NDH-PSI Supercomplex Assembly Precedes Full Assembly ......NDH-PSI Supercomplex Assembly Precedes Full Assembly of the NDH Complex in Chloroplast1 Yoshinobu Kato, Kazuhiko Sugimoto,

NDH-PSI Supercomplex Assembly Precedes FullAssembly of the NDH Complex in Chloroplast1

Yoshinobu Kato, Kazuhiko Sugimoto, and Toshiharu Shikanai2

Department of Botany, Graduate School of Science, Kyoto University, Sakyo-ku, Kyoto 606-8502, Japan

ORCID ID: 0000-0002-6154-4728 (T.S.).

The chloroplast NADH dehydrogenase-like (NDH) complex is structurally similar to respiratory complex I and mediates PSIcyclic electron flow. In Arabidopsis (Arabidopsis thaliana), chloroplast NDH is composed of at least 29 subunits and associateswith two copies of PSI to form the NDH-PSI supercomplex. Here, we found that CHLORORESPIRATORY REDUCTION3(CRR3) is an assembly factor required for the accumulation of subcomplex B (SubB) of chloroplast NDH. In Suc densitygradient centrifugation, CRR3 was detected in three protein complexes. Accumulation of the largest peak III complex wasimpaired in mutants defective in the SubB subunits PnsB2-PnsB5. The oligomeric form of CRR3 likely functions to assemblethe core of SubB to form the peak III complex as an assembly intermediate. A defect in the PnsL3 subunit increased the level ofthe peak III complex, suggesting that CRR3 was released from the assembly intermediate after PnsL3 binding. Unlike PnsB2-PnsB5 and PnsL3, PnsB1 was not absolutely necessary for stabilizing SubB. PnsB1 is likely incorporated into the intermediateat the final step during SubB assembly. Lhca6 is a linker protein mediating NDH-PSI supercomplex formation, and its site ofcontact with NDH was suggested to be SubB. In the lhca6 mutant, accumulation of the peak III complex was impaired,suggesting that SubB interacted with Lhca6 during the step of SubB assembly. The process of supercomplex formation wastriggered before the completion of the NDH assembly. Consistent with its predicted function, CRR3 accumulated in youngleaves, where the NDH complex was assembled.

Oxygenic photosynthesis was established in the an-cestors of cyanobacteria about 2.5 to 3.0 billion yearsago. Its central machineries that form the photosyn-thetic electron transport chain are well conservedamong cyanobacteria and chloroplasts. In contrast,great diversity is observed in the systems for acclima-tion to light environments, such as light-harvestingantennae or the defense system against excess lightand other stress conditions. The chloroplast NADHdehydrogenase-like (NDH) complex mediates cyclicelectron flow around PSI (Munekage et al., 2004). Al-though its contribution is smaller than that of anotherpathway that depends on PROTON GRADIENTREGULATION5 (PGR5) and PGR5-like PhotosyntheticPhenotype1 (PGRL1; Munekage et al., 2002; DalCorsoet al., 2008), the NDH-dependent pathway generatesproton motive force across the thylakoid membrane,consequently contributing to ATP synthesis (Wanget al., 2015). The mutant phenotypes also suggest thatchloroplast NDH rather than PGR5/PGRL1 contributes

to the formation of proton motive force under certainconditions (Yamori and Shikanai, 2016). In a rice mu-tant defective in chloroplast NDH, PSII yield and plantgrowth were slightly impaired in low light (Yamoriet al., 2015). The plastoquinone (PQ) pool was slightlymore reduced in a Marchantia polymorpha mutant de-fective in chloroplast NDH, also suggesting that chlo-roplast NDH functions in low light (Ueda et al., 2012).In rice (Oryza sativa), chloroplast NDH is also importantin alleviating oxidative stress in fluctuating light(Yamori et al., 2016).

In land plants, the PSI core forms a supercomplex(PSI-LHCI) with four light-harvesting complex Is(LHCIs), namely Lhca1-Lhca4 (Qin et al., 2015). TheNDH complex was originally independent of PSI-LHCI, because it exists as a monomer in M. poly-morpha (Ueda et al., 2012). During land plant evolution,the NDH complex began interacting with PSI-LHCI. Inangiosperms, the NDH complex binds to two copies ofPSI-LHCI (Peng et al., 2009; Kou�ril et al., 2014). For-mation of this supercomplex is required to stabilize theNDH complex, especially in strong light, but it is notessential for NDH activity (Peng and Shikanai, 2011).Two linker proteins, Lhca5 and Lhca6, each mediatesupercomplex formation with a copy of PSI-LHCI(Peng et al., 2009).

The chloroplast NDH complex shares several ho-mologous subunits with bacterial NADH dehydro-genase andmitochondrial complex I (respiratory NDH;Shikanai, 2016). However, subunits forming theNADH:FMN (flavin mononucleotide) oxidoreductase(N) module have not been found in chloroplast NDH.

1 Y.K. was supported by the Japan Society for the Promotion ofScience as a research fellow (grant no. 17J09745). T.S. was supportedby grants from the Japan Science and Technology Agency (CRESTprogram), the Japanese Society for the Promotion of Science(25251032) and the Human Frontier Science Program.

2 Address correspondence to [email protected] author responsible for distribution of materials integral to the

findings presented in this article in accordance with the policy de-scribed in the Instructions for Authors (www.plantphysiol.org) is:Toshiharu Shikanai ([email protected]).

www.plantphysiol.org/cgi/doi/10.1104/pp.17.01120

1728 Plant Physiology�, February 2018, Vol. 176, pp. 1728–1738, www.plantphysiol.org � 2018 American Society of Plant Biologists. All Rights Reserved.

https://plantphysiol.orgDownloaded on April 6, 2021. - Published by Copyright (c) 2020 American Society of Plant Biologists. All rights reserved.

http://orcid.org/0000-0002-6154-4728http://orcid.org/0000-0002-6154-4728http://orcid.org/0000-0002-6154-4728http://orcid.org/0000-0002-6154-4728http://crossmark.crossref.org/dialog/?doi=10.1104/pp.17.01120&domain=pdf&date_stamp=2018-01-27mailto:[email protected]://www.plantphysiol.orgmailto:[email protected]://www.plantphysiol.org/cgi/doi/10.1104/pp.17.01120https://plantphysiol.org

Instead, chloroplast NDH is equipped with an electron-donor binding subcomplex (SubE) including NdhS/CHLORORESPIRATORY REDUCTION31 (CRR31),which forms a ferredoxin (Fd) binding surface(Yamamoto et al., 2011; Yamamoto and Shikanai, 2013).Chloroplast NDH complex mediates electron transportfrom Fd to PQ, rather than NAD(P)H-dependent PQreduction.Chloroplast NDH is the largest protein complex

