Hypoxie-ischémie cérébrale néonatale : rôle de la voie de JNK et implication de l'autophagie dans la mort neuronale

Résumé : Malgré les immenses progrès réalisés depuis plusieurs années en médecine obstétricale ainsi qu'en réanimation néonatale et en recherche expérimentale, l'asphyxie périnatale, une situation de manque d'oxygène autour du moment de la naissance, reste une cause majeure de mortalité et de morbidité neurologique à long terme chez l'enfant (retard mental, paralysie cérébrale, épilepsie, problèmes d'apprentissages) sans toutefois de traitement pharmacologique réel. La nécessité de développer de nouvelles stratégies thérapeutiques pour les complications de l'asphyxie périnatale est donc aujourd'hui encore essentielle. Le but général de ce travail est l'identification de nouvelles cibles thérapeutiques impliquées dans des mécanismes moléculaires pathologiques induits par l'hypoxie-ischémie (HI) dans le cerveau immature. Pour cela, le modèle d'asphyxie périnatale (proche du terme) le plus reconnu chez le rongeur a été développé (modèle de Rice et Vannucci). Il consiste en la ligature permanente d'une artère carotide commune (ischémie) chez le raton de 7 jours combinée à une période d'hypoxie à 8% d'oxygène. Il permet ainsi d'étudier les lésions de type hypoxique-ischémique dans différentes régions cérébrales dont le cortex, l'hippocampe, le striatum et le thalamus. La première partie de ce travail a abordé le rôle de deux voies de MAPK, JNK et p38, après HI néonatale chez le raton à l'aide de peptides inhibiteurs. Tout d'abord, nous avons démontré que D-JNKI1, un peptide inhibiteur de la voie de JNK présentant de fortes propriétés neuroprotectrices dans des modèles d'ischémie cérébrale adulte ainsi que chez le jeune raton, peut intervenir sur différentes voies de mort dont l'activation des calpaïnes (marqueur de la nécrose précoce), l'activation de la caspase-3 (marqueur de l'apoptose) et l'expression de LC3-II (marqueur de macroautophagie). Malgré ces effets positifs le traitement au D-JNKI1 ne modifie pas l'étendue de la lésion cérébrale. L'action limitée de D-JNKI1 peut s'expliquer par une implication modérée des JNKs (faiblement activées et principalement l'isotype JNK3) après HI néonatale sévère. Au contraire, l'inhibition de la voie de nNOS/p38 par le peptide DTAT-GESV permet une augmentation de 20% du volume du tissu sain à court et long terme. Le second projet a étudié les effets de l'HI néonatale sur l'autophagie neuronale. En effet, l'autophagie est un processus catabolique essentiel au bien-être de la cellule. Le type principal d'autophagie (« macroautophagie » , que nous appellerons par la suite « autophagie ») consiste en la séquestration d'éléments à dégrader (protéines ou organelles déficients) dans un compartiment spécialisé, l'autophagosome, qui fusionne avec un lysosome pour former un autolysosome où le contenu est dégradé par les hydrolases lysosomales. Depuis peu, l'excès ou la dérégulation de l'autoptiagie a pu être impliqué dans la mort cellulaire en certaines conditions de stress. Ce travail démontre que l'HI néonatale chez le raton active fortement le flux autophagique, c'est-à-dire augmente la formation des autophagosomes et des autolysosomes, dans les neurones en souffrance. De plus, la relation entre l'autophagie et l'apoptose varie selon la région cérébrale. En effet, alors que dans le cortex les neurones en voie de mort présentent des caractéristiques mixtes apoptotiques et autophagiques, ceux du CA3 sont essentiellement autophagiques et ceux du CA1 sont principalement apoptotiques. L'induction de l'autophagie après HI néonatale semble donc participer à la mort neuronale soit par l'enclenchement de l'apoptose soit comme mécanisme de mort en soi. Afin de comprendre la relation pouvant exister entre autophagie et apoptase un troisième projet a été réalisé sur des cultures primaires de neurones corticaux exposés à un stimulus apoptotique classique, la staurosporine (STS). Nous avons démontré que l'apoptose induite par la STS était précédée et accompagnée par une forte activation du flux autophagique neuronal. L'inhibition de l'autophagie de manière pharmacologique (3-MA) ou plus spécifiquement par ARNs d'interférence dirigés contre deux protéines autophagiques importantes (Atg7 et Atg5) a permis de mettre en évidence des rôles multiples de l'autophagie dans la mort neuronale. En effet, l'autophagie prend non seulement part à une voie de mort parallèle à l'apoptose pouvant être impliquée dans l'activation des calpaïnes, mais est également partiellement responsable de l'induction des voies apoptotiques (activation de la