Projekte

The bioluminescent reaction of luciferase isolated from Vibrio harveyi, strain M17, was initiated by mixing the luciferase-bound flavin 4a-hydroperoxide intermediate, purified in advance, with a long-chain aldehyde (dodecanal or octanal) at -4 °C . Measurements of absorbance changes from 300 to 600 nm during the course of the reaction revealed the existence of three sets of isosbestic points and three kinetic phases, the second of which parallels kinetically the decay of bioluminescence, measured concurrently. The absorbance changes in this second step and the decay of light emission exhibited similar deuterium isotope effects; this is postulated to be the step giving rise to the excited state and the enzyme-bound flavin 4a-hydroxide. The first step of the reaction, however, did not show an isotope effect; the intermediate thereby formed, observed here for the first time, is postulated to correspond to the luciferase-bound flavin 4a-peroxyhemiacetal.

Bioluminescence catalyzed by bacterial luciferases was measured using FMN, iso-FMN (6-methyl-8-nor-FMN), and FMN analogs carrying the following substituents at position 8: -H, -Cl, -F, SMe, SOMe, S02Me, or -0Me. The first-order rate constants for the decay of light emission correlate with the one-electron oxidation potentials of the 4a,5-dihydro forms of the FMN analogs. To determine thevalues of these potentials, isoalloxazine (flavin) derivatives having the 4a,5-propano-4a,5-dihydros tructure and -H, -CH3, -C1, -OCH3, and -NH2 as substituents at position 8 have been synthesized as models for the 4a-peroxy-4a,5-dihydroflavin intermediates occurring during catalysis by the flavin-dependent monooxygenase luciferase. The tetrahydropyrrole ring between positions 4a and 5 of these isoalloxazine derivatives stabilizes the 4a,5-dihydroflavin by impeding formation of the thermodynamically more stable 1,5-dihydro form. One-electron oxidation potentials (Eobs) were measured by cyclic voltammetry and used to determine the empirical coefficients in the Swain equation. On the basis of this, the one-electron oxidation potentials of 4a,5-propano-4a,5-dihydraon alogs with other substituents in position 8 were calculated (Ecalc). The bioluminescence reaction rate is fastest with FMN analogs of lowest oxidation potential; Le., the slope of the correlation is negative. This indicates that in the rate-limiting step the 4a,5-dihydroflavin moiety donates negative charge. The results are compatible with an intramolecular, chemically initiated electron exchange luminescence mechanism for the bacterial luciferase reaction. A good correlation was also found between EOba nd the literature values of the 2-electron oxidation/reduction potentials (Eredox) of the couple FIox/1,5-dihydro-FIrf~o r the flavin derivatives having the same substituent at position 8. The effects of the substituent in position 8 on the redox properties of 1,5-dihydro- and 4a,5-dihydroflavins are thus essentially the same. This indicates that the earlier use of readily available redox potentials for FIox/1,5-dihydro-FIred for studying reactions involving the 4aS-dihydro isomer was sound.

Yellow fluorescent protein (YFP), which has a bound FMN, was isolated from the marine bacterium Vibrio fischeri strain Y-1b. Its presence in a luciferase [alkanal monooxygenase (FMN-linked); alkanal, reduced-FMN:oxygen oxidoreductase (1-hydroxylating, luminescing), EC 1.14.14.3] reaction mixture causes a striking color change, and an increase in bioluminescence intensity, as well as a faster rate of intensity decay, so that the quantum yield is not changed. The emission spectrum shows two distinct color bands, one at 490 nm attributed to the unaltered emission of the luciferase system, the other peaking in the yellow around 540 nm due to YFP emission. The kinetics of the two color bands differ, so the spectrum changes with time. The yellow emission reaches its initial maximum intensity later than the blue, and then both blue and yellow emissions decay exponentially with nearly the same pseudo-first-order rate constants, linearly dependent on [YFP] (from 0.01 sec-1 with no YFP to a maximum of ≈0.1 sec-1 at 4°C) but exhibiting a saturation behavior. The data can be interpreted by assuming the interaction of YFP with the peroxyhemiacetal intermediate in the luciferase reaction to form an unstable new complex whose breakdown gives the yellow emitter in its excited state. This simple model fits well the data at [YFP] < 15 µM. The results indicate that a single primary excited state cannot be responsible for the blue and the yellow emissions.