(about 700 kD) in the photosynthetic electron transportchain (Peng and Shikanai, 2011). To date, 29 subunitshave been identified in the NDH complex. These sub-units are categorized into five subcomplexes (Ifukuet al., 2011; Supplemental Fig. S1). Subcomplex A(SubA) consists of NdhH-NdhK and NdhL-NdhO,encoded by the plastid genome and the nuclear ge-nome, respectively. NdhH-NdhK form the core of SubAand are homologous to the quinone-reducing (Q)module subunits that hold Fe-S clusters in respiratoryNDH. Subunits of the membrane subcomplex (SubM)(NdhA-NdhG) are encoded by the plastid genomeand are homologous to proton pumping (P) modulesubunits, although NuoH (NADH ubiquinone oxi-doreductase)/ND1, corresponding to NdhA, is cat-egorized as a linker between the Q and P modules.Subcomplex B (SubB; PnsB1-PnsB5), luminal sub-complex (SubL; PnsL1-PnsL5), and SubE (NdhS-NdhV)consist of subunits encoded by the nuclear genome(Ifuku et al., 2011). Subunits of SubB, SubL, and SubE areunique to the chloroplast NDH complex and are notconserved in cyanobacterialNDH-1,which is believed tobe the origin of chloroplast NDH (Peltier et al., 2016).The process of assembly of chloroplast NDH has

been well documented, as in the case with other proteincomplexes mediating photosynthetic or respiratoryelectron transport. Because of the low abundance of theNDH complex in the thylakoid membrane, some bio-chemical approaches, including pulse-chase experi-ments, are not feasible for analyzing NDH assembly.However, a combination of genetics and proteomicsmakes the chloroplast NDH complex a target of as-sembly research. Because even complete loss of thechloroplast NDH complex does not have secondaryeffects on other photosynthetic machineries undergrowth chamber conditions, mutant screening focusingon NDH activity has been very efficient in isolating thegenes involved in assembly (Hashimoto et al., 2003;Peng et al., 2012). This strategy is powerful and hasbeen used in studies of the protein complexes involvedin major electron transport in respiration and photo-synthesis (Rochaix, 2011; Mimaki et al., 2012). Once theassembly factors were isolated, theywere used to probethe assembly intermediates of the protein complexes.This strategy was used very successfully to clarify thesteps of SubA assembly in the chloroplast NDH com-plex (Peng et al., 2012). Most parts of SubA are assem-bled in the stroma with the help of multiple assemblyfactors and then inserted into the thylakoid membrane(Peng et al., 2012). The remaining parts of the NDH-PSIsupercomplex are assembled in the thylakoid

membrane. Although two assembly factors havebeen reported in SubB assembly (Ishida et al., 2009;Armbruster et al., 2013), our knowledge of assembly ofthe other parts is limited. Furthermore, we have not yetfigured out the process of NDH-PSI supercomplexformation in the thylakoid membrane. In Chlamydo-monas reinhardtii, other supercomplexes consisting ofPSI-LHCI, LHCII, Cytochrome (Cyt) b6f, and PGRL1,mediating PSI cyclic electron flow in chloroplasts hasbeen reported (Iwai et al., 2010). However, there hasnot been an extensive study of how such super-complexes are assembled in the thylakoid membranein chloroplasts.

The chlororespiratory reduction3 (crr3) mutant wasisolated on the basis of its specific defect in NDH ac-tivity by monitoring transient postillumination in-creases in chlorophyll fluorescence. CRR3 has onetransmembrane domain in its C-terminal region. Wehave discussed the possibility that CRR3 is an NDHsubunit (Muraoka et al., 2006). However, we have notdetected CRR3, even in a highly sensitive mass spec-troscopic analysis of the NDH-PSI supercomplex sep-arated by blue-native (BN)-PAGE (Peng et al., 2009). Inthis study, we analyzed themolecular function of CRR3and found that CRR3 was an assembly factor requiredfor the formation of an assembly intermediate of SubB.We also propose an assembly model for SubB anddiscuss the process of interaction between NDH subu-nits and PSI-LHCI via Lhca6, which primes NDH-PSIsupercomplex formation.

RESULTS

Optimization of Experimental Conditions forCRR3 Detection

In our previous study (Muraoka et al., 2006), wewereunable to determine conclusively the exact molecularfunction of CRR3. One problem was the low quality ofthe antibody used to detect CRR3. In this study, wereprepared an anti-CRR3 antibody with a titer suffi-cient for biochemistry (Supplemental Fig. S2A). In thewild type, we detected two signals specifically lackingin the crr3 mutant by using this antibody. In the pres-ence of Cys protease inhibitors (E-64 and leupeptin),only the upper signal was detected, suggesting thatCRR3was sensitive to protease (Supplemental Fig. S2B).The inhibitors were essential for obtaining reproduc-ible results and were always added to the bufferduring thylakoid preparation in our subsequentexperiments.

CRR3 Is Required for Accumulation of SubB

The crr3 mutation results in the destabilization of anNDH subunit, NdhH, in the thylakoid membrane(Muraoka et al., 2006). To characterize the crr3 defectmore extensively, the NDH-PSI supercomplex wasseparated by BN-PAGE in wild-type and crr3 plants

Plant Physiol. Vol. 176, 2018 1729

Assembly of the NDH-PSI Supercomplex


http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1https://plantphysiol.org

(Fig. 1A). In crr3, the band corresponding to the NDH-PSI supercomplex was not visible, whereas other majorprotein complexes accumulated normally, as in thewild type. Consistent with our previous findings(Muraoka et al., 2006), CRR3 was essential for accu-mulation of the NDH complex.

The NDH complex is divided into five subcomplexes,namely SubA, SubB, SubE, SubL, and SubM (Ifuku et al.,2011; Supplemental Fig. S1). In general, a defect in sub-unit accumulationmost strongly affects the accumulationof proteins present in the same subcomplex (Peng et al.,2009; Armbruster et al., 2013). To specify the subcomplexwhose accumulation was most strongly affected by thecrr3 defect, we quantified the accumulation of six NDHsubunits present in four subcomplexes (Fig. 1B). No an-tibodies were available for monitoring the accumulationof the subunits of SubM. For comparison, we monitoredthe stability of these subunits in four NDH-less mutants,pnsb2 (SubB), pnsl1 (SubL), ndhl (SubA), and ndht (SubE;Sirpiö et al., 2009; Ishihara et al., 2007; Peng et al., 2009;Takabayashi et al., 2009; Yamamoto et al., 2011). The levelof PnsB2 was reduced more severely in the crr3 mutantthan in the pnsl1, ndhl, and ndhtmutants, suggesting thatthe crr3 defect most strongly affected the accumulation ofSubB (Fig. 1B). Because PnsL3 is a PsbQ-like protein, itwas categorized into SubL (Suorsa et al., 2010; Yabutaet al., 2010). PsbQ is a PSII subunit localized on the lu-minal side (Ifuku et al., 2010). However, the stability ofPnsL3 is linked more tightly to the SubB subunits than tothe SubL subunits (Yabuta et al., 2010). As was observedin PnsB2, the levels of PnsL3 were more severely affectedin crr3 and pnsb2 mutants than in pnsl1, ndhl, and ndhtmutants (Fig. 1B). In contrast, the accumulation of PnsL4/FKBP16-2 (SubL) or NdhT (SubE) was not drasticallyaffected in crr3. The levels of NdhH and NdhL (SubA)were affected in crr3, but the effect was milder than thatobserved in the SubA mutant ndhl. This is probablyexplained by a secondary effect of the loss of SubB, be-cause the same phenotype was observed in mutants de-fective in SubB subunits (Fig. 1B; Peng et al., 2009). Insummary, the immunodetection pattern of crr3 was clos-est to that of pnsb2, with the exception of the signals fromthe antibodies raised against PnsB2 and CRR3 (Fig. 1B),suggesting that crr3 is one of the SubB mutants.

