The multi-domain enzyme phenylalanine hydroxylase (PAH) catalyzes the hydroxylation of dietary I-phenylalanine (Phe) to I-tyrosine. Inherited mutations that result in PAH enzyme deficiency are the genetic cause of the autosomal recessive disorder phenylketonuria. Phe is the substrate for the PAH active site, but also an allosteric ligand that increases enzyme activity. Phe has been proposed to bind, in addition to the catalytic domain, a site at the PAH N-terminal regulatory domain (PAH-RD), to activate the enzyme via an unclear mechanism. Here we report the crystal structure of human PAH-RD bound with Phe at 1.8 Å resolution, revealing a homodimer of ACT folds with Phe bound at the dimer interface. This work delivers the structural evidence to support previous solution studies that a binding site exists in the RD for Phe, and that Phe binding results in dimerization of PAH-RD. Consistent with our structural observation, a disease-associated PAH mutant impaired in Phe binding disrupts the monomer:dimer equilibrium of PAH-RD. Our data therefore support an emerging model of PAH allosteric regulation, whereby Phe binds to PAH-RD and mediates the dimerization of regulatory modules that would bring about conformational changes to activate the enzyme.

Classic galactosemia is a potentially lethal disease caused by the dysfunction of galactose 1-phosphate uridylyltransferase (GALT). Over 300 disease-associated GALT mutations have been reported, with the majority being missense changes, although a better understanding of their underlying molecular effects has been hindered by the lack of structural information for the human enzyme. Here, we present the 1.9 Å resolution crystal structure of human GALT (hGALT) ternary complex, revealing a homodimer arrangement that contains a covalent uridylylated intermediate and glucose-1-phosphate in the active site, as well as a structural zinc-binding site, per monomer. hGALT reveals significant structural differences from bacterial GALT homologues in metal ligation and dimer interactions, and therefore is a zbetter model for understanding the molecular consequences of disease mutations. Both uridylylation and zinc binding influence the stability and aggregation tendency of hGALT. This has implications for disease-associated variants where p.Gln188Arg, the most commonly detected, increases the rate of aggregation in the absence of zinc likely due to its reduced ability to form the uridylylated intermediate. As such our structure serves as a template in the future design of pharmacological chaperone therapies and opens new concepts about the roles of metal binding and activity in protein misfolding by disease-associated mutants.