Determining the limitations and regulation of photosynthetic energy transduction in leaves


N. R. Baker. Fax: +44 1206 872592; e-mail:


The light-dependent production of ATP and reductants by the photosynthetic apparatus in vivo involves a series of electron and proton transfers. Consideration is given as to how electron fluxes through photosystem I (PSI), using absorption spectroscopy, and through photosystem II (PSII), using chlorophyll fluorescence analyses, can be estimated in vivo. Measurements of light-induced electrochromic shifts using absorption spectroscopy provide a means of analyzing the proton fluxes across the thylakoid membranes in vivo. Regulation of these electron and proton fluxes is required for the thylakoids to meet the fluctuating metabolic demands of the cell. Chloroplasts exhibit a wide and flexible range of mechanisms to regulate electron and proton fluxes that enable chloroplasts to match light use for ATP and reductant production with the prevailing metabolic requirements. Non-invasive probing of electron fluxes through PSI and PSII, and proton fluxes across the thylakoid membranes can provide insights into the operation of such regulatory processes in vivo.


cyclic electron flux around PSI


dark interval relaxation kinetics


electron transport rate


linear electron flux




NADP malate dehydrogenase


reaction centre chlorophyll of PSII


reaction centre chlorophyll of PSI


oxidized P700


non-oxidized P700


photosynthetically active radiation


transthylakoid proton motive force


photosynthetically active photon flux density




plastoquinol (reduced plastoquinone)


bound primary plastoquinone electron acceptor of PSII


bound secondary plastoquinone acceptor of PSII


free energy of ATP formation


pH component of pmf


electric field component of pmf


time constant for ECS decay after a brief dark interruption of light steady state


quantum efficiency of PSI electron transport


The major role of the photosynthetic apparatus of higher plant thylakoids is to transduce light energy into ATP and reductants (usually NADPH). Light is captured by an array of light-harvesting complexes, which absorb light and transfer excitation energy to the reaction centres of photosystem I (PSI) and photosystem II (PSII) to drive the primary photochemical reactions and create a separation of electrical charge. These light-driven charge separations at PSI and PSII effectively drive electron flux from water to terminal electron acceptors. The linear electron flux (LEF) from water through the PSII and PSI reaction centres is coupled to H+ release during water oxidation and the shuttling of H+ across the thylakoid membrane, which together establish a H+ electrochemical potential difference, or pmf, across the thylakoid membrane. A cyclic electron flux around PSI (CEF1) can also result in H+ transfer from the stroma to lumen, and contribute to the pmf. The pmf can be used to drive ATP synthesis by the transport of H+ through the ATP synthase back into the stroma. A schematic drawing outlining these processes is shown in Fig. 1.

Figure 1.

Schematic representation of the electron (orange arrows) and proton transfers (blue arrows), and associated processes that can occur as a result of light absorption by the thylakoid photosystems. Excitation of photosystem II (PSII) and photosystem I (PSI) oxidizes their reaction centres and drives LEF from water to NADPH (upper side of thylakoid). Electron flux from PSII and proton uptake from the stroma reduce PQ to PQH2. From PQH2, half of the electrons are transferred via the cytochrome b6f complex and plastocyanin (Pc) to PSI, which then transfers the electrons via ferredoxin (Fd), and a ferredoxin NADP oxidoreductase to NADP resulting in the generation of NADPH. The other half of the electrons from PQH2 is returned via the cytochrome b6f complex to PQ. Excitation of PSI can result in CEF around PSI via ferredoxin, PQ, cytochrome b6f complex and Pc (lower side of thylakoid). Reduction of PQ by ferredoxin is mediated by a plastoquinone reductase (PQR). Oxidation of water by PSII and PQH2 by the cytochrome b6f complex releases protons into the lumen creating a pmf across the membrane. Proton buffering will initially result in storage of the pmf as electric field (Δψ). However, the Δψ will be collapsed by counterion movements; with continued H+ influx into the lumen, the buffering capacity will be exceeded and the pH component (ΔpH) of the pmf will be formed. Movement of H+ from the lumen to the stroma via the ATP synthase results in ATP formation. The rate of electron flux from PQH2 through the cytochrome b6f complex I negatively regulated by an increasing pmf. The excitation density within PSII antennae can be regulated by energy-dependent quenching (qE; shown as loss of heat by a brown arrow). Creation of qE is associated with the pH-dependent activation of violaxanthin de-epoxidase (VDE), which reduces violaxanthin (V) to antheraxanthin (A) and zeaxanthin (Z), and the protonation of the Psbs protein. Consequently, excitation dissipation via qE is dependent on the accumulation of H+ in the lumen.

The composition, structure and functions of the photosynthetic apparatus of the thylakoid have been extensively studied and are now very well understood. In mature leaves, the primary role of the thylakoid photosynthetic apparatus is to provide ATP and reductants to meet the metabolic requirements for carbon assimilation and other energy-requiring processes. In the natural environment, rates of carbon assimilation by leaves can fluctuate markedly because of fluctuations in irradiance, or as the leaves experience various stresses, such as drought and high temperature, which restrict carbon assimilation. In situations where irradiance is limiting, efficiency needs to be maximized, placing a premium on efficient energy transfer to the reaction centres. When irradiance is not limiting, increases in LEF could produce increases in ATP and NADPH production that could thermodynamically or kinetically constrain LEF and result in over-excitation of the reaction centres, which could potentially lead to photodamage.Consequently, it is essential that the delivery of excitation energy to the reaction centres is regulated to prevent such damage, and this is achieved by the photosynthetic energy transduction systems having extremely flexible constraints on their efficiency. Besides using ATP and reductants for carbon assimilation, leaves use these photosynthetic products for a range of other metabolic processes, for example, nitrogen and sulphur metabolism, biosynthesis of macromolecules, which can differ in their stoichiometric requirements for ATP and reductants. Under suboptimal environmental conditions that reduce the ability to assimilate carbon, the proportion of electrons being consumed by such other electron sinks can increase markedly. Thus, besides having to regulate the rate of excitation of the reaction centres when CO2 assimilation is restricted, leaves also have to regulate the ratio of ATP to NADPH being produced by the thylakoid photosynthetic apparatus.

The regulation of the rate of excitation of the reaction centres and the ratio of ATP/NADPH production involves many factors associated with LEF and H+ transport in the thylakoid. In this review, we examine how information about PSI, PSII and H+ transport can be obtained and used to provide insights into the regulation of light energy transduction by thylakoids in vivo. Techniques that provide such information are described, and these allow analysis of the factors involved in regulation of light use and electron transport. Each of the techniques yields its own set of parameters. In a simple application, it may be that only one technique need be employed, for example, measurement of photodamage to PSII can be done using chlorophyll fluorescence alone. However, in many cases, particularly when examining the response of photosynthesis to the environment, a thorough analysis of limitations will require the concurrent application of several techniques. Correlation of the results from these different techniques allows identification of limiting processes and the regulation of fluxes. The review concludes with a summary of some of the current challenges in understanding the complexities of regulation.


Excitation of PSII results in oxidation of water and reduction of PQ. The excited reaction centre, P680*, can rapidly transfer an electron to a primary acceptor, pheophytin (Pheo), which will then transfer the electron to a bound plastoquinone (QA). QA- will then reduce a second plastoquinone (QB). P680+ is re-reduced by electron transfer from a tyrosine residue (Yz), which is re-reduced as a result of water oxidation. A second electron transfer from P680* to QB-, coupled with its protonation from the stroma, results in the formation of PQH2, which then dissociates from the PSII complex. Consequently, PSII acts as a water–plastoquinone oxidoreductase. PQH2 can now transfer the two electrons to the cytochrome b6f complex, and release the two protons into the thylakoid lumen, thus contributing to the formation of a H+ electrochemical potential difference across the thylakoid membrane.

Measurement of PSII electron transport

Modulated chlorophyll fluorescence measurements are widely used to estimate the quantum efficiency of PSII photochemistry and provide insights into the regulation of the rate of excitation of PSII reaction centres in leaves (Baker & Oxborough 2004). A leaf in continuous actinic light when monitored with a weak modulated fluorescence excitation beam has a fluorescence yield of F′, which rises to a maximal level Fm′ when the leaf is exposed to a saturating pulse of actinic light that maximally reduces QA. The difference between Fm′ and F′, designated Fq′, is the result of quenching of fluorescence by PSII photochemistry (Baker & Oxborough 2004). Genty, Briantais & Baker (1989) showed that Fq′/Fm′ was theoretically proportional to the quantum yield of PSII photochemistry, and this was confirmed empirically from direct measurements of oxygen evolution using mass spectrometry (Genty et al. 1992). Clearly, this fluorescence parameter could be used to provide a rapid and effective way of estimating the quantum efficiency of LEF through PSII in leaves under different conditions (hereafter referred to as the PSII operating efficiency); this parameter has previously been termed ΔF/Fm′ and ΦPSII in the literature. Assuming that a constant proportion of the reductants resulting from LEF is utilized for CO2 assimilation, then the PSII operating efficiency would be predicted to be directly proportional to the operating quantum efficiency of CO2 assimilation (Genty et al. 1989; Baker & Oxborough 2004). Empirically, this has been shown in leaves where photorespiration is absent or suppressed (Genty et al. 1989; Genty, Harbinson & Baker 1990a; Harbinson, Genty & Baker 1990; Krall & Edwards 1990, 1991; Cornic & Ghashghaie 1991; Krall et al. 1991; Edwards & Baker 1993; Siebke et al. 1997), and demonstrates the robustness of the use of fluorescence to investigate LEF through PSII. A list of the major fluorescence parameters used in analysis of PSII photochemistry is given in Table 1, together with their definitions and a brief explanation of their physiological relevance.

