Despite significant progress in clarifying the subunit compositions and functions of the multiple NADPH dehydrogenase (NDH-1) complexes in cyanobacteria, the subunit maturation and assembly of their NDH-1 complexes are poorly understood. By transformation of wild-type cells with a transposon-tagged library, we isolated three mutants of Synechocystis sp. PCC 6803 defective in NDH-1-mediated cyclic electron transfer and unable to grow under high light conditions. All the mutants were tagged in the same slr1097 gene, encoding an unknown protein that shares significant homology with the Arabidopsis protein chlororespiratory reduction 6 (CRR6). The slr1097 product was localized in the cytoplasm and was required for efficient assembly of NDH-1 complexes. Analysis of the interaction of Slr1097 with 18 subunits of NDH-1 complexes using a yeast two-hybrid system indicated a strong interaction with NdhI but not with other Ndh subunits. Absence of Slr1097 resulted in a significant decrease of NdhI in the cytoplasm, but not of other Ndh subunits including NdhH, NdhK and NdhM; the decrease was more evident in the cytoplasm than in the thylakoid membranes. In the ∆slr1097 mutant, NdhH, NdhI, NdhK and NdhM were hardly detectable in the NDH-1M complex, whereas almost half the wild-type levels of these subunits were present in NDH-1L complex; similar results were observed in the NdhI-less mutant. These results suggest that Slr1097 is involved in the maturation of NdhI, and that assembly of the NDH-1M complex is strongly dependent on this factor. Maturation of NdhI appears not to be crucial to assembly of the NDH-1L complex.
Cyanobacterial NADPH dehydrogenase (NDH-1) complexes localize in the thylakoid membrane (Ohkawa et al., 2001, 2002; Zhang et al., 2004; Xu et al., 2008; Battchikova et al., 2011a) and participate in a variety of bioenergetic reactions, including respiration, cyclic electron transport around photosystem I, and CO2 acquisition (Ogawa, 1991; Mi et al., 1992; Ohkawa et al., 2000). Structurally, the cyanobacterial NDH-1 complexes closely resemble energy-converting Complex I in eubacteria and the mitochondrial respiratory chain despite the absence in cyanobacterial genomes of homologs of the three subunits that constitute the catalytically active core of Complex I (Friedrich et al., 1995; Friedrich and Scheide, 2000; Arteni et al., 2006). Over the past few years, significant advancements have been made in resolving the subunit compositions and functions of the multiple NDH-1 complexes in several cyanobacterial strains (Battchikova and Aro, 2007; Ogawa and Mi, 2007; Ma, 2009; Battchikova et al., 2011b). Four types of NDH-1 have been identified in the cyanobacterium Synechocystis sp. strain PCC 6803 (hereafter Synechocystis 6803), and all four types are involved in NDH-1-dependent cyclic electron transport around photosystem I (NDH-CET) (Bernát et al., 2011). NDH-CET allows optimal functioning of photosynthesis by increasing the pH gradient and supplying additional ATP for CO2 assimilation. This function is particularly important under environmental stress conditions, such as high light (Endo et al., 1999; Battchikova et al., 2011a), in which the ATP demand is greatly increased. Moreover, the impairment of cyanobacterial NDH-CET caused by mutation of some auxiliary factors for NDH-1 complexes and Ndh subunits, such as NdhS (Battchikova et al., 2011a), results in high light-sensitive growth phenotypes. Therefore, use of high light conditions assists in identifying unknown factors or subunits that affect NDH-CET activity.
Cyanobacterial NDH-1 complexes together with the NADH dehydrogenase-like (NDH) complex from chloroplasts of green plants constitute sub-class of the Complex I family (Battchikova et al., 2011b). A common feature of this group, apart from absence of the three active subunits homologous to NuoE, NuoF and NuoG of the Escherichia coli enzyme, is the presence of several subunits specific for complexes originating from cells that perform oxygenic photosynthesis. However, the NDH complex of chloroplasts contains many subunits that are absent in cyanobacterial NDH-1 complexes (Ifuku et al., 2011; Peng et al., 2011a). These subunits, encoded by nuclear genes, form two NDH sub-complexes, the luminal complex and sub-complex B, which are specific to chloroplast NDH. Recently, in higher plants, reverse genetics studies identified many nuclear-encoded auxiliary proteins for stabilization and assembly of the chloroplast NDH complex, such as chlororespiratory reduction 6 (CRR6) (Suorsa et al., 2009; Ifuku et al., 2011; Peng et al., 2011a). However, homologs of most of these auxiliary proteins have not been found in cyanobacterial genomes. The severe alteration of the Ndh subunits and auxiliary proteins of the complexes during evolution from cyanobacteria to green plants (Suorsa et al., 2009; Ifuku et al., 2011) implies that the stability and assembly of chloroplast NDH complex and cyanobacterial NDH-1 complexes may be significantly different.
