Effect of PGR5 Impairment on Photosynthesis and Growth in Arabidopsis thaliana

Abstract

PGR5 has been reported as an important factor for the activity of the ferredoxin-dependent cyclic electron transport around PSI. To elucidate the role of PGR5 in C₃ photosynthesis, we characterized the photosynthetic electron transport rate (ETR), CO₂ assimilation and growth in the Arabidopsis thaliana pgr5 mutant at various irradiances and with CO₂ regimes. In low-light-grown pgr5, the CO₂ assimilation rate and ETR were similar to the those of the wild type at low irradiance, but decreased at saturating irradiance under photorespiratory conditions as well as non-photorespiratory conditions. Although non-photochemical quenching of chlorophyll fluorescence (NPQ) was not induced in the pgr5 mutant under steady-state photosynthesis, we show that it was induced under dark to light transition at low CO₂ concentration. Under low light conditions in air, pgr5 showed the same growth as the wild type, but a significant growth reduction compared with the wild type at >150 μmol photons m⁻² s⁻¹. This growth impairment was largely suppressed under high CO₂ concentrations. Based on the intercellular CO₂ concentration dependency of CO₂ assimilation, ETR and P700 oxidation measurements, we conclude that reduction of photosynthesis and growth result from (i) ATP deficiency and (ii) inactivation of PSI. We discuss these data in relation to the role of PGR5-dependent regulatory mechanisms in tuning the ATP/NADPH ratio and preventing inactivation of PSI, especially under conditions of high irradiance or enhanced photorespiration.

Introduction

During oxygenic photosynthesis, light energy is converted into chemical energy by two coupled photosystems, PSII and PSI, generating linear electron transport from water to NADP⁺. During this process, protons are released in the lumen by water oxidation at PSII and plastoquinol oxidation by the cytochrome b₆f complex, and the resulting generation of a proton gradient across the thylakoid membrane drives ATP synthesis. The stoichiometry of ATP/NADPH production during linear electron transport has been a matter of debate (Heber and Walker 1992, Bendall and Manasse 1995, Allen 2003). According to recent views which include Q cycle operation at the cytochrome b₆f complex, production of one molecule of NADP⁺ by linear electron transport is thought to be accompanied by the translocation of six protons into the lumen (Allen 2003). Assuming that ATP synthesis in the chloroplast requires 14 protons for the synthesis of three ATP molecules (Seelert et al. 2000), the ATP/NADPH production ratio is calculated as 1.29. The ATP/NADPH ratio required for CO₂ assimilation varies from 1.5 to 1.66 depending on the relative activities of C₃ Calvin–Benson and photorespiratory cycles, respectively (Osmond 1981). For CO₂ assimilation in air, the ATP/NADPH requirement drops to 1.43 if nitrate assimilation is considered (Kramer et al. 2004). This indicates a relative lack of ATP supply for the optimal CO₂ assimilation rate. Therefore, additional mechanisms have been considered to fulfill the the requirement for adequate ATP for CO₂ fixation. Cyclic electron transport around PSI, first evidenced as cyclic photophosphorylation, was proposed to produce extra ATP by recycling electrons from PSI acceptors such as ferredoxin or NADPH to plastoquinones (Arnon et al. 1954). Although inhibitor studies using antimycin A concluded that there was an involvement of cyclic electron transport in ΔpH formation and ATP production (Tagawa et al. 1963, Arnon et al. 1967, Crowther et al. 1979, Furbank and Horton 1987), the in vivo PSI cyclic activity is often considered as very low (Herbert et al. 1990, Fork and Herbert 1993, Bendall and Manasse 1995). Several reports suggest a larger contribution of cyclic electron transport in sustaining ΔpH formation and non-photochemical quenching of chlorophyll fluorescence (NPQ) in conditions such as low CO₂, high light and the induction phase of photosynthesis following a dark to light transition; however, the contribution of cyclic electron transport to steady-state photosynthesis for energy production is still a matter of debate (Harbinson and Foyer 1991, Golding et al. 2004, Joliot and Joliot 2005, Miyake et al. 2005, Laisk et al. 2007).