Table 1.  Definitions of the major modulated chlorophyll fluorescence parameters used in studies of PSII photochemical performance
Fluorescence parameterDefinitionPhysiological relevance
F, FFluorescence emission from dark- or light-adapted leaf, respectively.Provides little information on photosynthetic performance as they are influenced by many factors
Fo, FoMinimal fluorescence from dark- and light-adapted leaf, respectivelyLevel of fluorescence when PSII primary quinone electron acceptors are maximally oxidized (PSII centres ‘open’)
Fm, FmMaximal fluorescence from dark- and light-adapted leaf, respectivelyLevel of fluorescence when QA is maximally reduced (PSII centres ‘closed’)
Fv, FvVariable fluorescence from dark- and light-adapted leaves, respectivelyDemonstrates ability of PSII to perform primary photochemistry
FqDifference in fluorescence between Fm′ and FPhotochemical quenching of fluorescence caused by ‘open’ PSII centres
Fv/FmMaximum quantum efficiency of PSII photochemistryMaximum efficiency at which light absorbed by PSII is converted to chemical energy (QA reduction)
Fq′/FmPSII operating efficiencyEstimates the efficiency at which light absorbed by PSII is used for photochemistry (QA reduction); at a given light intensity, it provides an estimate of the quantum efficiency of linear electron transport through PSII; has previously been termed ΔF/Fm′ and ΦPSII in the literature
Fv′/FmPSII maximum efficiencyProvides an estimate of the maximum efficiency of PSII photochemistry at a given light intensity, which is the PSII operating efficiency if all the PSII centres were open (QA oxidized)
Fq′/FvPSII efficiency factorNon-linearly related to the proportion of PSII centres that are in the open state (with QA oxidized); relates the PSII maximum efficiency to the PSII operating efficiency; mathematically identical to the coefficient of photochemical quenching, qp
qLFraction of PSII centres which are in the open stateParameter estimating the fraction of PSII centres in open state (with QA oxidized) based on a lake model for the PSII photosynthetic apparatus; equates to (Fq′/Fv′)(Fo′/F′)
NPQNon-photochemical quenchingEstimates the non-photochemical quenching from Fm to Fm′; monitors the apparent rate constant for non-radiative decay (heat loss) from PSII and its antennae
qEEnergy-dependent quenchingAssociated with a light-induced development of ΔpH across the thylakoid membrane. Regulates the rate of excitation of PSII reaction centres
qIPhotoinhibitory quenchingAssociated with a photoinhibition of PSII photochemistry
qTQuenching associated with a state transitionAssociated with phosphorylation of LHCII

Because the PSII operating efficiency is directly related to the rate of LEF, it is possible in theory to estimate the rate of non-cyclic electron transport through PSII (ETR) using the following equation:


where I is the incident PPFD on the leaf, Aleaf is the spectral absorbance of the leaf and fractionPSII is the fraction of incident photons that are absorbed by PSII. It is frequently assumed that leaves absorb 84% of incident photons, and that 50% of these photons are absorbed by PSII, and consequently Eqn 1 is often modified to:


This equation is routinely used by commercial modulated fluorometers to estimate ETR. Although the assumption that leaves absorb 84% of incident photons is reasonably accurate for most mature green leaves, it is not always the case, and large deviations from this value can occur (Hodáňová 1985; Ehleringer 1991; Jones 1992). Similarly, the assumption that 50% of the light absorbed by the leaf is absorbed by PSII may be reasonable in many cases, but there will be many situations where this is not the case. The absorptivity of a leaf can be accurately determined using an integrating sphere with an appropriate light source and a spectroradiometer, although the effect on efficiency of non-photosynthetic blue-light absorbing pigments present in many leaves cannot be corrected for by means of a simple absorption measurement. However, it is very difficult to determine accurately the fraction of absorbed photons that are reaching PSII. Consequently, caution must be exercised when attempting to estimate ETR from measurements of PSII operating efficiency. Another important potential source of error when using Fq′/Fm′ to estimate ETR is the contribution of PSI fluorescence emission to measured fluorescence parameters (Baker & Oxborough 2004). As PPFD is increased, and Fq′/Fm′ decreases the relative contribution of PSI, fluorescence increases and will result in decreases in Fq′/Fm′ that are not associated with changes in PSII photochemistry and consequently ETR will be underestimated. Such errors can be minimized by measuring fluorescence associated with the 683 nm emission peak of PSII where the relative contribution of PSI fluorescence is minimized (Genty et al. 1990b; Pfündel 1998; Lawson et al. 2002; Itoh & Sugiura 2004). The use of short measuring wavelengths, however, is associated with greater reabsorption of the fluorescence with the result that the measurement is now biased towards the upper surface of the leaf (Lawson et al. 2002). With longer measuring wavelengths, there is still a bias in measurements, but this is not caused by reabsorption but to the greater penetration of the excitation light into the leaf. Commercial fluorometers often measure fluorescence primarily above 700 nm, and consequently have a high probability of PSI fluorescence and cells deeper within the leaf making a significant contribution to the fluorescence detected than when shorter measuring wavelengths are used.

Factors determining PSII operating efficiency

The PSII operating efficiency is the product of two important fluorescence parameters, Fv′/Fm′ and Fq′/Fv′, where Fv′ is the variable fluorescence yield of a light-adapted leaf (Genty et al. 1989). Fv′/Fm′ estimates the maximum quantum efficiency of PSII photochemistry in the illuminated leaf when QA is maximally oxidized and can be used to assess the contributions of non-photochemical quenching (NPQ) to changes in the PSII operating efficiency (Baker & Oxborough 2004). Fq′/Fv′ estimates the fraction of the maximum PSII operating efficiency that is realized in the leaf under the environmental conditions during the measurement; this relates to the proportion of PSII reaction centres with QA oxidized, i.e. the fraction of PSII centres that are ‘open’, and is mathematically identical to the frequently used coefficient of photochemical quenching, qp. Note, however, that because of interconnectivity between PSII photosynthetic units, the quantitative relationship between QA redox state and Fq′/Fv′ (and qp) is non-linear, so changes in Fq′/Fv′ can only be used to imply qualitative changes in QA redox state. By measuring both Fv′/Fm′ and Fq′/Fv′ in situations where changes are occurring in the PSII operating efficiency, it is possible to evaluate whether changes in LEF through PSII are attributable to changes in NPQ or the ability of an excited PSII reaction centre to perform photochemistry. A good example of this is during the induction of photosynthesis when a dark-adapted maize leaf is exposed to light (Fig. 2). Large changes in the PSII operating efficiencyare observed, which are clearly primarily associated with changes in Fq′/Fv′ as only small changes in Fv′/Fm′ are observed during the induction. Clearly, the fluctuations in PSII electron transport during induction of photosynthesis are almost entirely associated with changes in the ability to utilize the products of electron transport and not with changes in NPQ modifying the rate of excitation of PSII reaction centres.

Figure 2.

Changes in the photosystem II (PSII) operating efficiency (Fq′/Fm′, ●), maximum PSII quantum efficiency (Fv′/Fm′, ▴) and the fraction of the maximum PSII efficiency that is realized in the light (Fq′/Fv′, ○) during induction of photosynthesis in a dark-adapted mature maize leaf on exposure to a PPFD of 815 µmol m−2 s−1 over 60 min. Drawn from data of Oxborough & Baker (1997).

