Alternative electron flows (water-water cycle and cyclic electron flow around PSI) in photosynthesis: molecular mechanisms and physiological functions

An electron flow in addition to the major electron sinks in C(3) plants [both photosynthetic carbon reduction (PCR) and photorespiratory carbon oxidation (PCO) cycles] is termed an alternative electron flow (AEF) and functions in the chloroplasts of leaves. The water-water cycle (WWC; Mehler-ascorbate peroxidase pathway) and cyclic electron flow around PSI (CEF-PSI) have been studied as the main AEFs in chloroplasts and are proposed to play a physiologically important role in both the regulation of photosynthesis and the alleviation of photoinhibition. In the present review, I discuss the molecular mechanisms of both AEFs and their functions in vivo. To determine their physiological function, accurate measurement of the electron flux of AEFs in vivo are required. Methods to assay electron flux in CEF-PSI have been developed recently and their problematic points are discussed. The common physiological function of both the WWC and CEF-PSI is the supply of ATP to drive net CO(2) assimilation. The requirement for ATP depends on the activities of both PCR and PCO cycles, and changes in both WWC and CEF-PSI were compared with the data obtained in intact leaves. Furthermore, the fact that CEF-PSI cannot function independently has been demonstrated. I propose a model for the regulation of CEF-PSI by WWC, in which WWC is indispensable as an electron sink for the expression of CEF-PSI activity.     

Characteristics of photosynthesis and functions of the water-water cycle in rice (Oryza sativa) leaves in response to potassium deficiency

The mechanisms of photoprotection of photosynthesis and dissipation of excitation energy in rice leaves in response to potassium (K) deficiency were investigated. Net photosynthetic rate and the activity of ribulose-1,5-bisphosphate carboxylase/oxygenase decreased under K deficiency. Compared with the control, non-photochemical quenching of Chl fluorescence increased in K-deficient plant, whereas the efficiency of excitation transfer (F'(v)/F'(m)) and the photochemical quenching coefficient (q(P)) decreased. Thus, thermal dissipation of excitation energy increased as more excess electrons were accumulated in the photosynthetic chain. The electron transport rate through PSII (J(f)) was more sensitive to O2 concentration, and the fraction of electron transport rate required to sustain CO2 assimilation and photorespiration (J(g)/J(f)) was significantly decreased under K deficiency compared with the control. Furthermore, the alternative electron transport (J(a)/J(f)) was increased, indicating that a considerable amount of electrons had been transported to O2 during the water-water cycle in the K-deficient leaves. Although the fraction of electron transport to photorespiration (J(o)/J(f)) was also increased in the K-deficient leaves, it was less sensitive than that of the water-water cycle. With the generation of reactive oxygen species level, the activities of superoxide dismutase and ascorbate peroxidase, two of the key enzymes involved in scavenging of active oxygen species in the water-water cycle, also increased in K-deficient rice. Therefore, it is likely that a series of photoprotective mechanisms were initiated in rice plants in response to K deficiency and the water-water cycle might be critical for protecting photosynthetic apparatus under K deficiency in rice.     

Chlororespiration and cyclic electron flow around PSI during photosynthesis and plant stress response

Besides major photosynthetic complexes of oxygenic photosynthesis, new electron carriers have been identified in thylakoid membranes of higher plant chloroplasts. These minor components, located in the stroma lamellae, include a plastidial NAD(P)H dehydrogenase (NDH) complex and a plastid terminal plastoquinone oxidase (PTOX). The NDH complex, by reducing plastoquinones (PQs), participates in one of the two electron transfer pathways operating around photosystem I (PSI), the other likely involving a still uncharacterized ferredoxin-plastoquinone reductase (FQR) and the newly discovered PGR5. The existence of a complex network of mechanisms regulating expression and activity of the NDH complex, and the presence of higher amounts of NDH complex and PTOX in response to environmental stress conditions the phenotype of mutants, indicate that these components likely play a role in the acclimation of photosynthesis to changing environmental conditions. Based on recently published data, we propose that the NDH-dependent cyclic pathway around PSI participates to the ATP supply in conditions of high ATP demand (such as high temperature or water limitation) and together with PTOX regulates cyclic electron transfer activity by tuning the redox state of intersystem electron carriers. In response to severe stress conditions, PTOX associated to the NDH and/or the PGR5 pathway may also limit electron pressure on PSI acceptor and prevent PSI photoinhibition.     

