- © 2017 American Society of Plant Biologists. All Rights Reserved.
Abstract
During oxygenic photosynthesis, the reducing power generated by light energy conversion is mainly used to reduce carbon dioxide. In bacteria and archae, flavodiiron (Flv) proteins catalyze O2 or NO reduction, thus protecting cells against oxidative or nitrosative stress. These proteins are found in cyanobacteria, mosses, and microalgae, but have been lost in angiosperms. Here, we used chlorophyll fluorescence and oxygen exchange measurement using [18O]-labeled O2 and a membrane inlet mass spectrometer to characterize Chlamydomonas reinhardtii flvB insertion mutants devoid of both FlvB and FlvA proteins. We show that Flv proteins are involved in a photo-dependent electron flow to oxygen, which drives most of the photosynthetic electron flow during the induction of photosynthesis. As a consequence, the chlorophyll fluorescence patterns are strongly affected in flvB mutants during a light transient, showing a lower PSII operating yield and a slower nonphotochemical quenching induction. Photoautotrophic growth of flvB mutants was indistinguishable from the wild type under constant light, but severely impaired under fluctuating light due to PSI photo damage. Remarkably, net photosynthesis of flv mutants was higher than in the wild type during the initial hour of a fluctuating light regime, but this advantage vanished under long-term exposure, and turned into PSI photo damage, thus explaining the marked growth retardation observed in these conditions. We conclude that the C. reinhardtii Flv participates in a Mehler-like reduction of O2, which drives a large part of the photosynthetic electron flow during a light transient and is thus critical for growth under fluctuating light regimes.
Oxygenic photosynthesis, by reducing CO2 into biomass and using water as an electron donor, is responsible for the major entry of carbon into ecosystems. During oxygenic photosynthesis, electrons originating from water splitting at PSII are transferred to PSI, which reduces NADP+ into NADPH at its acceptor side. Electron transfer reactions generate as well a proton motive force (pmf) that drives the ATP synthesis. ATP and NADPH are then used to fuel the CO2 photoreduction cycle. Although the main flow of electrons generated by oxygenic photosynthesis is used for CO2 assimilation, it was early recognized that a significant part of electrons is diverted to molecular oxygen (Mehler, 1951; Radmer and Kok, 1976). Different mechanisms of light-dependent O2 consumption have been described in the chloroplast, including direct O2 photoreduction at the PSI acceptor side (also called Mehler reactions), the oxygenase activity of Rubisco (also called photorespiration), or the reduction of O2 by the plastidial terminal oxidase PTOX (also called chlororespiration). The respective contribution of these different pathways has been a matter of debate and may considerably vary depending on the experimental conditions and according to the organisms considered (Badger et al., 2000). The physiological function of the different O2 photoreduction pathways has been controversial, since they have been alternatively viewed as futile pathways resulting in a waste of energy, as protective mechanisms avoiding photooxidative damage, or as a mean to (re)-equilibrate the balance between reducing and phosphorylating powers through pseudocyclic photophosphorylations (Badger, 1985; Allen, 2003). In cyanobacteria, flavodiiron proteins catalyze NADPH-dependent reduction of O2 at the PSI acceptor side (Vicente et al., 2002; Helman et al., 2003; Allahverdiyeva et al., 2013). The Synechocystis sp. PCC 6803 genome harbors four Flv genes, two of them (Flv1 and Flv3) encoding proteins that form a functional heterodimer catalyzing O2 photoreduction (Helman et al., 2003) and allowing growth under fluctuating light (Allahverdiyeva et al., 2011). The two others (Flv2 and Flv4) code for proteins protecting PSII from over-reduction (Zhang et al., 2009, 2012) by using a yet-unidentified electron acceptor (Allahverdiyeva et al., 2015). Flv genes have been conserved in microalgae, mosses, and gymnosperms, but are notably absent from angiosperms genomes.
In Arabidopsis (Arabidopsis thaliana), the capacity to grow under fluctuating light depends on a strict regulation of the photosynthetic electron flow by the proton gradient, so-called photosynthetic control, which depends on the activity of cyclic electron flow (CEF) around PSI mediated by PGR5 and PGRL1 (Munekage et al., 2002; DalCorso et al., 2008; Suorsa et al., 2012). In cyanobacteria, the PGR5/PGRL1-dependent CEF is not functional, and it was proposed that growth under fluctuating light relies on the presence of Flv1 and Flv3 that would act as an electron sink preventing overreduction of PSI acceptors and production of reactive oxygen species (Allahverdiyeva et al., 2013). Algae and mosses harbor both Flvs and PGRL1/PGR5 CEF components (Tolleter et al., 2011; Dang et al., 2014; Johnson et al., 2014; Gerotto et al., 2016). Based on the accumulation of Flv proteins in a Chlamydomonas reinhardtii mutant (pgrl1) lacking photosynthetic control due to a deficiency in CEF, it was suggested that Flvs are induced to catalyze O2 photoreduction when PSI acceptors are overreduced (Dang et al., 2014). More recently, Flv mutants have been characterized in the moss Physcomitrella patens (Gerotto et al., 2016) and in the liverwort Marchantia (Shimakawa et al., 2017), showing that these proteins are involved in alternative electron flow and are required for growth of the moss under fluctuating light. However, so far there is no experimental evidence for a Flv-mediated oxygen photoreduction in eukaryotic microalgae (Curien et al., 2016).
With the aim to better understand the function of Flvs in microalgal photosynthesis, and particularly to determine whether oxygen is the electron acceptor of Flv and what part of the electron flow is diverted toward oxygen photoreduction during photosynthesis, we study here C. reinhardtii flvB insertion mutants of the Chlamydomonas Library Project (CLiP) library (Li et al., 2016) devoid of both FlvB and FlvA proteins. By performing photosynthetic gas exchange measurements using [18O]-labeled O2 and a membrane inlet mass spectrometer (MIMS), we show that O2 photoreduction is strongly decreased in flv mutants, particularly during the induction phase of photosynthesis. We conclude that FlvA and FlvB proteins are involved in a Mehler-like O2 photoreduction, which massively drives electrons to O2 during the induction phase of photosynthesis.
