The regulation of photosynthesis is crucial to efficiently support the assimilation of carbon dioxide and to prevent photodamage. One key regulatory mechanism is the pseudo-cyclic electron flow (PCEF) mediated by class-C flavodiiron proteins (FLVs). These enzymes use electrons coming from Photosystem I (PSI) to reduce oxygen to water, preventing over-reduction in the acceptor side of PSI. FLVs are widely distributed among organisms performing oxygenic photosynthesis and they have been shown to be fundamental in many different conditions such as fluctuating light, sulfur deprivation and plant submersion. Moreover, since FLVs reduce oxygen they can help controlling the redox status of the cell and maintaining the microoxic environment essential for processes such as nitrogen fixation in cyanobacteria. Despite these important roles identified in various species, the genes encoding for FLV proteins have been lost in angiosperms where their activity could have been at least partially compensated by a more efficient cyclic electron flow (CEF). The present work reviews the information emerged on FLV function, analyzing recent structural data that suggest FLV could be regulated through a conformational change.
Introduction: regulation of electron transport in photosynthesis
Oxygenic photosynthesis allows plants, algae and cyanobacteria to convert sunlight into chemical energy, essential for their metabolism. The resulting reduced carbon molecules support most life forms on our planet. Photosynthesis depends on the presence of four major complexes embedded in thylakoid membranes: photosystem II (PSII), photosystem I (PSI), cytochrome b6/f (Cyt b6f) and ATP synthase (ATPase) (Figure 1). PSII, Cyt b6f and PSI catalyze a linear electron flow (LEF) from water to NADP+, also involving the plastoquinone pool (PQ), the soluble protein plastocyanin, iron-sulfur proteins called ferredoxins (Fd) and the ferredoxin-NADP-oxidoreductase (FNR). The LEF is associated with the generation of a proton motive force, including both a proton concentration component (ΔpH) and an electric field component (Δψ), that drives ATP synthesis.
Schematic representation of photosynthetic electron transport pathways.
The products of the light phase (ATP and NADPH) fuel all cellular metabolism including the Calvin–Benson–Bassham (CBB) cycle that is responsible for the synthesis of organic molecules using carbon dioxide (CO2) as substrate. The activity of most enzymes of the CBB cycle is regulated by the redox state of the chloroplast and triggered in the presence of light via the thioredoxin system . CCB cycle activation is however slower than the onset of photosynthetic electron transport and, when light intensity changes, this causes a temporary unbalance between electron transport and carbon fixation rates. In addition to light dynamics, natural environments are subjected to continuous changes in temperature, water and CO2 availability, parameters with a strong influence on CBB cycle kinetics . Any reduction in CO2 fixation rates can cause over-reduction in the electron transport chain and drive to the harmful formation of reactive oxygen species, potentially damaging photosynthetic apparatus .
Alternative electron transport pathways, such as cyclic electron flow (CEF) and pseudo-CEF (PCEF), prevent over-reduction in photosynthetic machinery by recycling or diverting electrons from LEF (Figure 1), but sustaining ATP synthesis. Both CEF and PCEF thus influence the chloroplast ATP/NADPH ratio, allowing its adjustment to the demand of CBB cycle, of carbon-concentrating mechanisms in green algae or any other metabolic pathways requiring additional ATP [4,5]. CEF mediates stromal Fd oxidation, reinjecting electrons in the membrane PQ pool sustaining the proton pumping activity of Cyt b6f. Two major CEF pathways have been identified so far. The first depends on proton gradient regulator 5 (PGR5)  and PGR5-like 1 (PGRL1) proteins . In particular, PGR5 seems to be important for the tuning of CEF  while PGRL1 was suggested to mediate the tethering of PGR5 to thylakoid membrane  and to donate electrons to quinones in vitro, acting as a ferredoxin-quinone reductase . However, the direct or indirect role of PGR5/PGRL1 in CEF is still under debate . The second CEF pathway is mediated by the chloroplast NADH dehydrogenase-like (NDH) multimeric complex . NDH shares a common origin with respiratory complex 1 [12–14] and is capable of reducing PQ using electrons from Fd thanks to a stroma exposed specific subunit . In angiosperms FNR was shown to interact with NDH complex , PSI  and Cyt b6f . A PSI-Cyt b6f super-complex including FNR was also described in the green alga Chlamydomonas reinhardtii . FNR interaction with other complexes and the modulation of its binding to thylakoid membranes could play a significant role in modulating the fate of the electrons transported by Fd .
