Charge recombination and thermoluminescence in photosystem II

Abstract

In the recombination process of Photosystem II (S(2)Q(A)(-)-->S(1)Q(A)) the limiting step is the electron transfer from the reduced primary acceptor pheophytin Ph(-) to the oxidized primary donor P(+) and the rate depends on the equilibrium constant between states S(2)PPhQ(A)(-) and S(1)P(+)Ph(-)Q(A). Accordingly, mutations that affect the midpoint potential of Ph or of P result in a modified recombination rate. A strong correlation is observed between the effects on the recombination rate and on thermoluminescence (TL, the light emission from S(2)Q(A)(-) during a warming ramp): a slower recombination corresponds to a large enhancement and higher temperature of the TL peak. The current theory of TL does not account for these effects, because it is based on the assumption that the rate-limiting step coincides with the radiative process. When implementing the known fact that the radiative pathway represents a minor leak, the modified TL theory readily accounts qualitatively for the observed behavior. However, the peak temperature is still lower than predicted from the temperature-dependence of recombination. We argue that this reflects the heterogeneity of the recombination process combined with the enhanced sensitivity of TL to slower components. The recombination kinetics are accurately fitted as a sum of two exponentials and we show that this is not due to a progressive stabilization of the charge-separated state, but to a pre-existing conformational heterogeneity.

FIGURE 1

The reaction scheme in the Randall-Wilkins picture. C₁ is the charge-separated state (i.e., D⁺PA⁻, where P is the primary pigment donor, D is a secondary electron donor, and A an electron acceptor). It is in rapid equilibrium with the excited state C* (i.e., DP*A), from which recombination to the ground state C₀ (DPA) occurs with a limiting rate constant κ. ΔG, ΔH, and ΔS are the activation free energy, enthalpy, and entropy, respectively. A fixed fraction Φ of the recombination process is radiative, giving rise to the luminescence emission.

FIGURE 2

Thermoluminescence (Q-band, ascribed to the detrapping of the state) in whole cells of Chlamydomonas reinhardtii WT (solid squares) and mutant E130L (open circles). The algal suspensions were adjusted at the same chlorophyll concentration, corresponding to similar maximum fluorescence (F_m) yields. The electron transfer inhibitor DCMU (10 μM) and uncouplers (nonactin 1 μM and nigericin 1 μM) were added. The warming rate was B = 0.5° s⁻¹.

FIGURE 3

Various recombination routes from the state in PS II. Route 1 is the excitonic (radiative) pathway. Its activation enthalpy is ΔH_ex. Route 2 is the predominant (indirect) recombination route, with activation enthalpy ΔH_r. Route 3 is the direct route, with activation enthalpy ΔH_d. It is normally negligible with respect to route 2, but has an increased relative weight when the equilibrium constant between Q_A and Ph is increased (130L mutants).

FIGURE 4

Simulations according to Eq. 5 (solid lines) or Eq. 6 (dashed lines) of the TL band from the WT and E130L (see Fig. 2). The parameters used were ΔH_r = 625 meV and 680 meV for the WT and E130L, respectively; ΔH_ex = 670 meV; ΔH_d = 230 meV; s_r = 1.1 × 10¹⁰ s⁻¹; s_ex= 1.0 × 10⁹ s⁻¹; and s_d = 63 s⁻¹ (see Cuni et al., 2004); and B = 0.5° s⁻¹.

FIGURE 5

TL simulations according to Eq. 6. ΔH_ex was kept fixed at 670 meV. ΔH_r was taken equal to 625 meV (A), 650 meV (B), and 680 meV (C). The heating rate was B = 0.5° s⁻¹. The pre-exponential factor was s = 1.1 × 10¹⁰ s⁻¹, corresponding to a recombination half-time at 293 K of 2.8 s (A), 9.2 s (B), and 30.2 s (C). The top panel shows the corresponding n(T) functions and the unique k_ex(T) function (dashed line; the scale is arbitrary). The TL (T) curves (bottom panel) are the products n(T) k_ex(T). Notice the increase of the T_m and of the peak amplitude when ΔH_r increases. The inset is a plot of the TL integral (from T = 0 to 330 K) versus the half-time of the recombination reaction at 293 K while varying ΔH_r from 600 to 680 meV: a quasilinear dependence was obtained, in agreement with the data of Vavilin and Vermaas (2000).