RESULTS
Identification and Preliminary Characterization of Four Independent C. reinhardtii flvB Mutants
To investigate the function of Flvs in the unicellular alga C. reinhardtii, we searched for insertion mutants in the Flvb (Cre16.g691800) gene region in the CLiP library (https://www.chlamylibrary.org; Li et al., 2016). We obtained four putative flvB mutants, three holding insertion of the paromomycin resistance cassette in introns and one with a possible large deletion (Fig. 1A; Supplemental Table S1). All strains grew normally on Tris-acetate-phosphate or minimal medium under constant low light (40 µmol photons m−2 s−1). Insertion was confirmed by performing PCR on genomic DNA of flvB-14, flvB-208, and flvB-308 (Supplemental Fig. S1). We did not, however, succeed in mapping the insertion in the putative flvB-21 mutant, possibly due to atypical genomic rearrangement(s) that may occur with cassette insertion (Li et al., 2016). Immunodetection was then performed on total whole-cell protein extracts by using antibodies directed against recombinant FlvB and FlvA proteins (Fig. 1, B and C). FlvB and FlvA were not detected in the three independent flvB-21, flvB-208, and flvB-308 strains, while a low level was detected in flvB-14 (hereafter called knockdown mutant). The difference in protein accumulation observed in flvB-208 and flvB-14 is surprising, since insertions are located in the same intron (Fig. 1A). This might be due to a differential effect on intron splicing depending on the location of the insertion in the intron. Similar patterns of accumulation were previously observed for the Flv1 protein in the Synechocystis flv3 mutant (Allahverdiyeva et al., 2013) and for the FlvA protein in the P. patens flvb mutant (Gerotto et al., 2016), which were interpreted by the existence of a functional heterodimer between Flv1 and Flv3 in Synechocystis and between FlvA and FlvB in P. patens.
Characterization of independent C. reinhardtii mutants carrying an insertion in the flvB gene. A, Four mutant strains from the CLiP (www.chlamylibrary.org/) harboring putative insertions of the paromomycin resistance cassette in the FLVB locus (Cre16.g691800) were characterized. Exons, introns, and untranslated regions are shown as red boxes, black lines, and gray boxes, respectively. Genomic sequences flanking the insertion cassette of the four putative flvB mutant strains obtained from the CLiP Web site were used to design primers (F1/R2 and F8/R9) to confirm the location of insertion. B and C, Immunoanalysis of FlvB and FlvA protein amounts were carried out using antibodies produced against recombinant FlvB (B) and FlvA (C) proteins, respectively. A cytochrome f antibody was used as loading control.
Chlorophyll Fluorescence Transients Are Strongly Affected under High Light in flvB Mutants
To determine how electron transfer reactions are affected in flvB mutants, we recorded chlorophyll fluorescence transients under different light intensities. Under low actinic light (25 µmol photons m−2 s−1), no difference could be recorded between flvB mutants and the CC-4533 strain, hereafter referred to as the wild type (Fig. 2A). However, upon exposure to higher light intensities (100 or 500 µmol photons m−2 s−1), a marked difference was observed between flvB mutants and the control wild-type strain, the fluorescence signal transiently reaching a much higher value in the mutants than in the wild type (Fig. 2, B and C). The PSII operating yield measured after 20 s of actinic illumination was much lower in flvB mutants than in the wild type, but the difference partially vanished after 5 min of light exposure (Fig. 3). A smaller effect was observed in flvB-14, which accumulates low levels of FlvB and FlvA (Fig. 2B; Supplemental Fig. S2). Interestingly, relaxation of chlorophyll fluorescence to a steady-state level after a saturating flash was slower in the flvB mutants than in the wild-type strain (Supplemental Fig. S3), indicating that Flv proteins are functional during brief light transients. We conclude from these chlorophyll fluorescence measurements that Flv proteins are involved in photosynthetic electron transfer reactions, particularly during the induction phase of photosynthesis under moderate or high illumination.
Chlorophyll fluorescence measurements during dark to light transients in C. reinhardtii wild type (WT) and flvB mutant strains. Cells were grown in HSM medium under low light (40 µmol photon m−2 s−1) and harvested during exponential phase. Chlorophyll fluorescence measurements were performed using pulse amplitude modulated fluorimeter in the dark (black boxes) and under red actinic light (white boxes) of different intensities: 25 µmol photon m−2 s−1 (A), 100 µmol photon m−2 s−1 (B), and 500 µmol photon m−2 s−1 (C). Saturating flashes were supplied when indicated by red vertical lines. Data are normalized on initial FM measurements and traces of mutant strains are shifted few seconds to the right for clarity.
Light dependence of the PSII yield in the wild type (WT) and flvB mutant strains. Upon 10-min dark adaptation under constant air bubbling, algal suspensions of wild type and flvB mutant strains were introduced in the cuvette of a pulse amplitude-modulated fluorimeter. PSII yields were measured after 20 s (dotted lines) or 5 min (plain lines) of illumination at different light intensities. Shown are the mean values (±sd, n = 3).
Flv Proteins Interact with the Photosynthetic Electron Transport Chain at the Level of PSI Acceptors
To better characterize the site of interaction between Flv proteins and the photosynthetic electron transport chain, we first used DBMIB, a photosynthetic electron flow inhibitor acting at the Q0 site of the cytochrome b6f complex. Upon addition of DBMIB, the chlorophyll fluorescence level rapidly rose to the maximal fluorescence level (FM) with similar light induction curves in the wild type and flvB mutants (Fig. 4), thus indicating that Flv proteins interact with the photosynthetic electron transport chain downstream the cytochrome b6f. We then used oxidized methyl viologen (MV), which efficiently accepts electrons at the PSI acceptor side. The chlorophyll fluorescence induction curve of the wild type was only slightly affected by MV (Fig. 4). In contrast, the strong chlorophyll fluorescence increase observed in the flvB mutant was suppressed by the MV treatment, thus indicating that the site of action of Flv is located downstream the action site of MV, which is at the PSI acceptor side level. Altogether, these observations also show that Flv proteins are active in highly reducing conditions and that its activity is decreased when the reducing pressure is low.
