1. Biochemistry and Chemical Biology
  2. Plant Biology
Download icon

Flavodiiron proteins 1–to-4 function in versatile combinations in O2 photoreduction in cyanobacteria

Research Article
  • Cited 1
  • Views 566
  • Annotations
Cite this article as: eLife 2019;8:e45766 doi: 10.7554/eLife.45766

Abstract

Flavodiiron proteins (FDPs) constitute a group of modular enzymes widespread in Bacteria, Archaea and Eukarya. Synechocystis sp. PCC 6803 has four FDPs (Flv1-4), which are essential for the photoprotection of photosynthesis. A direct comparison of light-induced O2 reduction (Mehler-like reaction) under high (3% CO2, HC) and low (air level CO2, LC) inorganic carbon conditions demonstrated that the Flv1/Flv3 heterodimer is solely responsible for an efficient steady-state O2 photoreduction under HC, with flv2 and flv4 expression strongly down-regulated. Conversely, under LC conditions, Flv1/Flv3 acts only as a transient electron sink, due to the competing withdrawal of electrons by the highly induced NDH-1 complex. Further, in vivo evidence is provided indicating that Flv2/Flv4 contributes to the Mehler-like reaction when naturally expressed under LC conditions, or, when artificially overexpressed under HC. The O2 photoreduction driven by Flv2/Flv4 occurs down-stream of PSI in a coordinated manner with Flv1/Flv3 and supports slow and steady-state O2 photoreduction.

https://doi.org/10.7554/eLife.45766.001

Introduction

A-type flavodiiron proteins (Flvs or FDPs) were originally identified in strict and facultative anaerobes among Bacteria, Archaea and Protozoa and were considered to function in O2 and/or NO detoxification (Wasserfallen et al., 1998; Gonçalves et al., 2011; Folgosa et al., 2018). All FDPs share two conserved structural domains: the N-terminal metallo-β-lactamase-like domain, harboring a non-heme diiron center, where O2 and/or NO reduction takes place; and the C-terminal flavodoxin-like domain, containing a flavin mononucleotide (FMN) moiety. The structures of FDPs in anaerobic prokaryotes and eukaryotic protozoa have been resolved as homooligomers (dimer or tetramer comprised of two dimers) arranged in a ‘head-to-tail’ configuration, so that the diiron center of one monomer and the FMN of the other monomer are in close proximity to each other, which ensures rapid electron transfer between the two cofactors.

C-type FDPs, specific to oxygenic photosynthetic organisms, hold an additional flavin-reductase-like domain, coupled with extra cofactors (Romão et al., 2016; Folgosa et al., 2018). Synechocystis sp. PCC 6803 (hereafter, Synechocystis) possesses four genes encoding FDPs: sll1521 (Flv1), sll0219 (Flv2), sll0550 (Flv3) and sll0217 (Flv4). Recently resolved crystal structure of truncated Flv1 from Synechocystis revealed a monomeric form with a ‘bent’ configuration, however the organization of the additional flavin-reductase-like domain and the oligomeric structure remain unclear (Borges et al., 2019). Photosynthetic FDPs first gained attention in 2002, when recombinant Synechocystis Flv3 protein was shown to function in O2 reduction to water without producing ROS (Vicente et al., 2002). Later, it was demonstrated that Synechocystis Flv1 and Flv3 proteins function in vivo in the photoreduction of O2 downstream of Photosystem (PS) I (Helman et al., 2003). Since then, extensive research has been performed to reveal the crucial function of Flv1 and Flv3 (and their homologs, FLVA and FLVB in other photosynthetic organisms) as a powerful sink of excess photosynthetic electrons. This safeguards PSI and secures the survival of oxygenic photosynthetic organisms under fluctuating light intensities (Allahverdiyeva et al., 2013; Gerotto et al., 2016; Chaux et al., 2017; Jokel et al., 2018) or under short repetitive saturating pulses (Shimakawa et al., 2017). The Flv1- and Flv3-mediated light-induced alternative electron transport to O2 was named as the Mehler-like reaction, being a widespread pathway, operating in nearly all photosynthetic organisms from cyanobacteria up to gymnosperms, but lost in angiosperms (Allahverdiyeva et al., 2015; Ilík et al., 2017).

The Flv2 and Flv4 proteins are encoded by an operon, together with a small membrane protein, Sll0218. The flv4-sll0218-flv2 (hereafter flv4-2) operon is strongly induced in low inorganic carbon, Ci, (atmospheric 0.04% CO2 in air, LC) and high light conditions (Zhang et al., 2009). The operon structure is highly conserved in the genome of many β-cyanobacteria (Zhang et al., 2012; Bersanini et al., 2014). The flv4-2 operon-encoded proteins have been reported to function in photoprotection of PSII by acting as an electron sink, presumably transporting electrons from PSII or the plastoquinone (PQ) pool to an unknown acceptor (Zhang et al., 2009; Zhang et al., 2012; Bersanini et al., 2014; Chukhutsina et al., 2015). Since flv2, sll0218 and flv4 are co-transcribed, the contribution of each single protein of the operon to PSII photoprotection has been difficult to dissect. Recent data examining distinct and specific roles of the Flv2/Flv4 heterodimer and the Sll0218 protein (using a set of different mutants deficient only in Sll0218 or in Flv2 and Flv4) demonstrated that the majority of observed PSII phenotypes were actually due to the absence of Sll0218, thus leading to the conclusion that Sll0218 contributes to PSII repair and stability (Bersanini et al., 2017). However, the exact donor and acceptor of the Flv2 and Flv4 proteins have not yet been identified in vivo and possible cross-talk between all four FDPs has yet to be revealed, thus limiting our understanding of the function of FDPs on a cellular level.

In this work, to shed light on the in vivo function of Flv2 and Flv4 and to clearly separate the function of the Flv1/Flv3 heterooligomer from that of Flv2/Flv4, we employed a specific set of FDP mutants. These were: (i) the ∆flv1/∆flv3 mutant, deficient in both Flv1 and Flv3 proteins (Allahverdiyeva et al., 2011); (ii) ∆flv2 which does not express the Flv2 protein but retains a low amount of Flv4 and WT levels of Sll0218 (Zhang et al., 2012); (iii) ∆flv4 which is deficient in the accumulation of all three flv4-2 operon proteins (Zhang et al., 2012); (iv) ∆sll0218 which lacks the small Sll0218 protein, but expresses the Flv2 and Flv4 proteins (Bersanini et al., 2017); (v) ∆flv3/∆flv4 which is deficient in all four FDPs, whereby the absence of Flv3 results in a strong decrease in Flv1 (Mustila et al., 2016) and the inactivation of ∆flv4 affects the expression of the whole flv4-2 operon (Zhang et al., 2012); and, finally (vi) the flv4-2 operon overexpression strain, flv4-2/OE, expressing high amounts of Flv2, Flv4 and Sll0218 (Bersanini et al., 2014).

Here, we provide in vivo evidence for Flv2/Flv4 mediated O2 photoreduction in one of the most frequently studied cyanobacterial model organisms, Synechocystis. Unlike the powerful and rapid response proteins, Flv1 and Flv3, the Flv2 and Flv4 proteins are dispensable for survival under fluctuating light intensities. The expression of flv4 and flv2 under LC was found to be regulated by the pH of the growth media, with significant downregulation observed under strongly alkaline pH conditions. Results from this study provide important insights into the response of photosynthetic organisms to changes in Ci and how they regulate the availability of electron sinks.

Results

Extent and kinetics of the Mehler-like reaction in cells acclimated to low (LC) and high Ci (HC) conditions

Application of membrane inlet mass spectrometry (MIMS) with 18O-enriched oxygen allows differentiation between photosynthetic gross O2 production and O2 uptake under illumination. The flv4-2/OE cells, accumulating high amounts of Flv2, Sll0218 and Flv4 both in LC and HC (>1% CO2 in air, HC) conditions (Bersanini et al., 2014), demonstrated substantially higher O2 photoreduction rates compared to respective WT cells (Figure 1A, B and D). The Flv3 protein level was similar in flv4-2/OE and wild-type (WT) cells grown under both LC and HC (Figure 1C), strongly supporting the in vivo contribution of flv4-2 operon proteins to O2 photoreduction during illumination. Gross O2 evolution rates of flv4-2/OE and WT cells grown under LC did not differ significantly from each other. However, a significant increase in the gross O2 evolution rate was observed in HC grown flv4-2/OE cells (Figure 1—source data 2).

Figure 1 with 2 supplements see all
O2 reduction rates and Flv3 and Flv4 protein accumulation in cells grown in low (LC) and high CO2 (HC).

(A, B) O2 reduction rate of WT, flv4-2/OE and (D) the M55 mutant (∆ndhB) was recorded in darkness (gray background) and under illumination (white background). The experiment was conducted in three independent biological replicates and a representative plot is shown. (Figure 1—source data 1). (C) Immunoblot detection of Flv3 and Flv4 in WT and flv4-2/OE. Pre-cultures were grown in BG-11, pH 8.2 under 3% CO2 (HC) for 3 days, after that cells were harvested and resuspended in fresh BG-11, pH 8.2 at OD750 = 0.2. The experimental cultures were grown under HC or under LC. For the MIMS experiments the cells were harvested and resuspended in fresh BG-11, pH 8.2 at 10 µg Chl a mL−1. O2 photoreduction was recorded during the transition from darkness to high-light intensity of 500 µmol photons m−2s−1. In order to create comparable conditions for MIMS measurements, LC-grown cells were supplemented with 1.5 mM NaHCO3 prior to the measurements. Independent experiments performed on WT cells grown in BG-11 lacking Na2CO3, but supplied with 1.5 mM NaHCO3 prior to MIMS measurement showed no significant difference in O2 photoreduction rates (Figure 1—figure supplement 2), thus allowing confident comparison of the MIMS results. Different phases of O2 photoreduction kinetics are indicated as {I}, {II}, {III}. 50% WT, corresponds to 1:2 diluted WT total protein sample.

https://doi.org/10.7554/eLife.45766.002

As reported earlier, the Ci level has a remarkable effect on the expression of FDPs at both transcript and protein level: Flv2, Flv4 and Flv3 have been shown to be strongly upregulated under LC (Zhang et al., 2009; Wang et al., 2004; Battchikova et al., 2010), and down-regulated upon a shift to HC (Zhang et al., 2009; Hackenberg et al., 2009; Figure 1C). Nevertheless, a direct comparison of the efficiency and kinetics of the Mehler-like reaction in HC- and LC-acclimated cells has not been reported, thus the contribution of different FDPs to O2 photoreduction has been difficult to assess. Our initial approach to evaluating the contributions of the different FDPs was based on determining the activity of the Mehler-like reaction in Synechocystis cells grown under LC and HC (3% CO2) conditions, at pH 8.2.

After a shift from darkness, WT cells demonstrated a rapid light-induced O2 uptake under both LC and HC conditions (59 ± 6.4 and 56 ± 6.4 µmol O2 mg Chl a−1 h−1, respectively). This fast induction phase is designated as {I} in Figure 1A and B. Yet, the kinetics of O2 photoreduction in LC-grown cells differed from those grown under HC. In the LC-grown WT cells, the fast induction phase {I} was followed by a clear biphasic quenching of O2 reduction, namely by the strong decay phase {II}, which continued for about one minute, followed by a quasi-stable state, phase {III} (~33 ± 5.9 µmol O2 mg Chl a−1 h−1) during illumination. Contrasting this, in HC-grown WT cell, the light-induced O2 reduction rate achieved in phase {I} declined only slightly during the first 2–3 min (from ~56 ± 7.7 to~48 ± 6.3 µmol O2 mg Chl a−1 h−1). Thereafter, the rate remained relatively steady for at least 5 min (Figure 1B) of illumination. In flv4-2/OE cells, grown both in LC- and HC, light-induced O2 reduction was stronger that in the WT. Nevertheless, the kinetic phases of O2 photoreduction in flv4-2/OE cells resembled those of respective WT cells, being relatively stable under HC and demonstrating a strong biphasic quenching under LC.

Upon a shift from darkness to light, the ∆flv2 and ∆flv4 mutants grown under HC conditions demonstrated a similar O2 photoreduction pattern as the WT (Figure 1—figure supplement 1). A negligible amount of Flv2 and Flv4 protein in the WT cells grown under HC (Zhang et al., 2009; Zhang et al., 2012; Figure 1C) explains their lack of contribution to the Mehler-like reaction. The near absence of any light-induced O2 reduction in the ∆flv3/flv4 and ∆flv1/flv3 mutants (Figure 1—figure supplement 1) confirms that the small amount of the Flv1/Flv3 heterodimers (decreased Flv3 protein accumulation in HC compared to LC conditions, Figure 1C), is responsible for the constant Mehler-like reaction under the HC condition (Helman et al., 2003).

