How are memories encoded in the brain? How are memory centers organized? Recent models propose that anatomically distinct systems coexist, each storing specific aspects of memories. These systems are not independent, but rather interact with each other dynamically. The major challenge faced by neuroscientists studying memory is to define the links between the various levels of nervous system organization, starting with molecules and cells, and then neuronal circuits, and finally the global cognitive functions of the brain. Drosophila has many advantages for the study of brain function: its brain is small (100 000 cells) yet highly structured, and genetic tools allow us to manipulate gene expression in discrete neuronal assemblies while recording behavior. We use this power to perform an integrated analysis of the processes underlying associative learning and memory, studying not only the activity of individual genes and proteins, but also the behavior of cells, neural networks, and of the fly itself. In the past few years our team has made important discoveries as we have identified a long-term memory center (Pascual and Preat, 2001 Science), and shown that the two forms of Drosophila consolidated memory, anesthesia-resistant memory (ARM) and long-term memory (LTM) are exclusive (Isabel et al., 2004 Science). We have also characterized several genes involved in LTM formation and whose expression is regulated after conditioning (Comas et al., 2004 Nature; Didelot et al., 2006 Science). At last, we have shown that the Drosophila brain is asymmetric and that this asymmetry is necessary for LTM (Pascual et al., 2004 Nature). The present project aims at further integrating the various levels involved in memory formation, consolidation and retrieval, using in particular in vivo brain imaging. We have five major goals: 1) We have shown recently that dopamine plays a key role in the transformation of short-term memory into long-term memory (Isabel et al., unpublished). We will further analyze how dopamine regulates the activity of the mushroom bodies, the olfactory memory center. In particular we will analyze how dopamine and the DAMB receptor control the transition between short-term and LTM. A proteic probe has been introduced in Drosophila that will be used to record in vivo the activity of the protein kinase A, a downstream effector of the DAMB receptor. 2) The identification of new ARM mutants should help understanding how the two consolidated memories interact. A recent behavioral screen failed to produce bona fide ARM mutants that are affected only for this memory phase. We will perform a new and original screening to identify new genes involved in ARM formation. The success of this approach would contribute importantly to our understanding of memory dynamics, showing that indeed ARM acts as a gating mechanism that controls the transition between short-term and LTM. 3) We have accomplished a microarray analysis of the transcriptome after LTM conditioning. A short list of 112 genes has been generated. We will pursue the characterization of several pathways involved in LTM. In particular we will quantify gene expression in the brain of naïve and trained flies, and analyze the LTM of flies after adult inhibition of gene expression by RNA interference. 4) We plan to use fluorescent calcium probes to image the changes of brain activity linked to memory formation. Our team has acquired a two-photon microscope that allows precise imaging of brain circuits using genetically-encoded probes. We plan to further study how the different memory networks interact by imaging the activity of one neuronal ensemble while blocking the activity of another neuronal circuit with a thermosensitive inhibitory protein. We will investigate in particular the function of three newly identified neuronal ensembles that connect to the mushroom bodies and that are involved in LTM retrieval. 5) Formation of aversive LTM requires multiple presentations of the odorant paired with electric shocks. On the contrary we have shown recently that appetitive LTM is generated after a single training session. We will further compare the dynamics of aversive and appetitive memory phases, using memory mutants and calcium imaging. All together this program should have a major impact on our understanding of the dynamics of memory phase in Drosophila, and provide a link between the molecular and behavioral levels.

