Thermodynamics of the charge recombination in photosystem II

The temperature dependence of the electric field-induced chlorophyll luminescence in photosystem II was studied in Tris-washed, osmotically swollen spinach chloroplasts (blebs). The system II reaction centers were brought in the state Z⁺P⁺-Q_A ^-Q_B ^- by preillumination and the charge recombination to the state Z⁺PQ_AQ_B ^- was measured at various temperatures and electrical field strengths. It was found that the activation enthalpy of this back reaction was 0.16 eV in the absence of an electrical field and diminished with increasing field strength. It is argued that this energy is the enthalpy difference between the states IQ_A ^- and I^-Q_A and accounts for about half of the free energy difference between these states. The redox state of Q_B does not influence this free energy difference within 150 s after the photoreduction of Q_A. The consequences for the interpretation of thermodynamic properties of Q_A are discussed.Abbreviations DCMU 3(3,4-dichlorophenyl)-1,1-dimethylurea - I intermediary electron acceptor - Mops 3-(N-morpholino)propanesulphonic acid - P (P₆₈₀) primary electron donor - PS II photosystem II - Q_A and Q_B first and second quinone electron acceptors - Tricine N-tris(hydroxymethyl)methylglycine - Tris tris-(hydroxymethyl)aminomethane - Z secondary electron donor Dedicated to Professor L.N.M. Duysens on the occasion of his retirement     

Role of charge recombination processes in photodamage and photoprotection of the photosystem II complex

Light-induced damage of the photosynthetic apparatus is an important and complex phenomenon, which affects primarily the photosystem II (PSII) complex. Here, the author summarizes the current state of understanding, which concerns the role of charge recombination reactions in photodamage and photoprotection. The main mechanism of photodamage induced by visible light appears to be mediated by acceptor side modifications, which develop under light intensity conditions when the capacity of light-independent photosynthetic processes limits the utilization of electrons produced in the initial photoreactions. This situation facilitates triplet chlorophyll formation and singlet oxygen production in the reaction center of PSII, which initiates the damage of electron transport components and protein structure. This mechanism is an important, but not exclusive, pathway of photodamage, and light-induced inactivation of the Mn cluster of water oxidation may occur in parallel with the singlet oxygen-dependent pathway.     

Charge recombination and thermoluminescence in photosystem II

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.     

Multiple pathways of charge recombination revealed by the temperature dependence of electron transfer kinetics in cyanobacterial photosystem I

The kinetics of charge recombination in Photosystem I P₇₀₀-F_A/F_B complexes and P₇₀₀-F_X cores lacking the terminal iron?sulfur clusters were studied over a temperatures range of 310 K to 4.2 K. Analysis of the charge recombination kinetics in this temperature range allowed the assignment of backward electron transfer from the different electron acceptors to P₇₀₀⁺. The kinetic and thermodynamic parameters of these recombination reactions were determined. The kinetics of all electron transfer reactions were activation-less below 170 K, the glass transition temperature of the water-glycerol solution. Above this temperature, recombination from [F_A/F_B]^? in P₇₀₀-F_A/F_B complexes was found to proceed along two pathways with different activation energies (E_a). The charge recombination via A_1A has an E_a of ~290 meV and is dominant at temperatures above ~280 K, whereas the direct recombination from F_X^? has an E_a of 22 meV and is prevalent in the 200 K to 270 K temperature range. Charge recombination from the F_X cluster becomes highly heterogeneous at temperatures below 200 K. The conformational mobility of Photosystem I was studied by molecular dynamics simulations. The F_X cluster was found to ‘swing’ by ~30° along the axis between the two sulfur atoms proximal to F_A/F_B. The partial rotation of F_X is accompanied by significant changes of electric potential within the iron?sulfur cluster, which may induce preferential electron localization at different atoms of the F_X cluster. These effects may account for the partial arrest of forward electron transfer and for the heterogeneity of charge recombination observed at the glass transition temperature.     