CO2 response of cyclic electron flow around PSI (CEF-PSI) in tobacco leaves--relative electron fluxes through PSI and PSII determine the magnitude of non-photochemical quenching (NPQ) of Chl fluorescence

We hypothesized that cyclic electron flow around photosystem I (CEF-PSI) participates in the induction of non-photochemical quenching (NPQ) of chlorophyll (Chl) fluorescence when the rate of photosynthetic linear electron flow (LEF) is electron-acceptor limited. To test this hypothesis, the relationships among photosynthesis rate, electron fluxes through both PSI and PSII [Je(PSI) and Je(PSII)] and Chl fluorescence parameters were analyzed simultaneously in intact leaves of tobacco plants at several light intensities and partial pressures of ambient CO2 (Ca). At low light intensities, decreasing Ca lowered the photosynthesis rate, but Je(PSI) and Je(PSII) remained constant. Je(PSI) was larger than Je(PSII), indicating the existence of CEF-PSI. Increasing the light intensity enhanced photosynthesis and both Je(PSI) and Je (PSII). Je(PSI)/Je(PSII) also increased at high light and at high light and low Ca combined, showing a strong, positive relationship with NPQ of Chl fluorescence. These results indicated that CEF-PSI contributed to the dissipation of photon energy in excess of that consumed by photosynthesis by driving NPQ of Chl fluorescence. The main physiological function of CEF-PSI in photosynthesis of higher plants is discussed.     

Effects of light intensity on cyclic electron flow around PSI and its relationship to non-photochemical quenching of Chl fluorescence in tobacco leaves

We tested the hypothesis that plants grown under high light intensity (HL-plants) had a large activity of cyclic electron flow around PSI (CEF-PSI) compared with plants grown under low light (LL-plants). To evaluate the activity of CEF-PSI, the relationships between photosynthesis rate, quantum yields of both PSII and PSI, and Chl fluorescence parameters were analyzed simultaneously in intact leaves of tobacco plants which had been grown under different light intensities (150 and 1,100 micromol photons m(-2) s(-1), respectively) and with different amounts of nutrients supplied. HL-plants showed a larger value of non-photochemical quenching (NPQ) of Chl fluorescence at the limited activity of photosynthetic linear electron flow. Furthermore, HL-plants had a larger activity of CEF-PSI than LL-plants. These results suggested that HL-plants dissipated the excess photon energy through NPQ by enhancing the ability of CEF-PSI to induce acidification of the thylakoid lumen.     

Changes in the thermal dissipation and the electron flow in the water-water cycle in rice grown under conditions of physiologically low temperature

Effects of low temperature on chlorophyll (Chl) fluorescence, gas exchange rate, the amounts of xanthophyll cycle pigments (Xp) and the activities of several antioxidant enzymes were examined in the 8th leaf of two rice (Oryza sativa L.) cultivars (japonica and indica types) and rbcS antisense rice. All plants were grown hydroponically at 25/20 degrees C (day/night), and then exposed to 20/17 degrees C (day/night) after full expansion of the 8th leaf, or exposed to either 20/17 degrees C or 15/13 degrees C (day/night) during the expansion of the 8th leaf. All plants exposed to low temperatures showed a decrease in CO(2) assimilation rate without photoinhibition, and increases in the fraction of thermal dissipation in PSII, and in the electron flux through the water-water cycle (WWC) were observed. Although the increase of thermal dissipation was associated with increases in the ratio of carotenoids to Chl, the ratio of Xp to carotenoids and the de-epoxidation state of Xp, the increase of the electron flux of WWC was not accompanied by an increase in the activities of antioxidant enzymes. Such photoprotective responses did not differ between during and after full expansion of the leaf, and did not differ among the three genotypes. Quantitative analyses on the dissipation of excess light energy showed that thermal dissipation makes a larger contribution than WWC. Thus, although low temperature led to a decrease in CO(2) assimilation, rice potentially coped with the excess light energy by increasing the thermal dissipation and the electron flux of WWC under low temperature irrespective of leaf development and genotypes.     