In the field of biomedical ultrasound, conventional ultrasound imaging or ultrasound therapy techniques assume the investigated tissue to be homogeneous in terms of speed of sound in order to focus ultrasound. Such an assumption may be inadequate for some locations in the body. For instance, one of the most promising application in the fast emerging field of ultrasonic therapy is the possibility to non invasively treat brain tumours with high intensity focused ultrasound (HIFU). But focusing ultrasound into the brain requires focusing through the skull bone, which is a strongly aberrating medium with acoustic properties very different from those of soft tissue, and which exact shape and thickness are a priori unknown. Time-reversal of acoustic waves provides a way to automatically focus ultrasound in highly complex media, based on the recording and re-emission of the ultrasound waves generated by a localized ultrasound source. The different types of time-reversal experiments that have been performed so far, in several fields such as ocean acoustics, biomedical acoustics, seismology, have relied on the existence of some localized mechanical contrast or some exogenous active ultrasound source (explosive load in ocean acoustics, earthquakes in seismology, cavitating bubbles in tissue). In this project, we propose to use a new kind of source to perform time-reversal experiments: as opposed to all previous types of sources that have been used, we propose to use a photoacoustic source, which ultrasonic emission is related to the optical properties (absorption) of the source. The photoacoustic effect consists in the emission of sound by an illuminated optically absorbing region. Physically, photoacoustic waves are pressure waves generated by the thermoelastic expansion following the absorption of a light pulse. For optical wavelengths included in the so-called therapeutic window (700 nm to 900 nm), the absorption of light in biological tissue is relatively low (absorption length of the order of several centimetres), and multiply scattered photons can travel as deep as a few centimetres in tissue. Photoacoustic waves can therefore be generated deep in tissue, and have in particular been used so far as a way to perform passive acoustic imaging. In this proposal, we want to develop a prototype designed to perform photoacoustic-guided high-intensity focused ultrasound (PAG-HIFU) towards localized optical absorbers in biological tissue, based on the time-reversal of the photoacoustic waves emitted by the absorbers. Our final objective is to create PAG-HIFU lesions in vitro in biological tissue and in vivo in small animals (rats or mice). Towards this goal, we will : 1) develop a method to selectively detect the photoacoustic waves generated by targeted optical absorbers in biological tissue, in order to time-reverse only the desired photoacoustic waves. 2) design optical contrast agents that can be excited selectively and act as photoacoustic sources, based on absorbing quantum dots 3) design a 2D ultrasound array able both to detect photoacoustic signals and to emit high-intensity ultrasound. 4) generate high-intensity focused ultrasound based on time-reversed photoacoustic waves. 5) study and quantify the parameters which determine the feasibility of the technique in tissue. 6) test and validate the whole system in vitro on biological tissue and in vivo on small animal models (rats). The potential of time-reversal of photoacoustic waves to perform HIFU is twofold, as can be illustrated for instance in the case of tumor therapy: first, assuming the absorption of a tumor to be different from that of healthy tissue (either from natural blood contrast due to hyper-vascularization, or caused by exogenous optical contrast agent), the technique makes it possible to focus selectively on the tumour, while automatically compensating for possible aberration. This would not only benefit to brain tumours by correcting the aberrations induced by the skull bone but also to moving organs such as the liver as time-reversal performed in real time would always focus on the photoacoustic source. Second, the technique would more generally benefit to all HIFU treatments by focusing selectively on tumours that are known to have different optical properties from healthy tissues even though they have similar acoustic properties. The photoacoustic effect can therefore be used not only as a way to create a source for time-reversal, but to reveal the presence of an object (the tumor) which may have been invisible to ultrasound probing by purely ultrasonic methods based on mechanical contrast (conventional ultrasound imaging or elastography).

The « small GTPase structure/function relationship » project emerged one year ago in the Structural Motility group at the Curie Institute. Our goal is to gain 3D structural information on several small GTPases of the H-Ras superfamily that are involved in cancer. We have chosen to study these proteins in complex with their cellular partners in order to visualize the protein-protein interactions that are at the origin of their functions. Our structures will provide critical information leading to better understanding of their mode of action, their functions and their roles in cancer. The cellular biology environment of the Curie Institute is ideal for the development of this project. We are not only informed on unpublished results that involve small GTPases in the cancer field but we can develop very fruitful collaborations to perform a full structural/functional study of these proteins. In this context, we are interested in the study of three families of small GTPases : The RGK proteins, as atypical small GTPases. RGK proteins are critical regulators of cytoskeleton remodelling and voltage-dependent calcium channels (VDCCs). These small GTPases exhibit several atypical features in their primary sequence and their mode of regulation. To elucidate how this atypical family of small GTPases function, we propose to determine the structures of an RGK protein in complex with calmodulin which plays a role in its regulation and with one of its effectors that is involved in the VDCCs signalling pathway. The Arf proteins, structure/function studies. The Arf small GTPases regulate vesicular traffic and organelle structure. A new role for ARF6 in cytokinesis has been reported involving the JIP proteins (scaffold proteins regulating the MAP Kinase pathways), and two reverse microtubule motors, Kinesin1 and the Dynein-Dynactin complex. One intriguing observation is that ARF6 and the two reverse motors interact with the same region of the JIP proteins ; some interactions are exclusive, while others are cooperative. To better understand the role of each of these proteins in cytokinesis, we plan to determine the structures of the JIP proteins in complex with ARF6 and with the kinesin-1 light chain, respectively. The Rho proteins, as a target for bacterial toxins. Rho proteins are main regulators of the actin cytoskeleton organization and cell division. These critical functions make these small GTPases favorite targets of bacterial toxins. One of these toxins, the C3 exoenzyme, has become powerful as a pharmacological tool to study their functions, since it exhibits high specificity for one subfamily of Rho proteins. We are interested in understanding in detail, the structural basis that allows the C3 exoenzyme to specifically target and modify this Rho subfamily. To do so, we propose to solve the structure of several intermediates of the C3 exoenzyme-RhoA complex along the enzymatic reaction.