Overexpression of the RieskeFeS Protein Increases Electron Transport Rates and Biomass Yield1[CC-BY]

Overexpression of the Rieske FeS protein results in significant increases in the quantum efficiencies of PSI and PSII, increases in Amax, and has the potential to increase crop productivity. In this study, we generated transgenic Arabidopsis (Arabidopsis thaliana) plants overexpressing the Rieske FeS protein (PetC), a component of the cytochrome b6f (cyt b6f) complex. Increasing the levels of this protein resulted in concomitant increases in the levels of cyt f (PetA) and cyt b6 (PetB), core proteins of the cyt b6f complex. Interestingly, an increase in the levels of proteins in both the photosystem I (PSI) and PSII complexes also was seen in the Rieske FeS overexpression plants. Although the mechanisms leading to these changes remain to be identified, the transgenic plants presented here provide novel tools to explore this. Importantly, overexpression of the Rieske FeS protein resulted in substantial and significant impacts on the quantum efficiency of PSI and PSII, electron transport, biomass, and seed yield in Arabidopsis plants. These results demonstrate the potential for manipulating electron transport processes to increase crop productivity.

Increasing food and fuel demands by the growing world population has led to the need to develop higher yielding crop varieties (Fischer and Edmeades, 2010;Ray et al., 2012). Transgenic studies, modeling approaches, and theoretical considerations provide evidence that increasing photosynthetic capacity is a viable route to increase the yield of crop plants (Zhu et al., 2010;Raines, 2011;Long et al., 2015;von Caemmerer and Furbank, 2016). There is now a growing body of experimental evidence showing that increasing the levels of photosynthetic enzymes in carbon metabolism results in increased photosynthesis and plant biomass (Miyagawa et al., 2001;Lefebvre et al., 2005;Raines, 2006Raines, , 2011Rosenthal et al., 2011;Uematsu et al., 2012;Simkin et al., 2015Simkin et al., , 2017Driever et al., 2017). In addition, the manipulation of photosynthetic electron transport by the introduction of the algal cytochrome c 6 protein has been shown to improve the efficiency of photosynthesis and to stimulate plant growth in low light (Chida et al., 2007).
One endogenous target identified for manipulation is the cytochrome b 6 f (cyt b 6 f) complex, which is located in the thylakoid membrane and functions in both linear and cyclic electron transport, providing ATP and NADPH for photosynthetic carbon fixation. Initially, cyt b 6 f inhibitors (Kirchhoff et al., 2000), and later, transgenic antisense studies suppressing the accumulation of the Rieske FeS protein (PetC), a component of the cyt b 6 f complex, demonstrated that the activity of the cyt b 6 f complex is a key determinant of the rate of electron transport (Price et al., 1995(Price et al., , 1998Anderson et al., 1997;Ruuska et al., 2000;Yamori et al., 2011aYamori et al., , 2011b. The finding that the cyt b 6 f complex is a potential limiting step in the electron transport chain suggests that, by increasing the activity of this complex, it may be possible to increase the rate of photosynthesis. However, questions have been raised about the feasibility of manipulating this multiprotein, membrane-located complex, given that it is composed of eight different subunits, six being encoded in the chloroplast genome (PetA [cyt f], PetB [cyt b 6 ], PetD, PetG, PetL, and PetN) and two in the nucleus (PetC [Rieske FeS] and PetM; Willey and Gray, 1988;Anderson, 1992;Knight et al., 2002;Baniulis et al., 2009;Schöttler et al., 2015). Furthermore, this protein complex functions as a dimer, with the transmembrane domains of both the Rieske FeS and cyt b 6 proteins being implicated directly in the monomer-monomer interaction and stability of the complex and the petD gene product functioning as a scaffold (Hager et al., 1999;Schwenkert et al., 2007;Hojka et al., 2014). Essential roles in the assembly and stability of the cyt b 6 f complex also have been shown for the PetG, PetN, and PetM subunits, and a minor role in stability was assigned to the PetL gene product (Bruce and Malkin, 1991;Kuras and Wollman, 1994;Hager et al., 1999;Monde et al., 2000;Schöttler et al., 2007;Schwenkert et al., 2007;Hojka et al., 2014).
Notwithstanding both the genetic and structural complexity of the cyt b 6 f complex, it has been shown previously that it is possible to manipulate the levels of the cyt b 6 f complex by down-regulation of the expression of the Rieske FeS protein (Price et al., 1998;Yamori et al., 2011b). It also has been shown that the Rieske FeS protein is one of the subunits required for the successful assembly of the cyt b 6 f complex (Miles, 1982;Metz et al., 1983;Barkan et al., 1986;Anderson et al., 1997). Based on these results, we reasoned that overexpression of the Rieske FeS protein could be a feasible approach to take in order to increase the electron flow through the cyt b 6 f complex. In this article, we report on the production of Arabidopsis (Arabidopsis thaliana) with increased levels of the tobacco (Nicotiana tabacum) Rieske FeS protein, and we show that this manipulation resulted in increases in photosynthetic electron transport, CO 2 assimilation, and yield. This work provides evidence that the process of electron transport is a potential route for the improvement of plant productivity.

