2ème réunion de l'ARC MOCA
19 mai 2006
Institut de Biologie de Lille


Partie publique

Partie restreinte aux participants de l'ARC

Résumés et transparents

Regulation of non genomic functions of Mitogen Activated Protein Kinase (MAPK)

Underscoring the importance of MAPK (Mitogen Activated Protein Kinase) /Erk (E xtracellular signal Regulated kinases) pathway in tumorigenesis, activated MAPK/ Erk or elevated MAPK/Erk expression have been detected in a wide variety of huma n tumors and therefore, MAPK/Erk signalling has been identified as a potential t arget for anticancer strategies. The Xenopus oocyte is cellular context which enable to study biochemical, morp hological as well as ionic mechanisms involved in cell cycle progression. Using this meiotic model we analyzed the role of MAPK during cell reorganization at M- Phase. Using different approaches to inhibit the Mos - MEK - MAPK/Erk - Rsk pathway (morpholinos and phosphorothioate anti-Mos antisense, small molecul e inhibitor U0126), we demonstrated that activation of MEK - MAPK/Erk cascade induced by insulin is strictly dependent upon oncoprotein Mos accumulation. Mor eover, Ras dependent mechanisms triggered by insulin failed to fully phosphoryla te and activate Raf in absence of MEK activity. Either in oocyte stimulated by p rogesterone or insulin, we observed that even if chromosomes condense, oocytes f ailed to establish a bipolar spindle at the plasma membrane : microtubule aster- like structures are observed in deep cytoplasm. Such structures do not enable oo cytes to equally segregate their genomic content and suggest that cell cycle catastrophic events like unproper spindle m orphogenesis could be driven by MAPK/Erk inhibition. Spindle restoration experiment showed that (1) Mos - MEK - MAPK/Erk - Rsk pathway is required to establish a bipolar spindle of division and that (2) Mos and Rsk play essential roles in spindle morphogenesis : Spindle morphogenes is is restored if both Mos and Rsk are active and present. Moreover, Mos and Rsk appear to play complementary role in the formation of the spindle and the estab lishment of the bipolar axis. These results engage to consider the members of th e MAPK / Erk pathway not only as modules of a cascade but also as part of a netw ork where each module can play a particular role.

Modelling the dynamics of simple cyclical protein regulatory networks and its application to the budding yeast cell cycle [PPT]

In this talk, the dynamics of the cyclical organization of simple protein networks and its application to the budding yeast cell cycle will be discussed by taking simple nonlinear models. The protein network consists of two small cyclical loops, where each loop in the absence of interaction with the other exhibits dierent dynamical behavior. Bistability is exhibited by one loop in which the proteins are positively regulated by the preceding one and in turn regulates positively the subsequent one in a cyclic clockwise fashion. Limit cycle oscillations are exhibited by the second loop in which the proteins are negatively regulated by preceding one and in turn negatively regulates the subsequent one in a cyclic anticlockwise fashion. Coupling of both the cyclical loops by positive feedback loop displays complex behavior such as multi-stability, coexistence of limit cycle and multiple steady states. Also, the coupling brings in the notion of checkpoint in the model. The model exhibits domino-like behavior, where limit cycle oscillations takes place in a stepwise fashion. As an application, the events that govern the cell cycle of budding yeast is considered. In budding yeast, the feedback interactions among the important transcription factors, cyclins and their inhibitors in G1, S-G2 and M phases are taken for the construction of the biological circuit diagram. Surprisingly, the sequential activation of the transcription factors, cyclins and their inhibitors forms two independent cyclical loops, with transcription factors involved in the cyclic positive regulation in clockwise direction, while the cyclins and its inhibitors involved in the negative regulation in anticlockwise direction. The coupling of the transcription factors and the cyclin and its inhibitors by positive feedback loops generates rich bifurcation diagram that can be related to the dierent events in the G1, S-G2 and M phases of cell division cycle in terms of dynamical system theory. The dierent checkpoints in the cell cycle are accounted for by appropriately silencing the positive feedback loops that couple the transcription factors and the cyclin and their inhibitors.

Reaction kinetics of compositional gene circuits [PDF]

I derive the reaction kinetics underlying a recently proposed compositional approach to the stochastic dynamics of gene networks. The resulting rate equations are shown to be equivalent to commonly used ODE-based approaches. The results are interpreted with respect to compositional modeling decisions in stochastic and deterministic settings.

Logical modelling of the cell cycle in yeast and mammals [PDF]

Eukaryotic cell division is controlled by a complex, intertwinned regulatory network. This network is currently seen as a core oscillator to which are plugged several control modules that ensure that each step has been completed before the next one begins. The structure of the core engine and most of the control modules is well conserved among eukaryotes, though higher eukaryotes possess aditional levels of control and higher levels of redundancy and variability, in relation with the constraints imposed by multicellularity. This complexity is such that a modelling approach is necessary to fully understand the precise role of each component and regulatory loop in the global dynamics of the system. We use a logical formalism to model the mammalian cell cycle, taking inspiration in Novák and Tyson's (2004) ODE model. In parallel, the study of the simpler and better known budding yeast system, still based on Novák and Tyson's work, should help us to understand better the structure of the mammalian network. Using GINsim, the devoted software developped in our team, we should be able to rapidly extend our model, especially in terms of additional control modules.

Black magic constant [PPT]

Suppose we are given a piecewise linear model C, where one does not know the parameters, and a compatible (meaning sharing the same variables and influences) logical discrete model D. We're giving constraints on C's parameters for inheriting D's (qualitative) properties (meaning D simulates C).

This could be helpful in manipulating, and reusing continuous models (C) by making explicit some qualitative discrete infra-structure (D).

We work out an example where D, modeling p53's response to DNA damage, is due to Kaufman et al (ACI vicanne, Paris).

Hierarchies and modules in biological systems

I shall present several mathematical methods allowing to identify modules in systems biology models. The methods are based either on uniqueness of the input-output characteristic or on monotonicity. Then, I shall tackle the question of how the wiring of the modules influences the static response of the global system. I shall also discuss hierarchical models which define several levels of complexity. Passing from one level of complexity to another one is a many to one application of the parameters similar to renormalization from statistical mechanics. Renormalization allows to compare models with different levels of complexity and can be also useful for merging small modules into larger ones. I shall discuss the application of these theoretical ideas to various examples: signalling pathways (NFkB, TGFb) and cell cycle.

Coupling two numerical modules of the cell-cycle -- work in progress [PPT]

A partir de deux modèles publiés des transitions G1-S et G2-M du cycle cellulaire chez les mammifères, nous nous sommes intéressés aux problèmes émergeant du couplage de ces deux "modules". Nous avons pu ainsi établir une liste de difficultés auxquelles le modélisateur peut être confronté, pour ce cas précis ou pour tout type de couplage.
Il convient, en particulier, de bien définir ce qui est attendu du modèle couplé, de comprendre les questions auxquelles chaque module répond individuellement, d'adapter ces modules les uns aux autres, et de chercher les différents liens possibles entre eux.
Pour le moment, le couplage se fait "à la main" afin de déterminer des méthodes en vue d'un couplage automatique.

Last modified : Friday April 21, 2006