1Redox tuning in photosystem II 3John F. Allen1,* and Jon Nield2,* 51Research Department of Genetics, Evolution and Environment, Darwin Building, 6University College London, Gower Street, London, United Kingdom, WC1E 6BT 72School of Biological and Chemical Sciences, Queen Mary University of London, 8Mile End Road, London, United Kingdom, E1 4NS 10*Correspondence: j.f.allen@ucl.ac.uk (J. F. Allen); j.nield@qmul.ac.uk (J. Nield) 11http://jfallen.org (J. F. Allen); http://macromol.sbcs.qmul.ac.uk/ (J. Nield) 12@ProfJohnAllen (J. F. Allen) 13 14 15ABSTRACT 16 17In photosynthesis, oxygen is liberated from water, not from CO2, however this model 18has been silent on why photosynthesis requires bicarbonate. Rutherford and 19colleagues solve this problem elegantly: bicarbonate tunes wateroxidising 20photosystem II to make onward electron transfer efficient; absence of bicarbonate 21retunes, redirects, and safely shuts down energy flow. 22 23MAIN TEXT 24 25Electron transfer is the universal currency of biological energy transduction. Loss of 26an electron is oxidation; gain of an electron is reduction. Any reductionoxidation – 27“redox” – reaction has a direction determined by the electrochemical potentials of its 28participating atoms or molecules. Each has a “midpoint redox potential” (Em) – the 29electrical potential, measured in Volts (V), when the electrons leave as fast as they 30arrive. At equal concentrations of their oxidised and reduced states, an electron’s 31donor has a lower Em than its acceptor. The central event in photosynthesis breaks 1 32this rule because a photon absorbed by chlorophyll forces an electron from a reluctant 33donor, with a high (positive) Em, to a reluctant acceptor at lower (more negative) 34potential. Physical separation of electric charge captures the energy of the photon – 35light energy is converted into electrochemical potential 36 37Reaction centres 38At this photochemical “reaction centre” there is a tendency for the electron to return 39whence it came; to the newly photooxidised chlorophyll molecule. The chemistry 40employed by photosynthesis is to pass the electron, instead, along a chain of 41secondary acceptors, much like those in a respiratory chain. Each link in the chain 42accepts and donates at a slightly higher Em value than the last, while each forward 43electron transfer must be fast enough to outcompete the return of the electron to 44willing acceptors in wasteful and potentially destructive “back reactions” 45 46Photosystem II 47In most photosynthetic organisms a chlorophyll a molecule is the primary electron 48donor in both photosystems I and II. In photosystem II the Em of its redox couple is 49+1.2 V – enough to remove four electrons eventually from water, liberating oxygen, at 50an Em of +0.82 V. The primary electron acceptor of photosystem II, pheophytin at an 51Em of –0.5 V , passes one electron to a quinone, QA, giving QA•–. From there the 52electron passes to a second quinone, QB. A second electron, from QA•– to QB•–, allows 53formation of a stable, protonated molecule of plastoquinol, QH2 . This twoelectron 54gate thus accepts two protons from the bacterial cytoplasm or chloroplast stroma in a 55second type of energy coupling – from electron transport to transmembrane proton 56motive force. 57 58In order to understand how energy is conserved or dissipated it is essential to know 59the Em of the couple QA/QA•–. Frustratingly, the literature records widely differing 60measured values, all from carefully executed experiments on welldefined 61cyanobacterial or chloroplast membranes or membrane fractions. 62 2 63Brinkert et al. provide an explanation of these various Em values, and connect loose 64ends regarding the protection of photosynthesis from high light, the emergence of 65oxygenic from anoxygenic photosynthesis, and the strange dependency of oxygen 66evolution on the presence of carbon dioxide 67 68Two midpoints – the manganese cluster 69A difference of about 150 mV is seen in the Em of QA/QA•– centres in the presence or 70absence of the Mn4CaO5 inorganic catalyst of water oxidation at the electron donor 71side of photosystem II. A physiological analogue of this shift in potential occurs 72during assembly of photosystem II, when the manganese cluster is added to the 73reaction centre protein complex in a step that requires light . In photosystem II, 74reaction centres are heterodimers, and a bicarbonate ion is seen to sit adjacent to the 75iron atom that lies on the axis of symmetry between the two transmembrane 76polypeptide chains, D1 and D2, and therefore between QA and QB (Figure 1). The 77iron atom is not redox active. Nevertheless it plays a role in electron transfer between 78the two quinones. The iron atom is held in place by histidines of the bacterial L and M 79subunits and of their homologues, D1 