Mutations in the CD-loop region of the D2 protein in Synechocystis sp. PCC 6803 modify charge recombination pathways in photosystem II in vivo

The lumenal CD-loop region of the D2 protein of photosystem II contains residues that interact with the primary electron donor P680 and the redox active tyrosyl residue Y(D). Photosystem II properties were studied in a number of photoautotrophic mutants of Synechocystis sp. PCC 6803, most of which carried combinatorial mutations in residues 164-170, 179-186, or 187-194 of the D2 protein. To facilitate characterization of photosystem II properties in the mutants, the CD-loop mutations were introduced into a photosystem I-less background. According to variable fluorescence decay measurements in DCMU-treated cells, charge recombination of Q(A)(-) with the donor side was faster in the majority of mutants (t(1/2) = 45-140 ms) than in the control (t(1/2) = 180 ms). However, in one mutant (named C7-3), the decay of Q(A)(-) was 2 times slower than in the control (t(1/2) = 360 ms). The decay half-time of each mutant correlated with the yield of the Q-band of thermoluminescence (TL) emitted due to S(2)Q(A)(-) charge recombination. The C7-3 mutant had the highest TL intensity, whereas no Q-band was detected in the mutants with fast Q(A)(-) decay (t(1/2) = 45-50 ms). The correlated changes in the rate of recombination and in TL yield in these strains suggest the existence of a nonradiative pathway of charge recombination between Q(A)(-) and the donor side. This may involve direct electron transfer from Q(A)(-) to P680(+) in a way not leading to formation of excited chlorophyll. Many mutations in the CD-loop appear to increase the equilibrium P680(+) concentration during the lifetime of the S(2)Q(A)(-) state, for example, by making the midpoint potential of the P680(+)/P680 redox couple more negative. The nonradiative charge recombination pathway involves a low activation energy and is less temperature-dependent than the formation of excited P680 that leads to TL emission. Therefore, during the TL measurements in these mutants, the S(2)Q(A)(-) state can recombine nonradiatively before temperatures are reached at which radiative charge recombination becomes feasible. The results presented here highlight the presence of two charge recombination pathways and the importance of the CD-loop of the D2 protein in determination of the energy gap between the P680(+)S(1) and P680S(2) states.     

Oxygen evolution and chlorophyll fluorescence from multiple turnover light pulses: charge recombination in photosystem II in sunflower leaves

Oxygen evolution and Chl fluorescence induction were measured during multiple turnover light pulses (MTP) of 630-nm wavelength, intensities from 250 to 8,000?μmol quanta m(-2)?s(-1) and duration from 0.3 to 200?ms in sunflower leaves at 22?°C. The ambient O(2) concentration was 10-30?ppm and MTP were applied after pre-illumination under far-red light (FRL), which oxidized plastoquinone (PQ) and randomized S-states because of the partial excitation of PSII. Electron (e ( - )) flow was calculated as 4·O(2) evolution. Illumination with MTP of increasing length resulted in increasing O(2) evolution per pulse, which was differentiated against pulse length to find the time course of O(2) evolution rate with sub-millisecond resolution. Comparison of the quantum yields, Y (IIO)?=?e ( - )/hν from O(2) evolution and Y (IIF)?=?(F (m)?-?F)/F (m) from Chl fluorescence, detected significant losses not accompanied by fluorescence emission. These quantum losses are discussed to be caused by charge recombination between Q (A) (-) and oxidized TyrZ at a rate of about 1,000?s(-1), either directly or via the donor side equilibrium complex Q(A)?→?P (D1) (+) ??TyrZ(ox), or because of cycling facilitated by Cyt b (559). Predicted from the suggested mechanism, charge recombination is enhanced by damage to the water-oxidizing complex and by restricted PSII acceptor side oxidation. The rate of PSII charge recombination/cycling is fast enough for being important in photoprotection.     

