The roles of PDH complex and PFOR were studied in Synechocystis under different growth conditions. PDH could not be deleted from the genome indicating that this enzyme complex is essential, whereas pfor was knocked out in a previous study (Gutekunst et al., 2014). In line with this, we found that all fully sequenced diazotrophic and nondiazotrophic cyanobacteria with photosystem II (PSII contain genes coding for a PDH complex and that 56% of these cyanobacteria possess a PFOR as well). If we subtract from this group all diazotrophic cyanobacteria that contain a nitrogenase and might therefore utilize PFOR in the process of nitrogen fixation, 130 nondiazotrophic cyanobacteria remain. Within the group of nondiazotrophic cyanobacteria 42% possess a PFOR in addition to the PDH complex (Figure 1—figure supplement 1). This clearly shows that the property of holding both a PDH complex and a PFOR in cyanobacteria that live predominantly under oxic conditions is truly widespread. The analysis furthermore confirms our observation, that the PDH complex is preferred over the utilization of PFOR in cyanobacteria.We unexpectedly found that the Synechocystis Δpfor deletion mutant was impaired in its photomixotrophic growth under oxic conditions in continuous light. Growth impairment was typically visible starting around days 3–6 of the growth experiment (Figure 1A and 3A). In addition, maintenance in the stationary growth phase was affected in Δpfor. Under photoautotrophic conditions Δpfor grew similar to the WT (Figure 1A). The oxygen concentration in the photomixotrophic cultures was close to saturation around 250 µMol O₂ throughout the growth experiment (Figure 1—figure supplement 2). Studies on the transcription of pfor and the alpha subunit of the PDH (E1) of pdhA revealed that both genes are transcribed under photomixo- and photoautotrophic conditions (Figure 1—figure supplement 3). These observations raised two questions: Why is the PDH complex, which catalyzes the same reaction as PFOR, not able to compensate for the loss of PFOR? And how can PFOR, which is assumed to be oxygen sensitive, be of physiological relevance in the presence of oxygen?

Figure 1 with 4 supplements see all

Download asset Open asset

Figure 1—source data 1

Raw data of growth and NADH/NAD⁺ ratio of wild type (WT) and Δpfor under photoautotrophic and photomixotrophic conditions.

https://cdn.elifesciences.org/articles/71339/elife-71339-fig1-data1-v2.xlsx

Download elife-71339-fig1-data1-v2.xlsx

The most obvious assumption is that the PDH complex might get inactivated under photomixotrophic conditions. As the PDH complex gets inactivated at high NADH/NAD⁺ ratios in prokaryotes and eukaryotes (Kim et al., 2008; Sun et al., 2012; Kolobova et al., 2001), we wondered if NADH/NAD⁺ ratios might be increased under photomixotrophic conditions. Corresponding measurements confirmed this assumption. Whereas NADH/NAD⁺ ratios were stable under photoautotrophic conditions in WT and Δpfor they raised three to fourfold in the first 5 days of photomixotrophic growth, exactly in that period in which the growth impairment of Δpfor in the presence of glucose was most apparent (Figure 1B).

In addition, in vivo NAD(P)H fluorescence measurements and estimates for the reduction level of ferredoxin using a Dual-KLAS/NIR were performed, which show that in addition to the NADH/NAD⁺ ratio, also the NAD(P)H and ferredoxin pools are more reduced under photomixotrophic conditions in comparison to photoautotrophic conditions (Figure 1—figure supplement 4).

For prokaryotes it was shown that the PDH complex is inhibited by a distinct mechanism directly by NADH which binds to the dihydrolipoyl dehydrogenase (E3) subunit of the PDH complex (Kim et al., 2008; Sun et al., 2012). Therefore, the recombinant dihydrolipoyl dehydrogenase of Synechocystis (SynLPD) was tested in an in vitro assay with different NADH concentrations. The enzyme indeed loses activity at higher NADH/NAD⁺ ratios, whereas NADPH has no effect (Figure 2A). The SynLPD activity was completely inhibited by NADH with an estimated K_i of 38.3 µM (Figure 2A). Hence, the enzyme activity dropped to approximately 50% at a NADH/NAD⁺ ratio of 0.1 (e.g., at 0.2 mM NADH in the presence of 2 mM NAD⁺). Please note, that much higher NADH/NAD⁺ ratios (>0.4) were measured in photomixotrophic cells of Synechocystis (see Figure 1B). This points to an efficient inhibition of PDH activity via the highly decreased function of the E3 subunit (SynLPD). NADH/NAD⁺ ratios above 0.1 could not be tested in the enzyme assays due to the high background absorption of the added NADH, which prevented SynLPD activity detections.

Figure 2 with 3 supplements see all

Download asset Open asset

Figure 2—source data 1

Raw data of enzymatic in vitro test with the dihydrolipoyl dehydrogenase (E3) subunit (SynLPD) of the pyruvate dehydrogenase (PDH) complex and pyruvate:ferredoxin oxidoreductase (PFOR).

https://cdn.elifesciences.org/articles/71339/elife-71339-fig2-data1-v2.xlsx

Download elife-71339-fig2-data1-v2.xlsx

Taken together these measurements convincingly show that the PDH complex is most likely inhibited under photomixotrophic conditions at high NADH/NAD⁺ ratios, which provides evidence that pyruvate oxidation must be performed instead via PFOR and gives an explanation for the importance of PFOR under these conditions.

As the cyanobacterial PFOR is regarded as an oxygen-sensitive enzyme that exclusively supports fermentation under anaerobic conditions, we overexpressed the enzyme and purified it in the presence of oxygen in order to check for its stability under aerobic conditions (Figure 2—figure supplement 1, Figure 2—figure supplement 2, Figure 2—figure supplement 3). Enzymatic tests revealed that PFOR from Synechocystis was indeed stable under aerobic conditions in vitro, which means that the enzyme was not degraded and kept its activity but required anoxic conditions for the decarboxylation of pyruvate (Figure 2B) as reported for the oxygen stable PFORs of D. africanus and S. acidocaldarius (Witt et al., 2019; Pieulle et al., 1995).

In contrast to the PDH complex, PFOR transfers electrons from pyruvate to oxidized ferredoxin. In order to investigate if any of the low abundant ferredoxins (Fx) might be of importance for photomixotrophic growth, respective deletion mutants were generated (Supplementary file 1a, b, Figure 3—figure supplement 1) and tested for their ability to grow under photoautotrophic and photomixotrophic conditions. To this end, fx3 (slr1828), fx4 (slr0150), fx6 (ssl2559), fx7 (sll0662), and fx9 (slr2059) could be completely deleted from the genome, whereas the deleted alleles of fx2 (sll1382) and fx5 (slr0148) failed to segregate, keeping some wild-type copies. Furthermore, we did not succeed to delete fx8 (ssr3184). Flavodoxin (isiB; sll0284), which replaces ferredoxins functionally under Fe limitation was deleted as well. In addition, the double mutants Δfx7Δfx9 and Δfx9ΔisiB as well as the triple mutant Δfx7Δfx8Δfx9 were generated. Photoautotrophic growth of all these ferredoxin deletion mutants was similar to the WT (Figure 3—figure supplement 2). However, under photomixotrophic conditions deletion of either fx3, fx9, or flavodoxin (isiB) resulted in a growth behavior that was similar to Δpfor (Figure 3A).

Figure 3 with 2 supplements see all

Download asset Open asset

Figure 3—source data 1

Raw data from growth of wild type (WT), Δpfor, Δfx, ΔisiB, Δf-gogat, and Δn-gogat and electron transport with DCMU at photosystem I (PSI) in the absence and presence of glucose in the WT, ΔndhD1ΔndhD2, Δhk, and ΔglgP1ΔglgP2.

https://cdn.elifesciences.org/articles/71339/elife-71339-fig3-data1-v2.xlsx

Download elife-71339-fig3-data1-v2.xlsx

These results indicate that there might be a general shift to utilize the ferredoxin pool as soon as the NADH/NAD⁺ pool is over-reduced. Beside the PFOR/PDH complex couple, GOGAT (glutamine oxoglutarate aminotransferase) as well is present in form of two isoenzymes in Synechocystis that either utilizes reduced ferredoxin (F-GOGAT; sll1499) or NADH (N-GOGAT; sll1502). In line with our assumption that ferredoxin utilization is preferred in over-reduced cells after glucose addition, we hypothesized that F-GOGAT might be required for optimal photomixotrophic growth. Respective deletion mutants were generated (Supplementary file 1a, b, Figure 3—figure supplement 1) and revealed that neither Δf-gogat nor Δn-gogat were impaired in their growth under photoautotrophic conditions, whereas Δf-gogat displayed a strong growth impairment under photomixotrophic conditions in contrast to Δn-gogat and the WT (Figure 3B). These data indicate that cells indeed rely on a general switch from utilizing NAD(H) to utilizing ferredoxins for optimal photomixotrophic growth. It was recently shown that photosynthetic complex I (NDH1) exclusively accepts electrons from reduced ferredoxin instead of NAD(P)H (Schuller et al., 2019). Under photomixotrophic conditions photosynthesis operates in parallel to carbon oxidation. In addition to water oxidation at PSII, electrons from glucose oxidation can as well enter the respiratory/photosynthetic electron transport chain and eventually arrive at PSI. Photosynthesis based on PSI thus uses electrons from glucose oxidation that enter the respiratory/photosynthetic electron transport chain and are excited at PSI.

