Time-resolved fluorescence analysis of the mobile flavin cofactor in p-hydroxybenzoate hydroxylase

Abstract

Abstract. Conformational heterogeneity of the FAD cofactor in p-hydroxybenzoate hydroxylase (PHBH) was investigated with time-resolved polarized flavin fluorescence. For binary enzyme/substrate (analogue) complexes of wild-type PHBH and Tyr222 mutants, crystallographic studies have revealed two distinct flavin conformations; the 'in' conformation with the isoalloxazine ring located in the active site, and the 'out' conformation with the isoalloxazine ring disposed towards the protein surface. Fluorescence-lifetime analysis of these complexes revealed similar lifetime distributions for the 'in' and 'out' conformations. The reason for this is twofold. First, the active site of PHBH contains various potential fluorescence-quenching sites close to the flavin. Fluorescence analysis of uncomplexed PHBH Y222V and Y222A showed that Tyr222 is responsible for picosecond fluorescence quenching free enzyme. In addition, other potential quenching sites, including a tryptophan and two tyrosines involved in substrate binding, are located nearby. Since the shortest distance between these quenching sites and the isoalloxazine ring differs only little on average, these aromatic residues are likely to contribute to fluorescence quenching. Second, the effect of flavin conformation on the fluorescence lifetime distribution is blurred by binding of the aromatic substrates: saturation with aromatic substrates induces highly efficient fluorescence quenching. The flavin conformation is therefore only reflected in the small relative contributions of the longer lifetimes.

Monomer Formation and Function of p‐Hydroxybenzoate Hydroxylase in Reverse Micelles and in Dimethylsulfoxide/Water Mixtures

“…A single rotational correlation time ( ϕ prot ) of 32 ns was obtained for FAD‐PHBH in aqueous solution, and is in excellent agreement with that expected for a dimeric PHBH molecule 22. Although the 1.6 ns and 3.1 ns fluorescence lifetime components only contribute to the total fluorescence decay by 7 % and 2 %, respectively,38 they are the main carriers of the anisotropy signal in the nanosecond time region.…”

“…Although flavin fluorescence in PHBH is strongly quenched, the quenching is considerably less than in E. coli glutathione reductase (GR): the ratio of fluorescence quantum yields Q PHBH / Q GR ≈15. For substrate‐free PHBH in aqueous buffer the fluorescence lifetime pattern shows high heterogeneity, with at least four lifetime components covering a dynamic range between 15 ps and 3 ns 38. This could be due to the various dynamic states previously reported for PHBH 39.…”

“…From the similarity in structure and fluorescence characteristics between PHBH and other tyrosine‐containing flavoproteins such as glutathione reductase36 and NADPH peroxidase,40 the mechanism of quenching most probably involves photoinduced electron transfer from particular tyrosine residues to the light‐excited flavin. Tyr222 is very likely to be the quenching residue 38. In single‐molecule studies the Y222A variant does not show the fluctuations in fluorescence due to open‐to‐in positional dynamics 39…”

“…In many flavoenzymes the fluorescence of the flavin prosthetic group shows low molecular brightness due to severe quenching 36. 37 Recently, we have systematically investigated the experimentally observed total flavin fluorescence decay of wild‐type PHBH and its Y222A and Y222V muteins 38. Although flavin fluorescence in PHBH is strongly quenched, the quenching is considerably less than in E. coli glutathione reductase (GR): the ratio of fluorescence quantum yields Q PHBH / Q GR ≈15.…”

“…The changes in the average fluorescence lifetime of FAD as function of DMSO concentration were too gradual to give a reliable midpoint concentration, since the standard deviation is much larger than the actual value: [DMSO] 1/2 =4.9±35.3 % ( v / v ; http://www.wiley-vch.de/contents/jc_2268/2008/f700267_s.pdf). A plausible explanation is that the average fluorescence lifetime is dominated by the very short fluorescence lifetime component arising from picosecond quenching 38. This lifetime component then undergoes a gradual, noncooperative change with a midpoint at 20 %.…”