When light intensity is increased, the steady-state PSII operating efficiency decreases and is accompanied by decreases in both Fq′/Fv′ and Fv′/Fm′ (Fig. 3). The increases in NPQ, indicated by Fv′/Fm′, are saturated at much lower light intensity than the decreases in Fq′/Fv′, indicating that changes in NPQ are not contributing significantly to the large decreases in the PSII operating efficiency at higher light levels. Consequently, decreases in the PSII operating efficiency once NPQ is saturated are being determined by the ability of PSII reaction centres to transfer electrons to secondary electron acceptors. Confusingly, NPQ, in addition to being an abbreviation for a physiological process, is also used as a physiological parameter (we will use the italicized abbreviation NPQ to indicate the physiological parameter), which equates to (Fm/Fm′) − 1 (Bilger & Björkman 1990). This term is frequently used to assess levels of NPQ, and changes in NPQ are non-linearly related to and rise to higher values than Fv′/Fm′ for a given change in NPQ (Fig. 3). However, NPQ does not allow direct evaluation of the proportion of the change in the PSII operating efficiency that is attributable to changes in NPQ, whereas Fv′/Fm′ does. Furthermore, it should be noted that NPQ compares the NPQ from dark-adapted leaf at Fm to Fm′ in the light-adapted leaf. Consequently, care must be taken to only compare NPQ values from samples that have similar quenching characteristics in the dark-adapted state.

Figure 3.

Changes in the photosystem II (PSII) operating efficiency (Fq′/Fm′, ●), maximum PSII quantum efficiency (Fv′/Fm′, □), the fraction of the maximum PSII efficiency that is realized in the light (Fq′/Fv′, ▴), the fraction of PSII reaction centres that are ‘open’ (qL, ♦) and non-photochemical quenching (NPQ, ▵) as a function of PPFD in a tobacco leaf that was maintained in an atmosphere containing 100 µmol mol−1 CO2 and 2% O2 to reduce CO2 assimilation and eliminate photorespiration, respectively. Drawn from data of Kramer et al. (2004d).

A more detailed analysis of changes in the light-induced, down-regulatory quenching processes and other basal, non-photochemical losses in PSII that occur in dark-adapted leaves caused by non-radiative decays can be made (Hendrickson, Furbank & Chow 2004; Kramer et al. 2004d). This involves calculation of the quantum yields of NPQ (ΦNPQ) and basal non-radiative decays (ΦNO), and allows accurate estimation of the fraction of excitons that are lost in down-regulatory quenching and in basal, non-radiative decay processes. NPQ can comprise of three forms of quenching: energy-dependent quenching (qE), photoinhibitory quenching (qI) and state transition-related quenching (qT), with qE and qI being the major contributors in higher plants. The majority of qE is considered to be associated with quenching in the PSII antennae resulting from the development of a large ΔpH across the thylakoid membrane, which activates the violaxanthin de-epoxidase that converts violaxanthin to zeaxanthin (Demmig-Adams & Adams 1996; Yamamoto, Bugos & Hieber 1999), and also results in the protonation of carboxylic acid residues of the PsbS, a protein associated with the PSII antenna (Li et al. 2000, 2004). Protonation of PsbS and binding of zeaxanthin result in conformational changes in the PSII antennae that are associated with increases in the quantum yield of thermal dissipation of excitation energy (Krause & Jahns 2004; Pascal et al. 2005). PsbS protonation and violaxanthin de-epoxidase activation depend on the intrathylakoid (also known as the lumenal) pH, so factors that influence the steady-state value of this parameter are important in determining the extent of qE (see section on Proton Fluxes). Thermal dissipation plays an important regulatory role in regulating the rate of excitation of PSII reaction centres. Attempts have been to resolve qE from qI and qT on the basis of differences in their relaxation kinetics in the dark (Horton & Hague 1988; Quick & Stitt 1989; Walters & Horton 1991); however, caution should be exercised when attempting to do this as the rate of relaxation of these quenching components can be changed by the saturating light pulses under certain conditions and differ in response to long-term environmental stresses.

The relationship of Fq′/Fv′ with the redox state of QA is complex and depends on whether excitation energy transfer can occur between PSII reaction centres and the amount of NPQ at PSII (Baker & Oxborough 2004; Kramer et al. 2004d). Fq′/Fv′ is linearly related to the redox state of QA only if each PSII reaction centre has its own independent antenna system which cannot transfer excitation to the antennae of other reaction centres (Baker et al. 2001; Kramer et al. 2004d). It is clear that in leaves, this is not the case; excitation energy in PSII antennae can be competed for by a number of reaction centres (Lavergne & Trissl 1995), and Fq′/Fv′ is not linearly related to the fraction of PSII centres that are open (Baker & Oxborough 2004; Kramer et al. 2004d). Assuming a lake model for the photosynthetic apparatus where PSII photosynthetic units are connected and excitation energy can be competed for by a number of reaction centres, Kramer et al. (2004d) have demonstrated that the parameter qL, which is given by (Fq′/Fv′)(Fo′/F′), is linearly related to the redox state of QA. Consequently, qL can be used to monitor changes in the fraction of PSII centres that are open. For leaves exposed to a wide range of PPFDs, qL was consistently found to be lower than Fq′/Fv′, and at high PPFDs, Fq′/Fv′ was almost two times greater than qL (Kramer et al. 2004d), although the pattern of change in both parameters as a function of increasing PPFD was similar (Fig. 3).


PSI is associated with the formation of the reductants required for much of the metabolism that occurs in the stroma. PSI photochemistry is initiated by excitation energy transfer from antennae pigments to the reaction centre chlorophyll, P700. The intrinsically unstable, excited state of this pigment, P700*, is a powerful reductant that can reduce the primary acceptor of PSI, A0 (a pair of chlorophyll molecules symmetrically arranged within the PSI heterodimer) to form the strongest, stable, biologically generated reductant so far identified. The other product of this reaction is P700+, an oxidant. From A0, the electron is transferred to one of a pair of A1 (phylloquinone), Fx, FA and FB (iron–sulphur centres) before it leaves the PSI complex to reduce ferredoxin. Ferredoxin is a mobile, water-soluble protein containing an iron–sulphur centre that distributes electrons received from PSI to a diverse range of electron acceptors in the chloroplast stroma, of which quantitatively the most important is normally NADP. Meanwhile, P700+ is reduced by electron transfer from plastocyanin, which in turn receives electrons from the cytochrome b6f complex. There are two sources for the electrons that reduce the cytochrome b6f complex. They may be derived from PSII giving rise to the linear electron transport pathway, or they may be transferred from ferredoxin giving rise to a cyclic electron pathway. The relative contribution of these pathways is variable depending on the type of photosynthetic organism and the regulatory state of the electron transport chain. In C3, photosynthetic organisms whose main photosynthetic activity is CO2 assimilation and/or photorespiratory O2 reduction, PSII is the predominant source of electrons that ultimately reduce P700. A high yield of PSI-driven CEF1 in such plants would be inconsistent with their high yield of CO2 fixation under non-photorespiratory conditions (Genty & Harbinson 1996). Thus, in this type of system, linear electron transport is dominant and the electron transport activities of PSII and PSI are tightly coupled. In the bundle sheath cells of C4 plants, organisms with more flexible metabolism (e.g. green algae), and possibly, stressed C3 leaves, CEF1 can be a significant, or even the dominant, form of photosynthetic energy capture (Harbinson & Foyer 1991; Finazzi et al. 2002; Romanowska et al. 2006).

Ultimately, any conditions that prevent the flow of electrons away from P700* or into P700+ will block the PSI electron flux. To sustain electron flux through PSI, the following requirements must be met: (1) there must be molecules of P700 which can be photochemically oxidized; (2) an electron transport chain that is capable of transferring the electron from P700 to ferredoxin; (3) an electron donor system receiving electrons via either the linear or cyclic pathways that can re-reduce P700+; and (4) metabolism (or a non-metabolic electron acceptor activity, such as O2 reduction) that will reoxidize reduced ferredoxin. A limitation of any of these requirements will decrease the light-use efficiency of PSI.

