INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

During oxygenic photosynthesis, photosystem (PS) II and PS I cooperate to achieve a linear electron flow from H₂O to NADP⁺ and to generate a trans-membrane proton gradient driving ATP synthesis. However, ATP can also be produced by the sole PS I through cyclic electron transfer reactions (Arnon, 1959). This mechanism enables the generation of a proton gradient across the thylakoid membrane without NADP reduction by rerouting electrons of reduced PS I acceptors toward the intersystem carriers. Cyclic and linear electron transfers share a common sequence of electron carriers, namely the plastoquinone (PQ) pool, cytochrome b₆/f complex, and plastocyanin (for review, see Fork and Herbert, 1993; Bendall and Manasse, 1995). This alternative electron flow has been shown to occur in vivo in cyanobacteria (Carpentier et al., 1984), in algae (Maxwell and Biggins, 1976; Ravenel et al., 1994), and in bundle sheath cells of C₄ plants (Herbert et al., 1990; Asada et al., 1993). In cyanobacteria, cyclic electron flow around PS I has been shown to provide extra ATP for different cellular processes, e.g. adaptation to salt stress conditions (Jeanjean et al., 1993). In the bundle sheath cell chloroplasts of C₄ plants, PS II is low or undetectable (Woo et al., 1970) and ATP supply is totally dependent upon PS I-mediated cyclic electron transport (Leegood et al., 1981).

In C₃ plants, PS I-driven cyclic electron flow has been studied mainly in vitro on isolated chloroplasts or thylakoids with addition of artificial cofactors or reduced ferredoxin (Bendall and Manasse, 1995). Under those conditions, the redox poise was proposed to play an important role in the regulation of the rate of cyclic electron flow (Arnon and Chain, 1975; Heber et al., 1978; Fork and Herbert, 1993), with neither full reduction of the chloroplast electron transport chain (Ziem-Hanck and Heber, 1980) nor excessive oxidation allowing cyclic electron flow to occur in vitro. In intact leaves, PS I-mediated cyclic electron flow in far-red light was analyzed indirectly by measuring the light-scattering signal at 535 nm, which reflects changes in the trans-thylakoid pH gradient (Heber et al., 1992, 1995; Cornic et al., 2000). Cyclic electron transport around PS I can also be estimated indirectly by measuring the re-reduction rate of the oxidized primary electron donor in PS I (P₇₀₀⁺) after switching off the far-red light (Maxwell and Biggins, 1976; Asada et al., 1992). It was observed that this rate measured in leaves of C₃ plants (e.g. Burrows et al., 1998) was considerably much slower than that measured in the green alga Chlamydomonas reinhardtii (Maxwell and Biggins, 1976; Ravenel et al., 1994) or in cyanobacteria (Mi et al., 1992), suggesting a very slow recycling of electrons around PS I in vivo. The existence of cyclic electron transport in vivo in C₃ plants has also been questioned by photoacoustic measurements in far-red light (Herbert et al., 1990), which allow a direct and quantitative measure of energy storage (ES) by cyclic electron flow around PS I (for review, see Malkin and Canaani, 1994). This method has confirmed the existence of cyclic electron transfer reactions in C₄ plants, algae, and cyanobacteria (Herbert et al., 1990), but failed to show significant cyclic activity in C₃ plant leaves (Herbert et al., 1990; Havaux et al., 1991; Malkin et al., 1992).

New biochemical and genetic data support, however, the idea that cyclic electron flow around PS I occurs in vivo in C₃ plants. The plastid genome of higher plants contains ndh genes encoding peptides homologous to subunits of the proton-pumping NADH:ubiquinone oxidoreductase, a component of the mitochondrial respiratory chain (Ohyama et al., 1986; Shinozaki et al., 1986), and an NADH-dehydrogenase complex (Ndh) has been purified from pea (Pisum sativum) and barley (Hordeum vulgare) thylakoid membranes (Sazanov et al., 1998; Quiles et al., 2000). Inactivation of some ndh genes using plastid transformation of tobacco (Nicotiana tabacum cv Petit Havana) demonstrated the existence of a functional Ndh complex and its involvement in the transient nonphotochemical reduction of the PQ pool after a light to dark transition (Burrows et al., 1998; Cournac et al., 1998; Shikanai et al., 1998). Based on the study of chlorophyll fluorescence kinetics and the effects of inhibitors such as antimycin on tobacco leaf discs of an Ndh-less tobacco mutant, it was recently suggested that the Ndh complex could be involved in a PS I cyclic electron pathway operating in vivo in C₃ plants (Joët et al., 2000).