PCEF instead uses electrons from photosynthetic electron chain for oxygen (O2) reduction to water. Thus, PCEF dissipates part of the energy with a seemingly futile water–water cycle since water is oxidized at the level of PSII and to be later re-synthetized. Different PCEF pathways have been identified so far. (i) The Mehler reaction by which the O2 is directly reduced at PSI acceptor side to form superoxide (O2−) that is rapidly converted into hydrogen peroxide (H2O2) by superoxide dismutase [20,21]. The H2O2 is then efficiently scavenged by a chloroplast-located ascorbate peroxidase avoiding oxidative damage. (ii) The plastid terminal oxidase (PTOX) is an enzyme that resides on the thylakoid membranes and catalyze the transfer of four electrons from reduced PQ to O2 to synthesize water [22,23]. (iii) The reducing power generated in the plastid can be redirected to mitochondria to sustain the respiratory electron transport chain and alternative oxidase activity (AOX), again reducing O2 to form water [24,25]. (iv) Another PCEF pathway is mediated by Class-C flavodiiron proteins (FLVs), a group of enzymes that use electrons for the reduction in O2 to water and that was first described in cyanobacteria [26–28]. This work focuses on most recent advances in FLV biological role in photosynthetic eukaryotes, reviewing recent evidences on their interaction with other alternative electron transport pathways and their mechanisms of regulation.
Flavodiiron proteins in oxygenic phototrophs
FLVs belong to a larger family of Flavodiiron proteins that are in general shortened as FDPs and are widespread in Bacteria, Archaea and Eukarya [29,30]. FDPs are particularly important for anaerobic organisms because they reduce O2 to water and/or nitric oxide to nitrous oxide, protecting cells from oxidative and nitrosative stress . FDPs have a modular organization and they all share a common core composed by two functional domains, a metallo β-lactamase-like domain harboring a non-heme diiron catalytic site and a short-chain flavodoxin-like fold with a flavin mononucleotide (FMN) moiety (Figure 2A) . FDP proteins present a differentiated C-term functional domain (e.g. Rubredoxin-like domain, iron-sulfur clusters, NAD(P)H: flavin oxidoreductase-like domain or NADH: rubredoxin oxidoreductase-like domain) that allows distinguishing the members of the superfamily into eight classes . Some members of FDP superfamily are specific either for O2 or nitric oxide, whereas others have the ability to reduce both .
Phylogenetic tree of FLV.
The class-C flavodiiron proteins are typical of photosynthetic organisms and are named FLV. Unlike the other members of the family, FLV presents a C-term NAD(P)H:flavin oxidoreductase-like domain, suggesting they could be capable of accepting electrons directly from NADH/NADPH (Figure 2A) . Data from two-hybrid system showed however potential interactions between FLV and Fd isoforms in both Synechocystis PCC6803  and C. reinhardtii . These experiments suggest that Fd might also be involved in electron transport from PSI to FLVs but conclusive enzymatic studies proving the precise nature and affinity of electron donors are still needed.
The first evidences of FLV activity in photosynthetic organisms emerged in prokaryotes and in particular in the cyanobacterium Synechocystis sp. strain PCC 6803, where 4 FLV genes were identified based on sequence homology . FLV2 and FLV4 are only found in photosynthetic prokaryotes  and they showed O2 photo-reductive activity in vitro and in vivo [40,41]. FLV2 and FLV4 according to most recent reports accept electrons from PSI and their activity is relevant especially under limiting CO2 [40,41]. Homologous sequences of FLV1 and FLV3 have also been found in the genomes of eukaryotic organisms like green algae, liverworts, mosses and gymnosperms, where they are called respectively, FLVA and FLVB . Genes encoding FLVs, however, are not found in any available angiosperm genome nor in the genomes of red and brown algae , most likely due to a secondary gene loss.