The study of tobacco mutants deficient in the plastid NDH complex demonstrated the existence of a functional NDH complex involved in plastoquinone reduction (Burrows et al. 1998, Shikanai et al. 1998, Horvath et al. 2000), and it was proposed that this complex participates in cyclic electron transport by recycling electrons from NAD(P)H to plastoquinone. However, inactivation of the NDH complex did not lead to a decrease in the photosynthetic CO₂ assimilation rate under optimal conditions in air (Burrows et al. 1998, Shikanai et al. 1998, Horvath et al. 2000), but a significant effect was observed in conditions of high photorespiration induced by water stress (Horvath et al. 2000). By studying the effect of antimycin A on an ndhB knock-out mutant, Joët et al. (2001) concluded that two parallel cyclic pathways exist around PSI: one sensitive to antimycin A and the other involving the NDH complex. Although several nuclear-encoded subunits of the NDH complex have been identified recently (Munshi et al. 2005, Rumeau et al. 2005, Muraoka et al. 2006), subunits involved in NADH or NADPH binding have not been found yet.

From a genetic approach in Arabidopsis thaliana based on the screening of mutants impaired in NPQ, PGR5 was concluded to encode an essential component of the antimycin A-sensitive cyclic activity (Munekage et al. 2002). The pgr5 mutant showed over-reduction of stroma and a reduced electron transport rate (ETR) under high irradiance (Munekage et al. 2002). Furthermore, a severe decrease in electron transport and photoautotrophic growth was observed in a double mutant crr2 pgr5 impaired in both PGR5 and the NDH complex (Munekage et al. 2004). From these results, it was proposed that PGR5 is involved in cyclic electron transport around PSI, thereby contributing to lumenal acidification, and in turn NPQ induction and ATP synthesis (Munekage et al. 2002). Okegawa et al. (2007) showed that overexpression of PGR5 in A. thaliana enhanced the activity of cyclic electron transport. However, this interpretation was recently challenged by results showing large residual cyclic activity during induction in pgr5 and in crr2 pgr5 (Nandha et al. 2007). Moreover, as PGR5 is a small protein without an identified electron transport domain, its mechanism of action remains unelucidated and knowledge of the nature of electron carriers involved in the antimycin A-sensitive cyclic pathway is still lacking. Very recently, PGR5 has been shown to interact functionally and physically with a newly identified transmembrane protein PGRL1 that is associated with PSI in A. thaliana (Dalcorso et al. 2008). The authors proposed that the cyclic electron transport activity is dependent on the formation of a PGR5–PGRL1 complex.

With the aim of investigating further how PGR5 deficiency impacts on photosynthesis, we compare CO₂ assimilation, chlorophyll fluorescence, PSI absorbance change and growth of pgr5 mutant and wild-type A. thaliana under non-photorespiratory and photorespiratory conditions. We report a decrease of photosynthetic CO₂ assimilation in pgr5 under subsaturating and saturating irradiance, and a growth reduction which was dependent on CO₂ concentration and irradiance. Based on functional measurements, we conclude that PGR5 is required for optimum photosynthesis by sustaining the ATP supply and preventing inactivation of PSI especially under conditions of high irradiance or enhanced activity of photorespiration.

Results

CO₂ assimilation, photosynthetic electron transport and NPQ in pgr5

Since pgr5 has been reported to be sensitive to high light (Munekage et al. 2002), we used plants grown at low irradiance (50 μmol photons m⁻² s⁻¹) for the measurement of gas exchange. In this condition, pgr5 grew as well as the wild type, and the maximum photochemical yield of PSII (F_v/F_m) and the Chl a/b ratio were not affected (Table 1). Although the chlorophyll content was reduced by 25% in pgr5, leaf absorbance was only slightly reduced (Table 1). The irradiance dependency of net CO₂ assimilation and ETR was investigated in air conditions (350 μl l⁻¹ CO₂, 21% O₂) (Fig. 1). Respiration of pgr5 measured in dark conditions was not different from that of the wild type (Fig. 1A). At low irradiance (<100 μmol photons m⁻² s⁻¹), net CO₂ assimilation and ETR were similar in pgr5 and in the wild type, whereas at higher irradiance, both net CO₂ assimilation and ETR were lower (14 and 29%, respectively at 500 μmol photons m⁻² s⁻¹) in pgr5 than in the wild type (Fig. 1A, B). At steady state, the large increase of NPQ observed in response to increasing irradiance above 200 μmol photons m⁻² s⁻¹ in the wild type was not observed in pgr5 (Fig. 1C). (1 − q_p) was markedly higher in pgr5 than in the wild type at irradiances >200 μmol photons m⁻² s⁻¹ (Fig. 1D), indicating that the acceptor side of PSII is more reduced in pgr5.