In our previous study (Muraoka et al., 2006), on thebasis of results suggesting that CRR3was absent in someNDH-less mutants, we discussed the possibility thatCRR3might be a subunit of theNDHcomplex.Whenweprepared thylakoid samples with protease inhibitors,however, the level of CRR3 was unaffected in the fourNDH-less mutants (Fig. 1B). These results, taken to-gether, indicate that CRR3 is required for SubB ac-cumulation but that the stability of CRR3 does notdepend on NDH subunits including PnsB2. All SubBsubunits (PnsB1-PnsB5) depend on each other forstability, and thus the abundance of SubB subunitswas severely decreased in the absence of any of thePnsB1-PnsB5 subunits (Peng et al., 2009; Armbrusteret al., 2013). Our new results do not support the ideathat CRR3 is a subunit of the NDH complex.

CRR3 Is Absent from the NDH-PSI Supercomplex

Mass spectrometry of the NDH-PSI supercomplexseparated by BN gel electrophoresis has not detected

Figure 1. Characterization of crr3. A, Protein complexes in thechloroplast membrane isolated from wild-type (WT) and crr3 leaveswere separated by using BN-PAGE (left). The gel was stained withCoomassie Brilliant Blue (CBB; right). Each band was identifiedaccording to methods used in previous works (Darie et al., 2005; Penget al., 2008) and is indicated by black arrows at right. B, Immunoblotanalysis of chloroplast membrane proteins in representative mutantsdefective in each subcomplex except for SubM. NDH subunits andCRR3 were detected by using their specific antibodies. Samples wereprepared from total leaves and loaded on the basis of chlorophyllcontent. Cytf was detected as a loading control.

1730 Plant Physiol. Vol. 176, 2018

Kato et al.


http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1https://plantphysiol.org

CRR3 (Peng et al., 2009). To test whether CRR3 was ab-sent from the NDH-PSI supercomplex, we separated thethylakoid protein complexes by using 2D-BN-sodiumdodecyl sulfate-polyacrylamide gel electrophoresis(SDS-PAGE) and immunodetected CRR3 in the blot(Supplemental Fig. S3). Consistent with the results of themass analysis, CRR3 was not detected at the position oftheNDH-PSI supercomplex butwas detected probably asa free protein. This result also suggests that CRR3 is not asubunit of the NDH complex.In BN-PAGE analysis, some proteins are dissociated

from the complexes during electrophoresis, because thecombination of neutral detergent (n-dodecyl-b-D-malto-side) and negatively charged dye (Serva Blue G) mimicsanionic detergent conditions (Pfeiffer et al., 2003). In fact,NdhS is an NDH subunit that forms an Fd binding site,but it is detectedmainly as a free protein onBN-PAGEgels(Yamamoto et al., 2011). To address this potential problem,we separated the protein complexes includingCRR3moregently by using Suc density gradient (SDG) ultracentrifu-gation. As visible green bands in the SDG, LHCII (mon-omer and trimer), PSII monomer, PSI-LHCI, and theNDH-PSI supercomplex were separated (SupplementalFig. S4). The antibody detected three peaks of CRR3 on theSDG (Fig. 2A). At least in peaks II and III, CRR3 likelyformed protein complexes. The peak I and II complexeshad Mrs similar to those of LHCII trimer and the mon-omeric PSII core, respectively. The Mr of the peak IIIcomplexwas higher than that of PSI-LHCI.Although thefractions containing CRR3 partly overlapped with thosecontaining PnsB1, the intact NDH-PSI supercomplexmigrated with the higher molecular mass fractions,where CRR3 was not detected (Fig. 2A). From thesefindings, taken together with the stability of CRR3 in-dependent of the SubB subunits (Fig. 1B), we concludethat CRR3 is not a subunit of the NDH complex.

Some SubB Subunits Are Required for Formation of thePeak III Complex of CRR3

CRR3 formed unknown protein complexes, whichwere smaller than the NDH-PSI supercomplex (Fig. 2A).Because the crr3 defect most severely affected the ac-cumulation of SubB subunits (Fig. 1B), CRR3 maytransiently interact with SubB subunits. We separatedthe protein complexes including CRR3 in the pnsb2 andpnsb3 mutant backgrounds by using SDG ultracentri-fugation (Fig. 2, B and C). PnsB2/NDF2/NDH45 andPnsB3/NDF4 are SubB subunits (Sirpiö et al., 2009;Takabayashi et al., 2009; Ifuku et al., 2011). In bothmutants, peaks I and II were detected, as in the wildtype, but peak III had almost disappeared. PnsB2 andPnsB3 were required to form the peak III complex. Inthe wild type, PnsB2 and PnsB3 existed mainly in theintact NDH-PSI supercomplex. It was not easy to detectthe peak III complex in the wild type by using anti-bodies raised against PnsB2 or PnsB3, because the in-tensive signals of the NDH-PSI supercomplex maskedthe peak III region in the corresponding SDG fractions.

We also observed disruption of the peak III complex inthe pnsb4 and pnsb5 mutants (Supplemental Fig. S5).The SubB subunits PnsB2-PnsB5 were required forformation of the peak III complex of CRR3.

A SubB Assembly Intermediate Including CRR3 IsAccumulated at High Levels in the pnsl3 Mutant

Although PnsL3 is categorized as a subunit of SubL,the stability of PnsL3 was severely affected in the crr3

Figure 2. Separation of the NDH-PSI supercomplex and complexesincluding CRR3. Protein complexes of the chloroplast membrane iso-lated fromwild-type (WT) (A), pnsb2 (B), pnsb3 (C), and pnsl3 (D) leaveswere separated by SDG centrifugation. The left side is the top of thecentrifugation tube. After the centrifugation, the SDG was divided into30 fractions from top to bottom. Proteins were concentrated by tri-chloroacetic acid precipitation and subjected to immunoblot analysis.Positions of the NDH-PSI supercomplex and PSI-LHCI are indicated byblack arrows. Peaks of CRR3 are indicated below the blots. The cen-trifugation tube and blotting patterns were fitted according to theCoomassie Brilliant Blue-stained gel (Supplemental Fig. S4). Asteriskindicates nonspecific signal. Red arrow in D indicates the fraction usedin stoichiometric analysis in Supplemental Figure S7. E, Immunoblotanalysis of chloroplast membrane proteins isolated fromWTand pnsl3.Samples were prepared from total leaves. Sample loading was based onchlorophyll content. Cytf was detected as a loading control.




http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1https://plantphysiol.org

mutant, as it was in other mutants defective in SubB(Yabuta et al., 2010; Fig. 1B). In this study, we redefinePnsL3 as a SubB subunit. To test the possibility thatCRR3 interacts with PnsL3, we performed SDG ultra-centrifugation and immunoblot analysis using pnsl3thylakoids (Fig. 2D). Three peaks of CRR3 were alsodetected in the SDG fractions of pnsl3. CRR3 wasdetected mainly in peak III in pnsl3, whereas itwas detected mainly in peaks I and II in the wild type.In the pnsl3 mutant, the total level of CRR3 was notaffected (Fig. 2E), indicating that the distribution ofCRR3 was shifted from the peak I/II complexes to thepeak III complex.