In Figure 2B, an unrooted phylogenetic tree of sequences from various photosynthetic organisms shows two clearly distinguished groups that include either FLV1/FLVA or FLV3/FLVB sequences. In each group, the sequences of prokaryotes (FLV1 and FLV3) are well separated from those of eukaryotes (FLVA and FLVB). Sequences of Chlorophyta form a subgroup that is separated from those of Streptophyta (Figure 2B). In all cases, FLV coding sequences are found in couples (FLV1/FLV3 or FLVA/FLVB), suggesting that one protein needs the other to be fully functional. In fact, in mutants depleted of either FLV1/FLVA or FLV3/FLVB there is also a drastic decrease in accumulation of the other monomer [42–44]. Experimental evidences also suggest that a hetero-oligomeric association of FLVA and FLVB, as well as FLV1 and FLV3, is required for their ability of photo-reducing O2 to water [28,42]. As example, when FLV1 or FLV3 were overexpressed in a Δflv1Δflv3 double knock-out mutant of cyanobacterium Synechocystis sp. PCC 6803 , one protein alone was unable to sustain efficient electron transfer for O2 reduction, meaning that FLV1/FLV3 hetero-oligomerization is fundamental for this activity .
In vivo measurements of PSI redox kinetics can be exploited to obtain indirect information on FLV activity , comparing WT strains and specific KO mutants (e.g. Synechocystis, Marchantia polymorpha, C. reinhardtii and Physcomitrella patens), but also in the WT by comparing measurements performed under oxic and anoxic conditions, thus removing the electron acceptor. Results from these kinds of experiments are congruent with genomic data and confirmed that FLV-like activity is present in many groups of organisms from the green lineage (green algae, bryophytes, lycophytes, ferns and gymnosperms) but not in angiosperms [46,47]. Similar experiments performed for red and brown algae highlighted the existence of interesting exceptions: in the eustigmatophytes (Heterokontophyta) Vischeria punctate, the oxidation of P700 depends on FLV-like activity , even if so far FLV sequences have never been reported in these groups of algae.
Knock-out mutants defective in FLV1 and FLV3 were impaired in O2 photoreduction even if they were able to perform photosynthesis and respiration normally . Δflv1 and Δflv3 mutants in Synechocystis showed higher limitation at the acceptor side of PSI as compared with WT that could be alleviated adding an artificial electron acceptor like methyl-viologen . In eukaryotes, the use of [18O]-labeled O2 and membrane-inlet mass spectrometry allowed measuring real-time O2 photoreduction in vivo demonstrating that in C. reinhardtii FLVs are the functional enzymes responsible for this reaction . Normally upon exposition to saturating illumination P700 is in its oxidized state (P700+) and its activity is limited at the donor side by the rate of electron transport from Cyt b6f. In vivo measurements of PSI redox kinetics showed that FLVs act as an electron sink downstream of PSI and their influence is particularly evident under ﬂuctuating light where they promote a decrease in the acceptor side limitation and an increase in donor side limitation [48,49]. This suggests that FLV activity is instrumental to keep P700 in its oxidized form especially upon a sudden increase in light intensity .
A functional analysis carried on in the moss P. patens gave direct evidence of FLV contribution to photosynthetic electron transport activity  showing that flv KO mutants show a large reduction (∼60%) of their electron transport activity immediately after an increase in illumination . This amount of transported electrons is consistent with the estimations made in Synechocystis where, at light onset under inorganic carbon deprivation, 50–60% of the electrons are attributable to FLV1/FLV3 for O2 photoreduction . FLV activity is likely responsible for most O2 uptake that has previously been attributed to the Mehler reaction in the green alga Chlamydomonas  and in Gymnosperms . Such a massive contribution to electron transport activity, however, is limited to the first few seconds after the light is switched on (≈10 s in P. patens) and after that FLV contribution to electron transport strongly decreased and flv KO are indistinguishable from WT lines. In general, in both prokaryotes [27,44] and eukaryotes [43,53] intense and transient FLV activity is particularly relevant upon sudden increases in light intensity. FLV depletion causes a temporary over-reduction in the PQ pool and a delay in non-photochemical quenching activation attributable to a slower generation of ΔpH [42,43,53]. In addition to its electron transport activity, FLV thus appears to play a role in sustaining proton translocation and ATP production in the first few seconds after an increase in illumination [42,43,53]. In C. reinhardtii, FLV-mediated electron ﬂow to O2 contributes to supply extra-ATP also under steady-state illumination .
PSI is particularly stable and able to withstand strong illumination . In flv mutants of cyanobacteria, algae and mosses, after a sudden increase in illumination, there is a high acceptor side limitation that results in poor P700 oxidation [42,43,53]. In this case, PSI is over-reduced, electrons accumulate at the donor side of PSI where FeS clusters are damaged . FLVs contribute thus to protect PSI from over-reduction and from photodamage and consequently they safeguard photosynthetic activity and plant growth. [27,42,43,53,55–57].
