The 2-Methoxy Group Orientation Regulates the Redox Potential Difference between the Primary (QA) and Secondary (QB) Quinones of Type II Bacterial Photosynthetic Reaction Centers.

Recent studies have shown that only quinones with a 2-methoxy group can act simultaneously as the primary (QA) and secondary (QB) electron acceptors in photosynthetic reaction centers from purple bacteria such as Rb. sphaeroides. 13C HYSCORE measurements of the 2-methoxy group in the semiquinone states, SQA and SQB, were compared with DFT calculations of the 13C hyperfine couplings as a function of the 2-methoxy dihedral angle. X-ray structure comparisons support 2-methoxy dihedral angle assignments corresponding to a redox potential gap (ΔEm) between QA and QB of 175-193 mV. A model having a methyl group substituted for the 2-methoxy group exhibits no electron affinity difference. This is consistent with the failure of a 2-methyl ubiquinone analogue to function as QB in mutant reaction centers with a ΔEm of ∼160-195 mV. The conclusion reached is that the 2-methoxy group is the principal determinant of electron transfer from QA to QB in type II photosynthetic reaction centers with ubiquinone serving as both acceptor quinones.

Type II reaction centers (RCs) from anoxygenic and oxygenic photosynthetic RCs contain two quinones QA and QB that function in series as electron acceptors (Figure 1).[1,2] Following charge separation in the RC, QA is one-electron-reduced, generating the semiquinone radical form SQA, which then transfers the electron to QB, forming SQB. In the anoxygenic species Rhodobacter (Rb.) sphaeroides, and many others, QA and QB are chemically identical ubiquinones, and yet forward electron transfer is thermodynamically favorable by 60–75 mV.[3]

Figure 1

QA and QB quinones in the Rb. sphaeroides RC (coordinates from PDB ID: 3I4D). Hydrogen bond acceptance by the O4 atom of each quinone from the imidazole group Nδ of His-M219 (QA) and His-L190 (QB) is illustrated, as well as the Fe(II) atom that bridges the imidazoles. The 2-methoxy group of each quinone is circled in red.

The tuning of cofactor redox potentials is critically important to biological function and can often be accounted for by the electrostatic environment provided by the protein solvation.[4] This appears to be sufficient for electron transfer in oxygenic Photosystem II (PS II), where plastoquinone is active in both quinone sites. However, it cannot account for the unique ability of ubiquinone and other 2-methoxy-containing quinones to simultaneously fulfill QA and QB activity in Rb. sphaeroides RCs.[5] This was clearly demonstrated using two synthetic analogues of ubiquinone in which one or the other of the two methoxy groups was replaced by a methyl group, 2-methoxy-3,5-dimethyl-6-tetraisoprenyl-1,4-benzoquinone (2-MeO-Q) and 3-methoxy-2,5-dimethyl-6-tetraisoprenyl-1,4-benzoquinone (3-MeO-Q). Both can fully reconstitute QA function, but only 2-MeO-Q was also active as QB; 3-MeO-Q was completely inactive.[6] This points to a factor unique to the 2-methoxy group in determining functionality in the QB site.

The orientation that a methoxy group makes with the quinone ring plane has been previously investigated with regard to its influence on the quinone electron affinity and resultant redox potential (Em) value.[7−9] Qualitatively, it can be reasoned that when the methoxy group is out of the plane of the quinone ring, the main influence is the electron-withdrawing nature of the electronegative oxygen, leading to a relatively increased electron affinity. When the methoxy is in plane, the oxygen p orbitals can conjugate with the π-system of the quinone, causing electron donation to the ring, leading to a decreased electron affinity. The orientations of the methoxy groups for the QA and QB sites should, in principle, be obtainable from the atomic-level structural information available from crystal structure determinations on RC preparations. However, this is precluded by a lack of conformity on the methoxy orientations in QA and QB in the numerous available X-ray structures.[10]