Production and Selection of Rieske FeS Overexpression Transformants
The full-length tobacco Rieske FeS coding sequence (X64353) from the cyt b 6 f complex was used to generate an overexpression construct B2-NtRi (Supplemental Fig. S1). Following floral dipping, transgenic Arabidopsis plants were selected on both kanamycin-and hygromycin-containing medium (Nakagawa et al., 2007), and plants expressing the integrated transgenes were identified using reverse transcriptase-PCR (data not shown). Proteins were extracted from leaves of the T1 progeny, allowing the identification of three lines with increased levels of the Rieske FeS protein (PetC; Supplemental Fig. S2A). Immunoblot analysis of T3 progeny of lines 9, 10, and 11 showed them to have higher levels of the Rieske FeS protein when compared with the wild type ( Fig. 1; Supplemental Fig. S2B). The overexpression of the Rieske FeS protein (hereafter referred to as Rieske FeS ox) resulted in a concomitant Figure 1. Immunoblot analysis of leaf proteins of wild-type (WT) and Rieske FeS ox plants. Protein extracts were from leaf discs taken from two leaves per plant from three independent lines (9, 10, and 11) and separated on a 12% acrylamide gel, transferred to membranes, and probed with relevant antibodies. The cyt b 6 f complex subunits PSI and PSII, ATP synthase d-subunit (AtpD), Calvin-Benson cycle (CBC) proteins, and photorespiratory GDC-H protein were probed. A, Protein (6 mg) was loaded for all antibodies except for FBPA (3 mg) and Rubisco (1 mg). B, Proteins were loaded containing 0.63 to 10 mg of protein. C, Proteins were quantified using a Fusion FX Vilber Lourmat imager (Peqlab), as described previously (Vialet-Chabrand et al., 2017), and are presented as relative protein contents compared with the wild type.
increase in both cyt f (PetA) and cyt b 6 (PetB; Fig. 1A). An increase in the level of the PSI type I chlorophyll a/bbinding protein (LhcaI) and an increase in the core protein of PSI (PsaA) also were observed. Furthermore, the D1 (PsbA) and D2 (PsbD) proteins, which form the reaction center of PSII, also were shown to be elevated in Rieske FeS ox lines. Finally, an increase in the ATP synthase d-subunit (AtpD) also was observed in Rieske FeS ox lines (Fig. 1A). In contrast, no notable differences in protein levels for the chloroplastic FBP aldolase (FBPA), the mitochondrial Gly decarboxylase-H protein (GDC-H), or the Rubisco large subunit were observed (Fig. 1A). A quantitative estimate of the changes in protein levels was determined from the immunoblots of leaf extracts isolated from two to three independent plants per line. An example is shown in Figure 1B. These results showed a 2-to 2.5-fold increase in the Rieske FeS protein relative to wild-type plants, and a similar increase also was observed for cyt f, cyt b 6 , Lhca1, D2, and PsaA (Fig. 1C). No increase in the stromal FBPA protein was evident.