There are other situations in which FLV activity was shown to be physiologically relevant. As an example, underwater submersion of thalli of the liverwort M. polymorpha decreases CO2 diffusion and limits Rubisco activity. FLVs by consuming the reducing power of the chloroplast stroma to convert O2 into water releases the PSI acceptor site constraints . Another example is found in heterocyst-forming filamentous cyanobacteria like Anabaena, where FLV-dependent O2 reduction appears to be crucial for growth and for photosynthetic activity during nitrogen fixation . Another example of the physiological relevance of FLV activity was reported in C. reinhardtii. Sulfur deprivation inhibits photosynthesis and drives to anaerobic photo-fermentation with H2 production by [FeFe]-hydrogenase enzymes that accept electrons from PSI . In the first phase of sulfur deprivation, FLV accumulation and activity increases, when photosynthesis is still active and anaerobiosis is not established yet, hydrogenase is not active. This observation suggests that hydrogenase and FLVs have synergic activity in lowering reducing pressure downstream PSI and to produce ATP without generating NADPH at the onset of anaerobiosis  (Figure 3).
Schematic representation of flavodiiron protein activity, biochemical interaction with other pathways and physiological function.
The examples reported above show that FLVs can represent a strong electron sink in specific growing conditions where there is the risk of an acceptor side limitation in PSI. All cases described above have in common the characteristic that FLV activity is not constitutive but specifically modulated. One particularly clear example of this regulation is given by observing the first seconds after a change in illumination intensity in P. patens when FLVs reach their maximal electron transport capacity in ∼1 s and thereafter they are quickly inactivated after 10 s when LEF activity increases . This is a clear example of how FLVs are very effectively regulated and in a few seconds their activity quickly goes from being responsible of the largest fraction of electron transport to being barely detectable. Considering the very fast kinetics involved, FLV regulation should rely on post-translational controls like conformational changes altering protein activity or affinity for the substrates. In this context it is interesting to observe that FLV activity was found relevant when CBB cycle activity is limiting compared with light-driven electron transport (e.g. fluctuating light, submersion). This observation can be explained considering that CBB is the major pathway consuming ATP and NADPH, a decrease in its activity easily drives to an over-reduction in the PSI acceptors with a consequent strong risk of electron transport chain over-reduction (see Figure 3). One interesting hypothesis is thus that FLV activity, alike enzymes of the CBB cycle, could be sensitive to the redox state of the stroma. With increasing accumulation of reducing power in the stroma, FLV could be inhibited while CBB enzymes are activated and vice versa in oxidizing conditions. FLV sequences present several conserved cysteine residues that might be potentially involved in redox regulation [58,61]. Responding to the same signal would be particularly fitting since FLVs compete with CBB for electrons and its activity is only needed when the latter is inhibited.
Interaction between FLVs and other alternative electron transport pathways and their role in proton motif force regulation
As discussed above, FLVs are important in alleviating PSI acceptor side limitation, a role in which CEF pathways are also involved [62–64]. Both CEF and FLVs also sustain pmf across thylakoid membranes. Other mechanisms contributing to NADPH consumption (e.g. Mehler, photorespiration, rerouting to mitochondria, nitrate reduction) are also influencing PSI redox state, but several evidences suggest a strong functional overlap between CEF and PCEF. As an example, flv mutants of P. patens showed an enhanced CEF  and C. reinhardtii pgrl1 knock-out strains hyperaccumulate FLV proteins . In vivo functional analysis of C. reinhardtii also demonstrated that PSI is more limited at the acceptor side at onset of high light in both pgr5 and flvb mutants than in WT . PGR5/PGRL1 and FLVs have partially similar role in sustaining the growth and photoprotection during fluctuating light regimes [42,64,66]. A furthermore, even more striking, example is that the introduction of P. patens FLVs in A. thaliana restores rescue the photosensitivity of the pgr5 mutant [57,67].