The rate of charge recombination in Photosystem II

Loss by recombination of the charge separated state P₆₈₀⁺Q_A⁻ limits the performance of Photosystem II (PS II) as a photochemical energy converter. Time constants reported in literature for this process are mostly either near 0.17 ms or near 1.4 ms. The shorter time is found in plant PS II when reduction of P₆₈₀⁺ by the secondary electron donor Tyrosine Z cannot occur because Y_Z is already oxidized. The 1.4 ms recombination is seen in Y_Z-less mutants of the cyanobacterium Synechocystis. However, the rate of P₆₈₀⁺Q_A⁻ recombination that actually competes with the stabilization of the charge separation has not been previously reported. We have measured the kinetics of the flash-induced fluorescence yield changes in the microsecond time domain in Tris-washed spinach chloroplasts. In this way the kinetics and yield of P₆₈₀⁺ reduction by Y_Z were obtained, and the rate of the competing P₆₈₀⁺Q_A⁻ recombination could be evaluated. The recombination time was less than 0.5 ms; the best-fitting time constant was 0.1 ms. The presence of Y_Z^ox slightly decreased the efficiency of excitation trapping but did not seem to accelerate P₆₈₀⁺Q_A⁻ recombination. The two P₆₈₀⁺Q_A⁻ lifetimes in the literature probably reflect a significant difference between plant and cyanobacterial PS II.     

The rate of charge recombination in Photosystem II

Loss by recombination of the charge separated state P(680+)Q(A-) limits the performance of Photosystem II (PS II) as a photochemical energy converter. Time constants reported in literature for this process are mostly either near 0.17 ms or near 1.4 ms. The shorter time is found in plant PS II when reduction of P(680+) by the secondary electron donor Tyrosine Z cannot occur because Y(Z) is already oxidized. The 1.4 ms recombination is seen in Y(Z)-less mutants of the cyanobacterium Synechocystis. However, the rate of P(680+)Q(A-) recombination that actually competes with the stabilization of the charge separation has not been previously reported. We have measured the kinetics of the flash-induced fluorescence yield changes in the microsecond time domain in Tris-washed spinach chloroplasts. In this way the kinetics and yield of P(680+) reduction by Y(Z) were obtained, and the rate of the competing P(680+)Q(A-) recombination could be evaluated. The recombination time was less than 0.5 ms; the best-fitting time constant was 0.1 ms. The presence of Y(Z)(ox) slightly decreased the efficiency of excitation trapping but did not seem to accelerate P(680+)Q(A-) recombination. The two P(680+)Q(A-) lifetimes in the literature probably reflect a significant difference between plant and cyanobacterial PS II.     

Excitation energy transfer and charge separation in photosystem II membranes revisited

We have performed time-resolved fluorescence measurements on photosystem II (PSII) containing membranes (BBY particles) from spinach with open reaction centers. The decay kinetics can be fitted with two main decay components with an average decay time of 150 ps. Comparison with recent kinetic exciton annihilation data on the major light-harvesting complex of PSII (LHCII) suggests that excitation diffusion within the antenna contributes significantly to the overall charge separation time in PSII, which disagrees with previously proposed trap-limited models. To establish to which extent excitation diffusion contributes to the overall charge separation time, we propose a simple coarse-grained method, based on the supramolecular organization of PSII and LHCII in grana membranes, to model the energy migration and charge separation processes in PSII simultaneously in a transparent way. All simulations have in common that the charge separation is fast and nearly irreversible, corresponding to a significant drop in free energy upon primary charge separation, and that in PSII membranes energy migration imposes a larger kinetic barrier for the overall process than primary charge separation.     

Kinetics of the oxygen-evolving complex in salt-washed photosystem II preparations

The kinetics of flash-induced electron transport were investigated in oxygen-evolving Photosystem II preparations, depleted of the 23 and 17 kDa polypeptides by washing with 2 M NaCl. After dark-adaptation and addition of the electron acceptor 2,5-dichloro-p-benzoquinone, in such preparations approx. 75% of the reaction centers still exhibited a period 4 oscillation in the absorbance changes of the oxygen-evolving complex at 350 nm. In comparison to the control preparations, three main effects of NaCl-washing could be observed: the half-time of the oxygen-evolving reaction was slowed down to about 5 ms, the misses and double hits parameters of the period 4 oscillation had changed, and the two-electron gating mechanism of the acceptor side could not be detected anymore. EPR-measurements on the oxidized secondary donor Z⁺ confirmed the slower kinetics of the oxygen-releasing reaction. These phenomena could not be restored by readdition of the released polypeptides nor by the addition of CaCl₂, and are ascribed to deleterious action of the highly concentrated NaCl. Otherwise, the functional coupling of Photosystem II and the oxygen-evolving complex was intact in the majority of the reaction centers. Repetitive flash measurements, however, revealed P⁺Q⁻ recombination and a slow Z⁺ decay in a considerable fraction of the centers. The flash-number dependency of the recombination indicated that this reaction only appeared after prolonged illumination, and disappeared again after the addition of 20 mM CaCl₂. These results are interpreted as a light-induced release of strongly bound Ca²⁺ in the salt-washed preparations, resulting in uncoupling of the oxygen-evolving system and the Photosystem II reaction center, which can be reversed by the addition of a relatively high concentration of Ca²⁺.     