Effect of DBMIB and MV on chlorophyll fluorescence transients measured in wild type (WT) and flvB mutant strains. Samples were placed in the dark (black box) in the presence of 1 µm DBMIB, 1 mm MV, or without addition of any chemicals. Chlorophyll fluorescence was measured during a transient dark to light (100 µmol photon m−2 s−1) transient. Saturating flashes were supplied when indicated by red vertical lines. Shown are the traces representative of n = 2 independent measurements. Fluorescence data have been normalized on F0.
Oxygen Photoreduction Is Strongly Reduced in flvB Mutants
We then measured O2 exchange during light transients using a MIMS and [18O]-labeled O2 (Fig. 5). This technique allows discriminating unidirectional fluxes of O2 in the light: i, gross O2 evolution, which represents O2 produced by PSII, and ii, O2 uptake in the light, which consists in different mechanisms such as O2 photoreduction and mitochondrial respiration. In such experiments, O2-consuming processes take up all O2 species present in the medium including [18O]-labeled O2, while photosynthesis will essentially produce nonlabeled O2 from water-splitting at PSII. When light was switched on, the O2 uptake rate strongly increased in the wild type, and then progressively declined during the light period (Fig. 5A). In the flvB-21 mutant, the O2 uptake rate measured in the light was much lower than in the wild type and remained constant throughout the light period (Fig. 5B), thus showing that Flv proteins are involved in O2 photoreduction. The O2 uptake rate remained constant throughout the light period in flvB mutants and was close to the O2 exchange rate measured in the dark, which is mainly due in Chlamydomonas to persistent mitochondrial respiration (Peltier and Thibault, 1985b). Therefore, the large O2 uptake component observed in the wild type during the first minutes of illumination is mainly due to the activity of Flv proteins. Gross O2 evolution was also reduced in the flvB-21 mutant as compared to the wild type, indicating that in the wild type Flv-mediated O2 photoreduction promotes an extra-electron flow from PSII to O2. Similar effects were observed in the three independent flvB mutants (flvB-21, flvB-208, and flvB-308), while an intermediary effect was observed in the flvB-14 knockdown strain (Fig. 5, C and D). Dark O2 uptake rates measured in the different strains before and after the illumination period did not show significant differences (Supplemental Fig. S4). The net O2 evolution rate measured after 5 min of illumination, which reflects net photosynthesis due to CO2 reduction, was not much affected in the mutant as compared to the wild type (Fig. 5D). This shows that Flv-dependent O2 photoreduction does not efficiently compete with CO2 fixation during steady-state photosynthesis. Note that after 5 min of illumination, O2 uptake rates remained slightly higher in the wild type as compared to flvB mutant strains (Fig. 5D), indicating that the Flv-mediated electron flow to O2 also contributes under steady-state illumination. We conclude from these experiments that O2 photoreduction primes and replaces CO2 fixation during the induction phase of photosynthesis in C. reinhardtii, as was previously reported in Scenedesmus obliquus and Anacystis nidulans (Radmer and Kok, 1976), and we further establish that Flv proteins are involved in this phenomenon.
Oxygen exchange measurements performed using a MIMS and 18O-labeled O2 in the wild type (WT) and flvB mutant strains. A and B, Gross O2 evolution (black dots), O2 uptake (red dots), and net O2 production (blue dots) were determined in the wild type (A) and the flvB-21 mutant (B). C, Mean values of gross O2 evolution (dark gray boxes) and O2 uptake (red boxes) measured during the first minute of illumination (shown are means ± sd, n = 3) in the wild type, and in flvB-14, flvB-21, flvB-208, and flvB-308 mutant strains. D, Mean values of gross O2 evolution (dark gray boxes), O2 uptake (red boxes), and net O2 evolution (blue boxes) measured between 4 and 5 min of illumination (shown are means ± sd, n = 3). Values of O2 uptake in the dark prior to illumination (light gray boxes) are shown for the wild type and flvB-208 (shown are means ± sd, n = 3) (C and D). Letters above the bars (C and D) represent statistically significant differences (P value < 0.05) between strains for a given gas exchange measurement based on ANOVA analysis (Tukey adjusted P value); letters with single or double quote have been used to group oxygen uptake or net evolution data respectively.
Growth of flvB Mutants Is Delayed under Fluctuating Light Conditions Due to PSI Impairment
The strong effect observed in flvB mutants during the induction phase of photosynthesis and the growth delay previously reported in ∆flv cyanobacterial mutants grown under fluctuating light (Allahverdiyeva et al., 2011) prompted us to analyze growth performances under repeated light changes. When C. reinhardtii cells were grown under 5 min low light (50 µmol photons m−2 s−1) and 1 min high light (500 µmol photons m−2 s−1), a strong growth retardation was observed in the three flvB mutant strains (flvB-21, flvB-208, and flvB-308) as compared to the wild type, while an intermediary effect was observed in the flvB-14 strain (Fig. 6A). This drastic effect on growth did not, however, lead to cell death, since growth recovery was observed during a subsequent growth period under continuous light (Supplemental Fig. S5). In contrast, no growth difference could be observed between mutant and wild-type strains under constant illumination, even under high light intensity (800 µmol photons m−2 s−1; Fig. 6A). To understand the mechanism underlying growth retardation, we performed PSII and PSI activity measurements at different time points during exposure to fluctuating light conditions (Fig. 6B). Maximal PSII activity was slightly decreased (by about 20%) in the flvB mutant upon 48 h of fluctuating light exposure, but a drastic decrease of the PSI activity (about 80%) was observed in the mutant (Fig. 6B). Immunoanalysis of PSII and PSI subunits, respectively PsbA and PsaD, revealed no major change in the PsbA subunit amount, but a large decrease in PsaD in the flvB mutant upon 24 h of exposure to fluctuating light (Fig. 6C). We conclude from this experiment that repeated transitions from low to high light induce PSI photoinhibition in the absence of Flv-mediated electron flow.