To uncover the reason for the fast decay of O2 photoreduction observed under LC conditions (Figure 1A), we first tested putative competition between the NAD(P)H:quinone oxidoreductase (NDH-1) complex and FDPs for available photosynthetic electrons. The NDH-1 complex is a powerful machinery utilizing electrons for cyclic electron transport (CET) around PSI, CO2 uptake and respiration under LC conditions (Zhang et al., 2004; Schuller et al., 2019). To this end, O2 photoreduction was measured in the M55 mutant (ΔndhB), which is deficient in the hydrophobic NdhB subunit (Ogawa, 1991) and thus lacks all NDH-1 complexes (Zhang et al., 2004). The M55 mutant cells (grown under LC, pH 8.2 conditions) demonstrated a fast induction of O2 photoreduction (phase I) similar to the WT, which continued at steady-state, lacking the second phase of O2 photoreduction after the dark-to-light transition (Figure 1D). Importantly, the M55 mutant showed a slow induction (see phase I of gross O2 evolution in Figure 1—source data 2) and considerably lower gross O2 evolution rate compared to the WT cells (see phase III of gross O2 evolution in Figure 1—source data 2). This suggests that a steady-state O2 photoreduction in M55 is not due to increased electron flow from PSII. The lack of a strong second phase in O2 photoreduction kinetics resembles the situation in WT cells grown under HC (Figure 1B; Figure 1—figure supplement 1), where the expression of the NDH-1 complex is strongly reduced, and thus suggests competition for electrons between the NDH-1 complexes and FDPs under LC conditions.

The extent and kinetics of the Mehler-like reaction are strongly dependent on the pH and carbonate concentration of the growth medium

The pH and the presence of carbonate in the growth medium were evaluated as possible modulators of the extent and kinetics of the Mehler-like reaction and the accumulation of FDPs under LC conditions. Standard BG-11 medium containing sodium carbonate (Na2CO3) at a final concentration of 0.189 mM was used for all growth experiments, other than those indicated to be Ci limited. In these experiments, performed under atmospheric CO2, Ci limitation was achieved by omitting Na2CO3 from the BG-11 growth media.

The effect of pH

The WT cells grown at pH 9 demonstrated a strong but only transient Mehler-like reaction: the O2 photoreduction rate reached its maximum during the first 30 s of illumination, then quickly dropped (within 1 min) to the initial level of dark O2 uptake (Figure 2, right panel). Similarly to the WT, the ∆flv4 mutant cells demonstrated only a transient O2 photoreduction upon illumination. There was no significant O2 photoreduction detected for ∆flv1/flv3 and ∆flv3/flv4 mutants grown at pH 9.

Figure 2 with 1 supplement see all
O2 reduction rates of WT and FDP mutants grown at different pH levels.

O2 reduction rate was recorded in darkness (gray background) and under illumination with actinic white light at an intensity of 500 µmol photons m−2 s−1 (white background). Pre-cultures were grown in standard BG-11 medium (containing Na2CO3 at a final concentration of 0.189 mM) under HC for 3 days at different pH levels. For MIMS experiments, cells were shifted to LC at OD750≈0.2 (same pH) and grown for 4 days before measurements. Exceptions were: (i) pH 6 experimental cultures were inoculated from pH 8.2 pre-cultures; and (ii) pH 7.5 pre-culture was shifted to LC in standard BG-11 containing Na2CO3 at a final concentration of 0.189 mM or in BG-11 without Na2CO3 (dotted line ‘- Na2CO3’). The experiment was conducted in three independent biological replicates (except experiment at pH 6 with n = 2 independent biological replicates) and a representative plot is shown. (Figure 2—source data 1). In order to create comparable conditions for MIMS measurements, all cells were supplemented with 1.5 mM NaHCO3 prior to the measurements.

https://doi.org/10.7554/eLife.45766.007

Immunoblotting using specific antibodies showed that, as for WT cells grown under HC (Figure 1D), Flv2 and Flv4 proteins were almost undetectable in the WT grown under LC at pH 9 (Figure 3A).

Figure 3 with 1 supplement see all
The effect of the pH of growth medium on the protein and transcript accumulation.

(A, B) The effect of the pH and (B, C) sodium carbonate in the growth medium (A, C) on the protein and (B) transcript levels of FDP. (D) Protein immunoblots demonstrating the accumulation of bicarbonate transporter (SbtA) and NDH-1 subunits (NdhD3 and NdhJ) in the cells grown at different pH and CO2 concentration. Cells were pre-grown at different pH levels (+Na2CO3) under HC for 3 days, harvested, resuspended in fresh BG-11 (pH maintained), adjusted to OD750≈0.2 and shifted to LC for 4 days. At pH 7.5, the cells were grown at LC in the presence (+ Na2CO3, at final concentration of 0.189 mM) or in the absence (- Na2CO3) of sodium carbonate (B, C). Transcript abundance is presented as mean ± SD, n = 2–4 biological replicates, asterisks indicate a statistically significant difference to the WT (*p<0.05; ***p<0.001) (Figure 3—source data 1). Numbers 1–3 indicate different biological replicates. 25% and 50% correspond to 1:4, 1:2 diluted total protein sample, and 100% indicates undiluted total protein sample.

https://doi.org/10.7554/eLife.45766.010

In line with protein data, the transcript levels of both flv2 and flv4 were significantly down-regulated in the cells grown at pH 9 (Figure 3B), suggesting a pH-dependent transcriptional regulation of flv4 and flv2. This is consistent with earlier transcriptional profiling experiments reporting downregulation of flv2 and flv4 transcripts after transferring Synechocystis from pH 7.5 to pH 10 (Summerfield and Sherman, 2008). Importantly, the accumulation of Flv3 was not affected at pH 9. These results strongly suggest that the conspicuous but transient O2 photoreduction observed in the WT and ∆flv4 mutant cells at pH 9 originates mainly from the activity of Flv1/Flv3 heterodimer.

The WT cells grown at pH 6, at pH 7.5 (Figure 2, left and middle panels, respectively) and at pH 8.2 (Figure 1A) demonstrated a rapid induction of O2 reduction (phase {I}) followed by a biphasic decay during illumination: a fast decay phase (phase {II}) and a quasi-stable phase (phase {III}) (Figures 1A and 2). The highest O2 photoreduction rate was observed in the WT cells grown at pH 6 (Figure 2).

Importantly, the ∆flv1/flv3 mutant also showed residual O2 photoreduction: only a small O2 uptake was noticeable at pH 7.5, whereas at pH 6 the O2 photoreduction rate was substantial and constant during 5 min of illumination (Figure 2). Unlike the ∆flv1/flv3 mutant, both the ∆flv2 (Figure 2—figure supplement 1) and ∆flv4 (Figure 2) mutants showed a strong transient O2 photoreduction phase, peaking around the first 30 s of illumination and decaying quickly thereafter. This occurred at all tested pH levels. These results together with those demonstrating highly increased rates of O2 photoreduction in the overexpression strain flv4-2/OE (Figure 1B) collectively confirm the in vivo involvement of both Flv2 and Flv4 proteins in O2 photoreduction. The O2 photoreduction kinetics of the ∆sll0218 mutant resembled that of the WT (Figure 1—figure supplement 1 and Figure 2—figure supplement 1), indicating that the Sll0218 protein does not contribute to the Mehler-like reaction under the HC and LC conditions studied here. These results led us to exclude the ∆sll0218 mutant from any further experiments included in this section.

The data presented above allowed us to make preliminary conclusions about the origin of the different kinetic phases of O2 photoreduction. Since a transient O2 photoreduction was characteristic for the WT, ∆flv2 and ∆flv4 cells, but almost undetectable for ∆flv1/flv3, it is conceivable that the Flv1/Flv3 heterodimer is mostly responsible for the strong and transient O2 uptake during dark-light transitions, whilst Flv2/Flv4 contributes to steady-state O2 photoreduction under LC (see ∆flv1/flv3 particularly at pH 6, Figure 2). The complete lack of O2 photoreduction in the ∆flv3/flv4 mutant (representing deficiency of all four FDPs) is in line with this hypothesis. Importantly, there was no significant difference in the gross O2 evolution rates observed between the wild-type and the FDP mutants (Figure 1—source data 2).

It is not only FDPs, but also distinct variants of the NDH-1 complex as well as HCO3- transporters (Zhang et al., 2004) which are known to respond to CO2 and pH levels of the growth medium. Immunoblotting was performed to evaluate the abundances of NdhD3, representing a low Ci-inducible NDH-1MS complex, and SbtA, a high-affinity low Ci-inducible Na+/HCO3- transporter, in WT and different mutants under conditions used for the MIMS experiments.

As expected, in WT cells grown at pH 7.5, NdhD3 and SbtA were not detected under HC conditions, but both proteins were strongly accumulated in LC (Figure 3D). However, in LC conditions, the increase in alkalinity of the growth medium to pH 9 resulted in markedly lower levels of NdhD3 and SbtA accumulation compared to those observed at pH 7.5. The effect was more pronounced in the case of SbtA. Interestingly, the ∆flv2 and ∆flv4 mutants demonstrated a decrease of SbtA accumulation compared to WT even at pH 7.5 in LC, whereas in flv4/OE SbtA remained at the same level as in WT (Figure 3D).

The expression of the SbtA protein closely followed the changes in the expression of Flv2 and Flv4 proteins under all growth conditions, suggesting that Flv2/Flv4 and the Ci uptake mechanisms, particularly the inducible high-affinity Na+/HCO3- transporter, share a common regulatory pathway of protein expression.

Unlike the growth media at pH 6–8.2, the Ci-pool at pH 9 contains an additional species, CO32-. It is possible that a small amount of CO32- in the external growth medium acts as a signal to trigger the regulation of flv2 and flv4 expression via antisense RNA as1-flv4 and the master transcription factors, ndhR or cmpR (Eisenhut et al., 2012). Considering that the double negative charge of CO32- prevents its diffusion through the cell membrane, and the fact that an active carbonate uptake transporter is currently unknown, we cannot yet consider CO32- to be an internal sensor. To gain further insight to the carbonate effect on O2 photoreduction, MIMS experiments were performed on FDP mutants grown in BG-11 medium in the presence (0.189 mM) and absence of sodium carbonate.

The effect of sodium carbonate

Culturing the cells without Na2CO3 at pH 7.5 clearly enhanced O2 photoreduction in the WT and all studied FDP mutants (Figure 2, middle panel). Despite such a clear variation in O2 photoreduction rates in the WT, no significant difference in gene transcript (Figure 3B) and protein levels (Figure 3C) of FDPs were observed in the presence or absence of Na2CO3.

FDP induced O2 photoreduction does not occur at PSII or PQ-pool level

In order to establish where in the electron transport chain the Flv2/Flv4 heterodimer-related O2 photoreduction occurs, we focused on the flv4-2/OE mutant (grown at LC, pH 7.5, without carbonate). This mutant showed especially high accumulation of Flv2 and Flv4 proteins and a higher O2 photoreduction rate than the WT (Figure 1). When linear electron transport was blocked at Cytochrome b6f (Cyt b6f) level using DBMIB as an inhibitor (Draber et al., 1970; Yan et al., 2006), both the WT (Ermakova et al., 2016) and flv4-2/OE mutant cells demonstrated a strong light-induced O2 uptake (Figure 3—figure supplement 1). As expected, in the Δcyd mutant the light-induced O2 uptake was not detected in the presence of DBMIB (Ermakova et al., 2016), Figure 3—figure supplement 1). The addition of HQNO, an inhibitor of Cytochrome bd quinol oxidase (Cyd) (Pils et al., 1997) and Cyt b6f (Fernández-Velasco et al., 2001) to the DBMIB-treated WT and flv4-2/OE completely eliminated O2 photoreduction. These results confirmed that Cyd was solely responsible for the observed O2 photoreduction occurring at the PQ-pool level.

Growth phenotype of FDP deletion mutants under fluctuating light intensities

We have previously demonstrated that the Flv1/Flv3 heterodimer enables cell growth under fluctuating light, by functioning in the Mehler-like reaction as an efficient electron sink (Allahverdiyeva et al., 2013). However, the results of the current study clearly suggest an additional involvement of the Flv2/Flv4 heterodimer in the Mehler-like reaction, particularly under conditions of LC and at pH values of 8.2 or lower (Figures 1 and 2). These findings led us to more precisely examine the combined effects of the pH of the growth medium and the fluctuating growth light conditions (FL) on the growth performance of various FDP mutants. To this end, both severe (FL20/500, when 20 µmol photons m−2 s−1 background light was interrupted every 5 min by 30 s light pulse intensity of 500 µmol photons m−2 s−1) and mild (FL50/500, when 50 µmol photons m−2 s−1 background light was interrupted every 5 min by 30 s light pulse intensity of 500 µmol photons m−2 s−1) fluctuating lights were applied at different levels of pH. In line with our previous work, the ∆flv1/flv3 mutant (also ∆flv3/flv4) failed to grow under severe (FL20/500) light fluctuations, independent of the pH of the growth medium (Figure 4; Figure 4—figure supplement 1). Differently to the severe FL20/500 condition, under mild fluctuating light (FL50/500), the ∆flv1/flv3 mutant demonstrated slower growth than the WT under alkaline pH (pH 9, Figure 4 and pH 8.2 (Mustila et al., 2016), Figure 4—figure supplement 1)). Growth was similar to the WT at pH 7.5 (Mustila et al., 2016), Figure 4—figure supplement 1) and pH 6 (Figure 4). Importantly, the ∆flv4 mutant grew similarly to the WT at all studied pH levels, both under mild and severe FL conditions (Figure 4). The ∆flv2,sll0218 and flv4-2/OE mutants also demonstrated similar growth to the WT under severe FL20/500 at pH 7.5 and 8.2 (Figure 4—figure supplement 1).

