Search

The kinetic properties of three methylenetetrahydrofolate dehydrogenase-cyclohydrolase (D/C) enzymes (the NADP-dependent bifunctional domain of the human cytoplasmic trifunctional enzyme, the human mitochondrial NAD-dependent bifunctional enzyme, and the NAD(P)-dependent bifunctional enzyme from Photobacterium phosphoreum) were determined in both forward and reverse directions. In the forward direction, the enzymes possess widely different ratios of kcat C/Kcat D, but all channel methenylH4folate produced by the D activity to the C activity with approximately the same efficiency. A deuterium isotope effect is observed with the human NADP-dependent enzyme in both forward and reverse dehydrogenase assays, consistent with hydride transfer being rate limiting for the interconversion of methenyl- and methyleneH4folate. However, no kinetic isotope effect is observed for the overall reverse reaction (formylH4folate to methyleneH4folate). We devised an assay to measure the reverse cyclohydrolase activity independent of the dehydrogenase, and determined that the Kcat (overall reverse) for each enzyme is approximately equal to the Kcat for its reverse cyclohydrolase activity. Therefore, the rate-limiting step in the overall reverse reaction is not hydride transfer by the dehydrogenase, but the production of methenylH4folate catalyzed by the cyclohydrolase. The reverse cyclohydrolase activities of the NADP-dependent D/C and the P. phosphoreum enzymes, but not the mitochondrial NAD-dependent enzyme, can be stimulated 2-fold by the addition of 2',5'-ADP. The results suggest that the cyclohydrolases of the human NADP dependent and P. phosphoreum enzymes are optimized to catalyze the reverse reaction in the presence of bound coenzyme. These results imply that essentially all of the methenylH4folate produced by the cyclohydrolase in the reverse reaction is channeled to the dehydrogenase.

The Escherichia coli siderophore enterobactin is synthesized in response to iron starvation. 2,3-Dihydro-2,3-dihydroxybenzoate dehydrogenase (EntA) produces 2,3-dihydroxybenzoate (DHB), a biosynthetic intermediate. 2,3-Dihydroxybenzoate-AMP ligase (EntE) adenylates DHB, activating it for attachment to the NRPS substrate holo-EntB. Using analytical ultracentrifugation, we found that EntA undergoes concentration-dependent dimer-tetramer self-association (K(D) = 12.3 μM). We further found that EntA can form a specific complex with EntE. Pull-down assays revealed that recombinant EntA bait pulled down EntE from E. coli lysates, whereas recombinant EntE bait could pull down EntA. Addition of the SMCC cross-linker to a mixture of EntA and EntE resulted in a cross-linked product with a molecular mass of250 kDa, suggesting a complex stoichiometry of one EntA tetramer and four EntE monomers. The effect of EntA on EntE activity was also examined. Addition of a 4-fold excess of EntA to an EntE assay mixture resulted in a 6-fold stimulation of EntE activity. EntA was also found to perturb the FRET signal between EntE donor residues and EntE-bound DHB. By following the EntA-dependent decrease in the magnitude of the EntE-DHB FRET signal, EntA-EntE binding behavior was found to be sigmoidal, suggesting the presence of both low- and high-affinity binding sites. The EntA-EntE interaction was also directly measured by isothermal titration calorimetry at 10 °C. The resulting binding isotherm fit well to a model describing two binding sites, supporting our AUC and fluorescence data. Taken together, our data show that tetrameric EntA optimally interacts with EntE, resulting in an enhancement of EntE activity.

FhuA, an outer membrane receptor of Escherichia coli, facilitates transport of hydroxamate siderophores and siderophore-antibiotic conjugates. The cytoplasmic membrane complex TonB-ExbB-ExbD provides energy for transport via the proton motive force. This energy is transduced by protein-protein interactions between TonB and FhuA, but the molecular determinants of these interactions remain uncharacterized. Our analyses of FhuA and two recombinant TonB species by surface plasmon resonance revealed that TonB undergoes a kinetically limiting rearrangement upon initial interaction with FhuA: an intermediate TonB-FhuA complex of 1:1 stoichiometry was detected. The intermediate then recruits a second TonB protein. Addition of ferricrocin, a FhuA-specific ligand, enhanced amounts of the 2:1 complex but was not essential for its formation. To assess the role of the cork domain of FhuA in forming a 2:1 TonB-FhuA complex, we tested a FhuA deletion (residues 21-128) for its ability to interact with TonB. Analytical ultracentrifugation demonstrated that deletion of this region of the cork domain resulted in a 1:1 complex. Furthermore, the high-affinity 2:1 complex requires the N-terminal region of TonB. Together these in vitro experiments establish that TonB-FhuA interactions require sequential steps of kinetically limiting rearrangements. Additionally, domains that contribute to complex formation were identified in TonB and in FhuA.