Faster relaxation of nonphotochemical quenching (NPQ) in C4 than in C3 species

16 Photosynthesis, C4, NPQ, photoprotection 17 This PDF file includes: 18 Main Text 19

20 Figures 1 to 5 21 22 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 1, 2025. ; https://doi.org/10.1101/2025.04.01.646649doi: bioRxiv preprint 2

23 Acceleration of photoprotective non-photochemical quenching (NPQ) responses to changes in 24 light intensity has been suggested as a strategy to enhance crop yield. Despite many key crops 25 utilising C4 photosynthesis, our current understanding of NPQ overwhelmingly comes from C3 26 species. Using a series of experiments on three phylogenetically controlled C3 and C4 27 comparisons, we show that NPQ relaxation is faster in C4 species. Temporal analysis of NPQ 28 relaxation in leaves infiltrated with inhibitors to block proton motive force formation or xanthophyll 29 de-epoxidation showed that the faster relaxation observed in C4 species is driven by a greater 30 contribution of energy-dependent quenching (qE) to overall NPQ. We show that the C4-31 associated enhancement of qE is linked to altered regulation of lumen pH in C4 species, 32 reflecting increases in cyclic electron flow and membrane proton conductivity to meet the 33 increased ATP demands of the C4 pathway. Indeed, in two of the three tested C4 species, NPQ 34 relaxation became significantly slower and statistically indistinguishable from paired C3 species 35 when ATP and NADPH consumption was suppressed by performing measurements in CO 2-free 36 air. Altogether, our results suggest that NPQ responses in C4 species may already be optimised 37 to maintain high photosynthetic efficiency in the fluctuating light conditions typically found within 38 C4 canopies. Given the intrinsically faster NPQ in C4 photosynthesis, further acceleration of NPQ 39 may have limited scope to enhance crop photosynthetic efficiency. 40 41 Significance Statement 42 Acceleration of non-photochemical quenching has been proposed as a means to enhance crop 43 photosynthetic efficiency in C3 species but whether this strategy has potential in C4 species, 44 which include several major crops, remains unclear. We use three phylogenetically paired C3 and 45 C4 species to show that NPQ relaxation is significantly faster in species with the C4 pathway, 46 possibly aiding the maintenance of photosynthetic efficiency in fluctuating light environments. As 47 a result, accelerating the rate of NPQ relaxation in C4 crops may have a more limited scope to 48 enhance photosynthesis. 49 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 1, 2025. ; https://doi.org/10.1101/2025.04.01.646649doi: bioRxiv preprint 3

50 51 Nonphotochemical quenching (NPQ) refers to a collection of photoprotective mechanisms 52 wherein excess light energy in photosystem II (PSII) is dissipated as heat, preventing 53 overexcitation and the formation of reactive oxygen species that would otherwise damage the 54 photosynthetic machinery (1). NPQ components operate at different timescales and likely involve 55 conformational changes in PSII-associated antennae that trigger the quenched state (2). Energy-56 dependent quenching (qE) is activated within seconds to minutes by lumen acidification (3), 57 which also triggers xanthophyll cycle enzyme violaxanthin de-epoxidase (VDE) to convert 58 violaxanthin to zeaxanthin, enhancing qE (4). Zeaxanthin accumulation further supports a qE-59 independent quenching component (qZ) that activates and recovers over minutes to hours (5). 60 Even more sustained quenching (qH) relies on the plastid lipocalin LCNP and is negatively 61 regulated by suppressor of quenching SOQ1 (6). Finally, photoinhibitory quenching (qI) is 62 associated with photodamage and requires de novo synthesis of the D1 protein for recovery (7). 63 64 Our molecular understanding of NPQ as detailed above overwhelmingly comes from C3 species, 65 where NPQ relaxation has been found to significantly lag behind changes in irradiance, 66 temporarily lowering photosynthetic efficiency and leading to substantial losses in ‘foregone’ 67 canopy carbon assimilation (8). Accelerating recovery from photoprotection represents a 68 promising strategy for improving crop yield in C3 species (9, 10), yet whilst many key crops are 69 C4 (11) the specifics of the C4 NPQ response remain largely unknown (12). 70 71 In C3 photosynthesis, CO 2 is directly fixed in mesophyll (M) chloroplasts by ribulose-1,5-72 biphosphate carboxylase/oxygenase (Rubisco) into 3-carbon compound 3-phosphoglycerate (3-73 PGA). Higher photosynthetic rates are generally found in C4 species, where a carbon 74 concentrating mechanism (CCM) enhances photosynthesis by suppressing RuBP oxygenation 75 and concomitant photorespiration (13). The C4 pathway operates between morphologically 76 distinct M and bundle sheath (BS) cells, typically arranged in ‘Kranz’ anatomy: CO 2 is initially 77 converted to bicarbonate in the M and fixed into 4-carbon oxaloacetate that is further reduced or 78 transaminated into malate or aspartate before diffusion into the BS, where decarboxylating 79 enzymes release CO 2 around Rubisco and into the C3 cycle (14). Different C4 “subtypes” use 80 different decarboxylases, often in combination (15), and have additional cell and subtype-specific 81 ATP requirements for the regeneration of CCM biochemical intermediates (16-18). The 82 ATP:NADPH ratio generated by linear electron flow (LEF) from PSII to PSI is insufficient to satisfy 83 the demands of the C3 cycle, and the additional ATP demands by C4 metabolism further the 84 imbalance (19). Cyclic electron flow (CEF) helps balance energy budgets by recycling electrons 85 around PSI back to plastoquinone (PQ), contributing to the proton motive force (pmf) that powers 86 ATP synthesis without concurrent production of NADPH (20). Considerably higher ratios of 87 PSI:PSII (21, 22) and CEF:LEF (23) are found in C4 versus C3 species, reflecting the ATP cost of 88 the C4 pathway. 