The apparent discrepancy between those results and the absence of measurable cyclic activity in vivo remains to be elucidated. Cyclic PS I activity is usually measured under very special conditions (PS I excitation by far-red light, PS II inhibition by 3-[3,4-dichlorophenyl]-1,1-dimethylurea [DCMU]) in which linear electron flow is diminished or even abolished. We assumed that the adequate redox poise supposedly required for cyclic electron flow in vivo is not achieved under those experimental conditions. In the present study, we have used photoacoustic spectroscopy and kinetic spectrophotometry to monitor cyclic electron transport around PS I in C₃ plants. A rapid electron cycling around PS I was induced in vivo by increasing the reduction level of the stromal NADP pool and of the intersystem electron carriers using anaerobic conditions or respiration inhibitors. The high cyclic activity of PS I thus obtained was different in wild-type (WT) tobacco and in a mutant lacking the Ndh complex, demonstrating the involvement of the Ndh complex in cyclic PS I activity.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Cyclic Electron Transport around PS I in Different Organisms

PS I-mediated cyclic electron flow was monitored in vivo in Synechocystis sp., maize (Zea mays) and tobacco using the photoacoustic technique (Fig. 1A). This technique measures the conversion of light energy to heat in an absorbing sample and hence the storage of light energy as chemical energy (photochemical ES; see "materials and Methods"; Malkin and Canaani, 1994). Figure 1A shows a typical in vivo photothermal signal generated by Synechocystis sp. cells deposited on a nitrocellulose filter and irradiated with modulated far-red light. Addition of a strong nonmodulated far-red light to the modulated light beam saturates PS I photochemistry, causing a noticeable rise in the photoacoustic signal. Thus, the comparison of the actual and maximal heat-emission signals provides a measure of the amount of absorbed light energy that was stored in intermediates of the photochemical processes (Malkin and Canaani, 1994). ES measured in the cyanobacterium Synechocystis sp. PCC 6803 under such conditions was close to 18%. As far-red light is almost exclusively absorbed in PS I, the measured ES is specifically related to the PS I function, reflecting ES in photochemical products associated with the cycling of electrons around PS I (Canaani et al., 1989; Herbert et al., 1990). In higher plants, photoacoustic measurements must be conducted on water- or buffer-infiltrated leaves to eliminate the oxygen-evolution-related component of the photoacoustic signal (Malkin et al., 1992). In infiltrated leaves of maize (a C₄ plant), ES in far-red light (approximately 15%) was close to the activity measured in Synechocystis sp. In contrast, only a very small fraction (<5%) of the absorbed far-red light was used for photochemistry in infiltrated leaf discs of tobacco, a C₃ plant. This confirms a previous photoacoustic study of Herbert et al. (1990) who showed that C₃ plants exhibit no significant ES in far-red light.

View larger version (24K): [in this window] [in a new window]   Figure 1.   A, Photoacoustic signals generated by filter-deposited Synechocystis sp. cells and by water vacuum-infiltrated leaf discs of maize and tobacco, in wavelengths of measuring light absorbed predominantly by PS I (>715 nm; 10 Hz; 30 W m2). Upward-pointing arrows and downward-pointing arrows respectively indicate saturating far-red light (320 W m2) on and off. Thin wavy arrows represent the modulated measuring light. B, Dark re-reduction of P700+ after a far-red light period was monitored on the same samples and was expressed as  reflectance/reflectance (R/R). Data expressing the half-time (t1/2) of the dark P700+ reduction are means ± SD of six experiments.

Leaf absorbance measurements at 820 nm, which reflect changes in the redox state of the PS I reaction center (P₇₀₀), were also performed on the samples used for the photoacoustic measurements. P₇₀₀ was oxidized by far-red light, and its subsequent reduction in the dark by stromal reductants was recorded (Fig. 1B). The half-time (t_1/2) of the dark re-reduction of P₇₀₀⁺ after a period of far-red light was about 178 ± 12 ms in the cyanobacterium and 196 ± 29 ms in maize leaves. In tobacco, the re-reduction rate was considerably much slower, the t_1/2 value being close to 1,157 ± 113 ms. The latter value indicates a slow re-reduction of P₇₀₀⁺, which can be attributed to a slow kinetics of electron donation from the stroma to the intersystem electron transport chain and/or a small pool of electron donors in the stroma after the far-red light illumination (Mi et al., 1992). In tobacco, it is possible that this pool is oxidized in far-red light and is regenerated very slowly in the dark.