Despite significant achievements in the study of cyanobacterial NDH-1 complexes over the past few years, the auxiliary proteins for the complexes are still poorly understood. Here, we report the discovery of a maturation factor, Slr1097, in Synechocystis 6803, and this maturation factor was found to be specific for the NdhI subunit. We demonstrate that Slr1097 is essential for the assembly of the NDH-1M complex but is less important for the assembly of the NDH-1L complex.
Inactivation of slr1097 impairs NDH-CET activity
NDH-CET plays an important role in alleviating high-light stress in cyanobacteria (Battchikova et al., 2011a) and higher plants (Endo et al., 1999). Therefore, upon exposure of cells to high light, the growth of NDH-CET-defective mutants, such as ∆ndhS, is significantly retarded compared with the wild-type (WT) despite similar growth under growth light irradiation. To screen for NDH-CET-defective mutants, we transformed WT cells with a transposon-bearing library, thus tagging and inactivating many genes, and then cultured the mutant cells under high light. We isolated three mutant strains that grew slowly on plates under high light but similarly to WT under growth light conditions (Figure 1a).
To test whether the high light-dependent phenotype of the three mutants resulted from defective NDH-CET, we monitored the post-illumination increase in chlorophyll a fluorescence, which is extensively used to monitor NDH-CET activity in vivo in cyanobacteria (Mi et al., 1995; Deng et al., 2003; Ma and Mi, 2005; Battchikova et al., 2011a) and higher plants (Burrows et al., 1998; Shikanai et al., 1998; Hashimoto et al., 2003; Wang et al., 2006; Peng et al., 2009, 2011b, 2012; Sirpiö et al., 2009; Yamamoto et al., 2011). Figure 1(b) shows that the NDH-CET activity in all three mutants was lower than that in WT as determined by the height and the relative rate of post-illumination increase in chlorophyll fluorescence. The results indicate that NDH-CET is impaired in mutants 1, 2 and 3.
To identify the genes inactivated by transposon tagging, we analyzed the sites of transposon insertion in the three mutants. As shown by the PCR results (Figure 1c), these mutants were tagged in the same gene of unknown function (slr1097). The transposon insertion occurred at position 194 532 of the Synechocystis 6803 genome (gi 16329348) (Kaneko et al., 1996).
To confirm that inactivation of slr1097 impairs NDH-CET activity, we replaced the slr1097 coding region by a kanamycin resistance marker (KamR) (Figure 2a). PCR analysis of the slr1097 locus confirmed complete segregation of the ∆slr1097 mutant allele (Figure 2b). Immunoblotting analysis using an antibody specifically prepared against Slr1097 (see 'Slr1097 protein is required for efficient assembly of NDH-1 complexes in the thylakoid membranes') demonstrated absence of the gene product in the mutant (Figure 2c). As expected, the NDH-CET activity, as measured by the post-illumination increase in chlorophyll fluorescence, was lower in ∆slr1097 than in the WT. However, the activity remained relatively high compared with that in the M55 mutant (Figure 2d). A similar result was obtained by measuring the oxidation of P700 by far-red light after actinic light illumination. When actinic light was turned off after 30 sec illumination by actinic light (800 μE m−2 sec−1) supplemented with far-red light, P700+ was transiently reduced by electrons from the plastoquinone pool, and subsequently re-oxidized by background far-red light. Operation of the NDH-1 complexes, which transfer electrons from the reduced cytoplasmic pool to plastoquinone, hinders the re-oxidation of P700 (Shikanai et al., 1998; Battchikova et al., 2011a). The re-oxidation of P700 was markedly faster in ∆slr1097 compared with the WT, although it was still relatively slow compared with the M55 mutant (Figure 2e). We also measured the NDH-CET by monitoring the reduction rate of P700+ in darkness after illumination of cells with far-red light. The re-reduction of P700+ was much slower in ∆slr1097 compared with that in the WT, although it was still relatively fast compared with the rate in the M55 mutant (Figure 2f). Based on the above results, we conclude that inactivation of a cyanobacterial gene, slr1097, impairs NDH-CET activity.