Measurement of PSI electron transport

Measurements of PSI electron transport are often focused on analyzing to what extent, and by which means, donor and acceptor side processes limit PSI electron transport, and the relative contributions of linear and cyclic fluxes to the regeneration of P700 from P700+. Unlike PSII fluorescence, the yield of fluorescence from PSI is largely considered to be unaffected by the state of the PSI reaction centre in vivo at room temperature, so fluorescence cannot be used to measure PSI electron transport in vivo (Lavorel & Etienne 1977; Itoh & Sugiura 2004). Instead, the operation of PSI in vivo is monitored by means of a light-induced absorbance change, usually in the range 800–850 nm (Harbinson & Woodward 1987; Schreiber, Klughammer & Neubauer 1988). In this spectral region, the oxidation of P700 to P700+ creates an increase in absorbance. Scattering of the measuring beam by the leaf tissue increases its path length (Rühle & Wild 1979), so the absorbance increase is greater than that expected from the extinction coefficient of the absorbance change and the concentration of P700 present in leaves. This makes it impossible to use an unadjusted absorbance change to quantify the total amount of P700 in the leaf or otherwise use the absorbance change as an absolute measure of P700 oxidation. To circumvent this limitation, the absorbance increase developed during irradiance is calibrated by comparing it to the absorbance change produced during a far-red irradiance (around 720 nm) which will oxidize most of the (typically around 90%) P700 in the leaf. The far-red irradiance may also be combined with a flash of broad-band irradiance to ensure complete oxidation of the P700 pool (Kingston-Smith, Harbinson & Foyer 1999).The quantum yield of a PSI complex is zero when its P700 is oxidized; under these conditions, the reaction centre quenches the excitation energy, converting it to heat. So, in the case where there is no limitation of P700 oxidation caused by a shortage of electron acceptors, the relative amount of the P700 pool that is non-oxidized (P7000) is a measure of the ΦPSI, and is calculated from:


It is important to note that this is strictly a relative quantum efficiency; it is not known with certainty what the quantum yield of PSI electron transport is in absolute terms when no P700 is oxidized, although it is generally expected to be in the order of 0.95 (Lavergne & Trissl 1995). Thus the error implied by taking the relative yield to be an absolute yield is in most cases not significant. If P700 oxidation in some PSI reaction centres is limited by a shortage of electron acceptors, the measurement and calculation of PSI electron transport are more complicated because it is necessary to account for the effects of donor and acceptor side limitation (Klughammer & Schreiber 1994; Holtgrefe et al. 2003). For a wild-type (WT) leaf photosynthesizing in air, with open or closed stomata, or in 2% O2 with CO2 concentrations above 100 ppm, it is very unlikely that any acceptor limitation will be present, except transiently (e.g. following a large increase in irradiance). Most steady-state measurements of PSI electron transport do not, therefore, need to account for acceptor side limitation, and data obtained under these conditions are simpler to interpret in terms of changes in the quantum yield for PSI electron transport.

In leaves in darkness and at irradiances where photosynthesis is completely light limited, the relative amount of P700 that is in the non-oxidized state is 100%, and the efficiency of PSI electron transport is calculated to be 1 (Fig. 4). This implies that under completely light-limited conditions, electron transport into PSI is sufficient to reduce all photochemically oxidized P700. Over most of the PAR spectrum, the excitation of PSII has been calculated to exceed that of PSI (Evans 1987); and in most leaves, the operating efficiency of PSII decreases sharply from the dark-adapted value of about 0.8 by about 0.05–0.10 at PPFDs below 100 µmol m−2 s−1 because of reduction of QA (see Fig. 3) (Genty et al. 1989), consistent with PSII electron transport being limited by a lack of electron acceptors (P700+). This decrease of PSII efficiency under light-limiting conditions implies a loss of overall light-use efficiency. There are situations where ΦPSI will decrease sharply at low irradiances; for example, acute photodamage to PSII can reduce the activity of PSII to a point where electron transport from PSII is insufficient to reduce photochemically generated P700+ (Genty et al. 1990a), and some mutations that diminish the amount of chlorophyll b also produce the same effect by reducing the rate of excitation of PSII reaction centres (unpublished observations). Irradiance with wavelengths that preferentially excite PSI (far red: >700 nm), or following treatment with herbicides that affect PSII electron transport, will likewise produce an increase in the steady-state pool of P700+ at low irradiances and thus decrease ΦPSI (Harbinson & Woodward 1987). This implies that a comparison of PSI and PSII efficiencies under strictly light-limited irradiances can provide information about the balance of excitation of the two photosystems or of damage to the photosystems.

Figure 4.

A typical relationship between the quantum efficiency for electron transport by PSI (ΦPSI) and irradiance. The data were obtained from the leaf of a tropical epiphyte Juanulloa aurantiaca photosynthesizing in air and subjected to a regime of increasing irradiance. ΦPSI was calculated using Eqn 3, so the efficiency is relative and uncorrected for the actual maximum efficiency of PSI, although this is expected to be 0.95 or higher (see text).

In the absence of an acceptor side limitation, increasing irradiance results in a sigmoidal decrease in the quantum yield of PSI (Fig. 4). The sensitivity of ΦPSI to light intensity varies between leaves and also depends on the environmental and physiological conditions of the leaf at the time of measurement, for example, temperature, drought stress, leaf age, source/sink balance, CO2 and O2 concentration (Harbinson, Genty & Foyer 1990; Peterson 1991; Harbinson 1994; Laisk & Oja 1994). Decreases in ΦPSI at constant irradiance will also be produced by factors that decrease photosynthesis, such as decreasing CO2 concentration, decreasing temperature and drought. Although electron transport may be limited by metabolic processes, under steady-state conditions this limitation does not usually act directly to limit PSI electron transport on its acceptor side. In response to limited metabolic activity, electron transport is considered to be limited largely at the cytochrome b6f complex (Laisk & Oja 1994; Genty & Harbinson 1996). It is, however, important to remember that the extent to which a decrease in the potential rate of electron transport through the cytochrome b6f complex will limit electron transport as a whole will depend on irradiance. At low irradiance, where photosynthesis is limited by light-capture, inhibition of the cytochrome b6f complex has very little effect on the rate of electron transport, whereas at saturating irradiance electron transport is much more sensitive to inhibition of the cytochrome b6f complex (Heber, Neimanis & Dietz 1988). In contrast, electron transport at low irradiance is more sensitive to inhibition at QB than it is at high irradiance (Heber et al. 1988). NPQ of PSII will diminish the rate of reduction of the QA pool, and will increase with increasing irradiance above the region of light limitation of photosynthesis. By analogy with the effect of inhibition at QB on electron transport, it is possible that NPQ could exert a weak limitation on electron transport in the range of irradiances between complete light limitation and complete light saturation, but this remains to be demonstrated. In the absence of acceptor side limitations, ΦPSI can be used to estimate the electron flux through PSI (ETRPSI, also often termed JPSI):


where I is the incident PPFD on the leaf, Aleaf is the spectral leaf absorptance and fractionPSI is the fraction of the absorbed irradiance that is trapped by PSI complexes. It is difficult to determine fractionPSI experimentally, and consequently it is often assumed to be 0.5. As is the case for PSII (see previous text), the assumption that 50% of the photons absorbed by the leaf are absorbed by PSI will frequently be incorrect.

Kinetics of P700+ reduction

Removal of irradiance from a leaf results in the reduction of P700+, and analysis of the kinetics of this reduction can be used to provide information on the regulation of ΦPSI. In the absence of regulation on the donor side, an increase in PSI electron transport, for example, produced by increasing irradiance, would be limited by the approach to redox equilibrium between electron donors and acceptors on the acceptor side of PSI. This would result in the increasing reduction of the electron acceptor pool of the stroma to the point that forward electron transport from P700 would become impossible. Electron transport processes in the reaction centre would then be dominated by back reactions (Rutherford & Heathcote 1985). However, in vivo, the stroma does not become extensively reduced except transiently or under extreme conditions, for example, at CO2 concentrations below 100 ppm when the O2 concentration is 2% (Takahama, Shimuzu-Takahama & Heber 1981; Dietz & Heber 1984; Harbinson et al. 1990; Foyer, Lelandais & Harbinson 1992). The primary limitation of PSI electron transport therefore largely resides on the donor side at the cytochrome b6f complex rather than the acceptor side, even when metabolic demand for reductant is low. The rate of electron transport from the PQH2 pool to the cytochrome b6f complex is subject to short- (Tikhonov, Khomutov & Ruuge 1984; Nishio & Whitmarsh 1993) and long- (Onoda, Hikosaka & Hirose 2005) term control; short-term control is effected by changes in intrathylakoid pH, whereas long-term control is caused by changes in the amount of the cytochrome b6f complex. It is relatively easy to measure the extent of the controlled donor side limitation of electron transport by measuring the reduction kinetics of P700+ after the irradiance is removed. When irradiance is removed from the leaf, the rate of P700 oxidation falls to zero and the rate constant for P700+ reduction can then be obtained from the pseudo-first-order decay of the ΔA820 absorbance change (Fig. 5). This decay, which has a half-time of 3–4 ms or greater, reflects the rate-limiting supply of reductant passing from PQH2 via the cytochrome b6f complex and plastocyanin to P700+. The rate constant for this supply of reductant to P700+, ke, is a measure of the capacity for electron transport via this rate-limiting mechanism and can be treated like a conductance in leaf gas exchange models. A valuable feature of ke is that it is absolute, not relative like ΦPSI. This allows comparisons to be made between leaves and for the basis of changes in the relationship between ΦPSI and PPFD relationship to be analysed in terms of changes in ke (Riethmuller-Haage et al. 2006). Attempts to measure changes in the rate of electron transport into P700+ by measuring the initial slope of the millisecond decay component (Johnson 2005) were in error because they ignored the sub-millisecond kinetics caused by transfer from the reduced plastocyanin and cytochrome b6f pools which are not resolvable with instruments with a measuring beam-modulation frequency of 100 kHz (Sacksteder & Kramer 2000; Kramer et al. 2004a). The error will be greatest at lower PPFDs where the size of the reduced plastocyanin and cytochrome b6f pools will be greatest (Fig. 4) (Kirchhoff et al. 2004). A recent extension of these measurements of kinetics is the repeated application of light–dark intervals to leaves in a state of change. This general approach has been termed DIRK (Sacksteder & Kramer 2000). The analysis of the transients recorded during the DIRK procedure allows the changes in kinetics underlying the response to be resolved.

