Proton motive force partitioning links energy and redox balance to photoprotection and carbon gain

Fluctuating irradiance forces leaves to balance energy conversion with protection against reactive oxygen species (ROS) produced when light harvesting exceeds metabolic demand. In chloroplasts, this balance is strongly governed by the thylakoid proton motive force (pmf, ΔμH+) and by its partitioning between a pH gradient (ΔpH) and an electric field (Δψ). A proton-circuit framework in which proton deposition by linear and cyclic electron flow builds pmf, chloroplast ATP synthase spends pmf as ATP with an effective proton conductivity g(H+), and counter-ion fluxes reshape ΔpH:Δψ on seconds-to-minutes timescales. Δψ-relieving anion pathways (VCCN1, CLCe) promote rapid ΔpH expression during light increases, enabling timely engagement of PsbS-dependent qE and ΔpH-dependent photosynthetic control at cytochrome b6f, whereas the K+/H+ antiporter KEA3 accelerates ΔpH relaxation after transitions to lower light to speed recovery. These dynamics link to stromal metabolism by describing how stromal alkalinization and Mg²+/thioredoxin regulation activate Calvin–Benson–Bassham enzymes, how CEF pathways (PGR5/PGRL1 and NDH) increase pmf without net NADPH production, and how phosphate recycling and triose-phosphate utilization constrain ATP synthase flux. This review examines how thylakoid architecture could generate spatial heterogeneity in proton dynamics and highlight what remains inferred versus directly measured. Finally, we present an operating-regime map and a minimal diagnostic toolkit—multiwavelength ECS (pmf, ΔpH/Δψ, g(H+)) combined with NPQ, P700, and gas exchange—to translate mechanism into testable predictions and improve cross-study comparability. The unifying design principle is timing: rapid ΔpH formation to protect PSI during upshifts, followed by timely relaxation to minimize unnecessary quenching and sustain CO2 assimilation.

Photosynthesis converts light energy into chemical free energy, supporting most life on Earth by producing reduced carbon and O2 (Hohmann-Marriott and Blankenship, 2011). In leaves, the immediate energetic intermediate that links light-driven electron transfer to ATP synthesis is the proton motive force (pmf, ΔμH+) across the thylakoid membrane. Water oxidation at photosystem II (PSII) and proton-coupled plastoquinol oxidation at the cytochrome b6f complex deposit protons into the thylakoid lumen, generating a pmf with two thermodynamic components: a chemical gradient (ΔpH) and a transmembrane electric potential (Δψ). Although ΔpH and Δψ are thermodynamically interchangeable contributors to ΔμH+, they are not physiologically equivalent: ΔpH sets lumen pH (a chemical signal sensed by multiple regulatory modules), whereas Δψ sets the electric field across the membrane (Jagendorf and Uribe, 1966; Cruz et al., 2001). In intact leaves, pmf is often expressed predominantly as ΔpH under steady illumination, yet Δψ can contribute strongly during rapid transients or persist in contexts where counter-ion movements, buffering capacity, or membrane ion permeability constrain charge compensation (Kramer et al., 2004; Tikhonov, 2013). Thus, pmf partitioning reflects not only regulatory “intent,” but also physical and biochemical constraints that shape how quickly a given pmf can be converted into lumen acidification. pmf powers ATP synthesis by chloroplast ATP synthase and, through its partitioning between ΔpH and Δψ, influences the activation thresholds and kinetics of rapid photoprotective feedbacks in vivo (Avenson et al., 2005; Davis et al., 2017).