Growth and PSI activity of flvB mutants are impaired under fluctuating light. A, Cells were spotted on HSM agar plates and exposed for 10 d to continuous medium (125 µmol photon m−2 s−1), high (800 µmol photon m−2 s−1), or to fluctuating light intensity (alternatively 5 min at 50 µmol photon m−2 s−1 and 1 min at 500 µmol photon m−2 s−1). Shown are representative drops of three independent spot tests. B, PSII and PSI activity measured in response to fluctuating light exposure. Wild-type (WT) and flvB-308 cells were grown photoautotrophically in flasks under low light (50 µmol photon m−2 s−1) and exposed at t0 to fluctuating light cycles (5 min at 50 µmol photon m−2 s−1 and 1 min at 500 µmol photon m−2 s−1) for 48 h. Maximal PSII yields and maximal oxidizable PSI were respectively determined by means of chlorophyll fluorescence and P700 absorption changes measurements PSII. C, Immunoanalysis of PSII and PSI subunit amounts in the wild type and flvB-308 mutant in response to fluctuating light exposure. PsbA and PsaD antibodies were used to probe amounts of PSII and PSI complexes, respectively. AtpB antibody was used as loading control.
Defect in Flv Has Ambivalent Effects on Photosynthesis
To better characterize the function of Flv proteins during photosynthesis, we measured photosynthetic O2 exchange in cells during exposure to a fluctuating light regime (Fig. 7). Net photosynthesis was higher in flvB mutant cells during the first 30 min of a fluctuating light exposure as compared to the wild type (Fig. 7A), but the difference vanished after 1 h, and a negative effect was observed upon 4 h of fluctuating light exposure (Fig. 7C), in accordance with the inhibition of PSI (Fig. 6). In contrast, the O2 uptake rate measured in the light was initially much lower in the flvB mutant than in the wild type (Figs. 5, A and B, and 7, B and D), but progressively increased in the mutant until reaching a similar level as in the wild type (Fig. 7D). This indicates that during the time-course of an exposure to a fluctuating light regime, a Flv-independent O2 uptake process was progressively triggered in the mutant. Therefore, depending on the duration of a fluctuating light exposure, the loss of Flv proteins leads either to a positive or a negative effect on net photosynthesis, the positive effect being accompanied by a decrease in the O2 uptake (mediated by Flv) and the negative effect being accompanied by a progressive increase in an O2 uptake component. The induction of this Flv-independent, light-dependent O2 uptake phenomenon likely results in ROS production, thus explaining growth retardation and PSI photoinhibition observed under long-term exposure (Fig. 6).
Net O2 exchange measurements under fluctuating light in flvB mutants and in wild-type (WT) C. reinhardtii strains. Cells were grown photoautotrophically in air under continuous light (50 µmol photon m−2 s−1) and then exposed up to 4 h to a fluctuating light regime (1 min at 50 µmol photon m−2 s−1, 5 min at 500 µmol photon m-2 s−1) before performing O2 exchange measurements using a MIMS under similar fluctuating light conditions. A and B, Representative traces of net O2 evolution rate (A) and O2 uptake rate in the light (B) measured in the flvB-21 mutant and wild type under fluctuating light in cells previously cultivated under continuous light. C and D, Mean values of net O2 evolution (C) and O2 uptake rates in the light (D) measured in fluctuating light in flvB-21, flvB-308, and wild-type strains. Shown are means (±sd, n = 3 for t = 0 h, n = 2 for t > 0 h) of net O2 evolution rates relative to the wild type (at t = 0) and measured during the 5th light fluctuation period (as indicated in A and B by vertical dotted lines) in the wild type (red boxes) and flvB mutants (gray boxes). Letters (a and b) above the bars (C and D) represent statistically significant differences (P value < 0.05) between strains for a given gas exchange measurement based on ANOVA analysis; letters with simple or double quote have been used to group oxygen exchange data after 1 and 4 h or fluctuating light, respectively.
Nonphotochemical Quenching Induction Is Affected in flvB Mutant Strains
To better understand the impact of the Flv-mediated electron flow to O2 on photosynthesis, we measured the nonphotochemical quenching (NPQ) induction in high light-adapted cells (Fig. 8A). During a dark to light transient, NPQ rapidly rose in the wild type to reach a steady-state level. NPQ induction was slower in the flvB mutant, but reached a similar level after 1 to 2 min of illumination (Fig. 8A). In C. reinhardtii, NPQ relies on two factors: accumulation of the LHCSR3 protein and lumen acidification (Peers et al., 2009; Bonente et al., 2011). Since both wild-type and flvB mutant strains accumulated similar levels of the LHCSR3 protein (Fig. 8B), the slower induction of NPQ is likely due to a slower acidification of the lumen in flvB mutants as compared to the wild type. To further investigate effects on the proton gradient, we measured electrochromic carotenoid shift (ECS). In the flvB mutant, the pmf was decreased by one third and the ΔpH component by ∼50% as compared to the wild type (Fig. 8C), the ratio between ΔΨ and ΔpH components being unaffected (Supplemental Fig. S6). Moreover, time constant of ECS decay indicates that membrane conductivity to protons (gH+) is not altered in the flvB mutant, whereas proton flux (νH+ = pmf * gH+) is lower (Fig. 8D), suggesting a lower rate of ATP synthesis in the mutant during the induction of photosynthesis. We conclude from this experiment that the photosynthetic electron flow to O2 mediated by Flv proteins participates in the wild type to the establishment of pmf that allows fast induction of the NPQ and higher ATP production during a light transient (Fig. 9A).