The functional overlap between PGR5/PGRL1 and FLV was further studied in P. patens by the generation of a double mutant for PGRL1 and FLVA genes . In this moss FLV- and PGRL1-dependent electron transport pathways have a prominent role under fluctuating light and excess light conditions, respectively. When both proteins are simultaneously depleted in flva pgrl1 double KO plants, however, there is a further growth defect compared with the single flva KO even under fluctuating light, suggesting that PGRL1 partially compensate for the absence of FLV. In the other way around, under continuous high light, FLVs are expected to be inactive but the loss of FLVA in pgrl1 KO background leads to a further 30% growth reduction if compared with single pgrl1 KO plants . These results confirm that FLV and PGRL1 have synergic role in modulating PSI acceptor side limitation in P. patens and their activity can partially compensate the absence of the other while in case of combined depletion, phenotypes are amplified. Consistent with a partially overlapping biological role, the loss of FLVs in Angiosperms was suggested to be at least partially compensated by an increased CEF efficiency [55,56,64], although the modulation of all other mechanisms contributing to consumption of PSI electrons (e.g. Mehler, photorespiration, rerouting to mitochondria, nitrate reduction) may also be influential.
Structural bases for FLVs function and regulation
FLV activity and regulation described above clearly depends on the protein structural organization, and its analysis can thus be highly informative. In the case of FLVs there is no complete structure available but recently a partial structure of FLV1 lacking the C-term NAD(P)H:flavin oxidoreductase-like domain (Flv1-ΔFlR) from cyanobacterium Synechocystis sp. PCC6803 was described, providing new information on FLVs .
In the absence of direct structural information, structure from cyanobacterial protein was exploited for an in silico homology modeling to obtain a three-dimensional model of eukaryotic FLVA and FLVB enzymes from P. patens (Supplementary Table S1). Other available structures from FDPs were also used as templates, in particular, Anabaena FLV3 metallo-β-lactamase-like fragment (All0177; PDB 4FEK) and Anabaena FLV1 flavodoxin-like domain (All3895; PDB 3FNI). Further core structures of FDPs belonging to class A and B contributed to the model construction: rubredoxin oxygen:oxidoreductase (roo) from anaerobe Desulfovibrio gigas (1E5D), a homolog from Giardia intestinalis (2Q9U) or a nitric oxide reductase from Moorella thermoacetica (1YCH) and others (Supplementary Table S1). As expected, FLVA and FLVB sequences showed higher identities toward Synechocystis FLV1-ΔFlR (51 and 34%, respectively), and this occurs despite FLVA is characterized by a peculiar long linker between the two highly conserved core subdomains (314–363).
The third domain, specific to FLVs (residues 560-end of FLVA and 484-end of FLVB) has the lowest sequence identity with other proteins with known structure. Selected templates showed a satisfying coverage but never exceeded 25% sequence identity, the best being a putative oxidoreductase (sma0793) from Sinorhizobium meliloti 1021 (PDB:3rh7), NADH: FMN oxidoreductase from Paracoccus denitrificans (PDB: 4xhy) and a Streptomyces globisporus c-1027 NADH:FAD oxidoreductase sgce6 (PDB:4hx6). They all belong to NADH:flavin oxidoreductase subfamily and are often part of multi-component or multi-domain enzymes, as it is the case for FLV proteins. The modeled structures of C-term domain in FLVA and FLVB are presented in Supplementary Figure S1 but considering the low sequence identities (Supplementary Table S1), they are expected to have low accuracy. The reciprocal orientation of this domain with respect to the rest of FLVA or FLVB proteins is particularly uncertain.
Models for the other domains of the protein are instead more reliable thanks to the high homology with available structures. It should be considered, however, that so far FLVs have been isolated and structurally characterized only in the apo form, since neither metal ions nor flavin cofactors have been detected in the crystallized proteins. Iron binding sites are however identifiable since crystallized enzymes trapped organic anions such as citrate or phosphate in the metallo-beta-lactamase-like domain, where the metal is expected to be bound . Modeled FLVA and FLVB show a divergent nature of the putative diiron-binding site, similarly to what was described for FLV1 and FLV3. Iron binding cavities have comparable size and topological features except for the cap on top of the diiron center: here FLVB present a canonical two stranded β-sheet analogous to class-A/B FDP members such as E. coli or Giardia Intestinalis FDPs while in FLVA the top of the cavity is closed by a 310-helix followed by a turn. Even more relevant is that FLVA diiron-binding site of P. patens is heavily altered in aminoacidic composition, with a prevalence of basic residues (His170-x-Ser172-x-Lys174Arg175-x70-Arg246-x18-Lys265-x109-His375; PpFlvA; Figure 4A) while FLVB has a classical Asp/Glu/His rich pattern (His167-x-Glu169-x-Asp171His172-x61-His234-x18-Asp253-x56-His310; PpFlvB; Figure 4B) typical of Iron binding domains of other FDPs.
Structural model of FLVA/B binding site.