Slr1097 protein is required for efficient assembly of NDH-1 complexes in the thylakoid membranes
To reveal how mutation of slr1097 impairs NDH-CET activity, we compared the accumulation and assembly of the NDH-1L and NDH-1M complexes in the thylakoid membranes of the WT, ∆slr1097 and M55 strains. Inactivation of slr1097 reduced the abundance of NDH-1 subunits in the thylakoid membranes, predominantly NdhI protein (approximately 75% reduction) (Figure 3a). Furthermore, the inactivation almost completely abolished assembly of the NDH-1M complex but decreased assembly of the NDH-1L complex to almost half of the WT levels, as assessed by the protein abundance of NdhH, NdhI, NdhK and NdhM (Figure 3b,c). It seems likely that the significant reduction in NdhI protein mainly resulted in instability and subsequent disassembly of the NDH-1M complex, thereby impairing the NDH-CET activity.
Slr1097 is a cytoplasmic protein conserved in phototrophs
To understand how slr1097 deletion influenced the stability and assembly of NDH-1 complexes, we investigated the localization of the Slr1097 protein. To confirm purification of the soluble (SF) and total membrane (MF) fractions, we probed them using SF-specific RbcL and MF-specific D1 antibodies (Zhang et al., 2009, 2012). The results shown in Figure 4 demonstrate that there was little cross-contamination between SF and MF because the marker proteins RbcL and D1 were exclusively present in the SF and MF, respectively. The Slr1097 protein was detected only in the SF, in contrast to localization of the NDH-1 hydrophilic subunits NdhH and NdhK mainly in the MF (Figure 4). This clearly indicates that Slr1097 protein is confined to the cytoplasm.
As shown by Munshi et al. (2006), Slr1097 is a homolog of CRR6 (see Figure S1), and homologs are present in phototrophic organisms but not in Chlamydomonas reinhardtii, which lacks photosynthetic ndh genes (Maul et al., 2002). This fact, and the localization of Slr1097 and CRR6 in the cytoplasm and stroma (Peng et al., 2010), respectively, suggest that Slr1097 may play a similar role to CRR6 as an auxiliary factor in NDH-1 complex assembly.
Slr1097 interacts with NdhI
To clarify the role of Slr1097 in assembling cyanobacterial NDH-1 complexes, we determined the interaction of Slr1097 with 18 Ndh subunits that had been identified in NDH-1 complexes, using a yeast two-hybrid system (Figure S2). A very strong protein–protein interaction was found between Slr1097 and NdhI (Figure 5 and Figure S2). This implies that Slr1097 protein may be exclusively involved in maturation and/or assembly of the NdhI subunit.
To test this possibility, we analyzed the amounts of Slr1097, NdhI, NdhH, NdhK and NdhM proteins that accumulated in soluble fractions from the WT, ∆slr1097 and M55 strains. As shown in Figure 6, the absence of Slr1097 protein resulted in a significant reduction of NdhI but not NdhH, NdhK or NdhM in the soluble fraction. This suggests that Slr1097 is involved in maturation and/or assembly of NdhI. Furthermore, the degree of NdhI protein reduction caused by slr1097 mutation was more remarkable in the soluble fraction (less than one-eighth that of the WT; see Figure 6) than in the membrane fraction [approximately one-quarter that of the WT; see Figure 3(a)]. Therefore, we conclude that the Slr1097 protein exclusively participates in maturation of NdhI in Synechocystis 6803.
The decrease in mature NdhI protein destabilizes the NDH-1 complexes in the thylakoid membranes
Deletion of ndhB resulted in complete collapse of the NDH-1L and NDH-1M complexes in the thylakoid membranes (Ogawa, 1991; Zhang et al., 2004) (also see Figures 3 and 7). To test whether the reduction in NdhI protein caused by the absence of Slr1097 protein destabilizes the NDH-1 complexes, we attempted to construct a ∆ndhI mutant strain (Figure S3a–c and Methods S1). Although we were unable to achieve complete segregation of an inactivated ndhI gene, the amount of NdhI was reduced to less than half the WT value in the NdhI-less mutant, similar to the amount in ∆slr1097 mutant (Figure 3a and Figure S3c). The partial deletion of ndhI in the NdhI-less mutant greatly decreased the amount of the NDH-1M complex but did not reduce the amount of the NDH-1L complex so much (only to approximately half the WT level), as determined by the protein abundance of NdhH, NdhI, NdhK and NdhM (Figure 7a,b), in line with the result for the ∆slr1097 mutant (Figure 3b,c). This finding was reinforced by the observation of a high light-sensitive growth phonotype (Figure S4) and impaired NDH-CET activity (Figure S5) in the NdhI-less mutant compared with the WT. We therefore conclude that a sufficient level of NdhI protein is required for efficient assembly of the NDH-1M complex, but appears to be not essential for efficient assembly of the NDH-1L complex, and that Slr1097 is required to stabilize NdhI in the cytoplasm.
