Efficient Operation of NAD(P)H Dehydrogenase Requires Supercomplex Formation with Photosystem I via Minor LHCI in Arabidopsis

Abstract

In higher plants, the chloroplast NAD(P)H dehydrogenase (NDH) complex mediates photosystem I (PSI) cyclic and chlororespiratory electron transport. We reported previously that NDH interacts with the PSI complex to form a supercomplex (NDH-PSI). In this study, NDH18 and FKBP16-2 (FK506 Binding Protein 16-2), detected in the NDH-PSI supercomplex by mass spectrometry, were shown to be NDH subunits by the analysis of their knockdown lines. On the basis of extensive mutant characterization, we propose a structural model for chloroplast NDH, whereby NDH is divided into four subcomplexes. The subcomplex A and membrane subcomplex are conserved in cyanobacteria, but the subcomplex B and lumen subcomplex are specific to chloroplasts. Two minor light-harvesting complex I proteins, Lhca5 and Lhca6, were required for the full-size NDH-PSI supercomplex formation. Similar to crr pgr5 double mutants that completely lack cyclic electron flow activity around PSI, the lhca6 pgr5 double mutant exhibited a severe defect in growth. Consistent with the impaired NDH activity, photosynthesis was also severely affected in mature leaves of lhca6 pgr5. We conclude that chloroplast NDH became equipped with the novel subcomplexes and became associated with PSI during the evolution of land plants, and this process may have facilitated the efficient operation of NDH.

Full text

Efficient Operation of NAD(P)H Dehydrogenase Requires
Supercomplex Formation with Photosystem I via Minor
LHCI in Arabidopsis W
Lianwei Peng,
a
Yoichiro Fukao,
b
Masayuki Fujiwara,
b
Tsuneaki Takami,
c
and Toshiharu Shikanai
a,1
a
Department of Botany, Graduate School of Science, Kyoto University, Sakyo-ku, Kyoto 606-8502, Japan
b
Plant Science Education Unit, Graduate School of Biological Sciences, Nara Institute of Science and Technology, Takayama,
Ikoma, Nara 630-0101, Japan
c
Graduate School of Agriculture, Kyushu University, Higashi-ku, Fukuoka 812-8581, Japan
In higher plants, the chloroplast NAD(P)H dehydrogenase (NDH) complex mediates photosystem I (PSI) cyclic and
chlororespiratory electron transport. We reported previously that NDH interacts with the PSI complex to form a super-
complex (NDH-PSI). In this study, NDH18 and FKBP16-2 (FK506 Binding Protein 16-2), detected in the NDH-PSI super-
complex by mass spectrometry, were shown to be NDH subunits by the analysis of their knockdown lines. On the basis of
extensive mutant characterization, we propose a structural model for chloroplast NDH, whereby NDH is divided into four
subcomplexes. The subcomplex A and membrane subcomplex are conserved in cyanobacteria, but the subcomplex B and
lumen subcomplex are specific to chloroplasts. Two minor light-harvesting complex I proteins, Lhca5 and Lhca6, were
required for the full-size NDH-PSI supercomplex formation. Similar to crr pgr5 double mutants that completely lack cyclic
electron flow activity around PSI, the lhca6 pgr5 double mutant exhibited a severe defect in growth. Consistent with the
impaired NDH activity, photosynthesis was also severely affected in mature leaves of lhca6 pgr5. We conclude that
chloroplast NDH became equipped with the novel subcomplexes and became associated with PSI during the evolution of
land plants, and this process may have facilitated the efficient operation of NDH.
INTRODUCTION
In higher plants, photosystem I (PSI) cyclic electron transport
consists of both NAD(P)H dehydrogenase (NDH)-dependent and
PROTON GRADIENT REGULATION5 (PGR5)-dependent path-
ways (Joe
¨t et al., 2001; Munekage et al., 2004; Shikanai, 2007a).
The PGR5-dependent (main) pathway is required for both pho-
tosynthesis and photoprotection (Munekage et al., 2002). Addi-
tionally, chloroplast NDH helps to prevent the overreduction of
stroma, especially under stress conditions (Munekage et al.,
2004; Shikanai, 2007a). The existence of NDH in chloroplasts
was first suggested by the complete sequencing of the two
plastid genomes in tobacco (Nicotiana tabacum) and liverwort
(Marchantia polymorpha), in which 11 genes encode the homo-
logs of subunits in the mitochondrial complex I and eubacterial
NADH dehydrogenase (Matsubayashi et al., 1987; Shikanai,
2007b). However, the genes encoding three key components,
NuoE to G, which function in NADH binding and oxidation in
Escherichia coli NDH-1, are missing in the cyanobacterial and
higher plant genomes.
Several candidates for the electron donor and electron input
module in NDH in chloroplasts and cyanobacteria have been
proposed. An NDH subcomplex with a molecular mass of ;550
kD isolated from pea (Pisum sativum) oxidized NADH (Sazanov
et al., 1998). A hydrophilic part of NDH was also shown to contain
NADH oxidizing activity (Rumeau et al., 2005). NADPH and
ferredoxin (Fd) as well as NADH can be oxidized by the NDH
complex in cyanobacteria (Mi et al., 1995). However, NDH
purified from cyanobacteria favors NADPH as an electron donor
(Matsuo et al., 1998). In contrast with these results, no differ-
ences were detected in NADH or NADPH oxidizing activity
between the wild type and NDH-less mutants in Arabidopsis
thaliana (Sirpio
¨et al., 2009a). Additionally, Guedeney et al. (1996)
proposed that Fd-NADP
+
reductase binds to chloroplast NDH.
Inconsistently, Fd was required for NDH-dependent plastoqui-
none (PQ) reduction in our assay using ruptured chloroplasts
(Munekage et al., 2004). To resolve this long-debated issue,
researchers have tried to identify the missing subunits. So far,
four NDH subunits, NdhL to O, have been found in cyanobacteria
and chloroplast NDH (Prommeenate et al., 2004; Battchikova
et al., 2005; Rumeau et al., 2005; Shimizu et al., 2008). In
addition, six subunits specific to higher plants have been iden-
tified: PsbP-like protein 2 (PPL2), NDH-DEPENDENT FLOW6
(NDF6), NDF1 (NDH48), NDF2 (NDH45), NDF4, and At CYP20-2
(Ishihara et al., 2007; Ishikawa et al., 2008; Majeran et al., 2008;
Sirpio
¨et al., 2009a, 2009b; Takabayashi et al., 2009). However,
these NDH subunits do not contain an NAD(P)H binding motif.
Although CRR1 has an NAD(P)H binding domain, it is localized to
1
Address correspondence to shikanai@pmg.bot.kyoto-u.ac.jp.
The author responsible for distribution of materials integral to the
findings presented in this article in accordance with the policy described
in the Instructions for Authors (www.plantcell.org) is: Toshiharu Shikanai
(shikanai@pmg.bot.kyoto-u.ac.jp).
W
Online version contains Web-only data.
www.plantcell.org/cgi/doi/10.1105/tpc.109.068791
The Plant Cell, Vol. 21: 3623–3640, November 2009, www.plantcell.org ã2009 American Society of Plant Biologists

the stroma and unlikely to be an NDH subunit (Shimizu and
Shikanai, 2007). Chloroplast NDH may accept electrons from
donors other than NAD(P)H.
It is believed that chloroplast NDH originated from cyanobac-
terial NDH-1 (Shikanai, 2007b). Proteomics studies revealed that
three types of NDH-1 exist in cyanobacteria: NDH-1L, NDH-1M,
and NDH-1S, with molecular masses of ;460, 350, and 200 kD,
respectively (Herranen et al., 2004). Mass spectrometry analysis
revealed that NDH-1M consists of 13 subunits, including a
membrane-embedded arm (NdhA to C, E, G, and L) and a
hydrophilic connecting domain (NdhH to K and M to O). NDH-1L
includes NdhD1 and NdhF1 in addition to the NDH-1M complex
(Prommeenate et al., 2004; Battchikova et al., 2005; Zhang et al.,
2005). The NDH-1S complex comprises NdhD3, NdhF3, CupA,
and CupS (Ogawa and Mi, 2007) and interacts with NDH-1M to
form the functional complex NDH-1MS, which is induced under
low CO
2
conditions (Zhang et al., 2005). While NDH-1L is
involved in respiratory and PSI cyclic electron transport, the
NDH-1MS complex is considered to be participated in CO
2
uptake in cyanobacteria (Battchikova and Aro, 2007; Ogawa and
Mi, 2007). Although several copies of NdhD and NdhF genes
were found in cyanobacteral genomes, only NdhD1/D2 and
NdhF1 are related to chloroplast ndhD and ndhF genes, respec-
tively. Additionally, the CupA and CupS subunits of the cyano-
bacteral NDH-1S complex have no counterparts in higher plants.
These facts suggest that the structure of chloroplast NDH is
similar to the NDH-1L complex in cyanobacteria (Battchikova
and Aro, 2007; Ogawa and Mi, 2007; Shikanai, 2007b). However,
identification of several novel subunits specific to higher plants
and biochemical characterization of chloroplast NDH imply that
chloroplast NDH is equipped with additional devices compared
with cyanobacterial NDH-1L (Majeran et al., 2008; Peng et al.,
2008; Sirpio
¨et al., 2009a, 2009b; Suorsa et al., 2009; Takabayashi
et al., 2009). In particular, a 1000-kD bundle sheath cell-specific
NDH complex associated with more than 15 proteins was
suggested in maize (Zea mays), and the authors speculate that
this novel complex possibly functions in inorganic carbon con-
centration in addition to PSI cyclic electron transport (Majeran
et al., 2008). However, subunits included in the cyanobacterial
NDH-1S complex were not discovered in maize.
The x-ray crystal structure of plant PSI reveals that it is
composed of a reaction center (RC; PsaA to L and PsaN to P)
and light-harvesting complex I (LHCI), which is composed of four
different LHC proteins (Lhca1 to 4) (Amunts et al., 2007). The four
LHC proteins assemble into two dimers and attach to the PsaF
side of the PSI RC (reviewed in Melkozernov et al., 2006; Nelson
and Yocum, 2006). Two additional, minor LHCI-like proteins
(Lhca5 and Lhca6), with a high degree of similarity to Lhca1 to 4,
were identified in the Arabidopsis genome (Jansson, 1999).
Recently, Lhca5 was shown to be associated with PSI only in
substoichiometric amounts (Ganeteg et al., 2004). Chemical
cross-linking studies revealed that Lhca5 interacts with LHCI in
the Lhca2/Lhca3 site (Lucinski et al., 2006). The Arabidopsis
Lhca6 gene was originally classified as an Lhca2 gene because
of their high similarity (Zhang et al., 1994). The scarce information
on this protein still cannot resolve the question raised by Jansson
(1999): “is the unusual Lhca2 gene (Lhca6) a nonexpressed
pseudogene, or does it have a specific function”? The low
expression level and different expression pattern under different
conditions of Lhca6 compared with Lhca1-4suggest that Lhca6
would have a distinct function from the major Lhca (Lhca1-4)
(Klimmek et al., 2006).
Recently, it was found that thylakoid protein PGR5-Like
1 (PGRL1) was involved in PGR5-dependent PSI cyclic electron
transport (DalCorso et al., 2008). PGRL1 and PGR5 interact
physically, and this complex further associates with PSI, prob-
ably facilitating the operation of PSI cyclic electron transport
(DalCorso et al., 2008). In our previous study, we discovered a
novel NDH-PSI supercomplex with a molecular mass of >1000
kD in Arabidopsis and its putative subsupercomplex with a
slightly lower molecular mass in mutants lacking NdhL or NdhM
(Peng et al., 2008). Here, we give evidence that Lhca5 and Lhca6
are required for the formation of this full-size NDH-PSI super-
complex. Furthermore, we found that the interaction of NDH and
PSI favors the in vivo function of NDH.
RESULTS
Mass Analysis of the NDH-PSI Supercomplex
The NDH-PSI supercomplex was detected by blue native (BN)-
PAGE as a high molecular weight green band, band I, but it was
shifted to the smaller molecular weight position of band II in the
NdhL-defective ndhl/chlororespiratory reduction 23 (crr23) mu-
tant of Arabidopsis (Peng et al., 2008; Shimizu et al., 2008). To
investigate the components of the NDH-PSI supercomplex and
its putative subsupercomplex, we excised two bands containing
them from BN-PAGE (bands I and II). We digested the proteins in
the gel with trypsin and analyzed the extracted peptides by linear
ion-trap triple quadrupole (LTQ)-Orbitrap mass analysis, which
provides high mass accuracy, high resolution, and high sensi-
tivity. Hundreds of proteins were identified in the NDH-PSI
supercomplex corresponding to band I (Table 1; see Supple-
mental Data Set 1 online). Consistent with our previous conclu-
sion that band I contains the PSI-NDH supercomplex (Peng et al.,
2008), the identified proteins were classified mainly into three
groups (Table 1). The first group contains almost all of the PSI
subunits, including Lhca1-4 and two minor LHCI-like proteins
Lhca5 and Lhca6. NDH subunits conserved in the cyanobacterial
complex were classified into the second group; these included
all the NDH subunits, NdhA-O, except NdhG. In addition, our
mass analysis revealed some candidates for novel NDH sub-
units, including PPL2, NDF1 (NDH48), NDF2 (NDH45), NDF4, and
NDF6, which can be classified as genuine NDH subunits (Ishihara
et al., 2007; Ishikawa et al., 2008; Sirpio
¨et al., 2009a; Takabayashi
et al., 2009). Several proteins identified in maize and Arabidopsis
NDH complexes by proteomics analysis (Majeran et al., 2008;
Sirpio
¨et al., 2009a) were detected and classified into the third
group. CYP20-2 (Tlp20), which was recently shown to be an
auxiliary subunit of the NDH complex (Sirpio
¨et al., 2009b), is a
peptidyl-prolyl cis/trans isomerase of 20 kD present in the
thylakoid lumen (Edvardsson et al., 2003). FKBP16-2, two
PsbQ family proteins (PsbQ-F1 and PsbQ-F2), and transmem-
brane protein NDH18 were also detected (Table 1). FKBP16-2
and NDH18 were shown to be NDH subunits in this study (see
3624 The Plant Cell