Table 1

Chlorophyll contents (Chl a + b), chlorophyll a/b ratio (Chl a/b), leaf absorption (α) and F_v/F_m

Genotype	Chl a + b (mg m−2)	Chl a/b	α	Fv/Fm
Wild type	257 ± 22	2.45 ± 0.08	0.838 ± 0.004	0.765 ± 0.010
pgr5	194 ± 19*	2.49 ± 0.07	0.816 ± 0.010*	0.771 ± 0.008

Plants were grown at 50 μmol photons m⁻² s⁻¹ in a short-day photoperiod (8 h of light, 16 h of dark) for 40 d. Data are means ± SD (n = 5–6). *P < 0.005 (t-test, wild-type vs. pgr5).

Fig. 1

Irradiance dependency of net CO₂ assimilation and chlorophyll fluorescence parameters in leaves of the wild type (filled symbols) and pgr5 (open symbols) measured under ambient conditions (350 μl l⁻¹ CO₂, 21% O₂). (A) Net CO₂ assimilation. (B) Electron transport rate at PSII (ETR). (C) Non-photochemical quenching of chlorophyll fluorescence (NPQ). (D) Photochemical quenching expressed as (1 − q_p). Plants were grown at 50 μmol photons m⁻² s⁻¹. Each point represents the mean ± SD of three individual plants. Differences between the wild type and pgr5 were significant (P < 0.05, Student's t-test) for all parameters at irradiances ≥200 μmol photons m⁻² s⁻¹.

When measurements were performed under non-photorespiratory conditions (350 μl l⁻¹ CO₂ with 1.3% O₂ background), net CO₂ assimilation was decreased by approximately 17% and ETR was decreased by approximately 35% in pgr5 (Fig. 2). As in air, a higher reduction of ETR than CO₂ assimilation was observed in pgr5. Photoinhibition of PSII as monitored by F_v/F_m was not observed in pgr5 after gas exchange measurements (data not shown). However, we found that when ETR was measured at low irradiance after exposure at the higher irradiances its value significantly decreased. This inhibition that we attribute to PSI inactivation will be characterized in a later section.

Fig. 2

Irradiance dependency of net CO₂ assimilation and chlorophyll fluorescence parameters in leaves of the wild type (filled symbols) and pgr5 (open symbols) measured under non-photorespiratory conditions (350 μl l⁻¹ CO₂ 1.3% O₂). (A) Net CO₂ assimilation. (B) Electron transport rate at PSII (ETR). Each point represents the mean ± SD of three individual plants. Plants were grown at 50 μmol photons m⁻² s⁻¹. Differences between the wild type and pgr5 for net CO₂ assimilation and ETR were significant (P < 0.05, Student's t-test) at irradiances of ≥50 and 200 μmol photons m⁻² s⁻¹, respectively.

Intercellular CO₂ concentration (Ci) dependency of net CO₂ assimilation was investigated in the wild type and pgr5 under 21% O₂ conditions (Fig. 3A). We used subsaturating irradiance of 300 μmol photons m⁻² s⁻¹ and performed the lower and higher CO₂ range measurements separately on different leaves to minimize the high irradiance-induced inhibitory effect on pgr5. At low CO₂ concentration (at Ci <200 μl l⁻¹), where carboxylation efficiency limits net CO₂ assimilation, net CO₂ assimilation was similar in pgr5 and the wild type (Fig. 3A).

Fig. 3

Intercellular CO₂ concentration (Ci) dependency of net CO₂ assimilation and chlorophyll fluorescence parameters in leaves of the wild type (filled symbols) and pgr5 (open symbols) at 300 μmol photons m⁻² s⁻¹ in a 21% O₂ background. (A) Net CO₂ assimilation. (B) Electron transport rate at PSII (ETR). (C) Non-photochemical quenching of chlorophyll fluorescence (NPQ). (D) Reduced plastoquinone level (1 − q_p). Each point represents the mean ± SD of 9–13 separate leaves from six individual plants. Plants were grown at 50 μmol photons m⁻² s⁻¹. Differences between the wild type and pgr5 for net CO₂ assimilation were significant (P < 0.02, Student's t-test) at Ci ≥250 μl l⁻¹, those for ETR and 1 − q_p were significant at all Ci, and those for NPQ were significant at Ci ≤ 400 μl l⁻¹.

At high CO₂ concentration where ribulose-1,5-bisphosphate (RuBP) regeneration limits CO₂ assimilation, net CO₂ assimilation reached 13 μmol CO₂ m⁻² s⁻¹ in the wild type whereas it saturated at around 9 μmol CO₂ m⁻² s⁻¹ in pgr5 (Fig. 3A). In pgr5, the ETR was lower than in the wild type at every CO₂ concentration, even at low Ci concentration when assimilation is not affected (Fig. 3B). In pgr5, NPQ at steady state was not dependent on Ci, in contrast to the wild type in which a marked increase of NPQ was induced at low Ci (<400 μl l⁻¹) (Fig. 3C). (1 − q_p) in pgr5 was higher than in the wild type at every CO₂ concentration (Fig. 3D).