89 90 The functional differences between C3 and C4 photosynthesis are likely to affect NPQ. In C3 91 species, CEF plays a major photoprotective role by contributing to ∆ pH and thus qE activation, 92 with CEF-defective mutants having severely reduced NPQ (24, 25). This could suggest an 93 enhanced qE component in C4 species, given their higher CEF:LEF ratios. In C3 species, CEF is 94 predominantly mediated by proton gradient regulation 5 (PGR5) and PGR5-like photosynthetic 95 phenotype 1 (PGRL1), with a minor contribution by chloroplast NADH dehydrogenase-like 96 complex (NDH) in low and fluctuating light (26-28). In C4 species, although PGR5/PGRL1 is still 97 important, recent studies show CEF occurs primarily via NDH (29-31). Notably, NPQ amplitude 98 was lower in C4 PGR5 and PGRL1-deficient mutants, but higher in NDH mutants (30, 31). Based 99 on these results, it was suggested that PGR5/PGRL1 could play a similar photoprotective role in 100 C4 photosynthesis as in C3 species, with NDH mainly contributing to supplementing ATP 101 production (31), but the effect of CEF pathway interplay on C4 NPQ remains unclear. Finally, 102 NPQ modulation is also regulated by the proton conductivity of the thylakoid membrane (gH +, 103 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 1, 2025. ; https://doi.org/10.1101/2025.04.01.646649doi: bioRxiv preprint 4 indicative of ATP synthase activity), as changes to gH+ affect pmf formation and dissipation (32). 104 Higher ATP consumption in C4 species results in faster turnover of inorganic phosphate and 105 substrate availability for ATP synthase (33), potentially resulting in altered control of ∆ pH-106 dependent NPQ than in C3 photosynthesis. 107 108 The present work aimed to characterise differences in NPQ relaxation between C3 and C4 109 species using a combination of spectroscopic, chemical, and molecular approaches, including C4 110 CEF mutants. To account for the strong confounding effect of phylogenetic distance (34), we 111 compared phylogenetically linked pairs of C3 and C4 species from three evolutionarily distinct 112 genera operating three different C4 metabolic cycles. 113 114 115 116

117 118 Differences in NPQ relaxation between C3 and C4 species 119 To compare NPQ between photosynthetic pathways, we selected C3 and C4 species from 120 Alloteropsis (C3 A. semialata KWT, C4 A. semialata MDG), Flaveria (C3 F. cronquistii , C4 F. 121 bidentis), and Cleome (C3 T. hassleriana, C4 G. gynandra). This experimental design minimises 122 phylogenetic variation between each C3 and C4 pair, whilst maintaining substantial evolutionary 123 distance between the three genera. The selected species represent both monocots (Alloteropsis) 124 and dicots ( Flaveria, Cleome), three independent C4 origins (35, 36), and the three major 125 decarboxylating enzymes found across C4 photosynthetic species: NADP-ME/PEPCK in C4 A. 126 semialata MDG (37), NADP-ME in C4 F. bidentis (38), and NAD-ME in C4 G. gynandra (39). 127 128 NPQ induction and relaxation responses were measured during a 1 hour photoperiod (600 µmol 129 m-2 s-1 PFD) followed by 25 minutes of darkness. Since NPQ components can be resolved based 130 on their relaxation kinetics (1, 2) and given current interest in NPQ relaxation for improving 131 photosynthesis (40), we focused on NPQ following the light-to-dark transition ( Fig. 1A, full traces 132 in Fig. S1). NPQ relaxation was faster in all C4 species relative to their C3 counterparts, resulting 133 in significantly lower NPQ across the dark period. Integrated NPQ across the dark recovery 134 period was 40% lower in C4 Alloteropsis semialata MDG than in C3 Alloteropsis semialata KWT, 135 33% lower in C4 Flaveria bidentis than in C3 Flaveria cronquistii , and 22% lower in C4 136 Gynandropsis gynandra than in C3 Tarenaya hassleriana (values and statistics in Table S1). 137 138 The NPQ relaxation kinetics of C3 and C4 species were underpinned by clear differences in NPQ 139 composition (Fig. 1B, values and statistics in Table S2). NPQ components were analysed by 140 separating NPQ relaxation into different timescales of deactivation expressed as a function of 141 total NPQ. Fast-relaxing components (0-2 min) constituted a significantly greater proportion of 142 NPQ in all C4 species than in their C3 pairs and conversely, slower-relaxing NPQ components (2-143 15 min) were a larger part of NPQ in C3 species, contributing to a more exponential decay. 144 145 Effects of chemical inhibition of Δ pH and xanthophyll-dependent components 146 Fast-relaxing NPQ components include Δ pH-sensitive and xanthophyll-dependent qE in the first 147 two minutes, and qZ dissipation in 2- 15 minutes (2). To identify the elements responsible for 148 NPQ differences between C3 and C4 species, leaves were infiltrated with nigericin to collapse the 149 proton gradient or dithiothreitol (DTT) to inhibit the xanthophyll cycle, and integrated NPQ of 150 treated leaves during the light period was compared to a control to assess NPQ dependence on 151 both mechanisms ( Fig. 2 , full traces in Fig. S2 ). Nigericin infiltration resulted in significantly 152 greater suppression of NPQ in C4 than in C3 species (Fig. 2A, values and statistics in Table S3), 153 indicating higher reliance on Δ pH for C4 NPQ responses: 75±3% reduction in C4 A. semialata 154 MDG vs. 58±9% in C3 A. semialata KWT, 88±2% in C4 F. bidentis vs. 62±4% in C3 F. cronquistii, 155 and 94±3% in C4 G. gynandra vs. 68±3% in C3 T. hassleriana. Although both qE and xanthophyll 156 de-epoxidation depend on Δ pH, the lack of significant differences in NPQ suppression by DTT in 157 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 1, 2025. ; https://doi.org/10.1101/2025.04.01.646649doi: bioRxiv preprint 5 any C3 and C4 pairs ( Fig. 2B ) suggests that differences in C4 NPQ mostly stem from Δ pH-158 sensitive qE rather than qZ or xanthophyll-dependent qE. 159 160 Effects of the removal of photosynthetic and photorespiratory electron sinks 161 The higher rates of CEF to satisfy ATP:NADPH requirements (23) and larger electron sinks found 162 in C4 photosynthesis (33) may contribute to the formation and collapse of Δ pH. To test whether 163 fast NPQ relaxation in C4 species is linked to C4 metabolism, we repeated our NPQ 164 measurements in 2% O 2 + 0 ppm CO 2 air ( Fig. 3 ). The suppression of photosynthesis and 165 photorespiration slowed down NPQ relaxation and removed differences between C3 and C4 166 Alloteropsis and Cleome pairs. Unlike the step-like decay observed in ambient air (Fig. 1A), NPQ 167 in C4 A. semialata MDG and C4 G. gynandra followed the approximately exponential decay 168 pattern of their C3 counterparts in 2% O 2 + 0 ppm CO 2 (Fig. 3A, values and statistics in Table 169 S1), and had similar relative contributions of the 0-2 min and 2-15 min components to total NPQ 170 (Fig. 3B, Table S2). In contrast, differences in NPQ kinetics between C3 and C4 species seemed 171 enhanced in Flaveria in 2% O2 + 0 ppm CO2 air. Integrated NPQ was lower in C4 F. bidentis than 172 in C3 F. cronquistii; and the 0-2 min component still represented a significantly larger proportion 173 of NPQ in C4 F. bidentis than in C3 F. cronquistii , whilst the 2-15 min component showed the 174 opposite trend. 175 176 Based on the pH requirement of qE, we speculated that C4 F. bidentis may retain H + efflux 177 capacity in 2% O2 + 0 ppm CO2 air. We estimated gH+ from decay kinetics of electrochromic shift 178 (ECS) signal measurements during brief dark intervals across 20 minutes of light to encompass 179 the 0-2 and 2-15 min NPQ components (Fig. 4). Under ambient air, all C4 species tended to have 180 higher gH + than their C3 pairs ( Fig. 4A, C & E). However, C4 F. bidentis retained H + efflux 181 capacity in 2% O 2 + 0 ppm CO 2 (Fig. 4D), albeit diminished compared to 21% O2 + 410 ppm 182 CO2, whereas suppressing photosynthesis and photorespiration resulted in gH + values 183 approaching zero in all other species ( Fig. 4B, D & F ). These results show that fast NPQ 184 relaxation in C4 species is strongly linked to proton efflux capacity, and that an alternative 185 electron sink in C4 F. bidentis likely sustains proton efflux and qE even when photorespiration 186 and photosynthesis are suppressed. 187 188 Evaluating CEF in C3 and C4 species 189 Chlorophyll fluorescence reflects the PSII quinone redox state. In darkness, with no LEF-driven 190 reduction, a post-illumination chlorophyll fluorescence rise (PIFR) is attributed to residual CEF 191 transferring electrons from stromal donors to the PQ pool that then equilibrate with PSII-192 associated quinones (41). We measured PIFR as a proxy for CEF after 1 h light treatment, 193 periodically flashing far-red light to preferentially excite PSI and temporarily enhance PQ 194 oxidation (full PIFR trace in Fig. S3 ). After the far-red pulse, a larger PIFR response was 195 observed in C3 A. semialata KWT, C4 F. bidentis, and C4 G. gynandra than in their phylogenetic 196 pairs, indicating higher CEF activity in those species ( Fig. 5A, solid lines). This is consistent with 197 ATP:NADPH ratio requirements being higher in NADP-ME (C4 F. bidentis) and NAD-ME (C4 G. 198 gynandra) subtypes than in mixed PEPCK pathways (C4 A. semialata MDG), which have lower 199 ATP requirements (18). Leaves were also treated with PGR5/PGRL1 pathway inhibitor Antimycin 200 A (dashed lines), which resulted in significantly lower PIFR in C3 F. cronquistii and C3 T. 201 hassleriana, indicating a substantial contribution of PGR5/PGRL1 to PQ reduction. The addition 202 of Antimycin A also resulted in higher PIFR values in some cases – a secondary effect of 203 Antimycin A that has been previously attributed to additional inhibition of electron transport 204 downstream of PQ (31). Even with this effect, these results validate existing knowledge of C3 and 205 C4 CEF pathways– C3 species primarily use the PGR5/PGRL1 route while C4 species use both 206 the PGR5/PGRL1 and the NDH pathways (26, 30, 31). C3 A. semialata KWT appears to be an 207 exception, but given that the C3 subspecies may have reversed from a C3-C4 intermediate, CEF 208 operation might differ from other C3 plants (42). 209 210 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 1, 2025. ; https://doi.org/10.1101/2025.04.01.646649doi: bioRxiv preprint 6 The role of CEF in C4 NPQ relaxation was further studied in C4 F. bidentis FbPGRL1-RNAi and 211 FbNdhO-RNAi knockdown lines (31) – as previously observed, carbon assimilation was notably 212 lower only in FbNdhO-RNAi (Fig. S4), suggesting an impaired C4 cycle. NPQ in FbNdhO-RNAi 213 decayed more slowly than the step-wise drop observed in WT and FbPGRL1-RNAi (Fig. 5B, 214 values and statistics in Table S4). Whilst the 0–2 min component constituting most of WT NPQ 215 was significantly reduced in both FbPGRL1-RNAi and FbNdhO-RNAi, the slower NPQ relaxation 216 of FbNdhO-RNAi is evidenced by a more significant 2-15 min component than in WT. At the end 217 of illumination, NPQ was lower in FbPGRL1-RNAi than in WT and FbNdhO-RNAi, suggesting that 218 differences in NPQ composition stem from impaired relaxation in FbNdhO-RNAi, but from overall 219 NPQ suppression in FbPGRL1-RNAi. 220 221 222 223

224 225 C4 species have faster NPQ relaxation 226 This study sought to characterise differences in NPQ relaxation between C3 and C4 227 photosynthesis, by comparing phylogenetically linked Alloteropsis, Flaveria, and Cleome C3 and 228 C4 species. Despite considerable evolutionary distance between the three tested genera, all C4 229 species had significantly faster and overall greater NPQ relaxation than their C3 pairs ( Fig. 1), 230 showing that this is likely linked to the C4 pathway. The rapid return of PSII to the unquenched 231 state in C4 species would support higher photosynthetic quantum yields following decreases in 232 irradiance, and could be contributing to the more sustained rates of CO 2 assimilation that have 233 been observed in C4 versus C3 species during light-shade transitions (43, 44). Whilst increasing 234 the rate of photosynthetic efficiency has been remarkably successful at improving photosynthetic 235 efficiency in C3 species (9, 10), our results suggest that this approach may result in more limited 236 gains in carbon assimilation in C4 species given their intrinsically faster NPQ relaxation rate. 237 238 A greater proportion of C4 NPQ is qE 239 The comparatively faster NPQ relaxation found in C4 plants stemmed from