Figure 4 with 1 supplement see all
Growth curves of the different FDPs mutants under fluctuating light intensities.

Pre-cultures were grown in BG-11 medium under HC for 3 days illuminated with constant light of 50 µmol photons m−2 s−1. The cells pre-grown at pH 9 or pH 8.2 (for experimental culture at pH 6) were harvested, resuspended in fresh BG-11 (pH 9 or 6), adjusted to OD750 = 0.1 and shifted to LC. Experimental cultures were grown under FL 20/500 or 50/500 regime for 8 days. The experiment was conducted in two independent biological replicates and average values was plotted.

https://doi.org/10.7554/eLife.45766.013

The results above strongly suggest that, in contrast to the Flv1/Flv3-originated Mehler-like reaction, Flv2/Flv4-driven O2 photoreduction is not essential for the survival of cells under fluctuating light.

Effect of increasing light intensities on the Mehler-like reaction

In order to assess the response of the O2 photoreduction to different light intensities, the WT, ∆flv4 and ∆flv1/flv3 mutant cells were illuminated with 500, 1000 and 1500 µmol photons m−2 s−1 white light (Figure 5). Under LC conditions, increasing the light intensity from 500 to 1000 µmol photons m−2 s−1 resulted in a two-fold increase of the maximum O2 photoreduction rate in the WT (Figure 5A and D). The further increase (1500 µmol photons m−2 s−1) only slightly enhanced (2.3-fold) the maximum O2 photoreduction rate, suggesting that the applied light intensity was nearly saturating. Likewise, the ∆flv4 mutant demonstrated about 1.9- and 2.3-fold enhancements of the maximum rate of transient light-induced O2 reduction under 1000 and 1500 µmol photons m−2 s−1, respectively (Figure 5C and D). Contrasting this was the results of the ∆flv1/flv3 mutant, which showed lesser responses to increasing light intensities (1.6- and 1.8-fold enhancement in the maximum rate at 1000 and 1500 µmol photons m−2−1, respectively) (Figure 5B and D). It is important to note that both the ∆flv4 and ∆flv1/flv3 mutants accumulate nearly the WT level of the Flv3 or Flv4/Flv2 proteins, respectively (Zhang et al., 2009; Mustila et al., 2016). Moreover, increasing light intensity from 500 to 1500 µmol photons m−2 s−1 also resulted in enhancement of the O2 photoreduction rate in the WT cells grown under HC (Figure 5—figure supplement 1).

Figure 5 with 2 supplements see all
Rates of O2 reduction in response to increasing light intensity in WT, ∆flv1/flv3 and ∆flv4 mutant cells (A, B, C, respectively).

O2 reduction rate was recorded in darkness (gray background) and under illumination with actinic white light intensities of 500, 1000 and 1500 µmol photons m−2 s−1 (white background). In order to create comparable conditions for MIMS measurements, all cells were supplemented with 1.5 mM NaHCO3 prior to the measurements. Pre-cultures were grown in BG-11 medium (pH 7.5) under 3% CO2 (HC) for 3 days and then shifted to LC (atmospheric 0.04% CO2 in air) at OD750 = 0.2 and pH 7.5 for 4 days. For MIMS measurements, cells were harvested and resuspended in fresh BG-11 medium at a Chl a concentration of 10 µg mL−1. (D) Maximum rate of light-induced O2 uptake (O2 µmol mgChl a−1 hr−1) of WT, ∆flv1/flv3 and ∆flv4 mutant cells at different light intensities applied. The experiment was conducted in three independent biological replicates and a representative plot is shown (Figure 5—source data 1).

https://doi.org/10.7554/eLife.45766.016

The fast and transient response of ∆flv4 mutant cells to drastic increases in light intensity (Figure 5C) confirmed the high capacity of Flv1/Flv3-related O2 photoreduction to act as an electron sink. These results explain the essential role of Flv1/Flv3, unlike Flv2/Flv4, for the survival of cells under fluctuating light intensities. Intriguingly, both the fast induction phase {I} and quasi-stable phase {III} of O2 photoreduction rates of the WT were greater than the sum of the individual O2 photoreduction rates from ∆flv1/flv3 and ∆flv4, implying a strong enhancement of O2 photoreduction by various oligomer activities in the presence of all four FDPs.

Echoing trends seen in O2 photoreduction rates, gross O2 evolution rates of the WT strongly enhanced with increasing light intensities (1.6- and 1.8-fold increase in 1000 and 1500 µmol photons m−2 s−1, respectively), whereas the Δflv4 mutant showed only limited increases of gross O2 evolution rates (1.3- and 1.5-fold in 1000 and 1500 µmol photons m−2 s−1, respectively), and Δflv1/flv3 O2 evolution rates were already at maximum levels under the lowest light intensity of 500 µmol photons m−2 s−1 (Figure 1—source data 2). It is worth mentioning that, neither the Δflv1/flv3 nor Δflv4 mutant achieved a steady-state gross O2 evolution during the 5 min of illumination: Δflv1/flv3 demonstrated gradual increase, whereas Δflv4 showed gradual decrease in gross O2 evolution. Next, PSII (O2 evolving activity monitored in the presence of artificial electron acceptor, DMBQ) and PSI (maximum oxidizable amount of P700, Pm) activities were measured in cells grown under moderate light (50 µmol photons m−2 s−1) and exposed to high light (1500 µmol photons m−2 s−1) for 2 hr. After 2 hr of high light treatment, ∆flv1/flv3 showed no significant difference in the maximum oxidizable amount of P700 (Pm) and PSII activity compared to the WT and Δflv4 mutant (Figure 5—figure supplement 2). This is in line with previous studies proving that other photoprotective mechanisms are able to replace Flv1/Flv3 (Zhang et al., 2009) unless the cells experience abrupt fluctuations in light intensity (Allahverdiyeva et al., 2013). It has already been shown that a strong high light (1500 µmol photons m−2 s−1) causes slightly slow growth and a short high light treatment decreases PSII activity in the Δflv4 mutant compared to the WT (Figure 5—figure supplement 2; Zhang et al., 2009; Bersanini et al., 2014; Bersanini et al., 2017). Importantly, Δflv4 demonstrated a Pm level comparable to that of the WT after 2 hr of high-light treatment. This suggests the importance of the Flv2/Flv4 driven steady-state O2 photoreduction in photoacclimation, by the prevention of PSII photodamage caused by the over-reduction of the photosynthetic chain.

The functional expression of FDPs is highly modulated by Ci conditions and light penetration

The inoculum size (starting OD750 value) determines the extent of light penetration upon starting a cultivation. In previous studies, cells were pre-grown in HC, then harvested at late logarithmic phase and inoculated in fresh BG-11 (pH 8.2) at OD750≈0.4–0.5, before shifting to LC for the next 3 days (Allahverdiyeva et al., 2011; Allahverdiyeva et al., 2013; Ermakova et al., 2016). To ensure better light penetration of the cultures and to improve the acclimation of cells to the conditions used in this study, the experimental WT and ∆flv1/flv3 cultures were inoculated at a low OD750≈0.1–0.2 and then cultivated for 4 days (instead of 3 days in previous studies). The WT cells grown under LC from a lower OD (OD750≈0.2) demonstrated notably higher O2 uptake during illumination, compared to the cells shifted to LC at OD750≈0.5 (Figure 6A). Importantly, the ∆flv1/flv3 mutant cells shifted to LC at a lower OD (OD750≈0.2) also demonstrated a residual steady-state O2 photoreduction activity.

Effect of inoculum size on the O2 photoreduction and accumulation of FDPs in the WT and Δflv1/Δflv3 mutant cells.

(A) Rates of O2 uptake measured by MIMS during darkness (gray background) and under illumination with actinic white light at an intensity of 500 µmol photos m−2s−1 (white background). In order to create comparable conditions for MIMS measurements, all cells were supplemented with 1.5 mM NaHCO3 prior to the measurements. (B) Protein immunoblots showing the relative accumulation of different FDPs in the WT and Δflv1/Δflv3 mutant cells. Pre-cultures were grown in BG-11 (pH 8.2) under HC until late logarithmic phase (OD750≈2.5), then harvested and inoculated in fresh BG-11 under LC at OD750 = 0.2 for 4 days or OD750 = 0.5 for 3 days. The experiment was conducted in three independent biological replicates and a representative plot is shown in (A). WT_50% corresponds to 1:2 diluted total protein sample and 100% to undiluted total protein sample.

https://doi.org/10.7554/eLife.45766.020

Immunoblot analysis using specific FDP antibodies showed that the WT cells transferred from HC to LC at OD750 = 0.2 accumulated higher amount of the Flv2, Flv3 and Flv4 proteins compared to the cells shifted to LC at OD750 = 0.5 (Figure 6B). A similar trend was also observed in the ∆flv1/flv3 mutant, which accumulated more Flv2 and Flv4 when cultivated at LC from OD750 = 0.2. This is in line with previous results showing that the accumulation of flv2 and flv4 transcripts in Synechocystis (upon a shift from HC to LC, Zhang et al., 2009) and vegetative cell-specific flv1A and flv3A transcripts in Anabaena sp. PCC 7120 (upon a shift from dark to light, Ermakova et al., 2013) strongly depended on light intensity.

The results above highlight that Ci and light penetration upon a shift of cells from pre-culture conditions to different experimental conditions highly modulate the functional expression of FDPs.

Discussion

The Flv2/Flv4 heterodimer contributes to the Mehler-like reaction when naturally expressed under LC conditions or artificially overexpressed under HC

By characterizing Synechocystis mutants specifically affected in the accumulation of various FDPs, we show here that Flv2 and Flv4, together with Flv1 and Flv3 proteins, are involved in O2 photoreduction in vivo. Until recently, it has generally been accepted that the Flv1/Flv3 proteins safeguard PSI under both HC and LC conditions (Allahverdiyeva et al., 2013), whereas proteins encoded by the flv4-2 operon and being highly expressed under LC, function in the photoprotection of PSII, presumably by directing excess electrons from PSII to an as yet unknown acceptor (Zhang et al., 2009; Zhang et al., 2012; Shimakawa et al., 2015). The possibility of an Flv2/Flv4 contribution to O2 photoreduction in vivo was neglected due to a lack of evidence for light-induced O2 uptake in ∆flv1 and/or ∆flv3 mutants (Helman et al., 2003; Allahverdiyeva et al., 2011; Allahverdiyeva et al., 2013). Thus, Flv1 and Flv3 were assumed to be solely responsible for the Mehler-like reaction. Recently, it was demonstrated that Synechocystis Flv4 expressed in E. coli is capable of NADH-dependent O2-reduction in vitro (Shimakawa et al., 2015). However, the reported reaction rate was extremely low (almost residual) compared to the activity of FDP for example from anaerobic protozoa (Di Matteo et al., 2008) and the enzyme showed no affinity to NADPH. A similar scenario was previously presented for the Flv3 protein, where in vitro studies performed on recombinant Synechocystis protein led to a claim that Flv3 functions as a homodimer in NADH-dependent O2 reduction (very low affinity to NADPH) (Vicente et al., 2002), whilst subsequent study with ∆flv1-OEflv3 (or ∆flv3-OEflv1) mutants clearly demonstrated that homooligomers of Flv3 (or Flv1) do not function in O2 photoreduction in vivo (Mustila et al., 2016). Such discrepancies between the in vitro and in vivo results suggest that the in vitro assays conducted thus far have apparently failed to take into full consideration all the complex intracellular interactions, for example the involvement of Fed or FNR as an electron donor for FDPs, or the in vitro experiments do not necessarily demonstrate the processes occurring in vivo.

In this study, we provide compelling evidence for the in vivo contribution of Flv2/Flv4 to O2 photoreduction by applying 18O-labeled-oxygen and real-time gas-exchange measurements to distinct FDP deletion mutants. The inactivation of flv2 or flv4 is shown to result in a substantial decrease of O2 photoreduction in the mutants compared to the WT, while the overexpression of the flv4-2 operon increases the rate of O2 photoreduction approximately two-fold. In addition, the possibility that the small protein Sll0218 contributes to the Mehler-like reaction is excluded (Figure 1—figure supplement 1, compare Figure 2—figure supplement 1 and Figure 2).

It is noteworthy that both the ∆flv2 (deficient in Flv2 but retaining a low amount of Flv4) and ∆flv4 (deficient in both Flv2 and Flv4) mutants showed similar inhibition of O2 photoreduction rates, thus supporting the function of Flv2/Flv4 as a heterodimer in the Mehler-like reaction. The existence of the Flv2/Flv4 heterodimer has been proved biochemically in Synechocystis (Zhang et al., 2012). Nonetheless, our data do not exclude the possibility that Flv2/Flv2 and/or Flv4/Flv4 homooligomers are also involved in processes other than O2 photoreduction. Such a situation occurs with the Flv1 and Flv3 proteins, which contribute as homooligomers to the photoprotection of cells under fluctuating light conditions, probably via an unknown electron transport and/or regulatory network (Mustila et al., 2016).

