CHLORORESPIRATORY REDUCTION 9 is a Novel Factor Required for Formation of Subcomplex A of the Chloroplast NADH Dehydrogenase-Like Complex

Abstract

In vascular plants, the chloroplast NADH dehydrogenase-like (NDH) complex, a homolog of respiratory NADH:quinone oxidoreductase (Complex I), mediates plastoquinone reduction using ferredoxin as an electron donor in cyclic electron transport around PSI in the thylakoid membrane. In angiosperms, chloroplast NDH is composed of five subcomplexes and forms a supercomplex with PSI. The modular assembly of stroma-protruded subcomplex A, which corresponds to the Q module of Complex I, was recently reported. However, the factors involved in the specific assembly steps have not been completely identified. Here, we isolated an Arabidopsis mutant, chlororespiratory reduction 9 (crr9), defective in NDH activity. The CRR9 gene encodes a novel stromal protein without any known functional domains or motifs. CRR9 is highly conserved in cyanobacteria and land plants but not in green algae, which do not have chloroplast NDH. Blue native-PAGE and immunoblot analyses of thylakoid proteins indicated that formation of subcomplex A was impaired in crr9. CRR9 was specifically required for the accumulation of NdhK, a subcomplex A subunit, in NDH assembly intermediates in the stroma. Furthermore, two-dimensional clear native/SDS–PAGE analysis of the stroma fraction indicated that incorporation of NdhM into NDH assembly intermediate complex 400 was impaired in crr9. These results suggest that CRR9 is a novel factor required for the formation of NDH subcomplex A.

Introduction

The chloroplast NADH dehydrogenase-like (NDH) complex is a multisubunit protein complex embedded in the thylakoid membrane and returns electrons from PSI to the plastoquinone (PQ) pool via ferredoxin (Fd) as an electron donor in cyclic electron transport around PSI (CET) (Peltier et al. 2016, Shikanai 2016). CET produces a trans-thylakoid proton gradient (ΔpH) via the Q-cycle in the Cyt b₆f complex and possibly proton pumping by the NDH complex, contributing to the formation of the proton motive force, which energizes ATP synthesis (Shikanai 2014, Wang et al. 2015). Acidification of the thylakoid lumen by CET induces non-photochemical quenching of Chl fluorescence (NPQ) to dissipate excessively absorbed photon energy as heat (Müller et al. 2001) and decelerates electron transport via the down-regulation of Cyt b₆f activity, resulting in photoprotection of photosystems, especially under fluctuating light conditions (Munekage et al. 2002, Suorsa et al. 2012, Yamamoto et al. 2016). Genetics using Arabidopsis support the idea that in angiosperms, CET is mediated by at least two independent Fd:PQ oxidoreductase complexes, the antimycin A-sensitive PROTON GRADIENT REGULATION 5 (PGR5)/PGR5-like Photosynthetic Phenotype 1 (PGRL1) complex and the antimycin A-insensitive NDH complex (Munekage et al. 2004, Hertle et al. 2013, Sugimoto et al. 2013). Although the contribution of NDH to CET is smaller than that of the PGR5/PGRL1 complex, its physiological function in regulating photosynthesis emerges especially under low light conditions (Yamori et al. 2011, Ueda et al. 2012, Yamori et al. 2015) and also in fluctuating light conditions in rice (Yamori et al. 2016).

Chloroplast NDH was first identified on the basis of its homology to respiratory NADH:ubiquinone oxidoreductase (Complex I) in mitochondria (Matsubayashi et al. 1987). Its presence was confirmed through biochemical detection of subunits and analysis of tobacco ndh mutants (Burrows et al. 1998, Kofer et al. 1998, Sazanov et al. 1998, Shikanai et al. 1998). Subunits of chloroplast NDH are more similar to the subunits of cyanobacterial NDH-1 than to those present in the mitochondria of the same plant, suggesting the cyanobacterial origin of chloroplast NDH (Friedrich et al. 1995, Shikanai 2016). Recently, the fine tertiary structure of Complex I from the thermophilic bacterium Thermus thermophilus was solved (Baradaran et al. 2013). Thermus thermophilus Complex I forms an L-shaped structure, with a membrane arm embedded in the membrane and a peripheral arm extending into the cytosol. All cofactors involved in electron transport are localized to the peripheral arm. An NADH:FMN oxidoreductase module (N module) formed by the three subunits Nqo1–Nqo3/NuoE–NuoG is present on the top of the peripheral arm and accepts two electrons from NADH (Sazanov and Hinchliffe 2006, Baradaran et al. 2013). Electrons derived from NADH oxidation are transferred to Fe–S clusters present in the quinone-binding module (Q module) and finally to a quinone at the interface between the two arms. Coupled with reduction of a quinone to a quinol, a proton pumping module (P module) of the membrane arm, consisting of antiporter-like subunits, induces conformational changes and translocates four protons from the cytosol to the periplasmic space (Baradaran et al. 2013, Sazanov 2015). Fourteen subunits forming the L-shaped ‘catalytic core’ complex have been highly conserved during the evolution of respiratory Complex I, even though mitochondrial Complex I has acquired >30 supernumerary subunits (Letts and Sazanov 2015, Peltier et al. 2016).