Chlorophyll Fluorescence Imaging Reveals Increased Photosynthetic Efficiency in Young Rieske FeS ox Seedlings
In order to explore the impact of increased levels of the Rieske FeS protein on photosynthesis, the quantum efficiency of PSII (F q 9/F m 9) was analyzed using chlorophyll a fluorescence imaging (Baker, 2008;Murchie and Lawson, 2013). A small increase in F q 9/F m 9 was found in the Rieske FeS ox plants at irradiances of 310 and 600 mmol m 22 s 21 (Fig. 2). Leaf area, generated from these images, was significantly larger in all Rieske FeS ox lines compared with the wild type ( Fig. 2C), but no significant difference in leaf thickness was observed between the leaves of Rieske FeS lines 9 and 11 and the wild-type plants (Supplemental Table S1).

Photosynthetic CO 2 Assimilation and Electron Transport Rates Are Increased in the Rieske FeS ox Plants
The impact of overexpression of the Rieske FeS protein on the rate of photosynthesis in mature plants was investigated using combined gas-exchange and chlorophyll fluorescence analyses. Both the light-saturated rate of CO 2 fixation and the relative light-saturated rate of electron transport (ETRII) were increased in the Rieske FeS ox lines compared with the wild type when measured at 2% [O 2 ] (Fig. 3, A and B; Table I). Additionally, the light-saturated rate of CO 2 assimilation at ambient [CO 2 ] also was increased when measured at 21% [O 2 ] (Supplemental Fig. S3). No significant difference in leaf absorbance between the Rieske FeS ox and wild-type plants was found (Supplemental Table S1).
In plants grown at a light level of 130 mmol m 22 s 21 , no difference in the light-or CO 2 -saturated rate of CO 2 assimilation (A max ) was found. In contrast, in a second group of plants grown at 280 mmol m 22 s 21 , A max was greater in the Rieske FeS ox lines 9 and 11 relative to the wild type ( Fig. 3C; Table I). Further analysis of the A/C i curves revealed that maximum electron transport rate was significantly greater in the Rieske FeS ox plants when compared with the wild type (Table II), but no significant difference in the maximum rate of Rubisco (data not shown) was observed.

The Quantum Efficiency of PSI and PSII Was Increased in the Rieske FeS ox Plants
To further explore the influence of increases in the Rieske FeS protein on PSII and PSI photochemistry, dark-light induction responses were determined in the wild type and Rieske FeS ox (lines 11 and 10) using simultaneous measurements of P700 oxidation state and PSII efficiency. These results showed that the quantum yields of both PSI and PSII were increased in the Rieske FeS ox plants compared with the wild type and that the fraction of PSII centers that were open (q L ) also was increased, while the level of Qa reduction (1 2 q p ) was lower in leaves of 27-d-old plants from line 11 (Fig. 4). non-photochemical quenching (NPQ) levels also were shown to be lower in the Rieske FeS ox plants, together with a reduction in stress-induced limitation of NPQ (q N ) when compared with wild-type plants (Fig. 4). Similar results were obtained for both lines 10 and 11 when plants were analyzed later in development (34 DAP; Supplemental Fig. S4). The increases in the quantum yields of PSI and PSII observed here were accompanied by corresponding increases in electron transport rates (ETRI and ETRII; Supplemental Fig. S5, A and D); however, a bigger increase in ETRII relative to ETRI was observed, demonstrated by the wild type having a higher ETRI-ETRII ratio (Supplemental Fig.  S5, C and D).

Growth, Vegetative Biomass, and Seed Yield Are Increased in the Rieske FeS ox Plants
The leaf area of the Rieske FeS ox lines was significantly greater than that of the wild type as early as 10 DAP in soil, and by 18 d it was 40% to 114% larger (Fig. 5). Destructive harvest at day 25 showed that this increase in leaf area translated to an increase in shoot biomass of between 29% and 72% determined as dry weight (Fig. 5C). To determine the impact of increased Rieske FeS protein on seed yield and final shoot biomass, a second group of plants was grown in the same conditions as described in Figure 6. Interestingly, at 38 DAP, 40% of the Rieske FeS ox plants had flowered, in contrast to 22% in the wild-type plants (Fig. 6A). Following seed set (52 DAP), both vegetative biomass ( Fig. 6B) and seed yield ( Fig. 6C) were determined, and although a significant increase in biomass was observed in all of the Rieske FeS ox plants, a statistically significant increase in seed yield was evident only in line 11.