later). Besides the PSI and NDH subunits, our mass analysis also
found many proteins that were not known to be related to NDH
(see Supplemental Data Set 1 online), such as the photosystem II
(PSII) and ATPase subunits and other thylakoid membrane
proteins, a result that is inconsistent with our previous immuno-
blot studies (Peng et al., 2008). The samples were inevitably
contaminated by other protein complexes owing to the limited
resolution of BN-PAGE, and the LTQ-Orbitrap mass analysis was
sensitive enough to detect them. It is also possible that some
other previously unknown NDH subunits are also included in this
group.
All the PSI subunits detected in band I were identified in the
putative subsupercomplex corresponding to band II in ndhl
(Table 1), which is consistent with our previous report (Peng
et al., 2008). Interestingly, the membrane-embedded NDH sub-
units NdhA-F were detected in band II, but the hydrophilic
subunits, NdhH-O, were absent. NdhG may be also included in
band II, since no signals of this subunit were found in either band,
Table 1. Summary of the PSI and NDH Subunits Identified from LTQ-Orbitrap Mass Analysis of the NDH-PSI Supercomplex (Band I) and
Subsupercomplex (Band II)
Band I Band II
AGI Code Protein Name Morwse Score Protein Match Coverage (%) Morwse Score Protein Match Coverage (%)
PSI ATCG00350 PsaA 588 45 26.5 474 45 24.0
Complex ATCG00340 PsaB 1179 63 34.2 975 53 27.0
ATCG01060 PsaC 69 6 29.6 51 5 29.6
AT4G02770 PsaD-1 896 45 73.6 769 44 70.7
AT1G03130 PsaD-2 848 43 73.5 706 44 70.6
AT4G28750 PsaE-1 526 26 84.6 425 23 69.2
AT2G20260 PsaE-2 449 24 84.8 407 17 60.0
AT1G31330 PsaF 558 36 46.2 524 36 53.4
AT1G55670 PsaG 212 9 21.9 248 11 21.9
AT3G16140 PsaH-1 299 13 60.0 193 14 60.0
AT1G52230 PsaH-2 378 15 60.0 165 14 60.0
AT1G30380 PsaK 72 10 33.8 86 9 19.2
AT4G12800 PsaL 543 18 27.4 278 12 27.4
AT1G08380 PsaO 131 4 21.4 53 3 21.4
AT3G54890 Lhca1 305 16 24.1 336 19 24.1
AT3G61470 Lhca2 412 18 45.9 297 13 36.2
AT1G61520 Lhca3 736 24 27.5 667 25 27.5
AT3G47470 Lhca4 981 34 68.1 666 28 68.9
AT1G45474 Lhca5 720 32 42.6 421 22 36.3
AT1G19150 Lhca6 499 25 44.1 301 17 35.9
NDH ATCG01100 NdhA 716 22 35.0 431 14 23.1
Complex ATCG00890 NdhB 111 8 12.1 120 8 9.8
ATCG00440 NdhC 93 2 20.8 97 3 20.8
ATCG01050 NdhD 410 14 15.4 259 12 14.2
ATCG01070 NdhE 99 5 8.9 71 4 8.9
ATCG01010 NdhF 1137 43 32.6 966 31 29.9
ATCG01110 NdhH 2884 114 81.9 0 0 0.0
ATCG01090 NdhI 897 35 61.6 0 0 0.0
ATCG00420 NdhJ 603 33 48.7 0 0 0.0
ATCG00430 NdhK 556 30 56.4 0 0 0.0
AT1G70760 NdhL 113 14 33.0 0 0 0.0
AT4G37925 NdhM 780 22 49.3 0 0 0.0
AT5G58260 NdhN 786 29 49.3 0 0 0.0
AT1G74880 NdhO 237 12 43.0 0 0 0.0
NDH AT1G15980 NDF1 (NDH48) 1293 62 52.1 1228 63 53.6
Candidates AT1G64770 NDF2 (NDH45) 1374 57 61.5 1116 57 62.9
AT3G16250 NDF4 34 2 9.8 34 3 9.8
AT1G18730 NDF6 360 16 42.9 418 18 42.9
AT2G39470 PPL2 1071 51 60.1 838 49 63.9
AT1G14150 PsbQ-F1 891 34 56.8 627 36 56.8
AT3G01440 PsbQ-F2 813 27 41.4 627 28 41.4
AT5G13120 CYP20-2 769 33 51.0 671 29 51.0
AT4G39710 FKBP16-2 360 15 20.7 313 15 20.7
AT5G43750 NDH18 252 9 24.5 254 9 24.5
The complete list of proteins identified in bands I and II can be found in Supplemental Data Set 1 online. AGI, Arabidopsis Genome Initiative.
Function of the NDH-PSI Supercomplex 3625

possibly for technical reasons (Table 1). These facts imply that
NdhL is essential for stabilizing the hydrophilic subunits in
chloroplast NDH, and the membrane-embedded NDH subunits
stably accumulate even in the absence of the subcomplex
consisting of hydrophilic subunits and NdhL in chloroplasts.
NDH Subunits Encoded by At5g43750 and At4g39710
Our proteome analysis of the NDH-PSI supercomplex detected
many proteins with unknown function, among which NDH sub-
units may be included. Recent bioinformatics studies identified
several NDH subunits (Ishihara et al., 2007; Ishikawa et al., 2008;
Takabayashi et al., 2009) in a strategy based on the phenomenon
that the expression of genes encoding NDH subunits is coregu-
lated. To select candidates for novel subunits in our analysis, we
used a bioinformatics strategy to search for the genes coex-
pressed with nucleus-encoded NDH subunit genes in the
ATTED-II coexpression database (http://www.atted.bio.titech.
ac.jp/). Besides the genes already known to encode NDH sub-
units, two genes, At4g39710 and At5g43750, were coexpressed
with NDH subunit genes with high rvalues (see Supplemental
Table 1 online).
At5g43750 encodes a 212–amino acid protein with a 48–
amino acid N-terminal plastid-targeting peptide (predicted by
ChloroP; http://www.cbs.dtu.dk/services/ChloroP/) and a trans-
membrane domain (predicted by TMHMM; http://www.cbs.dtu.
dk/services/TMHMM-2.0/) (see Supplemental Figure 1 online).
We designated the protein encoded by this gene as NDH18
according to its apparent molecular mass. NDH18 is conserved
among higher plants (see Supplemental Figure 1 online), but no
homologs were found in cyanobacteria or Chlamydomonas
reinhardtii.
At4g39710 was designated as FKBP16-2 and is localized to
the lumen side of the thylakoid (He et al., 2004). FKBP16-2 shows
significant sequence similarity to FKBP13 (At5g45680), which
interacts with the Rieske Fe-S subunit of the cytochrome b
6
f
complex (Gupta et al., 2002). Gm FKBP16-2, Os FKBP16-2, and Z m
FKBP16-2 are classified into the same clade with At FKBP16-2
(66 to 73% identity) (Figure 1A), but At FKBP16-2 exhibits only
40 to 50% similarity to FKBP13s, and no proteins closely related
to FKBR16-2 were found in Chlamydomonas or cyanobacteria. It
is possible that, like NDH18, FKBP16-2 is an NDH subunit
specific to chloroplasts.
To further characterize the function of NDH18 and FKBP16-2,
we used RNA interference (RNAi) to decrease mRNA levels of the
NDH18 and FKBP16-2 genes. Three independent RNAi lines
were characterized in detail for each gene. NDH18 mRNA was
undetectable in the three ndh18 lines (L1 to L3) after 30 cycles of
RT-PCR (Figure 1B). In two fkbp16-2 RNAi lines (L2 and L3),
transcription was also below the detection limit by the same RT-
PCR, but the gene expression was mildly suppressed in L1. RT-
PCR analysis also showed that the expression of FKBP13 was
not affected in fkbp16-2 lines (Figure 1B), indicating that
FKBP16-2 was specifically knocked down.
A postillumination rise in chlorophyll fluorescence, which is
due to the NDH-dependent reduction of PQ by the stromal
electron pool in darkness, is widely used to monitor NDH activity
(Burrows et al., 1998; Shikanai et al., 1998). Although this method
does not analyze the rate of PSI cyclic electron transport in the
light, it reflects NDH activity in vivo and has been used to isolate
many mutants specifically defective in NDH activity (Hashimoto
et al., 2003). In ndh18 (L1 to L3) and fkbp16-2 (L2 and L3), the
transient increase in fluorescence was not detected, indicating
that NDH activity was impaired (Figure 1C). Consistent with the
result of RT-PCR (Figure 1B), NDH activity was detected in
fkbp16-2 L1 but was lower than that in the wild type (Figure 1C).
We conclude that NDH18 and FKBP16-2 are essential for NDH
activity.
The NDH-PSI supercomplex was analyzed in ndh18 and
fkbp16-2 lines by BN-PAGE (Figure 1D). Bands I and II are
present at the top of the gel in the wild type and ndhl, respec-
tively. However, both bands were missing in the ndh18 and
fkbp16-2 lines, except for the accumulation of band I in fkbp16-2
L1 (Figure 1D). To confirm the absence of the supercomplexes,
we excised the region corresponding to bands I and II from the
BN gels. The immunoblot results showed that PsaA, NdhH, and
NdhL were less than one-eighth of the wild-type levels except for
fkbp16-2 L1 (see Supplemental Figure 2 online). Immunoblot
analysis using antibodies against FKBP16-2 and NDH18 showed
that both proteins were below the detection limit in RNAi lines
except for fkbp16-2 L1 (Figure 1E). Consistent with the mRNA
level and activity (Figures 1B and 1C) and the supercomplex level
(see Supplemental Figure 2 online), fkbp16-2 L1 accumulated
one-quarter of the wild-type level of FKBP16-2 (Figure 1E).
Furthermore, the accumulation of NdhL and NDF2 was also
greatly decreased in the ndh18 and fkbp16-2 lines (Figure 1E).
These results demonstrate that NDH18 and FKBP16-2 are novel
NDH subunits and are essential for complex stability. Since
FKBP16-2 contains the FKBP domain, it may have another role in
protein folding or/and in the complex assembly as suggested by
Majeran et al. (2008).
Subunit Stability under the Different NDH
Mutant Backgrounds
NDH activity was completely lost in ndh18 and fkbp16-2 (Figure
1C, L2 and L3). NdhL accumulation was more severely affected
in fkbp16-2 than in ndh18, while that of NDF2 was more severely
affected in ndh18 than in fkbp16-2 (Figure 1E). It was also
reported that NDF2 accumulates stably in ndhl but unstably in
ndhB-defective crr2-2 (Sirpio
¨et al., 2009a; Takabayashi et al.,
2009). This genotype dependency of subunit stability may reflect
the different localization of each subunit in the supercomplex. By
characterization of subunit stability in different mutant back-
grounds, it may be possible to determine the supercomplex
structure. For this purpose, we performed a matrix analysis of
protein blots using eight genotypes and antibodies against seven
NDH subunits (Figure 2A). NdhM, NdhH, NDF1, and NDF2 are
hydrophilic proteins (Rumeau et al., 2005; Sirpio
¨et al., 2009a;
Takabayashi et al., 2009), but NdhL and NDH18 are predicted to
contain a transmembrane domain (Shimizu et al., 2008; see
Supplemental Figure 1 online). In addition, the translation of the
membrane subunit NdhD was impaired in crr4-3 (Kotera et al.,
2005). PPL2 and FKBP16-2 are localized to the lumen side of the
complex (He et al., 2004; Ishihara et al., 2007). Immunoblot
analysis showed that the stromal fraction does not contain NDH
3626 The Plant Cell