We have recently introduced an additional method of estimating methoxy group orientation by using hyperfine sublevel correlation (HYSCORE) measurements of the semiquinone radicals (SQA and SQB) in RCs containing ubiquinone 13C-labeled at the two methoxy groups.[11,12] The 2-methoxy groups in QA and QB were shown to give rise to quite distinct 13C isotropic hyperfine coupling (hfc) values with the magnitude of SQB exceeding that of SQA. Comparison of these couplings with quantum mechanically predicted values for a small model (6-methyl-ubisemiquinone) as a function of the methoxy orientation demonstrated that the larger value for SQB could be at least qualitatively explained by a more out-of-plane orientation of the 2-methoxy group compared with that of SQA. As this was also associated with a higher electron affinity value, the higher redox potential of the QB ubiquinone was easily rationalized.[11,12] However, other computational approaches have indicated that the midpoint potential difference between QA and QB can be accounted for by classical electrostatics,[13] such that the added effect of the 2-methoxy group orientation would be in excess of the experimental difference of 60–75 mV.

To address this, in this Letter we carry out a full quantitative analysis of the methoxy group orientation using a larger model (Figure S1a, Supporting Information), computed over the full range of ±180°. This allows us to fully explore the complete orientation dependence of the methoxy group and directly compare with experimental determinations. We also investigate models for 2-MeO-Q and 3-MeO-Q (Figures S1b and S1c, Supporting Information), which have been instrumental in experimentally demonstrating the key role played by the 2-methoxy group in controlling the redox potential of the ubiquinone.

The theoretical dependence of the 2-methoxy isotropic 13C hfc on the methoxy orientation for our ubisemiquinone model is shown in Figure 2. The 13C couplings (Aiso) for the 2-methoxy group in SQA (1.3 MHz) and SQB (5.7 MHz, adjusted to the same unpaired spin density (0.11) on C2)[12] are indicated by the solid horizontal lines. This defines four possible dihedral angles in the two SQs (see Figure 2 legend). In a survey of X-ray structures at resolutions of at least 2.8 Å, the average values for the 2-methoxy dihedral angles (COC2C1) of QA and QB were QA = +139 ± 25° and QB = −90 ± 9°, showing that the 2-methoxy group is located on opposite sides of the ring for the two quinone sites and that the QA site quinone has a 2-methoxy orientation relatively closer to the ring plane.[10] One can see that the best agreement when comparing our estimated dihedral angles with the experimental X-ray range is +155° for SQA and −82° for SQB. These are shown in Figure 2 by the solid vertical arrows.

Figure 2

Variation in the 2-methoxy 13C isotropic hfc as a function of the COC2C1 dihedral angle for the model shown in Figure S1a (Supporting Information). The estimates for the 2-methoxy dihedral angles, giving agreement with experimental determinations, are indicated by green (QA) and pink (QB) vertical arrows, QA (−155°, −50°, 50°, 155°) and QB (−130°,-82°, 82°,130°). The best agreement with X-ray COC2C1 dihedral angle values are highlighted in bold vertical arrows. Experimental QA and QB 2-methoxy 13C hfc values for the ubsemiquinone radical are indicated as solid horizontal lines.

Figure 3 gives the variation in electron affinity value as a function of the 2-methoxy dihedral angle. Again, the orientations corresponding most closely to the crystal structure analysis are indicated by the vertical arrows. For the dihedral angles given above, the QB site quinone is estimated to have an electron affinity 175 meV higher than that of QA. This is similar to our previous calculated value using a smaller model and restricted scan.[11,12] Also included in Table 1 is the ΔEA value calculated when the 3-methoxy group (COC3C4) is fixed at its midrange value from the crystal structure analysis, −77° for QA and +88° for QB.[10] This leads to an elevation of the ΔEA value to 193 meV compared with 175 meV using the optimized 3-methoxy dihedral angle values. This illustrates, as expected, that the 3-methoxy orientation influences the electron affinity as well but that the orientation of this group is similar for both QA and QB, in contrast to the 2-methoxy group where each has a significantly different orientation.

Figure 3

Variation in EA as a function of the COC2C1 dihedral angle for the model shown in Figure S1a (Supporting Information). Bold vertical arrows indicate dihedral angle values for SQA and SQB estimated from Figure 2. The 3-methoxy dihedral angles (COC3C4) for the QA and QB points are, respectively, −63.5 and −66.1°. The horizontal lines are the EA values (QA and QB) for ubiquinone.