Under steady-state photosynthesis, pgr5 almost completely lacks NPQ at high irradiance and low Ci (Figs. 1C and 3C, respectively; see also Munekage et al. 2002). However, following a dark to high light transition in CO₂-depleted air, a similar NPQ was induced in the wild type and pgr5 (Fig. 4). In the wild type, NPQ was induced at up to 1.1 within 7 min illumination and relaxed in the dark within 27 ± 5 s for the half reversion time (t_1/2). In pgr5, NPQ reached 1.0 and showed an almost 2-fold faster relaxation rate in the dark (t_1/2 = 14 ± 2 s). These results show that qE, the proton gradient-dependent type of NPQ, can be induced under particular conditions when PGR5 is impaired.

Fig. 4

Non photochemical quenching (NPQ) induction and relaxation under CO₂-depleted (70 μl l⁻¹) air at 500 μmol photons m⁻² s⁻¹ in the wild type and pgr5. Plants were grown at 50 μmol photons m⁻² s⁻¹. Maximum NPQ are represented as mean ± SD (n = 5).

Inactivation of PSI occurs in pgr5 at subsaturating irradiance

To investigate the inhibition of photosynthetic electron transport observed in pgr5 at subsaturating irradiance, the maximal and steady-state photochemical yield of PSII (F_v/F_m and Φ_PSII, respectively), the reduction level of Q_A (1 − q_P) and the maximal and steady-state P700 oxidation level (ΔA_max and ΔA/ΔA_max, respectively) were monitored after exposure to subsaturating irradiance of 300 μmol photons m⁻² s⁻¹ (Fig. 5). Before exposure, all parameters were similar in pgr5 and the wild type, except the maximal absorbance change at 820 nm (ΔA_max), which was approximately 79% of that of the wild type in pgr5 (Fig. 5). After exposure of pgr5 to 300 μmol photons m⁻² s⁻¹ for 30 min, F_v/F_m, steady-state photochemical yield of PSII, steady-state P700 oxidation and maximal absorbance change at 820 nm decreased by 3, 12, 34 and 20%, respectively, while (1 − q_P) increased. Those effects were enhanced after 4 h exposure (Fig. 5). In contrast, no changes occurred in the wild type in these conditions. These results indicate that following exposure to 300 μmol photons m⁻² s⁻¹ irradiance, linear electron transport impairment occurring in pgr5 is probably due to an inactivation of PSI. Since restoration of damaged PSI requires several days (Sonoike 1996), recovery was investigated after 1 d. F_v/F_m recovered after 1 d whereas the other parameters did not recover (Fig. 5). We conclude that when pgr5 is exposed to subsaturating irradiance (300 μmol photons m⁻² s⁻¹), photoinactivation of PSI occurs, thus explaining the observed decrease in the linear ETR.

Fig. 5

Effect of high irradiance on photosynthetic electron transport parameters in the wild type (white bar) and pgr5 mutant (gray bar). Plants grown at 50 μmol photons m⁻² s⁻¹ were exposed to 300 μmol photons m⁻² s⁻¹ for the indicated period. Recovery from 4 h light treatment was measured after a 1 d (8 h of light at 50 μmol photons m⁻² s⁻¹, 16 h of dark) period. F_v/F_m and ΔA_max were measured after 30 min dark adaptation. Φ_PSII, 1 − q_p and ΔA/ΔA_max (P700 oxidation ratio) were measured at 50 μmol photons m⁻² s⁻¹. Error bars represent the SD for n = 5–11. *P < 0.05 (t-test, wild type vs. pgr5); ^#P < 0.05 (t-test, before light treatment vs. after 30 min or 4 h light treatment).