Extensive studies of Arabidopsis ndh mutants by genetics, biochemistry and bioinformatics have revealed that chloroplast NDH consists of a ‘catalytic core’ subunit and many supernumerary subunits (Suorsa et al. 2009, Ifuku et al. 2011, Peng et al. 2011b), and further forms a supercomplex of >1 MDa with PSI (Peng et al. 2008, Peng et al. 2009). Mainly on the basis of the stability of each subunit in a series of ndh mutants, the chloroplast NDH complex has been divided into five subcomplexes, namely stroma-faced subcomplex A (SubA) and SubB, thylakoid-embedded membrane subcomplex (SubM), lumen subcomplex (SubL) and electron-donor-binding subcomplex (SubE) (Ifuku et al. 2011, Shikanai 2016). SubA (NdhH–NdhO) and SubM (NdhA–NdhG) correspond to the Q- and P-modules, respectively, although NdhA homologs are categorized as connecting subunits between the Q and P modules in respiratory Complex I. Although an L-shaped core consisting of Q and P modules is highly conserved between photosynthetic NDH (chloroplast and cyanobacterial NDH) and respiratory NDH, genes coding for N-module subunits have not been found in nuclear or plastid genomes in land plants or in the genomes of cyanobacteria (Friedrich et al. 1995, Shikanai 2007). Instead of the N module, chloroplast NDH has acquired SubE (Peltier et al. 2016). SubE is composed of at least four subunits, NdhS–NdhV, in Arabidopsis. NdhS/CRR31 has a Sarcoma homology 3 domain-like fold, and its positively charged pocket serves as the binding site for Fd (Yamamoto et al. 2011, Yamamoto and Shikanai 2013). Two J proteins, NdhT/CRRJ and NdhU/CRRL, and NdhV stabilize NdhS (Yamamoto et al. 2011, Fan et al. 2015), and SubE interacts with SubA (Yamamoto et al. 2011). Chloroplast NDH reduces PQ by using Fd as an electron donor rather than NAD(P)H; we have renamed the chloroplast NAD(P)H dehydrogenase complex as the chloroplast NADH dehydrogenase-like complex (Ifuku et al. 2011, Yamamoto et al. 2011).

SubA is composed of four plastid-encoded (NdhH–NdhK) and four nuclear-encoded (NdhL–NdhO) subunits. Peng et al. (2012) clarified the modular assembly of hydrophilic SubA subunits in the chloroplast stroma during the biogenesis of SubA. SubA subunits are sequentially assembled into high molecular mass NDH assembly intermediate (NAI) complexes of approximately 800 kDa (NAI800), approximately 500 kDa (NAI500) and approximately 400 kDa (NAI400) with the help of the chaperonin 60 (Cpn60) complex containing Cpn60β4 subunits and the NDH-specific assembly factors CRR1, CRR6, CRR7, CRR41 and CRR42 (Peng et al. 2010, Peng et al. 2011a, Peng et al. 2012). Hydrophilic SubA subunits are fully assembled in the stroma and are docked to other parts of NDH complexes embedded in the thylakoid membrane to form the NDH–PSI supercomplex. However, the steps from biogenesis to assembly are not yet completely understood, especially in relation to the membrane-embedded subunits.

Here, we report an Arabidopsis NDH-defective mutant, chlororespiratory reduction 9 (crr9). CRR9 encodes a novel protein without any known functional domains or motifs. CRR9 was localized to the chloroplast stroma, but its defect resulted in impaired SubA accumulation in the thylakoid membrane. Furthermore, accumulation of NdhK in the chloroplast stroma and incorporation of NdhM into NAI400 were severely disturbed in the crr9 mutants. These results indicate that CRR9 is a novel factor for formation of SubA of the chloroplast NDH complex.

Results

Arabidopsis crr9 is specifically impaired in chloroplast NDH activity

Chloroplast NDH is embedded in the thylakoid membrane and mediates electron donation from the stromal electron pool to PQ via Fd (Fig. 1A). After actinic light (AL) illumination, NDH still donates electrons to PQ in the dark, resulting in reduction of the PQ pool, which is monitored as a transient increase in Chl fluorescence (F_o rise) (Burrows et al. 1998, Shikanai et al. 1998) (Fig. 1B). Arabidopsis chlororespiratory reduction (crr) mutants defective in NDH activity were isolated on the basis of the lack of this post-illumination Chl fluorescence change (Hashimoto et al. 2003). crr9-1 was isolated by screening of Arabidopsis M₂ generation mutagenized by ethyl methanesulfonate using the technique of Chl fluorescence imaging (Shikanai et al. 1999, Hashimoto et al. 2003). The ndhl/crr23 mutant is defective in the gene encoding NdhL (Shimizu et al. 2008), whereas crr4-3 is defective in the RNA editing creating the translational initiation codon of ndhD (Kotera et al. 2005). In all crr mutants, including crr9-1, the post-illumination F_o rise was diminished (Fig. 1B). As in ndh mutants, the F_o rise was absent in crr9-1, indicating that CRR9 is required for NDH activity (Fig. 1B).

Fig. 1

Identification of the novel NDH-deficient mutant crr9. (A) Schematic model indicating electron transport from Fd to the PQ pool via NDH. PQ reduction was monitored by the Chl fluorescence emitted from PSII. PQ reduction in the dark depends on NDH activity and can be monitored as a transient increase in Chl fluorescence after AL illumination. PC, plastocyanin; FNR, Fd:NADP⁺ oxidoreductase. (B) Analysis of NDH activity in vivo by monitoring the transient increase in Chl fluorescence after turning off actinic light (AL). The bottom curve is a typical trace of Chl fluorescence in wild-type (WT) leaves. Leaves were exposed to AL (50 μmol photons m^–2 s^–1) for 5 min. AL was turned off and the subsequent change in Chl fluorescence (dashed boxed area) was monitored. Insets are magnified traces from the dashed boxed area. Fluorescence levels were normalized against F_m levels. a.u., arbitrary units. F_m, maximum Chl fluorescence; F_o, minimum Chl fluorescence; SP, saturating light pulse of white light. (C) Structure of CRR9. Open, gray and filled boxes indicate exons, untranslated regions and the transposon, respectively. Positions of the mutation and transposon insertion are indicated. Positions of primers used for RT–PCR in (D) are indicated by arrowheads. (D) RT–PCR analysis of CRR9 mRNA in WT and crr9 alleles. ACT8 was amplified as a control. U and C with arrowheads indicate the positions of cDNAs amplified from unspliced and spliced RNA, respectively. The red arrowhead indicates the position of cDNA amplified from crr9-1RNA. M, molecular size marker.

The contribution of the chloroplast NDH complex to photosynthetic electron transport is minor in Arabidopsis, and crr mutants specifically defective in NDH activity show no distinct phenotypes under growth chamber conditions at 50 μmol photons m^–2 s^–1 (Hashimoto et al. 2003, Shimizu et al. 2008, Yamamoto et al. 2011). Like other crr mutants, crr9-1 exhibited no visible phenotypes in the growth chamber. We analyzed several Chl fluorescence parameters that reflect even subtle alterations in photosynthetic electron transport (Supplementary Fig. S1). No differences were observed between crr9-1 and wild-type plants in terms of relative electron transport rate, NPQ, 1 – qL or F_v/F_m as previously reported in other crr mutants.