Induction of Cyclic PS I Activity in C₃ Plants by Anaerobiosis

Assuming that the redox poise is a key factor in the induction of cyclic electron flow in C₃ plants, ES and the P₇₀₀ dark reduction were measured under anaerobiosis, a condition known to reduce both the intersystem electron transport chain (Harris and Heber, 1993) and the stromal NADP pool (Joët et al., 1998). Anaerobiosis was reached in situ by placing tobacco leaf discs between two plastic wrap films impermeable to gas exchanges. Under those conditions, mitochondrial respiration consumes molecular O₂ in the gas phase of the leaf sample. This was checked by placing a leaf disc between a Clark-type O₂ electrode and a plastic film (data not shown). Oxygen was quickly consumed within 25 to 30 min, finally reaching the level obtained by flushing nitrogen in the chamber. We followed the changes in chlorophyll fluorescence emission (F_o and F_m levels) by tobacco leaf discs placed between two plastic films (Fig. 2). As previously reported by Harris and Heber (1993), obtention of anaerobiosis resulted in a noticeable increase in the dark level (F_o) of chlorophyll fluorescence (around +35%), indicating a partial reduction of the PQ pool. At the same time, the maximal fluorescence level (F_m) progressively decreased. The latter effect probably reflects a transition from state 1 to state 2, which modified the light energy distribution between PS II and PS I in response to the reduction of the PQ pool via lateral movement of a fraction of the light-harvesting complexes II (Allen, 1992). This was confirmed by measuring 77 K chlorophyll fluorescence spectra of tobacco leaf discs in air or in anaerobiosis. The ratio between the F730 fluorescence peak (corresponding to PS I-associated pigments) and the F685 peak (PS II-associated pigments) was 3.49 ± 0.27 for leaf discs dark-adapted in air and 4.84 ± 0.19 for leaf discs placed between two plastic films, indicating that the latter leaves underwent a transition to state 2 in which light energy is redistributed in favor of PS I. However, we cannot exclude that part of the F_m quenching involves also other factors such as nonphotochemical fluorescence quenching related to a trans-thylakoidal pH.

View larger version (13K): [in this window] [in a new window]   Figure 2.   Changes in chlorophyll fluorescence of a film-enclosed leaf disc. The dark chlorophyll fluorescence Fo () was monitored by rapidly switching the nonactinic modulated measuring light on and off. The maximal fluorescence level Fm () was determined by applying an 800-ms flash of saturating white light to the dark-adapted leaf discs.

Photoacoustic measurements of ES in far-red light were conducted in tobacco leaves subjected to anaerobiosis. Figure 3A shows that anaerobiosis led to a considerable increase in ES by cyclic PS I activity from less than 5% to approximately 25%. The effect of anaerobiosis was also measured on the rate of P₇₀₀ reduction after far-red light (Fig. 3B). In air, t_1/2 was about 1,157 ± 114 ms whereas anaerobic conditions drastically decreased the t_1/2 value to about 393 ± 89 ms, indicating a marked acceleration of the electron donation to P₇₀₀. The photoacoustic photothermal signal generated by tobacco leaf discs was measured at different fluence rates of far-red light, and Figure 4 shows the plot of ES versus the fluence rate. In air, ES was very low at any fluence rate of far-red light. In contrast, ES measured in anaerobiosis was much higher and significantly decreased with increasing far-red light fluence rate, indicating a progressive light saturation of cyclic PS I activity. The inset of Figure 4 represents the plot of ES¹ versus light fluence rate, which is linear (Havaux et al., 1989). The extrapolation of this plot to a fluence rate of zero gives an estimate of the maximal efficiency of photochemical ES. This maximal ES was close to 8% in aerobic conditions whereas it was increased to 30% to 50% in anaerobic conditions. ES measured in anaerobiosis was not inhibited by DCMU, thus confirming that ES measured in far-red light is specifically related to PS I photochemistry (Fig. 4).

View larger version (22K): [in this window] [in a new window]   Figure 3.   A, Photoacoustic signals (arbitrary units) generated by buffer vacuum-infiltrated (aerobiosis) or plastic film-enclosed (anaerobiosis) tobacco leaf discs were measured in wavelengths of measuring light absorbed predominantly by PS I (>715 nm; 10 Hz; 30 W m2). B, Dark re-reduction of P700+ after a far-red light period was measured on buffer vacuum-infiltrated (aerobiosis) and plastic film-enclosed (anaerobiosis) tobacco leaf discs. Data expressing the half-time (t1/2) of the P700+ dark reduction are means ± SD of six experiments.