Quantification of cyclic electron flow around Photosystem I in spinach leaves during photosynthetic induction

The variation of the rate of cyclic electron transport around Photosystem I (PS I) during photosynthetic induction was investigated by illuminating dark-adapted spinach leaf discs with red + far-red actinic light for a varied duration, followed by abruptly turning off the light. The post-illumination re-reduction kinetics of P700⁺, the oxidized form of the photoactive chlorophyll of the reaction centre of PS I (normalized to the total P700 content), was well described by the sum of three negative exponential terms. The analysis gave a light-induced total electron flux from which the linear electron flux through PS II and PS I could be subtracted, yielding a cyclic electron flux. Our results show that the cyclic electron flux was small in the very early phase of photosynthetic induction, rose to a maximum at about 30 s of illumination, and declined subsequently to <10% of the total electron flux in the steady state. Further, this cyclic electron flow, largely responsible for the fast and intermediate exponential decays, was sensitive to 3-(3,4-dichlorophenyl)-1,1-dimethyl urea, suggesting an important role of redox poising of the cyclic components for optimal function. Significantly, our results demonstrate that analysis of the post-illumination re-reduction kinetics of P700⁺ allows the quantification of the cyclic electron flux in intact leaves by a relatively straightforward method.     

Ferredoxin limits cyclic electron flow around PSI (CEF-PSI) in higher plants--stimulation of CEF-PSI enhances non-photochemical quenching of Chl fluorescence in transplastomic tobacco

We tested the hypothesis that ferredoxin (Fd) limits the activity of cyclic electron flow around PSI (CEF-PSI) in vivo and that the relief of this limitation promotes the non-photochemical quenching (NPQ) of Chl fluorescence. In transplastomic tobacco (Nicotiana tabacum cv Xanthi) expressing Fd from Arabidopsis (Arabidopsis thaliana) in its chloroplasts, the minimum yield (F(o)) of Chl fluorescence was higher than in the wild type. F(o) was suppressed to the wild-type level upon illumination with far-red light, implying that the transfer of electrons by Fd-quinone oxidoreductase (FQR) from the chloroplast stroma to plastoquinone was enhanced in transplastomic plants. The activity of CEF-PSI became higher in transplastomic than in wild-type plants under conditions limiting photosynthetic linear electron flow. Similarly, the NPQ of Chl fluorescence was enhanced in transplastomic plants. On the other hand, pool sizes of the pigments of the xanthophyll cycle and the amounts of PsbS protein were the same in all plants. All these results supported the hypothesis strongly. We conclude that breeding plants with an NPQ of Chl fluorescence increased by an enhancement of CEF-PSI activity might lead to improved tolerance for abiotic stresses, particularly under conditions of low light use efficiency.     

Enhancement of cyclic electron flow around PSI at high light and its contribution to the induction of non-photochemical quenching of chl fluorescence in intact leaves of tobacco plants

Non-photochemical quenching (NPQ) of Chl fluorescence is a mechanism for dissipating excess photon energy and is dependent on the formation of a DeltapH across the thylakoid membranes. The role of cyclic electron flow around photosystem I (PSI) (CEF-PSI) in the formation of this DeltapH was elucidated by studying the relationships between O2-evolution rate [V(O2)], quantum yield of both PSII and PSI [Phi(PSII) and Phi(PSI)], and Chl fluorescence parameters measured simultaneously in intact leaves of tobacco plants in CO2-saturated air. Although increases in light intensity raised V(O2) and the relative electron fluxes through both PSII and PSI [Phi(PSII) x PFD and Phi(PSI) x PFD] only Phi(PSI) x PFD continued to increase after V(O2) and Phi(PSII) x PFD became light saturated. These results revealed the activity of an electron transport reaction in PSI not related to photosynthetic linear electron flow (LEF), namely CEF-PSI. NPQ of Chl fluorescence drastically increased after Phi(PSII) x PFD became light saturated and the values of NPQ correlated positively with the relative activity of CEF-PSI. At low temperatures, the light-saturation point of Phi(PSII) x PFD was lower than that of Phi(PSI) x PFD and NPQ was high. On the other hand, at high temperatures, the light-dependence curves of Phi(PSII) x PFD and Phi(PSI) x PFD corresponded completely and NPQ was not induced. These results indicate that limitation of LEF induced CEF-PSI, which, in turn, helped to dissipate excess photon energy by driving NPQ of Chl fluorescence.     