Three entry points exist that can feed electrons from glucose oxidation into the plastoquinone (PQ) pool in the thylakoid membrane: the succinate dehydrogenase, which accepts electrons from the conversion of succinate to fumarate; NDH-2, which accepts electrons from NADH and photosynthetic complex I (NDH-1), which accepts electrons from reduced ferredoxin (see Figure 4B). Based on the observed shift from utilizing ferredoxin instead of NAD(P)H, we thus wondered if photosynthetic complex I (NDH-1) might be required for photosynthesis (involving only PSI) under photomixotrophic conditions as an entry point for electrons coming from glucose oxidation. Cells were incubated with DCMU that blocks the electron transfer from PSII to the PQ pool. Thereby, electron transfer from glycogen or glucose oxidation to PSI could be measured based on a recently developed protocol (Theune et al., 2021). According to this protocol electrons were counted that flow through PSI via DIRK_PSI measurements by the KLAS/NIR instrument. The electron transport at PSI was then measured in the absence and in the presence of glucose. In addition to the WT, several mutants were analyzed with deletions in entry points as well as glucose metabolizing enzymes. The mutant with a deleted photosynthetic complex I (ΔndhD1ΔndhD2) should no longer be able to feed electrons from reduced ferredoxin into the respiratory/photosynthetic electron transport chain, while the hexokinase mutant (Δhk) should no longer be able to metabolize external glucose. The glycogen phosphorylase mutant (ΔglgP1ΔglgP2) is unable to break down its internal glycogen reservoir (Theune et al., 2021; Doello et al., 2018; Makowka et al., 2020). As expected and in parts shown recently (Theune et al., 2021), addition of glucose resulted in higher donations of electrons to PSI in the WT and ΔglgP1ΔglgP2, whereas neither ΔndhD1ΔndhD2 nor Δhk were able to provide electrons from glucose oxidation to PSI (Figure 3C). Photosynthesis using glucose oxidation and PSI thus relies on the ferredoxin-dependent photosynthetic complex I. In line with this, it was shown earlier that ΔndhD1ΔndhD2 is not able to grow in the presence of glucose and DCMU under photoheterotrophic conditions (Ohkawa et al., 2000).

Figure 4

Download asset Open asset

Under photomixotrophic conditions, photosynthesis and glucose oxidation operate in parallel. The cells are thus flooded with electrons from water oxidation at PSII and electrons from glucose oxidation (Figure 4). This causes highly reducing conditions in the cells as visible in reduced NAD(P)H and ferredoxin pools (Figure 1B and Figure 1—figure supplement 4). Our data indicate that the PDH complex gets inhibited at high NADH levels under these conditions and is subsequently most likely functionally replaced by PFOR (Figures 1 and 2). Furthermore, the cells seem to rely on a general shift from utilizing NAD(H)- to ferredoxin-dependent enzymes under these conditions. In line with this, low abundant ferredoxins, whose functions are still only partly understood in detail, and ferredoxin-dependent F-GOGAT are required for optimal photomixotrophic performance (Figure 3). Photosynthetic complex I (NDHI) which accepts electrons from reduced ferredoxin (Schuller et al., 2019), is furthermore required to feed electrons from glucose oxidation into the photosynthetic electron chain and to thereby enhance electron flow at PSI (Figures 3 and 4).

PFORs are with a few reported exceptions highly oxygen-sensitive enzymes that work under strictly anaerobic conditions (Pieulle et al., 1995; Vita et al., 2008; Kerscher and Oesterhelt, 1981). We found that PFOR of Synechocystis is stable in the presence of oxygen, however, in vitro we could only measure the decarboxylation of pyruvate in the absence of oxygen (Figure 2). However, our data strongly indicate that this enzyme is active in vivo under aerobic and highly reducing conditions.

Similar results were recently reported for E. coli. E. coli possesses three enzyme systems to convert pyruvate to acetyl CoA: the PDH complex, PFOR, and pyruvate formate-lyase (PFL) (Blaschkowski et al., 2005). The PDH complex and PFL are the principle enzyme systems to convert pyruvate to acetylCoA in E. coli, whereas PFOR is expressed at very low levels (Blaschkowski et al., 2005). Transcription of PFOR was shown to be enhanced under oxidative stress (Nakayama et al., 2013). E. coli decarboxylates pyruvate via the PDH complex in the presence of oxygen. Under anaerobic conditions NADH levels rise and inhibit the PDH complex. PFL gets activated and the cells switch to fermentation. PFL activation requires reduced flavodoxin which is provided by PFOR (Blaschkowski et al., 2005). The regulation at the pyruvate node in E. coli is thus mainly regulated via the availability of oxygen and its concomitant requirement for redox control and ATP (Wang et al., 2010). By downregulation of glucose-6P dehydrogenase (ZWF), less NADPH was produced in E. coli, which activated the expression of PFOR and ferredoxin reductase (FPR) (Li et al., 2021). PFOR and FPR were shown to contribute to stationary-phase metabolism in aerobic cultures in this mutant probably by converting reduced ferredoxin to NADPH (Li et al., 2021). PFOR is thus obviously involved in redox control in these mutants in the presence of oxygen. This finding was highly unexpected, as PFOR activity in crude extracts from aerobically grown E. coli cells is only detectable in the absence of oxygen in vitro (Nakayama et al., 2013). There are several reports in prokaryotes and eukaryotes on the expression of enzymes under oxic conditions that are assigned to anaerobic metabolism (Gould et al., 2019; Schmitz et al., 2001; Gutekunst et al., 2005). One example is the production of hydrogen by the oxygen-sensitive FeFe-hydrogenase in air-grown Chlamydomonas reinhardtii algae in a fully aerobic environment, which is enabled by microoxic niches within the thylakoid stroma (Liran et al., 2016). Another example is the constitutive expression of PFOR and the oxygen-sensitive NiFe-hydrogenase under oxic conditions in cyanobacteria. By itself, the widespread presence of these enzymes in organisms that either live predominantly aerobically as for example cyanobacteria or are even obligate aerobes as for example S. acidocaldarius, which possesses a PFOR, indicates a misconception and lack of understanding. The PFOR of S. acidocaldarius could be isolated as stable enzyme in the presence of O₂, however, enzyme activity measurements required the consumption of oxygen in vitro (Witt et al., 2019). Does this mean, that anaerobic microniches are required within this obligate aerobe to activate an enzyme of its central carbon metabolism? It might alternatively be that living cells have the ability to maintain reducing conditions in the presence of oxygen by yet unknown mechanisms that for example consume oxygen, which is a challenge in enzymatic in vitro assays. Conclusions on in vivo enzyme activities based on in vitro experiments therefore should be made with caution. Even though we could measure decarboxylation of pyruvate via PFOR only in the absence of oxygen in vitro, our data strongly indicate that this enzyme is active in vivo under aerobic and highly reducing conditions. We assume that either anaerobic microniches or alternatively mechanisms within the cell that are not understood yet, keep the enzyme active in an aerobic environment.

Low abundant ferredoxins are required for optimal photomixotrophic growth (Figure 3), which is surprising when looking at glycolytic routes for glucose oxidation. Glucose is alternatively oxidized via different glycolytic routes in Synechocystis (Figure 4A). Flux analyses have shown that glycolytic intermediates enter the CBB cycle, eventually reach lower glycolysis, and finally provide pyruvate (Nakajima et al., 2014). Depending on the precise route taken, glucose oxidation yields distinct forms of reducing equivalents (Makowka et al., 2020). Three enzymes are involved in oxidation steps: Glc6P dehydrogenase (Zwf) and 6 PG dehydrogenase (Gnd) yield NADPH, whereas GAP dehydrogenase (GAPDH) yields NADH. NAD(P)H is furthermore provided downstream in the TCA cycle. PFOR is thus the only known direct source for reduced ferredoxin in glucose oxidation beside PSI (Figure 4). The wide network of low abundant ferredoxins in Synechocystis and the importance of these ferredoxins under photomixotrophic conditions on the one hand and the low number of known enzymes that directly reduce ferredoxins on the other hand unveils that our conception is not yet inherently consistent. An additional potential source of reduced ferredoxin could be the transfer of electrons from NAD(P)H. The transhydrogenase (PntAB), which is located in the thylakoid membrane utilizes proton translocation to transfer electrons from NADH to NADP⁺ (Kämäräinen et al., 2017). Electrons from NADPH could be further transferred to ferredoxin via ferredoxin-NADPH-oxidoreductase (FNR). Another potential turntable for the exchange of electrons is the diaphorase part of the NiFe-hydrogenase in Synechocystis, which was recently shown to shuttle electrons between NAD(P)H, flavodoxin and several ferredoxins in vitro (Artz et al., 2020).