The complete elimination of light-induced O2 reduction in WT cells grown at pH 8.2 (Ermakova et al., 2016) or at pH 7.5 (Figure 3—figure supplement 1) in the presence of electron-transport inhibitors DBMIB (blocks Qo site of Cytb6f; Roberts and Kramer, 2001) and HQNO (blocks Qi site of Cytb6f; Fernández-Velasco et al., 2001 and also Pils et al., 1997) suggests that FDP-driven O2 photoreduction (neither by Flv1/Flv3 nor by Flv2/Flv4) does not occur at the PSII or PQ-pool level. This conclusion is also supported by the fact that, differently to the WT and mutants deficient in FDPs, the ∆cyd mutant does not exhibit a light induced O2 uptake in the presence of DBMIB (Ermakova et al., 2016; Figure 3—figure supplement 1).

From the results discussed above, it can be concluded that both the Flv1/Flv3 and Flv2/Flv4 heterodimers have capacity to drive the Mehler-like reaction, functioning downstream of PSI.

The Flv1/Flv3 heterodimer drives a strong and steady-state O2 photoreduction under HC

It is generally accepted that under LC conditions, the slowing down of the Calvin-Benson cycle leads to a build-up of reduced stromal components (Cooley and Vermaas, 2001; Holland et al., 2015), which would stimulate the Mehler reaction to dissipate excess electrons (Ort and Baker, 2002). However, under HC conditions, the Mehler reaction would be expected to direct relatively low electron flux to O2. In this study, we provide evidence that HC-grown WT cells are capable of equally high O2 photoreduction as respective LC-grown WT cells, and that cells are capable of maintaining the steady-state activity at least during the first 5–10 min of illumination (Figure 1A). Compared to the WT, a drastically lower O2 photoreduction rate is observed in the ∆flv1/flv3 and ∆flv3/flv4 mutants grown in HC, confirming that O2 uptake under these conditions is mostly due to the Flv1/Flv3-driven Mehler-like reaction (Figure 1—figure supplement 1).

It is important to note that the O2 photoreduction capacity of Synechocystis generally correlates with the abundance of FDPs (Figures 1 and 6). However, protein abundance is not the only factor that determines O2 photoreduction capacity. Indeed, despite strong and steady-state O2 photoreduction, HC-grown cells demonstrate nearly undetectable levels of Flv2 and Flv4 and low amount of Flv3, compared to levels observed under LC conditions. Furthermore, the increase in O2 photoreduction rates (Figure 2, middle panel) obtained by omitting sodium carbonate from the BG-11 growth media at pH 7.5, does not correlate with any significant change in transcript and protein levels of the FDPs, thus suggesting a possible redox regulation of the enzyme activity.

Under LC, the Flv1/Flv3 heterodimer is a rapid, strong and transient electron sink whereas Flv2/Flv4 supports steady-state O2 photoreduction

The Mehler-like reaction of WT cells grown under LC at pH 6–8.2 exhibits triphasic kinetics of O2 photoreduction originating from the activity of both Flv1/Flv3 and Flv2/Flv4 heterodimers (Figure 2). In this study, we were able to unravel the contribution of Flv1/Flv3 and Flv2/Flv4 heterodimers to the O2 photoreduction kinetics: Flv1/Flv3 is mainly responsible for the rapid transient phase, whereas Flv2/Flv4 mostly contributes to the slow steady-state phase.

The almost complete absence of Flv2 and Flv4 proteins in WT cells grown under LC at pH 9 provides an excellent model system, where the Mehler-like reaction is naturally driven solely by the Flv1/Flv3 heterodimer, as is also the case under HC conditions. However, in contrast to HC-grown cells, where Flv1/Flv3 can drive a steady-state O2 photoreduction, the cells grown under LC at pH 9 demonstrate strong but only transient O2 photoreduction, which decays during the first 1–2 min of illumination (Figure 2). The identical O2 photoreduction kinetics of the WT cells grown at pH 9 (accumulating Flv3 but lacking both the Flv2 and Flv4 proteins) and the ∆flv4 mutant (accumulating Flv3 but lacking Flv4 and also Flv2), together with the complete absence of O2 photoreduction in the ∆flv3/flv4 mutant demonstrate that under LC, the Flv1/Flv3 heterodimer contributes to the Mehler-like reaction in a fast and transient manner (Figure 2). A similar conclusion was previously suggested for Synechocystis (Allahverdiyeva et al., 2013) and for the FlvA and FlvB proteins in Physcomitrella patens (Gerotto et al., 2016) and Chlamydomonas reinhardtii (Chaux et al., 2017; Jokel et al., 2018).

The sole contribution of Flv2/Flv4 to the Mehler-like reaction is clearly demonstrated as a steady-state O2 photoreduction by the ∆flv1/flv3 mutant grown under LC at pH 6 (Figure 2), whilst the same mutant cells grown at pH 7.5 and 8.2 show only residual steady-state O2 photoreduction. It is important to note that the Flv2/Flv4 heterodimer, when expressed, can readily contribute to O2 photoreduction under HC, as demonstrated by the flv4-2/OE strain (Figure 1), thus excluding all redox and structural hindrances for Flv2/Flv4 to function in O2 photoreduction under HC. However, such a contribution is naturally abolished in WT cells grown under high levels of CO2 by the down-regulation of the flv4-2 operon (Zhang et al., 2009; Zhang et al., 2012).

The rate of the Mehler-like reaction in WT cells exceeds the cumulative O2 photoreduction driven solely by Flv1/Flv3 (observed in ∆flv4) and Flv2/Flv4 (observed in ∆flv1/flv3). This demonstrates that all four FDPs are required for an efficient Mehler-like reaction in WT cells upon growth under LC (except at pH 9). A complex interaction between FDPs possibly arises from a coordinated inter-regulation of Flv1/Flv3 and Flv2/Flv4 heterodimers and on the possible occurrence of some active Flv1-4 oligomers (Figure 7). Despite detection of homotetrameric organization of Synechocystis Flv3 in vitro (Mustila et al., 2016), the direct biochemical demonstration of homo- or heterotetramer structures and function in vivo is still missing.

A schematic drawing of photosynthetic light reactions and alternative electron transport routes.

(A) A steady-state Mehler-like reaction in HC is carried out by the low-abundant, yet catalytically efficient Flv1/Flv3 heterodimer. The Flv3/Flv3 homooligomer is involved in photoprotection as an electron valve with unknown acceptor or as a component of a signaling/regulating network (Mustila et al., 2016). (B) In LC-grown cells the two pairs of FDP heterodimers are involved in the Mehler-like reaction: Flv1/Flv3 mainly drives rapid and transient O2 photoreduction and Flv2/Flv4 operates relatively slowly and provides a steady-state background O2 photoreduction. The soluble Flv1/Flv3 heterodimers function as an immediate acceptor of electrons presumable from reduced Fed, whereas association of Flv2/Flv4 with the thylakoid membrane (and/or Flv1/Flv3) is controlled by pmf and Mg2+. Several oligomeric forms of FDPs are hypothesized to exist, including a heterotetramer comprising different FDP protein compositions. The higher abundance of total NDH-1 complexes and FDPs oligomers in LC conditions, compared to HC conditions, is represented by larger size of the protein complexes.

https://doi.org/10.7554/eLife.45766.022

The growth inhibition of ∆flv1/flv3 cells under severe fluctuating light conditions (FL 20/500) at pH 8.2 (Allahverdiyeva et al., 2013), pH 7.5 (Mustila et al., 2016), pH 6 and pH 9 (Figure 5) demonstrate the essential role of Flv1 and/or Flv3 during drastic changes of light intensity, whereas Flv2 and Flv4 are dispensable under the same conditions (Figure 4, Figure 4—figure supplement 1). Here, we demonstrate that the crucial importance of Flv1/Flv3 heterodimers is based on their high capacity to rapidly and effectively respond to increasing light intensities (Figure 5). By adjusting their O2 photoreduction activity, the Flv1/Flv3 heterodimer works as an efficient and fast sink of electrons, whereas the responsiveness of Flv2/Flv4 is relatively limited and the heterodimer mostly functions on a slow time-scale in steady-state O2 photoreduction.

The intracellular location of these enzymes may partially contribute to the difference in O2 photoreduction: Flv1 and Flv3 are soluble cytosolic proteins able to quickly associate with soluble Fed and direct electrons towards O2 photoreduction. In line with this, the possible interaction between Synechocystis Flv1, Flv3 and Fed (Hanke et al., 2011), Flv3 and Fed9 (Cassier-Chauvat and Chauvat, 2014), Chlamydomonas reinhardtii FLVB and FED1 (Peden et al., 2013) have been reported. The Flv2/Flv4 heterodimer, specific for cyanobacteria, was suggested to bind to the thylakoid membrane upon increases in Mg2+ concentration on the cytoplasmic surface of the thylakoid membrane when lights are turned on (Zhang et al., 2012). It is likely that the association of Flv2/Flv4 with the membrane enhances electron transfer from Fed (or FNR) to Flv2/Flv4 and would probably result in a delayed and limited O2 photoreduction activity by Flv2/Flv4. However, the possibility that FDPs accept electrons from different and specific Fed paralogs cannot be excluded.

Traffic downstream of PSI affects the FDP-mediated Mehler-like reaction

Unlike WT cells demonstrating biphasic decay kinetics of O2 photoreduction under LC conditions (Figure 1A and Figure 2), the M55 mutant (deficient in NDH-1 mediated CET, CO2 uptake and respiration) (Ohkawa et al., 2000) shows steady-state O2 photoreduction, similar to the HC-grown WT (Figure 1B and D). This suggests that the strongly upregulated NDH-1 complex under LC in Synechocystis (Zhang et al., 2004) contributes to a rapid quenching of O2 photoreduction (Figure 1A, phase {II}) by efficient withdrawal of electrons from reduced Fed. Under such circumstances, the low but steady-state activity of the Flv2/Flv4 heterodimer is likely to be important for keeping linear electron transport in an oxidized state. This would explain why the PQ-pool is more oxidized in the presence of Flv2/Flv4 and more reduced in its absence, indirectly affecting PSII activity (Zhang et al., 2012; Bersanini et al., 2014) and Chukhutsina et al., 2015). Thus, by allocating different roles for FDPs between the two pairs of heterodimers (Flv1/Flv3 and Flv2/Flv4), the cells are well positioned to respond appropriately to changing Ci levels as well as to abrupt changes in light intensity, in a coordinated and energetically efficient manner.

Unlike prokaryotic cyanobacteria, chlorophytic algae (e.g. Chlamydomonas reinhardtii) and mosses rely not only on the FDP-driven pathway, but also harbor the PROTON GRADIENT REGULATION5 (PGR5)/PGR5‐LIKE PHOTOSYNTHETIC PHENOTYPE 1 (PGRL1) pathway which operates concomitantly to protect the cells under fluctuating light. It is noteworthy, however, that the PGR5/PGRL1 machinery in Chlamydomonas reinhardtii is neither fast nor strong enough to mitigate acceptor-side pressure under highly fluctuating light intensities. To complement this deficiency, the FDP-mediated pathway is indispensable for coping with sudden increases in light intensity (Jokel et al., 2018). Interestingly, the introduction of Physcomitrella patens FDPs rescues a fluctuating light phenotype of the PGR5 Arabidopsis thaliana mutant (Yamamoto et al., 2016; Yamamoto and Shikanai, 2019), and alleviates PSI photodamage in the PGR5-RNAi, crr6 (defective in NDH-dependent CET) and the PGR5-RNAi crr6 double mutants of Oryza sativa by acting as a safety valve under fluctuating light and substituting for CET without competing with CO2 fixation under constant light (Wada et al., 2018). Moreover, the expression of Synechocystis Flv1 and Flv3 in tobacco plants enhances photosynthetic efficiency during dark-light transitions by providing an additional electron sink (Gómez et al., 2018). Although data on Flv2/Flv4 proteins expressed in angiosperms is not yet available, our results collectively suggest that the FDP pathway(s) is important to consider in future high-yield crop development and microbial cell factories.

The question of how FDPs avoid competition with CO2 fixation is an interesting one. Relevant mechanisms may include post-transcriptional modifications of the FDPs, such as phosphorylation (Angeleri et al., 2016), and/or pmf based regulation systems.

Figure 7 provides a summary scheme of our understanding of the function and interaction of the different FDPs and their oligomers in photoprotection of the photosynthetic apparatus in the model cyanobacterium Synechocystis sp. PCC 6803. The importance of the available Ci species in the function and accumulation of FDPs is emphasized by separate schemes for the HC and LC growth conditions.

