Recently, the area of model predictive modification of biochemical pathways has received attention with the aim to increase the productivity of microbial systems. In this study, we present a generalization of previous work, where, using a sensitivity study over the fermentation as a dynamic system, the optimal selection of reaction steps for modification (amplification or attenuation) is determined. The influence of metabolites in the activity of enzymes has also been considered (through activation or inhibition). We further introduce a new concept in the dynamic modeling of biochemical reaction systems including a generalized continuous superstructure in which two artificial multiplicative terms are included to account for: (a) enzyme overexpression or underexpression (attenuation or amplification) for the whole enzyme pool; and (b) modification of the apparent order of a kinetic expression with respect to the concentration of a metabolite or any subset of metabolites participating in the pathway. This new formulation allows the prediction of the sensitivity of the pathway performance index (objective function) with respect to the concentration of the enzyme, as well as the interaction of the enzyme with other metabolites. Using this framework, a case study for the production of penicillin V is analyzed, obtaining the most sensitive reaction steps (or bottlenecks) and the most significant regulations of the system, due to the effect of concentration of intracellular metabolites on the activity of each enzyme.