In order to get a complete picture of the ferredoxin network and potential interaction partners, it would be essential to know the redox potentials of all low abundant ferredoxins in Synechocystis. Currently, they have been determined for Fx1 (−412 mV), Fx2 (−243 mV), and Fx4 (−440 mV), whereas the value for Fx4 is based on measurements of a homologue in Thermosynechococcus elongatus (Bottin and Lagoutte, 1992; Schorsch et al., 2018; Motomura et al., 2019). Fx1–Fx6 in Synechocystis possess 2Fe2S clusters for which redox potentials between −240 and −440 mV are typical (Liu et al., 2014). For 3Fe4S clusters as present in Fx8 (containing one 3Fe4S and one 4Fe4S cluster) redox potentials between −120 and −430 mV were determined and for 4Fe4S clusters as present in Fx7 (4FeFS) and Fx9 (containing two 4Fe4S clusters) redox potentials between −300 and −680 mV were found (Liu et al., 2014). Our data show that Fx9 is of importance under photomixotrophic conditions (Figure 3A). The redox potential of Fx9 in Synechocystis has been assumed to be around −420 mV based on its interaction partners (Cassier-Chauvat and Chauvat, 2014). However, this value requires experimental validation. Without yet knowing the exact values for all ferredoxins in Synechocystis, it is obvious that they span a wide range of redox potentials. Our data indicate that cells perform a general shift from utilizing NAD(H)- to ferredoxin-dependent enzymes under highly reducing photomixotrophic conditions.

The following lines include theoretical reflections based on this observation. However, as these conclusions are not entirely supported by the data, they should be regarded as hypothetical and are meant as thought-provoking impulses.

The replacement of FeS enzymes and ferredoxins by FeS-free alternatives and NADPH in the course of evolution is in general discussed with regard to the oxygen sensitivity of FeS clusters in connection with the shift from anoxic to oxic conditions on Earth (Imlay, 2006; Gould et al., 2019). Oxygen is without any doubt one important factor. However, the shift from anoxic to oxic conditions went along with a shift from reducing to more oxidizing conditions. This shift was among others achieved by the escape of hydrogen into space, which irreversibly withdrew electrons from Earth (Catling et al., 2001). The withdrawal of electrons and the establishment of oxidizing conditions might have been an additional important factor (independent of oxygen and the oxygen sensitivity of FeS clusters) that triggered these evolutionary changes by enabling reactions with higher driving forces. The idea is thus that PFOR and ferredoxins might have been replaced by the PDH complex and NADH due to their potential to release larger amounts of Gibbs-free energy (ΔG < 0). When competing with other organisms for resources an accelerated metabolism can be highly beneficial.

The decision to either utilize the PDH complex or alternatively PFOR and along this line, the replacement of PFOR by the PDH complex in the course of evolution might have been determined by the prioritization for high chemical driving forces.

On that note, we were unable to delete the PDH complex in Synechocystis, which points to its essential role. PFOR is in contrast dispensable under photoautotrophic conditions and cells obviously prefer to decarboxylate pyruvate via the PDH complex under these conditions. By transferring electrons to NAD⁺ instead of ferredoxin less Gibbs-free energy is stored. However, this comes along with a higher driving force that is visible when regarding the reaction Gibbs energies of Δ_rG′^m −39 kJ/mol for the reaction catalyzed by the PDH complex versus Δ_rG′^m −23 kJ/mol for the reaction catalyzed by PFOR (Figure 4C; Noor et al., 2013).

The idea is thus that the NADH/NAD⁺ pool gets reduced first prioritizing high driving forces. However, as the redox potential of the NADH/NAD⁺ pool turns slowly more negative, it might reach levels that are characteristic for ferredoxin couples. This might provoke a metabolic shift to transfer electrons to oxidized ferredoxin instead of NAD⁺ which should come along with lower metabolic rates (Figure 4C). This idea fits well with the observation, that PFOR and low abundant ferredoxins gain importance in the stationary growth phase (Figures 1A and 3A), which is characterized by a slowing down of metabolic reactions. The shift can be regulated on several levels. Among others, as shown in this study, high NADH/NAD⁺ ratios can inactivate enzymes that rely on this couple and thereby support the action of isoenzymes that interact with the Fx_red/Fx_ox couple instead. In addition, a shift to more reducing conditions (Figure 1—figure supplement 4), will alter the thermodynamic driving force of many redox reactions, and may in itself necessitate a shift in pathways. In addition, electron turntables as the transhydrogenase, FNR, and the diaphorase can support this shift (Artz et al., 2020; Kämäräinen et al., 2017).

By shifting their pools of reducing equivalents, cells are thus able to finetune their metabolism. They either liberate or save Gibbs-free energy and thereby either accelerate or slow down metabolic reactions, as required.

All data generated or analysed during this study are included in the manuscript and supporting file; Source data files have been provided for Figures 1, 2, 3 and accompanying figure supplements.

1. 1. Artz JH
   2. Tokmina-Lukaszewska M
   3. Mulder DW
   4. Lubner CE
   5. Gutekunst K
   6. Appel J
   7. Bothner B
   8. Boehm M
   9. King PW

   (2020) The structure and reactivity of the HoxEFU complex from the cyanobacterium Synechocystis sp. PCC 6803

   The Journal of Biological Chemistry 295:9445–9454.

   https://doi.org/10.1074/jbc.RA120.013136

   • PubMed
   • Google Scholar
2. 1. Blaschkowski HP
   2. Knappe J
   3. Ludwig-festl M
   4. Neuer G

   (2005) Routes of Flavodoxin and Ferredoxin Reduction in Escherichia coli

   European Journal of Biochemistry 123:563–569.

   https://doi.org/10.1111/j.1432-1033.1982.tb06569.x

   • Google Scholar
3. 1. Bottin H
   2. Lagoutte B

   (1992) Ferredoxin and flavodoxin from the cyanobacterium Synechocystis sp PCC 6803

   Biochimica et Biophysica Acta 1101:48–56.

   https://doi.org/10.1016/0167-4838(92)90465-p

   • PubMed
   • Google Scholar
4. 1. Bradford MM

   (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding

   Analytical Biochemistry 72:248–254.

   https://doi.org/10.1006/abio.1976.9999

   • PubMed
   • Google Scholar
5. 1. Cassier-Chauvat C
   2. Chauvat F

   (2014) Function and Regulation of Ferredoxins in the Cyanobacterium, Synechocystis PCC6803: Recent Advances

   Life (Basel, Switzerland) 4:666–680.

   https://doi.org/10.3390/life4040666

   • PubMed
   • Google Scholar
6. 1. Catling DC
   2. Zahnle KJ
   3. McKay CP

   (2001) Biogenic methane, hydrogen escape, and the irreversible oxidation of early Earth

   Science (New York, N.Y.) 293:839–843.

   https://doi.org/10.1126/science.1061976

   • PubMed
   • Google Scholar
7. 1. Doello S
   2. Klotz A
   3. Makowka A
   4. Gutekunst K
   5. Forchhammer K

   (2018) A Specific Glycogen Mobilization Strategy Enables Rapid Awakening of Dormant Cyanobacteria from Chlorosis

   Plant Physiology 177:594–603.

   https://doi.org/10.1104/pp.18.00297

   • PubMed
   • Google Scholar
8. 1. Evans MC
   2. Buchanan BB
   3. Arnon DI

   (1966) A new ferredoxin-dependent carbon reduction cycle in a photosynthetic bacterium

   PNAS 55:928–934.

   https://doi.org/10.1073/pnas.55.4.928

   • PubMed
   • Google Scholar
9. 1. Fuchs G

   (2011) Alternative pathways of carbon dioxide fixation: insights into the early evolution of life?