Materials and methods

Key resources table
Reagent type
(species) or
resource
DesignationSource or referenceIdentifiersAdditional
information
Strain, strain background (Synechocystis sp. PCC 6803)WT, Wild-typeWilliams, 1988
Genetic reagent (Synechocystis sp. PCC 6803)flv2Zhang et al., 2012
Genetic reagent (Synechocystis sp. PCC 6803)flv4Zhang et al., 2012
Genetic reagent (Synechocystis sp. PCC 6803)flv1/∆flv3Allahverdiyeva et al., 2011
Genetic reagent (Synechocystis sp. PCC 6803)flv3/∆flv4Helman et al., 2003
Genetic reagent (Synechocystis sp. PCC 6803)∆sll0218‐flv2Helman et al., 2003
Genetic reagent (Synechocystis sp. PCC 6803)flv4-2/OEBersanini et al., 2014
Genetic reagent (Synechocystis sp. PCC 6803)sll0218Bersanini et al., 2017
Antibodyα-Flv2 (rabbit polyclonal)AntiProt, against amino acids 521–535 of Synechocystis Flv2(1:500)
Antibodyα-Flv3 (rabbit polyclonal)AntiProt, against amino acids 377–391 of Synechocystis Flv3(1:2000)
Antibodyα-Flv4 (rabbit polyclonal)AntiProt, against amino acids 412–426 of Synechocystis Flv4(1:500)
Antibodyα-NdhD3 (rabbit polyclonal)Eurogentec, against amino acids 185 to 196 and 346 to 359 of Synechocystis NdhD3(1:1000)
Antibodyα-SbtAKind gift from T. Ogawa, against amino acids 184 to 203 of Synechocystis SbtA(1:5000)
Antibodyα-NdhJKind gift from J. Appel(1:1000)
AntibodySecondary antibody, Amersham ECL Rabbit IgG, HRP-linked F(ab')₂ fragment (from donkey)GE HealthcareNA9340-1ML(1:10000)
Commercial assay or kitAmersham ECL Western Blotting Detection ReagentGE HealthcareRPN2209
Commercial assay or kitiScript cDNA Synthesis KitBioRad, USACat. #170–8891
Commercial assay or kitiQ SYBR Green SupermixBioRad, USACat. #170–8882
Software, algorithmqbase + softwareBiogazelle,
Zwijnaarde, Belgium - www.qbaseplus.com

Strains and culture conditions

Request a detailed protocol

The glucose-tolerant Synechocystis sp. PCC 6803 was used as wild type (WT) strain (Williams, 1988). The FDP inactivation mutants ∆flv2, ∆flv4 (Zhang et al., 2012), and the double mutants ∆flv1/∆flv3 (Allahverdiyeva et al., 2011), and ∆flv3/∆flv4 (Helman et al., 2003), ∆sll0218‐flv2 (Helman et al., 2003) have been described previously. The flv4-2/OE and ∆sll0218 mutants were described in Bersanini et al. (2014); Bersanini et al. (2017).

Pre-experimental cultures were grown at 30°C in BG-11 medium, illuminated with continuous white light of 50 µmol photons m−2 s−1 (growth light: GL), under air enriched with 3% CO2 (high carbon: HC). BG-11 medium was buffered with 20 mM 2-(N-morpholino) ethanesulfonic acid (MES, pH 6.0), 20 mM HEPES-NaOH (pH 7.5), 10 mM TES-KOH (pH 8.2) or 10 mM N-Cyclohexyl-2-aminoethanesulfonic acid (CHES, pH 9.0), according to the pH of the experimental condition. Pre-cultures were harvested at logarithmic growth phase, inoculated in fresh BG-11 medium at OD750 = 0.2 (or OD750 = 0.5 when mentioned), measured with and shifted to low CO2 (atmospheric 0.04% CO2 in air, LC). OD750 was measured using Lambda 25 UV/VIS spectrometer (PerkinElmer, USA). HC experimental cultures were inoculated at OD750 = 0.1 and kept at HC for 3 days. During experimental cultivation, cells were grown under continuous GL at 30°C with agitation at 120 rpm and without antibiotics. For growth curves, cells pre-cultivated under continuous GL and HC were collected, inoculated at OD750 = 0.1 and shifted to LC under a light regime with a background light of 20 µmol photons m−2 s−1 interrupted with 500 µmol photons m−2 s−1 for 30 s every 5 min (FL 20/500) or 50 µmol photons m−2 s−1 interrupted with 500 µmol photons m−2 s−1 for 30 s every 5 min (FL 50/500). The standard BG-11 medium used in this work contains sodium carbonate (Na2CO3) at a final concentration of 0.189 mM and only when mentioned the sodium carbonate was omitted from the growth medium.

Absence of contamination with heterotrophic bacteria was checked by dropping liquid culture on LB and R2A agar plates and kept at 30°C.

Isolation of total RNA and Real-time quantitative PCR (RT-qPCR)

Request a detailed protocol

Total RNA was isolated from exponentially growing Synechocystis by hot-phenol method previously described (Tyystjärvi et al., 2001). After removing any residual genomic DNA, the RNA concentration and purity were measured with a NanoDrop spectrophotometer (Thermo Scientific, USA). RNA integrity was verified by agarose gel electrophoresis.

Complementary DNA was synthesized from 1 μg of purified RNA using the iScript cDNA Synthesis Kit (BioRad, USA) according to the manufacturer’s protocol. Synthesized cDNA was diluted four-fold and used as template for the RT-qPCR. The samples for RT‐qPCR were labeled by iQ SYBR Green Supermix (BioRad, USA) to detect accumulation of amplicons in 96-well plates. The primers to detect transcripts of flv1 and flv2 as well as for the reference genes rnpB and rimM are described in Mustila et al. (2016). The forward and reverse primers for flv3 were 5’-CAACTCAATCCCCGCATTAC-3’ and 5’-CAGTGGAGATTCGGAGCACT-3’ and for flv4 5’-ACGATGCCTGGAGTCAAAAC-3’ and 5’-GGGTATCCGCCACACTTAGA-3’. The PCR protocol was as follows: 3 min initial denaturation of cDNA at 95°C, followed by 40 cycles of 95°C for 10 s, annealing in 57°C for 30 s and extension in 72°C for 35 s. A melting curve analysis was performed at the end. Relative changes in the gene expression were determined using the qbase + software by Biogazelle. One-way ANOVA analysis performed with SigmaPlot was used to determine significant changes in gene expression.

MIMS experiments

Request a detailed protocol

In vivo measurements of 16O2 (mass 32) and 18O2 (mass 36) exchange was performed using a Membrane-inlet mass spectrometry (MIMS) as described previously in Mustila et al. (2016). Cells were harvested, adjusted to 10 µg Chl a mL−1 in fresh BG-11 medium and acclimated for 1 hr to the same experimental conditions as was applied for the cultivation.

Protein isolation, electrophoresis and immunodetection

Request a detailed protocol

Total cell extracts and the soluble fractions of Synechocystis cells were isolated as described (Zhang et al., 2009). Proteins were separated by 12% (w/v) SDS-PAGE containing 6 M urea and transferred onto a PVDF membrane (Immobilion-P; Millipore, Germany) and immunodetected by protein specific antibodies. Horseradish peroxidase (HRP) conjugated secondary antibody (anti-rabbit IgG from donkey) was used for recognizing the primary antibodies and Amersham ECL Western Blotting Detection Reagent (GE Healthcare) was used for the visualization of the antibodies.

References

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10
  11. 11
  12. 12
  13. 13
  14. 14
  15. 15
  16. 16
  17. 17
  18. 18
  19. 19
  20. 20
  21. 21
  22. 22
    Flavodiiron proteins and their role in cyanobacteria
    1. VL Gonçalves
    2. JB Vicente
    3. LM Saraiva
    4. M Teixeira
    (2011)
    In: G Peschek, C Obinger, G Renger, editors. Processes of Cyanobacteria. Dordrecht: Springer. pp. 631–653.
    https://doi.org/10.1007/978-94-007-0388-9_22
  23. 23
  24. 24
  25. 25
  26. 26
  27. 27
  28. 28
  29. 29
  30. 30
  31. 31
  32. 32
  33. 33
  34. 34
  35. 35
  36. 36
  37. 37
  38. 38
  39. 39
  40. 40
  41. 41
  42. 42
  43. 43
  44. 44
  45. 45
  46. 46
  47. 47
  48. 48
  49. 49
  50. 50
  51. 51
  52. 52
  53. 53

Decision letter

  1. Jürgen Kleine-Vehn
    Reviewing Editor; University of Natural Resources and Life Sciences, Austria
  2. Detlef Weigel
    Senior Editor; Max Planck Institute for Developmental Biology, Germany
  3. Aaron Kaplan
    Reviewer; The Hebrew University of Jerusalem, Israel

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

[Editors’ note: this article was originally rejected after discussions between the reviewers, but the authors were invited to resubmit after an appeal against the decision.]

Thank you for submitting your work entitled "Flv1-4 proteins function in versatile combinations in O2 photoreduction in cyanobacteria" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by a Reviewing Editor and a Senior Editor.

The reviewers carefully evaluated your interesting work and concluded that the manuscript is currently too preliminary for publication. They particularly pointed out that data on O2 evolution would be necessary to validate your conclusions on O2 uptake. It is the policy of eLife to avoid lengthy re-review processes with uncertain outcomes. Therefore, we decided to reject the paper at this stage. Please find the specific reviewer comments below.

Reviewer #1:

The authors analyze the functions of FLVs to catalyze Mehler reaction in S.6803 using MIMS. The reviewer cannot find any novel thing of FLVs in their physiological functions. Helman et al. (2003) already reported that FLVs catalyze O2-dependent electron flow in vivo using MIMS. Furthermore, the authors in the present manuscript already reported the functions of FLV2/4 to catalyze Mehler reaction. As an important, serious issue, the authors did not show any data for 16 O2 evolution in PSII, which reflects the electron flux in photosynthetic linear electron flow in vivo. The reviewer cannot understand why the authors did not show these data. The 162-evolution rate greatly affects photosynthetic electron transport rate. Miyake'group already showed the deficiency of FLV suppresses photosynthetic linear electron transport rate, and also suppresses CO2-dependent photosynthesis rate (Shaku et al., 2016). The deficiency of FLV induces the reduction of plastoquinone pool, which suppresses Q-cycle. This fact is known as RISE (Shimakawa et al., 2018). The authors did not test the possibility of RISE to regulate the activity of Mehler reaction. That is, the 16O2 evolution rate regulates 18O2-uptake rate in PSI. Therefore, the authors should show 16O2 evolution rate in the mutants used in the present manuscript.

From these points, the reviewer does not recommend the present manuscript for the publication in eLife.

References

1) Reduction-Induced Suppression of Electron Flow (RISE) in the Photosynthetic Electron Transport System of Synechococcus elongatus PCC 7942.

Shaku K, Shimakawa G, Hashiguchi M, Miyake C. Plant Cell Physiol. 2016 Jul;57(7):1443-1453. Epub 2015 Dec 26.PMID: 26707729

2) Reduction-Induced Suppression of Electron Flow (RISE) Is Relieved by Non-ATP-Consuming Electron Flow in Synechococcus elongatus PCC 7942.

Shimakawa G, Shaku K, Miyake C. Front Microbiol. 2018 May 7;9:886. doi:0.3389/fmicb.2018.00886. eCollection 2018. PMID: 29867800

3) Comparative analysis of strategies to prepare electron sinks in aquatic photoautotrophs.

Shimakawa G, Murakami A, Niwa K, Matsuda Y, Wada A, Miyake C. Photosynth Res. 2019 Mar;139(1-3):401-411. doi: 10.1007/s11120-018-0522-z. Epub 2018 May 29. PMID: 29845382

4)Oxidation of P700 Ensures Robust Photosynthesis.

Shimakawa G, Miyake C.Front Plant Sci. 2018 Nov 6;9:1617. doi: 10.3389/fpls.2018.01617. Collection 2018. Review. PMID: 30459798

5) Diversity of strategies for escaping reactive oxygen species production within photosystem I among land plants: P700 oxidation system is prerequisite for alleviating photoinhibition in photosystem I.

Takagi D, Ishizaki K, Hanawa H, Mabuchi T, Shimakawa G, Yamamoto H, Miyake C. Physiol Plant. 2017 Sep;161(1):56-74. doi: 10.1111/ppl.12562. Epub 2017 May 24. PMID: 28295410

Reviewer #2:

This is an important, comprehensive and interesting paper worth considering for publication in this journal. Using stable O2 isotopes and a MIMS the authors examined the routes of electrons in various mutants impaired in the 4 FLV proteins in a model cyanobacteria. These FLVs were shown to be involved in photoreduction of O2.

I do have several problems with the data presented in Figure 1. The cells hardly respired (higher under HC) during the first 4 minutes in the dark compared with that observed after darkening at the end of phase III. Is the rate of dark respiration (O2 uptake during the second dark phase in Figure 1) faster in flv4-2/OE mutant? And if so, why? This appear to be the case under both LC and HC.

What do you mean by 50% WT and why are the bands showing lower MW compared with the rest?

Why is phase II missing in mutant M55? Does it mean that electrons are re-routed in the NdhB mutant?. Can you check it in mutants ndhD1/3 vs ndhD2/4? I am not requesting this experiment but it may help understand the transient shape in the WT!!

In subsection “Extent and kinetics of the Mehler-like reaction in cells acclimated to low (LC) and high Ci (HC) conditions” we read "…confirms that the Flv1 and Flv3 proteins are responsible for the Mehler-like reaction under HC condition…" I don't see these data here, is it correct the Flv1 and Flv3 contributes under HC only?

Subsection “Extent and kinetics of the Mehler-like reaction in cells acclimated to low (LC) and high Ci (HC) conditions”: you are most probably correct but unlike mutant M55 the HC grown WT cells possess some NDH1, isn't it? It is able to utilize glucose, unlike M55.

