Evolution of selective neutrality: further considerations.

HARTL, DYKHUIZEN and DEAN (1985) have recently proposed a hypothesis for the evolution of selective neutrality. Their idea is largely based on the metabolic models of KACSER and BURNS (1973, 1979, 1981), which show that flux through a metabolic pathway will typically be a concave function of the participating enzyme activities. HARTL and co-workers suggest that, where natural selection favors maximal pathway flux, evolution will result in enzyme activities that lie on the plateau of the flux (fitness) us. activity curve. Once an enzyme’s activity has evolved to that plateau, further changes in activity will result in only negligible changes in flux (fitness). Consequently, mutant alleles with small effects on enzyme activity will then be effectively neutral. While this model provides important insight into the apparent selective neutrality of activity variation for enzymes that currently have low control (or sensitivity) coefficients, the proposed evolutionary process leading to low control coefficients has important limitations. In the course of evolving along the concave trajectory leading to maximal pathway flux, changes in enzyme activity will alter steady-state substrate pool sizes; these, in turn, may have important direct or indirect effects on fitness. Hence, while one might argue that a monotonic relationship between enzyme activity and fitness exists if all other parameters were fixed, the nature of metabolic systems suggests that this will rarely be the case. Here, we focus on the direct and indirect effects that changes in pool sizes of metabolic intermediates might have on fitness. Both theoretical considerations (KACSER and BURNS 198 1) and empirical evidence (e.g., POWERS, DIMICHELE and PLACE 1983; examples cited by KACSER and BURNS 1981) indicate that changes in enzymatic parameters can result in alteration of steady-state substrate pool sizes. In linear pathways, such changes in pool sizes can, in fact, compensate for changes in enzymatic parameters [both directly and through negative feedback (KACSER and BURNS 1981)], resulting in no net change in pathway flux. However, the changes in pool sizes themselves may directly affect fitness in at least three ways: (1) A number of metabolic intermediates are toxic at elevated concentrations [e.g., acetaldehyde (GARCIN et al. 1979; GELFAND and MCDONALD 1980) or ammonia (PROSSER 19’73)) changes in pool size for such intermediates could reduce fitness via deleterious effects on cell structure or function. (2) Changes in substrate or cofactor pool sizes could substantially influence the reaction rates for all enzymes using the common metabolite; in the case of NAD+ or NADP+, for