   Annual Review of Microbiology 65:631–658.

   https://doi.org/10.1146/annurev-micro-090110-102801

   • PubMed
   • Google Scholar
10. 1. Gibson DG
    2. Young L
    3. Chuang RY
    4. Venter JC
    5. Hutchison CA
    6. Smith HO

    (2009) Enzymatic assembly of DNA molecules up to several hundred kilobases

    Nature Methods 6:343–345.

    https://doi.org/10.1038/nmeth.1318

    • PubMed
    • Google Scholar
11. 1. Gould SB
    2. Garg SG
    3. Handrich M
    4. Nelson-Sathi S
    5. Gruenheit N
    6. Tielens AGM
    7. Martin WF

    (2019) Adaptation to life on land at high O2 via transition from ferredoxin-to NADH-dependent redox balance

    Proceedings. Biological Sciences 286:20191491.

    https://doi.org/10.1098/rspb.2019.1491

    • PubMed
    • Google Scholar
12. 1. Gutekunst K
    2. Phunpruch S
    3. Schwarz C
    4. Schuchardt S
    5. Schulz-Friedrich R
    6. Appel J

    (2005) LexA regulates the bidirectional hydrogenase in the cyanobacterium Synechocystis sp. PCC 6803 as a transcription activator

    Molecular Microbiology 58:810–823.

    https://doi.org/10.1111/j.1365-2958.2005.04867.x

    • PubMed
    • Google Scholar
13. 1. Gutekunst K
    2. Chen X
    3. Schreiber K
    4. Kaspar U
    5. Makam S
    6. Appel J

    (2014) The bidirectional NiFe-hydrogenase in Synechocystis sp. PCC 6803 is reduced by flavodoxin and ferredoxin and is essential under mixotrophic, nitrate-limiting conditions

    The Journal of Biological Chemistry 289:1930–1937.

    https://doi.org/10.1074/jbc.M113.526376

    • PubMed
    • Google Scholar
14. 1. Gutekunst K

    (2018) Hypothesis on the Synchronistic Evolution of Autotrophy and Heterotrophy

    Trends in Biochemical Sciences 43:402–411.

    https://doi.org/10.1016/j.tibs.2018.03.008

    • PubMed
    • Google Scholar
15. 1. Hoffmann D
    2. Gutekunst K
    3. Klissenbauer M
    4. Schulz-Friedrich R
    5. Appel J

    (2006) Mutagenesis of hydrogenase accessory genes of Synechocystis sp. PCC 6803

    FEBS Journal 273:4516–4527.

    https://doi.org/10.1111/j.1742-4658.2006.05460.x

    • PubMed
    • Google Scholar
16. 1. Imlay JA

    (2006) Iron-sulphur clusters and the problem with oxygen

    Molecular Microbiology 59:1073–1082.

    https://doi.org/10.1111/j.1365-2958.2006.05028.x

    • PubMed
    • Google Scholar
17. Book

    1. Jagannathan B
    2. Shen G
    3. Golbeck J

    (2012) The evolution of type I reaction centers: The response to oxygenic photosynthesis

    In: Burnap R, Vermaas W, editors. Functional Genomics and Evolution of Photosynthetic Systems. Springer. pp. 285–316.

    https://doi.org/10.1007/978-94-007-1533-2_12

    • Google Scholar
18. 1. Kämäräinen J
    2. Huokko T
    3. Kreula S
    4. Jones PR
    5. Aro E-M
    6. Kallio P

    (2017) Pyridine nucleotide transhydrogenase PntAB is essential for optimal growth and photosynthetic integrity under low-light mixotrophic conditions in Synechocystis sp. PCC 6803

    The New Phytologist 214:194–204.

    https://doi.org/10.1111/nph.14353

    • PubMed
    • Google Scholar
19. 1. Kerscher L
    2. Oesterhelt D

    (1977) Ferredoxin is the coenzyme of alpha-ketoacid oxidoreductases in Halobacterium halobium

    FEBS Letters 83:197–201.

    https://doi.org/10.1016/0014-5793(77)81004-1

    • PubMed
    • Google Scholar
20. 1. Kerscher L
    2. Oesterhelt D

    (1981) Purification and properties of two 2-oxoacid:ferredoxin oxidoreductases from Halobacterium halobium

    European Journal of Biochemistry 116:587–594.

    https://doi.org/10.1111/j.1432-1033.1981.tb05376.x

    • PubMed
    • Google Scholar
21. 1. Kim Y
    2. Ingram LO
    3. Shanmugam KT

    (2008) Dihydrolipoamide dehydrogenase mutation alters the NADH sensitivity of pyruvate dehydrogenase complex of Escherichia coli K-12

    Journal of Bacteriology 190:3851–3858.

    https://doi.org/10.1128/JB.00104-08

    • PubMed
    • Google Scholar
22. 1. Kim JD
    2. Rodriguez-Granillo A
    3. Case DA
    4. Nanda V
    5. Falkowski PG

    (2012) Energetic selection of topology in ferredoxins

    PLOS Computational Biology 8:e1002463.

    https://doi.org/10.1371/journal.pcbi.1002463

    • PubMed
    • Google Scholar
23. 1. Klughammer C
    2. Schreiber U

    (2016) Deconvolution of ferredoxin, plastocyanin, and P700 transmittance changes in intact leaves with a new type of kinetic LED array spectrophotometer

    Photosynthesis Research 128:195–214.

    https://doi.org/10.1007/s11120-016-0219-0

    • PubMed
    • Google Scholar
24. 1. Kolobova E
    2. Tuganova A
    3. Boulatnikov I
    4. Popov KM

    (2001) Regulation of pyruvate dehydrogenase activity through phosphorylation at multiple sites

    The Biochemical Journal 358:69–77.

    https://doi.org/10.1042/0264-6021:3580069

    • PubMed
    • Google Scholar
25. 1. Li S
    2. Ye Z
    3. Moreb EA
    4. Hennigan JN
    5. Castellanos DB
    6. Yang T
    7. Lynch MD

    (2021) Dynamic control over feedback regulatory mechanisms improves NADPH flux and xylitol biosynthesis in engineered E. coli

    Metabolic Engineering 64:26–40.

    https://doi.org/10.1016/j.ymben.2021.01.005

    • PubMed
    • Google Scholar
26. 1. Liran O
    2. Semyatich R
    3. Milrad Y
    4. Eilenberg H
    5. Weiner I
    6. Yacoby I

    (2016) Microoxic Niches within the Thylakoid Stroma of Air-Grown Chlamydomonas reinhardtii Protect [FeFe]-Hydrogenase and Support Hydrogen Production under Fully Aerobic Environment

    Plant Physiology 172:264–271.

    https://doi.org/10.1104/pp.16.01063

    • PubMed
    • Google Scholar
27. 1. Liu J
    2. Chakraborty S
    3. Hosseinzadeh P
    4. Yu Y
    5. Tian S
    6. Petrik I
    7. Bhagi A
    8. Lu Y

    (2014) Metalloproteins containing cytochrome, iron-sulfur, or copper redox centers

    Chemical Reviews 114:4366–4469.

    https://doi.org/10.1021/cr400479b

    • PubMed
    • Google Scholar
28. 1. Makowka A
    2. Nichelmann L
    3. Schulze D
    4. Spengler K
    5. Wittmann C
    6. Forchhammer K
    7. Gutekunst K

    (2020) Glycolytic Shunts Replenish the Calvin-Benson-Bassham Cycle as Anaplerotic Reactions in Cyanobacteria

    Molecular Plant 13:471–482.

    https://doi.org/10.1016/j.molp.2020.02.002

    • PubMed
    • Google Scholar
29. 1. Mall A
    2. Sobotta J
    3. Huber C
    4. Tschirner C
    5. Kowarschik S
    6. Bačnik K
    7. Mergelsberg M
    8. Boll M
    9. Hügler M
    10. Eisenreich W
    11. Berg IA

    (2018) Reversibility of citrate synthase allows autotrophic growth of a thermophilic bacterium

    Science (New York, N.Y.) 359:563–567.

    https://doi.org/10.1126/science.aao2410

    • PubMed
    • Google Scholar
30. 1. Motomura T
    2. Zuccarello L
    3. Sétif P
    4. Boussac A
    5. Umena Y
    6. Lemaire D
    7. Tripathy JN
    8. Sugiura M
    9. Hienerwadel R
    10. Shen JR
    11. Berthomieu C

    (2019) An alternative plant-like cyanobacterial ferredoxin with unprecedented structural and functional properties

    Biochimica et Biophysica Acta. Bioenergetics 1860:148084.

    https://doi.org/10.1016/j.bbabio.2019.148084

    • PubMed
    • Google Scholar
31. 1. Müller M
    2. Mentel M
    3. van Hellemond JJ
    4. Henze K
    5. Woehle C
    6. Gould SB
    7. Yu RY
    8. van der Giezen M
    9. Tielens AGM
    10. Martin WF