Radiative and non-radiative charge recombination pathways in Photosystem II studied by thermoluminescence and chlorophyll fluorescence in the cyanobacterium Synechocystis 6803

The mechanism of charge recombination was studied in Photosystem II by using flash induced chlorophyll fluorescence and thermoluminescence measurements. The experiments were performed in intact cells of the cyanobacterium Synechocystis 6803 in which the redox properties of the primary pheophytin electron acceptor, Phe, the primary electron donor, P(680), and the first quinone electron acceptor, Q(A), were modified. In the D1Gln130Glu or D1His198Ala mutants, which shift the free energy of the primary radical pair to more positive values, charge recombination from the S(2)Q(A)(-) and S(2)Q(B)(-) states was accelerated relative to the wild type as shown by the faster decay of chlorophyll fluorescence yield, and the downshifted peak temperature of the thermoluminescence Q and B bands. The opposite effect, i.e. strong stabilization of charge recombination from both the S(2)Q(A)(-) and S(2)Q(B)(-) states was observed in the D1Gln130Leu or D1His198Lys mutants, which shift the free energy level of the primary radical pair to more negative values, as shown by the retarded decay of flash induced chlorophyll fluorescence and upshifted thermoluminescence peak temperatures. Importantly, these mutations caused a drastic change in the intensity of thermoluminescence, manifested by 8- and 22-fold increase in the D1Gln130Leu and D1His198Lys mutants, respectively, as well as by a 4- and 2.5-fold decrease in the D1Gln130Glu and D1His198Ala mutants, relative to the wild type, respectively. In the presence of the electron transport inhibitor bromoxynil, which decreases the redox potential of Q(A)/Q(A)(-) relative to that observed in the presence of DCMU, charge recombination from the S(2)Q(A)(-) state was accelerated in the wild type and all mutant strains. Our data confirm that in PSII the dominant pathway of charge recombination goes through the P(680)(+)Phe(-) radical pair. This indirect recombination is branched into radiative and non-radiative pathways, which proceed via repopulation of P(680)(*) from (1)[P(680)(+)Ph(-)] and direct recombination of the (3)[P(680)(+)Ph(-)] and (1)[P(680)(+)Ph(-)] radical states, respectively. An additional non-radiative pathway involves direct recombination of P(680)(+)Q(A)(-). The yield of these charge recombination pathways is affected by the free energy gaps between the Photosystem II electron transfer components in a complex way: Increase of DeltaG(P(680)(*)<-->P(680)(+)Phe(-)) decreases the yield of the indirect radiative pathway (in the 22-0.2% range). On the other hand, increase of DeltaG(P(680)(+)Phe(-)<-->P(680)(+)Q(A)(-)) increases the yield of the direct pathway (in the 2-50% range) and decreases the yield of the indirect non-radiative pathway (in the 97-37% range).     

Surface charge asymmetry and a specific calcium ion effect in chloroplast photosystem II

We have used the decay kinetics of Signal IIf in Tris-washed chloroplasts as a direct probe to reactions on the oxidizing side of Photosystem II. A study of the salt concentration dependence of the rate of reduction of Z . + by the ascorbate monoanion has been interpreted by using the Gouy-Chapman diffuse double layer model and allows the calculation of an inner membrane surface charge density of -3.4 +/- 0.3 microC . cm-2 at pH = 8.0 in the vicinity of Photosystem II. We have also measured the outer membrane surface charge density at this pH in Tris- and sucrose-washed chloroplasts by monitoring the rate of potassium ferricyanide oxidation of Q-, and arrive at values of -2.2 +/- 0.3 microC . cm-2 and -2.1 microC . cm-2, respectively. From these experiments we conclude that in dark-adapted chloroplasts at pH 8.0 there exists a transmembrane electric field in the vicinity of Photosystem II which arises from this surface charge asymmetry. In the presence of 10 mM monovalent salts, the transmembrane potential difference is of the order of 20 mV, corresponding to a field of 4 . 10(4) V . cm-1 (negative inside) for a 50A membrane. It is both smaller in magnitude and in the opposite direction compared to the photoinduced transmembrane field which gives rise to the 515 nm absorption change. We have also found non-double layer Ca2+ effects on the decay kinetics of Signal IIf with both charged (ascorbate monoanion) and neutral (diphenylcarbazide) donors. These results suggest a change in the environment of Z from lipophilic to hydrophilic upon specific binding of Ca2+.     

