Biosynthesis of riboflavin: cloning, sequencing, mapping, and expression of the gene coding for GTP cyclohydrolase II in Escherichia coli.

GTP cyclohydrolase II catalyzes the first committed step in the biosynthesis of riboflavin. The gene coding for this enzyme in Escherichia coli has been cloned by marker rescue. Sequencing indicated an open reading frame of 588 bp coding for a 21.8-kDa peptide of 196 amino acids. The gene was mapped to a position at 28.2 min on the E. coli chromosome and is identical with ribA. GTP cyclohydrolase II was overexpressed in a recombinant strain carrying a plasmid with the cloned gene. The enzyme was purified to homogeneity from the recombinant strain. The N-terminal sequence determined by Edman degradation was identical to the predicted sequence. The sequence is homologous to the 3' part of the central open reading frame in the riboflavin operon of Bacillus subtilis.

Fluorescence study of the ligand stereospecificity for binding to lumazine protein.

6,7-Dimethyllumazine derivatives, substituted at the 8-position with aldityls or monohydroxyalkyl groups, have been examined for their binding ability to lumazine apo-protein from two strains of Photobacterium phosphoreum using fluorescence dynamics techniques. On the protein the lumazine has a nearly monoexponential decay of fluorescence with lifetime 13.8 ns (20 degrees C). In free solution the lifetime is 9.6 ns. The concentration of free and bound lumazine in an equilibrium mixture can be recovered readily by analysis of the fluorescence decay. Only the aldityl derivatives D-xylityl and 3'-deoxy-D-ribityl, having stereoconfigurations at the 2' and 4' positions identical to the natural ligand, 8-(1'-D-ribityl), show comparable dissociation constants (0.3 microM, 20 degrees C, pH 7.0). D-Erythrityl and L-arabityl have dissociation constants of 1-2 microM. All other ligands show no interaction at all or have dissociation constants in the range 6-80 microM, which can still be determined semi-quantitatively using the fluorescence decay technique. In the case of these very weakly bound ligands, unambiguous detection of bound ligand can be shown by a long correlation time (23 ns, 2 degrees C) for the fluorescence anisotropy decay. Examination of the bound D-xylityl compound's fluorescence anisotropy decay at high time resolution (< 100 ps) shows rigid association, i.e. no mobility independent of the macromolecule. All bound ligands appear to be similarly positioned in the binding site. The influence of the stereoconfiguration at the 8-position found for lumazine protein parallels that previously observed for the enzyme riboflavin synthase, where the lumazines are substrates or inhibitors. This is consistent with the finding of significant sequence similarity between these proteins. The binding rigidity may have implications for the mechanism of the enzyme.

Biosynthesis of riboflavin: mechanism of formation of the ribitylamino linkage.

Feeding experiments with Ashbya gossypii followed by NMR analysis of the resulting riboflavin showed incorporation of deuterium from D-[2-2H]ribose at C-2' and from D-[1-2H]ribose in the pro-R position at C-1' of the ribityl side chain. The results rule out an Amadori rearrangement mechanism for the reduction of the ribosylamino to the ribitylamino linkage and point to formation of a Schiff base that is reduced stereospecifically opposite to the face from which the oxygen has departed. As prerequisite for the analysis, the 1H NMR signals for the pro-R and pro-S hydrogens at C-1' of 6,7-dimethyl-8-ribityllumazine and riboflavin and its tetraacetate were assigned with the aid of synthetic stereospecifically deuteriated samples.