caspase-3 et translocation nucléaire d'AIF). En conclusion, ce travail a montré que l'inhibition de JNK par D-JNKI1 n'est pas un outil neuroprotecteur efficace pour diminuer la mort neuronale provoquée par l'asphyxie périnatalé sévère, et met en lumière deux autres voies thérapeutiques beaucoup plus prometteuses, l'inhibition de nNOS/p38 ou de l'autophagie. ABSTRACT : Despite enormous progress over the last«decades in obstetrical and neonatal medicine and experimental research, perinatal asphyxia, a situation of lack of oxygen around the time of the birth, remains a major cause of mortality and long term neurological morbidity in children (mental retardation, cerebral palsy, epilepsy, learning difficulties) without any effective treatment. It is therefore essential to develop new therapeutic strategies for the complications of perinatal asphyxia. The overall aim of this work was to identify new therapeutic targets involved in pathological molecular mechanisms induced by hypoxia-ischemia (HI) in the immature brain. For this purpose, the most relevant model of perinatal asphyxia (near term) in rodents has been developed (model of Rice and Vannucci). It consists in the permanent ligation of one common carotid artery (ischemia) in the 7-day-old rat combined with a period of hypoxia at 8% oxygen. This model allows the study of the hypoxic-ischemic lesion in different brain regions including the cortex, hippocampus, striatum and thalamus. The first part of this work addressed the role of two MAPK pathways (JNK and p38) after rat neonatal HI using inhibitory peptides. First, we demonstrated that D-JNKI1, a JNK peptide inhibitor presenting strong neuroprotective properties in models of cerebral ischemia in adult and young rats, could affect different cell death mechanisms including the activation of calpain (a marker of necrosis) and caspase-3 (a marker of apoptosis), and the expression of LC3-II (a marker of macroautophagy). Despite these positive effects, D-JNKI1 did not modify the extent of brain damage. The limited action of D-JNKI1 can be explained by the fact that JNKs were only moderately involved (weakly activated and principally the JNK3 isotype) after severe neonatal HI. In contrast, inhibition of nNOS/p38 by the peptide D-TAT-GESV increased the surviving tissue volume by around 20% at short and long term. The second project investigated the effects of neonatal HI on neuronal autophagy. Indeed, autophagy is a catabolic process essential to the well-being of the cell. The principal type of autophagy ("macroautophagy", that we shall henceforth call "autophagy") involves the sequestration of elements to be degraded (deficient proteins or organelles) in a specialized compartment, the autophagosome, which fuses with a lysosome to form an autolysosome where the content is degraded by lysosomal hydrolases. Recently, an excess or deregulation of autophagy has been implicated in cell death in some stress conditions. The present study demonstrated that rat neonatal HI highly enhanced autophagic flux, i.e. increased autophagosome and autolysosome formation, in stressed neurons. Moreover, the relationship between autophagy and apoptosis varies according to the brain region. Indeed, whereas dying neurons in the cortex exhibited mixed features of apoptosis and autophagy, those in CA3 were primarily autophagíc and those in CA1 were mainly apoptotic. The induction of autophagy after neonatal HI seems to participate in neuronal death either by triggering apoptosis or as a death mechanism per se. To understand the relationships that may exist between autophagy and apoptosis, a third project has been conducted using primary cortical neuronal cultures exposed to a classical apoptotic stimulus, staurosporine (STS). We demonstrated that STS-induced apoptosis was preceded and accompanied by a strong activation of neuronal autophagic flux. Inhibition of autophagy pharmacologically (3-MA) or more specifically by RNA interference directed against two important autophagic proteins (Atg7 and AtgS) showed multiple roles of autophagy in neuronal death. Indeed, autophagy was not only involved in a death pathway parallel to apoptosis possibly involved in the activation of calpains, but was also partially responsible for the induction of apoptotic pathways (caspase-3 activation and AIF nuclear translocation). In conclusion, this study showed that JNK inhibition by D-JNKI1 is not an effective neuroprotective tool for decreasing neuronal death following severe perinatal asphyxia, but highlighted two more promising therapeutic approaches, inhibition of the nNOSlp38 pathway or of autophagy.