Table 1

Electron Affinity Difference (ΔEA, QB – QA) for Model Ubisemiquinone (2,3-diMeO-Q: 2,3-dimethoxy-5-methyl-6-isoprenyl-1,4-benzoquinone) and Corresponding Monomethoxy Structures 3-MeO-Q (3-methoxy-2,5-dimethyl-6-isoprenyl-1,4-benzoquinone) and 2-MeO-Q (2-methoxy-3,5-dimethyl-6-isoprenyl-1,4-benzoquinone)a

	Electron Affinity Differenceb(ΔEA/meV)
	A	B
2,3-diMeO-Q	175	193
2-MeO-Q	175	175
3-MeO-Q	0	10

See Figure S1, Supporting Information.

These electron affinities were obtained by re-optimizing the geometry while keeping the 2-methoxy dihedral angles fixed at their values from Figure 2 (QA: 2-MeO (CmOmC2C1) = +155°; QB: 2-MeO (CmOmC2C1) = −82°) and the 3-methoxy dihedral angles (3-MeO (CmOmC3C4)) kept at either their optimized values (column A: 3-MeO −63.5° (QA) and −66.1° (QB)) or at the mid-range crystal structure values (column B: 3-MeO −77° (QA) and +88° (QB)).

The favorable electron affinity difference (ΔEA), resulting from the 2-methoxy orientation difference in QA and QB, is significantly larger than the experimentally measured ΔEm of 60–75 mV in wild-type RCs. This implies, somewhat surprisingly, that the protein-solvation contribution to the electron-transfer reaction may be at least 100 mV, unfavorable for QA to QB electron transfer in the wild type. This is in striking contrast to electrostatic calculations, which imply a favorable solvation influence.[4] This could be due to either an overestimation of the EA difference using our theoretical model or incorrect parametrization in the electrostatic calculations. To further explore this, we have calculated the electron affinity values for 2-MeO-Q and 3-MeO-Q quinone models (Figures S1b and S1c, Supporting Information). The values are given in Table 1. Replacement of the 3-methoxy group by a methyl group (2-MeO-Q) leads to the same EA difference value, 175 meV, while replacement of the 2-methoxy group by methyl (3-MeO-Q) effectively eliminates the EA difference, demonstrating its crucial contribution. The essentially same electron affinity for QA and QB (ΔEA = 0 or 10 meV) predicted for the 3-MeO-Q model explains the lack of QB reduction observed experimentally for this quinone upon substitution in wild-type RCs. In contrast, the maintenance of electron transfer for 2-MeO-Q in wild-type RCs is readily accounted for by an electron affinity difference very similar to that exhibited by the native ubiquinone (ΔEA = 175 or 193 mV).

Of special relevance are the data obtained for a mutant with isoleucine replaced by threonine at residue M265 in the QA site. Here, the Em of QA is decreased by 100–120 mV by a mechanism that does not involve the methoxy groups.[3] No electron transfer from QA to QB is observed experimentally for the 3-MeO-Q substituted form.[11] Thus, even though a favorable site ΔEm value of 100–120 mV has been engineered to facilitate QA to QB electron transfer, the quinone lacking the 2-methoxy group is still unable to manifest electron transfer. Only an out-of-plane-oriented 2-methoxy group can elevate the electron affinity of QB sufficiently to overcome a net unfavorable site solvation effect and render electron transfer from QA thermodynamically favorable. Strong corroboration comes from a recent experimental study in which naphthoquinones were tested for QB activity.[14] These quinones, lacking methoxy groups, were found to exhibit QB redox potentials 60–100 mV more negative than expected by comparison with the native ubiquinone and were only reducible when a low-potential quinone was present in the QA site.

The heterodimeric RCs present in PS II and purple bacteria are believed to have evolved from a common homodimeric system, with efficient QA to QB electron transfer providing a key driving force.[1] For bacteria, this was accomplished using the 2-methoxy group of its ubiquinone. In PS II, which uses plastoquinone, lacking methoxy groups, an alternative mechanism is required. Most simply, this would be the local electrostatic environment, as proposed in previous electrostatic calculations.[15] A recent theoretical study suggested that a complex switching mechanism using tyrosine residues and bicarbonate may also influence QA to QB electron transfer in PS II.[16]