Subsection “The effect of sodium carbonate”: How much sodium carbonate was used under the various experiments including the pH conditions? Altogether, I am confused regarding the concentration of sodium under the various growth conditions. Note that sodium is essential for two of the main bicarbonate transport systems. In fact, if sufficient bicarbonate is used the cells behave like HC-grown. Thus your finding that transcript is unaffected is quit surprising unless the concentration of carbonate used was negligible. Please make sure to provide the amounts of carbonate added in each of the experiments! throughout!!

Any reason the growth rates (as depicted in Figure 4) is that slow

Subsection “The FDP induced O2 photoreduction does not occur at PSII or the PQ-pool level”: With the current technologies you did the best one can but we must remember that when you block one electron route they are channeled to the others. In other words, you can't really assess the rate of electron transport in a specific tube of this multichannel system by blocking anyone of them either by a mutation or by most inhibitors.

Figure 5: Do you possess a similar data set for cells grown and maintained under HC?

To the best of my knowledge, this is the first report where the double flv1/3 mutant shows light-dependent O2 uptake.

Subsection “Growth history of cells has long-term consequences on the Mehler-like reaction”: It is not surprising that the growth conditions strongly affected the data but what you really show is that the cells you used were not fully acclimated to LC, with serious implications on the interpretations of the data!! in other words the experiments were conducted under wrong growth conditions? How much bicarbonate was present in the experiment presented in Figure 6?

I am a bit lost regarding the discrepancy between the in vivo and in vitro data. When expressed in E. coli FLV3 shows NADPH-dependent O2 uptake. Since this is a 4 electron business it has got to be a homodimer. But in vivo, a flv1 mutant is essentially unable to show light-dependent O2 uptake. Why can't it form the homodimer in vivo? Can FLV2 or 4 substitute for Flv1 in flv1 or flv3 mutants? Apparently not. Do you see a conservative NADPH binding motif in FLV2 or 4 or does it rely on NADH under LC?

What sets the kinetic limitation on the FLVs to avoid NADPH dissipation and hence inhibition of CO2 fixation? Any idea?

How about the possibility that the cyanobacterial PGR5 function differently that in green algae? And that it indeed plays a role here? Do you possess the needed mutant to be placed on the MIMS?

Is it possible that under the redox pressure in cells experiencing CO2 limitation a cyclic ET is activated within PSII?

Reviewer #3:

The present manuscript investigated the function of flavodiiron proteins (FDPs) in Synechocystis sp. PCC 6803. This organism possesses four FDPs (Flv1-4) essential for photoprotection of photosynthesis. The authors conclude that Flv2/Flv4 contribute to the Mehler-like reaction under defined conditions. Moreover, it is suggested that O2 photoreduction via Flv2/Flv4 occurs down-stream of PSI in a coordinated manner with Flv1/Flv3. The data as provided by the authors appear to support their conclusions.

However, a fundamental problem with this manuscript is that only O2 uptake rates are provided. As rate and extent of O2 photoreduction/uptake is also dependent on the capacity of light-driven photosynthetic electron transfer leading to O2 evolution, also O2 evolution rates need to be specified for the different conditions and mutants by using MIMS. For the moment it is not possible to exclude that differences observed in O2 photoreduction rates are simply due to difference in capacity of photosynthetic electron transfer, which in turn affects O2 uptake.

It would be important to determine the ratio of O2 evolution versus O2 uptake for the different conditions and mutants.

[Editors’ note: what now follows is the decision letter after the authors submitted for further consideration.]

Thank you for submitting your article "Flavodiiron proteins 1-to-4 function in versatile combinations in O2 photoreduction in cyanobacteria" for consideration by eLife. Your manuscript has been reviewed by two peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Detlef Weigel as the Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Michael Hippler (Reviewer #3).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

The reviewers are still very reserved and asked for further clarifications. They are particularly worried about the low O2 evolution rate of M55 and propose O2 evolution measurement in phase1 to substantiate the drawn conclusions. They are also generally concerned about the M55 mutant background, because the original mutant cannot grow under low C O2 levels, but does so under your condition.

Please, have also a look at the specific reviewer comments below, which may guide you to revise your manuscript.

Reviewer #2:

You may remember/note that my criticism of the earlier version was the modest among the three reviewers and that I recommended to accept their appeal. But when I started reading the new version I came across, in the fifth paragraph of the Results, that mutant M55 was grown under low level of CO2. This isn't a typo error – in their reply to the reviewers we read:

Q: You are most probably correct but unlike mutant M55 the HC grown WT cells possess some NDH1, isn't it? It is able to utilize glucose, unlike M55.

A: Yes, that is right, we are sorry for non-clarity. We have modified the sentence.

This mutant, originally raised by Teruo Ogawa many years ago, was characterized by its high CO2 requiring phenotype. It can only grow under very high CO2 level. One may grow it under high CO2 and then exposed to a lower one, this is fine. This could be the case in the earlier version though not define, but this is clearly not the case here. Possibly, under the stressing conditions it went through suppression mutations and is now able to grow under low CO2 level. You may ask the authors whether it is also able to grow on glucose (the original M55 mutant can't). The authors must re-sequence the mutant and find the suppression mutation. We can't proceed evaluating the manuscript unless we know what the nature of the mutation was and unless they compare the data shown here with those obtained in a high-CO2 requiring M55 mutant, they may have to construct it again.

Further, unlike earlier studies it is proposed that FLV2 and 4 take part in light-dependent O2 uptake. In view of the possibility that their growth conditions promoted accumulation of secondary mutations I examined the sequences of flv2 and 4 in cyanobase to find that it is unlikely that NADPH could bind there, questioning their conclusion. I am not clear whether they actually tested NADPH-dependent O2 uptake in vitro, shown to be very low in a study from another group. Given the in vitro results in the literature it is unlikely that FLV2 and 4 really take part. If they do I would test the sequence here too.

Altogether, while I can't recommend further assessment of the present version I strongly recommend resequencing of M55 mutant and their WT, it may end up with exciting results.

Reviewer #3:

Some aspects regarding the revised manuscript need to be addressed.

In Figure 1—source data 2, column G, the gross O2 uptake rate (µmol O2.mg Chla -1.hr-1) is shown. In the manuscript the authors refer to the gross O2 evolution rate, although this is not shown in this figure. Is it possible that instead of gross O2 uptake uptake, gross O2 evolution rates are shown in column G? This needs to be clarified.

Notable, the M55 mutant has a significantly reduced gross O2 evolution rate (considering, column G as gross O2 evolution rate) compared to WT, thus changes in O2 uptake rate might be linked to electron transfer capacity. Here it would be important to show O2 evolution in phase1.

HQNO also blocks b6f complexes. This should be considered in the interpretation of the results.

Looking at gross O2 evolution rates (considering, column G as gross O2 evolution rate) for Figure 5, ∆flv1/∆flv3 as well as ∆flv4 have a significant lower O2 evolution rates under 1000 and 1500 uE despite the fact that O2 uptake is compromised. Thus indicating that the absence of ∆flv1/∆flv3 as well as of ∆flv4 impacts O2 evolution. This should be considered in the respect of differences in O2 uptake and in regard to protection of the photosynthetic machinery. In this line, also data for PSII and PSI functionality at the different light intensities should be presented.

https://doi.org/10.7554/eLife.45766.025

Author response

The editorial decision letter and the comments of reviewer #1 and reviewer #3 clearly show that the main reason for a rejection of the manuscript is the absence of gross O2 production data. In fact, we have simultaneously monitored the O2 uptake and O2 production during all the MIMS experiments. We do have figures done depicting the O2 production data, which can be immediately incorporated into the manuscript. I would like to mention that, contrary to the assumptions of the reviewer #1, the O2 evolution data will not change the main conclusions and interpretation of the presented results, since no dramatic differences were observed between the wild-type and the FDP mutants. This was the main reason why we didn’t present the O2 evolution data, and instead, tried to focus primarily on the Mehler-like reaction. However, now we agree that presenting the gross O2 production would be necessary for a validation of the presented O2 uptake results.

[Editors’ note: the author responses to the first round of peer review follow.]

Reviewer #1:

The authors analyze the functions of FLVs to catalyze Mehler reaction in S.6803 using MIMS. The reviewer cannot find any novel thing of FLVs in their physiological functions. Helman et al. (2003) already reported that FLVs catalyze O2-dependent electron flow in vivo using MIMS. Furthermore, the authors in the present manuscript already reported the functions of FLV2/4 to catalyze Mehler reaction.

This is a wrong statement. There is no in vivo evidence on Flv2/Flv4 contribution to O2 photoreduction (Mehler-like reaction) in Helman et al., 2003. Moreover, we have not published nor hypothesized that Flv2/4 could be involved in the Mehler-like reaction (if this is what the reviewer criticizes with his confusing comment above “…the authors in the present manuscript already reported the functions of FLV2/4 to catalyze Mehler reaction”, and what we indeed are reporting now in the present manuscript). As to the Helman et al. paper, it describes O2 photoreduction in experimental cultures grown under high CO2 level (HC, 5% CO2 bubbled in air), where the expression of the flv4-flv2 operon is strongly downregulated (Zhang et al., 2009; Hackenberg et al., 2012, and Figure 1C in the present paper). Applying proper growth conditions (cultures grown for 4 days under air levels of CO2) and using different Flv mutants in this work we provide for the first-time evidence indicating the in vivo contribution of Flv2/Flv4 heterodimer to the Mehler-like reaction in Synechocystis sp. PCC 6308.

As mentioned in our manuscript, the group of Miyake has published a paper suggesting that Flv4 catalyzes O2 uptake but provided no direct evidence (Shimakawa et al., 2015). The suggestion was based on in vitro assay which demonstrated only residual O2 uptake activity of the recombinant Flv4 (20 min-1vs 40 s-1 from Giardia) and showed no affinity to NADPH but only to NADH.

As an important, serious issue, the authors did not show any data for 16O2 evolution in PSII, which reflects the electron flux in photosynthetic linear electron flow in vivo. The reviewer cannot understand why the authors did not show these data. The 162-evolution rate greatly affects photosynthetic electron transport rate. Miyake'group already showed the deficiency of FLV suppresses photosynthetic linear electron transport rate, and also suppresses CO2-dependent photosynthesis rate (Shaku et al., 2016). The deficiency of FLV induces the reduction of plastoquinone pool, which suppresses Q-cycle. This fact is known as RISE (Shimakawa et al., 2018). The authors did not test the possibility of RISE to regulate the activity of Mehler reaction. That is, the 16O2 evolution rate regulates 18O2-uptake rate in PSI. Therefore, the authors should show 16O2 evolution rate in the mutants used in the present manuscript.

We agree with both reviewers (reviewer #1 and #3) that the presentation of the gross O2 production would be necessary to show for validation of the O2 reduction results. In fact, we always monitor simultaneously the O2 production and O2 consumption rates during all the MIMS experiments. Since no drastic differences in the O2 evolution rates were observed between the wild-type and the Flv mutants, we did not present the O2 production data, and instead, tried to focus primarily on the Mehler-like reaction. However, we completely agree with the reviewers that, even though the O2 production rates were found not to change, this data should be added to the manuscript in order to facilitate the interpretation of obtained O2 photoreduction data and the conclusions made in the manuscript. We have now added this data and discuss the gross O2 values in the WT and the Flv mutant strains (Results section and the Figure 1—source data 2).

About RISE. We have used, as reviewer # 2 stated, “the best available technique” to study the Mehler-like reaction. The in-direct methods (e.g. RISE) used by Miyake team, cannot precisely address the scientific question related to the Mehler-reaction and, thus, are not commonly accepted by the scientific community. As an example “the RISE method” suggested by the Miyake team (Shaku et al., 2016)has so far been cited in 8 PubMed articles, and 7 of them are self-citations.

Reviewer #2:

This is an important, comprehensive and interesting paper worth considering for publication in this journal. Using stable O2 isotopes and a MIMS the authors examined the routes of electrons in various mutants impaired in the 4 FLV proteins in a model cyanobacteria. These FLVs were shown to be involved in photoreduction of O2.

I do have several problems with the data presented in Figure 1. The cells hardly respired (higher under HC) during the first 4 minutes in the dark compared with that observed after darkening at the end of phase III. Is the rate of dark respiration (O2 uptake during the second dark phase in Figure 1) faster in flv4-2/OE mutant? And if so, why? This appear to be the case under both LC and HC.

That’s right, higher post-illumination O2-uptake in the dark is a well-known phenomenon and can be observed also with Clark-type electrode during the measurements of O2 exchange in WT. Similar phenomenon occurs in all the studied Flv mutants, including flv4-2OE. This might be related to a light-induced activation of the respiratory pathway. However, the current work focuses on the activity of Flvs in O2 reduction occurring during illumination thus, the increase of post-illumination O2 uptake rate was not discussed in the text.

What do you mean by 50% WT and why are the bands showing lower MW compared with the rest?

In Figure 1C, “50% WT” corresponds to 1:2 diluted WT total protein sample (20 μg of total protein is normally loaded (100%) and 50% correspond to 10 μg total protein loaded). The dilution of protein samples is included in all the immunoblot figures (Figure 1, Figure 3 and 6B) to show the linear range of the detection for the specific antibodies used. Slightly lower apparent MW is related to less protein loaded. Now we have clarified this point in the corresponding Figure legends.

Why is phase II missing in mutant M55? Does it mean that electrons are re-routed in the NdhB mutant?. Can you check it in mutants ndhD1/3 vs ndhD2/4? I am not requesting this experiment but it may help understand the transient shape in the WT!!