and D2, respectively, of photosystem II. In 80reaction centres of purple bacteria , which are anoxygenic, a conserved glutamate side 81chain replaces the noncovalently bound bicarbonate of photosystem II (Figure 1) 82 83Two more midpoints – bicarbonate 84The standard method of determining Em values is potentiometric redox titration. This 85technique measures the ratio of oxidised and reduced form of a chemical species over 86a range of potentials, each obtained under strictly anoxic conditions so that chemical 87oxidants and reductants can be added to achieve the stable, poised potential. The 88routine way of achieving anoxia is to bubble the reaction vessel or cuvette with inert 89gas, nitrogen or argon, to displace air and, if required, to expel a sample. A different 90technique, using a transparent, thin, electrochemical cell, confirmed the 150 mV Em 91difference for QA/QA•– between photosystem II with and without the manganese 92cluster , while the two values were both about 80 mV lower than those previously 93agreed on the basis of conventional redox titration. Brinkert et al. reasoned that the 94bicarbonate ion could explain the difference – bubbling with inert gas would deplete 3 95the sample of carbon dioxide and therefore of bicarbonate, HCO3–, from the carbonic 96anhydrase reaction: 97 98 CO2 + H2O ⇌ HCO3– + H+ 99 100In contrast, the electrochemical cell of Shibamoto et al. would not be expected to 101remove the bicarbonate that interacts with the iron atom guiding electron transfer 102from QA (Figure 1). Accordingly, Brinkert et al. carried out their redox titration on 103photosystem II reaction centres, this time in the presence and absence of 1 mM 104bicarbonate. Without bicarbonate the Em value was –60 mV in the Mncontaining 105preparation and +64 mV with Mn depleted. With bicarbonate present, both Em values 106dropped, as predicted, to –124 mV with Mn and to –22 mV without . Brinkert et al. 107also demonstrate that accumulation of QA•– at high light decreases binding of 108bicarbonate to photosystem II . 109 110The safety valve 111A clear inference is that return of the electron from QA•– to pheophytin can occur when 112bicarbonate is in place and when QB is present as QBH2. This transfer is uphill, but the 113hill is less steep than the “cliff” presented when bicarbonate is absent. Furthermore, 114this reduction of pheophytin produces a high proportion of its triplet state, which 115converts groundstate oxygen to singlet oxygen, a toxic product. Depletion of CO2 116will thus do two things. First, a slowdown of the BensonCalvinBassham cycle of 117CO2 assimilation will feed back, preventing oxidation of QBH2 and so allowing QA•– 118no option for forward electron transport. Second, removal of the bicarbonate ion will 119increase the Em of QA, favouring a safer back reaction to chlorophyll or to the donor 120side of photosystem II through cytochrome b559 121 122Origins of oxygen 123Photosynthesis first evolved in a world devoid of free oxygen , and purple bacterial 124photosynthesis today occurs only under anoxic conditions. There the glutamate 125adjacent to the iron atom (Fig 1) is fixed in place – harmful singlet oxygen cannot be 126produced. When the first cyanobacterium learned to use its quinonecontaining 4 127reaction centre to make oxygen , QA•– had urgently to find a safe back reaction. 128Replacement of the glutamate by a removable bicarbonate ion provided the solution. 129 130In explaining the CO2requirement of photosystem II, Brinkert et al. also finally lay to 131rest the ghost of the longabandoned theory that primary photochemistry splits CO2 to 132give O2 plus a C atom that becomes hydrated. The real primary event is 133transmembrane electron transfer, and CO2 assimilation itself does not require light at 134all. Well, we knew that. But the wrong model persisted in popular science writing, 135and it is satisfying to understand, now, that CO2 tunes redox chemistry at the interface 136of our biosphere with energy from the sun 137 138Acknowledgements 139J.F.A. holds an Emeritus Research Fellowship of the Leverhulme Trust. J.N. 140acknowledges support from the CREST program of the Japan Science and 141Technology Agency 142 143References 1441 Umena, Y., et al. (2011) Crystal structure of oxygenevolving photosystem II at a 145resolution of 1.9 Ångström. Nature 473, 55U65 1462 Kato, Y., et al. (2009) Spectroelectrochemical determination of the redox potential 147of pheophytin a, the primary electron acceptor in photosystem II. Proceedings of the 148National Academy of Sciences of the United States of America 106, 1736517370 1493 Müh, F., et al. (2012) Lightinduced quinone reduction in photosystem II. 150Biochimica et Biophysica ActaBioenergetics 1817, 4465 1514 Brinkert, K., et al. (2016) Bicarbonateinduced redox tuning in Photosystem II