Natural irradiance is highly dynamic, especially within canopies where sunflecks and shadeflecks occur on seconds-to-minutes timescales (Kaiser et al., 2018). During rapid increases in light, electron input from the photosynthetic apparatus can temporarily exceed the capacity of stromal sinks to consume ATP and reducing power, increasing the risk of photosystem I (PSI) acceptor-side over-reduction and reactive oxygen species (ROS) formation (Havaux, 2020; Kono and Terashima, 2016). A key protective response is ΔpH-dependent “photosynthetic control” at cytochrome b6f, in which lumen acidification slows plastoquinol oxidation at the Qo site and restricts electron delivery to PSI when sinks lag (Tikkanen et al., 2014, 2015; Degen and Johnson, 2024). In parallel, lumen acidification activates non-photochemical quenching (NPQ)—especially the rapid, reversible qE component mediated by PsbS and supported by xanthophyll cycling—thereby reducing excitation pressure on PSII and reshaping upstream electron flow (Li et al., 2000, 2002; Fan et al., 2015). At the same time, rapid photoprotection under fluctuating light is not proton-dependent alone: antenna re-organization, protein phosphorylation/state transitions, and redox signaling can modulate excitation distribution and electron-transfer capacity in ways that interact with (but are not reducible to) ΔpH dynamics (Johnson and Ruban, 2014; Rochaix, 2014). Because PSI photoinhibition recovers slowly and can depress CO2 assimilation long after a transient, the timing of lumen acidification and its relaxation—together with how other fast regulatory layers are coordinated with these proton signals—is a central determinant of integrated carbon gain under fluctuating light (Zivcak et al., 2015; Kaiser et al., 2018).

Mechanistically, pmf dynamics during light transitions can be organized by three coupled levers: (i) proton deposition by linear electron flow and by cyclic electron flow (CEF) around PSI, (ii) proton efflux through chloroplast ATP synthase, and (iii) counter-ion fluxes that redistribute a given pmf between Δψ and ΔpH (Avenson et al., 2005; Davis et al., 2017). ATP synthase is not a passive consumer of pmf: redox regulation and kinetic control tune its effective proton conductivity—often parameterized as g(H+)—thereby setting the time constant for ΔpH decay after changes in light or metabolic demand (Kiirats et al., 2009; Hisabori et al., 2013). Conversely, stromal metabolism feeds back on proton-circuit dynamics because ATP synthase throughput depends on substrate availability, including ADP and inorganic phosphate (Pi); limitations in Pi recycling or triose-phosphate export can therefore constrain how effectively pmf is “spent” as ATP and can prolong ΔpH-dependent regulation even when light decreases (Kiirats et al., 2009). CEF adds a tunable proton flux on top of linear flow without net NADPH production; genetic and physiological studies implicate both the PGR5/PGRL1 pathway and the NDH complex in sustaining pmf and PSI safety under fluctuating light or high ATP demand (Hertle et al., 2013; Munekage et al., 2004; Hertle et al., 2013; Shikanai, 2007; Nawrocki et al., 2019; Yamori et al., 2016). Importantly, the relative dominance and physiological “purpose” of these CEF routes can differ across photosynthetic systems and cell types.

For example, in C4 species—particularly within bundle-sheath chloroplasts—NDH-dependent CEF can be quantitatively dominant and is often interpreted primarily as an ATP-supply mechanism for the CO2-concentrating cycle, rather than mainly as a rapid photoprotective buffer emphasized in many C3 leaf contexts (Yamori et al., 2016; Ermakova et al., 2019). This altered energetic context reshapes how pmf is generated, partitioned, and read out by downstream effectors, and it motivates a cautious, context-aware interpretation of ΔpH “timing” as a design principle across diverse photosynthetic systems.

Thylakoid ion transport provides the fast counter-ion pathways needed to reshape pmf partitioning on seconds-to-minutes timescales. The bestrophin-like Cl- channel VCCN1 and the CLC-family member CLCe contribute to Δψ dissipation, promoting rapid lumen acidification and timely engagement of qE and photosynthetic control during light increases (Duan et al., 2016; Herdean et al., 2016a; Dukic et al., 2019, 2022). By contrast, the thylakoid K+/H+ antiporter KEA3 accelerates ΔpH relaxation after transitions to lower light, helping qE relax and improving recovery of quantum yield when strong protection is no longer required (Armbruster et al., 2014, 2016; Wang and Shikanai, 2019). Evidence that KEA3 activity is modulated by stromal energy status reinforces that “ion transport” is not an isolated layer, but part of a feedback network linking metabolism, ATP synthase conductivity, and photoprotection (Armbruster et al., 2016; Correa Galvis et al., 2020). At the same time, the operational flexibility of ΔpH/Δψ partitioning is bounded by thylakoid permeability, buffering capacity, and charge-screening effects; these constraints can limit how independently ΔpH and Δψ can be tuned in vivo, especially under stress or unusual ionic/metabolic states (Kramer et al., 2004; Tikhonov, 2013). Finally, additional K+ pathways illustrate why transporter identity and localization must be treated cautiously: TPK3 was initially reported as thylakoid-localized, but later work found photosynthesis largely unaffected by loss of TPK3 function, consistent with a non-thylakoid role and underscoring unresolved aspects of the thylakoid ion transportome (Carraretto et al., 2013; Höhner et al., 2019).

