ABSTRACTS New Biotechnology · Volume 25S · September 2009 taine to L-(-)-carnitine biotransformation must be performed on non phosphorylated sugars such as glycerol since glucose inhibits the cAMP synthesis and the carnitine metabolism at the expression level, forcing the use of more expensive sources and increasing the final product price. Until now the attention in this as in most bioprocess optimiza- tion as referred above have been mainly focused at the kinetic structure of the metabolic pathways involved or at the bioreac- tor performance, but little attention have been addressed to the signalling structure regulating the enzymatic machinery. We present a model development using power-law formalism (GMA version) to describe the metabolic processes involved in the biotransformation of crotobetaine into L-carnitine by a diauxic E. coli continuous cell culture with glucose and glycerol as alternative carbon source and cAMP as a external regulatory metabolite. The system is running in a continuous bioreactor which is feed with crotonobetaine to produce L-carnitine. The model includes glu- cose and glycerol and the L-carnitine metabolisms as well as the involved signalling architecture which are transduced from the gene expression to the metabolic level. The proposed regulatory network has three levels a hierarchical structure: the metabolic, the gene expression and the regulatory one, which includes the modulon and the acting operons. The model’s purpose is, beyond the description and analysis of the glucose effect on the carnitine metabolism, to design carni- tine biosynthesis optimization strategies where glucose is the main energy and carbon source. doi:10.1016/j.nbt.2009.06.859 4.5.06 Captivity paralarves Octopus vulgaris growth: modelling and optimization J.A. Hormiga Cerde˜ na ∗ , I. Frías, N.V. Torres Darias, E. Almansa Universidad de La Laguna, San Cristobal de La Laguna, Spain In many countries, aquaculture production is restricted to a few crop species (sea bream, sea bass and turbot, etc.), which threatens to cause a rapid saturation of the market in few years. A solution is to diversify into other species. One of the main candidates is the common octopus (Octopus vulgaris) that has a strong market demand and a high growth rate. O. vulgaris is a Cephalopoda, order Octopoda, mollusk; their young tended by the female until complete embrionary development and hatching occurs. Cephalopods have a direct development, with no strict sense larval stages. Thus, the term “paralarve” has been established to replace the term larva to refer to the just hatched cephalopods. These paralarve, provided with eyes and a complex nervous system, become part of the plankton and behave as active predators. After a short development period they move to the sea bottom and became part of the bentonic ecosystem. Given its interest as a crop species, many studies have been conducted with the aim of closing the life cycle in captivity, but with scarce results mainly due to the high paralarve mortal- ity that makes the process not commercially profitable. Recently, nutritional problems have been identified as the main cause of this high mortality. In this work we propose a system biology approach, power-law formalism model, aimed to predict the diet dependent, temporal evolution of the paralarvas composition. For this purpose we have used paralarve and adult octopus composition data from other closely related species showing a less complex development patter (squid) to determine the diet that gives a best fit with the experi- mental data. Moreover we present the composition of the diet that ensures an optimal survival of as many individuals thus being this farming a profitable one. doi:10.1016/j.nbt.2009.06.860 4.5.07 In silico rotavirus-like particle production in the recom- binant baculovirus/insect cell system A. Roldão 1,∗ , M.J.T. Carrondo 1 , P.M. Alves 1 , R. Oliveira 2 1 ITQB-UNL/IBET, Oeiras, Portugal 2 REQUIMTE, FCT/UNL, Portugal In heterologous protein expression, the introduction of a recombi- nant gene into a host cellular system and its subsequent expression alters dramatically cell behaviors. The complexity of such cellular networks increases when the expression of more than one protein is intended, hampering the understanding of how functional regu- lation is achieved in a cell. This is the case of rotavirus-like particles (RLP). RLPs, a vaccine candidate against rotavirus disease responsible for 440,000 deaths per year in children <5 years, is produced in Sf-9 cells infected with recombinant baculovirus coding for three structural proteins of rotavirus, vp 2 , vp 6 and vp 7 . These intracellu- lar viral proteins assemble into triple layered structures, a process determined by thermodynamics with a considerable degree of ran- domness. At the end, many contaminants are formed (monomers and trimers of vp 6 and vp 7 , double layered particles of vp 2 and vp 6 ), with the RLP correctly formed representing less than 12% (w/w) of the total mass of proteins expressed. The aim of this work is to understand the interplays between the genetics of the viral vector, the infection strategy (cell den- sity, multiplicity of infection (MOI) and time of infection (TOI)) and the titer and structure of final particles. A bottom-up systems biology methodology was applied leading to an in silico stochastic, segregated, structured model combined with a thermodynamical- statistical description of the molecular assembly of the particles. Putting all pieces together, the model is able to simulate the effect of the infection strategy (a process degree of freedom) on the distribution of stable particle subunits with different molecular composition. Several infection strategies and methodologies (real-time quan- titative PCR, Western blot, ELISA, electronic microscopy, BCA protein quantification assay) were used to assess intra and extracel- lular data. Simulated dynamics of intracellular viral DNA, mRNA and viral proteins show: (i) maximum vDNA content is a trade- off between MOI and virus budding effect; (ii) vDNA and mRNA S356 www.elsevier.com/locate/nbt