Affinity and activity of non-native quinones at the Q(B) site of bacterial photosynthetic reaction centers

Abstract

Purple, photosynthetic reaction centers from Rhodobacter sphaeroides bacteria use ubiquinone (UQ10) as both primary (Q(A)) and secondary (Q(B)) electron acceptors. Many quinones reconstitute Q(A) function, while a few will act as Q(B). Nine quinones were tested for their ability to bind and reconstitute Q(A) and Q(B) functions. Only ubiquinone (UQ) reconstitutes both functions in the same protein. The affinities of the non-native quinones for the Q(B) site were determined by a competitive inhibition assay. The affinities of benzoquinones, naphthoquinone (NQ), and 2-methyl-NQ for the Q(B) site are 7 ± 3 times weaker than that at Q(A) site. However, di-ortho-substituted NQs and anthraquinone bind tightly to the Q(A) site (K d ≤ 200 nM), and ≥1,000 times more weakly to the Q(B) site, perhaps setting a limit on the size of the site. With a low-potential electron donor, 2-methyl, 3-dimethylamino-1,4-NQ, (Me-diMeAm-NQ) at Q(A), Q(B) reduction is 260 meV, more favorable than with UQ as Q(A). Electron transfer from Me-diMeAm-NQ at the Q(A) site to NQ at the Q(B) site can be detected. In the Q(B) site, the NQ semiquinone is estimated to be ≈60-100 meV higher in energy than the UQ semiquinone, while in the Q(A) site, the semiquinone energy level is similar or lower with NQ than with UQ. Thus, the NQ semiquinone is more stable in the Q(A) than in the Q(B) site. In contrast, the native UQ semiquinone is ≈60 meV lower in energy in the Q(B) than in the Q(A) site, stabilizing forward electron transfer from Q(A) to Q(B).

Figure 1

Electron transfer pathways in Rb. sphaeroides reaction centers with no external electron donor. P is the bacteriochlorophyll dimer, H is the bacteriophophytin. When P absorbs a photon, the excited P* reduces H in 3 ps. With UQ as Q_A and Q_B, the electron is transferred to Q_A in 200 ps (k_HA) and then to Q_B in 100 μs (k_AB). At each stage in the reaction cycle there are competing charge recombination reactions that transfer the electron back to P⁺ either by direct electron tunneling or via repopulation of a higher energy states (dashed lines). P⁺H⁻ decays to the ground state in ≈10 ns (k_HP is 7×10⁷ s⁻¹) if there is no Q_A. Once the electron reaches the quinones it can return to P⁺ by either direct tunneling (k_AP in P⁺Q_A⁻ or k_BP in P⁺Q_B⁻ RCs) or via a thermal back reaction. k_AHP is the rate of P⁺Q_A⁻ charge recombination via P⁺H⁻Q_A, while k_BAP is the back reaction rate from P⁺Q_AQ_B⁻ via P⁺Q_A⁻Q_B. The observed rates k_AP^obs or k_BP^obs are the sum of the rates by the two routes. In native RCs, if there is no Q_B, k_AP^obs ≈ k_AP is 10 s⁻¹ via direct electron tunneling from Q_A⁻ to P⁺, while the back reaction from UQ_B⁻ goes via reformation of P⁺Q_A⁻ at k_BP^obs ≈ k_BAP which is 1 s⁻¹.

Figure 2

Occupancy of Q_A and Q_B sites in a mixture of 2 quinones [RC_T]=1 M; K_d XQ_A = 0.06 M; K_d YQ_A = 8 M; K_d YQ_B = 80 M. The concentration of XQ is 10 M (light lines): or 50 M (darker lines):. The fraction of RCs as XQ_A:E_B (Dotted-Dashed lines); YQ_A:E_B + YQ_A:YQ_B (Dashed lines); XQ_A:YQ_B (Solid lines). E_B represents an empty Q_B site. XQ does not bind to the Q_B site so there is no XQ_A:XQ_B. Where there is the same quinone in Q_A and Q_B sites (YQ_A:YQ_B) there is no forward electron transfer from the semiquinone YQ_A⁻ to YQ_B. Desired XQ_A:YQ_B with 10 μM XQ_T (shaded darker region) or with 50 μM XQ_T (shaded lighter region). Mathematica equations used to find distribution of quinones bound to each site given in Supplementary Material (Eqn. s1).

Figure 3

Reconstitution of Q_B function with UQ₁ following addition of non-native quinones as competitive inhibitors. (○): UQ₁ alone; (▲): NQ; ( ): Me-NQ; (▽): diMe-NQ. Titration carried out with 1 μM RCs, 0.005% LDAO, 10mM Tris, pH 7.8 at 298 2 K. Theory lines show the best fit K_i derived from K_d^app and the fraction of P⁺Q_B⁻ at saturating UQ₁ using values given in Table 2 in equation 4.

Figure 4

Comparison of quinone binding affinity at Q_A and Q_B sites. (■) This work (Table 1). Ubiquinone with different tail lengths in (○): water-saturated hexane (Warncke et al. 1994); (●):10mM Tris, pH= 8.0, 0.1% LDAO, 298 K (McComb et al. 1990). Small numbers refer to the number of isoprenyl units in the UQ tail for measurements in either water or hexane. Straight line: $K_d^{QB}/K_d^{QA}=7$.

