Why Do Enzymes Have an Optimal Temperature? A Meta-Analysis Testing Three Models for Activity Decline at High T

Before the 1990s, the fact that almost all enzymes had an optimal temperature (Topt) above which catalytic activity declined was explained by thermal denaturation. However, since then it has been demonstrated for many enzymes that activity declines above Topt in the absence of measurable denaturation. Two main models have been proposed to explain this: the Inactivation Equilibrium Model proposed by Daniel and Danson, and the Macromolecular Rate hypothesis put forth by the Arcus group. Comparisons of these two models have, until now, used activity vs T data sets from a limited number of enzymes, with relatively few k(T) measurements. We present here a meta-analysis of 260 published data sets from 27 enzymes and 5 multienzyme metabolic processes. First, we note that fits for kcat vs T data sets give thermodynamic and activation parameters related specifically to the catalytic step; data sets with a more complex independent variable (e.g., kcat/Km, Vmax/Km, v0, enzyme activity) yield fitted parameters that could characterize catalysis, substrate binding, or both. Additionally, for both models, the fitted activation parameters display enthalpy/entropy compensation (ΔH‡ vs ΔS‡ plots are strictly linear, with R2 > 0.98) that is found to be “apparent”, i.e., statistical rather than thermodynamic. We also conclude that for symmetric activity vs T curves, the fits from the two models are equally good. However, for asymmetric activity vs T curves, the Inactivation Equilibrium Model yields significantly better fits. This problem was addressed recently by the Arcus group, which presented a “two-state” modification of their MMRt model with six adjustable parameters, which can fit asymmetric k vs T peaks. This compares with the Inactivation Equilibrium model which has four adjustable parameters.