The CRR9 gene encodes a chloroplast stromal protein conserved in cyanobacteria and land plants

The crr9-1 mutant (Col gl1 background) was crossed with the polymorphic wild-type plant [Lansberg erecta (Ler)]. On the basis of the recessive nature of crr9-1, the crr9-1 locus was mapped on chromosome 4 between two single polymorphic sequences present at 11,343,840 and 11,603,892. Sequence analysis of the candidate genes with putative plastid targeting signals identified a nucleotide substitution in At4g21445, which consisted of two exons and one intron (Fig. 1C). The 713G to A mutation was located on the acceptor splicing junction on the intron in the crr9-1 allele (Fig. 1C). To confirm that the defect in NDH activity was caused by the mutation discovered in At4g21445, we obtained the crr9-2 mutant (SM_3_24042), in which At4g21445 was disrupted by a transposon insertion in the first exon (Fig. 1C). In crr9-2, the F_o rise in Chl fluorescence was impaired in the same way as in crr9-1 (Fig. 1B).

Expression of CRR9 was analyzed by reverse transcription–PCR (RT–PCR) (Fig. 1D). crr9-2 did not accumulate any transcripts, indicating that crr9-2 is a knockout allele. crr9-1 accumulated a higher level of the unspliced transcript than did the wild type, suggesting that the mutation in the splicing junction reduced the efficiency of splicing. Furthermore, spliced transcripts in crr9-1 migrated slightly faster than that in the wild type in agarose gel electrophoresis (Fig. 1D). To determine the splicing site in crr9-1, cDNAs originating from the spliced transcripts were cloned from crr9-1 and the wild type, and their nucleotide sequences were analyzed (Supplementary Fig. S2). In crr9-1, the AG sequence, which was 22 nucleotides downstream from the authentic splicing site, was used as an acceptor splicing site, resulting in a 22 nucleotide truncation and a frameshift (Supplementary Fig. S2A). As a result, crr9-1 was expected to produce a shorter protein with a different amino acid sequence in the C-terminal half of CRR9 (Supplementary Fig. S2B).

To verify that the crr9 phenotype was due to the mutations in At4g21445, the wild-type genomic sequence was introduced into crr9-1 (crr9-1+CRR9) and crr9-2 (crr9-2+CRR9). This transformation fully complemented the F_o rise in both mutant alleles (Fig. 1B), confirming that At4g21445 was CRR9.

CRR9 encoded a protein composed of 154 amino acids (Supplementary Fig. S2B). The first 53 amino acids were predicted by ChloroP 1.1 Server (http://www.cbs.dtu.dk/services/ChloroP/) to be a plastid targeting signal. The molecular mass and pI of mature CRR9 were predicted to be 11.4 kDa and 4.49, respectively, suggesting that CRR9 was a small acidic protein. Domain prediction by Pfam 29.0 (http://pfam.xfam.org/) and SMART (http://smart.embl-heidelberg.de/) indicated that CRR9 had no known motifs or transmembrane domains. Secondary structure prediction by Jpred 4 (http://www.compbio.dundee.ac.uk/jpred/) indicated that CRR9 was an α-helical protein. BLAST analysis revealed that CRR9 orthologs were conserved in vascular plants and cyanobacteria. Green algae such as Chlamydomonas reinhardtii and Ostreococcus tauri, which lack chloroplast NDH, have no CRR9 homologs (Supplementary Fig. S3).

For the analysis of cellular localization of CRR9, a specific antibody was raised against the recombinant protein of mature CRR9. We prepared the chloroplast stroma fraction and the membrane fraction containing thylakoid membranes from intact chloroplasts and performed immunoblot analysis (Fig. 2). CRR9 antibody detected a protein of approximately 11 kDa in the chloroplast stroma isolated from wild-type leaves, although the NDH complex localizes to the thylakoid membrane. This molecular mass was consistent with that predicted for the mature form of CRR9 after removal of the transit peptide (11.4 kDa). This protein was not detected in crr9-1. CRR9 is a stromal protein and is unlikely to be a subunit of the chloroplast NDH complex.

Fig. 2

Immunoblot analysis of chloroplast localization of CRR9. Freshly isolated chloroplasts from wild-type (WT), crr9-1 and crr1 leaves were further fractionated into membrane and stromal fractions. Stromal and membrane protein extracts were loaded on an equal Chl basis. The large subunit of ribulose 1,5-bisphosphate carboxylase/oxygenase (RbcL) and Cyt b₆ were detected as loading and fractionation controls, respectively.

CRR9 is a novel factor specifically required for accumulation of SubA

To address the effect of crr9 mutations on NDH complex accumulation, blue native (BN)-PAGE and immunoblot analyses of the NDH–PSI supercomplex were conducted. Chloroplast NDH is composed of five subcomplexes, and an NDH monomer further associates with two copies of the PSI supercomplex to form the >1 MDa NDH–PSI supercomplex via minor LHCI proteins (Peng et al. 2009). This NDH–PSI supercomplex can be separated from other thylakoid protein complexes by using BN-PAGE (Shimizu et al. 2008, Peng et al. 2008). As previously reported, in the wild type, the NDH–PSI supercomplex was detected as a high molecular weight band (band I) in the top position of the BN gel (Fig. 3A, B). Band I was absent in crr4-3, in which the NDH–PSI supercomplex was destabilized in the absence of the NdhD membrane subunit (Kotera et al. 2005). In ndhl/crr23, band I was replaced by band II, corresponding to the NDH–PSI supercomplex lacking SubA and SubE (Shimizu et al. 2008, Peng et al. 2009). SubA and SubE are essential for NDH activity but are not required to stabilize the other parts of the NDH–PSI supercomplex. As in ndhl/crr23, band I was replaced by band II in crr9-1 and crr9-2, suggesting that SubA and SubE were destabilized in both mutant alleles and that the accumulation of other parts of the NDH–PSI supercomplex was not affected (Fig. 3A, B). In complemented lines, only band I was detected, as in the wild type (Fig. 3A, B).

