Detalhes bibliográficos
| Ano de defesa: |
2010 |
| Autor(a) principal: |
Montaño, Inti Doraci Cavalcanti |
| Orientador(a): |
Giordano, Roberto de Campos
 |
| Banca de defesa: |
Não Informado pela instituição |
| Tipo de documento: |
Dissertação
|
| Tipo de acesso: |
Acesso aberto |
| Idioma: |
por |
| Instituição de defesa: |
Universidade Federal de São Carlos
|
| Programa de Pós-Graduação: |
Programa de Pós-Graduação em Engenharia Química - PPGEQ
|
| Departamento: |
Não Informado pela instituição
|
| País: |
BR
|
| Palavras-chave em Português: |
|
| Palavras-chave em Inglês: |
|
| Área do conhecimento CNPq: |
|
| Link de acesso: |
https://repositorio.ufscar.br/handle/ufscar/4051
|
Resumo: |
This master's thesis project aims at studying the dynamic optimization of the operation of a bench scale (up to 5L) automated, agitated and aerated bioreactor, where the semi-continuous cultivation of recombinant Pichia pastoris is run. This yeast was cloned using the PGK1 promoter, which precludes the use of methanol as inducer, expressing constitutively the enzyme penicillin G Acylase (PGA) from Bacillus megaterium. While the group of molecular biology of DEQUFSCar is working on cloning the PGA, d P. pastoris expressing the enzyme - amylase from Bacillus subtilis was cultivated. This clone, provided by prof. Fernando Torres, UnB, uses the same construction and, therefore, its kinetics of growth and production should be very similar to the PGA s. Cultivation of recombinant Pichia pastoris was performed in flasks (skaker) using standard culture medium, aiming at obtaining kinetic data, which are the starting point for the escalation to a benchtop bioreactor. Following that, tests were performed in a 5L bioreactor in batch and fed batch operation modes. With the bioreactor data , kinetic parameters of growth, to be further used in the simulations, were estimated, using a hybrid algorithm (which combines the global method Simulated Annealing, with the local one Levenberg- Marquardt). This algorithm, is implemented in Matlab and available in the software library of Ladabio (Laboratory of Development and Automation of Bioprocesses ). From these data, models of microbial growth and of production were developed, following a classic approach (unstructured, non-segregated). Computer simulations using different feeding strategies and employing these models allowed mapping the dynamics of the system. From this information, optimal control strategies were proposed to define optimal feeding profiles. Cellular concentrations of 5.4 g/L (dry weight) were reached in shaker (20h of cultivation, when glucose is exhausted), expressing 218 U/mL of -amylase, compared to 11.4 g/L (dry weight) that were achieved in cultures in a bioreactor in batch simple (10h of cultivation, when glucose is exhausted), expressing 156 U/mL of -amylase In fed-batch cultures, cell concentrations of up to 45 g/L were achieved, expressing up to 260 U/mL of - amylase, with a productivity of 5.2 U/mL/ h. In fed-batch cultures of P. pastoris expressing PGA, cell concentrations of up to 35 g/L were achieved. Enzyme activity was not detected in the culture broth due to the effect of glycosylation. Immunodetection reaction confirmed the expression of the recombinant enzyme. Four specific growth rate equations were adjusted, with different types of inhibition by one product, detected at significant levels by liquid chromatography highperformance, but not yet identified. This metabolite was added as an inhibitor in kinetic models, using the peak areas, normalized as a pseudoconcentration. The best fit to the experimental data were the Monod kinetic model with non-competitive inhibition. Typical values obtained for the maximum specific growth and glucose/ cell conversion factor in bioreactor were max=0,24 h-1 and YX/S = 0,48. Algorithm for optimal control in open loop was developed and successfully implemented, providing a robust profiles of great power, whose validation is proposed as a continuation of this work. |