Pigment Content Was Altered in the Rieske FeS ox Plants
The pigment composition of the leaves of the Rieske FeS ox and wild-type plants was determined. Increases in the levels of chlorophyll a and b (14%-29%) were observed in the Rieske FeS ox compared with wild-type plants (Fig. 7). Plants with increased Rieske FeS protein also were found to have a small increase in the chlorophyll a/b ratio, from 2.96 to 3.12. These increases were accompanied by increases in the carotenoids, neoxanthin (+38%), violaxanthin (+59%), lutein (+75%), and b-carotene (+169%). No detectable change in the level of zeaxanthin was evident in the Rieske FeS ox plants. The increases in both total chlorophyll and total carotenoids also led to a significant decrease in the chlorophyll-to-carotenoid ratio, from 3.6 to 2.4 (Fig. 7).

DISCUSSION
In recent years, increasing the rate of photosynthetic carbon assimilation has been identified as a target for improvement to increase yield. Evidence to support this has come from the theory and modeling of the photosynthetic process, growth of plants in elevated CO 2 , and transgenic manipulation (Zhu et al., 2010). It was shown previously in antisense studies that reducing the levels of the Rieske FeS protein resulted in a reduction in levels of the cyt b 6 f complex, a decrease in photosynthetic electron transport, and, in rice (Oryza sativa), a decrease in both biomass and seed yield (Price et al., 1998;Yamori et al., 2016). These findings identified the cyt b 6 f complex as a limiting step in electron transport and would suggest that overexpression of the Rieske FeS protein may be a feasible route to increase photosynthesis and yield. In this study, we show that overexpression of the Rieske FeS protein in Arabidopsis results in increases in photosynthesis, vegetative biomass, and seed yield.
Increased Levels of the Rieske FeS Protein Increased Photosynthetic Electron Transport, CO 2 Assimilation, and Biomass Using chlorophyll fluorescence imaging, we have shown that overexpression of the Rieske FeS protein resulted in increases in photosynthesis and growth, which are evident from the early stages of seedling development. These observed increases in F q 9/F m 9 represent an early indication that the potential quantum yield of PSII photochemistry had been elevated in Rieske FeS ox lines (Genty et al., 1989(Genty et al., , 1992Baker et al., 2007). This early stimulation is maintained into maturity, and increases in the light-saturated rate of CO 2 assimilation and electron transport rates were evident in the Rieske FeS ox plants.
Our data also showed that quantum yields of both PSII and PSI were increased and that the fraction of PSII centers available for photochemistry was increased, as indicated by an increase in q L and a lower 1 2 q p (Baker and Oxborough, 2004;Kramer et al., 2004;Baker et al., 2007). These results are consistent with what would be predicted from results obtained from the Rieske FeS antisense studies where ETR was reduced (Price et al., 1998;Ruuska et al., 2000;Yamori et al., 2011b). However, the impact of overexpression of the Rieske FeS protein was clearly not restricted to increasing the activity of the cyt b 6 /f complex but resulted in changes in PSI, PSII, and the ATPase complex, which led to an increase in electron flow through the entire electron transport chain. Analysis of the ratio of PSI to PSII electron transport showed that, relative to wild-type plants, the Rieske ox lines have higher rates of PSII ETR relative to PSI, which may be related to cyclic electron flow, differences in energy distribution, and/or the PSI-PSII ratio (Kono et al., 2014).
In addition to increased rates of photosynthesis, a substantial and significant increase in the growth of the  Fig. 3