In this review, we discuss a pmf- and partitioning-centered (“proton-circuit”) view of leaf photosynthesis that connects thylakoid biophysics (“pumps, valves, and shapers”) to photoprotection and carbon gain under fluctuating light, while explicitly noting where energetic context, membrane properties, and metabolic state can shift the dominant control points. Section 1 defines how pmf is generated and parsed into ΔpH and Δψ in intact leaves and summarizes how electrochromic-shift (ECS) methods quantify pmf, partitioning, and ATP synthase conductivity in vivo, including key methodological caveats (Cruz et al., 2001; Davis et al., 2017). Section 2 discusses how proton pumping (linear flow and CEF), proton efflux through ATP synthase (g(H+)), and counter-ion transport (KEA3, VCCN1, CLCe) interact—within physiological and biophysical constraints—to set lumen pH and PSI redox poise during light transients. Section 3 then focuses on ΔpH-sensitive effectors—PsbS-mediated qE, xanthophyll-cycle enzymes, and ΔpH-dependent photosynthetic control at cytochrome b6f—and emphasizes how their kinetics and amplitudes are modulated by genotype, pigment state, and environmental history (Li et al., 2000, 2002; Fan et al., 2015; Demmig-Adams et al., 2020; Degen and Johnson, 2024). Section 4 connects proton-circuit dynamics to stromal activation, ATP/NADPH balancing, and CO2 assimilation kinetics during sunflecks and shadeflecks, including feedbacks mediated by Pi availability and sink strength. Section 5 evaluates evidence for spatial heterogeneity (“microdomain” pmf) imposed by thylakoid architecture, distinguishing what is directly measured from what is inferred.

Finally, Section 6 integrates these elements into an operating-regime map and a minimal diagnostic toolkit, and in section 7, highlights how mechanistic ordinary-differential-equation (ODE) models can complement experiments by formalizing the proton circuit, quantifying sensitivity to key parameters (e.g., ion flux conductances and g(H+)), and helping identify where the framework may oversimplify behavior. Guided by recent advances and persistent controversies, we organize the review around three questions: (1) what sets the dynamic ΔpH:Δψ partition of pmf in intact leaves, and how does it change across light regimes and during transients? (2) how do CEF, ATP synthase g(H+), and thylakoid ion transporters (KEA3, VCCN1, CLCe) interact to stabilize PSI during light fluctuations, and which interactions most strongly constrain (or enable) carbon gain? (3) what molecular features determine the pH-response curves and kinetics of PsbS-dependent qE and xanthophyll-cycle engagement, and how do these kinetics shape recovery after light decreases (Table 1)?

The thylakoid proton motive force (pmf, ΔμH+) is the electrochemical driving force for proton movement across the thylakoid membrane. It can be decomposed into two thermodynamic components: a transmembrane electric potential (Δψ) and a pH gradient (ΔpH). Although Δψ and ΔpH are thermodynamically interchangeable contributors to ΔμH+, they are not physiologically equivalent. ΔpH directly sets lumen pH and thus gates multiple pH-sensitive regulatory processes (e.g., photosynthetic control at cytochrome b6f and the activation of NPQ components), whereas Δψ reflects the electric field across the membrane and can dominate transiently when charge compensation is delayed, potentially altering recombination pathways and ROS risk even when total pmf is similar (Kaňa et al., 2019; Cruz et al., 2001; Davis et al., 2017; Tikhonov, 2013; Kaňa et al., 2019). Importantly, pmf partitioning should not be treated as a purely deterministic “regulatory choice”: the operational ΔpH:Δψ ratio is constrained by the conductance and kinetics of counter-ion pathways (anion influx and/or cation efflux), membrane capacitance and surface charge screening, and the buffering capacity and effective permeability of the thylakoid lumen and stroma (Cruz et al., 2001; Kramer et al., 2004; Tikhonov, 2013). Consequently, while many C3 leaves under typical ionic conditions express a large fraction of steady-state pmf as ΔpH, Δψ can persist or remain relatively large in contexts where counter-ion flux is limiting (e.g., ion-transport mutants, altered ionic strength, or particular stress states), yielding a situation in which total pmf may be substantial but lumen acidification—and therefore engagement of ΔpH-dependent effectors—lags behind (Cruz et al., 2001; Herdean et al., 2016; Davis et al., 2017; Dukic et al., 2019).