Effects of CO₂ and irradiance on growth in pgr5

To test how CO₂ and irradiance conditions impact on growth in relation to inactivation of ETR, the wild type and pgr5 mutant were grown at 150 μmol photons m⁻² s⁻¹ under three different atmospheric CO₂ concentrations, high (2,000 μl l⁻¹), ambient (350 μl l⁻¹) and low (150 μl l⁻¹). When plants were grown in high CO₂, pgr5 showed similar growth to the wild type, as shown by FW and DW (Fig. 6A, B), and no inactivation of photosystems was observed (Fig. 6B). In contrast, under ambient and low CO₂ conditions, pgr5 shows drastic growth reduction (Fig. 6A). The FW and DW of pgr5 were 39 and 37%, respectively, in ambient conditions and 36 and 35%, respectively, in low CO₂ conditions in the wild type (Fig. 6B). Although the maximal photochemical yield of PSII (F_v/F_m) was not affected at each CO₂ concentration, the steady-state P700 oxidation level was impaired in pgr5 when it was grown under low CO₂ conditions (Fig. 6B). To assess whether photoinactivation of PSI can be observed under low CO₂ concentration in the same way as it occurs under subsaturating irradiance, plants grown at 150 μmol photons m⁻² s⁻¹ under high CO₂ concentration were transferred to low CO₂ conditions (100 μl l⁻¹) at 150 μmol photons m⁻² s⁻¹ for 30 min. F_v/F_m slightly decreased (−4%) whereas the relative maximal absorbance change at 820 nm significantly decreased by 23% in pgr5, indicating that photoinactivation of PSI can occur at low irradiances under low CO₂ concentrations (data not shown).

Fig. 6

Growth of plants and activity of photosystems in response to different CO₂ concentrations at an irradiance of 150 μmol photons m⁻² s⁻¹. (A) Images of plants. (B) Fresh weight, dry weight, F_v/F_m measured in dark-adapted leaves and ΔA/ΔA_max measured at 50 μmol photons m⁻² s⁻¹ in the the wild type (white bar) and pgr5 (gray bar). Plants were grown in low CO₂ (150 μl l⁻¹), ambient (350 μl l⁻¹) and high CO₂ (2,000 μl l⁻¹) air conditions in a short-day photoperiod (8 h of light, 16 h of dark) for 45 d. Error bars represent the SD for six plants. *P < 0.05 (t-test, wild type vs. pgr5).

When plants were grown at ambient CO₂ under a higher irradiance (300 μmol photons m⁻² s⁻¹), pgr5 showed an even more drastic growth reduction than for growth at 150 μmol photons m⁻² s⁻¹ (Fig. 7A). The FW and DW of pgr5 were only 25 and 22%, respectively, those of the wild type. The npq4 mutant (Li et al. 2000), which was used as a control to evaluate the effect of NPQ deficiency in the conditions used, did not show growth reduction (Fig. 7A). Under high CO₂, growth reduction was largely diminished (FW and DW were 71 and 58%, respectively, those of the wild-type) (Fig. 7A, B). The steady-state P700 oxidation level measured in pgr5 was reduced relative to the wild type, and the extent of the PSI inactivation was larger at ambient CO₂ than at high CO₂ (Fig. 7B). In npq4, the P700 oxidation level or maximum activity of PSII were not affected, showing that impaired activity of PSII or PSI in pgr5 is not due to lack of NPQ by itself.

Fig. 7

Growth of plants and activity of photosystems in response to different CO₂ concentrations at an irradiance of 300 μmol photons m⁻² s⁻¹. (A) Images of plants. (B) Fresh weight, dry weight, F_v/F_m measured in dark-adapted leaves and ΔA/ΔA_max measured at 50 μmol photons m⁻² s⁻¹ in the wild type (white bar), pgr5 (gray bar) and npq4 (black bar). Plants were grown in ambient (350 μl l⁻¹) and high CO₂ (2,000 μl l⁻¹) air conditions in a short-day photoperiod (8 h of light, 16 h of dark) for 45 d. Error bars represent the SD for six plants. *P < 0.05 (t-test, wild-type vs. pgr5 or npq4).