    (2012) Biochemistry and evolution of anaerobic energy metabolism in eukaryotes

    Microbiology and Molecular Biology Reviews 76:444–495.

    https://doi.org/10.1128/MMBR.05024-11

    • PubMed
    • Google Scholar
32. 1. Mustila H
    2. Allahverdiyeva Y
    3. Isojärvi J
    4. Aro EM
    5. Eisenhut M

    (2014) The bacterial-type [4Fe-4S] ferredoxin 7 has a regulatory function under photooxidative stress conditions in the cyanobacterium Synechocystis sp. PCC 6803

    Biochimica et Biophysica Acta 1837:1293–1304.

    https://doi.org/10.1016/j.bbabio.2014.04.006

    • PubMed
    • Google Scholar
33. 1. Nakajima T
    2. Kajihata S
    3. Yoshikawa K
    4. Matsuda F
    5. Furusawa C
    6. Hirasawa T
    7. Shimizu H

    (2014) Integrated metabolic flux and omics analysis of Synechocystis sp. PCC 6803 under mixotrophic and photoheterotrophic conditions

    Plant & Cell Physiology 55:1605–1612.

    https://doi.org/10.1093/pcp/pcu091

    • PubMed
    • Google Scholar
34. 1. Nakayama T
    2. Yonekura S-I
    3. Yonei S
    4. Zhang-Akiyama Q-M

    (2013) Escherichia coli pyruvate:flavodoxin oxidoreductase, YdbK - regulation of expression and biological roles in protection against oxidative stress

    Genes & Genetic Systems 88:175–188.

    https://doi.org/10.1266/ggs.88.175

    • PubMed
    • Google Scholar
35. 1. Noor E
    2. Haraldsdóttir HS
    3. Milo R
    4. Fleming RMT

    (2013) Consistent estimation of Gibbs energy using component contributions

    PLOS Computational Biology 9:e1003098.

    https://doi.org/10.1371/journal.pcbi.1003098

    • PubMed
    • Google Scholar
36. 1. Ohkawa H
    2. Pakrasi HB
    3. Ogawa T

    (2000) Two types of functionally distinct NAD(P)H dehydrogenases in Synechocystis sp. strain PCC6803

    The Journal of Biological Chemistry 275:31630–31634.

    https://doi.org/10.1074/jbc.M003706200

    • PubMed
    • Google Scholar
37. 1. Pieulle L
    2. Guigliarelli B
    3. Asso M
    4. Dole F
    5. Bernadac A
    6. Hatchikian EC

    (1995) Isolation and characterization of the pyruvate-ferredoxin oxidoreductase from the sulfate-reducing bacterium Desulfovibrio africanus

    Biochimica et Biophysica Acta 1250:49–59.

    https://doi.org/10.1016/0167-4838(95)00029-t

    • PubMed
    • Google Scholar
38. 1. Pieulle L
    2. Magro V
    3. Hatchikian EC

    (1997) Isolation and analysis of the gene encoding the pyruvate-ferredoxin oxidoreductase of Desulfovibrio africanus, production of the recombinant enzyme in Escherichia coli, and effect of carboxy-terminal deletions on its stability

    Journal of Bacteriology 179:5684–5692.

    https://doi.org/10.1128/jb.179.18.5684-5692.1997

    • PubMed
    • Google Scholar
39. 1. Reinholdt O
    2. Schwab S
    3. Zhang Y
    4. Reichheld J-P
    5. Fernie AR
    6. Hagemann M
    7. Timm S

    (2019) Redox-Regulation of Photorespiration through Mitochondrial Thioredoxin o1

    Plant Physiology 181:442–457.

    https://doi.org/10.1104/pp.19.00559

    • PubMed
    • Google Scholar
40. 1. Russell MJ
    2. Martin W

    (2004) The rocky roots of the acetyl-CoA pathway

    Trends in Biochemical Sciences 29:358–363.

    https://doi.org/10.1016/j.tibs.2004.05.007

    • PubMed
    • Google Scholar
41. 1. Schmitz O
    2. Gurke J
    3. Bothe H

    (2001) Molecular evidence for the aerobic expression of nifJ, encoding pyruvate:ferredoxin oxidoreductase, in cyanobacteria

    FEMS Microbiology Letters 195:97–102.

    https://doi.org/10.1111/j.1574-6968.2001.tb10504.x

    • PubMed
    • Google Scholar
42. 1. Schorsch M
    2. Kramer M
    3. Goss T
    4. Eisenhut M
    5. Robinson N
    6. Osman D
    7. Wilde A
    8. Sadaf S
    9. Brückler H
    10. Walder L
    11. Scheibe R
    12. Hase T
    13. Hanke GT

    (2018) A unique ferredoxin acts as a player in the low-iron response of photosynthetic organisms

    PNAS 115:E12111–E12120.

    https://doi.org/10.1073/pnas.1810379115

    • PubMed
    • Google Scholar
43. 1. Schuller JM
    2. Birrell JA
    3. Tanaka H
    4. Konuma T
    5. Wulfhorst H
    6. Cox N
    7. Schuller SK
    8. Thiemann J
    9. Lubitz W
    10. Sétif P
    11. Ikegami T
    12. Engel BD
    13. Kurisu G
    14. Nowaczyk MM

    (2019) Structural adaptations of photosynthetic complex I enable ferredoxin-dependent electron transfer

    Science (New York, N.Y.) 363:257–260.

    https://doi.org/10.1126/science.aau3613

    • PubMed
    • Google Scholar
44. 1. Stanier RY
    2. Deruelles J
    3. Rippka R
    4. Herdman M
    5. Waterbury JB

    (1979) Generic Assignments, Strain Histories and Properties of Pure Cultures of Cyanobacteria

    Microbiology 111:1–61.

    https://doi.org/10.1099/00221287-111-1-1

    • Google Scholar
45. 1. Sun Z
    2. Do PM
    3. Rhee MS
    4. Govindasamy L
    5. Wang Q
    6. Ingram LO
    7. Shanmugam KT

    (2012) Amino acid substitutions at glutamate-354 in dihydrolipoamide dehydrogenase of Escherichia coli lower the sensitivity of pyruvate dehydrogenase to NADH

    Microbiology (Reading, England) 158:1350–1358.

    https://doi.org/10.1099/mic.0.055590-0

    • PubMed
    • Google Scholar
46. 1. Theune ML
    2. Hildebrandt S
    3. Steffen-Heins A
    4. Bilger W
    5. Gutekunst K
    6. Appel J

    (2021) In-vivo quantification of electron flow through photosystem I - Cyclic electron transport makes up about 35% in a cyanobacterium

    Biochimica et Biophysica Acta. Bioenergetics 1862:148353.

    https://doi.org/10.1016/j.bbabio.2020.148353

    • PubMed
    • Google Scholar
47. 1. Trautmann D
    2. Voss B
    3. Wilde A
    4. Al-Babili S
    5. Hess WR

    (2012) Microevolution in cyanobacteria: re-sequencing a motile substrain of Synechocystis sp. PCC 6803

    DNA Research 19:435–448.

    https://doi.org/10.1093/dnares/dss024

    • PubMed
    • Google Scholar
48. 1. Vita N
    2. Hatchikian EC
    3. Nouailler M
    4. Dolla A
    5. Pieulle L

    (2008) Disulfide bond-dependent mechanism of protection against oxidative stress in pyruvate-ferredoxin oxidoreductase of anaerobic Desulfovibrio bacteria

    Biochemistry 47:957–964.

    https://doi.org/10.1021/bi7014713

    • PubMed
    • Google Scholar
49. 1. Wächtershäuser G

    (1990) Evolution of the first metabolic cycles

    PNAS 87:200–204.

    https://doi.org/10.1073/pnas.87.1.200

    • PubMed
    • Google Scholar
50. 1. Wang Q
    2. Ou MS
    3. Kim Y
    4. Ingram LO
    5. Shanmugam KT

    (2010) Metabolic flux control at the pyruvate node in an anaerobic Escherichia coli strain with an active pyruvate dehydrogenase

    Applied and Environmental Microbiology 76:2107–2114.

    https://doi.org/10.1128/AEM.02545-09

    • PubMed
    • Google Scholar
51. 1. Williams JGK

    (1988) Construction of specific mutations in photosystem II photosynthetic reaction center by genetic engineering methods in Synechocystis 6803

    Methods in Enzymology 167:766–778.

    https://doi.org/10.1016/0076-6879(88)67088-1

    • Google Scholar
52. 1. Witt A
    2. Pozzi R
    3. Diesch S
    4. Hädicke O
    5. Grammel H

    (2019) New light on ancient enzymes - in vitro CO2 Fixation by Pyruvate Synthase of Desulfovibrio africanus and Sulfolobus acidocaldarius

    The FEBS Journal 286:4494–4508.

    https://doi.org/10.1111/febs.14981

    • PubMed
    • Google Scholar
53. 1. Zahnle KJ
    2. Catling DC
    3. Claire MW

    (2013) The rise of oxygen and the hydrogen hourglass

    Chemical Geology 362:26–34.

    https://doi.org/10.1016/j.chemgeo.2013.08.004

    • Google Scholar

Decision letter after peer review:

Thank you for submitting your article "Pyruvate:ferredoxin oxidoreductase and low abundant ferredoxins support aerobic photomixotrophic growth in cyanobacteria" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Gisela Storz as the Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Daniel C Ducat (Reviewer #1); Robert Burnap (Reviewer #2); Wolfgang Nitschke (Reviewer #3).

