On the involvement of single-bond rotation in the primary photochemistry of photoactive yellow protein

Abstract

Prior experimental observations, as well as theoretical considerations, have led to the proposal that C(4)-C(7) single-bond rotation may play an important role in the primary photochemistry of photoactive yellow protein (PYP). We therefore synthesized an analog of this protein's 4-hydroxy-cinnamic acid chromophore, (5-hydroxy indan-(1E)-ylidene)acetic acid, in which rotation across the C(4)-C(7) single bond has been locked with an ethane bridge, and we reconstituted the apo form of the wild-type protein and its R52A derivative with this chromophore analog. In PYP reconstituted with the rotation-locked chromophore, 1), absorption spectra of ground and intermediate states are slightly blue-shifted; 2), the quantum yield of photochemistry is ∼60% reduced; 3), the excited-state dynamics of the chromophore are accelerated; and 4), dynamics of the thermal recovery reaction of the protein are accelerated. A significant finding was that the yield of the transient ground-state intermediate in the early phase of the photocycle was considerably higher in the rotation-locked samples than in the corresponding samples reconstituted with p-coumaric acid. In contrast to theoretical predictions, the initial photocycle dynamics of PYP were observed to be not affected by the charge of the amino acid residue at position 52, which was varied by 1), varying the pH of the sample between 5 and 10; and 2), site-directed mutagenesis to construct R52A. These results imply that C(4)-C(7) single-bond rotation in PYP is not an alternative to C(7)=C(8) double-bond rotation, in case the nearby positive charge of R52 is absent, but rather facilitates, presumably with a compensatory movement, the physiological Z/E isomerization of the blue-light-absorbing chromophore.

Figure 1

Schematic representation of the structure of the chromophore binding pocket of wild-type PYP before (A) and after (B) the two modifications characterized in this study, namely, the R52A mutation and replacement of the coumaryl chromophore by a rotation-locked derivative, (5-hydroxy indan-(1E)-ylidene)acetic acid.

Figure 2

Characterization of the static UV-vis absorbance and fluorescence emission of the four PYP derivatives. (A) Absorption spectra of wild-type PYP (black), WTRL (red), R52A (blue), and R52ARL (green) in 20 mM Tris buffer, pH 8.0. The absorption maxima are centered at 446, 443, 450, and 445 nm, respectively. (B) The corresponding emission spectra, with excitation of all samples at their respective absorbance maxima.

Figure 3

Target-model-based analysis of the initial stages of the photocycle of the four PYP derivatives. I₀ decays into I₁ with a time constant of ∼1 ns and a yield of 100%. The fitted lifetimes and relative populations of the states accessible from the ES are shown. The excited states of wild-type and R52A were found to decay biexponentially with time constants of 0.7 and 6.8 ps and 1.9 and 27 ps, respectively, with the quantum yield of I₀ formation largest for the fastest fraction, in agreement with Larsen et al. (18). To facilitate comparison with the rotation-locked samples, the data for excited-state decay for all four proteins were fitted here with single exponents of 1.5 ps and 2.3 ps, respectively. The yield of the ground state in WTRL and R52ARL is very low, and can be fitted with yields between 0 and 10%. The reduction of this yield and that of I₀ in the rotation-locked samples is due to a four- to sixfold increase in the rate of ES → GSI.

Figure 4

Species-associated difference spectra (SADS) of the four PYP derivatives, obtained from fitting the target model of Fig. 3 to the data. The SADS of wild-type PYP are in good agreement with earlier reports in the literature (e.g., (18,39). The GSI spectrum of the rotation-locked samples is easily resolved, as it appeared pronounced in the data, as shown in Fig. S2. The quantum yields depicted in Fig. 3 were determined by scaling the SADS of each state to that of the ES, which may result in ∼10% uncertainty. An exception is the I₁ state, for which we have used a yield of 100% for the I₀ to I₁ transition, as determined with femtosecond midinfrared spectroscopy (14,19). Color code as in Fig. 2.

Figure 5

pH dependence of the ultrafast kinetics of wild-type PYP. At each pH, the dynamics could be fitted with four time constants. The EADS for each time constant are indicated: 0.9, 0.9, and 0.8 ps (black); 6, 5, and 5.5 ps (red); 1.4, 1.7, and 1.4 ns (blue), and infinite (green), for wild-type PYP at pH 5, 8, and 10, respectively.

Figure 6

Time traces of the absorbance changes during the later part of the photocycle of the four PYP derivatives, recorded at 360 nm. The pB intermediates absorb maximally at 360 nm. Their rise and decay can be fitted with the time constants, as shown in Table S2 (see Fig. S1). Color code as in Fig. 2.