Discussion

PGR5 deficiency impact on CO₂ assimilation

In this study, we provide evidence that PGR5 is essential for optimal photosynthesis at high rates of carboxylation, particularly when the rate of RuBP regeneration limits photosynthesis. At a CO₂ composition of air under both photorespiratory (21% O₂) and non-photorespiratory conditions (1.3% O₂), CO₂ assimilation and ETR were similar in the pgr5 mutant and the wild type at low irradiance, but up to 14–17 and 29–35% less active, respectively, in pgr5 at higher irradiance (Figs. 1, 2). This observation is very similar to the antimycin A response of CO₂-dependent O₂ evolution reported for protoplasts (Furbank and Horton 1987) and leaf discs (Cornic et al. 2000) in which antimycin A inhibits photosynthesis at high irradiance. It supports that pgr5 alters the antimycin A-sensitive cyclic pathway as proposed by Munekage et al (2002). The PGR5-dependent cyclic pathway seems to be required for ATP supply at subsaturating and saturating irradiances. At subsaturating irradiance, when CO₂ supply limits carboxylation, no difference in CO₂ assimilation was observed under photorespiratory conditions (Fig. 3). In contrast, at higher CO₂ when RuBP regeneration limits photosynthesis, a marked inhibition of CO₂ assimilation (up to 30%) was shown in pgr5 relative to the wild type (Fig. 3). Although we used subsaturating irradiance (300 μmol photons m⁻² s⁻¹) for measuring CO₂ dependency of photosynthesis to minimize light effects, we observed that PSI is predominantly inactivated in pgr5 (Fig. 5). The inactivation of PSI was induced within 5 min (data not shown) and was enhanced with time (Fig. 5) so that we could not eliminate a possible decrease in CO₂ assimilation due to PSI inactivation. A striking result is that, in pgr5, the reduction of ETR observed at subsaturating and saturating irradiances is higher than the reduction of CO₂ assimilation under both photorespiratory and non-photorespiratory conditions (Figs. 1–3). This phenomenon is observed in conditions where inactivation of PSI is revealed. This suggests the existence of an alternative sink for ETR in the wild type that is absent in pgr5, and does not depend on the oxygen concentration in a range between 1.3 and 21% O₂. Possible candidates for this alternative sink with high O₂ affinity are Mehler-type reactions (high affinity type; Badger 1985) and mitochondrial activity coupled to the malate–oxaloacetate shuttle (Gardeström and Lernmark 1995, Hoefnagel et al. 1998). However, this second possibility is unlikely, as Yoshida et al. (2007) have shown that the activities of mitochondrial oxidase and NADP-malate dehydrogenase (NADP-MDH) were up-regulated in pgr5 compared with the wild type.

PGR5 deficiency leads to low NPQ at steady state

The pgr5 mutant shows an almost complete lack of qE at high light and low Ci under steady-state photosynthesis (Figs. 1C, 3C), and low qE induction during dark to low light transition in air (Munekage et al. 2002). However, in particular conditions, such as dark to high light transition in low CO₂, qE was greatly induced in pgr5 at almost the same level as the wild type (Fig. 4), but vanished when the steady-state CO₂ assimilation rate was reached (not shown). Note that qE induction in pgr5 required both high light and low CO2 in order to be observed. These results suggest that in these conditions of photosynthetic induction, where the ATP demand is low, the occurrence of other electron transport mechanisms such as PGR5-independent cyclic pathways or Mehler-type reactions can induce significant lumenal acidification and subsequently large qE. This is consistent with data of Nandha et al. (2007) showing significant cyclic activity during induction in pgr5 under CO₂-depleted conditions. This suggests that PGR5 is more important during steady-state photosynthesis when ATP demand is high rather than during induction of photosynthesis. We observed a faster relaxation of qE in the dark in pgr5 (2-fold faster than the wild type). Since qE has been shown to be dependent on luminal acidification via the protonation of PsbS (Li et al. 2000) and it has been recently proposed that PGR5 deficiency leads to a more facile proton efflux from the lumen though ATP synthase (Avenson et al. 2005), it would be tempting to attribute faster relaxation of qE to a faster rate of proton efflux from the lumen in pgr5. However, this proposal is unlikely as the time scale of seconds for qE relaxation is largely slower than the time scale of milliseconds for the relaxation of bulk luminal acidification.

Growth phenotype in pgr5

Growth and CO₂ assimilation in pgr5 were similar to those in the wild type at low irradiance (50 μmol photons m⁻² s⁻¹) in air, whereas at higher irradiances >150 μmol photons m⁻² s⁻¹, growth and CO₂ assimilation in pgr5 were lower than those in the wild type. Since neither growth retardation (Fig. 7) nor inhibition of CO₂ assimilation (not shown) was observed in npq4, such effects are clearly not related to the impairment of NPQ, but specifically result from the impairment of cyclic electron transport. Interestingly, the growth reduction observed in pgr5 was dependent on the CO₂ supply, suppressed at high CO₂ (completely at 150 μmol photons m⁻² s⁻¹ irradiance and partially at 300 μmol photons m⁻² s⁻¹) and enhanced at low CO₂ (Figs. 6, 7). In a simple interpretation scheme, it is tempting to propose that such an effect of CO₂ supply on growth of pgr5 results from variations in the activity of photorespiration and correspondingly in the variable metabolic demand for ATP. The ATP/NADPH ratio required for steady-state CO₂ assimilation at the CO₂ compensation point, in ambient and in non-photorespiratory conditions is calculated as 1.66, 1.55 and 1.5, respectively (Osmond 1981). Assuming that electron flux for nitrate assimilation into glutamate contributes to 10% of total electron flux for photosynthesis (see Noctor and Foyer 1998, Kramer et al. 2004), the required ATP/NADPH ratio at the CO₂ compensation point, in ambient and in non-photorespiratory condition drops to 1.51, 1.42 and 1.37, respectively. From the predicted ATP/NADPH production ratio of 1.29 by linear electron flux alone, alternative ATP production is required to fulfill the deficit of the ATP/NADPH ratio for CO₂ assimilation (14.8, 9.1 and 5.8%, respectively). During steady-state photosynthesis, a deficit of the proton motive force has been reported in pgr5 compared with the wild type (∼13% of the total light-induced proton motive force according to Avenson et al. 2005). In this context, the reduction of CO₂ assimilation observed in pgr5 at high irradiance (>200 μmol photons m⁻² s⁻¹) can be explained by a reduction of ATP supply resulting from the proton motive force deficit. In pgr5, impaired cyclic electron transport would create a situation where the ATP requirement for photosynthesis is not fulfilled, leading to an inhibition of CO₂ assimilation and, consequently, to a reduction of growth.