View larger version (13K): [in this window] [in a new window]   Figure 4.   Plot of ES versus the fluence rate of the measuring far-red light (>715 nm, 10 Hz) for tobacco leaf discs under aerobic () or anaerobic conditions (). The effect of 50 µM DCMU under anaerobiosis is also shown (). Inset, Plot of the reciprocal of ES versus light fluence rate.

Mechanisms by which anaerobiosis leads to a stimulation of PS I-mediated cyclic activity were then investigated. First, the stimulatory effect of anaerobiosis on the P₇₀₀⁺ re-reduction rate was mimicked by inhibiting mitochondrial respiration either with myxothiazol or salicyl hydroxamic acid (SHAM; Table I). The use of two inhibitors is necessary to inhibit both the cytochrome respiratory pathway and the alternative oxidase pathway. Via chloroplast-mitochondria interactions, inhibition of respiration increased the reducing power in the chloroplasts leading to a rapid electron donation from the stroma to the intersystem chain, as did anaerobiosis. We observed the same phenomenon using antimycin A and SHAM (Table I), although antimycin A is also a potent inhibitor of the ferredoxin-dependent pathway of cyclic electron flow around PS I. This indicates that the latter pathway is not limiting for P₇₀₀ re-reduction under our experimental conditions and it can be compensated by the other pathways of nonphotochemical reduction of intersystem electron carriers. It is interesting that P₇₀₀⁺ re-reduction was also faster in leaves infiltrated with propyl gallate, a potent inhibitor of the newly discovered plastid terminal oxidase (Cournac et al., 2000). This protein has been recently overexpressed in tobacco leaves and was clearly shown to be involved in the dark oxidation of the PQ pool and to be sensitive to propyl gallate in planta (T. Joët, B. Genty, E.M. Josse, M. Kuntz, L. Cournac, and G. Peltier, unpublished data). This shows that electron flow to P₇₀₀ can be enhanced by either increasing nonphotochemical reduction of PQ or decreasing its oxidation, suggesting that the reduction state of the PQ pool play a central role in the induction of cyclic electron flow.

View this table: [in this window] [in a new window]   Table I.   Effects of various chemicals on the half-time (t1/2) of the dark reduction of P700+ after far-red illumination (6.5 W m2; >715 nm) The measurements were performed in air, unless specified otherwise. Data are means ± SD of six experiments.

Cyclic electron flow around PS I was searched in various C₃ plant species exposed to anaerobiosis. In air, ES was very low or even undetectable, whereas anaerobiosis caused a strong increase in ES in all species tested (Table II). As a consequence the phenomenon reported for tobacco is not restricted to this species but can be considered as a general response of C₃ plants to anaerobic conditions.

View this table: [in this window] [in a new window]   Table II.   Photoacoustically measured energy storage (%) by PS I cyclic activity in leaves of various C3 plants illuminated with far-red light (30 W·m2; 10 Hz) in air and in anaerobiosis nd, Not detected.

Anaerobiosis-Induced Cyclic Electron Transport around PS I in a Ndh-Less Tobacco Mutant

To further characterize PS I cyclic electron flow in C₃ plants, we performed photoacoustic and 820-nm absorbance measurements on leaves of a tobacco mutant lacking the plastid Ndh complex (Horvath et al., 2000). In Figure 5A, the reciprocal of ES measured in the WT and the mutant under anaerobic conditions was plotted versus the far-red light fluence rate. When photoacoustic measurements were performed using far-red light modulated at a low frequency of 10 Hz, no significant difference in ES was found between the WT and the Ndh-less mutant. This shows that ES by PS I cyclic electron flow can occur in the absence of the Ndh complex, via alternative electron transfer pathways. No significant difference was detected between WT and Ndh-deficient mutant on the basis of the half-time of the P₇₀₀ re-reduction under anaerobiosis (Fig. 5B). However, increasing the modulation frequency to 22 Hz revealed a noticeable difference between the two genotypes (Fig. 5A). The linear plot of the mutant had a much steeper slope than the plot of the WT, indicating that PS I-driven cyclic electron transport was more rapidly saturated with increasing far-red light intensity in the Ndh-deficient mutant. The frequency dependence of ES in far-red light was analyzed in more detail in Figure 6. In WT leaves placed in anaerobiosis, a progressive decrease in ES was observed with increasing frequency of the modulated light. A similar frequency dependence of ES by cyclic PS I activity was previously observed in C. reinhardtii (Canaani et al., 1989). ES depends on the modulation frequency because this parameter reflects energy stored in photochemical products that decay with a time constant larger than the modulation frequency of excitation (Malkin and Canaani, 1994). Thus, at very low frequency, ES reflects long-lived intermediates. More precisely, at a given frequency of modulation f, ES corresponds to photochemical products of electron transport that persist for 1/2f seconds after excitation (Canaani et al., 1989). Considering a maximal ES value of about 20% (at modulation frequencies close to zero), we estimated that the frequency corresponding to the half-value of the maximal ES was around 26 Hz in the WT. From this value, an apparent reaction half-time of 6 ms can be calculated for the limiting step of the reaction responsible for the measured ES. It is striking that ES measured in the Ndh-deficient mutant decreased much more rapidly with the modulation frequency, and the half-value of the maximal ES value was reached at around 12 Hz. In this case, the photochemical reaction responsible for ES in far-red light has a half-time of 13 ms. As expected, ES was low at any frequency in both WT and Ndh-less mutant leaves in air. The data of Figure 6 show that cyclic electron transfers around PS I in WT tobacco and in the Ndh-deficient mutant are kinetically different, the electron cycle being much slower in the latter one.

