Why is dihydrodipicolinate synthase a tetramer?

Dihydrodipicolinate synthase (DHDPS, E.C. 4.2.1.52) catalyses the final branch-point reaction in lysine biosynthesis and is feedback inhibited by lysine. The enzyme is a homotetramer of 125 kDa that is composed of a dimer of dimers, where two monomers associate strongly to form tight-dimer units and two of these units associate to form the wild-type tetrameric structure. Efficient catalysis by DHDPS requires the movement of protons between the active site and bulk solvent via a proton-relay triad. This motif is formed on association of the two monomers of the tight-dimer and is comprised of residues from both monomers, suggesting that establishment of the tight-dimer structure is required for optimal catalysis. However, there is no obvious function afforded by the association of tight-dimer units to form the tetramer. To understand the functional and mechanistic contributions of the tetrameric structure of DHDPS, a suite of amino acid mutations of residues at the dimer-dimer interface was designed to produce dimeric variants of the enzyme for comparative functional analysis. A complete re-evaluation of the composition of the dimer-dimer interface contact surface was undertaken, identifying five intersubunit contact residues and a water-blidging network that had not previously been reported. The variants DHDPS-L197D, DHDPS-L197Y, DHDPS-Q196D, DHDPS-D193A, DHDPS-D193Y, and DHDPS-Q234D were produced by site-directed mutagenesis. Comprehensive biophysical characterisation suggested that all variants had altered quaternary structure with respect to wild-type DHDPS. DHDPS-L197D and DHDPS-L197Y were found to exist as stable obligate dimers in solution. DHDPS-Q196D and DHDPS-D193A were shown to exist in equilibrium between dimer and tetramer with Ka values of 10⁵˙⁶ M-¹and 10⁵˙⁸ M-¹ respectively. Rigorous kinetic analysis showed that each variant retained catalytic activity, but that this activity was attenuated from the level of the wild-type enzyme. This was most pronounced in the dimeric variants, DHDPS-L197D and DHDPS-L197Y, which displayed 2.5% and 1.4% of wild-type activity, respectively. Each variant demonstrated a raised Km value for the first binding substrate, pyruvate. In DHDPS-L197D and DHDPS-L197Y, Kmpyr values were raised four-fold and seven-fold over that of wild-type DHDPS. Interestingly, it was also found that incubation of these variants at raised temperature gradually increased the catalytic turnover rate of each enzyme. Both DHDPS-L197D and DHDPS-L197Y were successfully crystallised and their structures solved to 2.0 Å and 1.7 Å resolutions, respectively. Both structures retained the tertiary and quaternary details of the wild-type tight-dimer with a very high degree of fidelity and no rerrangements at the active site were observed to account for the attenuation of catalysis in the variants. Electron density suggested that pyruvate remained covalently attached at the active site, despite crystallisation and purification in the absence of the substrate. The pyruvate adduct was identified as the trapped tetrahedral intermediate of the reaction between pyruvate and K161. The dehydration of this intermediate is catalysed by the proton-relay triad that includes a tyrosine residue, Y107, that interdigitates into the active site from the neighbouring monomer. In conclusion, it is proposed that a function of the tetrameric structure of E. coli DHDPS is to constrain dynamic movements of the monomers of the tight-dimer with respect to each other. This is thought to allow for correct positioning of Y107 in the proton-relay triad, resulting in optimal catalysis, and circumventing the trapping of the pyruvate reaction intermediate. Such constraint of protein dynamics may offer a general rationale for the oligomeric structure of many enzymes