subunits (NdhL, NDF1, NDF2, NDH18, and FKBP16-2) in the wild
type and even in various NDH-defective mutants (see Supple-
mental Figure 3 online), suggesting that the subunits are stable
only when they are associated with thylakoid membranes.
Although the accumulation of NdhH and NdhL was almost
completely impaired in ndhm and that of NdhH was also in ndhl,
the levels of other subunits were more than half of the wild type
(Figure 2A). BN-PAGE followed by second-dimensional (2D)
SDS-PAGE and immunoblot analysis showed that all the stable
subunits were present in band II in ndhl (Figures 2B and 2C),
consistent with the mass analysis (Table 1). The results suggest
that NdhH and NdhL form a subcomplex whose absence does
not greatly affect the stability of other parts of the supercomplex.
On the basis of the analogy with bacterial and mitochondrial
complex I and cyanobacterial NDH-1 (Shikanai, 2007b) as well as
our mass analysis of band II (Table 1), we consider that this
putative subcomplex further includes NdhI-K and NdhM-O. The
hydrophilic part of this subcomplex is likely to be anchored to the
Figure 1. Characterization of the ndh18 and fkbp16-2 Mutants.
(A) Phylogenetic tree of the FKBP13 and FKBP16-2 proteins. Sequences were retrieved from GenBank (http://www.ncbi.nlm.nih.gov/). The sequences
are named for each organism. The corresponding amino acid sequences were aligned with the ClustalW program with default settings, and an unrooted
tree was constructed using TreeView software.
(B) RT-PCR analysis of NDH18 and FKBP16-2 mRNA. After reverse transcription, the cDNA was analyzed by 30 cycles of amplification with specific
primers for NDH18,FKBP16-2, and FKBP13.ACT8 was used as an internal control.
(C) Monitoring of NDH activity by chlorophyll fluorescence. Four-week-old leaves were exposed to AL (50 mmol photons m
2
s
1
) for 5 min. After
illumination, the subsequent transient increase in chlorophyll fluorescence was monitored as an indicator of NDH activity. a.u.; arbitrary units. The
fluorescence levels were standardized by the Fm levels.
(D) Thylakoid protein complexes isolated from the wild type, and RNAi lines (ndh18 and fkbp16-2) were separated by BN-PAGE (top panel) and stained
with Coomassie Brilliant Blue (bottom panel). Band I, NDH-PSI supercomplex detected in the wild type; band II, subsupercomplex detected in ndhl.The
top part of the gel is compressed.
(E) Immunodetection of NDH subunits in the wild type (including indicated serial dilutions) and ndhl,ndh18, and fkbp16-2 mutants. Immunoblotting was
performed with antibodies against NDH18, FKBP16-2, NdhL, and NDF2 proteins. Thylakoid proteins were loaded on an equal chlorophyll basis to SDS-
PAGE. Cytfis a loading control. L1, L2, and L3 represent three independent RNAi lines.
Function of the NDH-PSI Supercomplex 3627

thylakoid membrane via NdhL (subcomplex A in Table 2; see
Discussion for more detail), since NdhL has transmembrane
domains and is specific to cyanobacteria and chloroplasts
(Ogawa, 1992; Shimizu et al., 2008). However, analysis of the
NdhL deletion mutant (M9) in Synechocystis sp PCC 6803
revealed that NdhL is not required for the interaction between
the hydrophilic and membrane subcomplexes but is essential for
NDH activity (Battchikova et al., 2005). These facts imply that the
additional transmembrane domain specifically present in chlo-
roplast NdhL may function in stabilizing the subcomplex A on the
membrane subcomplex (Shimizu et al., 2008).
In ppl2 and fkbp16-2, not only PPL2 and FKBP16-2, but also
NdhH and NdhL were absent (Figure 2A), indicating that PPL2
and FKBP16-2 are essential for stabilizing the subcomplex A.
Levels of NDH18, NDF1, and NDF2 in ppl2 and fkbp16-2 were
one-quarter to one-half of the wild-type levels. In the absence of
FKBP16-2, PPL2 is missing and vice versa (Figure 2A), implying
that PPL2 and FKBP16-2 form a subcomplex on the lumen side.
This lumen subcomplex may include lumen proteins CYP20-2,
PsbQ-F1, and PsbQ-F2, which were detected in our mass
analysis (Table 1) and also in previous proteomic analyses
(Majeran et al., 2008; Sirpio
¨et al., 2009a). The evidence of
CYP20-2 as an auxiliary NDH subunit in Arabidopsis also sup-
ports the existence of the lumen subcomplex (Sirpio
¨et al.,
2009b), although there is currently no direct biochemical data to
conclude that the lumen subcomplex is present. BN-PAGE
analysis detected two kinds of subsupercomplexes in fkbp16-2
(Figure 2D). The larger one includes PSI complex and may be the
partially stable subsupercomplex lacking both the subcomplex A
and lumen subcomplex. NDH subunits, NDF1, NDF2, and
NDH18, were also detected in the smaller subsupercomplex
(Figure 2D). The close migration of this subsupercomplex with
the main PSI complex makes it difficult to conclude that the
partially stable smaller subsupercomplex still includes PSI (Fig-
ure 2D). However, we showed that a partially stable subsuper-
complex with the similar molecular weight was detected in crr2-2
defective in the accumulation of membrane subunit NdhB, and
this subsupercomplex contains Lhca6, which is essential for the
formation of the NDH-PSI supercomplex, as well as NDF1 and
NDF2 (see below, Figure 6B, and Discussion). These results
imply that the smaller partially stable subsupercomplex detected
in fkbp16-2 may correspond to the subcomplex B associated
with PSI (Table 2; see Discussion for more detail).
NDH18, NDF1, and NDF2 were absent in crr4-3,ndh18, and
ndf2, yet approximately one-tenth of wild-type levels of NdhH,
NdhL, FKBP16-2, and PPL2 accumulated in these mutants
(Figure 2A). 2D/SDS-PAGE immunoblot also confirmed the
high molecular weight complexes containing PSI corresponding
to the supercomplexes detected in wild-type, ndhl, and fkbp16-2
plants were missing in the ndh18 mutant (Figure 2E). The
remaining NDH subunits in ndh18 also form a subcomplex with
a molecular mass of ;500 kD, which is smaller than that of the
PSI monomer (Figure 2E).
Stoichiometry of the NDH and PSI Subunits in
the Supercomplex
To study the structure of the NDH-PSI supercomplex further, we
determined the stoichiometry of the NDH and PSI subunits in the
supercomplex through the use of antibodies (Figure 3). We
purified the His-tagged recombinant PsaA, NDH18, and
FKBP16-2 proteins from E. coli and used them as quantitative
Figure 2. Stability of Each NDH Subunit in Different NDH Mutant Backgrounds.
(A) Immunoblot of thylakoid membrane proteins, indicated at right, isolated from the wild type (including indicated serial dilutions) and from indicated
mutants. Gels were loaded on an equal chlorophyll basis.
(B) to (E) Analysis of thylakoid protein complexes isolated from the wild type (B) and mutants defective in NDH subunits ([C],ndhl;[D],fkbp16-2;[E],
ndh18). Complexes were separated by BN-PAGE and further subjected to 2D SDS-PAGE. The proteins were immunodetected with specific antibodies.
Positions of band I (B), band II (C), and partially stable subsupercomplexes ([D] and [E]) are indicated by arrows.
3628 The Plant Cell