Figure 5

The rate of decay of P⁺, monitored at 430 nm after an actinic flash shows which quinone is in the Q_A and Q_B binding sites. The kinetic traces are normalized to give the same initial amplitude (ΔA₄₃₀) to facilitate comparison. The signal from RCs retaining residual UQ_A is subtracted. Residuals from the analysis of the kinetic traces are shown in the two lowest traces. 1 μM RCs, 0.005% LDAO, 10 mM Tris, pH 7.8 at 298±2 K. Top: faster trace: 300 μM NQ: k₁ =k_AP=7.8 s⁻¹, A₁ = 95%; constant A₀ = 5%; slower trace: 2 μM Me-diMeAm-NQ and 300 μM NQ: k₁= k_AP= 7.8 s⁻¹, A₁ = 60%; k₂ = k_BP^obs=0.50 s⁻¹, A₂ = 36%. A₀ = 4%. ΔA₄₃₀ 1.7 shown. Insert: Me-diMeAm-NQ at Q_A at 2000 fold faster data acquisition rate: k₁=k_AP^obs= 3000 s⁻¹, A₁ = 95%, A₀= 5%. No NQ is added here.

Figure 6

Fraction of Me-diMeAm-NQ in Q_A and NQ in Q_A or Q_B sites as a function of NQ concentration. The fraction of NQ_A:E_B, Me-diMeAm-Q_A:E_B and Me-diMeAm-Q_A:NQ_B RCs obtained from a three exponential fit of the back reaction kinetics with fixed rate constants (Fig. 5). ( ) RCs with Me-diMeAm-NQ_A and an empty Q_B site is derived from the initial amplitude of ΔA₄₃₀ decaying at k₁ = 3000 s⁻¹ normalized using ΔA₄₃₀/Φ_HA where Φ_HA is 60%; (▽) RCs with NQ_A and no Q_B from the initial amplitude of ΔA₄₃₀ decaying at k₂ = 7.8 s⁻¹; Φ_HA for NQ is 100%. (▲) RCs with Me-diMeAm-NQ_A:NQ_B obtained from the initial amplitude of ΔA₄₃₀ decaying at k₃ = 0.5 s⁻¹ normalized using ΔA₄₃₀/Φ_AB, with Φ_AB of 75%. Theoretical lines generated with Mathematica using Eqn. s1 in supplementary material with 1 M RCs and 50 M Me-diMeAm-NQ. K_d of Me-diMeAm-NQ at Q_A is 0.06 M. K_d^app of NQ at Q_A is 1640 μM when it is competing with 50 μM Me-diMeAm-NQ. The K_d for NQ is 175 μM in the Q_B site with either NQ_A or Me-diMeAm-NQ_A. There is negligible Me-diMeAm-Q_B since its K_d in the Q_B site is >100 M. The dashed line tracks the amplitude of the fast k_AP^obs indicative of RCs with Me-diMeAm-NQ_A without NQ_B. The dotted curve represents NQ_A. The solid line represents the hybrid Me-diMeAm-NQ_A:NQ_B RCs. The shaded region assumes the quantum yield for electron transfer from Me-diMeAm-NQ_A⁻ to NQ_B is 75%. The maximum NQ added represents the limit of its solubility.

Figure 7

The estimated energy for charge separated states in RCs with UQ, NQ and Me-diMeAm-NQ. Energies for UQ containing RCs from (Arata and Parson 1981; Gunner et al. 1982; Woodbury et al. 1986). Values for NQ and Me-diMeAm-NQ (abbreviated in the figure as dMA-NQ_A) containing RCs from tables 1 and 3. The lower limit for P⁺NQ_B⁻ of 55 meV above P⁺UQ_A⁻ (115 meV above P⁺UQ_BUQ_B⁻) is obtained from the negligible slowing of the back reaction rate k_AP^obs when NQ is added to UQ_A RCs. The ΔG°_AH of 260 meV with Me-diMeAmQ_A is obtained from the k_AP^obs of 3000 s⁻¹ with this quinone. Charge recombination from Me-diMeAm-NQ_A⁻ proceeds through the thermal back reaction via reformation of P⁺H⁻, so charge recombination from Me-diMeAm-NQ_A:NQ_B⁻ will also use this route (Eqn. 7). Using k_HP of 7×10⁷ s⁻¹ ΔG°_HB is from 480 (k_BP=0) to 520 (k_BP=0.4 s⁻¹) meV (see Fig. 8). ΔG°_AH is 520 meV for UQ_A (Arata and Parson 1981; Gunner et al. 1982; Woodbury et al. 1986), and ΔG°_BH is 580 meV with UQ_B. Thus, P⁺NQ_B⁻ is estimated to be 225–260 meV lower in energy then P⁺Me-diMeAm-NQ_A⁻ and 100–125 meV higher in energy than P⁺UQ_B⁻ (see equations 7 and s2-s4 in supplementary material).

Figure 8

Van’t Hoff plot for temperature dependence of the k_BP^obs in P⁺Me-diMeAm-NQ_A:NQ_B⁻ RCs assuming charge recombination occurs via both the direct route at k_BP and the thermal route at k_BHP. 1 μM RC, 2 μM Me-diMeAm-NQ, 300 μM NQ, 10 mM Tris pH=7.8, LDAO =0.005%. The solid lines represent the fit to Eqn. 8 with ΔG°_BH of 506 meV and k_BP varying from 0 to 0.4 s⁻¹. Temperature from 277K to 303K.