In the current work we hypothesize that during illumination, there is a competition between Flvs and the NDH-1 complex for the photosynthetic electrons. We speculate that the decay of O2 photoreduction rates in phase II corresponds to the redirection of electrons towards NDH-1, which consequently, results in a decrease of the Flv-driven O2 photoreduction. This hypothesis is supported by the constant O2 photoreduction in the M55 mutant, where no active NDH-1 complex competes for electrons with the Flvs, thus the Flv-driven O2 photoreduction does not decay during illumination and no phase II is observed. We have rephrased the text in Results section paragraph four to facilitate the reading. Please also see subsection “Under LC, the Flv1/Flv3 heterodimer is a rapid, strong and transient electron sink whereas Flv2/Flv4 supports steady-state O2 photoreduction”).

The ∆ndhD1/D2 mutant, lacking the NDH-1L complex involved in CET, shows only a small decay (phase II) followed by constant O2 photoreduction (phase III) during illumination of 5 min or 10 min. The data is not included in the paper and is a subject of a new manuscript.

In subsection “Extent and kinetics of the Mehler-like reaction in cells acclimated to low (LC) and high Ci (HC) conditions” we read "…confirms that the Flv1 and Flv3 proteins are responsible for the Mehler-like reaction under HC condition…" I don't see these data here, is it correct the Flv1 and Flv3 contributes under HC only?

Flv1/3 contributes to Mehler reaction also under low CO2, being responsible for the fast O2 photoreduction. Synechocystis cells grown under high CO2 levels (HC) have shown suppressed expression of the flv4-flv2 operon and strong downregulation of flv3 at the transcript and protein levels (Zhang et al., 2009; Hackenberg et al., 2012 and Figure 1C in this work). Therefore, in the absence of Flv2/Flv4 heterodimer, the constant O2 photoreduction rate observed in the WT cells grown under high CO2 will be mainly carried out by the small amount of Flv1/Flv3 heterodimers (decreased Flv3 accumulation in HC compared to LC conditions is shown in Figure 1C). This is supported by the drastic decrease of O2 photoreduction in the ∆flv1/3 and ∆flv3/4 mutants grown under HC conditions (see Figure 1—figure supplement 1). We have slightly reformulated the text to facilitate the reading (see subsection “Extent and kinetics of the Mehler-like reaction in cells acclimated to low (LC) and high Ci (HC) conditions”).

Subsection “Extent and kinetics of the Mehler-like reaction in cells acclimated to low (LC) and high Ci (HC) conditions”: you are most probably correct but unlike mutant M55 the HC grown WT cells possess some NDH1, isn't it? It is able to utilize glucose, unlike M55.

Yes, that is right, we are sorry for non-clarity. We have modified the sentence.

Subsection “The effect of sodium carbonate”: How much sodium carbonate was used under the various experiments including the pH conditions? Altogether, I am confused regarding the concentration of sodium under the various growth conditions. Note that sodium is essential for two of the main bicarbonate transport systems. In fact, if sufficient bicarbonate is used the cells behave like HC-grown. Thus your finding that transcript is unaffected is quit surprising unless the concentration of carbonate used was negligible. Please make sure to provide the amounts of carbonate added in each of the experiments! throughout!!

Thank you for the comment. In this work we used standard BG11 medium containing sodium carbonate (Na2CO3) at a final concentration of 0.189 mM for all experimental cultures. Only when indicated, strong limitation of inorganic carbon was achieved by omitting Na2CO3 from the BG11 media. This is now indicated in the Figure legends and in the main text more clearly. Sodium nitrate (NaNO3) macronutrient is the main source of sodium (Na+) in BG11 media and it is present at a final concentration of 17.6 mM in all experimental cultures in this work. Therefore, the cultures grown in BG11 media without sodium carbonate should not experience strong limitation of sodium.

In this work, we demonstrate that the concentration (at LC and HC conditions) and distribution of dissolved inorganic carbon species (DIC) in the medium (dependent on the pH) are essential factors determining the expression and/or function of Flv proteins in Synechocystis sp. PCC 6308. The stronger limitation of inorganic carbon by omitting 0.189 mM sodium carbonate from cultures grown in BG-11 at pH 7.5 under air levels of CO2 affected neither the expression nor accumulation of Flv proteins, however, this limitation was enough to enhance the O2 photoreduction activity by the Flv proteins, as shown in Figure 2.

We agree that the amounts of carbon must be repeated in each Figure legend to make it easy to read and follow the manuscript. As recommended, we have mentioned the concentration of sodium carbonate present in the BG11 medium used in this work throughout the text.

Any reason the growth rates (as depicted in Figure 4) is that slow

Figure 4A and 4B depict growth at low light background (20 μ µmol photons m-2 s-1 background light regularly punched with 30 sec 500 μ µmol photons m-2 s-1 pulse) therefore slow growth is not a surprise. At 50 μ µmol photons m-2 s-1 background the cells grow “normally” (at least in our lab conditions with our WT cells) reaching OD750 = 2 after 4 days at pH 9. The slow growth observed at pH 6 is due to the low solubility of CO2 at acidic pH.

Subsection “The FDP induced O2 photoreduction does not occur at PSII or the PQ-pool level”: With the current technologies you did the best one can but we must remember that when you block one electron route they are channeled to the others. In other words, you can't really assess the rate of electron transport in a specific tube of this multichannel system by blocking anyone of them either by a mutation or by most inhibitors.

We completely agree with the reviewer. The closest we can get to reality with deletion mutants is that we assess the replaceability of certain electron transport routes and redox components. The more profound the effect, the more important the component is. With the right setup, we can map which pathway substitutes for the deleted one and determine a condition-specific hierarchy.

Figure 5: Do you possess a similar data set for cells grown and maintained under HC?

According to the suggestion of the reviewer, we have performed a new experiment with HC grown cells. The data is added to Figure 5—figure supplement 1.

To the best of my knowledge, this is the first report where the double flv1/3 mutant shows light-dependent O2 uptake.

Subsection “Growth history of cells has long-term consequences on the Mehler-like reaction”: It is not surprising that the growth conditions strongly affected the data but what you really show is that the cells you used were not fully acclimated to LC, with serious implications on the interpretations of the data!! in other words the experiments were conducted under wrong growth conditions?

Not only acclimation to low carbon but also light intensity plays a role in accumulation level of Flvs. Rather than “wrong growth conditions”, we consider that in our previous studies, the accumulation of Flv2 and Flv4 in the cells was not enough to detect their O2 photoreduction activity using our MIMS setup. In the current work, we were able to induce higher accumulation of Flvs by reducing the inoculum size of the experimental cultures (allowing more light penetration) and extending the LC acclimation to 4 days. The increased accumulation of Flvs in the cells allowed us to demonstrate, for the first time, that O2 photoreduction is carried out by Flv2 and Flv4 proteins in the ∆flv1/3 mutant cells. We have reformulated the text to facilitate the reading.

How much bicarbonate was present in the experiment presented in Figure 6?

All the growth experiments were done in standard BG-11 medium containing sodium carbonate (Na2CO3) at a final concentration of 0.189 mM (including the cells studied in Figure 6). Prior to the MIMS measurements (including Figure 6), the cells were supplemented with 1.5 mM of sodium bicarbonate (NaHCO3). This was important to ensure comparable conditions between measurements with MIMS. Independent experiments performed on WT cells grown in the absence of Na2CO3, but supplied with 1.5 mM NaHCO3 prior to the MIMS measurement showed no significant difference in O2 photoreduction rates (see the Figure 1—figure supplement 2), and thus should not affect the interpretation of the results. As recommended, the concentration of carbonate/bicarbonate is mentioned throughout the text (see legend of all Figures (1-6).

I am a bit lost regarding the discrepancy between the in vivo and in vitro data. When expressed in E. coli FLV3 shows NADPH-dependent O2 uptake. Since this is a 4 electron business it has got to be a homodimer. But in vivo, a flv1 mutant is essentially unable to show light-dependent O2 uptake. Why can't it form the homodimer in vivo? Can FLV2 or 4 substitute for Flv1 in flv1 or flv3 mutants? Apparently not. Do you see a conservative NADPH binding motif in FLV2 or 4 or does it rely on NADH under LC?

Two in vitro studies have been performed on recombinant Synechocystis Flv3 or Flv4 proteins purified after expression in E. coli (Vicente et al., 2002, Shimakawa et al., 2015). Both studies claim that Flv3 (Vicente et al., 2002) or Flv4 (Shimakawa et al., 2015) function as a homo-dimer in O2 reduction. However, in both cases the enzyme activity is only residual (0.38 min-1 for Flv3, 20 min-1 for Flv4 compared to 40 s-1 of the Flv activity in Giardia). Moreover, studies have shown either little (Vicente et al., 2002) or no affinity (Shimakawa et al., 2015) to NADPH, a reducing equivalent produced by photosynthetic linear electron transport. Many years of work in our laboratory and in several other laboratories have shown that the isolation of functional Flvs from photosynthetic organisms is an extremely difficult task and so far none of the laboratories has managed to isolate highly active Flvs from photosynthetic organisms. Therefore, we believe that in vitro experiments were not reflecting a real situation. in vivo experiments clearly demonstrate that homooligomers cannot function in O2 uptake (Mustila et al., 2016).

Since, no crystal structures of the Flvs from oxygen evolving photosynthetic organisms are available (except recent structure of the truncated Flv1 lacking a flavodoxin-like domain, Borges et al., 2019), it is not possible to confidently conclude about the NADH or NAPDH preference just based on the amino acid sequence and homology searches.

What sets the kinetic limitation on the FLVs to avoid NADPH dissipation and hence inhibition of CO2 fixation? Any idea?

This is a key question and so far, an unanswered one. One possibility is post-transcriptional modification of the protein, such as phosphorylation (Angeleri et al., 2016) and/or pmf based regulation.

How about the possibility that the cyanobacterial PGR5 function differently that in green algae? And that it indeed plays a role here? Do you possess the needed mutant to be placed on the MIMS?

This is a logical question. In future we will probe also this mutant. However, in this work we concentrate on dissection of Flv1/3 and Flv2/4-originated O2 photoreduction and Ci-dependency of O2 photoreduction.

Moreover, Synechocystis Pgr5 deletion mutant, different from Chlamydomonas Pgr5 knockout mutant (Jokel et al., 2018), do not show any phenotype under fluctuating light intensities (Allahverdiyeva et al., 2013). Therefore we assume that in Synechocystis a strong crosstalk between these proteins is missing, at least under conditions studied so far.

Is it possible that under the redox pressure in cells experiencing CO2 limitation a cyclic ET is activated within PSII?

Unfortunately, we cannot answer this interesting question.

Reviewer #3:

The present manuscript investigated the function of flavodiiron proteins (FDPs) in Synechocystis sp. PCC 6803. This organism possesses four FDPs (Flv1-4) essential for photoprotection of photosynthesis. The authors conclude that Flv2/Flv4 contribute to the Mehler-like reaction under defined conditions. Moreover, it is suggested that O2 photoreduction via Flv2/Flv4 occurs down-stream of PSI in a coordinated manner with Flv1/Flv3. The data as provided by the authors appear to support their conclusions.

However, a fundamental problem with this manuscript is that only O2 uptake rates are provided. As rate and extent of O2 photoreduction/uptake is also dependent on the capacity of light-driven photosynthetic electron transfer leading to O2 evolution, also O2 evolution rates need to be specified for the different conditions and mutants by using MIMS. For the moment it is not possible to exclude that differences observed in O2 photoreduction rates are simply due to difference in capacity of photosynthetic electron transfer, which in turn affects O2 uptake.

It would be important to determine the ratio of O2 evolution versus O2 uptake for the different conditions and mutants.

We agree with the reviewer that the presentation of the gross O2 production would be necessary for validation of the O2 photoreduction results. We have simultaneously monitored both the O2 photoreduction and O2 production during all the MIMS experiments and the data have been added to the revised manuscript. Importantly, no significant difference in the gross O2 evolution rates was observed between the wild-type and the FDP mutants.

[Editors’ note: the author responses to the re-review follow.]

The reviewers are still very reserved and asked for further clarifications. They are particularly worried about the low O2 evolution rate of M55 and propose O2 evolution measurement in phase1 to substantiate the drawn conclusions. They are also generally concerned about the M55 mutant background, because the original mutant cannot grow under low CO2 levels, but does so under your condition.

Please, have also a look at the specific reviewer comments below, which may guide you to revise your manuscript.

Reviewer #2:

You may remember/note that my criticism of the earlier version was the modest among the three reviewers and that I recommended to accept their appeal. But when I started reading the new version I came across, in the fifth paragraph of the Results, that mutant M55 was grown under low level of CO2. This isn't a typo error – in their reply to the reviewers we read:

Q: You are most probably correct but unlike mutant M55 the HC grown WT cells possess some NDH1, isn't it? It is able to utilize glucose, unlike M55.

A: Yes, that is right, we are sorry for non-clarity. We have modified the sentence.