differences in NPQ 240 composition between C3 and C4 species. Separation of NPQ into components based on decay 241 timescales revealed that C4 species had a significantly higher proportion of a fast-relaxing (0-2 242 min) component compared to their C3 phylogenetic pairs, with C3 species exhibiting a 243 comparatively greater proportion of NPQ relaxation within the 2-15 min timeframe (Fig. 1B). NPQ 244 component qE operates within the 0-2 min timescale, responding to lumen acidification and 245 enhanced by zeaxanthin accumulation (2, 45). When infiltrated with ∆ pH-inhibitor nigericin, NPQ 246 in C4 species was significantly more depressed than in C3 species, whereas inhibiting 247 xanthophyll cycle activity with DTT did not result in significant differences in the extent of NPQ 248 reduction between C3 and C4 species ( Fig. 2). These results demonstrate that the acceleration 249 of NPQ identified in all three C4 species involved ∆ pH-dependent qE. Activation of qE by lumen 250 pH relies on the presence of photosystem II subunit S (PsbS), which initiates the quenched LHCII 251 state via a conformational switch upon protonation of lumen-exposed protonatable residues (46, 252 47). Thus, one hypothesis could be that C4 species may show slight alterations in PsbS relative 253 abundance, structure, or interactions with LHCs to explain the enhancement of qE. 254 255 The energetic requirements of the C4 pathway affect NPQ 256 When photosynthesis and photorespiration were suppressed, NPQ relaxation in C4 A. semialata 257 MDG and G. gynandra conformed to the slower exponential decay and composition of C3 A. 258 semialata KWT and T. hassleriana ( Fig. 3). In ambient air, C4 metabolism largely suppresses 259 photorespiration so these results suggest that the fast NPQ relaxation observed in Alloteropsis 260 and Cleome C4 species is specifically related to the ATP and NADPH demand by the C4 261 photosynthetic pathway. qE is regulated by lumen pH, which is in turn dependent on membrane 262 proton conductivity, gH+ (32). The higher gH+ found in all three C4 species in ambient conditions 263 (Fig. 4A, C & E) could contribute to accelerated relaxation of pH-dependent NPQ, primarily qE. 264 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 1, 2025. ; https://doi.org/10.1101/2025.04.01.646649doi: bioRxiv preprint 7 Different C4 pathways have distinct ATP:NADPH requirements: in NADP-ME/PEPCK subtypes 265 (C4 A. semialata MDG), the energetic balance is comparable to C3 species, whereas NADP-ME 266 (F. bidentis) and NAD-ME ( G. gynandra) pathways have elevated ATP:NADPH demands. The 267 increased ATP demand is localised in BS cells in NADP-ME subtypes but in M cells in NAD-ME 268 subtypes (16-18). M cells are positioned in the outer layer of Kranz anatomy, and therefore 269 represent the majority of the chlorophyll fluorescence signals used to determine NPQ. However, 270 in all three C4 species, the large metabolic pools of C4 cycle intermediates required to sustain 271 diffusion gradients between M and BS cells may underpin the observed increases in gH + relative 272 to C3 species. The high gH+ indicates upregulated proton efflux, be it ATP synthase dependent or 273 independent as photoprotective Δ pH has been found to be linked both to ATP synthase 274 regulation (32, 48) as well as to thylakoid antiporters like KEA3 (49), both mechanisms that could 275 differ in C4 photosynthesis compared to C3. Uniquely, in C4 F. bidentis NPQ relaxation was not 276 slowed by the suppression of photosynthesis and photorespiration ( Fig. 3), and gH + was much 277 higher than in the other C4 species ( Fig. 4D), indicating the presence of an alternative electron 278 sink. The identity of this electron sink remains unclear but it seems plausible that it relates to 279 specific attributes of the canonical NADP-ME C4 pathway, such as the significantly larger malate 280 pool sizes that could go towards ATP-consuming metabolic reactions (50), although sustained 281 leakiness of the membrane via ATP synthase-dependent and independent mechanisms is also 282 possible. 283 284 Contribution of CEF pathways to faster NPQ in C4 species 285 In C3 species, CEF via the PGR5/PGRL1 pathway is essential for photoprotection by contributing 286 to ∆ pH formation and qE (24, 25). Relative to the ancestral C3 pathway, PGR5/PGRL1 287 abundance increases in C4 species in both M and BS cells (17, 51), whilst NDH differentially 288 accumulates per ATP requirements: in M cells for NAD-ME pathways and in BS for NADP-ME 289 (16, 18). Our results point to both NDH and PGR5/PGRL1 pathways contributing to fast relaxation 290 in C4 NPQ. Both C4 FbNdhO-RNAi and FbPGRL1-RNAi lines showed altered NPQ relaxation 291 kinetics relative to WT ( Fig. 5B ). Consistent with our findings, previous work found the 292 PGR5/PGRL1 pathway to significantly contribute to NPQ induction in C4 F. bidentis (31). Here, 293 we also show that knocking down PGRL1 results in a specific decrease in qE activation (Fig. 5C), 294 probably due to deficient pmf formation. In contrast, reduction of NDH expression primarily 295 slowed down NPQ responses. C4 F. bidentis NDH is primarily found in the BS, accounting for 296 most of BS CEF (30). Although the chlorophyll fluorescence signal primarily comes from the M, 297 an impaired 0-2 min NPQ component was still observed in FbNdhO-RNAi mutants (Fig 5C). The 298 tissue-specific expression of NDH in C4 species and the deleterious effect of its suppression on 299 growth and carbon assimilation suggests NDH is the major route for ATP provision to C4 300 metabolism (30, 31). Accordingly, the rate of carbon assimilation in FbNdhO-RNAi was greatly 301 reduced relative to wild type ( Fig S4 ). An impaired C4 cycle in FbNdhO-RNAi would alter 302 inorganic phosphate availability and LEF:CEF energy balance, potentially diminishing CEF-303 related qE and leading to parallel decreases in gH+ and ∆ pH in M cells (32, 33). Since the role of 304 NDH in C4 photosynthesis has thus far only been studied in NADP-ME plants with NDH-enriched 305 BS, it remains unclear if NDH-mediated CEF generally also plays a photoprotective role similar to 306 PGR5/PGRL1. Future research could test this in NAD-ME species with NDH-enriched M (22), 307 such as C4 G. gynandra. 