However, this simple interpretation cannot explain all of our experimental data. The inhibition of CO₂ assimilation in pgr5 relative to the wild type is weakly dependent on the oxygen concentration and therefore on photorespiration (Figs. 1, 2). A complementary explanation would be that photoinhibitory effects including PSI inactivation contribute to growth retardation when pgr5 is exposed to either high irradiance or low CO₂. The PSI inactivation corresponded to a decrease of steady-state P700 oxidation as we observed in pgr5 grown at 300 μmol photons m⁻² s⁻¹ and at 150 μmol photons m⁻² s⁻¹ in low CO₂ (Figs. 6, 7). PSI photoinhibition has previously been observed to occur in pgr5 when plants were transferred from low irradiance to high irradiance (1,500 μmol photons m⁻² s⁻¹) (Munekage et al. 2002), but here it was observed to occur even at 300 μmol photons m⁻² s⁻¹ (Fig. 5). In wild-type plants, PSI photoinhibition is reported only for chilling conditions in the light, and is characterized by a slow recovery compared with PSII photoinhibition (Havaux and Davaud 1994, Terashima et al. 1994, Sonoike 1996). This phenomenon was attributed to a disequilibrium between reductants generated at PSI and their utilization which triggers PSI photoinhibition by increasing the lifetime of reduced electron carriers on the acceptor side of PSI, such as F_A/B⁻ and F_X⁻, such species being able to react with H₂O₂, producing the toxic hydroxyl radical (Sonoike 1996). A similar phenomenon is considered to occur in pgr5, i.e. an over-reduction of the PSI acceptor side triggers PSI photoinhibition, as we discussed previously (Munekage et al. 2002).

In conclusion, photosynthesis and growth in pgr5 are potentially limited by two phenomena: (i) ATP deficiency that would reduce photosynthesis and (ii) an inhibition of the linear electron transport activity due to PSI inactivation, which is slowly reversible within >1 d. We propose the following scenario to explain the inhibition of CO₂ assimilation and growth reduction occurring in pgr5. In the absence of PGR5, alternative mechanisms (such as the NDH pathway) may compensate ATP deficiency to a certain extent, but, in conditions promoting higher ATP demand, the impairment of ATP supply results in a significant inhibition of photosynthesis. An involvement of the NDH pathway is supported by the phenotype of the double mutant pgr5 crr2, which is impaired in both PGR5- and NDH-dependent cyclic electron pathways and shows drastic growth reduction and ETR impairment at low irradiances (Munekage et al. 2004). Moreover, it is likely that the high redox status of PSI acceptors in pgr5, that results from the unbalanced consumption of ATP and NADPH, leads to the enhanced generation of reactive oxygen species (ROS). In conditions where the generation of ROS exceeds the capacity of the detoxification processes, as in the case of chilling stress, PSI is photoinactivated (Asada 1996, Terashima et al. 1998). This effect would be enhanced in pgr5 depending on the activity of photorespiration due to the higher requirement for ATP than for NADPH in this process. However, it is important to note that under growth conditions where photosynthesis is mainly limited by irradiance (Fig. 6, in air), a marked growth reduction without significant PSI photoinactivation can occur in pgr5. This result suggests that PGR5 deficiency may induce other effects leading to growth reduction that cannot be explained within the frame of this study and requires further investigation.