Finally, even when ΔpH forms, its regulatory “effectiveness” depends on the sensitivity and availability of downstream modules (e.g., antenna composition, PsbS abundance, xanthophyll pool state) and can be modulated by additional fast regulatory layers that are not strictly proton dependent, including protein phosphorylation/state transitions and redox signaling (Johnson and Ruban, 2014; Rochaix, 2014).

The electrochemical potential difference for protons across the thylakoid membrane can be written (with sign conventions stated explicitly) as:

where ΔμH+ is the electrochemical potential difference for protons (the pmf, units of J mol−1); F is the Faraday constant (C mol−1); Δψ is the trans-thylakoid electric potential difference, defined as the electric potential in the lumen minus that in the stroma (ψlumenψstroma) and typically expressed in volts; R is the universal gas constant (J mol−1 k−1); T is absolute temperature (K); and ΔpH is the pH difference across the thylakoid membrane, defined as stroma minus lumen (pHstromapHlumen). The factor 2.303 converts between natural and base-10 logarithms. Recent re-analyses of ECS signals reinforce this conclusion and caution against over-attributing steady-state ECS plateaus to Δψ. Because Δψ and ΔpH are thermodynamically equivalent in the pmf, any biological process that changes charge screening or ion movement (e.g. K+, Cl-, H+/K+ exchange) will redistribute pmf between Δψ and ΔpH without necessarily changing the total ΔμH+. This is central to understanding both fast protection and ATP yield in fluctuating light (Juhaszova et al., 2022), so that at ~25 °C the chemical term contributes ~59 mV per pH unit. Because Δψ and ΔpH sum to the same total driving force, any process that changes charge screening or counter-ion movement (e.g., K+ fluxes, Cl- conductances, and K+/H+ exchange) can redistribute a given pmf between Δψ and ΔpH without necessarily changing the total Δp, with direct consequences for both ATP synthesis (which depends on total pmf) and rapid photoprotection (which depends strongly on lumen pH and thus on ΔpH expression).

This is why “pmf plateau” values and “Δψ fractions” inferred from electrochromic shift (ECS) measurements must be interpreted with care: the ECS signal primarily reports the pigment Stark effect driven by Δψ, but quantitative partitioning requires rigorous multi-wavelength controls and spectral deconvolution to avoid confounding by overlapping absorbance changes (e.g., carotenoid- and antenna-associated bands) and to avoid over-assigning steady-state ECS plateaus to Δψ when much of the steady-state pmf may reside in ΔpH (Joliot and Joliot, 2002; Davis et al., 2017; Uflewski et al., 2021; Wilson et al., 2021).

The sources of proton deposition into the lumen are defined by electron flow (Figure 1). Water oxidation at the PSII oxygen−evolving complex releases one lumenal proton per transferred electron. Reduced plastoquinone (PQH2) is then oxidized at the cytochrome b6f complex; when the full Q−cycle operates, b6f contributes two additional protons per electron to the lumen (i.e., four protons per PQH2 oxidized), yielding a nominal total of ~3 H+ per electron transferred from water to PSI acceptors (Hope, 2000; Sacksteder et al., 2000; Tikhonov, 2018). Because two electrons are required to reduce one NADP+, this corresponds to ~6 H+ translocated per NADPH formed by linear electron flow (LEF). Coupled with the proton/ATP ratio of chloroplast ATP synthase (set by CF1CF0 rotor stoichiometry and potentially varying among taxa), this stoichiometric baseline helps explain why LEF alone often produces an ATP/NADPH output near ~1.3 and why auxiliary ATP−biasing routes are required when stromal demand approaches ~1.5 ATP per NADPH (Joliot and Joliot, 2002; Munekage et al., 2004; Tikhonov, 2018). Notably, these “stoichiometric constraints” provide a useful energetic scaffold, but in vivo outcomes can deviate from simple stoichiometry because pmf utilization depends on ATP synthase conductance and substrate availability (ADP and Pi), and because pmf partitioning (Δψ vs ΔpH) affects how quickly lumen pH reaches the thresholds for rapid regulatory feedbacks during transients (Figure 1).

