14 minutes reading time (2899 words)

    How Do Plants Make Oxygen?

    David C. Weatherburn
    Victoria University of Wellington


    This article describes the current state of our knowledge of how plants and other photosynthetic organisms make oxygen. Our understanding is very incomplete and there is considerable doubt about much of the information presented below as facts. This is a very active area of research activity and this article will certainly be out of date by the end of this year.

    Photosynthetic organisms contain "antennae" that collect solar energy and convert it into chemical energy. The chemical energy is used to reduce CO2 to form organic compounds. The overall photosynthetic reaction involves the oxidation of water and the reduction of nicotinamide adenine dinucleotide phosphate, NADP.

    2NADP+ + 2H2O → 2NADPH + O2 + 2H+

    The net energy change for this reaction is 438 kJ mol-1 which corresponds to the energy of a photon with a wavelength of 223nm, i.e. light in the far UV. Even if photosynthesis were 100% efficient (which it is not), more than one photon of lower- energy visible light would be necessary to generate one molecule of 02. It is known experimentally that about ten photons are required.

    The biological apparatus that performs these functions is localised on the "thylakoid" membranes in the chloroplasts of the photosynthetic organisms. A diagrammatic representation of this membrane is shown in Figure 1. The photosynthetic apparatus comprises two components called photosystems, (photosystem I or PSI and photosystem II or PSII), and a cytochrome b6f complex. PSI is separated from PSII by about 100 Angstroms, and the cytochrome b6f complex is apparently located between the two photosystems. Also present in the membrane is the very important enzyme, ATP synthase, which makes ATP from ADP.

    The focus of this article is PSII. This photosystem contains at least 20 different protein molecules, all but three of which are intrinsic to (contained within) the membrane. Also present are about 400 chlorophyll molecules, (both type a and b, I and II) which collect the light, a "special pair" of chlorophyll molecules called P680, carotenoids (III), which are also involved in light harvesting, plastoquinones Qa and Qb (IV), Fe2+, a cluster of four Mnn+ ions; and Cl- , HCO3- and Ca2+. 

     Photosystem II

    PSII can be divided into two parts; the light-harvesting complex, proteins which bind most of the light absorbing pigments (chlorophylls and carotenoids); and a reaction centre. The structure of one type of light harvesting protein (the most abundant in chloroplasts) was determined in 1994. This protein binds seven chlorophyll a and five chlorophyll b molecules and two carotenoids. The chlorophylls are attached via bonds from the magnesium atom of the chlorophyll to amino acid side-chain oxygen or nitrogen atoms, or C=O groups from the amide groups of the protein. The chlorophylls and carotenoids are arranged in a manner that allows easy transfer of energy from an absorbed photon to the reaction centre. The reaction centre is the region in which oxygen evolution occurs. It contains at least thirteen proteins (Table 1), some chlorophyll molecules, the plastoquinones and the inorganic ions. The proteins are of two types; ten membrane bound proteins, which at least in some cases extend into the ceil on both sides of the membrane; and three extrinsic proteins, which are loosely bound to the intrinsic proteins on one side of the membrane. A cartoon of a possible arrangement of some of these proteins and cofactors is shown in Figure 2.

     Not all these proteins are necessary for O2 evolution. Experiments in which some of the proteins are removed suggest that the CP47, CP43, D1, D2, cyt b559 and the extrinsic proteins 33, 24 and 17 are necessary for O2 evolution. The functions of some of these proteins are not known; it is thought however that the extrinsic proteins assist the binding of Mnn+ and Cl-, but they probably do not act as ligands to these ions. Apart from providing the Fe2+ and Qa binding sites, the function of D2 is also unclear. The D1 and D2 proteins contain tyrosines (labelled YZ and YD respectively), which are redox active. The side chain of the tyrosine contains a phenol group, and this is able to lose one electron to form a radical. When this happens, the OH group of the phenol also loses a proton, so that the radical formed is neutral and relatively stable (Scheme 1). YZ is part of the electron transfer pathway in the water-oxidation reaction; the function of YD is not known. YZ is believed to be about 5 A from the manganese cluster and about 10-15 A from P680. The other tyrosine YD is apparently 27 A from the manganese cluster and about 40 A from YZ.

    The photochemistry of PSII

    The first step in the photochemistry of PSII involves absorption of a photon striking a chlorophyll b molecule bound to the light-harvesting protein. The energy of the photon is then transferred to a chlorophyll a, and then to the "special pair" of chlorophylls, (P680) bound to the CP47 protein of the reaction centre. This high-energy form of P680 is a strong reducing agent; it transfers an electron to another chlorophyll molecule (which does not have a bound Mg2+ and is called pheophytin) producing the cation P680+. The electron then moves from pheophytin along the membrane, via the plastoquinones, Qa and Qb, to the cytochrome b6f complex, to plastocyanin then to PSI and finally to NADP. The P680+ cation is reduced by tyrosine YZ on the D1 protein. The tyrosine radical that is formed (YZ.) is in turn reduced by an electron that comes from the cluster of manganese atoms which in turn obtains electrons from the oxidation of water. The electron flow is thus from water, to manganese, to YZ., to P680+, to pheophytin , to Qa, to Qb, to cytochrome b6f, to plastocyanin, to PSI and then finally to NADP.

