The Pair of Phylloquinones

One, or both, of the phylloquinones QK-A and QK-B are identical, with the secondary electron acceptor A1, Reduction of A1 by A- is very fast, occurring within 30

ps, thereby preventing charge recombination as a result of re-reduction of P700+ by A-. In order to keep as much as possible of the reducing power of P700*, the A^A- couple has the most negative redox potential ofknown quinone cofactors in nature, estimated to be <-700mV [1, 131]. This redox potential is roughly 600-700mV more negative than that of the quinones QA in type [II reaction centers from purple bacteria [ 132] and PSII [ 133][ Possible causes of the unusual redox potential can be found by comparing the protein-cofactor interactions into which the different quinones are involved (Figure 2.3c); (see also Chapters 4 and 12 ). Whereas, in each of the QA molecules, both carbonyl oxygens accept hydrogen bonds from the protein [35, 40], each of the phylloquinones in PSI forms only one hydrogen bond to the backbone NH groups of L722 in PsaA and L706 in PsaB of T. elongatus [17]. According to theoretical work, the lack of a single hydrogen bond in Qk-A and QK-B may contribute to lowering their redox potentials [134]. Another contribution to a negative shift in redox potential by 50-150mV [135] may be caused by the pronounced ^stacking which is found in the crystal structure of PSI between the quinones and the indole rings of W697 in PsaA and of W677 in PsaB [17]. Using quantum chemical methods, Kaupp proposed that a semiquinone radical anion which is formed upon reduction of a quinone would prefer a T-stacked arrangement with perpendicular ring systems and an N-H---rc bond donated by the Trp side chain to the quinone headgroup; n-stacking would destabilize the semiquinone. Transient EPR spectroscopy [ 136][ and modeling of the phylloquinone binding site based on the crystallographic structure at 4 A resolution and EPR data [137], already suggested that the n-stacked arrangement is also present in the reduced state of the quinone. Reorientation of the quinone is probably prevented by sterical restraints, of which the most important are the hydro-phobic interactions of the long phytyl chain of phylloquione with the protein. The idea that n-stacking plays a role in lowering the redox potential of the phylloquinone in PSI is supported by the fact that QA in bacterial reaction centers such as that from Rhodopseudomonas viridis [35] shows a weaker n-stacking with a tryptophan residue and that such an interaction is missing in PSII [40]. It was estimated that in PSI complexes from mutant cyanobacteria that cannot synthesize phylloquinone and bind plastoquinone in the QK-A and QK-B sites [138], the redox potential of the quinone is positively shifted by 135 mV [139]. This indicates that, indeed, the major contribution to the difference in quinone redox potential between PSI and type-II reaction centers must be caused by different protein environments. Structure based calculations suggest redox potentials of -531 mV for QK-A and of -686 mV for QK-B, and the authors ascribe a crucial role to the negatively charged FX in causing the negative redox potentials [140]. The geometric counterpart to FX in type-II reaction centers is the positively charged non-heme iron. A surprising result of the study by Ishikita and Knapp is the significant difference of circa 150 mV between the quinone redox potentials in both branches, which was attributed to different conformations of the polypeptide backbones and to a change in protonation state of D575 in PsaB upon quinone reduction, which induces asymmetry in the negatively charged state because this residue has a neutral glutamine as counterpart in PsaA (Q588).

That a fast (~10ns) and a slow (~200ns) phase for reoxidation of A- exist in PSI has been known for quite a long time, but different interpretations of these biphasic kinetics are possible [1]. The observation of similar kinetics of A- reoxidation in vivo in a green alga by optical spectroscopy led the authors to suggest that two phylloquinones in the symmetrically equivalent branches of the electron transfer chain within PSI are involved in electron transfer . 141]. This hypothesis was supported by a mutagenesis study on PSI from the green alga C.rheinhardtii [142]. Mutation ofthe tryptophan in PsaA which is n-stacked to QK-A slowed the slower phase of A1- reoxidation, whereas the faster phase was slowed in the corresponding PsaB-side mutant. Within the bidirectional electron transfer scheme, this would mean that reoxidation of the quinones by the next acceptor in the electron transfer chain, FX, is faster in the B- than in the A-branch. The time-resolution of EPR methods is insufficient to detect the fast kinetic phase which was observed by optical spectroscopy, but these techniques, when applied to the same mutants, which were used by Guergova. Kuras and coworkers . 142] . also showed that the slow phase is associated with the A-branch [143]. In more extensive mutagenesis studies on cyanobacterial PSI, amino acid residues were changed; they are part of the hydrogen- bonding network with which the quinones are involved. Based on optical spectroscopy and EPR spectroscopy on these mutants [144, 145], it was also concluded that the slow kinetic phase is associated with electron transfer along the A-branch and, because the relative amplitude of this phase is greater than 70%, the authors propose that electron transfer in PSI should be regarded as being strongly asymmetric in favor of the A-branch. They furthermore suggest that the asymmetry is determined by the asymmetric properties of P700.

As a possible origin of the different rates of forward electron transfer from A1-, different redox potentials for the two quinones were suggested [140, 146], requiring different local environments of these cofactors. Another possibility is different electronic couplings between each of the quinones and FX. A striking element of asymmetry, which was suggested as being responsible for different rates of electron transfer from the two branches to FX. is a tryptophan, W673 in PsaB of T. -elongatus, which is not conserved in PsaA [17, 27]. The aromatic side chain of this residue is located in a position that suggests its suitability for a role as an intermediate electron acceptor between A. and FX . 147]. this should facilitate faster electron transfer along the B .branch. In a more recent study, mutation of the tryptophan to glycine was shown to slow the reoxidation of A- [148]. The authors concluded that this mutation did not reduce the electronic coupling between Qk-B and FX, but affected the redox potential ofthe phylloquinone bound to PsaB. There are more asymmetric features in the region near the phylloquinones and FX that might contribute to the observed difference in electron transfer rates. At a distance of circa 14 A from Qk-A, the negatively charged phosphodiester group of a phosphatidylgycerol is located, having the neutral galactosyl residue of a mono-galactosyldiacylglycerol as its pseudosymmetric counterpart located at the same distance from QK-B [17, 27]. Another interesting and unusual feature is the presence of ordered water molecules in the intra- membrane space close to the qui-nones and FX. These water molecules are arranged in two clusters, one involved in hydrogembonds with PsaA, the other with PsaB. The clusters differ in their geometries; the PsaA[bound cluster consists of five molecules, the PsaB[bound has six molecules. The water clusters could clearly play a role in affecting the protonation state of D575 in PsaB, the residue which was discussed by Ishikita and Knapp [140] as contributing to the difference in the quinone redox potentials and is hydrogen bonded to the water cluster in PsaB. The asymmetry induced by the water clusters might also contribute to differences in activation and reorganization energies of electron transfer along the two branches [28] [ In fact, from the temperature dependence of the biphasic electron transfer from phylloquinone to FX, an activation energy of 110 meV was deduced for the slower phase, whereas the faster phase was shown to be activationless [146].

Was this article helpful?

0 0

Post a comment