Figure 5.

Typical decay kinetics of the absorption change at 820 nm (ΔA820) from a leaf produced by removing the irradiance. In the absence of irradiance, the steady-state of P700 oxidation and P700+ reduction is unbalanced by the absence of oxidation, and the pool of P700+ decays to zero following kinetics determined by the rate constant for electron transport from PQH2 and the cytochrome b6f complex. In addition to the millisecond time-scale kinetics shown in this figure, more rapid (sub-millisecond) kinetics of P700+ reduction will occur because of electron transfer from the fraction of pools of plastocyanin and cytochrome f that were already reduced at the point of cessation of irradiance. These kinetics, which increasingly dominate as the irradiance is decreased, are unresolved in this measurement.

Problems with measurement of P700

Two problems arise with the measurement of P700 oxidation state using light-induced absorbance changes. The first results from the overlap of absorbance changes caused by plastocyanin with those of P700 in the 800–850 nm spectral region. The second is the possible loss of PSI efficiency because of a shortage of electron acceptors; this loss of efficiency will not be detected by techniques that use the relative amount of P700+ to quantify ΦPSI as shown in Eqn 3.

The overlap between absorbance changes of plastocyanin and P700 is strong; and at around 820 nm, about 30% of the total absorbance change would be expected to derive from plastocyanin, with the proportion varying with wavelength (Klughammer & Schreiber 1991; Kirchhoff et al. 2004). This wavelength dependency has been exploited in deconvolution procedures to separate the contributions from plastocyanin and P700 to absorbance changes in vivo and in vitro (Kirchhoff et al. 2004). Under conditions where P700 and plastocyanin reach equilibrium, the absorbance changes reflect the expected electrochemical equilibrium between P700 and plastocyanin. Upon switching off the actinic light, first P700 is reduced, followed by plastocyanin and cytochrome f (Klughammer & Schreiber 1991; Sacksteder & Kramer 2000; Kirchhoff et al. 2004). If equilibrium is achieved on the time-scale of the normal turnover of the cytochrome b6f complex, and the equilibrium constant for sharing electrons is constant, a simple model can be used to allow measurements of the absorbance change around 820 nm to yield accurate information about the electron flux through the cytochrome b6f complex and P700 (Sacksteder & Kramer 2000). However, in many cases the apparent equilibrium constant changes suggesting partial disequilibrium among the electron carriers (Sacksteder & Kramer 2000; Kirchhoff et al. 2004). In this case, it is necessary to consider the kinetics of reduction of all of the carriers. It is clear in many cases that the estimates of ΦPSI based on absorbance changes around 820 nm correlate well with other estimates of leaf photosynthetic efficiency (Harbinson, Genty & Baker 1989; Genty & Harbinson 1996). This contradiction can be resolved in two ways. Firstly, in systems with rapid electron transport, there appears to be a restriction in the equilibration between plastocyanin and P700, and the apparent equilibrium constant is reduced from a value in the range 64–312 expected from the redox potentials to one in range of 12–4 (Kirchhoff et al. 2004), dependent on the rate constant for P700+ reduction; the lower value is reached with rate constants of over 100 s−1, which would be normal for plants with a high rate of CO2 fixation, such as crop plants. Consequently, in vivo absorbance changes caused by P700 and plastocyanin vary in parallel. Secondly, leaves and other photosynthetic systems are usually optically dense, for example, the average absorption of a leaf is around 84% which results in large gradients of irradiance through the system. Along such gradients, there will be a continuum of photochemically generated couples of P700+/P700 and plastocyanin+/plastocyanin. Even if these couples are at equilibrium, the effect of the irradiance gradient is such that when the apparent equilibrium between P700 and plastocyanin is calculated from measurements that integrate over all these couples, it tends to unity as the absorbance approaches infinity (Harbinson & van Vliet 1994). This would also result in parallel changes in absorbance caused by plastocyanin and P700 oxidation.

To verify that P700 oxidation is possible and not limited by a shortage of electron acceptors, it is necessary to examine the oxidizability of the P700 pool using a saturating light-pulse technique similar to that used to measure PSII efficiency (Klughammer & Schreiber 1994). Results from this technique verify that under most conditions, there is no shortage of PSI acceptors. Only during photosynthetic induction (Harbinson & Hedley 1993; Klughammer & Schreiber 1994), low-carbon dioxide concentrations under non-respiratory conditions (Genty & Harbinson 1996), or when the pool of PSI acceptors has been diminished (Holtgrefe et al. 2003) does the pool of acceptors appear to limit P700 oxidation. Under these conditions, the measurement of ΦPSI needs to take account of the decrease in efficiency caused not only by P700+ but also to those PSI reaction centres where photochemistry is impossible because of a shortage of acceptors. This can be done using the saturating flash technique to determine the proportion of PSI that is non-oxidizable and combining this with the conventional estimate of ΦPSI based on the proportion of P700+. A possible source of error with the saturating pulse technique is that the multiple turnovers of PSI induced by the flash could close some open reaction centres by over-reducing their acceptor pools. There is, therefore, the risk of overestimating the degree of reaction centre closure using this technique, especially at low irradiances where the degree of reduction of high-potential PSI donors (plastocyanin and cytochrome f) will be high and the metabolic activity of the stroma low.


The proton circuit of photosynthesis plays a central role in energy transduction by the thylakoids, but is often given less attention than the associated electron transfer reactions. The light-driven fluxes of protons act not only to store energy for the synthesis of ATP, but also as a key regulatory component: activating the down-regulation of PSII-associated antenna and governing electron transfer by controlling the oxidation of PQH2 at the cytochrome b6f complex.

Generation of pmf by electron transfer-coupled reactions

The light reactions drive the energy-requiring generation of an electrochemical potential difference, or pmf, across the thylakoid membrane. To accomplish this, proton translocation is coupled to (i.e. powered by) electron transfer reactions at four key points, making up a ‘Mitchellian’ chemiosmotic energy storage system. As shown in Fig. 1, protons are released at two sites, the oxygen-evolving complex (OEC) of PSII and the plastoquinol-oxidizing (Qo) site of the cytochrome b6f complex, and are taken up at the plastoquinone reductase site of PSII (QB site) and the plastoquinone reductase site of the cytochrome b6f complex (usually termed the Qi site, but sometimes the Qn site).

In the OEC, four electrons are extracted from a pair of water molecules, resulting in the generation of one molecule of O2 and the release of four protons into the lumenal space. The electrons extracted from water are transferred across the membrane through a chain of redox carriers within the PSII complex. The accumulation of two electrons (i.e. after two photochemical excitations of P680) on a PQ bound at the QB site of PSII results in the uptake of two protons from the stromal side of the membrane, followed by the release of a neutral plastoquinol (PQH2) into the thylakoid membrane. The PQH2 is free to diffuse around, with some restrictions (Kirchhoff et al. 2004), until it is oxidized at the Qo site via turnover of the cytochrome b6f complex via the Q-cycle.

The Q-cycle is catalysed by the cytochrome b6f complex. The presentation of two high-resolution structures (Kurisu et al. 2003; Stroebel et al. 2003) indicated that the overall structure of the cytochrome b6f complex was similar to that of the related mitochondrial and bacterial cytochrome bc1 complexes, strongly supporting the operation of very similar mechanisms in both systems. Most importantly, the structure is broadly consistent with the Q-cycle mechanism previously proposed by a number of laboratories (Rich 2004; Cape, Bowman & Kramer 2006). The key step in the Q-cycle is the ‘bifurcated’ oxidation of PQH2 into two distinct chains of electron carriers. In most Q-cycle models, one electron from the PQH2 bound to the Qo site is transferred to the ‘high-potential’ chain consisting of the Rieske FeS cluster and cytochrome f, followed in chloroplasts by the mobile carrier plastocyanin (Cape et al. 2006). This process leaves a reactive semiquinone radical in the Qo site, which reduces the ‘low potential’ chain comprising of two cytochrome b hemes and the newly discovered c-type heme, heme ci (Kurisu et al. 2003; Stroebel et al. 2003). When two electrons have accumulated in the low potential chain, PQ is reduced to PQH2 at the Qi site of the cytochrome b6f complex with uptake of protons from the stroma. The Q-cycle appears to operate continuously in vivo, with very low rates of side reactions (Rich 1988, 2004; Kramer & Crofts 1993; Sacksteder et al. 2000). Consequently, it has been suggested that the overall proton pumping stoichiometry (H+/e- ratio) for LEF is 3, i.e. one proton released into the lumen at the level of water oxidation and two released during PQH2 oxidation at the Qo site of the cytochrome b6f complex (Allen 2003; Kramer et al. 2004a), rather than the value of 2 expected in the absence of a Q-cycle.