Characterization of photosynthetic electron transport in bundle sheath cells of maize. I. Ascorbate effectively stimulates cyclic electron flow around PSI

Redox changes of the reaction-center chlorophyll of photosystem I (P700) and chlorophyll fluorescence yield were measured in bundle sheath strands (BSS) isolated from maize (Zea mays L.) leaves. Oxidation of P700 in BSS by actinic light was suppressed by nigericin, indicating the generation of a proton gradient across the thylakoid membranes of BSS chloroplasts. Methyl viologen, which transfers electrons from photosystem I (PSI) to O₂, caused a considerable decrease in the reduction rate of P700⁺ in BSS after turning off actinic light, showing that electron flow from the acceptor side of PSI to stromal components is critical for this reduction. Ascorbate (Asc), and to a lesser extent malate (Mal), caused a lower level of P700⁺ in BSS under aerobic conditions in far-red light, implying electron donation from these substances to the intersystem carriers. When Asc or Mal was added to BSS during pre-illumination under anaerobic conditions in the presence of 3-(3,4-dichlorophenyl)-1,1-dimethyl urea (DCMU), the far-red-induced level of P700⁺ was lowered. The results suggest Asc and Mal can cause reduction of stromal donors, which in turn establishes conditions for rapid PSI-driven P700⁺ reduction. Addition of these metabolites also strongly stimulated the development of a proton gradient in thylakoids under aerobic conditions in the absence of DCMU, i.e. under conditions analogous to those in vivo. Ascorbate was a much more effective electron donor than Mal, suggesting it has a physiological role in activation of cyclic electron flow around PSI.     

Photosystem II cycle and alternative electron flow in leaves

Sunflower (Helianthus annuus L.) and tobacco (Nicotiana tabacum L.) were grown in the laboratory and leaves were taken from field-grown birch trees (Betula pendula Roth). Chlorophyll fluorescence, CO2 uptake and O2 evolution were measured and electron transport rates were calculated, J(C) from the CO2 uptake rate considering ribulose-1,5-bisphosphate (RuBP) carboxylation and oxygenation, J(O) from the O2 evolution rate, and J(F) from Chl fluorescence parameters. Mesophyll diffusion resistance, r(md), used for the calculation of J(C), was determined such that the in vivo Rubisco kinetic curve with respect to the carboxylation site CO2 concentration became a rectangular hyperbola with Km(CO2) of 10 microM at 22.5 degrees C. In sunflower, in the absence of external O2, J(O) = 1.07 J(C) when absorbed photon flux density (PAD) was varied, showing that the O2-independent components of the alternative electron flow to acceptors other than CO2 made up 7% of J(C). Under saturating light, J(F), however, was 20-30% faster than J(C), and J(F)-J(C) depended little on CO2 and O2 concentrations. The inter-relationship between J(F)-J(C) and non-photochemical quenching (NPQ) was variable, dependent on the CO2 concentration. We conclude that the relatively fast electron flow J(F)-J(C) appearing at light saturation of photosynthesis contains a minor component coupled with proton translocation, serving for nitrite, oxaloacetate and oxygen reduction, and a major component that is mostly cyclic electron transport around PSII. The rate of the PSII cycle is sufficient to release the excess excitation pressure on PSII significantly. Although the O2-dependent Mehler-type alternative electron flow appeared to be under the detection threshold, its importance is discussed considering the documented enhancement of photosynthesis by oxygen.     

Rates and roles of cyclic and alternative electron flow in potato leaves

Measurements of 810 nm transmittance changes in leaves, simultaneously with Chl fluorescence, CO(2) uptake and O(2) evolution, were carried out on potato (Solanum tuberosum L.) leaves with altered expression of plastidic NADP-dependent malate dehydrogenase. Electron transport rates were calculated: J(C) from the CO(2) uptake rate considering ribulose-1,5-bisphosphate (RuBP) carboxylation and oxygenation, J(O) from the O(2) evolution rate, J(F) from Chl fluorescence parameters and J(I) from the post-illumination re-reduction speed of PSI donors. In the absence of external O(2), J(O) equaled (1.005 +/- 0.003) J(C), independent of the transgenic treatment, light intensity and CO(2) concentration. This showed that nitrite and oxaloacetate reduction rates were very slow. The Mehler-type O(2) reduction was evaluated from the rate of electron accumulation at PSI after the O(2) concentration was decreased from 210 to 20 mmol mol(-1), and resulted in <1% of the linear flow. J(F) and J(I) did not differ from J(C) while photosynthesis was light-limited, but considerably exceeded J(C) at saturating light. Then, typically, J(F) = 1.2 J(C) and J(I) = 1.3 J(C), and J(F) -J(C) and J(I) -J(C) depended little on CO(2) and O(2) concentrations. The results showed that the alternative and cyclic electron flow necessary to compensate variations in the ATP/NADPH ratio were only a few percent of the linear flow. The data do not support the requirement of 14H(+)/3ATP by the chloroplast ATP synthase. We suggest that the fast PSI cyclic electron flow J(I) - J(C), as well as the fast J(F) - J(C) are energy-dissipating cycles around PSI and PSII at light saturation.     