View larger version (12K): [in this window] [in a new window]   Figure 5.   A, Plot of the reciprocal of ES versus the fluence rate of the measuring far-red light (>715 nm) under anaerobic conditions. ES was measured in WT (, ) and Ndh-less mutant (, ) at 10 Hz (, ) and 22 Hz (, ). B, Dark re-reduction of P700+ after a far-red light period was measured on WT and Ndh-less mutant tobacco leaf discs under anaerobiosis. Data expressing the half-time (t1/2) of the P700+ dark reduction are means ± SD of 12 experiments.

View larger version (17K): [in this window] [in a new window]   Figure 6.   Plot of ES versus the frequency of the modulated light for WT () and Ndh-less mutant leaves (). ES was measured under aerobic (dotted lines) or anaerobic conditions (full lines) using far-red modulated light (>715 nm; 40 W m2). The experimental points are the results of three independent experiments.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Rapid Cyclic Electron Flow around PS I in Vivo in C₃ Plants

In this study, cyclic electron flow around PS I was triggered in vivo by placing tobacco leaf discs in anaerobiosis. The maximal ES level measured in tobacco leaves in far-red light was equivalent to that measured in cyanobacteria and C₄ plants. The frequency dependence of ES measured in anaerobiosis showed that the frequency corresponding to one-half of the maximal ES was about 26 Hz, indicating that the rate constant of the photochemical reaction responsible for the ES was about 6 ms. This is to compare to the 2 ms previously observed in C. reinhardtii (Canaani et al., 1989). To the best of our knowledge, this is the first time that such a high and rapid cyclic electron flow around PS I is measured in vivo in C₃ plants. Moreover, anaerobiosis-induced cyclic PS I activity was observed in several C₃ plant species, indicating that this phenomenon is a general feature of C₃ plants (Table II). The half-time of P₇₀₀⁺ re-reduction measured in tobacco leaf discs was also accelerated in anaerobiosis from 1,200 ms (in air) to 400 ms which is very close to the t_1/2 reported for isolated thylakoids under anaerobiosis (Scheller, 1996). This value corresponds to a rapid electron donation to P₇₀₀ which is of the same magnitude of that measured in Synechocystis sp. and maize (respectively close to 180 and 200 ms) or that reported for C. reinhardtii (220 ms; Ravenel et al., 1994) and other unicellular algae (150 ms; Maxwell and Biggins, 1976). Our results are in accordance with previous studies performed by Heber et al. (1978) who observed a significant increase in the light scattering of spinach leaves illuminated with far-red light by flushing with nitrogen. Since this signal is indirectly related to the thylakoid pH gradient, these authors proposed that PS I-mediated cyclic activity was stimulated by anaerobiosis.