standards. The NDH-PSI supercomplex was separated from
thylakoid membranes by BN-PAGE, denatured in the gel, and
then directly used for SDS-PAGE. The levels of PsaA, NDH18,
and FKBP16-2 were approximately estimated by the quantitative
immunoblotting to be 0.21, 0.20, and 0.19 pmol in the NDH-PSI
supercomplex corresponding to thylakoid membranes con-
taining 10 mg chlorophyll, respectively (Figure 3). Since the
recovery of proteins from BN gel may depend on the nature of
each protein, we cannot conclude the stoichiometry exactly.
But the levels of PsaA, NDH18, and FKBP16-2 are estimated to
be roughly equimolar (Figure 3), consistent with the idea that
NDH18 and FKBP16-2 are subunits of the NDH-PSI super-
complex.
Lhca6 and Lhca5 Are Required for the Full-Size NDH-PSI
Supercomplex Formation
Our discovery raised a question of how NDH interacts with PSI.
Besides NDH18 and FKBP16-2,Lhca6 was also found to be
coexpressed with NDH subunit genes (see Supplemental Table 1
online). It is possible that Lhca6 is specifically required for
stabilizing the NDH-PSI supercomplex. To study this possibility,
we generated Lhca6 RNAi lines of Arabidopsis. For vector
construction, we chose the region encoding the 59-untranslated
region and the plastid-targeting signal of Lhca6, which shows
low similarity to Lhca2. Several independent lines (lhca6) showed
no visibly different phenotype (see Supplemental Figure 4 online).
Lhca6 expression was specifically knocked down, but the Lhca2
mRNA level was not affected (Figure 4A).
To check whether NDH activity was affected in lhca6,we
measured the transient increases in chlorophyll fluorescence
after actinic light (AL) was turned off. NDH activity was detected
in immature leaves of lhca6 (leaf age is shown in Supplemental
Figure 4 online), although it seems slightly lower than that in the
wild type (Figure 4B). By contrast, mature leaves of lhca6 showed
no chlorophyll fluorescence increase (Figure 4B). To confirm that
this phenotype is due to the defective expression of Lhca6,we
introduced the genomic Lhca6 sequence of rice (Oryza sativa;Os
Lhca6) into Arabidopsis lhca6 lines. Owing to its sequence
differences in the region encoding the 59-untranslated region
and plastid-targeting signal, the rice gene escaped RNAi and
fully restored the transient increase in chlorophyll fluorescence in
mature leaves (Figures 4A and 4B, lhca6c).
Immunoblotting analysis provided direct evidence that NDH
content was reduced slightly in immature leaves of lhca6 (;80%)
compared with the wild type and more so (;60%) in mature
leaves (Figures 4C and 4D). However, subunits of the other
protein complexes, D1, PsaA, and Cytf, were not affected
(Figures 4C and 4D), indicating that accumulation of the NDH
complex was specifically affected in the absence of Lhca6,
especially in mature leaves. Surprisingly, BN-PAGE showed that
the NDH-PSI supercomplex corresponding to band I was com-
pletely absent in both immature and mature leaves of lhca6
(Figure 5A; see Supplemental Figure 5 online), and NDH subunits
were present in a complex of ;1000 kD (Figure 5B). We
conclude that Lhca6 is essential for the full-size NDH-PSI super-
complex formation but not for the activity detected in the
chlorophyll fluorescence analysis in immature leaves (Figure 4B).
Table 2. Summary of the Chloroplast NDH Subunits
Subcomplex Name Annotation Molecular Mass (kD) Reference
Membrane NdhA ATCG01100 40.0
subcomplex NdhB ATCG00890 42.0
NdhC ATCG00440 14.0
NdhD ATCG01050 57.0
NdhE ATCG01070 11.3
NdhF ATCG01010 85.0
NdhG ATCG01080 19.0
Subcomplex A NdhH ATCG01110 45.5
NdhI ATCG01090 20.0
NdhJ ATCG00420 20.0
NdhK ATCG00430 27.0
NdhL AT1G70760 17.0 Shimizu et al. (2008)
NdhM AT4G37925 22.0 Rumeau et al. (2005)
NdhN AT5G58260 18.0 Rumeau et al. (2005)
NdhO AT1G74880 13.0 Rumeau et al. (2005)
Subcomplex B NDF1 (NDH48) AT1G15980 48.0 Takabayashi et al. (2009); Sirpio
¨et al. (2009a)
NDF2 (NDH45) AT1G64770 45.0 Takabayashi et al. (2009); Sirpio
¨et al. (2009a)
NDF4 AT3G16250 22.0 Takabayashi et al. (2009)
NDF6 AT1G18730 18.0 Ishikawa et al. (2008)
NDH18 AT5G43750 18.0 This work
Lumen PPL2 AT2G39470 17.0 Ishihara et al. (2007)
subcomplex FKBP16-2 AT4G39710 16.0 This work
PsbQ-F1 AT1G14150 17.0 Majeran et al. (2008)
PsbQ-F2 AT3G01440 17.0 Majeran et al. (2008)
CYP20-2 AT5G13120 20.0 Majeran et al. (2008); Sirpio
¨et al. (2009b)
The chloroplast NDH complex was divided into four subcomplexes. The localization of the subcomplexes is presented in Figure 9.
Function of the NDH-PSI Supercomplex 3629

Since Lhca5 was also detected in the NDH-PSI supercomplex
by mass analysis (Table 1), we analyzed Lhca5 knockout lines of
Arabidopsis and found that the NDH-PSI supercomplex corre-
sponding to band I was absent and the levels of NDH subunits
were slightly decreased (Figures 5C and 5E). 2D BN/SDS-PAGE
and immunoblotting studies showed that the NDH subunits were
present mainly in a complex of ;1000 kD and in trace amounts in
the NDH-PSI supercomplex corresponding to band I in lhca5
(Figure 5D). To estimate the levels of residual supercomplex in
lhca5 and lhca6 further, we excised the region corresponding to
the NDH-PSI supercomplex from BN gels. Immunoblot results
showed that signals of PsaA, NdhH, and NdhL were below the
detection limit in lhca6 RNAi lines (at least less than one-sixteenth
of the wild-type levels), and approximately one-sixteenth of the
supercomplex was still present in the lhca5 mutant compared
with the wild type (see Supplemental Figure 5 online).
In BN gel, NDH subunits were detected in the 1000-kD
complex exclusively in lhca6 and mainly in lhca5 (Figures 5B
and 5D). Unexpectedly, this 1000-kD complex still includes
PsaA, although the PsaA level is reduced compared with that
in the NDH-PSI supercomplex (Figures 5B and 5D). These results
can be explained by an idea that the NDH complex interacts with
multiple copies of the PSI complex (see Discussion). Further
analyses confirmed that this 1000-kD complex also contains
Lhca3 and NDH subunits, NDH18 and NDF1 (see Supplemental
Figure 6 online), suggesting that the complex observed in lhca5
and lhca6 is a smaller version of the NDH-PSI supercomplex,
which contains entire subunits of NDH and at least a single copy
of PSI.
Specific Function of Lhca6 in the NDH-PSI Supercomplex
EST databases included 22 and 18 putative orthologs of Arabi-
dopsis Lhca6 and Lhca5, respectively, in flowering plants, but no
homologs were found in Chlamydomonas, indicating that Lhca5
and Lhca6 are conserved among flowering plants. Although
Lhca6 is homologous to Lhca2 (Jansson, 1999), the Lhca6
orthologs were not clustered with At Lhca2 (Figure 6A). Further-
more, as indicated by the phenotype of lhca6 (Figures 4 and 5),
Lhca2 clearly could not complement the function of Lhca6, even
though the level of Lhca2 mRNA was much higher than that of
Lhca6 (Figure 4A). This fact supports the idea that Lhca6 is not an
isoform of Lhca2 and has a different physiological function.
Protein alignment revealed that mature Lhca6 has an N-terminal
extension (see Supplemental Figure 7 online), implying a specific
function of Lhca6 in this region.
Lhca5 is associated with PSI monomer in substoichiometric
amounts (Ganeteg et al., 2004). To study whether Lhca6 also
binds to PSI monomer, we excised the band corresponding to
PSI monomer from the BN gel and used it in LTQ-Orbitrap mass
analysis. Although Lhca1-5 were detected in PSI monomer
isolated from the wild type, crr2-2, and lhca6, no Lhca6 signal
was found in our mass analysis (see Supplemental Data Set 2
online). We also constructed a chimeric gene in which the
C-terminal end of Os Lhca6 was fused with HA (influenza
hemagglutinin protein epitope) tag (Os Lhca6-HA) under the
control of the Os Lhca6 promoter. The construct was then
introduced into Arabidopsis lhca6 lines and complemented NDH
activity. BN-PAGE and immunoblot studies showed that Lhca6
was associated mainly with the NDH-PSI supercomplex (Figure
6B), indicating that Lhca6 is predominantly present in the NDH-
PSI supercomplex. We also transformed wild-type Arabidopsis
and crr2-2 with Lhca6-HA tag under the control of the cauliflower
mosaic virus (CaMV) 35S promoter. BN-PAGE and immunoblot
studies showed that Lhca6 was associated with both the NDH-
PSI supercomplex and PSI monomer in the wild type probably
due to the result of overaccumulation but exclusively with PSI
monomer in crr2-2 (Figure 6B). These results indicate that Lhca6
is stable in the absence of NDH and can interact with PSI
monomer in the same way as Lhca5 (Ganeteg et al., 2004).
However, PSI monomer is ;100 times as common as NDH-PSI
supercomplex (Peng et al., 2008), indicating that Lhca6 has a
much higher affinity for NDH than for PSI monomer. We conclude
that Lhca6 is an LHCI specifically required for NDH-PSI super-
complex formation, consistent with the fact that the gene is not
conserved in Chlamydomonas, which does not contain chloro-
plast NDH.
The trace level of Lhca6-HA was detected in the NDH-PSI
supercomplex position in crr2-2 probably because of the leaky
accumulation of the NDH complex in this mutant (Figure 6B;
Figure 3. Stoichiometric Analysis of NDH and PSI in the Supercomplex.
The NDH-PSI supercomplex corresponding to band I was excised from
the BN-PAGE and denatured in gel. Dilution series of the purified His-
tagged recombinant PsaA (A), NDH18 (B), and FKBP16-2 (C) proteins
were used to estimate the amount of NDH subunits and PsaA in the
supercomplex. The signal was visualized by an LAS3000 chemilumines-
cence analyzer (Fuji Film) and analyzed by Imagemaster software
(Amersham Pharmacia Biotech). The result is representative of three
experiments using independently isolated thylakoid membranes. Control
immunoblots confirmed that the antibodies did not recognize the His-tag
or Nus-tag. I, NDH-PSI supercomplex isolated from thylakoids contain-
ing 10 mg chlorophyll.
3630 The Plant Cell

Hashimoto et al., 2003; Peng et al., 2008). Besides PSI monomer
and the NDH-PSI supercomplex, Lhca6 was also associated
with a putative subsupercomplex that contains NDF1, NDF2, and
NDH18 but does not contain NdhL, NdhH, or FKBP16-2 (Figure
6B). These results suggest that the subcomplex B, including
NDF1, NDF2, and NDH18, may be still associated with PSI via
Lhca6 even in the absence of the membrane subcomplex
including NdhB.
Lhca6 Is Required for Efficient Operation of NDH
Although lhca6 lines contained 60 to 80% of the wild-type levels
of NDH subunits, NDH activity was undetectable in mature
leaves (Figure 4B). To study the link between the supercomplex
formation and activity, we further analyzed the photosynthetic
electron transport in mature leaves of lhca6. We determined two
chlorophyll fluorescence parameters, electron transport rate
(ETR) and nonphotochemical quenching (NPQ), since they reflect
even subtle defects in photosynthetic apparatus (Shikanai et al.,
1999). ETR was slightly reduced and NPQ was not affected in
mature leaves of lhca6 (Figures 7A and 7B). The P700
+
(oxidized
RC chlorophyll of PSI) level, which is a sensitive indicator of the
capacity for electron acceptance from PSI, in lhca6 mature
leaves was similar to that in ndhl, but slightly lower than that in the
wild type (Figure 7C). These phenotypes are consistent with
those of other Arabidopsis mutants defective in chloroplast NDH
(Munekage et al., 2004). We also assayed Fd-dependent PQ
reduction activity, which is involved in PSI cyclic electron trans-
port in vivo, using ruptured chloroplasts isolated from mature
leaves of lhca6 (Figure 7D). As in crr2-2, PQ reduction activity
was slightly lower in lhca6 than in the wild type. Antimycin A
inhibits the PGR5-dependent pathway of PQ reduction by Fd
(Munekage et al., 2004) and can be used to discriminate NDH-
dependent activity from PGR5-dependent activity. Antimycin A
decreased the PQ reduction activity slightly less in lhca6 than in
crr2-2. The remaining Antimycin A–resistant PQ reduction activ-
ity may depend on the smaller version of NDH-PSI supercom-
plex, which is still present in lhca6 mature leaves (Figures 4 and
Figure 4. Characterization of the lhca6 RNAi Lines.
(A) RT-PCR analysis of Lhca6 and Lhca2 mRNA. Total RNA (5 mg) was reverse transcribed, and the resulting cDNA was used in 30 cycles of PCR with
specific primers for Lhca6,OsLhca6,andLhca2.ACT8 was used as an internal control. “lhca6c” indicates lhca6 RNAi mutant complemented by the
introduction of Os Lhca6 gene.
(B) Monitoring of NDH activity by chlorophyll fluorescence as in Figure 1C.
(C) Immunodetection of chloroplast proteins from immature and mature leaves of the wild type and lhca6. The thylakoid membrane proteins were
separated by SDS-PAGE and immunodetected with specified antibodies. Thylakoid proteins were loaded on an equal chlorophyll basis. These
experiments were repeated three times independently, and similar results were obtained. Results from a representative experiment are shown.
(D) Analysis of thylakoid proteins. Immunoblot results of three independent isolations of thylakoid membranes were analyzed with Imagemaster
software (Amersham Pharmacia Biotech). The protein levels in the wild-type mature leaves and lhca6 immature and mature leaves are shown relative to
those in the wild-type immature leaves (100%). Means 6SD (n=3).
Function of the NDH-PSI Supercomplex 3631