PnsB2-PnsB5 were required for formation of the peakIII complex (Fig. 2, B and C; Supplemental Fig. S5), im-plying that these SubB subunits formed the assemblyintermediate including CRR3 (peak III complex). If thiswere the case, SubB subunits would accumulate in thepeak III complex in pnsl3. In fact, PnsB2-PnsB4 weredetected in similar fractions as the peak III complex inpnsl3 (Fig. 2D). An exception was PnsB1, which wasdetected in lower-molecular-weight fractions. The peakIII complex is likely to be an assembly intermediate ofSubB. This complex is unlikely to be a fully assembledSubB, because PnsB1 and PnsL3 itself were absent fromthe complex detected in pnsl3 (Fig. 2D). Moreover, thepeak ofNdhL,whichwas a component of SubA,was notcoincident with that of PnsB2, indicating that the com-plete NDH monomer was not formed in the pnsl3 mu-tant (Supplemental Fig. S6). Hence, the peak III complexis not a fully assembled NDH monomer with CRR3.

To analyze the components of the peak III com-plex, we determined the stoichiometry of CRR3 andPnsB2 in the peak III fraction of pnsl3 (SupplementalFig. S7; Supplemental Table S1). Because theaccumulation level of CRR3 depended on leaf age(see Fig. 6), we did not determine the absolute level ofCRR3 in a mixture of leaves. Instead, we determinedthe ratio of CRR3 to PnsB2 to be 6.29 (SD, 1.28) in theSDG fractions containing the peak III complex. Al-though we cannot exclude the possibility that someCRR3 or PnsB2 proteins were dissociated from thepeak III complex during the ultracentrifugation,CRR3 likely functioned as an oligomer in the peak IIIcomplex.

PnsB1 Is Not a Core Component of SubB and Its AssemblyDepends on PnsB2-PnsB5

PnsB1 was detected only in low-molecular-weightfractions of SDG in pnsl3, whereas PnsB2-PnsB4 weredetected at the position of the peak III complex ofCRR3, probably reflecting the assembly intermediateincluding PnsB2-PnsB4 and oligomeric CRR3(Fig. 2D). It is possible that PnsB1 is incorporated intothe assembly intermediate at a relatively late stage ofSubB assembly. To test this possibility, we checked thestatus of assembly of PnsB2, PnsL3, and CRR3 in theabsence of PnsB1 by using SDG ultracentrifugation

and immunoblotting (Fig. 3A). In the pnsb1 mutant,PnsB2 and CRR3 were detected in the peak III frac-tions. PnsL3 also showed the similar peak with thepeak of PnsB2 and CRR3. This indicated that SubBwasassembled to a certain step beyond the PnsL3 incor-poration even in the absence of PnsB1.

Moreover, the impact of the pnsb1 defect on theaccumulation of other SubB subunits was milder thanthat of defects in the other SubB subunits (Fig. 3B;Supplemental Fig. S8). PnsB1 is unlikely to form a coreof SubB, which is essential for the stabilization of othersubunits.

We further analyzed PnsB1 assembly in pnsb2-pnsb5mutants (Fig. 3C). PnsB1 was detected in lower-molecular-weight fractions than that corresponding

Figure 3. Characterization of PnsB1. A, Protein complexes of the thy-lakoid membrane isolated from pnsb1 were separated by using SDGcentrifugation. PnsB2, PnsL3, and CRR3 were detected by immuno-blotting, as in Figure 2. Position of PSI-LHCI is indicated by a blackarrow. Asterisk indicates nonspecific signal. B, Matrix analysis of fiveSubB subunits under five mutant backgrounds. Five SubB subunits andCRR3 were detected by using specific antibodies. Samples were loadedon the basis of chlorophyll content, along with a dilution series of wild-type (WT) proteins. Cytf was detected as a loading control. C, Proteincomplexes of the chloroplast membrane isolated fromWTand pnsb2-5leaveswere separated by using SDGcentrifugation. PnsB1was detectedby immunoblotting, as in Figure 2. The centrifugation tube of the WT isshown as a representative pattern.


Kato et al.


http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1https://plantphysiol.org

to PSI-LHCI in these mutants, as in the pnsl3 mutant(Fig. 2D), whereas in the wild type it was detected inthe NDH-PSI supercomplex. PnsB1 assembly dependson all the other SubB subunits (PnsB2-PnsB5 andPnsL3).

The SubB Assembly Intermediate Including CRR3Interacts with Lhca6

The peak III complex including CRR3 was smallerthan the intact NDH-PSI supercomplex but slightlylarger than PSI-LHCI (Fig. 2A). In the NDH-PSIsupercomplex, the NDH complex interacts with twocopies of PSI-LHCI, one via Lhca5 and the other viaLhca6 (Peng et al., 2009). On the basis of its molecularsize, the peak III complex may include a single copy ofPSI-LHCI. To test this possibility, we separated thethylakoid protein complexes in the lhca5 or lhca6 singlemutant, which lacks a single copy of PSI-LHCI, andtheir double mutant lhca5 lhca6, in which NDH exists asa monomer (Peng and Shikanai, 2011). In the lhca5mutant, all of the peaks of CRR3were detected, as in thewild type (Fig. 4, A and B). In contrast, peak III of CRR3was not detected in the lhca6 and lhca5 lhca6 mutants(Fig. 4, C and D). The peak III complex did not includeLhca5, but most likely it included Lhca6. This resultsuggests that the SubB assembly intermediate contain-ing CRR3 includes Lhca6, although it is unclearwhether a single copy of PSI-LHCI is also associated viaLhca6.Notably, the level of CRR3 in the lhca6 and lhca5

lhca6 mutants was approximately double that in thewild type (Fig. 4E). To compensate for destabiliza-tion of the NDH complex due to the lhca6 defect,expression of CRR3 may have been activated inthe mutants. However, the transcription level ofCRR3 was not upregulated in the lhca6 mutant(Supplemental Fig. S9).To further test whether the peak III complex con-

tained Lhca6, we crossed the lhca6 mutant with thepnsl3 mutant, in which CRR3 was enriched in thepeak III complex including some SubB subunits. Inthe pnsl3 lhca6 double mutant, the level of CRR3 wasstill higher than that in the wild type, as observed inthe lhca6 single mutant (Supplemental Fig. S10B).However, CRR3 was no longer detected in the peakIII complex, and there was also an absence of PnsB2 inthe fractions corresponding to the peak III complex(Supplemental Fig. S10A). This result suggests thatPnsB2, CRR3, and Lhca6 are components of the samepeak III complex.

PnsB4 Is Not Incorporated into the Putative AssemblyIntermediate Including PnsB2 and PnsB3 in thecrr3 Mutant

To assess the role of CRR3 in SubB assembly, weanalyzed the crr3 mutant by using SDG centrifugation

(Fig. 5). In crr3, low levels of PnsB1 and PnsB2 weredetected in the fractions corresponding to the NDH-PSIsupercomplex. Formation of the NDH-PSI super-complex occurred in the absence of CRR3, but the effi-ciency of this formation was extremely low. PnsB3,PnsB4, and PnsL3 were not detected in the fractionscorresponding to the NDH-PSI supercomplex, proba-bly because of the low titers of the antibodies.