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

The work demonstrates that PFOR-mediated decarboxylation of pyruvate replaces the (more canonical) PDH-based pathway cells of the cyanobacterium Synechocystis sp. PCC6803 are shifted from photoautotrophic to photomixotrophic growth conditions. The reviewers recognized the novelty and potential importance of the work, in particular on the implications for the evolution of different enzymes and metabolic pathways under aerobic and anaerobic conditions.

1. On the other hand, all of the reviewers had similar criticisms that the text was greatly over-interpreted the results, while ignoring some possible alternative interpretations, and needs to be extensively rewritten. There needs to be a complete overhaul of both the Introduction and Discussion to address the points made by the reviewers.

2. Multiple reviewers specifically cited problems with the data in Figure S4, which appeared unconvincing or even contradictory to the stated interpretation. At the very least, this issue needs to be addressed in the text, if not in the data itself.

3. Most of the results are based on growth data, but was interpretated in the context of cellular redox states. It was disappointing that the efforts to probe cellular redox states were not very strong, as described by Reviewer 1, weakening the arguments (see below).

4. The consensus of the reviewers was that such measurements would greatly strengthen the paper. Alternatively, it may be possible to modify the text to reflect the alternative models and potential ambiguities inherent in the approach.

5. A major component of the proposed model is that PFOR is specifically active highly aerobic conditions, but this claim is never directly tested, and evidence from simple deletion mutant without tested for redox states activity does not fully test this possibility, see comments by Reviewer 2.

6. Further, the data was interpreted in the context of redox state, but as Reviewer 3 points out, the available thermodynamic driving force should be very different under the different growth conditions, and may in itself necessitate a shift in pathway, Thus, the results can be reasonably interpreted without the need to involve "inactivation" of enzymes. This, or course, could be settled by enzymatic assays and redox probes.

Reviewer #2 (Recommendations for the authors):

Most of my comments are already present in the public review. However, the authors should seriously consider making measurements of cells in vivo where they're tracking pyridine nucleotide (blue-green) fluorescence. Such measurements are primarily tracking and NADPH rather than NADH, nevertheless it could prove very informative regarding the ability for the cells to assimilate photosynthetic reductant as the light comes on. Such measurements can be performed using a PAM fluorometer, for example, with the appropriate attachments. Also, contemporaneous measurements of chlorophyll fluorescence can provide valuable information on linear and cyclic photosynthetic electron flow.

Reviewer #3 (Recommendations for the authors):

I very much like the performed experiments and find that they are well-planned, professionally performed and that they provide solid evidence for the discussed switch of pathways. However, I feel that the discussion is unconvincing in certain places and tries to convey concepts which I would consider doubtful at the least. I will be happy if the authors can prove me wrong.

– Overly affirmative preconception: on line 46 and also subsequently in the manuscript, you claim that ferredoxins are older than NAD. You base your statement on the observation that ferredoxins are present in all three domains (I think that we have moved on from Woose's royalist kingdoms to the term "domains";-) of life. Isn't this also true for NAD? You may well be right that FeS-clusters preceeded NAD during life's emergence but so far I don't think that we have any evidence for LUCA lacking NAD. In your scenario NAD would even come in only much later when the environment became more oxidizing.

Reducing and oxidizing conditions:

– eg. l 289 ll: it is difficult for me to see how you imagine maintaining reducing conditions in the presence of oxygen (in any setting, cells or in vitro or whatever). Of course, it all depends on what you call reducing … However, in your manuscript, reducing is always connoted with NADH or ferredoxin, so well below the O2/superoxide transition. If you have O2 under such "reducing" conditions and some undefined one-electron compound (e.g. ferredoxins) floating around, they have no choice but reduce O2 to O2.-. It then all depends on who is dominant (in abundancy), the reductant or O2. If it is O2, then you will loose your reducing conditions, if it is the reductant, you will turn anaerobic. So I honestly don't understand what you mean by "reducing conditions in the presence of oxygen". By contrast, spatial separation (your first alternative) is of course an option.

– l303 ll: You claim that loss of hydrogen to space will "withdraw electrons and …[thereby] … establish oxidizing conditions". You have reducing conditions if there is an abundance of reductant and oxidizing conditions if you have an abundance of oxidant. Reducing the amount of reductant therefore doesn't take you to oxidizing conditions! On an O2-free planet, the reductants may have been H2, Fe2+, CH4, formate etc and the oxidants CO2, SO42-, Fe3+ and the likes. On line 305 you say that the loss of H2 would lead to higher driving forces. How? You don't add any new (more positive) oxidant, you only take out reductant. Therefore the driving force for any individual electron transfer reaction will stay put while the deltaG of the bulk reaction will even decrease (since you now have lower amounts of reductant).

– l366 ll: I can see what you want to say on these lines and you are basically correct. However, this is how you would explain the situation to a colleague from the humanities (no condescension intended!). I would hold that in a life science article you can assume that the reader is able to digest the Nernst equation and therefore comprehend that there is a mathematical relationship between the standard (or midpoint) potential at equal amount of oxidized and reduced form of a redox couple and the effective potential at given ratios of ox to red.

– 370 ll: "slow down the back reaction and speed up the forward reaction". Ambiguous and potentially wrong. The reaction ks are of course unaffected but mass-action results in shifted equilibrium concentrations of donor and acceptor when deltaE changes.

– 372 ll: (my main concern regarding the entire manuscript!) You go on the conclude that upon increasing NADH over NAD+ you shift the effective potential more negative. Perfectly correct! Now here is my question: On lines 183 – 198 of the Results section, you hypothesize on an inactivation of PDH via binding of NAD to the E3 subunit. Isn't it enough that you have increasing levels of NADH? (they call this product inhibition) To my mind, the switch from PDH (using NAD+) to PFOR (using Fd) is a thermodynamic necessity since PDH would cease to operate at very high NADH/NAD+ ratios anyway. Oxygenic photosynthesis in general has a problem of redox equilibration: There is an essentially infinite pool of electron donor (water) and a very limited pool of electron acceptor (CO2 to generate biomass). Hence all the regulation and cyclic transfer and overflow valves and what have you. Of course the situation is aggravated when the beasts go photomixotrophic since they are flooded with viciously reducing electrons (without having seen a single photon). Their intracellular redox poise will rapidly go to red alert reducing (NB: life needs electron transfer from red to ox to make ATP on the way; when all gets reduced through, doom looms!). Which brings me to my final question: Have you checked that your photomixotrophic cultures don't produced H2 by any chance? Channeling electrons through PFOR into the Fd pool would potentially open up the way to hydrogenase and thereby help getting rid of excess reducing equivalents via the emergency valve "proton reduction".

The work demonstrates that PFOR-mediated decarboxylation of pyruvate replaces the (more canonical) PDH-based pathway cells of the cyanobacterium Synechocystis sp. PCC6803 are shifted from photoautotrophic to photomixotrophic growth conditions. The reviewers recognized the novelty and potential importance of the work, in particular on the implications for the evolution of different enzymes and metabolic pathways under aerobic and anaerobic conditions.

1. On the other hand, all of the reviewers had similar criticisms that the text was greatly over-interpreted the results, while ignoring some possible alternative interpretations, and needs to be extensively rewritten. There needs to be a complete overhaul of both the Introduction and Discussion to address the points made by the reviewers.

We thank all reviewers for reading our manuscript thoroughly and for raising very interesting discussion points. We addressed all raised points either by responding to the reviewers directly and by overhauling both the introduction and the discussion part.

The introduction was extended by alternative interpretations. Furthermore, animated by the requests of the reviewers we searched the literature again and found a recent study from E. coli, which was added both in the introduction and discussion. The authors found that PFOR in E. coli supports metabolism during stationary phase in the presence of oxygen in an E coli. mutant (1).

The discussion was completely reorganized and overhauled and now starts with a summary of the findings that are best supported by our data. We then added two sentences to stress that the following lines include hypothetical consideration, as suggested:

“The following lines include theoretical reflections based on the presented data. However, as these conclusions are not entirely supported by the data, they should be regarded as hypothetical and are meant as thought-provoking impulses.”