5). We also compared the oxidation kinetics of P700 by far-red
light (FR) after AL illumination between wild-type, ndh18, and
lhca6 mature leaves (Figure 7E). After 2-min illumination of AL
(900 mmol photons m
22
s
21
) supplemented with FR, AL was
turned off and P700
+
was transiently reduced by electrons from
the PQ pool, and subsequently P700 was reoxidized by back-
ground FR. The operation of NDH retards the reoxidation of P700
by transferring electrons from the reduced stromal pool to PQ
(Shikanai et al., 1998). The reoxidation of P700 was slower in
wild-type leaves than ndh18 and lhca6 mature leaves (Figure 7E),
which is consistent with the phenotypes observed in the DndhB
tobacco mutant (Shikanai et al., 1998). From these results, we
conclude that NDH activity was affected in the mature leaves of
lhca6 mutant.
To further investigate the physiological significance of the
NDH-PSI supercomplex formation, we constructed the lhca6
pgr5 double mutant. At a light intensity of 50 mmol photons m
22
s
21
,lhca6 pgr5 plant growth was impaired, as in crr4-2 pgr5 and
crr3 pgr5, but slightly better than in crr2-2 pgr5 (Figure 8A).
Among the three crr mutants, the NDH level was most severely
affected in crr2-2, resulting in a smaller plant (Figures 8A and 8D;
Munekage et al., 2004). Steady state chlorophyll fluorescence
captured by a CCD camera was compared among mutants at a
light intensity of 100 mmol photons m
22
s
21
(Figure 8B). The
chlorophyll fluorescence level was similar between the wild type
and lhca6, suggesting that the NDH-PSI supercomplex forma-
tion is not essential for photosynthesis (Munekage et al., 2004).
However, pgr5 displayed slightly higher chlorophyll fluorescence
Figure 5. Analysis of Thylakoid Protein Complexes from Wild-type, lhca6, and lhca5 Plants.
(A) Thylakoid protein complexes isolated from immature and mature leaves of wild-type and lhca6 plants were separated by BN-PAGE (left) and stained
with Coomassie Brilliant Blue (CBB) (right). Band I position is indicated.
(B) Thylakoid membrane complexes separated by BN-PAGE in (A) were further subjected to 12.5% 2D SDS-PAGE, and the proteins were
immunodetected with specific antibodies against PsaA, NdhH, and FKBP16-2. Positions of NDH-PSI supercomplex and the smaller NDH-PSI
supercomplex are indicated by red and black arrows, respectively.
(C) Thylakoid protein complexes isolated from wild-type and lhca5 plants were separated by BN-PAGE (left) and stained with Coomassie blue (right).
Band I position is indicated.
(D) Thylakoid protein complexes isolated from the lhca5 plants were separated by BN-PAGE and further subjected to 2D SDS-PAGE. The proteins were
immunodetected with specific protein antibodies against PsaA, NdhH, and FKBP16-2. Positions of the NDH-PSI supercomplex and the smaller NDH-
PSI supercomplex are indicated by red and black arrows, respectively.
(E) Immunodetection of chloroplast proteins from the wild type and lhca5. The thylakoid membrane proteins were separated by SDS-PAGE and
immunodetected with antibodies against the indicated proteins. Thylakoid proteins were loaded on an equal chlorophyll basis.
3632 The Plant Cell

than the wild type, indicating a mild impairment of photosynthetic
electron transport at this light intensity. Although NDH is dis-
pensable for photosynthesis under these conditions, it is essen-
tial for efficient photosynthesis in the pgr5 mutant background,
thus resulting in the drastic phenotype in the double mutants
(Figures 8A and 8B; Munekage et al., 2004). The chlorophyll
fluorescence level was high in lhca6 pgr5, suggesting a defect in
photosynthesis, as in the other double mutants (Figure 8B).
To further characterize the photosynthetic activity in lhca6
pgr5, we analyzed the light intensity dependence of ETR (Figure
8C). Although the maximum ETR in pgr5 decreased to ;60% of
the wild-type level, it was reduced to 10 to 20% in the crr4-2
pgr5,crr3 pgr5, and crr2-2 pgr5 double mutants. Although ETR
was lower in mature leaves of lhca6 pgr5 than in pgr5, it was
higher than in the other double mutants (Figure 8C). ETR was
significantly higher in immature leaves than in mature leaves, as
reflected in the chlorophyll fluorescence image (Figures 8A and
8B). The NDH subunit levels in lhca6 pgr5 were more than three-
quarters of those in the wild type (Figure 8D), and the levels were
even higher than those in the lhca6 single mutant (Figure 4D). As
expected, the other double mutants accumulated less than half
of the wild-type level of NDH subunits (Figure 8D). Although lhca6
pgr5 accumulated >75% of the wild-type level of the NDH
complex, plant photoautotrophic growth and photosynthesis
were severely impaired in the pgr5 mutant background.
DISCUSSION
The discovery of several chloroplast-specific NDH subunits
(Figure 1; Ishihara et al., 2007; Ishikawa et al., 2008; Majeran
et al., 2008; Sirpio
¨et al., 2009a; Takabayashi et al., 2009) makes
it likely that chloroplast NDH has a more complex structure than
cyanobacterial NDH. Taking together all the available informa-
tion, we present a structural model of the NDH-PSI supercom-
plex (Table 2, Figure 9). On the basis of the analogy with E. coli
and cyanobacterial NDH-1 (Herranen et al., 2004; Zhang et al.,
2004, 2005; Shikanai 2007b), it is likely that membrane-spanning
subunits corresponding to chloroplast NdhA-G form the mem-
brane subcomplex, and hydrophilic subunits corresponding to
NdhH-K and NdhM-O form one of the stroma-side subcomplex-
es (Table 2, Figure 9). Although NdhL contains three transmem-
brane domains (Shimizu et al., 2008), our results suggest the
direct interaction among NdhL and hydrophilic subunits NdhH-K
and NdhM-O in chloroplast NDH. We grouped these subunits,
including NdhL, into the subcomplex A to distinguish it from
subcomplex B, which consists of chloroplast-specific subunits
(Table 2, Figure 9).
Our results suggest that lumen proteins PPL2 and FKBP16-2
form a putative subcomplex on the lumen side (Figure 2A).
CYP20-2, PsbQ-F1, and PsbQ-F2 may also be included in this
lumen subcomplex (Table 2, Figure 9). Subunits of subcomplex A
are unstable without subunits of the lumen subcomplex (Figure
2A), suggesting that subcomplex A may interact with the lumen
subcomplex. This idea is consistent with the discovery of a
500-kD complex including the subunits of subcomplex A and
lumen subcomplex in ndh18 (Figure 2E) and with trace levels of a
complex of similar molecular weight in mature leaves of lhca6
Figure 6. Characterization of Lhca6.
(A) Phylogenetic tree of Lhca family proteins. Sequences of Lhca5 and
Lhca6 were retrieved from GenBank (http://www.ncbi.nlm.nih.gov/). The
sequences are named for each organism. The sequences of Lhca1-9 of
Chlamydomonas and Lhca1-4 of Arabidopsis were also included in the
analysis. The corresponding amino acid sequences were aligned with the
ClustalW program with default settings, and an unrooted tree was
constructed using TreeView software.
(B) Immunodetection of Lhca6 using a monoclonal antibody against the
HA tag. The lhca6 plants were transformed with the Os Lhca6 gene fused
to the sequence encoding the HA tag. Wild-type and crr2-2 plants were
transformed with At Lhca6 cDNA fused to the region encoding the HA tag
expressed under the control of the CaMV 35S promoter. Thylakoid
complexes isolated from transformants were separated by BN-PAGE and
further subjected to 2D SDS-PAGE. The proteins were immunodetected
with specific antibodies. The positions of NDH-PSI supercomplex and PSI
monomer are indicated by closed arrows. The position of the putative
subsupercomplex detected in crr2-2 is indicated by an open arrow.
Function of the NDH-PSI Supercomplex 3633

(Figure 5B). NdhL and/or other unidentified transmembrane
subunits may intermediate in the binding of the two subcom-
plexes separated by thylakoid membranes. The subunits of the
lumen subcomplex do not exist in cyanobacteria, implying that
higher plants developed a novel mechanism to stabilize sub-
complex A.
NDF1 (NDH48), NDF2 (NDH45), and NDF4 are attached to the
membrane subcomplex on the stromal side (Sirpio
¨et al., 2009a;
Takabayashi et al., 2009). NDH18, NDF1, and NDF2 were miss-
ing in ndf2 and ndh18 and in NdhD-defective crr4-3 (Figure 2A).
These results suggest that the subcomplex including NDF1,
NDF2, and NDH18 is associated, directly or indirectly, with the
membrane subunit NdhD. This idea is further supported by the
similar NDH subunit accumulation profile in crr4-3,ndh18, and
ndf2: the complete lack of subcomplex B including NDF1, NDF2,
and NDH18, the low-level accumulation of subcomplex A
Figure 7. In Vivo and in Vitro Analysis of Electron Transport Activity.
(A) The ETR is depicted relative to the maximum value of F
PSII
3light intensity in the wild type (100%).
(B) Dependence of NPQ of chlorophyll fluorescence on light intensity.
(C) Light intensity dependence of the P700 oxidation ratio (DA/DA
max
)inndhl,lhca6, and wild-type mature leaves.
(D) Increases in chlorophyll fluorescence by addition of NADPH (0.25 mM) and Fd (5 mM) under weak illumination (1.0 mmol photons m
2
s
1
) were
monitored in osmotically ruptured chloroplasts (20 mg chlorophyll/mL) of wild-type, lhca6,crr2-2,andpgr5 mature leaves. Ruptured chloroplasts were
incubated with 10 mM Antimycin A before measurement. All values are mean 6SD (n=5)in(A) to (C). This is a representative result of three experiments
using thylakoid membranes independently isolated.
(E) Redox kinetics of P700 after termination of AL illumination (900 mmol photons m
2
s
1
for 2 min) under a background of FR. The leaves were
illuminated by AL supplemented with FR to store electrons in the stromal pool. After termination of AL illumination, P700
+
was transiently reduced by
electrons from the PQ pool; thereafter, P700 was reoxidized by background FR. The redox kinetics of P700 was recorded. The P700
+
levels were
standardized by their maximum levels by exposing FR. The results using two independent plants for each genotype are overlapped.
3634 The Plant Cell