Over the past few years, numerous auxiliary proteins that are mainly involved in splicing and editing of chloroplast ndh mRNAs as well as stabilization and assembly of the NDH complex have been discovered and characterized in higher plants (Suorsa et al., 2009; Peng et al., 2012). However, such auxiliary proteins are poorly understood in cyanobacteria, even though homologous sequences of auxiliary proteins for chloroplast NDH complexes, such as CRR6 and CRR7 (Munshi et al., 2005, 2006; Takabayashi et al., 2009), have been found in cyanobacterial genomes. Here, we successfully demonstrate that a cytoplasmic auxiliary protein, Slr1097, plays a similar role to CRR6 in stabilization of the NdhI subunit in the cytoplasm.
Knockout of the crr6 gene resulted in complete impairment of NDH-CET and collapse of sub-complex A of chloroplast NDH (Munshi et al., 2006; Takabayashi et al., 2009; Peng et al., 2012). In contrast, deletion of slr1097 only partially decreased NDH-CET activity (Figures 1 and 2). Although deletion of slr1097 almost completely abolished assembly of the NDH-1M complex, it only partially reduced assembly of the NDH-1L complex (approximately 50% reduction) (Figure 3). Thus, there is a difference between cyanobacteria and higher plants regarding the role of Slr1097/CRR6 in assembly of NDH-1 complexes. Although absence of CRR6 completely abolished the accumulation of NdhI in the stroma (Peng et al., 2012), absence of Slr1097 reduced the amount significantly but not completely in the cytoplasm (Figure 6). Thus, stabilization of NdhI in the stroma is completely dependent on CRR6 in higher plants, but a certain amount of NdhI stably exists in the cytoplasm even in the absence of Slr1097 in cyanobacteria. NdhI may be preferentially utilized for assembly of the NDH-1L complex, which explains the significant reduction in the NDH-1M complex and the partial reduction in the NDH-1L complex. This was reinforced by the results for the NdhI-less mutant (Figure 7). In this mutant, the amount of NdhI was reduced to almost one-third of the WT level (Figure S3), and the reduction of the NDH-1M complex was much more significant than that of the NDH-1L complex (Figure 7). The presence of the NDH-1L complex in ∆slr1097 but the absence of its counterpart in ∆crr6 explains the different NDH-CET activities in these two mutants.
In chloroplasts, a High Chlorophyll Fluorescence 101 (HCF101) protein was co-purified with CRR6. This stromal protein was previously reported to function as a scaffold for assembly of the [4Fe–4S] cluster in chloroplasts (Schwenkert et al., 2010). The crystal structure of the hydrophilic domain of complex I from Thermus thermophilus showed that Nqo9, which corresponds to NdhI in chloroplasts and cyanobacteria, binds three [4Fe–4S] clusters (Sazanov and Hinchliffe, 2006; Efremov et al., 2010; Baradaran et al., 2013). Co-purification of HCF101 with CRR6 suggested that HCF101 is required for the formation of [4Fe–4S] clusters in NdhI. In Synechocystis 6803, Slr0067 is homologous to HCF101 (Lezhneva et al., 2004). Moreover, Slr0067 interacts with NdhI, as determined by the results of the yeast two-hybrid assay (Figure S6). Thus, Slr0067 may be essential for formation of [4Fe–4S] clusters in NdhI in the cyanobacterium Synechocystis 6803, as schematically represented in Figure 8.
Analysis of the crystal structure of the hydrophilic domain of NDH-1 from T. thermophilus showed that Nqo9 and Nqo6, which correspond to NdhI and NdhK, respectively, in cyanobacterial NDH-1 complexes, bind three [4Fe–4S] clusters, N6a, N6b and N2 (Sazanov and Hinchliffe, 2006; Efremov et al., 2010; Baradaran et al., 2013), and form the terminal part of the electron transfer pathway to quinone (red arrows in Figure 8). Although the source of electrons for cluster N6a in the NdhI subunit is still unknown in cyanobacteria because of a lack of knowledge regarding three active subunits that constitute the electron donor of Complex I (question mark in Figure 8), the decrease in NdhI caused by the absence of Slr1097, predominantly in the NDH-1M complex and only partially in the NDH-1L complex (Figure 3b,c), will certainly impair the activity of electron transfer to quinone. Therefore, the residual NDH-CET activity in ∆slr1097 may mainly originate from the NDH-1L complex.