Radiative and non-radiative charge recombination pathways in Photosystem II studied by thermoluminescence and chlorophyll fluorescence in the cyanobacterium Synechocystis 6803

The mechanism of charge recombination was studied in Photosystem II by using flash induced chlorophyll fluorescence and thermoluminescence measurements. The experiments were performed in intact cells of the cyanobacterium Synechocystis 6803 in which the redox properties of the primary pheophytin electron acceptor, Phe, the primary electron donor, P₆₈₀, and the first quinone electron acceptor, Q_A, were modified. In the D1Gln130Glu or D1His198Ala mutants, which shift the free energy of the primary radical pair to more positive values, charge recombination from the S₂Q_A⁻ and S₂Q_B⁻ states was accelerated relative to the wild type as shown by the faster decay of chlorophyll fluorescence yield, and the downshifted peak temperature of the thermoluminescence Q and B bands. The opposite effect, i.e. strong stabilization of charge recombination from both the S₂Q_A⁻ and S₂Q_B⁻ states was observed in the D1Gln130Leu or D1His198Lys mutants, which shift the free energy level of the primary radical pair to more negative values, as shown by the retarded decay of flash induced chlorophyll fluorescence and upshifted thermoluminescence peak temperatures. Importantly, these mutations caused a drastic change in the intensity of thermoluminescence, manifested by 8- and 22-fold increase in the D1Gln130Leu and D1His198Lys mutants, respectively, as well as by a 4- and 2.5-fold decrease in the D1Gln130Glu and D1His198Ala mutants, relative to the wild type, respectively. In the presence of the electron transport inhibitor bromoxynil, which decreases the redox potential of Q_A/Q_A⁻ relative to that observed in the presence of DCMU, charge recombination from the S₂Q_A⁻ state was accelerated in the wild type and all mutant strains. Our data confirm that in PSII the dominant pathway of charge recombination goes through the P₆₈₀⁺Phe⁻ radical pair. This indirect recombination is branched into radiative and non-radiative pathways, which proceed via repopulation of P₆₈₀^* from ¹[P₆₈₀⁺Ph⁻] and direct recombination of the ³[P₆₈₀⁺Ph⁻] and ¹[P₆₈₀⁺Ph⁻] radical states, respectively. An additional non-radiative pathway involves direct recombination of P₆₈₀⁺Q_A⁻. The yield of these charge recombination pathways is affected by the free energy gaps between the Photosystem II electron transfer components in a complex way: Increase of ΔG(P₆₈₀^* ↔ P₆₈₀⁺Phe⁻) decreases the yield of the indirect radiative pathway (in the 22-0.2% range). On the other hand, increase of ΔG(P₆₈₀⁺Phe⁻ ↔ P₆₈₀⁺Q_A⁻) increases the yield of the direct pathway (in the 2-50% range) and decreases the yield of the indirect non-radiative pathway (in the 97-37% range).     

Multiple redox-active chlorophylls in the secondary electron-transfer pathways of oxygen-evolving photosystem II

