair, we will also utilize a model already extensively used to study the role of chromatin in transcription: the drosophila 'dosage compensation' process. Indeed, in drosophila, the male X chromosome exihibit very specific chromatin features, to ensure proper equalisation of X genes transcription between male and female, through a process called 'dosage compensation'. We want to take advantage of this different chromatin structure between X chromosome and autosomes, to study the influence of chromatin on DSB repair. We plan to perform these studies in male (SL2) and female (Kc) drosophila cell lines, stably transfected with the AsiSI-ER fusion enzyme. We will also examine the tri-dimensional genome reorganisation upon DSB induction, using GCC (Genome Conformation Capture), a very new technique developed by Dr O'Sullivan (Rodley, C.D.F et al, submitted), in order to better understand the clustering of DSB that occur in the whole nucleus and the function of such clustering events. Finally we will develop whole organism models (transgenic flies and mice) expressing the AsiSI-ER fusion enzyme, in a spatially and temporally controlled manner, in order to dissect, at a molecular level, DSB repair in vivo in various cell types and across differentiation (nervous system, germ cells, stem cells...) The completion of this project will undoubtedly lead to important breakthroughs in our understanding of the role of chromatin in DSB repair. Since the impairment of DNA repair processes is involved in many human diseases such as cancer, neurodegenerative diseases, ageing and sterility, a better understanding of the molecular processes involved in DNA repair will assist in the design of new therapeutic strategies targeting such diseases. An additional benefit of the successful completion of the proposed analyses stems from our desire to deposit all data and computational methodology into the public domain. Both information about chromatin structure and its relationship to DNA damage repair as well as methods used to harness the massive amounts of data, will be invaluable to researchers in many fields including DNA repair, transcription and chromatin structure.

Partial discharges (PD ' a discharge affecting a small distance as compared to the distance between electrodes to which the voltage is applied) in solid insulated electrical systems and the way they affect insulation quality has for decades been a rewarding subject for many researchers. This is because electrical systems reliability is deeply affected by PD occurrence in weakening the insulation ability to sustain functional stresses (electrical, thermal, mechanical, etc.). The slow erosion of the insulating material ' especially in the case of organic polymers that are highly sensitive to exposition of PD, can lead to breakdown of the electrical system with large practical consequences, for example; interruption in the energy supply (which is specifically critical for the electricity network where large electrical generators can be stopped); blackout in part of the European grid (breakdown of solid insulated power cable has been at the origin of major blackout within the European power grid); lost of part of power generation in systems where the failed components are embedded. Two important areas of research in LAPLACE are modeling gas discharge phenomena and studies related to aging in dielectrics. Up to now, these have been very separate activities, but it is clear that there is considerable interest in developing an activity which combines both. The reason is that the standard scenario describing aging in dielectrics is gas breakdown in voids ' small, gas filled voids inevitably present in dielectrics which, over the course of some years, and after repeated breakdown events in the voids, lead to damage and eventual failure of the dielectric. State of the art gas discharge modeling has not been applied to this situation. In the dielectric community, the modeling of gas breakdown in the voids has been limited to very approximate treatments. The gas discharge physics community is very actively looking microdischarges ' discharges generated in very small geometries. A few of the outstanding issues which could be addressed by gas discharge modeling are (a) quantification of electron multiplication in voids for different conditions typical of those encountered in dielectric operation for different missions (b) evaluation of radical and photon generation in voids (c) energy distributions of electrons and ions on the walls. Closely related is the issue of charge transport in dielectrics, where our current understanding is largely empirical. Models of charge transport in dielectrics have been developed, but these are still very macroscopic. They could be developed, but before that some systematic clean experiments are needed. Experience at LAPLACE in the gas discharge groups in microdischarges, plasma display panels will be a part of this project. Also participating in the project will be the local experts in dielectric aging, who will help define reference conditions and evaluate the range of parameters of interest. A young CR CNRS who has considerable experience already in the modeling of radical and UV generation in gas discharges in dielectric barrier discharges and in streamer breakdown will pilot the project. Other young researchers participating in this project are experts in charge transport in dielectrics and in measuring conductivity in dielectrics. This project will borrow extensively from existing activities being carried in the framework of other projects, and a slight reorientation and reorganization will provide valuable information. Although the proposed project is 3 years, longer-term objectives can be identified, the details of which will depend on the results obtained over the next few years. Among the potential follow-ups that would deserve to be built on the grounds of the current project one can name: extension and adaptation of the modeling to the case of cavity size with a very high aspect ratio, i.e. treeing filaments with lengths of some 100 'm, and diameter of a few 'm; development of a model of tree propagation involving the physics of the driving process, i.e. discharges along the dielectric walls; development of a degradation model involving the chemistry induced by the discharge; investigation of charging mechanisms in cellular polymer electrets; development of diagnosis tools dedicated to low activity partial discharges.

Magnetized plasma transport plays a key role not only in hot fusion plasmas but also in low-temperature plasma sources operating at low pressure, in which magnetic fields are used to limit charged particle losses to the walls and/or obtain special kinds of energy coupling. Such plasma sources are widely used in applications like materials processing, space propulsion, and neutral beam injection, and are generally developed by a combination of experimental research and numerical modeling. However, in contrast to the progress made in fusion plasma research, the knowledge of magnetized transport in low-temperature plasma sources has hardly evolved since the 1950s-1960s and is insufficient to meet the modern needs for modeling. Due to the presence of chamber walls, magnetized low-temperature plasmas can show ill-understood complex behavior even in a non-turbulent regime, while most available experimental data is too application-oriented and not detailed enough for model validation. This problem has become particularly urgent in recent efforts to model the magnetic filter stage of the negative ion source of the neutral beam heating system for ITER. In this context, the objective of the METRIS project is to improve the general understanding and modeling of transport in magnetized low-temperature plasma sources, in particular negative ion sources for fusion, but with strong benefits for low-temperature plasma modeling in general. In order to achieve this, we will study the magnetized transport as a problem of its own, as much as possible isolated from other aspects of the discharge operation, for different basic magnetic field configurations. We will combine numerical modeling, theoretical analysis, and basic experiments. These different approaches will be developed by a closely-interacting team, working at the same site of LAPLACE, which will naturally lead to efficient synergy between the approaches, while at the same time establishing an intensive collaboration between several young researchers with complementary backgrounds. The modeling part of the METRIS program concerns the development of robust numerical model schemes for plasma transport in arbitrary magnetic field configuration, as well as simplified analytical models that show the interdependence of the relevant parameters. Our main focus is fluid modeling, which we will back up with more detailed particle-in-cell simulations. As various aspects of this model development are known also in fusion plasma research, we will explore the methods used in that field and profit as much as possible from existing knowledge. Since it is impossible to clarify the magnetized transport problem by modeling alone, an essential part of the METRIS program consists of basic experiments. In order to avoid the practical constraints of application-oriented plasma source designs, we will build a new dedicated laboratory set-up with flexible magnetic field and simple but detailed diagnostics of the plasma transport. In addition to probe diagnostics and imaging, we will develop and perform innovative space- and time-resolved wall-current measurements, allowing detailed model validation. Our aim is to characterize different transport regimes governed by classical cross field mobility, magnetic drifts, and instabilities/turbulence, with the accent on non-turbulent transport and the onset of turbulence. The METRIS project can be expected to yield numerous scientific publications in leading international journals and conferences, as well as numerical codes and methods directly exploitable for plasma-based technologies such as negative ion sources for fusion. The project will allow LAPLACE to maintain and reinforce its position among the international leaders in the field of low-temperature plasma simulation.