NPQ induction and proton motive force (pmf) measurements in the wild type (WT) and flvB-21 mutant. A, For NPQ measurements, cells were exposed to 200 µmol photon m−2 s−1 for 4 h to induce accumulation of the LHCSR3 prior to chlorophyll fluorescence measurements. The wild type and flvB-21 mutant were then exposed to 500 µmol photon m−2 s−1 in the PAM cuvette (white box) and NPQ determined from chlorophyll fluorescence measurements. NPQ values are the mean (±sd, n = 3) in the wild type (red symbols) and flvB-21 mutant (black symbols). B, Immunodetection of LHCSR3 protein in experimental conditions as described in (A) in wild type and different flvB mutants. C, Different components of the pmf (ΔΨ and ΔpH) were determined from ECS measurements in wild type (red) and flvB-21 mutant (gray) from similar experiments as described in (Supplemental Fig. S5A). D, Membrane proton conductivity (gH+) and proton flow (νH+) were determined from ECS measurements. Numbers above bars (C and D) represent uncorrected P values as determined by ANOVA using Fischer’s lsd.
DISCUSSION
Flv Proteins Are Involved in O2 Photoreduction during a Light Transient
By characterizing four independent insertion mutants of C. reinhardtii in the Flvb gene region affected in the accumulation of FlvB and FlvA proteins, we show here that Flv are involved in O2 photoreduction at the acceptor side of PSI. Forty years ago, Radmer and Kok used [18O]-labeled O2 and mass spectrometry to show that in Scenedesmus and Anacystis, O2 photoreduction primes and replaces CO2 fixation during the induction of photosynthesis (Radmer and Kok, 1976), but mechanisms of O2 photoreduction have remained elusive for a long time. Later studies showed that different mechanisms contribute to O2 uptake in the light. Under steady-state conditions, persistent mitochondrial respiration in the light (Peltier and Thibault, 1985b) and a small contribution of photorespiration could be detected in C. reinhardtii (Peltier and Thibault, 1985a), the latter operating in microalgae at a much lower rate than in C3 plants (Badger et al., 2000). We have shown here that Flv-mediated O2 photoreduction drives most of the photosynthetic electron flow during the induction phase of photosynthesis in wild-type strains, when CO2 assimilation has not started yet. Flv activity appears to be restricted to highly reducing conditions, which might result from a lower affinity to NADPH than the Calvin cycle enzymes, as recently proposed (Shikanai and Yamamoto, 2017).
Chlamydomonas FlvA/FlvB Proteins Function in a Similar Manner as Cyanobacterial Flv1/Flv3
Phylogenetic analysis of the Flv family has shown that algal and mosses FlvB proteins are homologous to cyanobacterial Flv3 and Flv4, while FlvA proteins are homologous to cyanobacterial Flv1 and Flv2 (Zhang et al., 2009; Peltier et al., 2010). In Synechocystis PCC6803, Flv1 and Flv3 have been proposed to catalyze O2 photoreduction into water without generation of ROS, and protect PSI from inhibition under fluctuating light (Allahverdiyeva et al., 2013). Based on O2 photoreduction measurements in ∆flv1 and ∆flv3 cyanobacterial mutants, it was concluded that Flv3 function as a hetero-dimer with Flv1 (Helman et al., 2003; Allahverdiyeva et al., 2011). However, a functional Flv3 homo-oligomer may also be formed, as deduced from the overproduction of Flv3 in Synechocystis (Mustila et al., 2016). On the other hand, Flv2 and Flv4 protect PSII by a different mechanism and using a yet-uncharacterized electron acceptor (Zhang et al., 2012; Bersanini et al., 2014). From our data, we conclude that C. reinhardtii FlvA and FlvB are involved in O2 photoreduction at the PSI acceptor side and therefore likely functions in a similar manner as cyanobacterial Flv1 and Flv3. Because the presence of Flv proteins prevents PSI photodamage, it is most likely that O2 is fully reduced into water without production of ROS, as previously proposed for cyanobacteria (Helman et al., 2003). Based on the fact that FlvA could not be detected in flvB knockout mutants (Fig. 1C), we conclude that C. reinhardtii FlvA and FlvB proteins form in vivo a functional heterodimer, as previously proposed for Flv1 and Flv3 in Synechocystis (Helman et al., 2003; Allahverdiyeva et al., 2013) and FlvA and FlvB in P. patens (Gerotto et al., 2016).
Flv Proteins Are Involved in a Pseudocyclic Electron Flow to O2 and Participates in the Establishment of a pmf
By mediating an electron flow from water to O2 Flv proteins are involved in a pseudocyclic electron flow and therefore participates in the generation of a pmf (Fig. 8, C and D). In addition to its role in ATP synthesis, the pmf triggers two main mechanisms, the qE-type NPQ (LHCSR3-dependent) and the photosynthetic control, which participate in the regulation of the photosynthetic electron flow and protect PSI from acceptor-side limitations and PSI photoinhibition (Chaux et al., 2015, 2017). The CEF also participates in the establishment of a pmf, and its role in triggering both NPQ and the photosynthetic control has been established (Munekage et al., 2002; Joliot and Johnson, 2011; Tolleter et al., 2011; Suorsa et al., 2012). As shown here by the delayed induction of NPQ (Fig. 7A) measured in flvB mutants during a light transient, the Flv-mediated pseudocyclic electron flow to O2 is involved in the rapid establishment of NPQ (Fig. 8A).