Based on its conserved sequence and structural features, FLVB is confidently expected to bind iron and be able to catalyze O2 reduction similarly to other FDPs (Figure 4B). On the other hand, the binding site of FLVA is heavily altered and metal loading, if any, would demand totally different coordination features (Figure 4A). The latter hypothesis cannot be excluded, since basic and/or acidic side chains can play the role of ligands in metal centers coordination, as reported for other enzymes , but such an altered binding could also have possible consequences on enzymatic activity.
According to our FMN docking trials, the flavodoxin-like subdomain in both FLVA and FLVB can in principle accommodate flavins upon minimal conformational adjustments. Flavodoxins and flavodoxin-like enzymes are characterized by several conserved residues involved in the network of interactions that stabilize FMN . Modeled P. patens FLVA and FLVB structures present both the hydrophilic residues necessary for orthophosphate recruitment through a network of hydrogen bonds (FLVA Ser327, Ser331, Ser411), as well as aromatic residues (FLVA Trp414, Tyr 329, Phe 412) in the loops devoted to isoalloxazine moiety engagement. In particular, FLVA exposes a tryptophan in a motif of the type F412GW414S that could strengthen the FMN-binding through aromatic stacking analogously to G. intestinalis FDP and others .
While the structures of β-lactamase- and flavodoxin-like domains show a highly conserved structure in all the models independently from the template, an intriguing divergence has been observed in their reciprocal orientation. Indeed, FDPs members fold in an open and extended architecture, with both iron and FMN-binding sites pointing toward the solvent (Figure 5A). In these proteins, FMN-binding site, located in the second subdomain of the FDP core, is thus too far (more than 30 Å apart) from diiron center to allow a direct electron exchange between the two moieties within the same monomer. Indeed, in class-A and -B FDPs the protein can function by forming a head-to-tail dimeric assembly, that has been observed in multiple FDPs structures. In this oligomeric organization FMN-binding domain of one monomer places the flavin in front of the diiron center of a second monomer. This forms an extended catalytic cavity comprising both the diiron and FMN-binding sites where the two cofactors are at a close distance of roughly 6 Å, allowing for electron transport. This configuration is also suggested by structural models of FLVB and is consistent with the experimental indications that FLVA and FLVB are active as hetero-oligomers. Given the different amino acid sequences of FLVA and B, the two diiron-FMN merged catalytic sites will likely not be identical, opening the way to distinct roles for the two independent reaction centers defined in the dimeric assembly (Figure 5A). One of them is conserved and strongly superimposable to classical FDP active sites (Figures 4B and 5A), while the second presents several differences in the diiron coordination center (Figure 4A) that could result in a different tuning of the two active sites in terms of redox-potential.
Putative oligomerization model for FLVA and FLVB of Physcomitrella patens.
It is surprising to observe that class-C Synechocystis FLV1-ΔFlR show instead a different orientation of the domains and thus FLVA structural model based on this template assembles in a more closed and compact manner, as a result of the remarkable shift and rotation of the two subdomains (Figure 5B). Based on final Q-mean values obtained for the calculated models, both open (Figure 5A) and close (Figure 5B) architectures are plausible for FLVs. The ‘closed’ conformation, however, strongly alters the possible hetero-oligomeric association for FLVA and FLVB proteins. Indeed, in the ‘closed’ conformation the dimeric head-to-tail association cannot be established and the two redox centers, FMN and diiron loaded, could not define reciprocal intimate and extended interactions anymore, impairing their catalytic activity because of larger distances between reactive cofactors (Figure 5).
Given the structural data that suggest two possible conformations, it is inspiring to suggest that FLVA/B can assume different conformation that would correspond to different active/inactive states. These models thus suggest a plausible hypothesis on how FLV activity can be fast regulated by switching between two conformations in response to an external signal such as redox state. According to this hypothesis in one conformation FLV would be ‘open’ and capable of catalyzing the O2 reduction while int the ‘closed’ conformation, the protein would be inactive.
alternative oxidase activity
- CBB cycle
cyclic electron flow
- Cyt b6f
linear electron flow
chloroplast NADH dehydrogenase-like complex
proton gradient regulator 5
proton motive force
A.A. and T.M. conceived the project. L.C. performed protein modeling. All authors contributed equally to this manuscript.
A.A. acknowledges the support from Università degli Studi di Padova, Dipartimento di Biologia (BIRD173749/17).
The Authors declare that there are no competing interests associated with the manuscript.