308 309 The complexity of NPQ molecular mechanisms, evolution of CEF pathways, and biochemical 310 diversity within C4 species leave many outstanding questions with relation to C4 NPQ. 311 Nevertheless, the results presented here demonstrate that the higher rates of CEF and larger 312 electron sinks of C4 metabolism result in enhancement of the qE component, leading to faster 313 relaxation kinetics. The increased contribution of qE to total NPQ in C4 species mimics the 314 engineering attempts to accelerate NPQ responses in C3 crops via single overexpression of 315 PsbS (52, 53) or in combination with violaxanthin de-epoxidase and zeaxanthin epoxidase (9, 316 10), engineering attempts which may have less scope to further accelerate NPQ responses and 317 photosynthetic efficiency in C4 species. 318 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 1, 2025. ; https://doi.org/10.1101/2025.04.01.646649doi: bioRxiv preprint 8 319 320

321 322 Plant material and growth conditions 323 C4 F. bidentis, C3 F. cronquistii, C4 G. gynandra and C3 T. hassleriana were grown in soil under 324 growth chamber conditions at 20ºC and 150 µmol m -2 s-1 PFD over a 16-hour photoperiod. A. 325 semialata MDG and A. semialata KWT were grown in a 4:1 mix of soil and vermiculite under 326 semi-controlled glasshouse conditions at 18-25 ºC with supplemental lightning provided to a 327 minimum of 140-160 µmol m -2 s-1 PFD over a 16-hour photoperiod (further detail in (54)). 328 Measurements were conducted on fully expanded leaves during vegetative stage: at 8-10 weeks 329 for both Flaveria species and G. gynandra, 4-6 weeks for T. hassleriana , and 2 weeks after 330 vegetative propagation for both Alloteropsis species. C4 F. bidentis WT, FbPGRL1-RNAi and 331 FbNdhO-RNAi RNAi plants with expression ~10% of WT (31), were grown in a 3:2 mix of soil and 332 vermiculate, in a growth chamber at 24ºC and 250 µmol m -2 s-1 PFD over a 12-hour photoperiod. 333 Young, fully-expanded FbNdhO-RNAi leaves were measured after 12-16 weeks and of all other 334 plants after 8-10 weeks. 335 336 Chlorophyll fluorescence measurements 337 Chlorophyll fluorescence was measured with a gas exchange system (LI-6400XT, LI-COR, 338 Lincoln, NE, USA) equipped with a leaf chamber fluorometer (6400-40 LCF, LI-COR). Chamber 339 conditions were controlled at 410 or 0 ppm sample CO 2 concentration (the latter with 2% O 2 air), 340 40-60% relative humidity, 25ºC block temperature, 300 µmol s-1 flow rate, and 10% blue (470 nm) 341 and 90% red actinic light (630 nm). The LCF used a 0.25 Hz modulated measuring light and a 342 multiphase flash (55) to measure chlorophyll fluorescence parameters. 343 344 Leaves were dark-adapted until stomatal conductance and net CO2 exchange rate stabilised (30-345 60 min), illuminated with 600 µmol m -2 s-1 PFD for 1 hour, and returned to darkness for 25 346 minutes. Multiphase saturating flashes at 4000 µmol m -2 s-1 PFD were used to measure steady 347 and maximal fluorescence after dark-adaption ( F and F m), and during photoperiod and dark 348 recovery (F’ and Fm’), occurring five minutes before illumination (for F v/Fm); after 3, 5, 10, 15, 25, 349 35, 45, and 60 minutes of light exposure; and 30s after dark transition, then every 90s thereafter. 350 NPQ was derived from fluorescence measurements (56). NPQ relaxation was compared by 351 calculating integrated NPQ from area under the curve (AUC); and 0–2 min and 2–15 min 352 component contributions were expressed as the integrated NPQ within each phase relative to 353 total post-illumination NPQ. 354 355 Chlorophyll fluorescence rise during dark recovery (following (41)) was monitored for 2.5 minutes 356 in darkness after the photoperiod, interspersed by 5s of far-red (FR) light to oxidise the PQ pool 357 (740 nm, ~50 μ mol of photons m−2 s−1) every 20-25s from instrument variation. We compared the 358 PIFR of the dark period after FR, normalised to the final fluorescence value post-oxidation ( F0*, 359 see Fig. S3 ). Representative data from five biological replicates of the subsequent rise in 360 fluorescence is presented. 361 362 Nigericin, DTT, and Antimycin A leaf infiltration 363 Dark adapted leaves were vacuum infiltrated with NPQ inhibitors in a syringe with a buffer (20 364 mM HEPES/KOH pH 7.0) supplemented with either 100 µM nigericin, 5 mM DTT, or 250 µM 365 Antimycin A. Controls were infiltrated with buffer and equivalent volume of solvent without 366 inhibitor. The effect of NPQ inhibitors was estimated by calculating NPQ AUC during the light 367 period and expressing as a proportion of control NPQ AUC (Fig S2 for full NPQ traces). 368 369 Electrochromic Shift measurements 370 The ECS signal was measured as absorptance changes at 515 nm using 535 nm as an 371 isosbestic waveband (Dual-KLAS fitted with a P515/535 module, Heinz Walz GmbH, Effeltrich, 372 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 1, 2025. ; https://doi.org/10.1101/2025.04.01.646649doi: bioRxiv preprint 9 DE) (57). The fore-optics were integrated in a custom measuring chamber (GFS-3000 measuring 373 chamber for DUAL-KLAS, Walz) with temperature controlled at 25 °C. Chamber conditions were 374 controlled via the console of a LI-6800 gas exchange system (LI-COR, USA) at 410 or 0 ppm 375 sample CO2 concentration (the latter with 2% O2 air), 60% relative humidity, and 200 µmol s-1 flow 376 rate. Leaves were dark-adapted, and illuminated with 600 µmol m -2 s-1 PFD for 20 minutes. Dark 377 Interval Relaxation Kinetics (DIRK) measurements of the ECS signal (58) were taken at 1, 3, 5, 378 10, and 20 minutes of illumination. Estimates of thylakoid membrane proton conductivity (gH +) 379 were obtained from the inverse of the decay time constant ( τ ECS) of a single exponential decay 380 fitted to the first 300 milliseconds of the dark interval (59). 