Structural and mechanistic work has consolidated the basis for lumenal proton release and its coupling to electron transfer. High-resolution cryo-EM reconstructions of plant cytochrome b6f reveal quinones poised at the Qp site and features consistent with a plastoquinone channel, strengthening the mechanistic linkage between electron bifurcation in the Q-cycle and proton release/uptake across the membrane (Johnson, 2011; Tikhonov, 2018). Complementary time-resolved and mutational studies on PSII have identified hydrogen-bonded proton-egress pathways from the oxygen-evolving complex, consistent with the idea that protein architecture and lumenal microenvironment can shape how rapidly lumen acidification develops after light onset. Together, these insights rationalize why ΔpH-dependent slowdown of plastoquinol oxidation at b6f (“photosynthetic control”) provides a direct chemical coupling between lumen proton activity and electron-transfer throughput (Herdean et al., 2016; Davis et al., 2017; Tikhonov, 2018).

Cyclic electron flow (CEF) superimposes a tunable proton-pumping contribution on this LEF baseline without net NADPH production (Miller et al., 2022). In the PGR5/PGRL1-dependent route, electrons from the PSI acceptor side return to the plastoquinone pool, increasing b6f turnover and thereby increasing proton deposition to the lumen primarily through the Q-cycle (Figure 1). In the NDH-dependent route, a NAD(P)H dehydrogenase-like complex contributes additional proton pumping on top of b6f turnover, increasing the effective H+/e- yield of cyclic flow relative to the PGR5/PGRL1 route (Shikanai, 2007; Nawrocki et al., 2019; Yamori et al., 2016). The physiological interpretation of CEF therefore depends on context: in many C3 leaf settings it is often discussed as supporting rapid PSI safety by strengthening pmf-linked control during upshifts, whereas in high-ATP-demand contexts—most prominently C4 bundle-sheath chloroplasts—NDH-dependent CEF can be quantitatively dominant and may be better viewed primarily as an ATP-supply strategy for the CO2-concentrating mechanism (Yamori et al., 2016; Ermakova et al., 2019). In either case, increased proton deposition via CEF does not automatically translate into faster lumen acidification: whether added pmf appears mainly as ΔpH versus Δψ depends on concurrent counter-ion conductances and on ATP synthase proton efflux at that moment. This conditionality is central for linking “pmf generation” to downstream regulation and for avoiding over-deterministic interpretations in which ΔpH timing alone is assumed to govern all protection–productivity trade-offs across photosynthetic systems.

Plant ATP synthase typically assembles a C-ring of ~14 subunits, so one full rotation returns ~14 protons and synthesizes ~3 ATP, giving ~4.67 H+ per ATP as a practical coupling ratio. Species variation (c13-15) and engineered rotor expansions illustrate the trade-off: more protons per ATP lower energy efficiency and force compensations elsewhere in the photosynthetic network, while tighter coupling raises the ATP yield per unit pmf but can limit protective ΔpH formation unless counter-ion routes keep Δψ in check (Davis et al., 2017; Roach and Krieger-Liszkay, 2019).

Pmf partitioning into ΔpH and Δψ comprise two distinct features. First, within the first minute of illumination, leaf pmf is usually carried predominantly as ΔpH, a state that is inherently safer because ΔpH activates qE and engages photosynthetic control at b6f, whereas sustained Δψ favors charge-recombination pathways that generate singlet oxygen at PSII. Second, interpreting Δψ from ECS requires care: the pigment Stark effect at ~515–520 nm overlaps with xanthophyll-associated bands (ΔA505, ΔA535), so quantitative partitioning demands standard protocols, spectral references, and transparent reporting to make data comparable across laboratories (Joliot and Joliot, 2002; Shikanai, 2007; Kato et al., 2012).