The Synechocystis 6803 glucose-tolerant strain (wild-type) and the mutants ∆slr1097, NdhI-less and M55 (Ogawa, 1991) were cultured at 30°C in BG-11 medium (Allen, 1968) buffered with Tris/HCl (5 mM, pH 8.0) and bubbled with 2% v/v CO2 in air. The solid medium used was BG-11 supplemented with 1.5% agar. Continuous illumination was provided by fluorescence lamps at 40 μE m−2 sec−1.
Isolation and construction of mutants
Using a GPS-1 genomic priming system (New England Biolabs, http://www.neb.com/), a library of cosmids was obtained by random insertion of a transposon containing a chloramphenicol resistance (CmR) cassette into 1092 genes of Synechocystis 6803 (Ozaki et al., 2007). To avoid isolating already known mutants, we selected 662 genes with the CmR insertion, most of which encode hypothetical or unknown function proteins, and made a new library. The wild-type strain of Synechocystis 6803 was transformed with this new inactivation library, and CmR mutants that displayed a slow growth phenotype under high light conditions but not under growth light conditions were isolated. Genomic DNA isolated from each mutant was digested using HhaI, and, after self-ligation, was used as a template for inverse PCR with primers complementary to the N- and C-terminal regions of the CmR cassette (Table S1). The exact position of the cassette in the mutant genome was determined by sequencing the PCR product.
A fragment of 389 bp between the SalI and XbaI sites was removed from the slr1097 gene, and a fragment containing the kanamycin resistance (KamR) cassette was inserted between the two restriction sites using appropriate PCR primers, slr1097-G and slr1097-H (Table S1). The constructed vector was used to transform WT cells of Synechocystis 6803 as described by Williams and Szalay (1983). The transformants were spread on agar plates containing BG-11 medium and kanamycin (10 μg ml−1) buffered at pH 8.0, the plates were incubated in 2% v/v CO2 in air, and continuous illumination was provided by fluorescent lamps generating photosynthetically active radiation of 40 μE m−2 sec−1. The mutated slr1097 in the transformants was segregated to homogeneity (by successive streak purification) as determined by PCR amplification and immunoblotting.
Chlorophyll fluorescence and P700 analysis
The transient increase in chlorophyll fluorescence after actinic light had been turned off was monitored as described by Ma and Mi (2005). The redox kinetics of P700 were measured as previously described (Battchikova et al., 2011a). The re-reduction of P700+ in darkness was measured using a DUAL-PAM-100 measuring system (Walz, http://www.walz.com/) with an ED-101US/MD emitter detector unit, by monitoring absorbance changes at 830 nm and using 875 nm as a reference. Cells were kept in the dark for 2 min, and 10 μm (3-(3,4-dichlorophenyl)-1,1-dimethylurea) was added to the cultures prior to measurement. P700 was oxidized by far-red light at a maximum wavelength of 720 nm from an LED lamp for 30 sec, and the subsequent re-reduction of P700+ in the dark was monitored.
Isolation of crude thylakoid membranes
Cell cultures (800 ml) were harvested at the logarithmic phase and washed twice for 2 min each at 4°C by suspension in 50 ml of fresh BG-11 medium, and the thylakoid membranes were isolated as described by Gombos et al. (1994) with some modifications as follows. Cells suspended in 5 ml disruption buffer (10 mm HEPES/NaOH, 5 mm sodium phosphate, pH 7.5, 10 mm MgCl2, 10 mm NaCl and 25% v/v glycerol) were supplemented with zirconia/silica beads and broken by vortexing 15 times at the highest speed for 20 sec at 4°C with a Bead-beater (Biospec, http://www.biospec.com/) followed by 5 min cooling on ice between runs. The crude extract was centrifuged at 5000 g for 5 min to remove the glass beads and unbroken cells. We obtained crude thylakoid membranes from the precipitate by further centrifugation at 20 000 g for 30 min.
Isolation of membrane and soluble cell fractions
Total membrane and soluble fractions of Synechocystis 6803 cells were isolated as described previously (Zhang et al., 2009, 2012) with slight modifications. In brief, the cells were pelleted by 5 min centrifugation at 5000 g from batch cultures and broken with glass beads by vortexing at 4°C in disruption buffer. The cell debris and glass beads were removed by 5 min centrifugation at 5000 g. The total membrane and soluble fractions were separated by centrifugation at 110 000 g for 30 min.