for 152regulation and protection. Proceedings of the National Academy of Sciences 113, 1531214412149 1545 Johnson, G.N., et al. (1995) A change in the midpoint potential of the quinone Q(A) 155in photosystem II associated with photoactivation of oxygen evolution. Biochimica et 156Biophysica ActaBioenergetics 1229, 202207 5 1576 Deisenhofer, J., et al. (1995) Crystallographic refinement at 2.3 Å Resolution and 158Refined Model of the Photosynthetic Reaction Centre from Rhodopseudomonas 159viridis. Journal of Molecular Biology 246, 429457 1607 Shibamoto, T., et al. (2010) Speciesdependence of the redox potential of the 161primary quinone electron acceptor Q(A) in photosystem II verified by 162spectroelectrochemistry. FEBS Letters 584, 15261530 1638 Nishimura, T., et al. (2016) The Nterminal sequence of the extrinsic PsbP protein 164modulates the redox potential of Cyt b(559) in photosystem II. Scientific Reports 6 1659 Knoll, A.H., et al. (2016) Life: the first two billion years. Philosophical 166Transactions of the Royal Society B: Biological Sciences 371 16710 Allen, J.F. (2005) A redox switch hypothesis for the origin of two light reactions in 168photosynthesis. FEBS Letters 579, 963968 16911 Cardona, T., et al. (2012) Charge separation in Photosystem II: A comparative and 170evolutionary overview. Biochimica et Biophysica ActaBioenergetics 1817, 2643 17112 Fischer, W.W., et al. (2016) Evolution of Oxygenic Photosynthesis. In Annual 172Review of Earth and Planetary Sciences, Vol 44 (Jeanloz, R. and Freeman, K.H., eds), 173pp. 647683 174 175 176Figure legend 177 178Figure 1 179 180Comparison of structures of oxygenic and anoxygenic type II reaction centres at the 181electron acceptor side. Panel A is a view, within the membrane plane, of the 182cyanobacterial photosystem II reaction centre 3WU2.pdb based upon its entry in the 183Orientations of Proteins in Membranes (OPM) database (http://opm.phar.umich.edu/) 184A 25 Å sphere of key sidechain residues is centred on the nonhaem iron, Fe, of the 185D1 protein. Panel B is the corresponding view of 1PRC.pdb model of the 186photosynthetic reaction centre of Rhodopseudomonas viridis 3WU2.pdb where D1 is 187subunit L, D2 is M, and QA and QB are menaquinone (MQ7) and ubiquinone (UQ1). 6 188Panels C and E show the structure in A rotated 90º; panels D and F show structure 189in B rotated 90º. C and D are “top” views normal to the membrane plane; E and F are 190views parallel to the membrane plane and along the direction of electron transfer. 191Using the defaults of the MatchMaker structure comparison tool of UCSF Chimera 192(https://www.cgl.ucsf.edu/chimera/), the best aligning pair of chains within 1933WU2.pdb and 1PRC.pdb were overlaid and found to be D2 protein of 3WU2 with 194the M chain of 1PRC. Side chain residues in A and B are labelled based upon chain, 195name and amino acid position e.g. D1His 272. BCT = bicarbonate ion; Fe = Fe (II) 196ion; QA, QB = Plastoquinone; MQ7 = Menaquinone7; UQ1 = Ubiquinone1. 197Coloured labels reflect the 'residues' they represent; Glutamate, cornflower blue; 198Tyrosine, magenta; Lysine, blue; BCT the bicarbonate ion, brown with oxygen atoms 199in red; Histidine, cyan; Isoleucine, yellow; Glycine, purple; Serine, red; Alanine, 200green; Fe (II) ion, brown; the quinones QA/QB/MQ7/UQ1, hot pink. For an 201appreciation of scale, the centretocentre distance of the benzene rings of the 202quinones QA and QB is ~ 22 Å; panels A to F are at the same zoom level, a sphere of ~ 20325 Å, centred on the Fe (II) ion. Panel G is zoomed out to encompass the quinones' 204isoprenoid chains, the intriguing Fe (II) ion of the PsbF subunit of Cytochrome b559 , 205which is 18 to 19 Å distant from the QB isoprenoid tail, and the protein backbone is 206shown as ribbons; 3WU2.pdb is coloured transparent brown, 1PRC.pdb transparent 207blue. The slab view depth ~ 75 Å and overall placement of this alignment within the 2083WU2.pdb photosystem II dimer are shown as the inset (a) 209 7 ... 150Biochimica et Biophysica ActaBioenergetics 1817, 4465 1514 Brinkert, K., et al. (2016) Bicarbonateinduced? ?redox? ?tuning? ?in? ?Photosystem? ?II for 152regulation and protection. Proceedings of the National Academy of Sciences 113, ... 198Tyrosine, magenta; Lysine, blue; BCT the bicarbonate ion, brown with oxygen atoms 19 9in? ?red; Histidine, cyan; Isoleucine, yellow; Glycine, purple; Serine, red; Alanine, 200green; Fe (II) ion, brown; the quinones QA/QB/MQ7/UQ1, hot pink. For an 201appreciation of scale, the centretocentre distance of the benzene rings of the ... 40employed by photosynthesis is to pass the electron, instead, along a chain of 41secondary acceptors, much like those? ?in? ?a respiratory chain. Each link? ?in? ?the chain 42accepts and donates at a slightly higher Em value than the last, while each forward