[Selection criteria in metabolic systems].

In a given environment the physico-chemical parameters of a metabolic system are externally prescribed constant quantities. The living cell, however, is able to regulate the catalytic activities through its programme of gene expression. The paper investigates to what extent this programme is amenable to optimization by evolutionary selection. Abstract, but typical mathematical models of the cellular metabolic system together with the epigenetic metabolism of biomacromolecules (enzyme protein, structural protein) were formulated and evaluated. Selection needs a criterion and the existence of an optimum of parameter values with respect to that criterion. As selection criterion served the growth rate of the cell. Parameters were the rate constants of metabolic reactions whose values may be chosen by regulation of gene expression. Two main types of parameters were distinguished: those which influence the growth rate in a monotonous manner, and those whose positive influence has a maximum. It was found that a maximum is defined if the parameter is part of an autocatalytic cycle limited by competition. The most common such situation is that of a biomacromolecule contributing directly or indirectly, but in an essential manner, to biosynthesis and growth, while competing at the same time for the share in biosynthesis. As the rate of biosynthesis has a natural upper bound, and as most proteins as end products of biosynthesis play an essential role in other parts of metabolism, the conclusion was reached that most of the controllable parameters of gene expression are amenable to evolutionary optimization.