In Chlamydomonas, Flv-Mediated Electron Flow to O2 Has Ambivalent Effects on Photosynthesis
Although no effect of Flv deficiency was observed on photosynthesis and growth under continuous illumination, ambivalent effects were observed upon exposure to fluctuating light. During a short-term exposure (<1 h) of fluctuating light, net photosynthesis was enhanced in flvB mutants as compared to the wild type. However, this advantage vanished thereafter, net photosynthesis being strongly decreased in the mutant (Fig. 7C). To understand such an ambivalent effect, we need to consider the different factors involved in the limitation of photosynthesis. Photosynthetic CO2 fixation requires both NADPH and ATP, and a fine tune in the supply of both energy sources is needed for an optimal functioning of photosynthesis (Kramer and Evans, 2011). In other words, photosynthetic CO2 fixation can be either limited by the NADPH supply or by the ATP supply. The Flv-mediated electron flow from PSII to O2 has two effects on photosynthesis, one is to divert electrons toward O2, therefore limiting the NADPH supply and the other to generate pmf, which may either be used to produce ATP and down-regulate the photosynthetic electron flow. Lack of Flv would therefore favor NADPH supply by two means: by avoiding the loss of electrons toward O2 and by lowering the down-regulation of photosynthetic electron flow. A positive effect of a Flv defect on photosynthesis would therefore indicate that photosynthesis is limited by the supply of reducing power (high ATP/NADPH ratio), when cells switch from a continuous to a fluctuating light regime. However, in conditions where ATP is lacking (low ATP/NADPH ratio), the defect in Flv would enhance the disequilibrium by lowering ATP and increasing NADPH.
The ambivalent effect of Flv may therefore reveal the existence of different limitation regimes of photosynthesis (NADPH-limited versus ATP-limited) under fluctuating light conditions, a progressive switch from a NADPH-limited regime toward an ATP-limited regime occurring upon long-term exposure to fluctuating light. In the absence of Flv, cells would produce more NADPH than strictly needed to make photosynthesis working at is optimal regime and would progressively accumulate an excess of reducing power. The 1-h retardation in the appearance of the negative effect would be due to the buffering capacity of intracellular metabolic pools and electron sinks. When the excess of NADPH produced exceeds the buffering capacity, overaccumulation of reducing power would result in the production of ROS and in turn to PSI photoinhibition (Fig. 6, B and C). A similar scenario based on a progressive disequilibrium between NADPH and ATP production was recently proposed to explain the delay between PSI photoinhibition and CO2 when a double C. reinhardtii mutant deficient in CEF and qE was exposed to a sudden light increase (Chaux et al., 2017).
Compensation between Flvs-Mediated O2 Photoreduction and PGRL1/PGR5-Dependent CEF
In the absence of CEF, Arabidopsis pgr5 and pgrl1 mutants are prone to PSI photoinhibition and severe growth retardation under high light (DalCorso et al., 2008; Munekage et al., 2008). The wild-type phenotype was restored in the Arabidopsis pgr5 mutant by the expression of P. patens FlvA and FlvB proteins, thus showing that artificially rewiring the photosynthetic electron flow to O2 through Flvs can compensate deficiency in CEF (Yamamoto et al., 2016). In Synechocystis, which lacks PGR5-PGRL1 CEF, deletion of Flv1 and Flv3 results in a severe decrease (about 60%) in net photosynthesis under constant high light (Allahverdiyeva et al., 2013). This strongly contrasts with C. reinhardtii, where no decrease in net photosynthesis was observed under constant light in flvB mutants (Fig. 5D). This difference may result from a compensation by the PGRL1/PGR5-dependent CEF, which like pseudocyclic electron flow generates extra-pmf and extra-ATP during photosynthesis. Indeed, deficiency in CEF in C. reinhardtii was partially compensated by an increase in Flv-dependent O2 photoreduction (Dang et al., 2014). Therefore, Flv-dependent O2 photoreduction and CEF appear as partially redundant mechanisms in microalgae, both able to supply extra-ATP during steady-state photosynthesis.
Under fluctuating light conditions, growth is severely impaired in C. reinhardtii flvB mutants (Fig. 6), as previously observed in Synechocystis Δflv1 and Δflv3 mutants (Allahverdiyeva et al., 2013). This shows that in C. reinhardtii the PGR5-PGRL1 CEF is not able to compensate for the loss of Flvs under fluctuating light. In contrast, Arabidopsis growth under fluctuating light relies on the PGRL1/PGR5-mediated CEF, which protects of PSI from photoinhibition in the absence of Flvs (Suorsa et al., 2012), indicating that the overall process of CEF may be more efficient in angiosperms than in algae.
CONCLUSION
While the role of Flv in driving O2 photoreduction in a photosynthetic organism has been first demonstrated in cyanobacteria (Helman et al., 2003), recent studies performed in the moss P. patens (Gerotto et al., 2016) and liverwort (Shimakawa et al., 2017), and mainly based on chlorophyll fluorescence measurements, concluded that Flv drive alternative electron flow and protect PSI particularly under fluctuating light. However, there was no experimental evidence until now of an Flv-catalyzed O2 photoreduction in microalgae (Curien et al., 2016). Here, by using [18O]-labeled O2 and MIMS, we clearly establish that Flv are functional in microalgae and involved in O2 photoreduction. Moreover, quantification of O2 photoreduction rates showed that Flv massively drives electrons toward O2 during the induction of photosynthesis. Positive or negative effect of Flv observed on net photosynthesis depending on the fluctuating light conditions, may supply an experimental basis to understand why Flvs were conserved in some species like gymnosperms and algae and discarded in angiosperms. Finally, the loss of Flv in angiosperms could be related to the existence of a more efficient CEF.
MATERIALS AND METHODS
Chlamydomonas Cultures
The Chlamydomonas reinhardtii wilt-type strain CC-4533 and flvB mutants were obtained from the CLiP (Li et al., 2016). Upon reception, strains were plated on Tris-acetate-phosphate medium and streaked until exhaustion. After a 1-week growth in the dark, three single-clone derived colonies were randomly chosen for conservation and subsequent characterization for each strain. For further liquid culture experiment, cells were grown in flasks at 25°C in HSM medium under dim light (30–40 µmol photon m−2 s−1). Unless otherwise stated, experiments presented throughout this manuscript were performed on three single colony-derived lines for the wild type and for each of the four strains carrying insertions in the FLVB gene; thus, SDs account for ses of biological triplicates calculated with Prism (GraphPad Software).