The thylakoid proton motive force (pmf, ΔμH+) comprises two thermodynamically interchangeable components: the transmembrane electric potential (Δψ) and the pH gradient (ΔpH). However, Δψ and ΔpH are not physiologically equivalent, because ΔpH directly sets lumen pH—the chemical signal that engages rapid photoprotective feedbacks—whereas Δψ reflects the electric field across the membrane and can dominate transiently when counter-ion fluxes lag behind proton deposition (Cruz et al., 2001; Davis et al., 2017). In intact leaves, the regulatory question is rarely “how much pmf?” alone, but how quickly pmf is built and how it is partitioned into ΔpH versus Δψ during light transitions, because that partitioning controls (i) ATP synthesis and (ii) the activation thresholds and kinetics of ΔpH-dependent effectors, including qE and photosynthetic control at cytochrome b6f (Davis et al., 2017; Correa Galvis et al., 2020).

A mechanistically useful way to organize pmf partitioning is by three coupled levers:

proton deposition (“supply”) by linear electron flow (LEF) and cyclic electron flow (CEF),

proton efflux (“valve”) through chloroplast ATP synthase, often summarized as an effective proton conductivity g(H+), and

counter-ion pathways (“shapers”) that dissipate Δψ and thereby allow a given pmf to be expressed more as ΔpH (Cruz et al., 2001; Davis et al., 2017).

On the input side, proton deposition increases with irradiance because PSII water oxidation releases protons to the lumen and cytochrome b6f transfers additional protons to the lumen during plastoquinol oxidation via the Q-cycle. In vivo, the resulting pmf rise is opposed by proton return through ATP synthase and by any dissipative leaks; thus, the instantaneous pmf reflects the net balance between proton deposition and proton efflux (Sacksteder et al., 2000; Davis et al., 2017). Importantly, the initial electrical component (Δψ) can rise rapidly at light onset or during an upshift, but counter-ion fluxes (anion influx and/or cation efflux) typically dissipate Δψ on short timescales, allowing continued proton deposition to manifest as lumen acidification (ΔpH) (Cruz et al., 2001; Davis et al., 2017).

Because many of the fastest photoprotective controls are ΔpH-dependent, not “pmf-dependent in general,” the capacity to convert Δψ into ΔpH is a key determinant of how effectively the proton circuit protects PSI during rapid light increases (Davis et al., 2017; Herdean et al., 2016b).

CEF around PSI increases pmf without net NADPH accumulation by recycling electrons from the PSI acceptor side back to the plastoquinone (PQ) pool, increasing proton deposition predominantly through additional b6f turnover and (for the NDH branch) direct proton pumping (Shikanai, 2007; Nawrocki et al., 2019). Two routes are commonly distinguished: a PGR5/PGRL1-dependent pathway and an NDH-dependent pathway (Figure 2) (Shikanai, 2007; Nawrocki et al., 2019).

CEF returns electrons to PQ, its impact on PQ redox is conditional, reflecting the balance among electron input, b6f throughput, PSI acceptor-side capacity, and downstream stromal sink activity. A more defensible framing is that CEF can support PSI integrity indirectly by (i) increasing pmf/ΔpH and thereby strengthening photosynthetic control at b6f (slowing electron delivery to PSI when sinks lag) and/or (ii) improving ATP supply to stimulate stromal metabolism, which can accelerate electron consumption and relieve acceptor-side limitation (Figure 2) (Shikanai, 2007; Tikkanen et al., 2015; Davis et al., 2017; Degen and Johnson, 2024).

The key point for pmf partitioning is conditional: CEF increases the supply of pmf, but whether that increase appears mainly as ΔpH (versus Δψ) depends on counter-ion dissipation of Δψ and on ATP synthase proton efflux (g(H+)) at that moment (Cruz et al., 2001; Davis et al., 2017).

Chloroplast ATP synthase converts pmf into ATP, but its throughput is strongly regulated. Redox regulation of the γ-subunit (a thiol/disulfide switch controlled by thioredoxin) suppresses wasteful ATP hydrolysis in the dark and modulates enzyme activity in the light, thereby changing the effective rate at which protons return to the stroma (Hisabori et al., 2013). In vivo, pmf discharge kinetics are often parameterized as an effective proton conductivity, g(H+), estimated from the electrochromic shift decay after dark intervals (DIRK/ECS approaches) (Kiirats et al., 2009; Davis et al., 2017).

When g(H+) is high, pmf is spent rapidly, ΔpH relaxes quickly after a downshift, and ATP supply can keep pace—but ΔpH may not build as strongly during upshifts if counter-ion flux and pumping do not outpace efflux.