The role of the pmf in ATP synthesis

The mechanism of ATP synthesis driven by proton transfer through the ATP synthases has been extensively reviewed (Junge 1999; Stock, Leslie & Walker 1999; Seelert et al. 2000; Stock et al. 2000; Herbert 2002). High-resolution structural information and a likely molecular mechanism have allowed precise (but putative because they are based on a presumed mechanism) estimates of the H+/ATP ratio for steady-state ATP synthesis (Junge 1999; Stock et al. 2000; Ort & Baker 2002; Allen 2003). A full rotation of the γ-subunit within the (C)F1α/β trimer of the coupling factor should form three molecules ATP from ADP and phosphate. The transfer of a single proton is thought to rotate the α subunit assemblage by a single c-subunit of the (C)F0 ring. It follows that the number of c-subunits in the ring determines the H+/ATP ratio. This ratio varies depending upon species, but was found by atomic force microscopy to be 14 c-subunits/ring in chloroplasts (Seelert et al. 2000). Overall, this implies an H+/ATP ratio of 4.66, and taken together with the H+/e- stoichiometry of 3, indicates an ATP/NADPH ratio of ca. 1.3 for LEF (Allen 2003; Kramer et al. 2004a). Consequently, this has generated a renewed interest in mechanisms that can alter this ratio.

The role of the pmf in regulating light processing and electron transfer

As discussed in the section on PSII, chloroplasts can down-regulate the rate of excitation of P680 by developing an energy-dependent quenching, qE, in the light-harvesting antennae. Initiation of qE requires conversion of violaxanthin to zeaxanthin by violaxanthin de-epoxidase (Demmig-Adams & Adams 1996, which is located on the lumenal face of the thylakoid membrane (Yamamoto et al. 1999), and protonation of the antenna PsbS protein (Li et al. 2000, 2004); both are activated by acidification of the thylakoid lumen, i.e. the ΔpH component of the pmf. Consequently, pmf not only drives the synthesis of ATP, but also is a key feedback signal for the processing of excitation energy in the PSII antennae.

It has also been proposed that LEF is regulated by lumen pH at the level of PQH2 oxidation at the cytochrome b6f complex (see Fig. 1) (Haehnel 1984; Hope 1992; Kramer et al. 2004a). Regulation of electron flow could act to decrease superoxide generation and photodamage at the reducing side of PSI (Tjus et al. 1998; Asada 1999) or to prevent over-acidification of the lumen, i.e. acting as a pH governor.

Electrochromic shift (ECS) measurements for analysis of pmf: methods for probing the proton circuit of photosynthesis in vivo

A series of non-invasive spectroscopic tools have been introduced to probe the proton transfer circuit of photosynthesis in vivo. These techniques take advantage of the electrochromic shift (ECS, sometimes called ΔA520 or ΔA518 because the absorption measurements are made at 518 or 520 nm) of certain pigment (mostly carotenoid) species that occur naturally in thylakoid membranes. The ECS gives a linear indication of changes in the transthylakoid electric field, Δψ (Witt 1975, 1979). The ECS is also a useful tool for probing proton fluxes because it responds to net movements of charges (including that on the proton) across the transthylakoid membrane.

These techniques also make use of perturbation of steady-state photosynthesis by short, dark intervals, which allow the photosynthetic apparatus to relax in ways that reveal information about the system in the steady state and is an example of DIRK spectroscopy (Sacksteder & Kramer 2000).

The basis of DIRK spectroscopy is the principle of balanced steady-state fluxes. During steady-state condition, the flux of protons into the lumen is counterbalanced by efflux, predominantly through the ATP synthase. It is important to note that, even though influx and efflux are balanced, there will still be a light-driven pmf, because a certain level of pmf is required to drive ATP synthesis (Kramer, Avenson & Edwards 2004b). The actual level of pmf sustained under steady-state conditions will reflect the kinetics of influx and efflux (described as follows). Switching off the light from a steady state will inhibit the proton influx, but allow efflux to continue until the pmf comes into equilibrium with the free energy of ATP formation (ΔGATP) via reversible ATP hydrolysis at the ATP synthase. Because protons are charged, the net outward proton flux during the dark intervals will affect membrane potential, and thus can be followed by the ECS, yielding several interesting parameters as described as follows. The definitions and physiological relevance of the key ECS parameters are given in Table 2.

Table 2.  Definitions of the parameters determined from electrochromic shift (ECS) measurements used in studies of thylakoid pmf
ECS measurementDefinitionPhysiological relevance
  1. Refer to Fig. 6 for illustration of the kinetics of the ECS signals.

ECSinvThe difference in ECS signals from the quasi-stable (tens to hundreds of ms) and fully relaxed (∼2 min) after a light–dark transitionProportional to the light–dark difference in ΔpH component of transthylakoid pmf
ECSssThe difference in ECS signals from steady state and fully relaxed (∼2 min) after a light–dark transitionProportional to the light–dark difference in Δψ component of transthylakoid pmf
ECStThe total rapid (<1 s) change in ECS signal upon rapidly switching off actinic light from steady stateProportional to the total light–dark difference in transthylakoid pmf
inline imageThe inverse of the lifetime of the rapid decay of ECS, (τECS) upon rapidly switching off actinic light from steady stateProportional to the aggregate conductivity (or permeability) of the thylakoid membrane to protons, predominantly determined by ATP synthase activity
pmfLEFLEF (measured by fluorescence changes) divided by inline imageAn estimate of the pmf that would be generated by LEF alone
inline image(sometimes termed DIRKECS)The initial rate of decay of the ECS signal to a quasi-stable state, tens to hundreds of ms after rapidly switching off actinic light from steady stateProportional to proton flux through the photosynthetic apparatus and thus ATP synthesis; can be used to estimate changes in proton translocation by CEF1

Estimates of steady-state proton fluxes

Measuring the initial rate of decay of ECS upon a rapid light–dark transition from steady-state conditions yields a parameter termed inline image, sometimes termed DIRKECS (Fig. 6) (Sacksteder et al. 2000). Under appropriate conditions, inline imageshould reflect the rate of proton translocation by the entire electron transfer chain, because all protons exit the lumen through the same pathways, and usually predominantly through the ATP synthase. Care must be taken when the rate of decay of ECS changes quickly, because proton pumping by the cytochrome b6f turnover can then overlap in time with the Δψ decay via the ATP synthase (Sacksteder & Kramer 2000; Joliot & Joliot 2002; Avenson et al. 2005b), requiring the use of more elaborate deconvolution using simulations, such as those used in Sacksteder & Kramer (2000) and Cruz et al. (2001).

Figure 6.

Electrochromic shift (ECS) probes of the proton circuit of photosynthesis in vivo. ECS is measured as the absorption changes occurring around 520 nm (ΔA). A rapid light–dark transition from steady-state conditions leads to several phases of ECS decay that can be analysed to give information about the proton circuit. (a) The initial rate (taken over the ms range) yields inline image, which is proportional to the flux of protons through the photosynthetic apparatus. The inverse of the lifetime for ECS decay to a quasi-stable state on the tens to hundreds of ms time-scale, termed τECS, reflects the conductivity of the thylakoid membrane to protons, inline image, which is predominantly controlled by the activity of the ATP synthase. The amplitude of the ECS signal from steady state to the quasi-stable state is proportional to the total pmf. (b) Over a longer time-scale, the quasi-stable state decays, reflecting the movements of counter-ions, which relax the portion of the ECS signal caused by the ΔpH component of pmf, termed ECSinv, and leaving the portion because of Δψ, termed ECSss.

Comparisons of inline imagewith LEF (measured via fluorescence changes as described earlier) have been used to indicate the extent of proton pumping by CEF (Sacksteder et al. 2000; Avenson, Cruz & Kramer 2005a; Joliot & Joliot 2005). Linear plots of inline imageagainst LEF indicate that changes in proton translocation by CEF1 are either negligible or a constant fraction of LEF. Deviations in linearity, particularly during induction of photosynthesis, have been used to indicate activation and high rates of CEF1 (Avenson, Cruz & Kramer 2005a; Joliot & Joliot 2005).

Estimates of ATP synthase activity

The first-order decay time of the ECS decay towards a quasi-stable state, termed τECS, is also very informative. Because the proton efflux is pseudo-first order above the pmf required to activate the ATP synthase (Avenson et al. 2005b; Cruz et al. 2001), the inverse of the lifetime yields a relative measure of the aggregate conductivity of the ATP synthase to protons, termed inline image(Kanazawa & Kramer 2002). In effect, inline imagedefines the relationship between pmf and proton flux through the ATP synthase. When inline imageis large, a high efflux is seen for a relatively small pmf. When inline imagedecreases, a larger pmf is needed to attain the same proton efflux.