Fig. 3

Accumulation of chloroplast NDH complex in the thylakoid membrane. (A) Analysis of thylakoid protein complexes separated by BN-PAGE. Chloroplasts were isolated from various genotypes, as indicated at the top. Thylakoid protein extract corresponding to 10 μg of Chl was loaded onto each lane. BN-PAGE gel was stained with Coomassie Brilliant Blue. The positions of bands I and II are indicated. (B) Magnification of the top part of the BN-PAGE gel in (A). Positions of bands I and II are indicated. Asterisks indicate bands probably corresponding to aggregated proteins. (C) Immunodetection of NDH subunits and photosynthetic protein complexes. The thylakoid and stromal fractions were prepared from chloroplasts of various genotypes, as indicated at the top. Thylakoid protein extracts corresponding to 1 or 2 μg of Chl were loaded onto each lane as well as a dilution series of wild-type (WT) proteins. A 6 μg aliquot of stromal proteins was loaded onto each lane for CRR9 detection. Antibodies used are indicated on the right: NdhH and NdhK–M (NDH SubA); PnsB1 and PnsB2 (NDH SubB); PnsL4 (NDH SubL); NdhT (NDH SubE); PsbB, PsbC, and Lhcb1 (PSII); PsaA, PsaB, and Lhca2 (PSI); Cyt f (Cyt b₆f); AtpB (ATP synthase).

To confirm the accumulation of the band II complex, the BN gel was subjected to two-dimensional (2D) SDS–PAGE, and NDH subunits were probed using specific antibodies (Supplementary Fig. S4). Trace levels of NdhL and NdhT were detected in the high molecular weight position, suggesting that CRR9 was not absolutely necessary for the accumulation of the fully assembled NDH complex corresponding to band I. However, both proteins were mainly detected in low molecular weight positions probably as monomers. In contrast, PnsB1 and PnsL4 were detected in the high molecular weight position, consistent with the accumulation of the band II complex.

Immunoblot analyses using antibodies against various NDH subunits indicated that the accumulation of SubA subunits NdhH, K, L and M was substantially decreased in crr9 mutant alleles, as in ndhl/crr23 (Fig. 3C). The level of NdhT/CRRJ (SubE) was slightly reduced, but the levels of the other subcomplex subunits, PnsB1, PnsB2 (SubB) and PnsL4 (SubL), were unaffected. Identical levels of PsbB, PsbC and Lhcb1 (PSII), PsaA, PsaB and Lhca2 (PSI), Cyt f (Cyt b₆f) and AtpB (ATP synthase) were detected in crr9 and wild-type plants (Fig. 3C). The CRR9 level was not affected in ndhl/crr23, crr4-3 and crr1 (Figs. 2, 3C), suggesting that the stability of CRR9 was independent of the accumulation of NDH. NdhL is a subunit of SubA, whereas CRR1 is a non-subunit factor required for the assembly of NdhK, NdhM or both (Shimizu and Shikanai 2007, Shimizu et al. 2008, Peng et al. 2012). In both mutants, accumulation of SubA was specifically impaired. In contrast, the NDH complex was totally destabilized in crr4-3, which is defective in the expression of ndhD encoding a SubM subunit (Kotera et al. 2005). These results indicate that CRR9 is a novel non-subunit factor specifically required for the accumulation of SubA.

CRR9 is necessary for NdhK accumulation in the stroma

Modular assembly of hydrophilic SubA subunits occurs in the stroma, and final assembly with the membrane-embedded part of the NDH complex occurs in the thylakoid membrane with the aid of the assembly factor CRR7 (Peng et al. 2012). In mutants defective in the assembly factors required for SubA assembly, the accumulation of hydrophilic SubA subunits in NAIs in the stroma is significantly disturbed (Peng et al. 2010, Peng et al. 2012). To test whether CRR9 is required for the modular assembly of SubA, the accumulation of stroma-localized SubA subunits was analyzed by immunoblotting (Fig. 4A). In crr9-1, accumulation of hydrophilic SubA subunits was drastically reduced in the thylakoid membrane (Fig. 3). Wild-type and crr7 plants accumulated similar levels of NdhH, J, K, M and O in the stroma, as previously reported (Peng et al. 2010, Peng et al. 2012), although the level of NdhI was slightly reduced in crr7. NdhN could be detected in the thylakoid membrane, but not in the stroma, consistent with the model in which NdhN is incorporated into the NDH complex when SubA interacts with the membrane-embedded part of the NDH complex (Peng et al. 2012). In crr9-1, NdhK accumulation in the stroma was reduced to 25% of the wild-type level (Fig. 4A). The crr9-1 mutant accumulated normal levels of other SubA subunits, including NdhJ, which is encoded by the same ndhC–K–J operon in plastid DNA as is NdhK (Matsubayashi et al. 1987). This result indicates that CRR9 is specifically required for NdhK accumulation in NAIs in the stroma.

Fig. 4

Accumulation of SubA subunits in the stroma. (A) Immunodetection of stromal SubA subunits in the wild type (WT), crr7 and crr9-1. Removal of lanes from a blot for NdhJ detection is indicated by a black line. (B) Immunodetection of stromal SubA subunits in the wild type, crr1 and ndhm. In (A) and (B), stromal proteins were isolated from leaves, and immunoblot analyses were performed with antibodies raised against SubA subunits. Antibodies used are indicated on the right. Stromal proteins were loaded onto each lane by equal protein content, as well as a dilution series of WT proteins. Thylakoid proteins from the WT and crr9-1 were loaded as controls.

NdhK accumulation depends on NdhM and CRR1 in the stroma (Peng et al. 2012). CRR1 is an assembly factor required for the incorporation of NdhK, NdhM or both into NAI500. In ndhm, the NdhK level was drastically reduced, although accumulation of the other SubA subunits NdhH–J and NdhO was not affected (Fig. 4B). Accumulation of NdhK depends on NdhM in the stroma, although NdhM accumulates in the absence of NdhK, as in crr9-1. In crr1, the levels of both NdhK and NdhM were reduced to about 50% of those in the wild type (Fig. 4B). This information, taken together, indicates that CRR9 is essential for the accumulation of NdhK in the stroma. CRR1 is required for the accumulation of stromal NdhM, although we cannot eliminate the possibility that CRR1 is also directly required for the accumulation of stromal NdhK.

