ABCE1 has two nucleotide-binding sites, nearly symmetric. Mutate one — hydrolysis slows by half. Mutate the other — hydrolysis speeds up tenfold. The same kind of damage, applied to an apparently equivalent site, produces opposite outcomes.
Schäffner, Smith, Tampé, and Grubmüller (JCTC 2026) show this is Braess' Paradox in biochemistry. In traffic networks, adding a road can slow everyone down by creating a new route that individually rational drivers overuse, congesting the system. Removing the road restores flow. In ABCE1, the wild-type enzyme has access to a kinetic trap — a conformational state that the enzyme visits because both sites are functional, but that slows the overall cycle. The mutation at one site eliminates access to that trap. The enzyme loses an option and gains speed.
No direct cooperativity is needed. The asymmetry emerges entirely from population balance: the opening and closing of both sites is structurally coupled (they move together), so changing one site's kinetics shifts the population distribution across all conformational states. A 13-state Markov model reproduces the experimental data using only this mechanism. The mutation doesn't accelerate catalysis. It redirects traffic away from the slow lane.
The thing that looks like damage is a shortcut. The intact enzyme is stuck in a trap that exists only because it has more options.