This mutant, originally raised by Teruo Ogawa many years ago, was characterized by its high CO2 requiring phenotype. It can only grow under very high CO2 level. One may grow it under high CO2 and then exposed to a lower one, this is fine. This could be the case in the earlier version though not define, but this is clearly not the case here. Possibly, under the stressing conditions it went through suppression mutations and is now able to grow under low CO2 level. You may ask the authors whether it is also able to grow on glucose (the original M55 mutant can't). The authors must re-sequence the mutant and find the suppression mutation. We can't proceed evaluating the manuscript unless we know what the nature of the mutation was and unless they compare the data shown here with those obtained in a high-CO2 requiring M55 mutant, they may have to construct it again.

We understand the concern of the reviewer 2 related to the background of the strains. Indeed, in the original paper it was shown that M55 cannot grow under air level CO2 (Ogawa T. 1991 “A gene homologous to the subunit-2 gene of NADH dehydrogenase is essential to inorganic carbon of Synechocystis PCC 6803”, PNAS 88:4275–427). However, in this paper the growth experiment was performed with very diluted cell culture (OD730=0.03) and at relatively high light intensity (120 umol m-2 s-1). It is not surprising that M55, the mutant strain deficient in important bioenergetic processes such as Ci assimilation, respiration and cyclic electron transfer, shows strong growth retardation under such stressful condition.

Most importantly, the same author later published 2 papers where he precisely described the growth characteristics of the M55mutant. In both papers it is clearly shown that the M55 mutant is able to grow (at a slower rate compared to WT) under air level CO2 conditions (pH 8.0) in liquid medium and on BG-11 agar plates under moderate light intensity of 60 umol m-2 s-1 (see Ohkawa, Price, Badger, Ogawa “Mutation of ndh genes leads to inhibition of CO2 uptake rather than HCO3−uptake in Synechocystis sp. strain PCC6803”, 2000, Journal of Bacteriology p. 2591–2596, in Figure 3 and Figure 4; Ohkawa, Pakrasi, Ogawa ‘Two types of functionally distinct NAD(P)H Dehydrogenases in Synechocystis sp. strain PCC6803’ 2000 JBC 275, p. 31630-4; in Figure 1).

There are a bunch of papers where M55 was grown under air level CO2, high pH conditions. In our lab, we routinely cultivate M55 at pH 8.2 at air level CO2 (e.g. Zhang et al., 2004; Zhang et al., 2012).

Following the suggestion of the reviewer 2, we performed a growth trial with M55 in the presence of 5mM glucose + 10uM DCMU. As expected, M55 was unable to grow under photoheterotrophic conditions (Author response image 1), which is in line with the previous observations of photoheterotrophic growth deficiency of M55 (Ohkawa et al., 2000b in Figure 1, Zhao et al., 2015 in Figure 5). Moreover, we have cultivated the M55 strain at air-level CO2 at pH 8.2 and we were able to reproduce the results published in Ohkawa et al., 2000a, where M55 demonstrates slower growth compared to wildtype (Author response image 1).

Author response image 1
Photoheterotrophic growth of wild-type (WT) and M55 cells on BG-11 agar plates.

The WT and M55 mutant cells of Synechocystis were resuspended in BG-11 median at pH8.2. Three microliters of cell suspensions at OD750nm of 0.1 (top row), 0.01 (middle row), and 0.001 (bottom row) was spotted on agar plates. Five millimolars of Glucose and 10 μm DCMU were added to the plates for photoheterotrophic growth (right side) or were not added for photoautotrophic growth (left side). The plates were grown under ambient air for 8d at 50 μmol photons m-2 S-1.

In light of this, we respectfully disagree that resequencing of M55 mutant is necessary for this manuscript. Considering the fact that M55 was generated decades ago, there is no doubt that many other mutations have likely occurred during the cultivation but in respect of Ci assimilation and respiration, which is the focus of this manuscript, we still observe the original phenotype.

Further, unlike earlier studies it is proposed that FLV2 and 4 take part in light-dependent O2 uptake. In view of the possibility that their growth conditions promoted accumulation of secondary mutations I examined the sequences of flv2 and 4 in cyanobase to find that it is unlikely that NADPH could bind there, questioning their conclusion. I am not clear whether they actually tested NADPH-dependent O2 uptake in vitro, shown to be very low in a study from another group. Given the in vitro results in the literature it is unlikely that FLV2 and 4 really take part. If they do I would test the sequence here too.

The reason why earlier studies did not link the Flv2 and Flv4 proteins to the light-induced O2 uptake is simple: the Flv2 and Flv4 deletion mutants were not probed by MIMS (previous works from our lab) or high CO2 grown mutant cells were used in the experiments (Helman et al., 2003). Since, under high CO2 conditions the flv4-sll0218-flv2 operon is strongly downregulated, the similar phenotype between WT and the Flv2 and Flv4 deletion mutants is logical (Helman et al., 2003 and in this manuscript in Figure 1—figure supplement 1). We believe that re-sequencing here is unnecessary. Involvement of Flv2 and Flv4 in the light induced O2 uptake is solid: the ΔFlv2/4 lacks light-induced O2 uptake, whereas the Flv2/4 overexpression strain demonstrates significantly increased O2 uptake (see Figure 1 in this manuscript). In our lab we are well aware that secondary mutations frequently occur in cyanobacteria. In order to keep the original genome pool, we maintain our strains in cryogenic freezers, carefully avoid long-term handling and frequently start new cultures from the original cryo-stored stocks.

As reviewer pointed out “…it is proposed that FLV2 and 4 take part in light-dependent O2 uptake”. Indeed, this is one of the main messages of this paper and how we revealed the involvement of Flv2 and Flv4 in light-induced O2 uptake in vivo, is thoroughly discussed and explained. Therefore, we firmly believe that our observations do not require further supporting experiments.

Confidentconclusions about the absence or presence of NAD(P)H-binding sites, based solely on the amino acid sequence, are dubious. Well defined tertiary structural elements (see https://www.ebi.ac.uk/interpro/entry/IPR016040) form the NAD(P)H-binding motif and in the coordinative binding of NADPH, several amino acid residues take part all along the polypeptide. However, based on secondary structural element prediction and sequence conservation of NAD(P)H-binding proteins, attempts were made to predict NAD(P)H-binding (http://crdd.osdd.net/raghava/nadbinder/)and based on that, Flv4 might be able to bind NAD.

Importantly, both Flv4 and Flv2 conserve a Flavin reductase domain (IPR002563) at the C-termini, a domain found in NAD(P)H-flavin oxidoreductases. Moreover, the sequence similarity of Flv4 to Flv3 and of Flv2 to Flv1 (62% and 71% similarity, respectively) implies that Flv2/4 carries out similar reactions to Flv1/3. Besides, Flv3 and Flv1 conserve the same Flavin reductase domain as Flv2, and Flv4 (IPR002563) at the C-termini.

However, in the manuscript we do not make conclusion that Flv2/4 accepts electrons from NADPH. We hypothesize that Flv2/4 takes electrons from Fed or FNR (see Figure 7 in this manuscript) but for making such a strong conclusion, the in vitro assay with the recombinant proteins would be necessary. Unfortunately, we do not have functional recombinant protein (see our previous answer to the reviewers and the manuscript) therefore we cannot perform in vitro assays. Anyhow, the exact electron donor of Flv2/4 would only refine and not contradict our conclusions.

Reviewer #3:

Some aspects regarding the revised manuscript need to be addressed.

In Figure 1—source data 2, column G, the gross O2 uptake rate (µmol O2.mg Chla -1.hr-1) is shown. In the manuscript the authors refer to the gross O2 evolution rate, although this is not shown in this figure. Is it possible that instead of gross O2 uptake uptake, gross O2 evolution rates are shown in column G? This needs to be clarified.

We thank the reviewer for pointing out this mistake. Indeed, there was a typo in the title of column G in Figure 1—source data 2. We have now corrected gross O2 uptake to gross O2 evolution.

Notable, the M55 mutant has a significantly reduced gross O2 evolution rate (considering, column G as gross O2 evolution rate) compared to WT, thus changes in O2 uptake rate might be linked to electron transfer capacity. Here it would be important to show O2 evolution in phase1.

This is a correct observation. Following the reviewer's suggestion, we have added gross O2 evolution data to Figure 1—source data 2. During the dark-to-light transition (Phase I), M55 demonstrates a slow induction of O2 evolution. Indeed, in M55, the light-induced O2 uptake is higher, whereas gross O2 evolution is notably lower, compared to WT grown under the same condition. This is mentioned in a revised manuscript.

HQNO also blocks b6f complexes. This should be considered in the interpretation of the results.

We thank the reviewer for this information. We have mentioned the inhibitory effect of HQNO to Cytb6f in the revised manuscript. However, the inhibitory effect of HQNO does not affect our conclusions, since it is used in combination with DBMIB, which is also a known inhibitor of Cytb6f.

Looking at gross O2 evolution rates (considering, column G as gross O2 evolution rate) for Figure 5, ∆flv1/∆flv3 as well as ∆flv4 have a significant lower O2 evolution rates under 1000 and 1500 uE despite the fact that O2 uptake is compromised. Thus indicating that the absence of ∆flv1/∆flv3 as well as of ∆flv4 impacts O2 evolution. This should be considered in the respect of differences in O2 uptake and in regard to protection of the photosynthetic machinery. In this line, also data for PSII and PSI functionality at the different light intensities should be presented.

Involvement of Flv2 and Flv4 in photoprotection of PSII complex has been previously reported (Zhang et al., 2009, Hakkila et al., 2013, Bersanini et al., 2014, Chukhutsina et al., 2015, Bersanini et al., 2017), whereas the Flv1 and Flv3 proteins are linked mainly to photoprotection of PSI under fluctuating-light intensities (Allahverdiyeva et al., 2013). Following the reviewer's suggestion, we have performed new experiment, where the mutant strains grown at moderate light (50 µmol photons m-2s-1, same condition used in this work) were exposed to high-light (1500 µmol photons m-2 s-1) for 2 hours. PSII activity was probed as O2 evolution in the presence of artificial electron acceptor, DMBQ and PSI was evaluated as maximum oxidazable P700, Pm (Figure 5—figure supplement 2). In the revised manuscript we shortly mention about the photoprotective role of FLVs under high light intensities.

https://doi.org/10.7554/eLife.45766.026

Article and author information

Author details

  1. Anita Santana-Sanchez

    Molecular Plant Biology, Department of Biochemistry, University of Turku, Turku, Finland
    Contribution
    Conceptualization, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing—original draft, Writing—review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-1556-0321
  2. Daniel Solymosi

    Molecular Plant Biology, Department of Biochemistry, University of Turku, Turku, Finland
    Contribution
    Conceptualization, Validation, Investigation, Methodology, Writing—original draft, Writing—review and editing
    Competing interests
    No competing interests declared
  3. Henna Mustila

    Molecular Plant Biology, Department of Biochemistry, University of Turku, Turku, Finland
    Present address
    Department of Chemistry–Ångström Laboratory, Uppsala University, Uppsala, Sweden
    Contribution
    Conceptualization, Investigation, Writing—review and editing
    Competing interests
    No competing interests declared
  4. Luca Bersanini

    Molecular Plant Biology, Department of Biochemistry, University of Turku, Turku, Finland
    Present address
    Biophysics of Photosynthesis, Department of Physics and Astronomy, Vrije Universiteit Amsterdam, Amsterdam, The Netherlands
    Contribution
    Conceptualization, Investigation, Writing—review and editing
    Competing interests
    No competing interests declared
  5. Eva-Mari Aro

    Molecular Plant Biology, Department of Biochemistry, University of Turku, Turku, Finland
    Contribution
    Conceptualization, Resources, Funding acquisition, Writing—review and editing
    For correspondence
    evaaro@utu.fi
    Competing interests
    No competing interests declared
  6. Yagut Allahverdiyeva

    Molecular Plant Biology, Department of Biochemistry, University of Turku, Turku, Finland
    Contribution
    Conceptualization, Resources, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing—original draft, Project administration, Writing—review and editing
    For correspondence
    allahve@utu.fi
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-9262-1757

Funding

Suomen Akatemia (315119)

  • Yagut Allahverdiyeva

NordForsk (82845)

  • Eva-Mari Aro
  • Yagut Allahverdiyeva

Koneen Säätiö (7fa491)

  • Yagut Allahverdiyeva

Suomen Akatemia (307335)

  • Eva-Mari Aro

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

The authors would like to thank Dr. Lauri Nikkanen and Dr. Natalia Battchikova for critical reading of the manuscript. Dr. Duncan Fitzpatrick is acknowledged for technical assistance and maintenance of MIMS.

Senior Editor

  1. Detlef Weigel, Max Planck Institute for Developmental Biology, Germany

Reviewing Editor

  1. Jürgen Kleine-Vehn, University of Natural Resources and Life Sciences, Austria

Reviewer

  1. Aaron Kaplan, The Hebrew University of Jerusalem, Israel

Publication history

  1. Received: February 4, 2019
  2. Accepted: July 10, 2019
  3. Accepted Manuscript published: July 11, 2019 (version 1)
  4. Version of Record published: July 25, 2019 (version 2)

Copyright

© 2019, Santana-Sanchez et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 566
    Page views
  • 126
    Downloads
  • 1
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)

Further reading

    1. Biochemistry and Chemical Biology
    Zuanning Yuan et al.
    Research Article
    1. Biochemistry and Chemical Biology
    2. Cell Biology
    Leslie A McCauliff et al.
    Research Article