We very much hope that this overhaul is satisfying.

2. Multiple reviewers specifically cited problems with the data in Figure S4, which appeared unconvincing or even contradictory to the stated interpretation. At the very least, this issue needs to be addressed in the text, if not in the data itself.

We agree that data from Figure S4 (kinase mutants) as well as data concerning the phosphorylation of the PDH complex are not convincing. We were uncertain from the very beginning if it would be a good idea to include these data sets and having this in mind discussed the data very cautiously. However, these data are not really required for the statement of our manuscript, as the enzymatic tests on the E3 subunit of the PDH complex at different NADH concentration convincingly show, that the PDH complex gets inhibited at high NADH levels. Therefore, we now decided to take these data sets (kinase mutants Figure S4 and phosphorylation of the PDH complex) out of the manuscript. We thank the reviewers for their criticism as we feel that the manuscript improved and is more focused now.

3. Most of the results are based on growth data, but was interpretated in the context of cellular redox states. It was disappointing that the efforts to probe cellular redox states were not very strong, as described by Reviewer 1, weakening the arguments (see below).

This is an important point. Thank you very much. We agree with this criticism and now addressed it by including in vivo measurements of the redox state of NAD(P) and ferredoxin. The presented results strongly support our original interpretation that the redox state is more reduced under photomixotrophic conditions compared to photoautotrophic conditions (see new Figure 1 —figure supplement 4).

4. The consensus of the reviewers was that such measurements would greatly strengthen the paper. Alternatively, it may be possible to modify the text to reflect the alternative models and potential ambiguities inherent in the approach.

The new results were added in Figure 1 —figure supplement 4 and the text was modified accordingly, stating that both NAD(P)H and ferredoxin pools are reduced to a greater extent under photomixotrophic conditions.

5. A major component of the proposed model is that PFOR is specifically active highly aerobic conditions, but this claim is never directly tested, and evidence from simple deletion mutant without tested for redox states activity does not fully test this possibility, see comments by Reviewer 2.

We absolutely agree that direct testing of the PFOR under different conditions and oxygen concentrations would be much more convincing. Such tests have been unsuccessful in our hands for the cyanobacterial enzyme until now. It can be very difficult to detect activity in the presence of oxygen in vitro for redox enzymes, as there might be mechanisms inside the cell that continuously reactivate the enzyme upon inactivation. This is e.g. known for FeFe-hydrogenases that are irreversibly inactivated by oxygen in vitro but stay active or can be reactivated in vivo. The mimic of such reactivation was not accomplished in vitro, yet. Therefore, for the time being we are unfortunately unable to provide further evidence. However, we think that our collected data strongly indicates that PFOR is an important enzyme also under aerobic conditions and discuss the presented evidence with appropriate caution. In addition, animated by suggestions of the reviewers we found a recent report on an E. coli mutant, in which PFOR contributes to the metabolic stationary phase of aerobic cultures even though the enzyme requires anaerobic conditions in the in vitro test (1, 2). This report is now mentioned in the introduction and discussion.

6. Further, the data was interpreted in the context of redox state, but as Reviewer 3 points out, the available thermodynamic driving force should be very different under the different growth conditions, and may in itself necessitate a shift in pathway, Thus, the results can be reasonably interpreted without the need to involve "inactivation" of enzymes. This, or course, could be settled by enzymatic assays and redox probes.

We agree with the point raised by Reviewer 3. Our data show that high NADH levels inactivate the PDH complex. However, this does not rule out that the general shift to more reducing conditions as shown by our new measurements (Figure 1 —figure supplement 4), which will change the thermodynamic driving force of redox reactions, in itself necessitates shifts in pathways and might be e.g. valid for the shift between the importance of N-GOGAT relative to F-GOGAT. Both mechanisms might be of importance. This point was added in line 363 of the discussion.

“In addition, a shift to more reducing conditions (Figure 1 —figure supplement 4), will alter the thermodynamic driving force of many redox reactions, and may in itself necessitate a shift in pathways.”

Reviewer #2 (Recommendations for the authors):

Most of my comments are already present in the public review. However, the authors should seriously consider making measurements of cells in vivo where they're tracking pyridine nucleotide (blue-green) fluorescence. Such measurements are primarily tracking and NADPH rather than NADH, nevertheless it could prove very informative regarding the ability for the cells to assimilate photosynthetic reductant as the light comes on. Such measurements can be performed using a PAM fluorometer, for example, with the appropriate attachments. Also, contemporaneous measurements of chlorophyll fluorescence can provide valuable information on linear and cyclic photosynthetic electron flow.

Thank you very much for this comment. As already stated, the NAD(P)H-module of the PAM does not distinguish NADPH and NADH. However, using this device we now show that NAD(P)H is more strongly reduced under photomixotrophy. In addition, with the Dual-KLAS/NIR we also have evidence that this is also true for the ferredoxins (newly added results in Figure 1 —figure supplement 4).

Reviewer #3 (Recommendations for the authors):

As stated in the public review parts, I very much like the performed experiments and find that they are well-planned, professionally performed and that they provide solid evidence for the discussed switch of pathways. However, I feel that the discussion is unconvincing in certain places and tries to convey concepts which I would consider doubtful at the least. I will be happy if the authors can prove me wrong.

– Overly affirmative preconception: on line 46 and also subsequently in the manuscript, you claim that ferredoxins are older than NAD. You base your statement on the observation that ferredoxins are present in all three domains (I think that we have moved on from Woose's royalist kingdoms to the term "domains";-) of life. Isn't this also true for NAD? You may well be right that FeS-clusters preceeded NAD during life's emergence but so far I don't think that we have any evidence for LUCA lacking NAD. In your scenario NAD would even come in only much later when the environment became more oxidizing.

Thank you very much for this point. We did not mean to say that LUCA lacked NAD. The point we wanted to make was rather, that NAD gained importance over oxygen sensitive ferredoxins during the evolution from an anaerobic to an aerobic metabolism as also stated by Gould et al., 2019 (6). In order to be more precise, we now stressed that NAD(P)H gained importance to make sure that it was as well present already.

Accordingly, we changed the former sentence “In addition, ferredoxins have in general been complemented or replaced by NAD(P)H as alternative, oxygen-insensitive reducing agents in aerobes” in line 80 to: “In addition, NAD(P)H has gained importance as alternative, oxygen-insensitive reducing agent in aerobes and thereby complemented or replaced oxygen sensitive ferredoxins, that are useful for anaerobes.”

Reducing and oxidizing conditions:

– eg. l 289 ll: it is difficult for me to see how you imagine maintaining reducing conditions in the presence of oxygen (in any setting, cells or in vitro or whatever). Of course, it all depends on what you call reducing … However, in your manuscript, reducing is always connoted with NADH or ferredoxin, so well below the O2/superoxide transition. If you have O2 under such "reducing" conditions and some undefined one-electron compound (e.g. ferredoxins) floating around, they have no choice but reduce O2 to O2.-. It then all depends on who is dominant (in abundancy), the reductant or O2. If it is O2, then you will loose your reducing conditions, if it is the reductant, you will turn anaerobic. So I honestly don't understand what you mean by "reducing conditions in the presence of oxygen". By contrast, spatial separation (your first alternative) is of course an option.

This is again a good point. Thank you very much. We agree fully. We changed the sentence by adding, that mechanisms might be present that consume the oxygen.

“It might alternatively be that living cells have the ability to maintain reducing conditions in the presence of oxygen by yet unknown mechanisms that e.g. consume oxygen, which is a challenge in enzymatic in vitro assays.” (line 287 following)

– l303 ll: You claim that loss of hydrogen to space will "withdraw electrons and …[thereby] … establish oxidizing conditions". You have reducing conditions if there is an abundance of reductant and oxidizing conditions if you have an abundance of oxidant. Reducing the amount of reductant therefore doesn't take you to oxidizing conditions! On an O2-free planet, the reductants may have been H2, Fe2+, CH4, formate etc and the oxidants CO2, SO42-, Fe3+ and the likes. On line 305 you say that the loss of H2 would lead to higher driving forces. How? You don't add any new (more positive) oxidant, you only take out reductant. Therefore the driving force for any individual electron transfer reaction will stay put while the deltaG of the bulk reaction will even decrease (since you now have lower amounts of reductant).

We indeed think that the loss of hydrogen established oxidizing conditions. The oxidant was present by water splitting at PSII and the evolution of oxygen. However, the accumulation of oxygen was only possible due to the loss of reductant (H2 into space on the one hand and burial of reduced carbon on the other hand). Without this loss of reductant, respiration would have consumed all the oxygen quickly and nothing would have changed. However, by the withdrawal of H2 and reduced carbon from the equation, oxygen could accumulate. Which means that the high abundance of oxidant (O2) indeed required the removal of reductant (H2).