NdhM is not incorporated into NAI400 in crr9

In the modular assembly of hydrophilic SubA subunits in the stroma, the formation of three high molecular weight NAI complexes, NAI800, NAI500 and NAI400—assembled in that order—was detected by 2D clear native (CN)/SDS–PAGE followed by immunoblot analysis using anti-NdhH antibody (Peng et al. 2010). Formation of NAI800 containing non-native NdhH before the folding of NdhH occurs with the aid of the Cpn60 complex that includes Cpn60β4 (Peng et al. 2011a). In subsequent assembly steps, native NdhH serves as a scaffold for other SubA subunits. In addition to NdhH, NAI500 contains at least NdhO and an assembly factor, CRR41. NAI500 is transformed into NAI400 by incorporating NdhI, J, K and M. CRR42 is required for NAI400 formation. CRR6 is required for the incorporation of NdhI into NAI400. At the last step of SubA assembly, CRR7 supports the incorporation of NdhN and the assembled SubA into the membrane-embedded part of the NDH complex (Peng et al. 2012).

Immunoblot analysis suggested that CRR9 was required for the incorporation of NdhK into NAI400 (Fig. 4). To test this idea, NAI complexes present in the stroma fraction were separated by 2D CN/SDS–PAGE and were detected by immunoblot using antibodies against hydrophilic SubA subunits (Fig. 5). In the stromal fractions from crr9-1 plants, three complexes, corresponding to NAI800, NAI500 and NAI400, were detected by using anti-NdhH antibody, as in the wild type (Fig. 5). In the wild type, NAI400 contained NdhH, I, J, K, M and O, as previously reported (Fig. 5; Peng et al. 2012). In crr9-1, however, NAI400 lacked NdhK and NdhM, and NdhM was detected only as a free protein (Fig. 5), suggesting that CRR9 is required for incorporation of NdhK and NdhM into NAI400. We also tested the possibility that CRR9 was incorporated into NAI complexes; the same blot was probed with antibody against CRR9 (Fig. 5). CRR9 was detected as free protein and did not co-migrate with any of the NAI complexes. An aliquot of NdhJ co-migrated with CRR9 in CN-PAGE in the wild type. Although both signals probably reflect monomers, we do not eliminate the possibility that CRR9 interacts with NdhJ in the stroma.

Fig. 5

Formation of NAI complexes in the stroma. Immunodetection of NAI complexes in the stroma from wild-type (WT) and crr9-1 chloroplasts. Stromal proteins were isolated from WT and crr9-1 chloroplasts and separated by 2D CN/SDS–PAGE. Positions of NAI complexes, NAI800, NAI500 and NAI400 are indicated by broken lines. Antibodies used for immunoblot analyses are indicated on the right. Stromal proteins were loaded by equal protein content.

Although CRR6 did not co-migrate with any NAI complexes in CN-PAGE, its transient interaction with NAI500 and NAI400 was confirmed by co-immunoprecipitation using antibodies raised against CRR6, CRR41 and CRR42 combined with highly sensitive proteomics in the form of liquid chromatography–tandem mass spectrometry (LC-MS/MS) (Peng et al. 2012). To investigate the proteins interacting with CRR9, crr9-1 was transformed with binary vectors to express CRR9 fusion proteins with a 4×cMyc tag or synthetic green fluorescent protein (sGFP) at the C-terminus. These transformations fully complemented in vivo NDH activity (Fig. 1B) and NDH–PSI supercomplex formation in crr9-1, indicating that the 4×cMyc tag and sGFP at the C-terminus did not affect the function of CRR9 (Supplementary Fig. S5A, B). CRR9 complexes were isolated by affinity chromatography with anti-cMyc tag or anti-GFP antibodies from the stromal fractions of complemented crr9-1. Co-purified proteins in the eluate were analyzed by LC-MS/MS (Supplementary Fig. S5C, D). Neither NDH subunits nor assembly factors were purified from wild-type plants, which were used as negative controls (Supplementary Fig. S5D). In crr9-1+35S;CRR9-sGFP plants, CRR9 was not concentrated in the eluate (data not shown). Although CRR9–4×cMyc fusion protein was successfully purified from the crr9-1+CRR9-4×cMyc plants with the highest Mascot scores, no NDH subunits or NDH assembly factors were detected in the eluate (Supplementary Fig. S5D; Supplementary Table S1). Co-purified proteins included components of the nuclear spliceosome (AT2G38770, AT2G47640 and AT3G03920), laccase (AT5G48100) and an unknown protein (AT4G30830), which were not predicted to be localized to chloroplasts (Supplementary Table S1). Although Cpn60α2 (AT5G18820) was a chloroplast protein, its assembly partners Cpn60α1 or Cpn60β1–4 were not detected. The low Mascot score did not strongly suggest that Cpn60α2 was associated with CRR9 in chloroplasts. From these results, we concluded that the mass analysis had detected contaminating proteins that were not associated with SubA assembly. In accordance with the results of the immunoblot and 2D CN/SDS–PAGE analyses, CRR9 might transiently interact with NdhK to stabilize it, or help with the incorporation of NdhK and NdhM into the NAI400 complex, or both.

Discussion

CRR9 is required for the accumulation of SubA of the thylakoid-localized NDH–PSI supercomplex (Fig. 3), although CRR9 is localized to the stroma (Fig. 2). The stromal localization of CRR9 is consistent with the results of recent proteomics analyses that detected CRR9 in the stroma (Peltier et al. 2006, Zybailov et al. 2008, Ferro et al. 2010). The NdhK level in the stroma was specifically and severely reduced in crr9 (Fig. 4A), suggesting that CRR9 is a non-subunit factor specifically required for the accumulation of NdhK in the stroma in NAIs.