Name Parameters
Light  Table II. Maximum electron transport rate (J max ) and maximum assimilation (A max ) in wild-type and Rieske FeS ox lines Results were derived from the A/C i response curves shown in Figure  3C  Wild type 181 6 11.6 24.4 6 2.31 9 210 6 10.2* 31.1 6 1.14** 10 194 6 8.9 28.3 6 0.67 11 216 6 11.9** 31.9 6 0.68** rosette area was observed in the Rieske FeS ox plants in the early vegetative phase, which resulted in an approximately 30% to 70% increase in biomass yield in the different lines. Importantly, seed yields in line 11, which showed the biggest increases in shoot biomass, also were shown to be increased relative to the wild type. Given that increases in leaf area are evident in the Rieske ox plants from early in development, the improvements in biomass are likely due to a combination of increased light capture and photosynthesis due to the greater leaf area.

Rieske FeS ox Plants Have Increased Levels of Proteins in the cyt b 6 f, PSI, and PSII Complexes
In keeping with our analysis of the electron transport processes in the Rieske FeS ox plants, increases in the levels of cyt b 6 and cyt f, core proteins of the cyt b 6 f complex, were evident. Furthermore, increases in the levels of proteins in both PSII and PS1 and the d-subunit of the ATPase complex also were observed. This result was unexpected, given that no changes in the components of PSII or PSI were observed in the Rieske FeS antisense plants. In contrast, in keeping with this study, the hcf mutant, in which the biogenesis of the cyt b 6 f complex was inhibited, also had a reduced accumulation of components of both PSI and PSII (Lennartz et al., 2001). In addition, a recent study reported increases in cyt b 6 f protein levels in Arabidopsis plants grown under square-wave light compared with plants grown under fluctuating light, and these were accompanied by increased levels of PSII, PSI, and the d-subunit of the ATPase proteins (Vialet-Chabrand et al., 2017). Taken together, these results provide further demonstration of the ability of the thylakoid membrane to respond to changes in the external and internal environments (Foyer et al., 2012). Although there are many examples of coordination of the accumulation of photosynthetic complexes in the thylakoid membrane, the underlying regulation of the synthesis and assembly of components of the thylakoid membrane, the factors determining the accumulation of these complexes, is still poorly understood (Schöttler et al., 2015). Recently, a nucleus-encoded chloroplast RNA-binding protein, photosystem biogenesis regulator1 (PBR1), was identified and shown to be involved in the coordination of the biogenesis of the PSI, cyt b 6 f, and NDH complexes (Yang et al., 2016). It will be interesting to explore the role of PBR1 in the plants produced in this study, as these are likely useful tools to investigate fundamental questions on factors controlling both the biogenesis and activities of the photosynthetic complexes in the electron transport chain.
A recent study has shown that, for rice plants growing in elevated CO 2 , a small decrease in Rubisco protein levels resulted in an increase in the electron transport components and an increase in biomass (Kanno et al., 2017). Given the increase in thylakoid protein composition of the Rieske ox plants, it might be expected that there would be some tradeoff, and a reduction in other proteins may occur to compensate. Although no comprehensive analysis of proteins has been undertaken in these plants, we found no evidence of changes in the levels of Rubisco, which is a major N sink or FBPA and GDC. What is not yet clear is the impact of this manipulation on nitrogen use efficiency (NUE), and a full analysis of the performance of these plants under different N regimes will be needed to explore this question.