– l366 ll: I can see what you want to say on these lines and you are basically correct. However, this is how you would explain the situation to a colleague from the humanities (no condescension intended!). I would hold that in a life science article you can assume that the reader is able to digest the Nernst equation and therefore comprehend that there is a mathematical relationship between the standard (or midpoint) potential at equal amount of oxidized and reduced form of a redox couple and the effective potential at given ratios of ox to red.

We deleted this paragraph as suggested, when we rewrote and overhauled the discussion.

– 370 ll: "slow down the back reaction and speed up the forward reaction". Ambiguous and potentially wrong. The reaction ks are of course unaffected but mass-action results in shifted equilibrium concentrations of donor and acceptor when deltaE changes.

We fully agree with this statement. However, we think that our text was not clear enough as there might to be a misunderstanding. The point we wanted to make is related to a shift from using ferredoxin instead of NAD⁺ as a redox acceptor. In this case reaction ks are affected and deltaE decreases. This might not have been clear enough. However, as we rewrote and completely overhauled the discussion as suggested, these sentences were deleted.

– 372 ll: (my main concern regarding the entire manuscript!) You go on the conclude that upon increasing NADH over NAD+ you shift the effective potential more negative. Perfectly correct! Now here is my question: On lines 183 – 198 of the Results section, you hypothesize on an inactivation of PDH via binding of NAD to the E3 subunit. Isn't it enough that you have increasing levels of NADH? (they call this product inhibition) To my mind, the switch from PDH (using NAD+) to PFOR (using Fd) is a thermodynamic necessity since PDH would cease to operate at very high NADH/NAD+ ratios anyway.

In E. coli an alteration in a single amino acid of the E3 subunit of the PDH complex reduced the sensitivity of the PDH complex to inhibition via NADH and allowed decarboxylation of pyruvate at high NADH levels (9). The respective amino acid was not part of the binding pocket for NAD⁺. The mechanisms is thus not a simple product inhibition. NADH seems to have a regulatory function here and seems to bind as a regulator at a site that does not belong to the active site of the enzyme. The enzyme activity measurements that we did with the E3 subunit of Synechocystis at different NADH concentration does not allow to determine if NADH inhibited the enzyme in a regulatory manner or in the manner of classical product inhibition. We therefore left this point open.

Oxygenic photosynthesis in general has a problem of redox equilibration: There is an essentially infinite pool of electron donor (water) and a very limited pool of electron acceptor (CO2 to generate biomass). Hence all the regulation and cyclic transfer and overflow valves and what have you. Of course the situation is aggravated when the beasts go photomixotrophic since they are flooded with viciously reducing electrons (without having seen a single photon). Their intracellular redox poise will rapidly go to red alert reducing (NB: life needs electron transfer from red to ox to make ATP on the way; when all gets reduced through, doom looms!). Which brings me to my final question: Have you checked that your photomixotrophic cultures don't produced H2 by any chance? Channeling electrons through PFOR into the Fd pool would potentially open up the way to hydrogenase and thereby help getting rid of excess reducing equivalents via the emergency valve "proton reduction".

Yes, we checked the cultures for hydrogen production, but we could not detect any. The hydrogenase is very likely not able to produce hydrogen due to the presence of oxygen. A valve to get rid of excess reducing equivalents that the cells use under photomixotrophic conditions is e.g. to enhance their glycogen levels.

References

1. S. Li et al., Dynamic control over feedback regulatory mechanisms improves NADPH flux and xylitol biosynthesis in engineered E. coli. Metab Eng 64, 26-40 (2021).

2. T. Nakayama, S. Yonekura, S. Yonei, Q. M. Zhang-Akiyama, Escherichia coli pyruvate:flavodoxin oxidoreductase, YdbK – regulation of expression and biological roles in protection against oxidative stress. Genes Genet Syst 88, 175-188 (2013).

3. A. Witt, R. Pozzi, S. Diesch, O. Hädicke, H. Grammel, New light on ancient enzymes – in vitro CO2 Fixation by Pyruvate Synthase of Desulfovibrio africanus and Sulfolobus acidocaldarius. The FEBS Journal 286, 4494-4508 (2019).

4. C. Cassier-Chauvat, F. Chauvat, Function and Regulation of Ferredoxins in the Cyanobacterium Synechocystis PCC6803: Recent Advances. Life 4, 666-680 (2014).

5. M. Müller et al., Biochemistry and Evolution of Anaerobic Energy Metabolism in Eukaryotes. Microbiology and Molecular Biology Reviews 76, 444 (2012).

6. S. B. Gould et al., Adaptation to life on land at high O2 via transition from ferredoxin-to NADH-dependent redox balance. Proceedings of the Royal Society B: Biological Sciences 286, 20191491 (2019).

7. O. Schmitz, J. Gurke, H. Bothe, Molecular evidence for the aerobic expression of nifJ, encoding pyruvate : ferredoxin oxidoreductase, in cyanobacteria. FEMS Microbiol. Lett. 195, 97-102 (2001).

8. K. Gutekunst et al., LexA regulates the bidirectional hydrogenase in the cyanobacterium Synechocystis sp. PCC 6803 as a transcription activator. Molecular Microbiology 58, 810-823 (2005).

9. Z. Sun et al., Amino acid substitutions at glutamate-354 in dihydrolipoamide dehydrogenase of Escherichia coli lower the sensitivity of pyruvate dehydrogenase to NADH. Microbiology 158, 1350-1358 (2012).

Author details

1. Yingying Wang

   Department of Biology, Botanical Institute, Christian-Albrechts-University, Kiel, Germany

   Contribution

   Conceptualization, Data curation, Investigation, Methodology, Writing - original draft, Writing - review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-0603-6691

2. Xi Chen

   Department of Biology, Botanical Institute, Christian-Albrechts-University, Kiel, Germany

   Contribution

   Data curation, Investigation, Methodology, Writing - review and editing

   Competing interests

   No competing interests declared

3. Katharina Spengler

   Department of Biology, Botanical Institute, Christian-Albrechts-University, Kiel, Germany

   Contribution

   Data curation, Investigation, Methodology, Writing - review and editing

   Competing interests

   No competing interests declared

4. Karoline Terberger

   Department of Biology, Botanical Institute, Christian-Albrechts-University, Kiel, Germany

   Contribution

   Data curation, Investigation, Methodology, Writing - review and editing

   Competing interests

   No competing interests declared

5. Marko Boehm

   1. Department of Biology, Botanical Institute, Christian-Albrechts-University, Kiel, Germany
   2. Department of Molecular Plant Physiology, Bioenergetics in Photoautotrophs, University of Kassel, Kassel, Germany

   Contribution

   Data curation, Investigation, Methodology, Supervision, Writing - review and editing

   Competing interests

   No competing interests declared

6. Jens Appel

   1. Department of Biology, Botanical Institute, Christian-Albrechts-University, Kiel, Germany
   2. Department of Molecular Plant Physiology, Bioenergetics in Photoautotrophs, University of Kassel, Kassel, Germany

   Contribution

   Data curation, Investigation, Methodology, Supervision, Writing - review and editing

   Competing interests

   No competing interests declared

7. Thomas Barske

   Plant Physiology Department, University of Rostock, Rostock, Germany

   Contribution

   Data curation, Investigation, Methodology, Writing - review and editing

   Competing interests

   No competing interests declared

8. Stefan Timm

   Plant Physiology Department, University of Rostock, Rostock, Germany

   Contribution

   Data curation, Investigation, Methodology, Writing - review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-3105-6296

9. Natalia Battchikova

   Department of Biochemistry, Molecular Plant Biology, University of Turku, Turku, Finland

   Contribution

   Conceptualization, Data curation, Investigation, Writing - review and editing

   Competing interests

   No competing interests declared

10. Martin Hagemann

    Plant Physiology Department, University of Rostock, Rostock, Germany

    Contribution

    Conceptualization, Formal analysis, Funding acquisition, Supervision, Writing - review and editing

    Competing interests

    No competing interests declared

11. Kirstin Gutekunst

    1. Department of Biology, Botanical Institute, Christian-Albrechts-University, Kiel, Germany
    2. Department of Molecular Plant Physiology, Bioenergetics in Photoautotrophs, University of Kassel, Kassel, Germany

    Contribution

    Conceptualization, Formal analysis, Funding acquisition, Project administration, Supervision, Writing - original draft, Writing - review and editing

    For correspondence

    kirstin.gutekunst@uni-kassel.de

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0003-4366-423X

• 2,161

  Page views

• 293

  Downloads

• 9

  Citations

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