The inline imageparameter was used to demonstrate that physiological control of ATP synthase activity is a major mechanism for modulating lumen pH-dependent regulation of the light reactions (Kanazawa & Kramer 2002; Kramer et al. 2004c). When LEF becomes limited, for example, by lowering CO2 or O2, or by imposing water stress, inline imagedecreases by up to sixfold (i.e. the activity of the ATP synthase decreases). Under these conditions, a larger pmf is produced for a given proton flux (e.g. with the same LEF), resulting in a more sensitive qE response. It is not yet known what factors regulate the ATP synthase, although stromal inorganic phosphate (Sharkey 1990; Kanazawa & Kramer 2002) and phosphorylation of the ATP synthase (Kanekatsu et al. 1998; Bunney, van Walraven & de Boer 2001; del Riego et al. 2006) have been proposed.

Estimates of pmf generated by LEF

Dividing LEF by inline imagegives a new parameter, pmfLEF, which is an estimate of how much pmf is produced by LEF alone (Kanazawa & Kramer 2002). This estimate is based on a constant proton pumping stoichiometry (H+/e- ratio) for LEF, together with an ATP synthase that obeys Ohm's law above its activation threshold, which it appears to do under steady-state conditions (Kanazawa & Kramer 2002). Deviation from linearity in the relationship between pmf (estimated by ECSt) and pmfLEF could indicate contributions from CEF1 or changes in H+/e- (Avenson, Cruz & Kramer 2004; Avenson et al. 2005a,b).

Measurements of pmf components

Estimates of the light–dark differences in the components of the pmf can be obtained by analyzing the kinetics of relaxation of the ECS signal (Cruz et al. 2001). To understand these parameters, it is instructive to describe what happens to the components of pmf during a typical measurement, as illustrated in Fig. 6. The physical bases of the ECS parameters were derived in Cruz et al. (2001) and described in more accessible terms in (Kramer, Cruz & Kanazawa 2003). Upon switching off the actinic light, the outward proton flux will begin to dissipate any ‘positive’Δψ (in–out), leaving the ΔpH component of pmf, which relaxes much more slowly because of the high-proton buffering capacity of the chloroplast compartments. The ΔpH continues to drive further proton efflux, further altering the Δψ, even forming a ‘negative’Δψ (in–out). This will continue until a quasi-stable state is reached, where ΔpH is nearly precisely counterbalanced by the energetic sum of Δψ formed by movements of protons and ΔGATP, because ATP hydrolysis at the ATP synthase will move protons into the lumen. As long as ΔGATP remains relatively constant over the measurement, as it should considering the relatively large pool of ATP + ADP (Avenson et al. 2005b), the amplitude of ECS changes above the baseline, termed ECSss (for steady-state ECS), will reflect the light–dark difference in Δψ, the ‘inverted’ ECS signal, or ECSinv, will reflect the light–dark difference in ΔpH, while the sum, ECSt, should reflect the light–dark difference in total pmf. By comparing qE with ECSinv, estimates of ΔpH, it was shown that the short-term (hours) response of qE regulation to lumen pH remained relatively constant over a wide range of conditions (Kanazawa & Kramer 2002; Avenson et al. 2004, 2005a,b; Cruz et al. 2005). This suggests that over the short term, the responses of violaxanthin de-epoxidase and PsbS protonation are probably not modulated. These components are likely to play a role in regulation of antenna responses over longer developmental time-scales. It is known, for example, that thylakoid xanthophyll content may be subject to large changes, and that this has a substantial effect on the extent of qE responses (Demmig-Adams & Adams 1996). Likewise, genetic manipulation of PsbS expression has been shown to affect the amplitude of qE and presumably this may play a role in acclimation or adaptation to changing environmental conditions (Li et al. 2002, 2004; Niyogi et al. 2005).

There are, of course, caveats to the ECS-derived probes that need to be taken into consideration, as reviewed in detail in Avenson et al. (2005b) and Cruz et al. (2005). Firstly, it is important to note that most measurements using ECS-yield extrinsic parameters, i.e. they scale with the transmembrane electrical response of the ECS signal as well as the density of thylakoid membranes being probed by the spectrophotometer measuring beam. Thus, typically the ECS measurements are normalized to the single turnover flash-induced ECS responses to account for changes in chloroplast or pigment compositions. Provided that light absorption is similar, the parameter pmfLEF can be considered to be an intrinsic probe because it relies solely on the kinetics, and not the extent, of ECS decay, together with LEF, which is derived by intrinsic fluorescence changes as described earlier. Changes in the slope of the relationship between pmfLEF and inline imagecould indicate changes in the ECS response or pigment content, and thus pmfLEF is a very useful control, and this should be taken into consideration when comparing different species, leaves grown under different environmental conditions and mutant strains which can have substantial differences in antennae composition (Avenson et al. 2005a).


To examine the regulation of electron and proton fluxes, it is essential to combine the techniques discussed earlier. However, this is technically not a simple task. These biophysical techniques are frequently used concurrently with gas analysis, usually CO2 fixation, although O2 evolution and water vapour release are often also measured. Consequently, a leaf cuvette is required, which permits gas exchange measurements to be made in an accurately controlled environment. It must also allow the reliable application of the biophysical techniques to the same complete area of leaf whose gas exchange is being measured. All of the measurement systems must be able to function in the presence of the others, so interference between the systems needs to be either controlled or eliminated. The same area of leaf needs to be measured by each system; otherwise, it is not possible to reliably compare these measurements. This requirement is particularly evident from images of PSII operating efficiency of stressed leaves within which there are large variations in efficiency over small distances (e.g. Leipner, Oxborough & Baker 2001). When biochemical analyses are required to provide information of the metabolic status of the leaf tissue, then the photosynthesis measurement systems need to be combined with a freeze-clamp system to preserve the biochemical status of the tissue.

Even when a leaf area is being equally measured by several different systems, there remains an unavoidable heterogeneity problem. A typical leaf is optically dense and will absorb ca. 84% of the incident PPFD. As discussed in the section on Measurement of PSII electron transport, chlorophyll fluorescence measurements are prone to bias because of light penetration and absorption of fluorescence. As light intensity decreases as a function of depth in the leaf, there will be mismatch in the correlation between fluorescence and techniques, such as gas exchange, which have no such intrinsic bias. When quantitative analyses ofmismatched results are attempted, phenomena may be identified which are purely artefacts and not indicative of chloroplast physiology. Absorbance-based measuring systems have a similar problem; they give different results when used in the transmissive compared to the back-scatter modes of use. There is no solution for these problems other than careful experimentation and interpretation of the data. The light intensity gradient through the leaf also makes some calculations problematic, for example, the equilibrium constant between P700+ and its electron donors. If a procedure requires that the sample be in all respects homogeneous, then applying that procedure to a non-homogeneously illuminated leaf will produce a result that is best viewed as apparent not real. The user must always be confident that the procedure used can reliably integrate heterogeneity, or that the errors arising from heterogeneity can be managed within the analysis. Nonetheless, the use of such apparent values may be useful in identifying short-term changes in the leaf, as occur in response to the environment, but it must be emphasized that the changes in apparent values do not provide definitive information about processes.

Although difficult, the successful integration of the techniques is essential to provide a better understanding of the limitations and regulation of leaf photosynthesis.


The primary function of the thylakoid photosynthetic apparatus is to generate reductants and ATP. If the rate of production of these products exceeds the metabolic demand for them, then inevitably this will impose thermodynamic constraints on photosynthetic electron and proton transfer processes. Consequently, a persistent question is: how are the rates of synthesis of reductants and ATP coordinated with the fluctuating demands of metabolism?

An interesting conundrum associated with the operation of the C3 photosynthetic reduction cycle is that calculations have shown that insufficient ATP is generated by LEF associated with the carboxylation and oxygenation reactions of ribulose 1·5-bisphosphate carboxylase/oxygenase (Rubisco) to sustain all of the metabolic reactions associated with CO2 assimilation, photorespiration and related nitrogen metabolism (Cornic & Baker 2007). Consequently, leaves must operate a mechanism other than LEF to meet such an ATP deficit, or they must by some means utilize the excess reductant formed. The requirement for regulation of NADPH and ATP supply becomes compounded when it is considered that in leaves the proportion of these products of LEF consumed by processes other than CO2 assimilation can increase substantially when leaves are at different developmental stages or operating under stress. Many of these processes have quite different demands for ATP and reductants than CO2 assimilation via the C3 photosynthetic reduction cycle (Baker et al. 2004). Clearly, leaves must operate mechanisms that can modify the supply of ATP and NADPH to meet changing metabolic demands. The following sections examine possible mechanisms for achieving this.