Physiology of PSI cyclic electron transport in higher plants

Having long been debated, it is only in the last few years that a concensus has emerged that the cyclic flow of electrons around Photosystem I plays an important and general role in the photosynthesis of higher plants. Two major pathways of cyclic flow have been identified, involving either a complex termed NDH or mediated via a pathway involving a protein PGR5 and two functions have been described-to generate ATP and to provide a pH gradient inducing non-photochemical quenching. The best evidence for the occurrence of the two pathways comes from measurements under stress conditions-high light, drought and extreme temperatures. In this review, the possible relative functions and importance of the two pathways is discussed as well as evidence as to how the flow through these pathways is regulated. Our growing knowledge of the proteins involved in cyclic electron flow will, in the future, enable us to understand better the occurrence and diversity of cyclic electron transport pathways. This article is part of a Special Issue entitled: Regulation of Electron Transport in Chloroplasts.     

The water-water cycle in leaves is not a major alternative electron sink for dissipation of excess excitation energy when CO(2) assimilation is restricted

Electron flux from water via photosystem II (PSII) and PSI to oxygen (water-water cycle) may provide a mechanism for dissipation of excess excitation energy in leaves when CO(2) assimilation is restricted. Mass spectrometry was used to measure O(2) uptake and evolution together with CO(2) uptake in leaves of French bean and maize at CO(2) concentrations saturating for photosynthesis and the CO(2) compensation point. In French bean at high CO(2) and low O(2) concentrations no significant water-water cycle activity was observed. At the CO(2) compensation point and 3% O(2) a low rate of water-water cycle activity was observed, which accounted for 30% of the linear electron flux from water. In maize leaves negligible water-water cycle activity was detected at the compensation point. During induction of photosynthesis in maize linear electron flux was considerably greater than CO(2) assimilation, but no significant water-water cycle activity was detected. Miscanthus × giganteus grown at chilling temperature also exhibited rates of linear electron transport considerably in excess of CO(2) assimilation; however, no significant water-water cycle activity was detected. Clearly the water-water cycle can operate in leaves under some conditions, but it does not act as a major sink for excess excitation energy when CO(2) assimilation is restricted.     

The water-water cycle is more effective in regulating redox state of photosystem I under fluctuating light than cyclic electron transport

Photosynthetic electron flux from water via photosystem II (PSII) and PSI to oxygen (water-water cycle) may act as an alternative electron sink under fluctuating light in angiosperms. We measured the P700 redox kinetics and electrochromic shift signal under fluctuating light in 11 Camellia species and tobacco leaves. Upon dark-to-light transition, these Camellia species showed rapid re-oxidation of P700. However, this rapid re-oxidation of P700 was not observed when measured under anaerobic conditions, as was in experiment with tobacco performed under aerobic conditions. Therefore, photo-reduction of O₂ mediated by water-water cycle was functional in these Camellia species but not in tobacco. Within the first 10 s after transition from low to high light, PSI was highly oxidized in these Camellia species but was over-reduced in tobacco leaves. Furthermore, such rapid oxidation of PSI in these Camellia species was independent of the formation of trans-thylakoid proton gradient (ΔpH). These results indicated that in addition to ΔpH-dependent photosynthetic control, the water-water cycle can protect PSI against photoinhibition under fluctuating light in these Camellia species. We here propose that the water-water cycle is an overlooked strategy for photosynthetic regulation under fluctuating light in angiosperms.     