    How has nature coupled the photochemistry, which is a one photon/one electron process, with the oxidation of water which is a four electron process?

    2H2O → O2 + 4H+ + 4e-

    The site of the water oxidation reaction is the cluster of four manganese atoms, which together with the required ions (Cl- and Ca2+) is able to accumulate sufficient oxidising power. The first real clue as to how this cluster works came from light-flash experiments, in which PSII preparations were held in the dark for 30 minutes and then subjected to intense short flashes of light. O2 evolution did not occur until the third flash, and then peaks of oxygen evolution were observed every fourth flash. The interpretation of these experiments is that dark adapted PSII preparations exist in a stable state called S1. The first three flashes of light each remove one electron from the manganese cluster to produce new states S2, S3 and S4 with increasing oxidising ability. S4 is able to oxidise water to O2 at the same time gaining four electrons to form a state called S0. The next flash of light returns the manganese centre to the stable S1 state, and then the cycle is repeated. The overall scheme is illustrated in Figure 3. 

     There are a number of questions which spring to mind given this scenario.

    • Which protein or proteins bind the manganese atoms?
    • What are the oxidation states of the manganese atoms?
    • Does the increase in oxidation state of the cluster correspond to oxidation of a manganese ion?
    • What is the structure of the manganese cluster?
    • At what stage(s) are the protons evolved?
    • What is the role of the essential Cl-and Ca2+ ions?
    • At what stage does water bind to the manganese cluster?
    • What is the mechanism of dioxygen formation?
    • Are intermediates such as hydrogen peroxide formed during the oxidation?

    At this time it is fair to say that the only answer we can give to all of these questions is that we do not know. We have partial answers, we have hints, we have intelligent guesses, and we have a lot of speculation from researchers of this topic.

    The structure and role of the manganese cluster
    The four manganese atoms are apparently bound to amino acids in the D1 protein but bonds to other reaction-centre proteins (such as CP47) cannot be ruled out. The amino acids bound to manganese ions are believed to be in the portion of the D1 protein which extends into the lumen i.e. the manganese cluster is not buried in the cell membrane. Apparently Ca2+ is also bound to the D1 protein since Ca2+ depleted samples do not proceed beyond the S3 state, and removal of the Ca2+ affects the structure of the manganese cluster in the S1 state. There is considerable debate about how close the Ca2+ and Mn cluster are to one another - some experiments suggest that the metals are within 5A of each other, but others suggest they are much further apart.

    Apparently the S0 → S1 and S1 → S2 transitions involve oxidation of manganese, but the S2 → S3 transition does not. The S4 state lasts for too short a time for us to gain any information about the S3 → S4transition. If manganese is not oxidised in the S2 → S3 transition then what is? A number of suggestions have been made including, YZ, a histidine residue on D1, an oxo ligand bound to Mn, or a water molecule. But the truth is we do not know.

    The structure of the manganese cluster has been the subject of many PhD theses and much debate over the last 15 years. There have been two approaches to the problem. The first tries to determine the structure of the core complex in the biological system, and the second consists of endeavours to synthesise tetranuclear manganese complexes, which may serve as models for the core complex. The model makers have been spectacularly successful in preparing many different types of manganese complexes with four (and up to nineteen) manganese atoms, but they have not yet made compounds that can oxidise water. A selection of the types of structures formed in the laboratory are shown in structures VI-VIII.

    From the evidence available, only VIII is close to the structure of the manganese cluster in the reaction centre. Comparison of the properties of the model complexes VI and VII with those of the reaction centre shows that they are not good models. Researchers studying the biological system have shown that in the S2 state, each Mn has two Mn-O bonds with lengths of approximately 1.8A. Studies show this is only consistent with two oxide ions bridging two manganese atoms. There are Mn neighbours to each Mn atom at a distance of 2.7A which is consistent with the Mn- Mn distance in compounds like VI and VIII but is inconsistent with structure VII.

     The roles of other species

    The most recent evidence suggests that each of the steps from S0 to S4 results in the liberation of one proton. Such a scenario has the advantage that charge build-up is avoided - one proton and one electron are released in each change of S state.

    Cl- is essential for the oxidation of water and only one Cl- is required per reaction centre. Chloride is apparently only required for the S2 → S3 and S3 → S0 transitions. Br can replace Cl- but F- inhibits the reaction. The evidence is that the Cl- is bound to either Mn or ca is tenuous however, we do not know where it is bound.