Spot Tests
Cells were harvested during exponential phase and resuspended in fresh HSM to 2, 10, or 20 µg chlorophyll mL−1. Ten-microliter drops were spotted on plates and exposed to different light regimes. Homogeneous light was supplied by a panel of 400 cool-white LEDs (1 cm distant of each other) and placed 33 cm above plates for growth tests. Temperature was maintained at 25°C at the level of plates by means of fans. The LED panel was powered via a capacitor voltage transformer giving a direct current with a variable voltage between 0 and 54 V. Voltage was monitored via an Arduino UNO microcontroller board to obtain the appropriate light fluctuation regimes.
PCR Procedures
Total DNA was extracted using Chelex kit (Sigma-Aldrich) as described (Dang et al., 2014). Putative insertions were confirmed in three of the four strains by PCR using TaKaRa LA Taq DNA polymerase with GC buffer (Clontech). The following set of primers was designed according to Cre16.g691800 gene sequence (Phytozome v5.5; www.phytozome.jgi.doe.gov): F1: 5′-GAGGCATGCGACA-GGCCCGAATTGCAC and R2: 5′-GCACGGCACCATCTCCGACCTAGCC for FLVB first intron region, F8: 5′-GTACCTGGCTGCAGCGTGTTCACTGCC and R9: 5′-CACCTCGC-AGTAGGTGACCCAGTGGTCG for FLVB seventeenth intron region. Cycles were as follows: 2 min at 94°C / 35 cycles: 20 s at 94°C, 20 s at 64°C, 2 min at 72°C/1 min at 72°C. PCR products were separated on 1% (w/v) agarose gels.
Production of FlvB and FlvA Antibodies
Synthetic FLVA and FLVB genes were cloned respectively into the pLIC7 and the Champion pET151 Directional TOPO (Invitrogen) expression vectors, allowing the production of a recombinant FLVA fused to TEV-cleavable His-tagged Escherichia coli thioredoxin and the recombinant FLVB protein fused to a N-ter (His)6. Production was performed in the E. coli BL21 Star (DE3) strain grown at 37°C in TB medium. Induction was initiated at an OD600 = 0.6 by adding 0.5 mm isopropyl β-d-thiogalactoside (Sigma-Aldrich). Following overnight incubation at 25°C, cells were centrifuged and pellets resuspended in a lysis buffer containing an antiprotease inhibitor cocktail (SigmaFast tablet), 0.25 mg mL−1 lysosyme, and 10 µg mL−1 DNaseI. Following incubation (30 min at 4°C), cells were sonicated and centrifuged (12,000g for 30 min at 4°C). Crude protein extracts were loaded on a His-Trap HP column (GE Healthcare) and eluted with 250 mm imidazole. The thioredoxin fused to FLVA was cleaved off with overnight incubation with TEV protease. FLVA and FLVB fractions were loaded onto a HiPrep 26/60 Sephadex S-200 HR size-exclusion column (GE Healthcare) and recovered with 10 mm Tris, pH 8.0, buffer containing 300 mm NaCl. The protein peaks containing recombinant FLVA in one case and the (His)6-FlvB in the other case were controlled for purity on SDS-PAGE and concentrated using a Amicon-Ultra device (Millipore). Polyclonal antibodies against FlvA and FlvB were raised in rabbits (ProteoGenix).
Immunodetection
Cells (10–15 mL) were harvested from liquid cultures and centrifuged at 3,000g for 2 min. Pellets were then frozen in liquid nitrogen and stored at −20°C until use. Pellets were resuspended in 400 µL 1% SDS and then 1.6 mL acetone (−20°C) was added. After overnight incubation at −20°C, samples were centrifuged (14,000 rpm, 10 min). Supernatant was removed and used for chlorophyll quantification using SAFAS UVmc spectrophotometer (SAFAS). Pellets were resuspended to 1 µg chlorophyll mL−1 in LDS in the presence of NuPAGE reducing agent (ThermoFischer) and loaded on 10% PAGE Bis-Tris SDS gel. To load equal protein amounts for immunoblot analysis, protein contents were estimated from Coomassie Brilliant Blue staining using an Odyssey IR Imager (LICOR). After gel electrophoresis, proteins were transferred to nitrocellulose membranes for 90 min at 25 V and 200 mA as described (Dang et al., 2014). Membranes were blocked in milk for 1 h then incubated overnight in the presence of the following antibodies: anti-Cyt f, anti-AtpB, anti-PsaD, anti-PsbA, anti-LHCSR3 (Agrisera), or anti-FLVB (see previous paragraph). After 4°C overnight incubation, primary antibody was removed by several rinsings in TBS, and a peroxidase-coupled secondary antibody was added for at least 1 h before detection with a Gbox imaging system (Syngene).
Chlorophyll Fluorescence Measurements
Chlorophyll fluorescence measurements shown in Figures 2, 4, 6, and 7 were performed using a pulse amplitude-modulated fluorimeter (Dual-PAM 100) upon 5-min dark-adaptation under continuous stirring. Detection pulses (10 µmol photon m−2 s−1 blue light) were supplied at a 100-Hz frequency. Basal fluorescence (F0) was measured in the dark prior to the first saturating flash. Red saturating flashes (6,000 µmol photon m−2 s−1, 600 ms) were delivered to measure FM (in the dark) and FM’ (in the light). PSII maximum yields were calculated as (FM-F0)/FM. In the experiment shown in Figure 4, DBMIB and MV (Sigma-Aldrich) were added 1 min prior to measurements at final concentrations of 1 µm and 1 mm, respectively. For NPQ measurements (Fig. 7), cells were grown in high light (200 µmol photon m−2 s−1) for 24 h to induce LHCSR3 accumulation. Measurements were then performed under 500 µmol photon m−2 s−1 for 10 min and NPQ calculated as (FM−FM’)/FM’.
Chlorophyll fluorescence measurements shown in Figure 3 were performed using a Joliot-type spectrofluorimeter (JTS 10; BioLogic). Cells were harvested from flask cultures, centrifuged (3,000 rpm, 2 min), and resuspended in a 3-mL glass cuvette in a buffer containing 20% Ficoll buffered with 20 mm HEPES (pH 7.2). Upon 10 min in the dark, fluorescence was obtained after nonactinic detection pulses of blue light given in the dark (F0 parameter), during green light saturating flashes (4,000 µmol photon m-2 s−1, 250 ms; FM and FM’ parameters), and during green light illumination at 40, 190, 370, 640, 1,220, and 1,800 µmol photon m−2 s−1 (F parameter). Maximal PSII yield was calculated from (FM−F0)/FM before illumination, and PSII yield for each light intensity was calculated from (FM’−F)/FM’.