PnsB2 and PnsB3 formed peaks at the same higher-molecular-weight fractions than those containing PSI-LHCI in crr3 (Fig. 5B). The fractions were slightlysmaller than those containing the peak III complex ofCRR3 (Figs. 2–4). We could not detect PnsB4 in anySDG fractions of crr3 (Fig. 5B). In the absence of CRR3,PnsB2 and PnsB3 may have still interacted, but the as-sembly intermediate was unlikely to have containedPnsB4. PnsL3 and PnsB1 were detected mainly in thelower-molecular-weight fractions in crr3. This is prob-ably because the early steps of SubB assembly, includ-ing the incorporation of PnsB4, are retarded in theabsence of CRR3.

Figure 4. Separation of CRR3 complexes in lhca5 and lhca6mutants. Ato D, Protein complexes of the chloroplast membrane isolated fromwild-type (WT) (A), lhca5 (B), lhca6 (C), and lhca5 lhca6 (D) leaveswereseparated by SDG centrifugation. CRR3 was detected by immuno-blotting, as in Figure 2. (E) Immunoblot analysis of chloroplast mem-brane proteins isolated from WT, lhca5, lhca6, and lhca5 lhca6 leaves.Samples were prepared from total leaves. Sample loading was based onchlorophyll content. Cytf was detected as a loading control.





CRR3 Accumulates in the Early Leaf-Development Stages

Because CRR3 is an assembly factor of the NDHcomplex, the accumulation of CRR3 might depend onthe plant or leaf age. To test this possibility, leaves ofArabidopsis (Arabidopsis thaliana) plants grown for 28 dafter germination were allocated to four groupsaccording to leaf stage (Fig. 6A). Thylakoid proteinswere loaded on an equal chlorophyll basis onto SDS-PAGE gels. Total protein compositions and levels weresimilar among the four leaf stages, suggesting that se-nescence had not taken place, even in stage 4 (Fig. 6B).Consistently, the accumulation of PsaA (PSI subunit)and Cytf (Cyt b6f subunit) did not change dramatically(Fig. 6C). Although the level of PnsB1 was also constantin the course of leaf development, the CRR3 levelgradually decreased as the leaves became older. TheCRR3 level was not stoichiometric with the levels ofNDH subunits during the course of leaf development.This observation is consistent with the idea that CRR3 isnot a subunit but an assembly factor of the NDHcomplex. Indeed, we used young leaves (stages 1 and 2)in the SDG experiment to stably detect the peak III ofCRR3 (Figs. 2–4).

DISCUSSION

The Peak III Complex Is an Assembly Intermediate ofSubB Including Lhca6

Supplemental Table S2 summarizes the impact ofeach mutation on the accumulation of the CRR3complexes. Formation of the peak III complex of

CRR3 was impaired in pnsb2-pnsb5 mutants (Fig. 2;Supplemental Fig. S5). Most likely, CRR3 forms aputative SubB assembly intermediate includingPnsB2-PnsB5. This idea is consistent with the crr3phenotype being similar to that of the mutant de-fective in SubB subunits (Fig. 1). To show the directinteraction of CRR3 with SubB subunits, we coim-munoprecipitated CRR3 complexes by using anti-body against CRR3 in the pnsl3 mutant, in which thelevel of the peak III complex was higher than inthe wild type. However, we did not detect PnsB2 inthe precipitation (Supplemental Fig. S11).

In addition to the SubB core consisting of PnsB2-PnsB5, Lhca6, which was a linker protein for super-complex formation between NDH and PSI-LHCI,was required for accumulation of the peak III com-plex (Fig. 4). This finding can be explained by theidea that the peak III complex also includes Lhca6.Consistently, the peak III complex was disassembledin the lhca6 pnsl3 double mutant, despite the higherCRR3 level (Supplemental Fig. S10). In the crr2 mu-tant defective in SubM, SubB subunits migrated tothe same position in the BN gel as Lhca6 and a PSIcore protein, PsaA (Peng et al., 2009). On the basis ofthis observation, it was proposed that SubB formedthe contact site for Lhca6, and this link was stillmaintained in the absence of SubM. This is consistentwith the discovery that Lhca6 was also necessary forpeak III complex accumulation (Fig. 4). Although wefailed to detect PsaA in coimmunoprecipitation withthe anti-CRR3 antibody (Supplemental Fig. S11), thepeak III complex of CRR3 may contain the core ofSubB (PnsB2-PnsB5), which, as revealed by its mo-bility in SDG, attached to a single copy of PSI-LHCIvia Lhca6.

Because of the resistance of PnsB2/NDH45 to trypsindigestion of the thylakoidmembrane, Sirpiö et al. (2009)predicted that PnsB2 was partly buried in the mem-brane on the stromal side of the membrane. We do noteliminate the possibility that PnsB2 and PnsB3 are lo-calized on the luminal side of the thylakoid membrane.In contrast, PnsB1/NDH48 is likely localized on thestromal side of the thylakoid membrane, as suggestedpreviously (Sirpiö et al., 2009). In any case, PnsB2 andPnsB3, like CRR3 and Lhca6, may form a scaffold forSubB assembly. This idea is consistent with our previ-ous proposal that SubB forms the contact site for Lhca6(Peng et al., 2009). The stromal side of SubB may in-teract with the stromal loop of Lhca6, which is neces-sary and sufficient for the linker function of Lhca6(Otani et al., 2017).

The Assembly Model of SubB

On the basis of the accumulation of some putativeassembly intermediates in the different mutant back-grounds, we propose amodel for the SubB assembly. Inthe crr3 mutant, PnsB2 and PnsB3 were detected in thesame fractions of SDG, although any peaks of PnsB4

Figure 5. Analysis of SubB assembly intermediates in crr3. Proteincomplexes of chloroplast membrane isolated from wild-type (WT) (A)and crr3 (B) leaves were separated by using SDG centrifugation, as inFigure 3. SubB subunits were detected by using specific antibodies.Positions of PSI-LHCI and the NDH-PSI supercomplex are indicated byblack arrows. Asterisks indicate nonspecific signals.


Kato et al.


http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1https://plantphysiol.org

were not detected (Fig. 5). A high-molecular-weightcomplex including PnsB2 and PnsB3 was detected alsoin the pnsb4 mutant (Supplemental Fig. S12). Interactionof PnsB2 and PnsB3 is at least partly independent ofCRR3 and PnsB4. CRR3 function seems to be re-quired for the efficient incorporation of PnsB4, PnsB5, orboth into the putative assembly intermediate includingPnsB2 and PnsB3 as an assembly factor (Fig. 7, i). Wecould not determine whether this assembly interme-diate also included a single copy of PSI-LHCI viaLhca6. It is possible that the putative SubB assemblyintermediate includes only Lhca6 and interacts withPSI-LHCI at a later step of NDH-PSI supercomplexassembly.In the pnsl3 mutant, PnsB2-PnsB4 were detected in

the same fractions of SDG. Assembly of SubB pro-ceeded to the step including PnsB2-PnsB4 in the ab-sence of PnsL3 (Fig. 2D). In the pnsl3mutant, CRR3wasenriched in the peak III complex, although the majorityof CRR3 was detected in the peak I and II complexes inthe wild type (Fig. 2A). In our model, CRR3 is releasedfrom the assembly intermediate after recruiting PnsL3,possibly in oligomeric form (Fig. 7, iii). However, theactual order of events in the wild type cannot be de-duced from characterization of the mutants. We cannot

eliminate the possibility that the order of steps (i) to (iii)is flexible in the wild type.