Overexpression of Rieske FeS Significantly Modifies the Pigment Content of Leaves
In parallel with the increase in components of the thylakoid membranes, plants with increased Rieske FeS protein were found to have increases in the levels of both chlorophyll a and b and a small increase in the chlorophyll a/b ratio, from 2.96 to 3.12. The increases in chlorophyll a and b suggest a greater investment in both light capture and PSII reaction centers and would fit with the increase in photosynthetic electron transport capacity in the Rieske FeS ox plants. In previous work, plants with reduced levels of the Rieske FeS protein had a lower chlorophyll a/b ratio (Hurry et al., 1996;Price et al., 1998), which is the opposite of what was observed in the Rieske FeS ox plants. In addition to increases in chlorophyll, significant increases in the carotenoid pigments also were seen with b-carotene (+169%), violaxanthin (+59%), lutein (+75%), and neoxanthin (+37%). b-Carotene is a component of both the reaction centers (RC) and lightharvesting complex (Kamiya and Shen, 2003;Ferreira et al., 2004;Loll et al., 2005;Litvin et al., 2008;Janik et al., 2016), and the increase in these pigments observed in the Rieske FeS ox plants is in agreement with the investment in both light harvesting and increasing RC efficiency. We also observed a significant decrease in the chlorophyllto-carotenoid ratio in Rieske FeS ox plants, which would suggest that there is a greater investment in photoprotection relative to the increased chlorophyll for light harvesting. This increase in carotenoids relative to chlorophyll may contribute to the lower NPQ observed in Rieske FeS ox plants (Kaňa et al., 2016).
Lutein, neoxanthin, and violaxanthin are the main xanthophyll pigment constituents of the largest lightharvesting pigment-protein complex of photosystem II . Two leaf discs, collected from two different leaves, were immersed in N,N-dimethylformamide (DMF) at 4°C for 48 h and separated by ultra-performance liquid chromatography. Results are represented as mg g 21 fresh weight (FW). Statistical differences are shown by asterisks (*, P , 0.1; **, P , 0.05; and ***, P , 0.001). Error bars represent SE. (Thayer and Björkman, 1992;Ruban et al., 1994Ruban et al., , 1996Ruban et al., , 1999Ruban, 2012;Janik et al., 2016). Acidification of the thylakoid lumen as a result of electron transport (and driven in particular by the activities of the cyt b 6 f complex) is accompanied by the deepoxidation of violaxanthin and an accumulation of zeaxanthin (Björkman and Demmig-Adams, 1994;Müller et al., 2001;Ruban, 2012) as well as protonation of carboxylic acid residues of the PsbS protein associated with PSII antennae (Li et al., 2000(Li et al., , 2004. Protonation of PsbS and binding of zeaxanthin increase NPQ and the thermal dissipation of excitation energy (Baker, 2008;Jahns and Holzwarth, 2012). However, we found that the increase in the level of the Rieske FeS protein led to small but significantly lower steady-state levels of NPQ. The absence of an increase in NPQ in the presence of significant increases in electron transport rates suggests that the Rieske FeS ox plants also have increased rates of ATP synthesis. Although we provide no direct support for this, we did observe an increase in the level of the ATP synthase d-subunit protein in the Rieske FeS ox plants. Further support for this suggestion comes from earlier work on Rieske FeS antisense plants, which showed that the level of ATP synthase and the transthylakoid pH gradient were both reduced (Price et al., 1995(Price et al., , 1998Ruuska et al., 2000).

CONCLUSION
A number of studies have shown that increasing photosynthesis through the manipulation of CO 2 assimilation can improve growth (Miyagawa et al., 2001;Lefebvre et al., 2005;Rosenthal et al., 2011;Uematsu et al., 2012;Simkin et al., 2015Simkin et al., , 2017. This work, together with a study in which cytochrome c 6 from the red alga Porphyra was expressed in Arabidopsis (Chida et al., 2007), provides direct evidence that there is also an opportunity to improve the efficiency of the electron transfer chain. Here, we have shown not only that overexpression of the Rieske FeS protein resulted in increases in cyt f and cyt b 6 proteins of the cyt b 6 f complex but that components of PSI, PSII, and the ATPase also were increased. Under the conditions of growth used in this study, these changes led to improvements in photosynthesis and biomass obtained from the Rieske ox plants relative to the wild type.

Rieske FeS Protein of cyt b 6 f
The full-length coding sequence of the Rieske FeS protein of the cyt b 6 f (X64353) complex was amplified by reverse transcription-PCR using primers NtRieskeFeSF (59-caccATGGCTTCTTCTACTCTTTCTCCAG-39) and NtRieske-FeSR (59-CTAAGCCCACCATGGATCTTCACC-39). The resulting amplified product was cloned into pENTR/D (Invitrogen) to make pENTR-NtRieskeFeS, and the sequence was verified and found to be identical. The full-length cDNA was introduced into the pGWB2 Gateway vector (Nakagawa et al., 2007;AB289765) by recombination from the pENTR/D vector to make pGW-NtRieske (B2-NtRi). cDNAs are under transcriptional control of the 35S tobacco mosaic virus promoter, which directs constitutive high-level transcription of the transgene, followed by the nos 39 terminator. Full details of the B2-NtRi construct assembly can be seen in Supplemental Figure S1.