    There is evidence that there are two different water binding sites in the S3 state, and some evidence that suggests that water is not bound to Mn in the S2 state. Most of the current evidence suggests that intermediates such as peroxide are not formed during the S-state progression, O2 is formed from water or manganese ligands in a single four-electron step. The oxidation states of the Mn atoms are also uncertain, but a consensus has emerged that the S2 state contains Mn in the III,IV,IV and IV oxidation states. We cannot be sure about the oxidation states in the higher S states until we know whether Mn is oxidised in the S2 → S3 transition.

    The mechanism of water oxidation
    There have been many proposals made for the mechanism of water oxidation (chemists love to speculate about mechanisms). Many of the earlier proposals have been shown to be inconsistent with experimental results acquired subsequent to the postulate. In 1995 and 1996 two new mechanisms were suggested based on structure VIII.

    The first of these proposals (Scheme 2) [1] has one Mn2O2 dimer as a redox-inert unit, retaining a formal oxidation state of IV,IV throughout the reaction. The other pair of manganese atoms is in the III,II oxidation state in S0 and in the III,IV oxidation state in S2. It is proposed that going from S2 to S3 removes an electron from one of the oxygen atoms bridging the Mn atoms forming an oxyl radical. A second oxyl radical is formed in going to the S4 state. Formation of a bridging peroxo group (O22-) follows immediately, and this species then spontaneously decomposes in the presence of water to form S0, O2 and two protons. This mechanism predicts the loss of two protons in each of the oxidation steps, S0 → S1 and S4 → S0.

    Some of the objections that can be made to this mechanism are that proton release does not occur during the S-state transitions in the suggested manner, and that it does not explain the need for four Mn ions to oxidise water. Clearly if the sequence of proton release can be established with certainty then this will provide a good test for this mechanism.

    The second mechanism, Scheme 3, was proposed by Babcock [2]. This mechanism assumes that the Ca2+ ion shares at least one ligand with a Mn ion, or is at least in close proximity to the manganese cluster. Cl- is also assumed to act as a ligand to both Mn and Ca and YZ is assumed to be an integral part of the mechanism. The absorption of a photon generates YZ and results in the abstraction of both H+ and e- from the manganese cluster, and thus each S state advance is charge neutral. Generation of a terminal oxo ligand in the S2 state results in a ligand shuttling process, in which Mn and Ca exchange H2O and Cl- ligands. It is supposed that Cl- is bound to the Mn in the lower S states in order to prevent the formation of two Mn-bound OH- species in the S2 state, which could collapse to form H2O2. In the S4 state, it is supposed that the two oxo groups are close enough together to form the O2 molecule. The distance between the two water-binding Mn atoms in the model complex VIII is 5.5 A and so the two bound oxo groups in the S4state should be close enough to form the O2 molecule.

    Objections which can be made to this mechanism are that the involvement of either or both the Cl-and Ca2+ ions is speculative, and the involvement of Cl- before the S2 state is reached is not in accord with the evidence. Formation of a Mn complex with three oxo groups is unprecedented in the model chemistry. However, it could be argued that this is a plus for the proposal as none of the model complexes oxidise water! The apparent lack of change in the oxidation state of the manganese atoms in the S2 → S3 transition is also not currently accommodated by this scheme, although the mechanism could probably be modified to do so. 


    Both of the mechanisms described are probably incorrect. However they should provide stimuli to devise new experiments that will ultimately lead to better and eventually a correct hypothesis. We probably won't know the complete answer until somebody succeeds in producing suitable crystals of the PSII reaction centre and then manages to solve the structure using X-ray crystallography. If that should happen, it is not certain that we would then know how plants make oxygen - but we would be standing on firmer ground when we speculate about the mechanism.


    1. Yachandra, V. K., Sauer, K., and Klein, M. P., Chem. Rev.,1996,96,2927- 2950.
    2. Babcock, G. T. in Photosynthesis from Light to Biosphere Voll/ ed. Mathis P., Kluwer Academic Publishers, Amsterdam, 1995 p 209-215.

    The author would like to thank Ms. Teresa Gen for the production of the diagrams.

    About the author
    David was born and educated in Sydney. After completing a PhD in Inorganic Chemistry at Sydney University, working with Neville Gibson, he held a post-doctoral fellowship at Purdue University (Indiana USA) working with Dale Margerum. He returned to Australia to take up a Senior Tutorship at the University of Queensland and then moved to Wellington, New Zealand, where he is currently a Senior Lecturer in the School of Chemical and Physical Sciences. During his time at Wellington, he has been fortunate in being able to spend time conducting research at the University of Leeds, Ruhr Universitat Bochum, Germany, University of Texas at Austin, Oxford University and the University of California at San Diego. His research interests are related to transition metal coordination compounds. They include the preparation of macrocyclic ligands and their complexes, magnetic interactions in polynuclear complexes, bioinorganic chemistry, in particular manganese model compounds for PSII and other manganese containing enzymes, and the use of coordination complexes to destroy pollutants of various types.
    The 1999 Nyholm Youth Lecture: Closet Chemistry - ...


    No comments made yet. Be the first to submit a comment
    Already Registered? Login Here
    Sunday, 22 July 2018

    My Account

    Get Social