We propose for the following reasons that incorpo-ration of PnsB1 is the final step of SubB assembly: (1)The process of SubB assembly proceeds to the step ofPnsL3 binding in the absence of PnsB1 (Fig. 3A); (2)Incorporation of PnsB1 into the high-molecular-weightcomplexes was arrested in other SubB subunit mutants(Fig. 3C); and (3) PnsB1 was not essential for the accu-mulation of other SubB subunits (Fig. 3B).

Function of CRR3

Trace levels of the NDH-PSI supercomplex accumu-lated in the crr3 mutant (Fig. 5B), indicating that CRR3was not absolutely necessary for assembly of the NDH-PSI supercomplex. Although the exact molecularfunction of CRR3 is still unclear, we propose that CRR3is required for efficient SubB assembly. In the crr3mutant, the NDH activity was no longer detectable intransient postillumination increases in chlorophyllfluorescence (Muraoka et al., 2006). The effect of theNDH defect was enhanced in the pgr5 mutant back-ground, in which the main pathway of PSI cyclic

Figure 6. Leaf-stage-dependent accumulationof CRR3. A, Wild-type (WT) seedling grown for28 d after germination (left). Detached leaveswere allocated to four groups according to leafstage, with the exception of cotyledons. B andC, Chloroplast membrane proteins of each leafstage were subjected to SDS-PAGE, and the gelwas stained by Coomassie Brilliant Blue (B) oranalyzed by immunoblotting (C). Sampleloading was based on chlorophyll content.




http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1https://plantphysiol.org

electron transport was impaired. The severe pheno-type observed in the crr3 pgr5 double mutant, as wasobserved in other double mutants (Munekage et al.,2004). CRR3 function is definitely required for theoperation of NDH-dependent PSI cyclic electron flowin vivo via the assembly of sufficient NDH-PSIsupercomplex.

Although the copy number of PnsB2 in the peak IIIcomplex was not experimentally determined, the stoi-chiometry of CRR3 to PnsB2 in the peak III fractions ofthe pnsl3 mutant was around six (Supplemental TableS1). The molecular mass of putative mature CRR3 waspredicted to be 18 kD. If the oligomeric form containedsix molecules of CRR3, its molecular mass would be108 kD. This size corresponds roughly to that of LHCIItrimer (;100 kD) and the peak I complex (Fig. 2A;Supplemental Fig. S3). Because peaks I and II of CRR3were detected in all the mutants analyzed in this study,both complexes were unlikely to have contained SubBsubunits or Lhca6.We speculate that the peak II complexcontains oligomeric CRR3 and some unknown factors.We also speculate that both complexes represent CRR3oligomer recycled from the SubB assembly intermediate(Fig. 7, iv). Consistent with this idea, the peaks of CRR3were shifted from I and II to III in the pnsl3 mutant, inwhich release of CRR3 from the assembly intermediatewas probably arrested (Figs. 2D and 7).

CRR3 accumulated mainly in younger leaves, andthe level of the protein decreased in the course of leafdevelopment (Fig. 6). Given that CRR3 is an assemblyfactor for NDH complex, our observation suggests thatde novo synthesis of the NDH complex occurs mainlyin immature leaves (Fig. 6A, stage 1). The rate of syn-thesis may slow when a leaf is fully expanded (Fig. 6A,

stages 3 and 4). Because the NDH subunit level wasconstant during leaf development, the NDH complexlooked fairly stable once it had been synthesized inimmature leaves.

Supercomplex Formation Can Occur Before theCompletion of NDH Assembly

In our model, Lhca6 is incorporated into the SubBassembly intermediate (Fig. 7). This model may be in-consistent with a previous observation in lhca6 andlhca5 lhca6 mutants (Peng and Shikanai, 2011). TheNDH complex is fully assembled in the form of partialsupercomplexes associated with either a single copy ofPSI-LHCI via Lhca5 (in the lhca6 mutant) or NDH freefrom PSI (in the lhca5 lhca6 mutant; Peng and Shikanai,2011). This finding indicates that SubB is fully assem-bled and incorporated into the main NDH complex inthe absence of Lhca6. This is not surprising, because theNDH complex is originally a monomer and does notform the supercomplex in M. polymorpha (Ueda et al.,2012). The interaction with Lhca6 is not essential forSubB assembly. On the other hand, our study suggeststhat, in angiosperms, the process of assembly of theNDH complex has been flexibly modified: The processof supercomplex formation via Lhca6 is triggered be-fore assembly of the entire NDH complex is completed.

MATERIALS AND METHODS

Plant Materials and Growth Conditions

Arabidopsis (Arabidopsis thaliana; Columbia gl1) was grown in soil in agrowth chamber (50 mmol photons m22 s21, 16-h photoperiod, 23°C) for

Figure 7. Assembly model of SubB. Oligomeric CRR3 is in the peak III complex, including PnsB2-B5 and Lhca6 (i and ii). Lhca6may interact with a single copy of PSI-LHCI in this complex. Release of CRR3may be coupledwith the incorporation of PnsL3 intothe peak III complex (iii). Because themodel is based on themutant phenotypes, the order of assembly in thewild type (WT)mightnot be correct in all parts. The peak I and peak II complexes likely represent CRR3 released from the assembly intermediate (iv).The peak II complex may contain some unknown factors and be reused for new cycles of the SubB assembly. Finally, PnsB1 isattached to themembrane-spanning part of SubB (v). This SubB assembly process is not absolutely dependent on CRR3, but CRR3is definitely required for efficient assembly. Lhca6 is dispensable for assembly, although it is essential for stabilizing the assembledNDH complex.


Kato et al.



4 weeks. The T-DNA insertion line SALK_203766 (pnsb5) was provided by theSalk Institute Genomic Analysis Laboratory.

Purification of Recombinant CRR3 and PnsB2 Proteins andAntibody Preparation

cDNAs encoding the soluble part of CRR3 (amino acid positions 55 to 143) ortruncated PnsB2 (amino acid positions 18 to 283) were amplified by PCR withsynthetic oligo nucleotides (see Supplemental Table S3). The amplified sequencewas digested with NdeI and XhoI and cloned into pET-22b(+) (Novagen). Ex-pression of the recombinant proteins was induced by 1 mM isopropyl b-D-thio-galactopyranoside at 37°C for 3 h in host Escherichia coli (E. coli) strain Rosetta(DE3) pLysS cells (Novagen). After induction, the cells were harvested in 20 mMpotassium phosphate buffer (pH 7.4) containing 40 mM imidazole, 500 mM NaCl,and cOmplete EDTA-free protease inhibitor cocktail (Roche). The inclusion bodieswere pelleted from the sonicated cells at 3000g for 30min and solubilized in 20mMpotassium phosphate buffer (pH 7.4) containing 40 mM imidazole, 500 mM NaCl,and urea. The concentration of urea was 4 M and 6 M for solubilizing recombinantCRR3 and PnsB2, respectively. The recombinant proteins were purified withNi-NTA Agarose (Qiagen) in accordance with the manufacturer’s protocol. BothHis-tagged fusion proteins were eluted with 20 mM potassium phosphate buffer(pH 7.4) containing 500 mM imidazole, 500 mM NaCl, and 4 M urea. Polyclonalantisera were raised against the purified CRR3 protein in a rabbit. For the stoi-chiometric analysis, the protein concentrations were determined by Bradfordassay using bovine serum albumin as a standard.