Absorption Change Measurements
P700 absorption and carotenoid ECS were measured at 705 and 520 nm, respectively, using a JTS-10 (BioLogic). Cells were harvested from flask cultures, centrifuged at 3,000 rpm for 2 min, and resuspended in a 3-mL glass cuvette in a buffer containing 20% Ficoll buffered with 20 mm HEPES (pH 7.2). For P700 absorption change measurements, PSII was inhibited by addition of 3-(3,4-dichlorophenyl)-1,1-dimethylurea (10 µm final concentration), and PSI acceptor side limitations were prevented by addition of oxidized MV, 10 mm final concentration). Measurements were carried out in the dark to reach a baseline level and in response to saturating red light (1,200 µmol photons m−2 s−1), P700+ was accumulated within 1 to 2 s. Concentration of oxidizable P700 was calculated from steady-state (∼5 s) relative absorbance levels and molar extinction coefficient: (Alight−Adark)/(−50*2.3) (Hiyama and Ke, 1972; Joliot and Joliot, 2005). P700 absorption changes measured in the absence of MV showed similar results (not shown). For ECS measurements, low light-grown cells were dark-adapted for 5 min in an open cuvette manually bubbled with a syringe (aerobic conditions) and rapidly introduced in the JTS sample holder (BioLogic). ECS was measured during 5 s under orange light (1,170 µmol photons m−2 s−1) and normalized to the ECS value measured in response to 6 ns single-turnover flash. Total pmf values (ECSt) were calculated from the difference between the ECS value after 5 s in light and the minimum value obtained 300 to 500 ms after light was switched off (Supplemental Fig. S6A). ΔΨ and ΔpH components of the pmf were calculated according to the difference between ECSt and maximal ECS value 3 to 5 s after the light was switched off. Using Prism (GraphPad Software), data points of the last seconds of illumination and 0 to 500 ms of dark relaxation were fitted to the “Plateau followed by one-phase decay” function, from which time constants were extracted. Membrane proton conductivity gH+ was calculated as the inverse of the time constant and proton flow (νH+) as the product of pmf and gH+ (Cruz et al., 2005).
MIMS Measurements
O2 exchanges were monitored using a water-jacketed, thermoregulated (25°C) reaction vessel coupled to a mass spectrometer (model Prima dB; Thermo Electron) through a membrane inlet system (Tolleter et al., 2011). The cell suspension (1.5 mL) was placed in the reaction vessel, and bicarbonate (4 mm final concentration) was added to reach CO2 saturation. [18O]-enriched O2 (99% 18O2 isotope content; Euriso-Top) was bubbled at the top of the suspension until reaching approximately equal concentrations of 16O2 and 18O2. Upon closure of the reaction vessel, O2 exchanges were measured during a 2-min dark period, then light was switched on (800 μmol photons m−2 s−1). Isotopic O2 species [18O18O] (m/e = 36) and [16O16O] (m/e = 32) were monitored, and O2 exchange rates were determined as described previously (Radmer and Kok, 1976).
Statistical Analysis
Statistical significance was assessed by ANOVA using GraphPad Prism (GraphPad Software). P values were computed by multiple comparison tests using Fischer’s lsd test (uncorrected P values) or using Tukey corrections for large datasets (adjusted P values). Shown are the P values or the grouping of strains into statistical families according to P values. We defined the statistical significance cutoff as 0.05 (5%).
Supplemental Data
The following supplemental materials are available.
Supplemental Table S1. Wild-type and mutant strains from the CLiP used in this study.
Supplemental Figure S1. PCR characterization of the insertion region in the three flvB mutants flvB-14, flvB-208, and flvB-308.
Supplemental Figure S2. Light dependence of PSII yield in wild type and flvB mutant strains.
Supplemental Figure S3. Chlorophyll fluorescence kinetics in response to a saturating flash in the wild type and flvB mutants.
Supplemental Figure S4. Measurements of dark- and light-induced oxygen uptake rates in the wild type and flvB mutant strains.
Supplemental Figure S5. Growth recovery of flvB mutants after long exposure to fluctuating light.
Supplemental Figure S6. ECS measurements in the wild-type and flvB-21 mutant strains.
Acknowledgments
We thank Claire Riot for preliminary investigations and Drs. Jean Alric and Xenie Johnson (CEA Cadarache) for helpful discussions. Experimental support was provided by the HélioBiotec platform, funded by the European Union (European Regional Development Fund), the Région Provence Alpes Côte d’Azur, the French Ministry of Research, and the CEA.
Footnotes
The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantphysiol.org) is: Gilles Peltier (gilles.peltier{at}cea.fr).
F.C., A.B., and G.P. conceived the original research plans; F.C, A.B., M.M., P.A., and S.B. performed the experiments; F.C., A.B., P. R., and G.P. analyzed the data; F.C., A.B., and G.P. wrote the article.
1 This work was supported by the ERA-SynBio project Sun2Chem, and by the A*MIDEX (ANR-11-IDEX-0001-02) project. F.C. was recipient of a PhD grant of the CEA Tech Division of the “Commissariat à l’Energie Atomique et aux Energies Alternatives.”
2 These authors contributed equally to the article.
[OPEN] Articles can be viewed without a subscription.
Glossary
- CEF
- cyclic electron flow
- CLiP
- Chlamydomonas Library Project
- ECS
- electrochromic carotenoid shift
- MIMS
- membrane inlet mass spectrometer
- MV
- methyl viologen
- NPQ
- nonphotochemical quenching
- Received March 29, 2017.
- Accepted May 7, 2017.
- Published May 9, 2017.