Electrophoresis and immunoblotting
Blue native PAGE (BN-PAGE) of Synechocystis 6803 membranes was performed as described previously (Kügler et al., 1997) with slight modifications (Battchikova et al., 2011a). Isolated membranes were prepared for BN-PAGE as follows. Membranes were washed for 2 min at 4°C with 330 mm sorbitol, 50 mm Bis/Tris, pH 7.0, 0.5 mm phenylmethanesulfonyl fluoride (Sigma, http://www.sigmaaldrich.com/), and resuspended in 20% w/v glycerol, 25 mM Bis/Tris, pH 7.0, 10 mm MgCl2, 0.1 units RNase-free DNase RQ1 (Promega, http://www.promega.com/) at a chlorophyll a concentration of 0.3 mg ml−1, and 0.5 mm phenylmethanesulfonyl fluoride. The samples were incubated on ice for 10 min, and an equal volume of 3% n-dodecyl β-d-maltoside was added. Solubilization was performed for 40 min on ice. Insoluble components were removed by centrifugation at 18 000 g for 15 min. The collected supernatant was mixed with 1/10 volume of sample buffer, 5% Serva Blue G (Serva, http://www.serva.de/), 100 mm Bis/Tris, pH 7.0, 30% w/v sucrose, 500 mm ε-amino-n-caproic acid and 10 mm EDTA. Solubilized membranes were then applied to a 0.75 mm thick 5–12.5% acrylamide gradient gel (Mighty Small mini-vertical unit, Hoefer, http://hoeferinc.com/). Samples were loaded on an equal chlorophyll a basis per lane. Electrophoresis was performed at 4°C by increasing the voltage gradually from 50 to 200 V during the 5.5 h run. The lanes of the BN gel were cut out and incubated in Laemmli SDS sample buffer containing 5% β-mercaptoethanol and 6 m urea for 1 h at 25°C.
SDS–PAGE of Synechocystis 6803 membrane and soluble cell fractions was performed on a 12% polyacrylamide gel with 6 m urea as described previously (Laemmli, 1970).
For immunoblotting, the proteins were electrotransferred to a polyvinylidene difluoride membrane (Immobilon-P; Millipore, http://www.millipore.com/) and detected using protein-specific antibodies with an ECL assay kit (Amersham Pharmacia, http://www.gelifesciences.com/) according to the manufacturer's protocol. Antibodies against the Slr1097 protein of Synechocystis 6803 were raised in our laboratory. To amplify the slr1097 gene, primer sequences were designed and are listed in Table S1. The PCR products were ligated into vector pET32a (Novagen, http://www.novagen.com/), and the construct was amplified in E. coli DH-5α. The plasmid was used to transform E. coli strain BL21 (DE3) pLysS for expression. The gene expression products from E. coli were purified and used as antigens to immunize rabbits to produce polyclonal antibodies. The NDH-1 complexes were detected using antibodies against NdhH, NdhI, NdhK and NdhM, respectively, that were previously raised in our laboratory (Ma and Mi, 2005).
Yeast two-hybrid assay
The yeast two-hybrid assay was performed using the LexA system (Clontech, http://www.clontech.com/). A PCR-amplified slr1097 gene fragment was cloned in-frame into XhoI sites of pLexA (Clontech) to form the bait construct using primers shown in Table S1. The fragments containing 18 ndh genes encoding NdhA–NdhQ and NdhS were amplified by PCR, and inserted into the EcoRI and XhoI or XhoI sites of pJG4-5 (Clontech) to form the prey construct. The bait and prey constructs, together with reporter vector pSH18-34 (Clontech), were co-transformed into yeast strain EGY48 according to the manufacturer's instructions for the Matchmaker LexA two-hybrid system (Clontech). The bait and prey constructs for CCT1 and COP1 were prepared as described previously (Yang et al., 2001). Transformed yeast was diluted and dropped onto synthetic drop-out (SD) medium, leucine-negative (–Leu) SD medium or X-gal medium, then grown at 30°C in darkness as described previously (Sun et al., 2009).
We thank Hongquan Yang (Shanghai Jiaotong University, China) for supplying the yeast two-hybrid system. This work was partially supported by the National Basic Research Program of China (grant number 2009CB118500), and the Project of Shanghai Education Committee (grant number 11YZ89 and 12ZZ132).