Chloroplast Membrane Preparation, BN-PAGE, CBBStaining, and Immunoblot Analysis

Chloroplasts were isolated as described (Munekage et al., 2002). To preparechloroplast membranes, purified chloroplasts were ruptured in 20 mM HEPES-KOH (pH 7.6) containing 5 mM MgCl2, 2.5 mM EDTA, 10 mM E-64, and 100 mMleupeptin. Insoluble fractions were separated by centrifugation at 15,000g for2 min. BN and 2D-PAGE were performed as described before (Shimizu et al.,2008). BN and SDS gels were stained with Bio-Safe Coomassie stain (Bio-Rad).For immunoblot analysis, membrane proteins were loaded on an equal chlo-rophyll basis. Signals were detected with ECL Prime Western Blotting Detec-tion Reagent (GE Healthcare) and visualized with an ImageQuant LAS4000(GE Healthcare).

SDG and Protein Precipitation

Young leaves (Fig. 6A, stages 1 and 2) were used for the SDG experiment.Chloroplast membranes were washed with 5 mM Tricine-NaOH (pH 8.0) con-taining 10 mM E-64 and 100 mM leupeptin and then solubilized with 5 mMTricine-NaOH (pH 8.0) containing 0.9% (w/v) n-dodecyl-b-D-maltoside, 10 mME-64, and 100 mM leupeptin. The concentration of chlorophyll was adjusted to1.0mg/mL. Chloroplast membraneswere dissolved for 5min on ice. Theywerethen loaded on the top of a linear Suc gradient (5%–40%) prepared with 25 mMMES-NaOH (pH 6.8) containing 5 mM MgCl2, 10 mM NaCl, 0.02% n-dodecyl-b-D-maltoside, 10 mM E-64, and 100 mM leupeptin. The photosynthetic com-plexes were separated by ultracentrifugation for 24 h by using an SW32.1-Tirotor (Beckman) at 28,700 rpm. The gradients were fractionated from top tobottom into 30 fractions by using a Gradient Fractionator (BIOCOMP). Proteinsfrom equal amounts of fractions were precipitated by using trichloroacetic acidand washed twice with 99% (v/v) ice-cold acetone. The pellets were dissolvedin 13 Laemmli buffer and used for further immunoblot analysis. Representa-tive results from at least three and two independent thylakoid preparations areshown in the main and supplemental figures, respectively.

Supplemental Data

The following supplemental materials are available.

Supplemental Figure S1. Structural model of the NDH-PSI supercomplex.

Supplemental Figure S2. CRR3 is susceptible to endogenous protease dur-ing sample preparation.

Supplemental Figure S3. Immunodetection of CRR3 in 2D-BN-SDS-PAGE.

Supplemental Figure S4. Separation of photosynthetic electron transportcomplexes by SDG ultracentrifugation.

Supplemental Figure S5. Separation of CRR3 complexes in pnsb4 andpnsb5 mutants.

Supplemental Figure S6. Separation of complexes including PnsB2 orNdhL in the pnsl3 mutant.

Supplemental Figure S7. Stoichiometric analysis of CRR3 and PnsB2 in thepeak III complex.

Supplemental Figure S8. Characterization of mutants defective in SubBsubunits including PnsL3.

Supplemental Figure S9. Expression analysis of CRR3 in wild type andlhca6.

Supplemental Figure S10. Separation of CRR3 and PnsB2 complexes andanalysis of the CRR3 accumulation in the pnsl3 lhca6 double mutant.

Supplemental Figure S11. Immunoprecipitation using antibody againstCRR3 in pnsl3 and crr3 mutants.

Supplemental Figure S12. Separation of PnsB2 and PnsB3 complexes inthe pnsb4 mutant.

Supplemental Table S1. Stoichiometric analysis of CRR3 and PnsB2 of thepeak III fraction in the pnsl3 SDG.

Supplemental Table S2. Impact of each mutation on the accumulation ofCRR3 complexes.

Supplemental Table S3. The primer sequences used for the expression ofrecombinant proteins in E. coli and qRT-PCR.

ACKNOWLEDGMENTS

The authors thank Tsuyoshi Endo for the lines pnsb1, pnsb2, pnsb3, and pnsb4and for antisera against PnsB1, PnsB2, PnsB3, and PnsB4; Kentaro Ifuku for thelines pnsl1 and pnsl3 and for antisera against PnsL3; Hualing Mi for antiseraagainst NdhH; and AmaneMakino for antisera against Cytf. The authors thankHiroshi Yamamoto for helpful discussions and Tetsuki Kuniyoshi for technicaladvice.

Received August 9, 2017; accepted November 30, 2017; published December 4,2017.

LITERATURE CITED

Armbruster U, Rühle T, Kreller R, Strotbek C, Zühlke J, Tadini L,Blunder T, Hertle AP, Qi Y, Rengstl B, et al (2013) The photosynthesisaffected mutant68-like protein evolved from a PSII assembly factor tomediate assembly of the chloroplast NAD(P)H dehydrogenase complexin Arabidopsis. Plant Cell 25: 3926–3943

DalCorso G, Pesaresi P, Masiero S, Aseeva E, Schünemann D, Finazzi G,Joliot P, Barbato R, Leister D (2008) A complex containing PGRL1 andPGR5 is involved in the switch between linear and cyclic electron flow inArabidopsis. Cell 132: 273–285

Darie CC, Biniossek ML, Winter V, Mutschler B, Haehnel W (2005) Iso-lation and structural characterization of the Ndh complex from meso-phyll and bundle sheath chloroplasts of Zea mays. FEBS J 272: 2705–2716

Hashimoto M, Endo T, Peltier G, Tasaka M, Shikanai T (2003) A nucleus-encoded factor, CRR2, is essential for the expression of chloroplast ndhBin Arabidopsis. Plant J 36: 541–549

Ifuku K, Endo T, Shikanai T, Aro E-M (2011) Structure of the chloroplastNADH dehydrogenase-like complex: nomenclature for nuclear-encodedsubunits. Plant Cell Physiol 52: 1560–1568

Ifuku K, Ishihara S, Sato F (2010) Molecular functions of oxygen-evolvingcomplex family proteins in photosynthetic electron flow. J Integr PlantBiol 52: 723–734

Ishida S, Takabayashi A, Ishikawa N, Hano Y, Endo T, Sato F (2009) Anovel nuclear-encoded protein, NDH-dependent cyclic electron flow 5,is essential for the accumulation of chloroplast NAD(P)H dehydrogen-ase complexes. Plant Cell Physiol 50: 383–393




http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1http://www.plantphysiol.org/cgi/content/full/pp.17.01120/DC1https://plantphysiol.org

