Summary
Glutamate dehydrogenases (EC 1.4.1.2–4) catalyse the oxidative deamination of l-glutamate to α-ketoglutarate using NAD+ and/or NADP+ as a cofactor. Subunits of homo-hexameric bacterial enzymes comprise a substrate-binding Domain I followed by a nucleotide binding Domain II. The reaction occurs in a catalytic cleft between the two domains. Although conserved residues in the nucleotide-binding domains of various dehydrogenases have been linked to cofactor preferences, the structural basis for specificity in the glutamate dehydrogenase (GDH) family remains poorly understood. Here, the refined crystal structure of Escherichia coli GDH in the absence of reactants is described at 2.5Å resolution. Modelling of NADP+ in Domain II reveals the potential contribution of positively charged residues from a neighbouring α-helical hairpin to phosphate recognition. In addition, a serine residue that follows the P7 aspartate is presumed to form a hydrogen bond to the 2’-phosphate. Mutagenesis and kinetic analysis confirms the importance of these residues in NADP+ recognition. Surprisingly, one of the positively charged residues is conserved in all sequences of NAD+ dependent enzymes, but the conformations adopted by the corresponding regions in proteins whose structure has been solved preclude their contribution toward the co-ordination of the 2’-ribose phosphate of NADP+. These studies clarify the sequence/structure relationships in bacterial glutamate dehydrogenases, revealing that identical residues may specify different coenzyme preferences, depending on the structural context. Primary sequence alone is therefore not a reliable guide for predicting coenzyme specificity. We also consider how it is possible for a single sequence to accommodate both coenzymes in the dual specificity GDHs of animals.