including NdhH and NdhL (<10%), and the milder effect on the
lumen subcomplex including PPL2 and FKBP16-2 (<20%) (Fig-
ure 2A). We classified NDF1, NDF2, and NDH18 into the
subcomplex B, in which all subunits are specific to chloroplasts
(Table 2, Figure 9). NDF4 and NDF6 probably belong to this group
(Ishikawa et al., 2008; Takabayashi et al., 2009). In the absence of
the subcomplex A and lumen subcomplex, the subcomplex B
and membrane subcomplex still associate with PSI (larger
subsupercomplex in fkbp16-2 mutant), although this subsuper-
complex is partially unstable (Figures 2D and 9). The subcomplex
B still associates with Lhca6 without the membrane subunit
NdhB (Figures 6B) forming a partially stable subsupercomplex
whose molecular weight is similar to that of the smaller sub-
supercomplex detected in fkbp16-2 (Figure 2D). These results
imply that the NDH complex interacts with PSI probably via the
subcomplex B and also possibly via membrane subunits NdhD
and/or NdhF (Figure 9) based on the analogy with cyanobacterial
NDH-1L whose membrane complex easily dissociates into the
core domain consisting of NdhA-C, E, G, and L and the NdhD/F
subcomplex (Battchikova and Aro, 2007).
Immunoblots showed that Lhca6 was mainly detected in the
NDH-PSI supercomplex (Figure 6B). In agreement with our
results (Table 1; see Supplemental Data Set 2 online), Lhca6
was not detected in the PSI monomer in various plant species by
mass spectrometry analysis (Zolla et al., 2007). These facts
suggest that Lhca6 is mainly localized to the NDH-PSI super-
complex, although we do not eliminate the possibility that an
extremely low level of Lhca6 associates with monomeric PSI
since they can interact in the Lhca6 overexpressors (Figure 6B).
Furthermore, Lhca6 is specifically detected in species containing
chloroplast NDH (Figure 6A), implying that this key gene was
acquired during the evolution of land plants, allowing NDH to
interact with PSI. By contrast, Lhca5 was found in both the NDH-
PSI supercomplex and PSI monomer (Table 1; see Supplemental
Data Set 2 online), which is consistent with previous reports
(Ganeteg et al., 2004; Storf et al., 2004; Zolla et al., 2007). Lhca6
overaccumulating in thylakoid membranes also can bind to PSI
monomer (Figure 6B), suggesting that Lhca6 intermediates di-
rectly in the binding of NDH and PSI or indirectly via other
proteins, such as Lhca5.
Majeran et al. (2008) reported that the subunits of PSI and NDH
complex showed average bundle sheath cell (BSC)/mesophyll
cell (MC) accumulation ratios of 1.6 and 3.0, respectively, in Z.
mays. Given the apparent role of Lhca5 and Lhca6 in the
Figure 8. Characterization of the lhca6 pgr5 Double Mutant.
(A) Visible phenotype of the double mutants. Seedlings were cultured at 50 mmol photons m
2
s
1
for 3 weeks after germination.
(B) High chlorophyll fluorescence phenotype of lhca6 pgr5. Dark-adapted seedlings of the wild type and knock out mutants were illuminated at 100 mmol
photons m
2
s
1
for 1 min and then a chlorophyll fluorescence image was captured by CCD camera. Arrows indicate immature leaves of lhca6 pgr5
emitting less fluorescence than mature leaves.
(C) Light intensity dependence of the relative ETR. ETR is shown relative to the maximum ETR in the wild type (100%); means 6SD (n=5).
(D) Immunodetection of chloroplast proteins. The thylakoid membrane proteins were separated by SDS-PAGE and immunodetected with antibodies
against NdhH, NdhL, and Cytfproteins. Thylakoid proteins were loaded on an equal chlorophyll basis.
Function of the NDH-PSI Supercomplex 3635

formation of the NDH-PSI supercomplex, these two minor LHCI
proteins should accumulate more in the BSC chloroplasts, which
include higher levels of NDH subunits. Consistent with this idea,
Lhca6 and Lhca5 showed BSC/MC ratios of 3.5 and 2.5, re-
spectively (Majeran et al., 2008). From these results, we conclude
that both Lhca5 and Lhca6 are required for the NDH-PSI super-
complex formation.
Normal state transition occurs in the lhca6 mutant (see Sup-
plemental Figure 8 online; Kouril et al., 2005; Jensen et al., 2007).
State transition is also not essential for NDH-PSI supercomplex
formation (Peng et al., 2008), excluding the possibility that NDH
interacts with PSI via LHCII. Since Lhca5 interacts with PSI on the
Lhca2/Lhca3 site (Lucinski et al., 2006), it is likely that NDH
associates with PSI via Lhca5/Lhca6 and the LHCI complex.
BN-PAGE showed that NDH subunits accumulate in smaller
versions of the NDH-PSI supercomplex (Figure 5) in both lhca5
and lhca6, in which the major photosynthetic protein complexes,
including PSI, PSII, and LHCII, were not affected (Figure 5; see
Supplemental Figure 5 online). NDH activity was still detected in
immature leaves of lhca6 (Figure 4B) in which accumulation of the
NDH-PSI supercomplex was completely impaired (Figures 5B
and 5D; see Supplemental Figure 5 online), suggesting that the
smaller NDH-PSI supercomplex detected in lhca5 and lhca6
contains at least a minimum set of NDH subunits. Immunoblot
analysis showed that the smaller NDH-PSI supercomplex con-
tains all the subunits of the PSI complex tested (Figures 5B and
5D; see Supplemental Figure 6 online). The smaller versions of
NDH-PSI supercomplex migrated faster in BN gel than the
subsupercomplex corresponding to band II detected in ndhl
(Figure 5; Peng et al., 2008), suggesting that the size difference
between the smaller NDH-PSI supercomplex and the intact
NDH-PSI supercomplex is larger than 280 kD (based on our
estimation of the molecular mass of the subcomplex A). One
possible explanation for the results is that NDH interacts with two
copies of the PSI complexes possibly via Lhca5 and Lhca6. Our
rough estimation of the NDH/PSI stoichiometry does not exclude
this possibility (Figure 3). However, we do not include this
information in the model of the supercomplex (Figure 9) since
the biochemical information is still lacking especially on the
docking sites. The most straightforward approach is coimmu-
noprecipitation using antibodies against subunits of the super-
complex, but this trial was unsuccessful so far probably because
of a low abundance and fragility of the NDH-PSI.
In the absence of Lhca5 and Lhca6, NDH exists in thylakoids
as a smaller NDH-PSI supercomplex at a level mildly reduced
compared with that in the wild type (Figures 4 and 5). The levels of
this smaller NDH-PSI supercomplex depend on the leaf devel-
opment in lhca6 (Figures 4C and 4D). The pgr5 defect did not
decrease the content of NDH subunits (Figure 8D), implying that
the smaller NDH-PSI supercomplex may not be sensitive to
oxidative stress. The level of protein is determined by the balance
of synthesis and degradation, and active synthesis of NDH
subunits may overcome its instability more efficiently in imma-
ture leaves than in mature leaves.
What is the determinant of the severe phenotype observed in
lhca6 pgr5 (Figures 8A to 8C)? At least NDH activity detected in
the postillumination rise of chlorophyll fluorescence reflects the
level of NDH in immature and mature leaves (Figures 4B to 4D).
The in vitro Fd-dependent PQ reduction assay using ruptured
chloroplasts also suggests that the smaller NDH-PSI super-
complex still retains some activity (Figure 7D). It is possible that
under the certain threshold level the in vivo function of NDH
complex is drastically impaired, leading to the phenotype of
lhca6 pgr5 (Figure 8). Consistent with this idea, the fluorescence
level was lower in immature leaves of lhca6 pgr5 than its mature
leaves (Figures 8A and 8B). In this case, Lhca6 is required for the
supercomplex formation and consequently for full NDH activity in
vivo via its function stabilizing NDH. However, the level of NDH
was only mildly affected in lhca6 pgr5, including at least >75%
levels of NDH subunits (Figure 8D). It is also probable that the
supercomplex formation is required for the efficient operation of
NDH activity.
As a conclusion, we cannot clearly explain the discrepancy
between the partial impairment of NDH activity in lhca6 (Figures
4B, 7D, and 7E) and the drastic phenotype in lhca6 pgr5 (Figure
8). The problem is related to the difficulty in monitoring NDH-
dependent PSI cyclic electron transport in the light, and our
methods rely on the measurement in the dark (Figures 4B and 7D)
Figure 9. A Schematic Model of the NDH-PSI Supercomplex in Chlo-
roplasts.
The chloroplast NDH is divided into four subcomplexes (Table 2). The
membrane subcomplex contains of plastid-encoded subunits NdhA-G.
The subcomplex A consists of plastid-encoded NdhH-K and nuclear-
encoded NdhL-O. The NdhL subunit is a transmembrane protein and is
required for stabilizing this subcomplex. The lumen subcomplex may
include PPL2, PsbQ-F1, PsbQ-F2, AtCYP20-2, and FKBP16-2 and is
essential for stabilizing the subcomplex A, probably via NdhL or other
unidentified membrane proteins. The subcomplex B contains NDF1
(NDH48), NDF2 (NDH45), NDF4 (Sirpio
¨et al., 2009a; Takabayashi et al.,
2009), and two transmembrane proteins, NDF6 and NDH18. The
subcomplex A partially protects NDF1 from protease attack, suggesting
that it also interacts with the subcomplex B (Sirpio
¨et al., 2009a). The
subcomplex B is partially unstable in the absence of the subcomplex A
and lumen subcomplexes. Without the membrane subunit NdhB, the
subcomplex B still interacts with Lhca6. However, the subcomplex B is
totally unstable without the membrane subunit NdhD in the crr4-3
mutant, suggesting that it also interacts with NdhD or/and NdhF. Sub-
units of the subcomplex B and lumen subcomplex are specific to
chloroplast NDH. The two minor LHCI proteins Lhca5 and Lhca6 are
required for the full-size NDH-PSI supercomplex formation. The model
does not include the information of stoichiometry, which is discussed in
the main text in detail.
3636 The Plant Cell