Involvement of the Ndh Complex in Cyclic Electron Flow

The rapid cyclic electron transport around PS I in anaerobiosis is related to the Ndh complex activity. The cyclic PS I activity in Ndh-less mutant was kinetically different from that of the WT as shown by the more abrupt decrease of ES with increasing frequency. This indicates a slower cycling of electrons in the absence of functional Ndh complex. The photochemical reaction responsible for the measured ES in far-red light had a calculated half-time of approximately 13 ms, which is much higher than the value measured in the WT (approximately 6 ms). It is interesting that PS I-mediated cyclic electron flow was not inhibited in Ndh-less mutant leaves of tobacco at very low frequencies of modulated light (<10 Hz) indicating the existence of cyclic electron pathways independent of the Ndh complex. Those alternative pathways could be the antimycin-sensitive pathway (Cleland and Bendall, 1992; Joët et al., 2000) or a pathway involving alternative NAD(P) H dehydrogenases (Corneille et al., 1998; Cournac et al., 1998). The rapid disappearance of ES with increasing frequency in the Ndh-less mutant shows that these additional pathways are slow and that the predominant pathway in the WT is dependent on the Ndh complex. It should incidentally be noted that the photoacoustic method revealed differences in kinetics of cyclic electron flow whereas the P₇₀₀⁺ re-reduction rate measured in the dark was equivalent in the WT and the Ndh-less mutant. This can be explained by the fact that the photoacoustic technique measures the energy stored during the complete electron cycle in far-red light whereas the P₇₀₀ reduction rate in the dark reflects electron donation to P₇₀₀ from stromal donors. This rate depends not only on the kinetics of electron transfer from stromal donors to the intersystem chain but also on the size of the stromal electron donor pool. Thus, the t_1/2 value is expected to depend on the experimental conditions (Mi et al., 1992). It has been shown in cyanobacteria that t_1/2 measured after continuous far-red light represents the donation from respiratory donors, and depletion of respiratory donors in dark-starved cells resulted in a very slow reduction of P₇₀₀⁺ (Mi et al., 1992). Then, t_1/2 cannot be taken as an absolute value of the turnover time of P₇₀₀ during cyclic electron transport around PS I. Rather, it should be considered as an indicator of the potential capacity of electron donation to P₇₀₀ from stroma electron donors. In tobacco, re-reduction of oxidized P₇₀₀ was possibly determined mainly by the re-generation of stromal electron donors after switching off the far-red light. This phenomenon was probably more limiting for the P₇₀₀ reduction rate than the absence of Ndh, thus explaining why the WT and the Ndh-deficient mutant could not be distinguished on the basis of the P₇₀₀ reduction kinetics. In contrast, in the photoacoustic experiments, the frequency dependence of ES allows the kinetics of the electron cycle to be analyzed and provides information on the lifetime of the intermediates that limit the energy-storage reaction (Malkin and Canaani, 1994; Malkin, 1996). Also, it cannot be excluded that some phenomena needed for cyclic electron flow are deactivated in the dark and are thus not observable in the P₇₀₀ redox change experiments. Further studies are probably required to confirm our interpretations.

The involvement of the Ndh complex in cyclic electron transport around PS I is thus demonstrated in this study, confirming a number of previous reports that have hypothesized the role of the Ndh complex in cyclic electron transport via measurements of nonphotochemical reduction of the PQ pool in the dark (Burrows et al., 1998; Shikanai et al., 1998; Cournac et al., 1998). It was recently shown that photosynthesis of the Ndh-less mutant leaves was highly sensitive to antimycin A, and it was concluded to the participation of the Ndh complex in cyclic electron flow (Joët et al., 2000). Cyclic electron flow via the NADPH pool has already been described in the cyanobacterium Synechocystis sp., where lesions in the ndh genes were observed to cause a strong slowdown of the P₇₀₀ turnover in far-red light (Mi et al., 1992, 1995) and a marked inhibition of ES in far-red light (Jeanjean et al., 1998). At last, an NAD(P) H dehydrogenase activity involved in PS I cyclic activity has also been suggested in C. reinhardtii where several independent pathways may coexist in vivo (Ravenel et al., 1994). It should be pointed out that the involvement of Ndh in cyclic electron transport around PS I can be direct by allowing rapid recycling of electrons from NADPH to the PQ pool or indirect as a redox poise regulator (compare with below).

The Redox Poise Controls Cyclic PS I Activity in C₃ Plants

It is likely that PS I-mediated cyclic electron transport is controlled by the atmospheric O₂ concentration in C₃ plants via changes in the redox state of intersystem electron carriers and/or stromal reductants. The PQ pool was found to be partially reduced in tobacco leaves exposed to anaerobiosis in plastic wrap films. In a previous study, the NADPH to NADP ratio, as measured in vivo by the blue-green fluorescence emission, was observed to increase noticeably in anaerobiosis (Joët et al., 1998). A control of PS I-driven cyclic electron transfer by the reduction state of intersystem electron carriers and through the NADP⁺ to NADPH ratio was previously suggested from in vitro data obtained on isolated C₃ chloroplasts (Arnon and Chain, 1975, 1979; Slovacek et al., 1980; Takahama et al., 1981; Hosler and Yocum, 1985, 1987) and from light scattering measurements on leaves (Heber et al., 1978). In those experiments, it was shown that cyclic activity is modulated by varying the O₂ partial pressure to modify the redox poise of the intersystem electron carriers (Arnon and Chain, 1975, 1979; Scheller, 1996). This suggests that adequate redox poise was not achieved in leaves placed in air and illuminated with far-red light, thus explaining why PS I-mediated cyclic activity was not detectable in vivo in C₃ plants.