What is the molecular function of CRR9 in SubA biogenesis? NdhK corresponds to Nqo6/NuoB, an Fe–S protein of the Q module of bacterial Complex I (Shikanai 2016). The ndhK gene is encoded in the ndhC–K–J operon in the plastid genome and co-transcribed with ndhC and ndhJ as tri-cistronic precursor RNA (Matsubayashi et al. 1987). ndhC and ndhJ encode a hydrophobic SubM subunit and a hydrophilic SubA subunit, respectively. In tobacco plants, disruption of ndhC by plastid transformation severely reduces the accumulation of the SubB subunits PnsB1 and PnsB2 in the thylakoid membrane (Takabayashi et al. 2009). CRR9 accumulated normal levels of these SubB subunits in the thylakoid membrane (Fig. 3C), and the NdhJ level in the stroma was also normal (Fig. 4A), suggesting that expression of the upstream ndhC or ndhJ was not affected in crr9. We cannot eliminate the possibility that CRR9 is involved in the translation of ndhK, but it is more plausible that CRR9 is involved in the post-translational step of NdhK biogenesis, as discussed below.

We did not observe the interaction between CRR9 and NdhK by CN-PAGE or affinity chromatography analyses in this study (Fig. 5B;Supplementary Fig. S5D). Furthermore, CRR9 has not been co-purified with other known assembly factors by co-immunoprecipitation using CRR1, CRR6, CRR7, CRR41 and CRR42 as baits (Peng et al. 2012). We also failed to observe the interaction between CRR9 and NdhK by yeast two-hybrid analysis (data not shown). However, interaction between the Synechocystis CRR9 homolog Ssr2781 and cyanobacterial NdhK has been indicated in a yeast cell in the IntAct molecular interaction database (http://www.ebi.ac.uk/intact/). On the basis of the pI values predicted from protein sequences, Arabidopsis CRR9 is negatively charged (pI = 4.49), whereas NdhK is positively charged (pI = 9.27). CRR9 may transiently interact with NdhK in the stroma via electrostatic interaction. It is also possible that CRR9 easily dissociates from NdhK during CN-PAGE or affinity chromatography. We do not eliminate the possibility that CRR9 does not interact with NdhK directly but is involved in the stability of NdhK via the function of the third unknown protein.

2D CN/SDS–PAGE and subsequent immunoblot analysis indicated that the formation of NAI800, NAI500 and NAI400 was not impaired in crr9 (Fig. 5). This result is consistent with the concept that CRR9 is not a core component of NAI complexes. However, NAI400 lacks NdhK and NdhM, and NdhM was detected as a free protein (Fig. 5). Among the nuclear-encoded SubA subunit mutants ndhl–ndhO, only the ndhm mutant severely reduced the NdhK level in the stroma (Fig. 4B; Peng et al. 2012). Recently, direct interaction between NdhK and NdhM in the NDH-1 complex of the cyanobacterium Synechocystis sp. PCC 6803 was indicated by pull-down and yeast two-hybrid assays (He et al. 2015). The results suggest that NdhK and NdhM interact in the NDH–PSI supercomplex and also in NAI400 in Arabidopsis. This idea is consistent with the fact that NdhK is unstable in the stroma in the absence of NdhM (Fig. 4; Peng et al. 2012). Although NdhK was required for the incorporation of NdhM into NAI400, NdhM was stable in the absence of NdhK in the stroma (Fig. 4A). NdhK is conserved in all NDH-related complexes as a core component of the Q module. In contrast, NdhM is specific to photosynthetic NDH and is likely to be needed to stabilize NdhK, even in assembly intermediates. CRR9 is likely to be involved in a similar process—stabilization of NdhK—in photosynthetic NDH but as a non-subunit assembly factor. A similar case is observed in the core component of NAIs. NdhH is conserved in all NDH-related complexes and its stability depends on NdhO specific to photosynthetic NDH in NAIs (Peng et al. 2012). CRR41 is a non-subunit assembly factor required to stabilize both NdhH and NdhO. Although CRR41 behaves as a scaffold for the further assembly steps, CRR9 probably transiently interacts with NdhK in the stroma.

The exact molecular function of CRR9 is still unclear, but the mutant phenotype is most closely related to that of crr1 (Shimizu and Shikanai 2007, Peng et al. 2012). However, accumulation of CRR9 in the stroma was not affected by the lack of CRR1 (Fig. 2), and vice versa (Fig. 3C), suggesting that the two proteins function independently in SubA assembly. Both assembly factors are required to incorporate NdhK and NdhM into NAI400. Because CRR1 is co-purified with CRR41, the putative protein complex including NdhK, NdhM and CRR1 transiently interacts with NAI500 (Peng et al. 2012). In contrast, the interaction of CRR9 with NdhK has been suggested only in cyanobacteria. CRR9 may be involved in the folding of NdhK before the interaction of NdhK with NdhM and CRR1.

Materials and Methods

Plant materials and growth conditions

Arabidopsis thaliana wild type (Col gl1) and mutants were grown in soil in a growth chamber (50 μmol photons m^–2 s^–1, 16 h photoperiod, 23 °C) for 3–4 weeks. crr9-1 was mutagenized by using ethyl methanesulfonate (Hashimoto et al. 2003). The SIGnAL T-DNA express database (http://signal.salk.edu/cgi-bin/tdnaexpress) was used to find T-DNA insertion mutants for crr9. The transposon insertion line SM_3_24042 (crr9-2) was provided by The Arabidopsis Information Resource (TAIR) (http://www.arabidopsis.org/index.jsp).

Map-based cloning

The crr9-1 mutation was mapped with molecular markers based on a cleaved amplified polymorphic sequence (CAPS) (Konieczny and Ausubel 1993). Genomic DNA was isolated from F₂ plants derived from a cross between crr9-1 and the wild type (Ler). Genomic sequences of candidate genes from crr9-1 were amplified by PCR using ExTaq DNA polymerase (TAKARA BIO INC.). The resulting PCR product was directly sequenced with a dye terminator cycle sequencing kit and an ABI PRISM 3100 sequencer (Perkin-Elmer).