Photosystem II (PS II) is unique among photosynthetic reaction centers in having secondary electron donors that compete with the primary electron donors for reduction of P680(+). We have characterized the photooxidation and dark decay of the redox-active accessory chlorophylls (Chl) and beta-carotenes (Car) in oxygen-evolving PS II core complexes by near-IR absorbance and EPR spectroscopies at cryogenic temperatures. In contrast to previous results for Mn-depleted PS II, multiple near-IR absorption bands are resolved in the light-minus-dark difference spectra of oxygen-evolving PS II core complexes including two fast-decaying bands at 793 and 814 nm and three slow-decaying bands at 810, 825, and 840 nm. We assign these bands to chlorophyll cation radicals (Chl(+)). The fast-decaying bands observed after illumination at 20 K could be generated again by reilluminating the sample. Quantization by EPR gives a yield of 0.85 radicals per PS II, and the yield of oxidized cytochrome b 559 by optical difference spectroscopy is 0.15 per PS II. Potential locations of Chl(+) and Car(+) species, and the pathways of secondary electron transfer based on the rates of their formation and decay, are discussed. This is the first evidence that Chls in the light-harvesting proteins CP43 and CP47 are oxidized by P680(+) and may have a role in Chl fluorescence quenching. We also suggest that a possible role for negatively charged lipids (phosphatidyldiacylglycerol and sulfoquinovosyldiacylglycerol identified in the PS II structure) could be to decrease the redox potential of specific Chl and Car cofactors. These results provide new insight into the alternate electron-donation pathways to P680(+).     

Low-temperature electron transfer in photosystem II: a tyrosyl radical and semiquinone charge pair

The states induced by illumination at 7 K in the oxygen-evolving enzyme (PSII) from Thermosynechococcus elongatus were studied by EPR. In the S(0) and S(1) redox states, two g approximately 2 EPR signals, a split signal and a g = 2.03 signal, respectively, were generated by illumination with visible light. These signals were comparable to those already reported in plant PSII in terms of their g value, shape, and stability at low temperatures. We report that the formation and decay of these signals correlate with EPR signals from the semiquinone of the first quinone electron acceptor, Q(A)(-). The light-induced EPR signals from oxidized side-path electron donors (Cyt b(559), Car, and Chl(Z)) were also measured, and from these and the signals from Q(A)(-), estimates were made of the proportion of centers involved in the formation of the g approximately 2 signals (approximately 50% in S(0) and 40% in S(1)). Comparisons with the signals generated in plant PSII indicated approximately similar yields for the S(0) split signal. A single laser flash at 7 K induced more than 75% of the maximum split and g = 2.03 EPR signal observed by continuous illumination, with no detectable oxidation of side-path donors. The matching electron acceptor side reactions, the high quantum yield, and the relatively large proportion of centers involved support earlier suggestions that the state being monitored is Tyr(Z)(*)Q(A)(-), with the g approximately 2 EPR signals arising from Tyr(Z)(*) interacting magnetically with the Mn complex. The current picture of the photochemical reactions occurring in PSII at low temperatures is reassessed.     

Secondary pair charge recombination in photosystem I under strongly reducing conditions: temperature dependence and suggested mechanism.

Photoinduced electron transfer in photosystem I (PS I) proceeds from the excited primary electron donor P700 (a chlorophyll a dimer) via the primary acceptor A0 (chlorophyll a) and the secondary acceptor A1 (phylloquinone) to three [4Fe-4S] clusters, Fx, FA, and FB. Prereduction of the iron-sulfur clusters blocks electron transfer beyond A1. It has been shown previously that, under such conditions, the secondary pair P700+A1- decays by charge recombination with t1/2 approximately 250 ns at room temperature, forming the P700 triplet state (3P700) with a yield exceeding 85%. This reaction is unusual, as the secondary pair in other photosynthetic reaction centers recombines much slower and forms directly the singlet ground state rather than the triplet state of the primary donor. Here we studied the temperature dependence of secondary pair recombination in PS I from the cyanobacterium Synechococcus sp. PCC6803, which had been illuminated in the presence of dithionite at pH 10 to reduce all three iron-sulfur clusters. The reaction P700+A1- --> 3P700 was monitored by flash absorption spectroscopy. With decreasing temperature, the recombination slowed down and the yield of 3P700 decreased. In the range between 303 K and 240 K, the recombination rates could be described by the Arrhenius law with an activation energy of approximately 170 meV. Below 240 K, the temperature dependence became much weaker, and recombination to the singlet ground state became the dominating process. To explain the fast activated recombination to the P700 triplet state, we suggest a mechanism involving efficient singlet to triplet spin evolution in the secondary pair, thermally activated repopulation of the more closely spaced primary pair P700+A0- in a triplet spin configuration, and subsequent fast recombination (intrinsic rate on the order of 10(9) s(-1)) forming 3P700.