or under low FR light (Figure 7E). Consistent with the mutant
phenotype (Munekage et al., 2004), PGR5-dependent PSI cyclic
electron transport is predominate in thylakoids in the light and
NDH-dependent PSI cyclic electron transport is under the de-
tection limit (Okegawa et al., 2008). Although we cannot evaluate
the activity experimentally, the most straightforward discussion
for the clear mutant phenotype (Figure 8) is the compensatory
contribution of NDH to PSI cyclic electron transport in pgr5. The
supercomplex formation via Lhca6 is required for the process,
but the exact molecular mechanism remains for future analysis.
We propose that the NDH-PSI supercomplex is a minimal
functional unit for the efficient in vivo function of NDH, as evident
in the lhca6 pgr5 phenotype. Our findings relate to a long debate
on the electron donor to NDH, since it is clear now that the
previous biochemical approaches focused on NDH monomer or
NDH subcomplex (Guedeney et al., 1996; Sazanov et al., 1998).
What was a reason why chloroplast NDH acquired the subcom-
plex B and lumen subcomplexes and used Lhca5 and Lhca6 to
form the supercomplex? This process may have facilitated the
novel electron transport in chloroplasts required for stress toler-
ance. It may be necessary to reevaluate the biochemistry of
chloroplast NDH in the supercomplex.
METHODS
Plant Material and Growth Conditions
Arabidopsis thaliana (ecotype Columbia gl1) plants were grown in soil in a
growth chamber (50 mmol photons m
22
s
21
, 16-h photoperiod, 238C) for 3
to 4 weeks. The lhca5 mutant was obtained from the RIKEN Bioresource
Center (http://www.brc.riken.jp/lab/epd/Eng/catalog/seed.shtml). The
ppl2 and ndf2 mutants were kindly provided by Kentaro Ifuku and
Tsuyoshi Endo, respectively, of Kyoto University.
Thylakoid Membrane Preparation, BN-PAGE, and
Immunoblot Analysis
Chloroplasts and thylakoids were isolated as described (Munekage et al.,
2002). BN-PAGE and subsequent 2D/SDS-PAGE immunoblot analysis
was performed as described (Peng et al., 2008). For immunoblot analysis,
thylakoid proteins were loaded on an equal chlorophyll basis. The signals
were detected using an ECL Advance Western Blotting Detection Kit for
NdhH (GE Healthcare) or an ECL Plus Western Blotting Detection Kit for the
others (GE Healthcare) and visualized by an LAS3000 chemiluminescence
analyzer (Fuji Film). Immunoblotswere quantified by Imagemaster software
(Amersham Pharmacia Biotech) in three independent experiments.
Peptide Preparation for Tandem Mass Spectrometry Analysis
Thylakoid membrane complexes isolated from wild-type and ndhl mutant
plants were solubilized and separated by BN-PAGE. Bands I and II
(described in Peng et al., 2008) and the PSI monomer (described in Peng
et al., 2006) were excised from the gel. Peptide preparation and liquid
chromatography–tandem mass spectrometry (LC-MS/MS) analyses
were performed as previously described (Fujiwara et al., 2009). The
excised bands were treated twice with 25 mM ammonium bicarbonate in
30% (v/v) acetonitrile for 10 min and 100% (v/v) acetonitrile for 15 min and
then dried in a vacuum concentrator. The dried gel pieces were treated
with 0.01 mg/mL trypsin (sequence grade; Promega)/50 mM ammonium
bicarbonate and incubated at 378C for 16 h. The digested peptides in the
gel pieces were recovered twice with 20 mL 5% (v/v) formic acid/50% (v/v)
acetonitrile. The extracted peptides were combined and then dried in a
vacuum concentrator.
MS Analysis and Database Searching
LC-MS/MS analyses were performed on an LTQ-Orbitrap XL-HTC-PAL
system. Trypsin-digested peptides were loaded on the column (diameter
75 mm, 15 cm; L-Column, CERI) using a Paradigm MS4 HPLC pump
(Michrom BioResources) and an HTC-PAL autosampler (CTC Analytics)
and were eluted by a gradient of 5 to 45% (v/v) acetonitrile in 0.1% (v/v)
formic acid over 70 min. The eluted peptides were introduced directly into
the LTQ-Orbitrap XL MS at a flow rate of 300 nL/min and a spray voltage
of 2.0 kV. The range of MS scan was m/z 450 to 1500, and the top three
peaks were analyzed by MS/MS analysis. MS/MS spectra were com-
pared by the MASCOT server (version 2.2) against TAIR8 (The Arabidop-
sis Information Resource) with the following search parameters: set-off
threshold at 0.05 in the ion score cutoff; peptide tolerance, 10 ppm; MS/
MS tolerance, 60.8 D; peptide charge, 2+ or 3+; trypsin as enzyme
allowing up to one missed cleavage; carboxymethylation on cysteines as
a fixed modification, and oxidation on Met as a variable modification.
Chlorophyll Fluorescence and P700 analysis
The transient increase in chlorophyll fluorescence after AL had been
turned off was monitored as described (Shikanai et al., 1998). An image of
chlorophyll fluorescence was captured by a CCD camera after 1 min
illumination with AL (100 mmol photons m
22
s
21
) as described (Shikanai
et al., 1999). Chlorophyll fluorescence was measured with a MINI-PAM
portable chlorophyll fluorometer (Walz). F
PSII
was calculated as (Fm9–
Fs)/Fm9, where Fm9is the maximum fluorescence level in the light, and Fs
is the steady state fluorescence level. ETR was calculated as F
PSII
3
photon flux density (mmol photons m
22
s
21
). NPQ was calculated as (Fm –
Fm9)/Fm9. The redox change of P700 was assessed by monitoring
absorbance at 830 nm with a PAM101 chlorophyll fluorometer (Walz)
equipped with an emitter-detector unit (ED P700DW) as described
(Munekage et al., 2004). The redox kinetics of P700 was measured
according to previously described methods (Shikanai et al., 1998). Fd-
dependent PQ reduction activity was measured in ruptured chloroplasts
as described (Endo et al., 1998), with minor modifications (the pH of the
assay medium was changed to 8.0). As electron donors, 5 mM maize Fd
(Sigma-Aldrich) and 0.25 mM NADPH (Sigma-Aldrich) were used. Anti-
mycin A (Sigma-Aldrich) at 10 mM was added before measurement.
RNAi, Complementation, and Plant Transformation
For RNAi vector construction, short sequences of Arabidopsis Lhca6,
NDH18,andFKBP16-2 were cloned into the pHANNIBAL vector (Wesley
et al., 2001) between the XbaI-BamHI sites in sense orientation and
between the XhoI-KpnI sites in antisense orientation. The primers are
listed in Supplemental Table 2 online. The expression cassette was
excised with NotI and cloned into the NotI site of the binary vector
pART27 (Gleave, 1992). For complementation of the Lhca6 RNAi mutant,
3.7-kb Oryza sativa Lhca6 genomic DNA amplified by primers
59-CCAAAGCTTTAGGAGTATTCACTGCTCAG-39and 59-AATCTCGA-
GAATGGGACGTGAATGCCTGC-39was cloned into the pGWB-NB1
vector. The genomic sequence of O. sativa Lhca6 used for the comple-
mentation was modified to carry the sequence encoding the HA tag
(YPYDVPDYAG). The fusion gene was cloned into the pGWB-NB1 vector.
For overexperssionof Lhca6-HA in wild-type and crr2-2p lants, the cD NA of
Arabidopsis Lhca6 carrying the sequence encoding the HA tag was
subcloned into the pBI121 vector under the control of the CaMV 35S
promoter. The vectors were transferred into Agrobacterium tumefaciens
C58C by electroporation, and the bacteriawere used to transformwild-type
Arabidopsis or Lhca6 RNAi lines by floral dipping (Clough and Bent, 1998).
Function of the NDH-PSI Supercomplex 3637

Nucleic Acid Preparation and RT-PCR analysis
Total RNAs were isolated from Arabidopsis leaves with an RNeasy Plant
Mini Kit (Qiagen). Total RNA (5 mg) was reverse transcribed with a
SuperScript III first-strand synthesis system (Invitrogen) in a total volume
of 20 mL. The cDNA was used in 30 cycles of PCR. The PCR primers are
listed in Supplemental Table 2 online. Each set of primers covered at least
one intron sequence to eliminate amplification of the genomic DNA
sequence. RT-PCR products were separated in agarose gels and were
detected by ethidium bromide staining.
Production of Polyclonal Antisera against NDH18 and FKBP16-2
The nucleotide sequences encoding the soluble parts of NDH18 (amino
acids 41 to 131 and 155 to 212) and the mature protein of FKBP16-2
(amino acids 51 to 217) were amplified and cloned into the pET30a vector
(Novagen). Expression of the recombinant proteins was induced in
Escherichia coli BL21 (DE3) cells by 1 mM isopropylthio-b-galactoside
for 2 h, and then the cells were harvested in 300 mM NaCl and 50 mM Tris-
HCl, pH 8.0. After incubation for 30 min at 48C in the presence of 1 mg/mL
lysozyme, the inclusion bodies were pelleted from the sonicated cells by
centrifugation at 3000gfor 30 min. Recombinant proteins were then
purified from the inclusion bodies in Ni
2+
-NTA columns (Qiagen) under
denaturing conditions according to the manufacturer’s protocol. Poly-
clonal antisera were raised in a rabbit from purified recombinant protein.
Immunochemical Quantification of NDH Subunits and PSI in
the Supercomplex
Antiserum against Arabidopsis PsaA was produced in rabbits using the
N-terminal part of recombinant PsaA protein (amino acids 1 to 77) as
antigen. The nucleotide sequence encoding this N-terminal part was
amplified by PCR using the primers 59-GGCGAATTCATGATTAT-
TCGTTCGCCGG-39and 59-GATCTCGAGTTGGCCGAAATGGGCAC-39.
The amplified sequence was fused to the Nus-tag in the pET43.1a vector
(Novagen). The PsaA-Nus recombinant protein was induced in DE3 cells
and purified in a Ni
2+
-NTA column according to the manufacturer’s
protocol. The protein contents were determined with a Bio-Rad protein
assay kit. The band corresponding to the NDH-PSI supercomplex was
excised from BN gel and further denatured in gel as described (Peng
et al., 2008) and then directly used for SDS-PAGE together with the
recombinant protein. Immunoblot analysis using an ECL Advance West-
ern Blotting Detection Kit (GE Healthcare) was performed according to
standard procedures. The signal was visualized by an LAS3000 chemi-
luminescence analyzer (Fuji Film). Immunoblots from three independent
experiments were quantified by Imagemaster software (Amersham
Pharmacia Biotech).
Phylogenetic Analysis
Protein sequences of FKBP13 and FKBP16-2 proteins (shown in Sup-
plemental Data Set 3 online) and Lhca family proteins (shown in Supple-
mental Data Set 4 online) were aligned using the ClustalW program with
default settings (http://clustalw.ddbj.nig.ac.jp/top-e.html) and adjusted
manually. The phylogenetic tree was constructed using TreeView soft-
ware (version 1.6.6) (http:/taxonomy.zoology.gla.ac.uk/rod/treeview.
html).
Accession Numbers
Sequence data from this article can be found in the Arabidopsis Genome
Initiative or GenBank/EMBL databases under the following accession
numbers: At (Arabidopsis thaliana) FKBP13 (AT5G45680), At FKBP16-2
(AT4G39710), Os (Oryza sativa) FKBP13 (OS06G0663800), Os FKBP16-2
(OS02G0751600), Zm (Zea mays) FKBP13 (EU958134), Zm FKBP16-2
(EU955427), Gm (Glycine max) FKBP16-2 (DB960344), At NDH18
(AT5G43750), Gm NDH18 (CD394214), Nt (Nicotiana tabacum) NDH18
(EB679832), Os NDH18 (OS01G0929100), Zm NDH18 (DV514173), At
Lhca6 (AT1G19150), Rr (Raphanus raphanistrum) Lhca6 (EV549333), Br
(Brassica rapa) Lhca6 (EX086756), Vv (Vitis vinifera) Lhca6 (EC925446), Mt
(Medicago truncatula) Lhca6 (EV260067), Vu (Vigna unguiculata) Lhca6
(FG880635), Aa (Artemisia annua) Lhca6 (EY082250), Cs (Citrus sinensis)
Lhca6 (EY664891), Ta (Triticum aestivum) Lhca6 (CJ883523), Sb (Sor-
ghum bicolor) Lhca6 (CN151166), So (Saccharum officianrum) Lhca6
(CA294625), Hv (Horheum vulgare) Lhca6 (BI953315 and AJ432207),
Nt Lhca6 (DW000027), Gm Lhca6 (EH258354), Zm Lhca6 (DV507315),
Os Lhca6 (AK067780), Os Lhca2 (AK104651), At Lhca5 (At1g45474),
Rr Lhca5 (EV538184), Br Lhca5 (EX087930), Vv Lhca5 (EC934896), Mt
Lhca5 (CX519116), Vu Lhca5 (FG880536), Aa Lhca5 (EY103979),
Cs Lhca5 (EN184748), Ta Lhca5 (CJ723109), Sb Lhca5 (CN147414), Hv
Lhca5 (BE422210 and BI950547), At Lhca4 (AT3G47470), At Lhca3
(AT1G61520), At Lhca2 (AT3G61470), At Lhca1 (AT3G54890), Cr (Chla-
mydomonas reinhardtii) Lhca1 (AAD03734), Cr Lhca2 (XP_001691031),
Cr Lhca3 (XP_001701405), Cr Lhca4 (EDP08012), Cr Lhca5
(XP_001702730), Cr Lhca6 (XP_001698070), Cr Lhca7 (XP_001691959),
Cr Lhca8 (EDP08179), and Cr Lhca9 (XP_001692548).
Supplemental Data
The following materials are available in the online version of this article.
Supplemental Figure 1. Amino Acid Sequence Alignment of NDH18.
Supplemental Figure 2. NDH-PSI Supercomplex Content in the
ndh18 and fkbp16-2 Lines.
Supplemental Figure 3. Localization Analysis of NDH Subunits in
Chloroplasts of the Wild Type, ndhl,ppl2,crr4-3, and ndh18 Mutants.
Supplemental Figure 4. Visible Phenotype of lhca6 Mutant.
Supplemental Figure 5. NDH-PSI Supercomplex Content in lhca5
and lhca6.
Supplemental Figure 6. Analysis of Thylakoid Protein Complex from
Wild-type and lhca6 Mature Leaves.
Supplemental Figure 7. Amino Acid Sequence Alignments of Lhca6
and Lhca2.
Supplemental Figure 8. State 1–State 2 Transitions in Wild-Type,
stn7,andlhca6 Plants.
Supplemental Table 1. The rValues between NDH Complex-Related
Genes with Lhca6,FKBP16-2,andNDH18.
Supplemental Table 2. Primers Used in This work.
Supplemental Data Set 1. The Total Proteins Identified from Bands
I and II.
Supplemental Data Set 2. The Total Proteins Identified from PSI
Monomer from Wild Type, crr2-2, and lhca6.
Supplemental Data Set 3. Text File of Alignment Corresponding to
the Phylogenetic Tree in Figure 1A.
Supplemental Data Set 4. Text File of Alignment Corresponding to
the Phylogenetic Tree in Figure 6A.
ACKNOWLEDGMENTS
We thank Tsuyoshi Endo (Kyoto University, Kyoto, Japan), Amane
Makino (Tohoku University, Sendai, Japan), and Kentaro Ifuku (Kyoto
University, Kyoto, Japan) for giving us antibodies. This work was
3638 The Plant Cell