Ishihara S, Takabayashi A, Ido K, Endo T, Ifuku K, Sato F (2007) Distinctfunctions for the two PsbP-like proteins PPL1 and PPL2 in the chloro-plast thylakoid lumen of Arabidopsis. Plant Physiol 145: 668–679

Iwai M, Takizawa K, Tokutsu R, Okamuro A, Takahashi Y, Minagawa J(2010) Isolation of the elusive supercomplex that drives cyclic electronflow in photosynthesis. Nature 464: 1210–1213

Kou�ril R, Strouhal O, Nosek L, Lenobel R, Chamrád I, Boekema EJ, ŠebelaM, Ilík P (2014) Structural characterization of a plant photosystem I andNAD(P)H dehydrogenase supercomplex. Plant J 77: 568–576

Mimaki M, Wang X, McKenzie M, Thorburn DR, Ryan MT (2012) Un-derstanding mitochondrial complex I assembly in health and disease.Biochim Biophys Acta 1817: 851–862

Munekage Y, Hashimoto M, Miyake C, Tomizawa K, Endo T, Tasaka M,Shikanai T (2004) Cyclic electron flow around photosystem I is essentialfor photosynthesis. Nature 429: 579–582

Munekage Y, Hojo M, Meurer J, Endo T, Tasaka M, Shikanai T (2002)PGR5 is involved in cyclic electron flow around photosystem I and isessential for photoprotection in Arabidopsis. Cell 110: 361–371

Muraoka R, Okuda K, Kobayashi Y, Shikanai T (2006) A eukaryotic factorrequired for accumulation of the chloroplast NAD(P)H dehydrogenasecomplex in Arabidopsis. Plant Physiol 142: 1683–1689

Otani T, Yamamoto H, Shikanai T (2017) Stromal loop of Lhca6 is re-sponsible for the linker function required for the NDH-PSI super-complex formation. Plant Cell Physiol 58: 851–861

Peltier G, Aro E-M, Shikanai T (2016) NDH-1 and NDH-2 plastoquinonereductases in oxygenic photosynthesis. Annu Rev Plant Biol 67: 55–80

Peng L, Fukao Y, Fujiwara M, Shikanai T (2012) Multistep assembly of chloroplastNADH dehydrogenase-like subcomplex A requires several nucleus-encodedproteins, including CRR41 and CRR42, in Arabidopsis. Plant Cell 24: 202–214

Peng L, Fukao Y, Fujiwara M, Takami T, Shikanai T (2009) Efficient operationof NAD(P)H dehydrogenase requires supercomplex formation with photo-system I via minor LHCI in Arabidopsis. Plant Cell 21: 3623–3640

Peng L, Shikanai T (2011) Supercomplex formation with photosystem I isrequired for the stabilization of the chloroplast NADH dehydrogenase-like complex in Arabidopsis. Plant Physiol 155: 1629–1639

Peng L, Shimizu H, Shikanai T (2008) The chloroplast NAD(P)H dehy-drogenase complex interacts with photosystem I in Arabidopsis. J BiolChem 283: 34873–34879

Pfeiffer K, Gohil V, Stuart RA, Hunte C, Brandt U, Greenberg ML,Schägger H (2003) Cardiolipin stabilizes respiratory chain super-complexes. J Biol Chem 278: 52873–52880

Qin X, Suga M, Kuang T, Shen JR (2015) Photosynthesis. Structuralbasis for energy transfer pathways in the plant PSI-LHCI supercomplex.Science 348: 989–995

Rochaix J-D (2011) Assembly of the photosynthetic apparatus. PlantPhysiol 155: 1493–1500

Shikanai T (2016) Chloroplast NDH: A different enzyme with a structuresimilar to that of respiratory NADH dehydrogenase. Biochim BiophysActa 1857: 1015–1022

Shimizu H, Peng L, Myouga F, Motohashi R, Shinozaki K, Shikanai T(2008) CRR23/NdhL is a subunit of the chloroplast NAD(P)H dehy-drogenase complex in Arabidopsis. Plant Cell Physiol 49: 835–842

Sirpiö S, Allahverdiyeva Y, Holmström M, Khrouchtchova A, Haldrup A,Battchikova N, Aro E-M (2009) Novel nuclear-encoded subunits of thechloroplast NAD(P)H dehydrogenase complex. J Biol Chem 284: 905–912

Suorsa M, Sirpiö S, Paakkarinen V, Kumari N, Holmström M, Aro E-M(2010) Two proteins homologous to PsbQ are novel subunits of thechloroplast NAD(P)H dehydrogenase. Plant Cell Physiol 51: 877–883

Takabayashi A, Ishikawa N, Obayashi T, Ishida S, Obokata J, Endo T,Sato F (2009) Three novel subunits of Arabidopsis chloroplastic NAD(P)H dehydrogenase identified by bioinformatic and reverse genetic ap-proaches. Plant J 57: 207–219

Ueda M, Kuniyoshi T, Yamamoto H, Sugimoto K, Ishizaki K, Kohchi T,Nishimura Y, Shikanai T (2012) Composition and physiological func-tion of the chloroplast NADH dehydrogenase-like complex in March-antia polymorpha. Plant J 72: 683–693

Wang C, Yamamoto H, Shikanai T (2015) Role of cyclic electron transportaround photosystem I in regulating proton motive force. Biochim Bio-phys Acta 1847: 931–938

Yabuta S, Ifuku K, Takabayashi A, Ishihara S, Ido K, Ishikawa N, EndoT, Sato F (2010) Three PsbQ-like proteins are required for the function ofthe chloroplast NAD(P)H dehydrogenase complex in Arabidopsis. PlantCell Physiol 51: 866–876

Yamamoto H, Peng L, Fukao Y, Shikanai T (2011) An Src homology3 domain-like fold protein forms a ferredoxin binding site for the chlo-roplast NADH dehydrogenase-like complex in Arabidopsis. Plant Cell23: 1480–1493

Yamamoto H, Shikanai T (2013) In planta mutagenesis of Src homology3 domain-like fold of NdhS, a ferredoxin-binding subunit of the chlo-roplast NADH dehydrogenase-like complex in Arabidopsis: a conservedArg-193 plays a critical role in ferredoxin binding. J Biol Chem 288:36328–36337

Yamori W, Makino A, Shikanai T (2016) A physiological role of cyclicelectron transport around photosystem I in sustaining photosynthesisunder fluctuating light in rice. Sci Rep 6: 20147

Yamori W, Shikanai T (2016) Physiological functions of cyclic electrontransport around photosystem I in sustaining photosynthesis and plantgrowth. Annu Rev Plant Biol 67: 81–106

Yamori W, Shikanai T, Makino A (2015) Photosystem I cyclic electron flowvia chloroplast NADH dehydrogenase-like complex performs a physi-ological role for photosynthesis at low light. Sci Rep 5: 13908


Kato et al.

https://plantphysiol.org

NDH-PSI Supercomplex Assembly Precedes Full Assembly ......NDH-PSI Supercomplex Assembly Precedes Full Assembly of the NDH Complex in Chloroplast1 Yoshinobu Kato, Kazuhiko Sugimoto,

Documents