FIGURE 6

Effect of the membrane potential on thermoluminescence. (Top panel) TL in whole cells of the Chlamydomonas reinhardtii mutant E130L in the absence (solid symbols) and presence (open symbols) of uncouplers (nonactin 1 μM and nigericin 1 μM). The electron transfer inhibitor DCMU (10 μM) was added. The warming rate was B = 0.5° s⁻¹. (Bottom panel) Simulated TL curves according to Eq. 6, assuming B = 0.5° s⁻¹, s_r = 1.1 × 10¹⁰ s⁻¹, ΔH_ex = 670 meV, ΔH_r = 690 meV (dashed curve), and ΔH_ex = 625 meV, ΔH_r = 665 meV (solid curve). The latter case corresponds to a membrane potential of 50 mV, assuming that the formation of the P⁺Ph⁻ state accounts for 50% of the total electrogenicity.

FIGURE 7

Experimental dependence of k_r upon T and predicted TL behavior. (Top panel) An Arrhenius plot of k_r, measured from the decay of the fluorescence yield after a weak flash in WT Chlamydomonas cells. k_r was computed from the effective half-time of the kinetics, as ln(2)/t_1/2. The linear regression (line) yielded ΔH_r = 625 meV and s = 1.1 × 10¹⁰ s⁻¹. (Bottom panel) The n(T) function computed with this value of ΔH_r from Eq. 4, with B = 0.5° s⁻¹. The inset shows the predicted T_m as a function of ΔH_ex, using Eq. 7.

FIGURE 8

Decay of the fluorescence yield after a weak flash in benzoquinone-treated, DCMU-inhibited Chlamydomonas cells. The initial amplitude (used for normalization) was 4.5%, induced by a saturating flash. The dashed line is a best fit as a single-exponential decay. The solid line is the best fit as a sum of two exponentials, with a fast phase (t_1/2 ≈ 0.81 s) accounting for 32% of the amplitude and a slow phase with t_1/2 ≈ 4.73 s (68%). The dotted line is a best fit as a sum of three exponentials: the two major components are little-modified and the third phase is equivalent to a constant offset by 0.013. The bottom panel is a plot of the residuals for the three types of fits.

FIGURE 9

Effect of multiple flashing. (Top panel) Kinetics of the normalized fluorescence yield in benzoquinone-treated WT Chlamydomonas cells in the presence of 100 μM DCMU, after a series of saturating flashes triggered 3-s apart. Kinetics after flash 1 (solid circles), 2–4 (open circles), and 23 (open diamonds) are shown. (Bottom panel) A simulation according to the competition model (2) described in the text. The curves show (from bottom to top) the kinetics after flashes 1–4, and after flash 8. Eq. 8 was applied, using for f₁(t) the fitted bi-exponential function to the fluorescence kinetics induced by a weak flash. The family of curves thus obtained for the decay of closed centers was converted to fluorescence-yield for a saturating flash, using a hyperbolic fit (see Lavergne and Trissl, 1995) of the experimental plot: (saturating flash kinetics) versus (weak flash kinetics).

FIGURE 10

TL simulation for two PSII recombination components using Eq. 6. (Left panel) The s-factors were kept constant and the two ΔH_r were adjusted to yield half-times of 0.9 s and 5 s at 293 K. Curves A and B (dotted) are the TL bands for the 0.9-s and 5-s components, respectively. Curve C (solid line) is a combination of both with 0.35/0.65 weights. Curve D (dashed) is the TL band for a 293-K half-time of 3.4 s (i.e., the effective half-time of the biexponential decay). Similar results were obtained when assuming that the fast and slow components correspond to different values of s, with a fixed ΔH_r. (Middle and right panels) A comparison between the simulation of the TL bands with Chlamydomonas reinhardtii WT cells (middle panel) and spinach thylakoids (right panel). The accident at ∼−7°C in the thylakoids curve is due to melting of the glycerol-containing medium. Curves C and D were replotted from the left panel for comparison. Both curves were normalized to the amplitude of the experimental TL bands. In the experiments and simulations, B = 0.5° s⁻¹.