Materials and Methods

Plant material

Arabidopsis thaliana wild type (ecotype Columbia gl1) and pgr5 (Munekage et al. 2002) were pot grown on soil for 9–10 weeks in a phytotron (50 μmol photons m⁻² s⁻¹, 8 h light/16 h dark at 23°C). For growth analysis experiments, the wild type, pgr5 and npq4 (Li et al. 2000) were grown for 45 d in CO₂-controlled growth chambers (Fabreguettes et al. 1994) at 150 or 300 μmol photons m⁻² s⁻¹, 8 h light/16 h dark at 23°C.

CO₂ assimilation and chlorophyll fluorescence measurements

CO₂ assimilation and chlorophyll fluorescence were simultaneously measured by using a LI6400 equipped with leaf chamber fluorometer LI6400-40 (LI-COR, Lincoln, NE, USA) in detached leaves fed with water. CO₂ gas exchange was performed at 25°C, 50% relative humidity. Oxygen concentration was controlled by mixing N₂ gas and air using mass flow mixing systems. Saturating flashes of red light-emitting diodes (LEDs; >7,000 μmol photons m⁻² s⁻¹, 0.8 s duration) were applied to determine the maximum chlorophyll fluorescence at closed PSII centers in the dark (F_m) and during actinic light illumination (F_m′). The steady-state chlorophyll fluorescence level (F_s) was recorded during actinic light illumination provided by 90% red LEDs plus 10% blue LEDs. Maximal (F_v/F_m) and steady state PSII (Φ_PSII) efficiencies were calculated by (F_m − F_o)/F_m and (F_m′ − F_s)/F_m′, respectively (Genty et al. 1989). The ETR was estimated as Φ_PSII × irradiance (μmol m⁻² s⁻¹) × α × 0.5, where α is fraction of incident light absorbed by the leaf measured. NPQ and the reduction level of Q_A (1 − q_p) were estimated as (F_m − F_m′)/F_m′ following Stern–Volmer expression and (F_s − F_o′)/(F_m′ − F_o′) (Dietz et al. 1985), respectively.

NPQ induction and relaxation experiments, were performed in low CO₂ air (70 μl l⁻¹) using 500 μmol photons m⁻² s⁻¹ actinic light (90% red LEDs plus 10% blue LEDs). The half reversion time (t_1/2) of NPQ, calculated as (F_m − F_m′)/F_m′, was obtained by supplying a train of saturating red LEDs pulse (0.4 s duration).

For CO₂ response measurements, we performed the experiments with lower CO₂ (from 350 to 75 μl l⁻¹) and higher CO₂ (from 350 to 1,500 μl l⁻¹) ranges with different leaves within 1 h per leaf. In each CO₂ range, leaves were exposed for 5–10 min.

Leaf absorptance and chlorophyll content

Leaf absorptance was measured using a Taylor integrating sphere. LED red light similar to the one used for gas exchange and chlorophyll fluorescence measurements was used as a light source, and a silicon photodiode was used as a sensor. Leaf absorptance (α) is calculated as 1 − T − R (T, transmittance; R, reflectance). Chl a and b contents were determined according to Arnon (1949).

Absorbance change of 820 nm

Redox changes of P700 were measured by monitoring the absorbance change at 820 nm with a dual-wavelength pulse-modulation system, ED-P700DW (Heinz-Walz, Effeltrich, Germany) combined with PAM101 (Heinz-Walz, Effeltrich, Germany) as previously described (Klughammer and Schreiber 1998). ΔA was recorded under low irradiance using red LEDs (650 nm, 50 μmol photons m⁻² s⁻¹) supplemented with far-red light (720 nm, 26 Wm⁻²). The maximum oxidation of P700 (ΔA_max) was measured in the presence of strong far-red light (720 nm, 100 Wm⁻²). For the data in Fig. 5, a saturating blue green light flash (20,000 μmol photons m⁻² s⁻¹, 50 ms duration, rise time <1 ms) has been used in the presence of the strong far-red light to ensure maximal oxidation of P700 (Klughammer and Schreiber 1994).

Funding

The Human Frontier Science Program Organization (Y.N.M.).

Acknowledgments

The authors thank the GRAP team (CEA Cadarache) for support in the management of controlled growth chambers.

Abbreviations:

Abbreviations:
 
• LED

  light-emitting diode

 
• NDH

  NAD(P)H dehydrogenase

 
• PGR5

  PROTON GRADIENT REGULATION 5

 
• ROS

  reactive oxygen species

 
• RuBP

  ribulose-1,5-bisphosphate.

Author notes

²Present address: Graduate School of Biological Sciences, Nara Institute of Science and Technology, 8916-5 Takayama, Ikoma, 630-0192 Japan.