Oxygen and electron flow in C4 photosynthesis: Mehler reaction, photorespiration and CO2 concentration in the bundle sheath

The photosynthetic linear electron transport rate in excess of that used for CO₂ reduction was evaluated in Sorghum bicolor Moench. [NADP-malic enzyme (ME)-type C₄ plant], Amaranthus cruentus L. (NAD-ME-type C₄ plant) and Helianthus annuus L. (C₃ plant) leaves at different CO₂ and O₂ concentrations. The electron transport rate (J _F) was calculated from fluorescence using the light partitioning factor (relative PSII cross-section) determined under conditions where excess electron transport was assumed to be negligible: low light intensities, 500 μmol CO₂ · mol⁻¹ and 2% O₂. Under high light intensities there was a large excess of J _F/4 at 10–100% O₂ in the C₃ plant due to photorespiration, but very little in sorghum and somewhat more in amaranth, showing that photorespiration is suppressed, more in the NADP-ME- and less in the NAD-ME-type species. It is concluded that when C₄ photosynthesis is limited by supply of atmospheric CO₂ to the C₄ cycle, the C₃ cycle becomes limited by regeneration of ribulose 1,5-bisphosphate (RuBP) which in turn limits RuBP oxygenase activity and photorespiration. The rate of excess electron transport over that consumed for CO₂ fixation in C₄ plants was very sensitive to the presence of O₂ in the gas phase, rapidly increasing between 0.01 and 0.1% O₂, and at 2% O₂ it was about two-thirds of that at 21% O₂. This shows the importance of the Mehler O₂ reduction as an electron sink, compared with photorespiration in C₄ plants. However, the rate of the Mehler reaction is still too low to fully account for the extra ATP which is needed in C₄ photosynthesis. Received: 8 November 1997 / Accepted: 26 December 1997     

Ferredoxin-quinone reductase benefits cyclic electron flow around photosystem 1 in tobacco leaves upon exposure to chilling stress under low irradiance

The function of chloroplast ferredoxin quinone reductase (FQR)-dependent flow was examined by comparing a wild type tobacco and a tobacco transformant (ΔndhB) in which the ndhB gene had been disrupted with their antimycin A (AA)-fed leaves upon exposure to chilling temperature (4 °C) under low irradiance (100 μmol m⁻² s⁻¹ photon flux density). During the chilling stress, the maximum photochemical efficiency of photosystem (PS) 2 (F_v/F_m) decreased markedly in both the controls and AA-fed leaves, and P700⁺ was also lower in AA-fed leaves than in the controls, implying that FQR-dependent cyclic electron flow around PS1 functioned to protect the photosynthetic apparatus from chilling stress under low irradiance. Under such stress, non-photochemical quenching (NPQ), particularly the fast relaxing NPQ component (q_f) and the de-epoxidized ratio of the xanthophyll cycle pigments, (A+Z)/(V+A+Z), formed the difference between AA-fed leaves and controls. The lower NPQ in AA-fed leaves might be related to an inefficient proton gradient across thylakoid membranes (ΔpH) because of inhibiting an FQR-dependent cyclic electron flow around PS1 at chilling temperature under low irradiance.     

Is there significant cyclic electron flow around photoreaction 1 in cyanobacteria?

Evidence for a cyclic electron flow has been sought by study of the steady-state poise of P700 and rate of photoreaction 1 in three cyanobacteria. Under an actinic light 1 (440 or 680 nm) the rate of photoreaction 1 is limited by the rate of electron supply provided by photoreaction 2 and by all return electron flow from low potential donors such as ferredoxin and NAD(P)H. Plots of p, the steady-state fraction of P700 reduced, versus the reciprocal intensity, 1/I, yield linear segments of slope Ip. From considerations of a simple model the slopes and extrapolated intercepts of the linear segments provide estimates of the rate of return electron flow. Analysis shows that the total return electron flow cannot be large, by one estimate not more than three times the rate of dark respiration. This result leads to a conclusion that cyclic electron flow (and any dependent phosphorylation) is not a significant process in these cyanobacteria at ordinary light intensities.Abbreviations DAD diaminodurene - PMS phenazine methosulfate     

Redox modulation of cyclic electron flow around photosystem I in C3 plants

We have investigated the occurrence of cyclic electron flow in intact spinach leaves. In particular, we have tested the hypothesis that cyclic flow requires the presence of supercomplexes in the thylakoid membrane or other strong associations between proteins. Using biochemical approaches, we found no evidence of the presence of supercomplexes related to cyclic electron flow, making previous structural explanations for the modulation of cyclic flow rather unlikely. On the other hand, we found that the fraction of photosystem I complexes engaged in cyclic flow could be modulated by changes in the redox state of the chloroplast stroma. Our findings support therefore a dynamic model for the occurrence of linear and cyclic electron flow in C3 plants, based on the competition between cytochrome b(6)f and FNR for electrons carried by ferredoxin. This would be ultimately regulated by the balance between the redox state of PSI acceptors and donors during photosynthesis, in a diffusing system.