Generation of Transgenic Plants
The recombinant plasmid B2-NtRi was introduced into wild-type Arabidopsis (Arabidopsis thaliana) by floral dipping (Clough and Bent, 1998) using Agrobacterium tumefaciens GV3101. Positive transformants were regenerated on Murashige and Skoog medium containing kanamycin (50 mg L 21 ) and hygromycin (20 mg L 21 ). Kanamycin/hygromycin-resistant primary transformants (T1 generation) with established root systems were transferred to soil and allowed to self-fertilize.

Plant Growth Conditions
Wild-type T2 Arabidopsis plants resulting from self-fertilization of transgenic plants were germinated in sterile agar medium containing Murashige and Skoog salts, selected on kanamycin, and grown to seed in soil (Levington F2; Fisons), and lines of interest were identified by western blot and quantitative PCR. For experimental study, T3 progeny seeds from selected lines were germinated on soil in controlled-environment chambers at an irradiance of 130 mmol m 22 s 21 in an 8-h/16-h square-wave photoperiod with an air temperature of 22°C and a relative humidity of 60%. Plant position was randomized, and the position of the trays was rotated daily under the light. Leaf areas were calculated from photographic images using ImageJ software (imagej.nih.gov/ij). Wild-type plants used in this study were a combined group of the wild type and null segregants from the transgenic lines, verified by PCR for nonintegration of the transgene, as no significant differences in growth parameters were seen between them (Supplemental Fig. S2).

Chlorophyll Fluorescence Imaging
Chlorophyll fluorescence measurements were performed on 10-d-old Arabidopsis seedlings that had been grown in a controlled-environment chamber at a photosynthetic photon flux density (PPFD) of 130 mmol m 22 s 21 with ambient CO 2 at 22°C. Images of the operating efficiency of PSII photochemistry (F q 9/F m 9) were taken at PPFDs of 310 and 600 mmol m 22 s 21 using a chlorophyll fluorescence imaging system (Technologica; Barbagallo et al., 2003;Baker and Rosenqvist, 2004). F q 9/F m 9, was calculated from measurements of steady-state fluorescence in the light (F9), and maximum fluorescence in the light (F m 9) was obtained after a saturating 800-ms pulse of 6,200 mmol m 22 s 21 PPFD using the following equation: F q 9/F m 9 = (F m 9 2 F9)/F m 9 (Oxborough and Baker 1997a; Baker et al., 2001).

A/C i Response Curves
The response of A to C i was measured using a portable gas-exchange system (CIRAS-1; PP Systems). Leaves were illuminated using a red-blue light source attached to the gas-exchange system, and light levels were maintained at saturating PPFD of 1,000 mmol m 22 s 21 with an integral light-emitting diode light source (PP Systems) for the duration of the A/C i response curve. Measurements of A were made at ambient CO 2 concentration (C a ) of 400 mmol mol 21 , before C a was decreased in a stepwise manner to 300, 200, 150, 100, and 50 mmol mol 21 before returning to the initial value and increased to 500, 600, 700, 800, 900, 1,000, 1,100, and 1,200 mmol mol 21 . Leaf temperature and vapor pressure deficit were maintained at 22°C and 1 6 0.2 kPa, respectively. The maximum rate of Rubisco and the maximum rate of electron transport for ribulose 1,5bisphosphate regeneration were determined and standardized to a leaf temperature of 25°C based on equations from Bernacchi et al. (2001) and McMurtrie and Wang (1993), respectively.