Diabetes mellitus and cancer risk in a network of case-control studies.

Diabetes has been associated to the risk of a few cancer sites, though quantification of this association in various populations remains open to discussion. We analyzed the relation between diabetes and the risk of various cancers in an integrated series of case-control studies conducted in Italy and Switzerland between 1991 and 2009. The studies included 1,468 oral and pharyngeal, 505 esophageal, 230 gastric, 2,390 colorectal, 185 liver, 326 pancreatic, 852 laryngeal, 3,034 breast, 607 endometrial, 1,031 ovarian, 1,294 prostate, and 767 renal cell cancer cases and 12,060 hospital controls. The multivariate odds ratios (OR) for subjects with diabetes as compared to those without-adjusted for major identified confounding factors for the cancers considered through logistic regression models-were significantly elevated for cancers of the oral cavity/pharynx (OR = 1.58), esophagus (OR = 2.52), colorectum (OR = 1.23), liver (OR = 3.52), pancreas (OR = 3.32), postmenopausal breast (OR = 1.76), and endometrium (OR = 1.70). For cancers of the oral cavity, esophagus, colorectum, liver, and postmenopausal breast, the excess risk persisted over 10 yr since diagnosis of diabetes. Our data confirm and further quantify the association of diabetes with colorectal, liver, pancreatic, postmenopausal breast, and endometrial cancer and suggest forthe first time that diabetes may also increase the risk of oral/pharyngeal and esophageal cancer. [Table: see text] [Table: see text].

Two-sided learning in New Keynesian models: Dynamics, (lack of) convergence and the value of information

This paper investigates the role of learning by private agents and the central bank(two-sided learning) in a New Keynesian framework in which both sides of the economyhave asymmetric and imperfect knowledge about the true data generating process. Weassume that all agents employ the data that they observe (which may be distinct fordifferent sets of agents) to form beliefs about unknown aspects of the true model ofthe economy, use their beliefs to decide on actions, and revise these beliefs througha statistical learning algorithm as new information becomes available. We study theshort-run dynamics of our model and derive its policy recommendations, particularlywith respect to central bank communications. We demonstrate that two-sided learningcan generate substantial increases in volatility and persistence, and alter the behaviorof the variables in the model in a significant way. Our simulations do not convergeto a symmetric rational expectations equilibrium and we highlight one source thatinvalidates the convergence results of Marcet and Sargent (1989). Finally, we identifya novel aspect of central bank communication in models of learning: communicationcan be harmful if the central bank's model is substantially mis-specified.

Selective hiring and welfare analysis in labor market models

Firms select not only how many, but also which workers to hire. Yet, in standardsearch models of the labor market, all workers have the same probability of being hired.We argue that selective hiring crucially affects welfare analysis. Our model is isomorphicto a search model under random hiring but allows for selective hiring. With selectivehiring, the positive predictions of the model change very little, but the welfare costsof unemployment are much larger because unemployment risk is distributed unequallyacross workers. As a result, optimal unemployment insurance may be higher and welfareis lower if hiring is selective.

Bridging DSGE models and the raw data

A method to estimate DSGE models using the raw data is proposed. The approachlinks the observables to the model counterparts via a flexible specification which doesnot require the model-based component to be solely located at business cycle frequencies,allows the non model-based component to take various time series patterns, andpermits model misspecification. Applying standard data transformations induce biasesin structural estimates and distortions in the policy conclusions. The proposed approachrecovers important model-based features in selected experimental designs. Twowidely discussed issues are used to illustrate its practical use.

Calibration of stochastic volatility models via second order approximation: the Heston model case

Using a suitable Hull and White type formula we develop a methodology to obtain asecond order approximation to the implied volatility for very short maturities. Using thisapproximation we accurately calibrate the full set of parameters of the Heston model. Oneof the reasons that makes our calibration for short maturities so accurate is that we alsotake into account the term-structure for large maturities. We may say that calibration isnot "memoryless", in the sense that the option's behavior far away from maturity doesinfluence calibration when the option gets close to expiration. Our results provide a wayto perform a quick calibration of a closed-form approximation to vanilla options that canthen be used to price exotic derivatives. The methodology is simple, accurate, fast, andit requires a minimal computational cost.

Solving non-linear stochastic models by parameterizing expectations: An application to asset pricing with production

A new algorithm called the parameterized expectations approach(PEA) for solving dynamic stochastic models under rational expectationsis developed and its advantages and disadvantages are discussed. Thisalgorithm can, in principle, approximate the true equilibrium arbitrarilywell. Also, this algorithm works from the Euler equations, so that theequilibrium does not have to be cast in the form of a planner's problem.Monte--Carlo integration and the absence of grids on the state variables,cause the computation costs not to go up exponentially when the numberof state variables or the exogenous shocks in the economy increase. \\As an application we analyze an asset pricing model with endogenousproduction. We analyze its implications for time dependence of volatilityof stock returns and the term structure of interest rates. We argue thatthis model can generate hump--shaped term structures.