Adequate redox poise was also induced in air either by inhibiting mitochondrial oxidases using myxothiazol or antimycin and SHAM or by inhibiting a chloroplastic PQ oxidation pathway using propyl gallate. In the former case, the cytosolic NAD(P) H, which cannot be oxidized by mitorespiration, is re-routed toward chloroplasts because of the existence of redox interactions between mitochondria and plastid (for review, see Hoefnagel et al., 1998), hence stimulating nonphotochemical reduction of the intersystem electron carriers (Gans and Rebeillé, 1990). In leaves treated with propyl gallate, inhibition of a plastid oxidase involved in nonphotochemical oxidation of the PQ pool, led to an over-reduction of the intersystem electron carriers. Those mechanisms, by which the PQ pool is nonphotochemically reduced and subsequently reoxidized in the dark using molecular oxygen as a terminal acceptor, are commonly described as chlororespiration (Bennoun, 1982; Peltier et al., 1987; Nixon, 2000). From our data, we conclude that the entire chlororespiratory electron transfer chain, i.e. the nonphotochemical reduction as well as the nonphotochemical oxidation of the PQ pool, may control the redox poise of intersystem electron transport chain in vivo in C₃ plants, which in turn controls PS I-mediated cyclic electron flow.

Anaerobic conditions were used in this study as an experimental trick that allows measurement of cyclic electron transport in far-red light. Under physiological conditions, the adequate redox state of the PQ pool may be achieved by the PS II activity, especially when PS I acceptors are not fully available. Under our experimental conditions, however, cyclic electron flow is measured in far-red light, which cannot stimulate PS II activity, and cyclic electron flow is then artificially activated in anaerobiosis by simultaneous stimulation of nonphotochemical PQ reduction and inhibition of PQ oxidation by the plastid oxidase. The situation described here in higher plant chloroplasts is close to that occurring in cyanobacteria where respiration and photosynthesis electron transfer chains share the PQ pool in common (Scherer, 1990). It is interesting that Schubert et al. (1995) reported a stimulation of PS I-driven cyclic electron flow, estimated by the photochemical ES in far-red light and by the P₇₀₀⁺ dark reduction in the cyanobacterium Fremyella diplosiphon when treated with KCN, a well-known inhibitor of cytochrome c oxidase.

Transition to the light state 2, which is activated when the intersystem redox carriers are reduced (Allen, 1992), took place in tobacco leaves exposed to anaerobiosis. During transition to state 2, the cytochrome b₆/f complex was shown to accumulate in the stroma lamellae of maize and C. reinhardtii (Vallon et al., 1991) where both PS I and the Ndh complex are located (Horvath et al., 2000; Sazanov et al., 1998). This could possibly facilitate PS I cyclic activity. One may then suppose that reduction of the PQ pool can indirectly favor cyclic electron flow via state-transition-related migration of cyt b₆f to the vicinity of PS I. It is interesting that inhibitors of mitochondrial respiration were also reported to induce reduction of the PQ pool and transition to state 2 in C. reinhardtii (Bulté et al., 1990).

One intriguing question is why the adequate redox poise for cyclic electron transport requires specific conditions in C₃ plants and is naturally observed in other types of plants under far-red light conditions. Under normal conditions, it is possible that nonphotochemical reduction of the PQ pool is not sufficient to achieve the adequate redox poise of intersystem electron carriers. Cyclic electron flow will be triggered by a reduction of the PQ pool mediated either by an imbalance in chlororespiration activity between nonphotochemical reduction and oxidation of PQs or by PS II activity. This could be the case for example during induction of photosynthesis or in high light, where the intersystem electron carriers are partly reduced. The Calvin cycle uses more ATP than NADPH so that a high photosynthetic rate will lead to an increased NADPH to ATP ratio (Osmond, 1981). It has been suggested that one function of cyclic electron transfer around PS I is to synthesize extra ATP to adjust the NADPH to ATP ratio (Bendall and Manasse, 1995). Our observation that PS I cyclic activity is controlled by the redox poise of the chloroplasts is consistent with this function: An accumulation of NADPH will trigger the cyclic electron flow, thus compensating the ATP deficit.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Plant Material and Preparation of Leaf Samples