In vivo Chl fluorescence analysis

Chl fluorescence was measured with a MINI-pulse-amplitude modulation portable Chl fluorometer (MINI-PAM, Walz) as described previously (Yamamoto et al. 2011). A saturating pulse of white light (800 ms, 8,000 μmol photons m⁻² s⁻¹) was applied to determine the maximum fluorescence at closed PSII centers in the dark-adapted state (F_m) and during AL illumination (F_m'). The steady-state fluorescence level (F_s) was recorded during AL illumination (5–1,000 μmol photons m⁻² s⁻¹). Maximum quantum yield of PSII was calculated as F_v/F_m. NPQ was calculated as (F_m – F_m')/F_m'. The quantum yield of PSII, Y(II), was calculated as (F_m' – F_s)/F_m' (Genty et al. 1989). The relative electron transport rate was calculated as Y(II)×light intensity (μmol photons m⁻² s⁻¹). qL, the fraction of open PSII center, was calculated as [Y(II)/(1 – Y(II)] × [(1 – F_v/F_m)/(F_v/F_m)] × (NPQ + 1) (Kramer et al. 2004, Miyake et al. 2009). The transient increase in Chl fluorescence after AL had been turned off was monitored as described in the legend to Fig. 1 (Shikanai et al. 1998).

RT–PCR analysis

Expression of CRR9 was analyzed by RT–PCR. Total RNA was prepared from rosette leaves by using an RNeasy plant mini kit (QIAGEN). Contaminating DNA was digested with DNase I. Total RNA (2 μg) was reverse transcribed with random hexamers with PrimeScript for RT–PCR (TAKARA BIO INC.) in a total volume of 20 μl. After 10-fold dilution of the reaction mixture, a 1 μl aliquot containing cDNA was used for subsequent PCR with TAKARA Ex Taq DNA polymerase (TAKARA BIO INC.). PCRs were performed in a final volume of 50 μl containing 2.5 U of DNA polymerase and 10 pmol of each primer. PCR primer sets used for amplification of CRR9 and ACT8 were as follows. For CRR9, 5'-ATGAACATTGCCATTCATAATTACCGTCAC-3' and 5'-TTAAAAAGGAAAATCGGGATCATCCCATTG-3'; and for ACT8, 5'-GGTGCTGAGAGATTCAGGTGCCCAG-3' and 5'-AAGAGCGAGAGCGGGTTTTCAAACC-3'. PCRs consisted of 30 s denaturation at 94°C, 20 s annealing at 60°C and 90 s extension at 72°C. RT–PCR products were separated on a 1.4% agarose gel and detected by ethidium bromide staining. The number of cycles was optimized so that the abundance of products could be compared within the linear phase of amplification. The cDNAs amplified by RT–PCR from spliced and non-spliced CRR9 transcripts in crr9-1 were recovered from the agarose gels and cloned into pGEM-T Easy (Promega). DNA sequences of three independent cDNA clones for non-spliced transcripts and 13 independent cDNA clones for spliced transcripts were confirmed by DNA sequencing.

Complementation of crr9

The wild-type genomic sequence containing CRR9 was amplified by PCR with the primers 5'-ggggacaagtttgtacaaaaaagcaggctTGTGGAACCTTTCGAAAACCCTACG-3' and 5'-ggggaccactttgtacaagaaagctgggtTTTTCCTAGTCAGACCACAATAGAC-3' (lower case indicates attB1 or attB2 sequences). The amplified DNA fragments were cloned into pDONR/Zeo (Invitrogen) by using the BP clonase reaction (Invitrogen) and then transferred to the binary vector pGWB1 (Nakagawa et al. 2007) by LR clonase reaction (Invitrogen). To construct CRR9-4×cMyc, the genomic sequence containing CRR9 lacking the translational termination codon was amplified with the primers 5'-ggggacaagtttgtacaaaaaagcaggctTGTGGAACCTTTCGAAAACCCTACG-3' and 5'-ggggaccactttgtacaagaaagctgggtAAAAAGGAAAATCGGGATCATCCC-3'. The amplified sequence was finally cloned into the binary vector pGWB16 as described above. To construct 35S;CRR9-sGFP, cDNA containing the 5'-untrranslated region and the CRR9 coding sequence was amplified by RT–PCR using the primers 5'-ggggacaagtttgtacaaaaaagcaggctGTTATCCTCATCTAGTCATCTTCAC-3' and 5'-ggggaccactttgtacaagaaagctgggtAAAAAGGAAAATCGGGATCATCCC-3'. The amplified sequence was finally cloned into the binary vector pGWB5 (Nakagawa et al. 2007), as described above. The resultant vector was transformed into Agrobacterium tumefaciens C58 by electroporation, and the bacteria were used to transform crr9-1 and crr9-2 plants by using the floral dip method (Martinez-Trujillo et al. 2004). Transformed plants were selected on solidified 1/2 Murashige and Skoog medium containing 50 μg ml^–1 kanamycin. T₃ generations of transgenic lines were used for physiological and biochemical analyses.

Production of polyclonal antisera against CRR9

To express His-tagged fusion protein composed of the tandem repeat of the CRR9 sequence, cDNA encoding the mature form of CRR9 (amino acids Ala54–Phe154) was amplified by PCR with the primers 5'-aaggagatatacatatgGCAAGTGCCGAGACCGGAGAAGAAG-3' and 5'-gctcgaattcggatccAAAGGAAAATCGGGATCATCCC-3'. Lower case letters indicate sequences involved in homologous recombination in the subsequent cloning reaction. The amplified DNA fragments were cloned into pET22b (Novagen) digested with NdeI and BamHI by using an In-Fusion HD cloning kit (Clontech). The cDNA for the mature form of CRR9 was amplified by PCR with the primers 5'-tcgaaggtaggcatatgGCAAGTGCCGAGACCGGA-3' and 5'-gcaggtcgacaagcttTTAAAAAGGAAAATCGGGA-3' and cloned into the resultant plasmid digested with BamHI and XhoI by using the In-Fusion HD kit. Escherichia coli Rosetta (DE3) pLysS cells (Novagen) were transformed with the resultant plasmid. The transformed E. coli cells were grown at 37 °C for 8 h in 500 ml of Terrific broth (1.2% tryptone, 2.4% yeast extract, 0.4% glycerol, 72 mM K₂HPO₄ and 17 mM KH₂PO₄), supplemented with 100 μg ml^–1 carbenicillin and 30 μg ml^–1 chloramphenicol, in the presence of 1.5 mM isopropyl-β-d-thiogalactopyranoside to induce the expression of His-tagged tandem CRR9 protein. The cells were collected by centrifugation, and the collected cells were disrupted by sonication for 10 min on ice in 25 ml of binding buffer [50 mM Tris–HCl (pH 7.5), 5 mM imidazole, 2 mM MgCl₂, 10 mM dithiothreitol, 200 mM NaCl and cOmplete Protease Inhibitor Cocktail (Roche)]. The lysate was centrifuged at 12,000×g for 10 min at 4 °C, and then the supernatant was centrifuged at 48,000×g for 1 h at 4 °C. The resulting supernatant was filtered through a 0.45 μm filter and incubated for 2 h at 4 °C with 2 ml of cOmplete His-Tag Purification Resin (Roche) equilibrated with binding buffer. The resin was loaded into a plastic column and washed with 50 ml of binding buffer plus 0.2% Triton X-100. After washing off the resin with 50 ml of washing buffer [50 mM Tris–HCl (pH 7.5), 6 M urea, 5 mM imidazole, 5 mM dithiothreitol, 200 mM NaCl and 0.2% Triton X-100], bound His-tagged fusion proteins were eluted from the resin with 5 ml of elution buffer [50 mM Tris–HCl (pH 7.5), 4 M urea, 400 mM imidazole and 200 mM NaCl]. The purity of the fusion proteins was examined by SDS–PAGE, and the proteins were used as antigens.