381 382 Gas concentration manipulation 383 To suppress photosynthesis and photorespiration, a pre-mixed 2% O 2 and 98% N 2 gas mixture 384 (BOC Ltd., Woking, UK) was supplied to the LI-6400XT or LI-6800 using a mass flow controller 385 (EL-FLOW, Bronkhorst Hight-tech BV, Ruurlo, NL), and CO2 controlled at 0 ppm. 386 387 Statistical analysis 388 One-way ANOVA was conducted on independent experiments comparing each C3 and C4 389 phylogenetic pair, and comparing FbPGRL1-RNAi and FbNdhO-RNAi mutants with WT. 390 Assumptions of normality and homogeneity of variance were tested for, respectively with a 391 Shapiro-Wilk and Bartlett’s test. 392 393 394 395 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 1, 2025. ; https://doi.org/10.1101/2025.04.01.646649doi: bioRxiv preprint 10 Acknowledgments and funding sources 396 397 For providing the original plant material, we thank Dr. Luke Dunning ( Alloteropsis cuttings), Dr. 398 Marjorie Lundgren ( F. cronquistii cuttings), Prof. Peter Westhoff ( F. bidentis seeds) and Prof. 399 Julian Hibberd ( T. hassleriana seeds). Additional thanks to Prof. Hibberd and Prof. David M. 400 Kramer for kindly consulting on our results, and to Dr. Gustaf Degen for his helpful guidance on 401 ECS. 402 403 LAC was jointly funded by the Cambridge Trust; and by Consejo Nacional de Ciencia y 404 Tecnología (CONACyT). This work was supported by the Biotechnology and Biological Sciences 405 Research Council (BBSRC) via grant BB/T007583/1 awarded to JK. For the purpose of open 406 access, the authors have applied a Creative Commons Attribution (CC BY) licence to any Author 407 Accepted Manuscript version arising from this submission. 408 409 Author Contributions 410 411 JK and LAC conceived the study and designed the experiments. AN measured the CEF F. 412 bidentis mutants developed by YM. LAC carried out all other experiments, data analysis and 413 interpretation, and wrote the manuscript. RLV helped with the 2% O 2 experimental setup and 414 provided support with gas exchange experiments. JW and CRGS helped with developing ECS 415 protocols, and JW with chemical infiltration protocols. ELB procured the initial plant material. All 416 authors contributed to and reviewed the final manuscript. Descriptive statistics and figures were 417 created using R 4.1.1 on RStudio 2023.03.1+446. 418 419 Competing Interest Statement 420 421 The authors declare no conflict of interest. 422 423 424 Supporting Information 425 426 Fig. S1: NPQ induction and relaxation in phylogenetic pairs 427 Fig. S2: NPQ induction and relaxation with chemical inhibitors 428 Fig. S3: Sample full trace of Post-Illumination Fluorescence Rise 429 Fig. S4: Carbon assimilation of CEF mutants 430 Table S1: Values and statistics of integrated NPQ across dark recovery in phylogenetic pairs 431 Table S2: Values and statistics of NPQ composition in phylogenetic pairs 432 Table S3: Values and statistics of NPQ with chemical inhibitors in phylogenetic pairs 433 Table S4: Values and statistics of NPQ composition in CEF mutants 434 435 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 1, 2025. ; https://doi.org/10.1101/2025.04.01.646649doi: bioRxiv preprint 11

436 437 1. P. Müller, X. P. Li, K. K. Niyogi, Non-ph oto chemical quenching. A r esponse to e xces s light 438 energy. Plant Physiology 125 , 1558-1566 (2001). 439 2. E. H. Murchie , A. V. Rub an, Dynamic non-photochemical qu enching in plants: fro m 440 molecular mechanism t o product ivity. The Plant Journal 101 , 885-896 (2020). 441 3. C. A. Wraigh t, A . R. Crofts, Ene rgy-depen dent que nching of chlorophyll alpha 442 fluorescence in isola ted chlo roplas ts. European Journal of Biochemistry 17, 319-327 443 (1970). 444 4. B. Demmig-Adams, Caroten oids and pho topro tect ion in plan ts: A rol e for the 445 xanth ophyll zea xant hin. Biochimica et Biophysica Acta (BBA) - Bioenergetics 1020 , 1-24 446 (1990). 447 5. M. Nilkens et al. , Iden tification of a slowl y inducible zeax anthin-d epend ent comp onent 448 of non-photochemical qu enching of chlorophyll fluorescence gen era ted un der st eady-449 stat e conditions in A rabido psis. Biochimica et Biophysica Acta 1797 , 466-475 (201 0). 450 6. A. Malno ë et al. , The Plastid Lipocali n LCNP Is Require d for Sustain ed Photop rot e ctive 451 Energy Dissipation in Arabid opsis. The Plant Cell 30, 196-208 (2017). 452 7. N. Ker en, A. Kri eger-Liszkay, Photoinhibi t ion: molecular m echanisms and physiolo gical 453 significance. Physiologia Plantarum 142 , 1-5 (2011 ). 454 8. S. P. Long et al. , I nto the Shad ows and Ba ck into Sunlight: Photosyn thesis in Fluctu ating 455 Light. Annual Review of Plant Biology 73, 617-648 (2022 ). 456 9. J. Kromdijk et al. , Improving pho tosynthe sis and crop productivity by accele rating 457 recovery from photo pro tectio n. Science 354 , 857-861 (2016 ). 458 10. A. P. De Souza et al. , Soyb ean pho tosynt hesis and crop yield ar e improved by 459 accelera ting recove ry from photopr otec t ion. Science 377 , 851-854 (2022). 460 11. FAO, "Pr oduction , Trade an d Prices of Commodities" in W orld Food and Agricult ur e - 461 Statis tical Yearb ook 2020. (FAO, Rom e, I t aly, 2020), pp. 10-13. 462 12. L. Guidi, E. Lo Piccolo, M. Landi, Chlor oph yll Fluorescence, Pho toinhibi tion and Ab i otic 463 Stress: Does i t Make Any Difference t he Fact to B e a C3 or C4 Species? Frontiers in Plant 464 Science 10, 174 (2019). 465 13. R. F. Sage, Th e evolutio n of C4 photosynt hesis. New Phytologist 161 , 341-370 (200 4). 466 14. J. A. Langd ale, C4 cycles: past, pr esen t, a nd future r esearch o n C4 photosynthesis . Plant 467 Cell 23, 3879-3892 (2011 ). 468 15. R. T. Furbank, Evoluti on of the C4 phot osynthetic mechanism: are ther e re ally thr ee C4 469 acid decarbo xylatio n types? Journal of Experimental Botany 62, 3103-3108 (2011). 470 16. X. Yin, P. C. Struik, The ene rgy budget in C4 photosynthesis: insights from a cell-ty pe-471 specific electron transp ort mod el. New Phytologist 218 , 986-998 (2018). 472 17. Y. N. Mun ekage, Y. Y. Taniguchi, Promoti on of Cyclic Electron Transpo rt Ar ound 473 Photosystem I with th e Developmen t of C4 Photosynthesis. Plant and Cell Physiology 57, 474 897-903 (2016 ). 475 18. X. Yin, P. C. Struik, Exploiti ng differences in the en ergy budget among C 4 subtypes to 476 improve crop produc tivity. New Phytologist 229 , 2400-2409 (2021 ). 