WT tobacco (Nicotiana tabacum cv Petit Havana) and ndhB-inactivated mutant (see Horvath et al., 2000) were grown on compost in a phytotron (25°C day/20°C night) under a photon flux density of 350 µmol photons m² s¹ (photoperiod, 12 h) supplied by quartz halogen lamps (HQI-T 400W/DV, Osram, Germany). Plants were watered with one-half-strength Hoagland nutritive solution. Leaf samples were taken from plants aged 5 to 8 weeks. Potato (Solanum tuberosum L. cv Desirée), Arabidopsis cv Colombia, barley (Hordeum vulgare L. cv Plaisant), and Brassica napus cv Orphée were cultivated under the same conditions, whereas poplar (Populus koreana × Populus trichocarpa Torr. & Gray cv Peace) was grown in a greenhouse. Samples from Acer pseudoplatanus were harvested outside.

Anaerobiosis was induced by placing leaf discs (12 mm in diameter) between two plastic films of barrier food wraps (Saran, Dow Chemical, Midland, MI) and keeping them in the dark for 60 min before photoacoustic or absorbance measurements. Because of respiration, O₂ was rapidly depleted in the leaf discs, as controlled with an O₂ electrode at 25°C.

Treatments with Inhibitors

DCMU treatment was performed on whole leaves that were infiltrated with 50 µM DCMU via their petiole through the transpiration flux for 5 h. PS II inhibition by DCMU was checked by chlorophyll fluorescence yield measurements with a PAM-2000 fluorometer (Walz, Effeltrich, Germany). For other treatments, tobacco leaf discs were stripped by removal of lower epidermis and soaked for 90 min in petri dishes containing distilled water and various inhibitors. The inhibitors were added diluted in methanol (maximal final methanol concentration was 0.5%). Control leaf discs were soaked in petri dishes containing water and 0.5% (v/v) methanol.

Photoacoustic Measurements of Photochemical ES

ES by cyclic electron flow around PS I was measured in vivo using the photoacoustic technique (Herbert et al., 1990; Havaux et al., 1991; Ravenel et al., 1994). The photoacoustic spectrometer is described in Ravenel et al. (1994). Leaf discs, placed in the photoacoustic chamber, were illuminated with modulated far-red light (>715 nm). The far-red light fluence rate was measured with a LI-COR radiometer (Li-185B/Li-200SB, LI-COR, Lincoln, NE). PS I photochemistry was saturated with a strong background far-red light (>715 nm, 320 W m²). ES was calculated from the amplitude of the maximal photothermal photoacoustic signal (Apt⁺ measured when the strong far-red light was added to the modulated measuring light) and the actual photothermal signal amplitude (Apt):

(1)

The plastic wrap film is impermeable to O₂ and, therefore, abolishes the gas-exchange-related photoacoustic signal that can appear at low modulation frequencies. When the plastic film was not used, leaf discs were soaked for 3 h in an osmoticum buffer (25 mM phosphate buffer, pH 7.0, 200 mM sorbitol, 10 mM KCl, and 2 mM MgCl₂) to eliminate the photobaric component of the photoacoustic signal as described by Malkin et al. (1992). Synechocystis sp. cells were filtered under pressure through an MF-Millipore filter (cellulose nitrate/acetate, SS type, 3-µm pore size, Millipore, Bedford, MA). The cells deposited on the filter (diameter, 1.2 cm) were then placed in the photoacoustic cell for measurement.

Redox State of P700

Changes in the redox state of the reaction center P₇₀₀ of PS I were monitored via leaf absorbance changes at around 820 nm (Schreiber et al., 1988). A Walz PAM-101 system connected to an ED-800-T emitter/detector unit (Walz) was used in the reflection mode (Schreiber et al., 1988). The rate of P₇₀₀ re-reduction was measured with a storage oscilloscope (Tektronix 5111A, Tektronix, Guernsey, Channel Islands) after a period of far-red light (6.5 W m², >715 nm).

Room Temperature and Low Temperature Chlorophyll Fluorescence Measurements

Chlorophyll fluorescence spectra were recorded in liquid nitrogen (77 K) with leaf discs dark-adapted for 30 min using a bifurcated light guide connected to a Perkin-Elmer LS50B spectrofluorometer (Perkin-Elmer, Beaconsfield, UK). The wavelength of the excitation light beam was 440 nm.

Room temperature chlorophyll fluorescence was measured with a PAM-2000 chlorophyll fluorometer (Walz). F_o was excited with a dim red light modulated at 600 Hz. Fm was induced by a 800-ms pulse of intense white light.