Photosynthetic Capacity
Photosynthesis as a function of PPFD (A/Q response curves) was measured using the 6400XT portable gas-exchange system (Li-Cor). Cuvette conditions were maintained at a leaf temperature of 22°C, relative humidity of 50% to 60%, and ambient growth CO 2 concentration of 400 mmol mol 21 for plants grown in ambient conditions. Leaves were initially stabilized at saturating irradiance of 1,000 mmol m 22 s 21 , after which A and stomatal conductance were measured at the following PPFD levels: 0, 50, 100, 150, 200, 250, 300, 350, 400, 500, 600,

PSI and PSII Quantum Efficiency
The photochemical quantum efficiency of PSII and PSI in transgenic and wildtype plants was measured following a dark-light induction transition using a Dual-PAM-100 instrument (Walz) with a DUAL-DR measuring head. Plants were dark adapted for 20 min before placing in the instrument. Following a darkadapted measurement, plants were illuminated with 220 mmol m 22 s 21 PPFD. The maximum quantum yield of PSII was measured following a saturating pulse of light for 600 ms at an intensity of 6,200 mmol m 22 s 21 . The PSII operating efficiency was determined as described by the routines above. PSI quantum efficiency was measured as an absorption change of P700 before and after a saturating pulse of 6,200 mmol m 22 s 21 for 300 ms (which fully oxidizes P700) in the presence of far-red light with a far-red preillumination of 10 s. Both measurements were recorded every 1 min for 5 min. q p or F v 9/F m 9 was calculated from measurements of steady-state fluorescence in the light (F9) and maximum fluorescence in the light (F m 9), while minimal fluorescence in the light (F o 9) was calculated following the equation of Oxborough and Baker (1997b). The fraction of open PSII centers (q L ) was calculated from q p 3 F o 9/F (Baker, 2008).

Pigment Extraction and HPLC Analysis
Chlorophylls and carotenoids were extracted using DMF as described previously (Inskeep and Bloom, 1985), which was shown subsequently to suppress chlorophyllide formation in Arabidopsis leaves (Hu et al., 2013). Briefly, two leaf discs collected from two different leaves were immersed in DMF at 4°C for 48 h and separated by ultra-performance liquid chromatography as described by Zapata et al. (2000).

Leaf Thickness
Leaves of equivalent developmental stage were collected from plants after 28 d of growth. Strips were cut from the center of the leaf, avoiding the midvein, preserved in 5% glutaraldehyde, stored at 4°C for 24 h, followed by dehydration in sequential ethanol solutions of 20%, 40%, 80%, and 100% (v/v). The samples were placed in LR White acrylic resin (Sigma-Aldrich), refrigerated for 24 h, embedded in capsules, and placed at 60°C for 24 h. Sections (0.5 mm) were cut using a Reichert-Jung Ultracut microtome (Ametek), fixed, stained, and viewed with a light microscope (López-Juez et al., 1998). Leaf thickness was determined by measuring leaves from two to three plants from lines 9 and 11 compared with leaves from four wild-type plants.

Statistical Analysis
All statistical analyses were done by comparing ANOVA using Sys-stat (University of Essex). The differences between means were tested using the posthoc Tukey's test (SPSS).

Accession Numbers
Sequence data from this article can be found in the GenBank/EMBL data libraries under accession number X64353.

Supplemental Data
The following supplemental materials are available.
Supplemental Figure S1. Schematic representation of the Rieske FeS overexpression vector pGWRi used to transform Arabidopsis.
Supplemental Figure S2. Determination of photosynthetic efficiency and leaf area of the wild type versus azygous segregating controls using chlorophyll fluorescence imaging.
Supplemental Figure S3. Immunoblot analysis of wild-type and Rieske FeS ox proteins.
Supplemental Figure S4. Light response curves of the Rieske FeS plants at 21% [O 2 ].
Supplemental Figure S5. Determination of the efficiency of electron transport in leaves of mature Rieske FeS ox plants.
Supplemental Figure S6. Determination of the efficiency of electron transport in leaves of young and mature Rieske FeS ox plants.
Supplemental Table S1. Physiological parameters of Rieske FeS ox plants compared with the wild type.