477 19. J. F. All en, Cyclic, pseudocyclic and noncyclic photophospho rylation : new links in t he 478 chain. Trends in Plant Science 8 , 15-19 (2003). 479 20. T. Shikanai, Cent ral rol e of cyclic electron transpor t ar ound pho tosystem I in th e 480 regulati on of photosynt hesis. Current Opinion in Biotechnology 26, 25-30 (2014). 481 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 1, 2025. ; https://doi.org/10.1101/2025.04.01.646649doi: bioRxiv preprint 12 21. W. Majer an et al. , S tructu ral and me tabo lic transitio ns of C4 leaf development an d 482 differentia tion d efined by microscopy and quantit ative pr ot eomics in maize. Plant Cell 483 22, 3509-3542 (2010 ). 484 22. A. Takabayashi, M . Kishine, K. Asad a, T. E ndo, F. Sa to, Differenti al use of two cyclic 485 electr on flows around pho tosystem I for driving CO 2 -concentratio n mechanism in C 4 486 photosynth esis. Proceedings of the National Academy of Sciences 102 , 16898-16903 487 (2005). 488 23. J. V. Sagun, M . R. Badg er, W . S. Chow, O. Ghannoum, Cyclic elect ron flow and light 489 parti tioning be tween the two pho tosyste ms in leaves of plants with different func tional 490 types. Photosynthesis Research 142 , 321- 334 (2019). 491 24. Y. Munekage et al., PG R5 is involved in cyclic electron flow around pho tosystem I and is 492 essential for ph otop rot ectio n in Arabid o psis. Cell 110 , 361-371 (2002 ). 493 25. H. Nakano , H. Yamamot o, T. Shikanai , Contribu tion of NDH-depend ent cyclic elect ron 494 transpo rt a round pho tosystem I to th e genera tion of pro ton mo tive force in the weak 495 mutant all ele of pgr5. Biochimica et Biophysica Acta (BBA) - Bioenergetics 1860 , 369-374 496 (2019). 497 26. Y. Munekage et al., Cyclic electr on flow around pho tosystem I is essen tial for 498 photosynth esis. Nature 429 , 579-582 (20 04). 499 27. M. Suorsa et al. , PROTON GRADIE NT REG ULATION5 is esse ntial for p rope r acclimation of 500 Arabidopsis ph otosyst em I to na turally a nd artificially fluctuating light condi tions . Plant 501 Cell 24, 2934-2948 (2012 ). 502 28. W. Yamori, A . Makino , T. Shikanai, A phy siological role of cyclic electron transp ort 503 around pho tosystem I in sust aining phot osynthesis under fluctu ating light in ric e. 504 Scientific Reports 6 , 20147 (2016). 505 29. R. B. Pet erson, N. P. Schult es, N . A. McHa le, I. Zeli tch, Evidence for a Rol e for NAD(P)H 506 Dehydrogenase in Concen tra tion of CO2 in the Bundl e Shea th Cell of Zea mays. Plant 507 Physiology 171 , 125-138 (2016). 508 30. M. Ermakova et al., Chloropl ast N ADH de hydrogenase-like comple x-media ted cyclic 509 electr on flow is the main elec tron transp ort rou te in C4 bundle she ath cells . New 510 Phytologist 243 , 2187-2200 (2024). 511 31. T. Ogawa et al., Two cyclic electron flows around pho tosystem I differ entially pa rti cipate 512 in C4 photosynthesis. Plant Physiology 191 , 2288-2300 (2023 ). 513 32. A. Kanazawa , D. M. Krame r, In vivo modu lation of nonpho tochemical excit on quen ching 514 (NPQ) by regulation of the chloropl ast AT P synthase. Proceedings of the National 515 Academy of Sciences 99, 12789-12794 (2002). 516 33. O. Kiira ts, D. M. Krame r, G . E. Edwards, Co-regulatio n of dark and light r eactions i n thre e 517 biochemical subtypes of C4 species. Photosynthesis Research 105 , 89-99 (2010). 518 34. S. H. Taylor et al. , Ecophysiological t raits in C 3 and C 4 grasses: a phylogenetical ly 519 controll ed screeni ng exp erimen t. New Phytologist 185 , 780-791 (2010). 520 35. P.-A. Christin, C. P. Osbor ne, R . F. Sage, M. Arakaki, E . J. Edwards , C4 eudicots ar e not 521 younger than C4 monocots . Journal of Experimental Botany 62, 3171-3181 (2011). 522 36. M. R. Lundgren et al. , Photosynth etic inn ovation bro adens th e niche within a singl e 523 species. Ecology Letters 18, 1021-1029 (2015). 524 37. O. Uen o, N . Sent oku, Comparison of leaf structur e and pho tosynth etic char acte ris tics of 525 C3 and C4 Alloteropsis semialata subspe cies. Plant, Cell & Environment 29, 257-2 68 526 (2006). 527 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 1, 2025. ; https://doi.org/10.1101/2025.04.01.646649doi: bioRxiv preprint 13 38. U. Gowik, A . Bräu tigam, K. L. Web er, A. P. M. We ber , P. Westh off, Evolution of C4 528 Photosynthesis in th e G enus Flaveria: Ho w Many and Which Gen es Does It Take t o 529 Make C4? The Plant Cell 23, 2087-2105 (2011). 530 39. A. Brä utigam et al. , An mR NA Blu eprin t for C4 Photosynthesis Derived from Comparative 531 Transcriptomics of Closely Relat ed C3 and C4 Species. Plant Physiology 155 , 142-156 532 (2010). 533 40. Y. Wang, S. J . Burgess, E . M. Becke r, S. P. Long, Photosynthesis in th e fleeting shad ows: 534 an overlooked o ppor tunity for incre asing crop productivity? The Plant Journal 101 , 874-535 884 (2020). 536 41. D. D. Strand, N . Fishe r, D. M. Krame r, Th e higher plan t plastid NAD(P)H dehydrogenase-537 like complex (NDH) is a high efficiency pr oton pump th at incr eases ATP producti o n by 538 cyclic electron flow. Journal of Biological Chemistry 292 , 11850-11860 (2017). 539 42. M. R. Lundgren et al. , Evolution ary implications of C3-C4 intermediat es in the gr as s 540 Alloteropsis semialata . Plant, Cell & Environment 39, 1874-1885 (2016). 541 43. L. Arce Cubas et al., Lessons from rela tiv es: C 4 photosynthesis en hances CO 2 assi milation 542 during the low-light phas e of fluctuation s. Plant Physiology 193 , 1073-1090 (2023). 543 44. M. S. Lee , R. A . Boyd, D. R. O rt, The ph ot osynthetic r esponse of C3 and C4 bioene rgy 544 grass species to fluctua ting light. GCB Bioenergy 14, 37-53 (2022). 545 45. R. Bassi, L. Dall' Osto, Dissipat ion of Light Energy Absorbed in Exc ess: The Mol ecul ar 546 Mechanisms. Annual Review of Plant Biology 72, 47-76 (2021). 547 46. X. P. Li, P. Muller-Moule, A. M. Gilmor e, K. K. Niyogi, PsbS-dependen t enhanc eme nt of 548 feedback de-exci tat ion pro tects ph otosy stem II from pho toinhibi tion . Proceedings of the 549 National Academy of Sciences 99, 15222-15227 (2002). 