Cyclic electron flux

Cyclic electron flow around PSI (CEF1) has frequently been suggested to augment the ATP/NADPH output ratio. The general CEF1 pathway (Fig. 1) starts with the transfer of electrons from P700 in PSI to ferredoxin, which reduces PQ to PQH2 at a plastoquinone reductase (PQR). Normal oxidation of PQH2 by the cytochrome b6f complex and plastocyanin then returns the electrons to P700, completing the cycle. The process results in no net reduction of NADP, but protons are taken up from the stroma upon reduction of PQ at PQR and released upon oxidation of PQH2 at the cytochrome b6f complex, generating a pmf, which can drive ATP synthesis. The identity and proton pumping capacity of PQR are key unknowns, and at least four versions of PQR have been proposed (Bendall & Manasse 1995; Kramer et al. 2004a; Avenson et al. 2005b; Cruz et al. 2005), possibly operating in parallel (Holser & Yocum 1987; Scheller 1996; Munekage et al. 2004; Ducruet et al. 2005) or in species-specific manners (Havaux, Rumeau & Ducruet 2005). Consequently, the H+/e- ratio for CEF1 is not known because the particular pathways operating have not been identified.

Although substantial rates of CEF1 in the steady state have been demonstrated in Chlamydomonas and C4 mesophyll chloroplasts, the situation in C3 vascular plants remains confusing, mainly because of difficulties in measuring cyclic processes (Kramer et al. 2004a; Avenson et al. 2005b; Cruz et al. 2005; Ducruet et al. 2005). Some groups have presented evidence for high rates under special conditions, such as during induction of photosynthesis (Joet et al. 2002; Joliot & Joliot 2002, 2005), while others found evidence for only small contributions (or contributions proportional to LEF) to steady-state photosynthesis (Baker & Ort 1992; Kramer et al. 2004a,b; Avenson et al. 2005a,b; Cruz et al. 2005). It would appear that the capacity for CEF1 could be substantial, but that it is highly regulated to prevent imbalances in ATP/NADPH, yielding contributions that are usually small or a constant fraction of LEF. There is evidence that the rate of CEF1 in C3 chloroplasts is set by the competition with other processes for electrons from PSI (Joet et al. 2002; Joliot & Joliot 2005), which may act hierarchically diverting reductant through different pathways.

Oxygen as an alternate electron acceptor to CO2

Oxygen photoreduction by LEF from PSI, often termed the Mehler reaction, is an attractive mechanism for increasing the ATP/NADPH produced. Operation of the water–water cycle results in the generation of a pmf in the absence of NADPH production. Oxygen is photoreduced by the PSI iron–sulphur centres, ferredoxin and other redox centres in the stroma, in amounts stoichiometrically equivalent to the oxygen evolved by PSII-mediated photooxidation of water (Asada 1999). The superoxide radicals resulting from the one-electron reduction of molecular oxygen are rapidly dismutated by superoxide dismutase to hydrogen peroxide, which is detoxified by ascorbate peroxidase oxidizingascorbate to monodehydroascorbate (MDA). To sustain operation of the water–water cycle, ascorbate has to be regenerated from MDA and this requires reduction of MDA by electrons transferred from ferredoxin. Consequently, the operation of the water–water cycle generates a large sink for electrons and a pmf. The presence in chloroplasts of high concentrations of ascorbate in excess of 20 mm (Smirnoff, Conklin & Loewus 2001) and of 25 mm glutathione, a key metabolic intermediate in the regeneration of ascorbate from MDA (Asada 1999) would support the contention that the water–water cycle can operate and be sustained at high rates (Polle 2001).

Although there are rather few direct measurements of oxygen photoreduction in leaves, there is a considerable amount of circumstantial evidence that supports the operation of the water–water cycle in stressed leaves (Ort & Baker 2002). Large increases in the ratio of PSII electron transport:CO2 assimilation have been observed in stressed leaves (Cheeseman et al. 1997; Fryer et al. 1998; Farage et al. 2006). Under such stress conditions, operation of the water–water cycle would serve to protect PSII from photoinhibition by maintaining LEF through PSII. However, such water–water cycle activity would be expected to be associated with the creation of a large pmf and ATP production, and possible damage caused by reactive oxygen species escaping the superoxide detoxification mechanism. The proton translocation-coupled steps of the water–water cycle are identical to those of LEF, and thus, the H+/e- ratios for the two processes would be identical. Consequently, under such conditions, significant sinks for ATP must exist, or alternatively a mechanism for uncoupling LEF from ATP synthesis must operate.

Thylakoids have a plastid terminal oxidase (PTOX) that can transfer electrons from PQH2 directly to oxygen, which has been associated with the process of chlororespiration (Nixon 2000; Peltier & Cournac 2002). However, in the light electron flux from PSII through PTOX to oxygen could generate a pmf and potentially be involved as a potential sink for electrons serving to aid in photoprotection of PSII. Electron flux to oxygen via chlororespiratory pathways is generally considered to be very low, being estimated at ca. 0.3% of light-saturated electron flux in sunflower leaves (Field, Nedbal & Ort 1998). However, recently, very high levels of PTOX have been found to occur in leaves of some alpine species, which were found to exhibit high ratios of rates of LEF through PSII to CO2 assimilation and very low capacities for antioxidant scavenging of reactive oxygen species (Streb et al. 2005). Consequently, in some plants, PTOX may offer a mechanism for dissipating PSII excitation to oxygen that does not involve electron flux through PSI, although to maintain such an electron flux through PSII to oxygen the pmf generated must be dissipated by production and consumption of ATP or uncoupling of electron flux from ATP synthesis, as is the case for PSI-mediated photoreduction of oxygen in the Mehler reaction.

The malate shuttle

In addition, or as an alternative, to the balanced generation of ATP and reductant, excess reducing power can be exported from the chloroplast to the cytosol via the combined actions of chloroplast NADP–malate dehydrogenase (NADP–MDH) and the chloroplast envelope malate–oxaloacetate shuttle (Backhausen, Kitzmann & Scheibe 1994; Fridlyand, Backhausen & Scheibe 1998). NADP–MDH reduces oxaloacetate to malate using NADPH, and the malate is exchanged for cytosolic oxaloacetate by the shuttle. Once in the cytosol, the malate can be reoxidized to oxaloacetate with the release of reductant, which can then be used outside the chloroplast, to reduce, for example, nitrate to nitrite, or depending on mitochondrial/cytosolic redox potentials, be used by the mitochondrial electron transport chain (Hanning & Heldt 1993). Consequently, mitochondria could be a significant sink for excess reductants transported out of the chloroplasts. To what extent reductant can be exported from the chloroplast and how important this would be in balancing ATP/reductant of the chloroplast depend on the regulation of NADP–MDH and how this is related to the other mechanisms that act to adjust ATP/reductant.

The regulation and properties of NADP–MDH and the malate transporter determine the flux of reductant out of the cytosol. The maximum capacity of NADPH consumption by NADP–MDH is about 50% that of NADPH consumption by CO2 fixation (Fridlyand et al. 1998), which suggests that at least transiently the rate of oxaloacetate reduction can be a relatively substantial sink for reductant. The capacity for malate export from the chloroplast does not appear to be limiting, although it is not known for how long oxidative processes in the cytosol or malate storage by the vacuole could support this rate of reductant export. In unstressed leaves, the rate of reductant export from the chloroplast via malate export is expected to be around 5% of the rate of reductant used by CO2 assimilation fixation. Importantly, in chloroplasts under conditions of diminishing CO2 fixation, the activation of NADP–MDH, and thus reductant export from the chloroplast, has been shown to precede activation of cyclic electron transport and the Mehler reaction (Backhausen et al. 2000). It may be, therefore, that the export of reductant to the cytosol is the principal means by which the rates of formation of ATP and reductant by the chloroplast are brought into balance.


The evolution of the non-destructive, quantitative techniques described earlier offers plant scientists the opportunity to probe effectively the functioning and limitations of the photosynthetic apparatus of thylakoids in vivo. When these are used appropriately with other non-invasive techniques, such as gas analyses and infrared imaging thermometry, the relationships between light-use efficiency and metabolism, which determine leaf photosynthetic productivity, can be definitively resolved.

Although current knowledge of the structure, composition and function of the components of the photosynthetic apparatus is highly advanced, quantitative understanding of the cooperation of these components in vivo under physiological relevant conditions is surprisingly limited. Simultaneous applications of the non-destructive techniques described earlier provide the opportunity to address these important questions.


We are grateful to the numerous colleagues who have contributed to our studies in developing the techniques and approaches described in this review. The Biotechnology and Biological Research Council has supported many of N.R.B.'s studies on this topic; D.M.K. was supported by the U.S. Department of Energy (DE-FG02-04ER15559).