Isolation of intact chloroplasts and chloroplast membranes

Intact chloroplasts were purified from leaves of 3- to 4-week-old plants as previously described (Munekage et al. 2002). The purified chloroplasts were ruptured in a buffer [20 mM HEPES-KOH (pH 7.6), 5 mM MgCl₂ and 2.5 mM EDTA]. The insoluble fraction containing thylakoids and envelopes was separated from the soluble fraction by centrifugation for 10 min at 15,000×g. The stromal fraction was obtained by further centrifugation of the soluble fraction at 15,000×g for 10 min. The concentration of Chl was determined as described previously (Porra et al. 1989). The concentration of protein was determined as reported by Bradford (1976), using bovine serum albumin as the standard.

SDS–PAGE and immunoblot analyses

Proteins separated by 12.5% (w/v) SDS–PAGE or Tricine-SDS–PAGE (for CRR9 detection) were electrotransferred onto polyvinylidene fluoride membranes. The antibodies were added, and the protein–antibody complexes were labeled by using an ECL Prime Western blotting detection system (GE Healthcare). Chemiluminescence was detected with an LAS3000 lumino-image analyzer (Fujifilm) and analyzed with Multi Gauge Version 3.0 software (Fujifilm).

BN- and CN-PAGE analyses

BN-, CN- and 2D-PAGE were performed as described before (Shimizu et al. 2008, Peng et al. 2012). Thylakoid membranes solubilized with 1% (v/v) β-dodecyl-maltoside corresponding to 10 μg of Chl were subjected to 5–12% BN-PAGE. Images of BN-PAGE gels were captured with an image scanner, and the gels were stained with Bio-Safe Coomassie stain (Bio-Rad). For 2D-PAGE analysis, stromal proteins were subjected to CN-PAGE, and lanes excised from CN gels were applied to 12.5% SDS–PAGE.

Co-immunoprecipitation and LC-MS/MS analysis

Chloroplast stromal fractions were prepared from wild-type, crr9-1+35S;CRR9-sGFP and crr9-1+CRR9-4×cMyc plants, as described above. Co-immunoprecipitation of CRR9-4×cMyc or CRR9–sGFP from the stromal fraction (300 μg protein) was done with a μMACS c-myc isolation kit or μMACS GFP isolation kit (Miltenyi Biotech) as previously reported (Peng et al. 2012). The eluate containing the CRR9 complex was subjected to 5–20% SDS–PAGE, and the gel was stained with SYPRO Ruby Protein Gel Stain (Bio-Rad). After in-gel digestion of proteins with trypsin (Promega), the digested peptides were recovered from the gel pieces and subjected to LC-MS/MS analyses using LTQ-Orbitrap XL MS (Thermo Scientific) as previously reported (Peng et al. 2012). MS/MS spectra were compared by using the Mascot server (version 2.4) against TAIR10 (TAIR), with the following search parameters: set-off threshold 0.05 for the ion-score cut-off; peptide tolerance, 10 p.p.m.; MS/MS tolerance, 0.5 Da; peptide charge, monoisotopic; trypsin as the enzyme; carboxymethylation on cysteine residues as a fixed modification; and oxidation of methionine as a variable modification.

Funding

This study was supported by the Japan Science and Technology Agency [grants from the CREST program to T.S.]; the Japan Society for the Promotion of Science [25251032 to T.S.]; the Network of Centers of Carbon Dioxide Resource Studies in Plants [to T.S.]; the National Natural Science Foundation of China [31270289 to L.P].

Acknowledgments

We thank Yuki Okegawa for her contribution to mapping the mutation. We are grateful to Amane Makino (Tohoku University), Tsuyoshi Endo (Kyoto University) and Hualing Mi (Shanghai Institutes for Biological Sciences, Chinese Academy of Science) for their gifts of antibodies. We also thank Tsuyoshi Nakagawa (Shimane University) and TAIR for providing the pGWB vectors and the transposon insertion line SM_3_24042.

Disclosures

The authors have no conflicts of interest to declare.

Abbreviations

Abbreviations
 
• AL

  actinic light

 
• BN

  blue native

 
• CET

  cyclic electron transport around PSI

 
• cMyc

  avian myelocytomatosis virus oncogene cellular homolog

 
• CN

  clear native

 
• Cpn60

  chaperonin 60

 
• crr

  chlororespiratory reduction

 
• ΔpH

  trans-thylakoid proton gradient

 
• Fd

  ferredoxin

 
• LC-MS/MS

  liquid chromatography–tandem mass spectrometry

 
• NAI

  NDH assembly intermediate

 
• NDH

  NADH dehydrogenase-like (complex)

 
• PGR5

  PROTON GRADIENT REGULATION 5

 
• PGRL1

  PGR5-like Photosynthetic Phenotype 1

 
• PQ

  plastoquinone

 
• RT–PCR

  reverse transcription–PCR

 
• sGFP

  synthetic green fluorescent protein

 
• Sub

  subcomplex