550 47. M. Krishnan-Schmiede n, P. E. Konold, J . T. M. Kennis, A . Pandit, Th e molecular pH-551 response mech anism of the plan t light-st ress sensor PsbS. Nature Communications 12, 552 2291 (2021). 553 48. G. E. Degen et al., High cyclic elect ron tr a nsfer via the PGR5 pa thway in the abse n ce of 554 photosynth etic con trol. Plant Physiol 192 , 370-386 (2023 ). 555 49. M. Uflewski et al. , The thylakoid pro ton a ntipor ter KEA3 r egulat es photosyn thesis in 556 response to th e chloropl ast en ergy statu s. Nature Communications 15, 2792 (202 4). 557 50. E. Romanowska et al., Differences in pho tosynthe tic respons es of NADP-ME type C4 558 species to high light. Planta 245 , 641-657 (2017). 559 51. N. Nakamu ra, M . Iwano, M . Havaux , A. Y okota, Y. N . Mun ekage, Promoti on of cyclic 560 electr on tr anspor t arou nd phot osystem I during the evolu tion of NADP-malic enzyme-561 type C 4 photosynthesis in th e genus Flaveria . New Phytologist 199 , 832-842 (2013). 562 52. K. Głowacka et al., Photosyst em II Subu ni t S overe xpressio n increas es the efficienc y of 563 water use in a field-grown crop . Nature Communications 9 , 868 (2018). 564 53. S. Hubbar t et al., Enhanc ed thylakoid ph otopr otec tion can incr ease yield and can opy 565 radiati on use efficiency in rice. Communications Biology 1 , 22 (2018). 566 54. L. Arce Cubas et al., Activa tion of CO 2 assimilation during pho tosynth etic induct io n is 567 slower in C 4 than in C 3 photosynthesis in thre e phylogene tically contr olled e xpe ri ments. 568 Frontiers in Plant Science 13 (2023). 569 55. S. D. Loriaux et al. , Closing in on maximum yield of chlorophyll fluorescence using a 570 single multiphase flash of sub-satur ating intensity . Plant, Cell & Environment 36, 1 755-571 1770 (2013). 572 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 1, 2025. ; https://doi.org/10.1101/2025.04.01.646649doi: bioRxiv preprint 14 56. B. Ge nty, J .-M. Bri antais , N. R . Bake r, The relati onship be tween th e quan tum yield of 573 photosynth etic el ectr on tra nspor t and q uenching of chlorophyll fluorescenc e. 574 Biochimica et Biophysica Acta (BBA) - General Subjects 990 , 87-92 (1989). 575 57. C. Klughammer, K. Siebke, U . Schreib er, Continuous ECS-indicated r ecording of th e 576 proton-mo tive charge flux in le aves. Photosynthesis Research 117 , 471-487 (2013) . 577 58. C. A. Sacksted er, D. M. Kr amer, Dark-int e rval relax atio n kinetics (DIRK) of absorbance 578 changes as a quanti ta tive probe of stead y-state el ectron transfe r. Photosynthesis 579 Research 66, 145-158 (2000). 580 59. N. R. B aker, J. Har binson, D. M. Kr amer, Determining th e limita tions and r egulati on of 581 photosynth etic en ergy transd uction in le aves. Plant, Cell & Environment 30, 1107- 1125 582 (2007). 583 584 585 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 1, 2025. ; https://doi.org/10.1101/2025.04.01.646649doi: bioRxiv preprint 15 586 587 .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 1, 2025. ; https://doi.org/10.1101/2025.04.01.646649doi: bioRxiv preprint Figures Fig. 1: Differences in NPQ relaxation in phylogenetically linked C3 and C4 species (n=5). A) NPQ relaxation after 1h of illumination at 600 µmol m -2 s-2 PFD preceded by 5m of darkness for Fv/Fm. Ribbons represent standard error of the mean. B) NPQ composition based on time relaxation kinetics as a percentage of total NPQ. Asterisks indicate significant differences between C3 and C4 species found by one-way ANOVA (* P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001). .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 1, 2025. ; https://doi.org/10.1101/2025.04.01.646649doi: bioRxiv preprint Fig. 2: Effect of Δ pH disruption and xanthophyll cycle inhibition on C3 and C4 NPQ. AUC of NPQ induction of leaves infiltrated with A) 100 µM nigericin to collapse the proton gradient and B) 5 mM dithiothreitol to inhibit the xanthophyll cycle, as a percentage of a control (n=5). Asterisks indicate significant differences between C3 and C4 species found by one-way ANOVA (* P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001). .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 1, 2025. ; https://doi.org/10.1101/2025.04.01.646649doi: bioRxiv preprint Fig. 3: Differences in NPQ relaxation in phylogenetically linked C3 and C4 species when suppressing photosynthesis and photorespiration with 2% O 2 and 0 ppm CO 2 air (n=5). A) NPQ relaxation after 1h of illumination at 600 µmol m -2 s-2 PFD preceded by 5m of darkness for Fv/Fm. Ribbons represent standard error of the mean. B) NPQ composition based on time relaxation kinetics as a percentage of total NPQ. Asterisks indicate significant differences between C3 and C4 species found by one-way ANOVA (* P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001). .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 1, 2025. ; https://doi.org/10.1101/2025.04.01.646649doi: bioRxiv preprint Fig. 4: Proton conductivity of the thylakoid membrane (gH +) in C3 and C4 phylogenetically linked species, under ambient and 2% O 2 air and 0 ppm CO 2 (n=5), during 20 minutes of illumination at 600 µmol m -2 s-2 PFD. Asterisks indicate significant differences between species at a given timepoint found by one-way ANOVA (n=5, * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001). .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 1, 2025. ; https://doi.org/10.1101/2025.04.01.646649doi: bioRxiv preprint Fig. 5: CEF in C3 and C4 phylogenetically linked species. A) Representative trace (n=5) of post- illumination fluorescence rise (PIFR) in leaves infiltrated with a control buffer (solid lines) or with Antimycin A (dashed lines), an inhibitor of the CEF PGR5 pathway. B) NPQ relaxation in C4 F. bidentis WT and FbNdhO and FbPGRL1 RNAi knockdown mutants. Ribbons represent standard error of the mean (n=3). C) NPQ composition based on time relaxation kinetics as a percentage of total NPQ. Asterisks indicate significant differences between each mutant and the WT found by one-way ANOVA (n=3, * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001). .CC-BY 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted April 1, 2025. ; https://doi.org/10.1101/2025.04.01.646649doi: bioRxiv preprint