supported by Grant 17GS0316 for Creative Science Research from the
Ministry of Education, Culture, Sports, Science, and Technology of
Japan and a grant from the Ministry of Agriculture, Forestry, and
Fisheries of Japan (Genomics for Agricultural Innovation; GPN0008).
This work was also supported by a grant from the Japan Society for the
Promotion of Science (JSPS-19-07142) to L.P.
Received May 19, 2009; revised October 1, 2009; accepted October 23,
2009; published November 10, 2009.
REFERENCES
Amunts, A., Drory, O., and Nelson, N. (2007). The structure of a plant
photosystem I supercomplex at 3.4 A resolution. Nature 447: 58–63.
Battchikova, N., and Aro, E.-M. (2007). Cyanobacterial NDH-1 com-
plexes: Multiplicity in function and subunit composition. Physiol.
Plant. 131: 22–32.
Battchikova, N., Zhang, P., Rudd, S., Ogawa, T., and Aro, E.-M.
(2005). Identification of NdhL and Ssl1690 (NdhO) in NDH-1L and
NDH-1M complexes of Synechocystis sp. PCC 6803. J. Biol. Chem.
280: 2587–2595.
Burrows, P.A., Sazanov, L.A., Svab, Z., Maliga, P., and Nixon, P.J.
(1998). Identification of a functional respiratory complex in chloro-
plasts through analysis of tobacco mutants containing disrupted
plastid ndh genes. EMBO J. 17: 868–876.
Clough, S.J., and Bent, A.F. (1998). Floral dip: A simplified method for
Agrobacterium-mediated transformation of Arabidopsis thaliana. Plant
J. 16: 735–743.
DalCorso, G., Pesaresi, P., Masiero, S., Aseeva, E., Schu¨ nemann, D.,
Finazzi, G., Joliot, P., Barbato, R., and Leister, D. (2008). A complex
containing PGRL1 and PGR5 is involved in the switch between linear
and cyclic electron flow in Arabidopsis. Cell 132: 273–285.
Edvardsson, A., Eshaghi, S., Vener, A.S., and Andersson, B. (2003).
The major peptidyl-prolyl isomerase activity in thylakoid lumen of
plant chloroplasts belongs to a novel cyclophilin TLP20. FEBS Lett.
542: 137–141.
Endo, T., Shikanai, T., Sato, F., and Asada, K. (1998). NAD(P)H
dehydrogenase-dependent, antimycin A-sensitive electron donation
to plastoquinone in tobacco chloroplast. Plant Cell Physiol. 39:
1226–1231.
Fujiwara, M., Hamada, S., Hiratsuka, M., Fukao, Y., Kawasaki, T.,
and Shimamoto, K. (2009). Proteome analysis of detergent resistant
membranes (DRMs) associated with OsRac1 mediated innate immu-
nity in rice. Plant Cell Physiol. 50: 1191–1200.
Ganeteg, U., Klimmek, F., and Jansson, S. (2004). Lhca5 – An LHC-
type protein associated with photosystem I. Plant Mol. Biol. 54:
641–651.
Gleave, A.P. (1992). A versatile binary vector system with a T-DNA
organisational structure conducive to efficient integration of cloned
DNA into the plant genome. Plant Mol. Biol. 20: 1203–1207.
Guedeney, G., Corneile, S., Cuine, S., and Peltier, G. (1996). Evidence
for an association of ndhB, ndhJ gene products and ferredoxin-NADP
reductase as components of a chloroplastic NAD(P)H dehydrogenase
complex. FEBS Lett. 378: 277–280.
Gupta, R., Mould, R.M., He, Z., and Luan, S. (2002). A chloroplast
FKBP interacts with and affects the accumulation of Rieske subunit of
cytochrome bf complex. Proc. Natl. Acad. Sci. USA 99: 15806–15811.
Hashimoto, M., Endo, T., Peltier, G., Tasaka, M., and Shikanai, T.
(2003). A nucleus-encoded factor, CRR2, is essential for the expres-
sion of chloroplast ndhB in Arabidopsis. Plant J. 36: 541–549.
He, Z., Li, L., and Luan, S. (2004). Immunophilins and parvulins.
Superfamily of peptidyl prolyl isomerases in Arabidopsis. Plant
Physiol. 134: 1248–1267.
Herranen, M., Battchikova, N., Zhang, P., Graf, A., Sirpio
¨,S.,
Paakkarinen, V., and Aro, E.-M. (2004). Towards functional proteo-
mics of membrane protein complexes in Synechocystis sp. PCC
6803. Plant Physiol. 134: 470–481.
Ishihara, S., Takabayashi, A., Ido, K., Endo, T., Ifuku, K., and Sato, F.
(2007). Distinct functions for the two PsbP-like proteins PPL1 and
PPL2 in the chloroplast thylakoid lumen of Arabidopsis. Plant Physiol.
145: 668–679.
Ishikawa, N., Takabayashi, A., Ishida, S., Hano, Y., Endo, T., and
Sato, F. (2008). NDF6: A thylakoid protein specific to terrestrial plants
is essential for activity of chloroplastic NAD(P)H dehydrogenase in
Arabidopsis. Plant Cell Physiol. 49: 1066–1073.
Jansson, S. (1999). A guide to the Lhc genes and their relatives in
Arabidopsis. Trends Plant Sci. 4: 236–240.
Jensen, P.E., Bassi, R., Boekema, E.J., Dekker, J.P., Jansson, S.,
Leister, D., Robinson, C., and Scheller, H.V. (2007). Structure,
function and regulation of plant photosystem I. Biochim. Biophys.
Acta 1767: 335–352.
Joe
¨t, T., Cournac, L., Horva
´th, E.M., Medgyesy, P., and Peltier, G.
(2001). Increased sensitivity of photosynthesis to antimycin A induced
by inactivation of the chloroplast ndhB gene. Evidence for a partic-
ipation of the NADH-dehydrogenase complex to cyclic electron flow
around photosystem I. Plant Physiol. 125: 1919–1929.
Klimmek, F., Sjo
¨din, A., Noutsos, C., Leister, D., and Jansson, S.
(2006). Abundantly and rarely expressed Lhc protein genes exhibit
distinct regulation patterns in plants. Plant Physiol. 140: 793–804.
Kotera, E., Tasaka, M., and Shikanai, T. (2005). A pentatricopeptide
repeat protein is essential for RNA editing in chloroplasts. Nature 433:
326–330.
Kouril, R., Zygadlo, A., Arteni, A.A., de Wit, C.D., Dekker, J.P.,
Jensen, P.E., Scheller, H.V., and Boekema, E.J. (2005). Structural
characterization of a complex of photosystem I and light-harvesting
complex II of Arabidopsis thaliana. Biochemistry 44: 10935–10940.
Lucinski, R., Schmid, V.H., Jansson, S., and Klimmek, F. (2006).
Lhca5 interaction with plant photosystem I. FEBS Lett. 580: 6485–
6488.
Majeran, W., Zybailov, B., Ytterberg, A.J., Dunsmore, J., Sun, Q.,
and van Wijk, K.J. (2008). Consequences of C
4
differentiation for
chloroplast membrane proteomes in maize mesophyll and bundle
sheath cells. Mol. Cell. Proteomics 7: 1609–1638.
Matsubayashi, T., Wakasugi, T., Shinozaki, K., Yamaguchi-
Shinozaki, K., Zaita, N., Hidaka, T., Meng, B.Y., Ohto, C., Tanaka,
M., Kato, A., Maruyama, T., and Sugiura, M. (1987). Six chloroplast
genes (ndhA-F) homologous to human mitochondrial genes encoding
components of the respiratory chain NADH dehydrogenase are
actively expressed: determination of the splice sites in ndhA and
ndhB pre-mRNAs. Mol. Gen. Genet. 210: 385–393.
Matsuo, M., Endo, T., and Asada, K. (1998). Properties of the respi-
ratory NAD(P)H dehydrogenase isolated from the cyanobacterium
Synechocystis PCC6803. Plant Cell Physiol. 39: 263–267.
Melkozernov, A.N., Barber, J., and Blankenship, R.E. (2006). Light
harvesting in photosystem I supercomplexes. Biochemistry 45:
331–345.
Mi, H., Endo, T., Ogawa, T., and Asada, K. (1995). Thylakoid
membrane-bound, NADPH-specific pyridine nucleotide dehydrogenase
complex mediates cyclic electron transport in the cyanobacterium
Synechocystis sp. PCC 6803. Plant Cell Physiol. 36: 661–668.
Munekage, Y., Hashimoto, M., Miyake, C., Tomizawa, K., Endo, T.,
Tasaka, M., and Shikanai, T. (2004). Cyclic electron flow around
photosystem I is essential for photosynthesis. Nature 429: 579–582.
Munekage, Y., Hojo, M., Meurer, J., Endo, T., Tasaka, M., and
Function of the NDH-PSI Supercomplex 3639

Shikanai, T. (2002). PGR5 is involved in cyclic electron flow around
photosystem I and is essential for photoprotection in Arabidopsis. Cell
110: 361–371.
Nelson, N., and Yocum, C.F. (2006). Structure and function of photo-
systems I and II. Annu. Rev. Plant Biol. 57: 521–565.
Ogawa, T. (1992). Identification and characterization of the ictA/ndhL
gene product essential to inorganic carbon transport of Synechocystis
PCC6803. Plant Physiol. 99: 1604–1608.
Ogawa, T., and Mi, H. (2007). Cyanobacterial NADPH dehydrogenase
complexes. Photosynth. Res. 93: 69–77.
Okegawa, Y., Kagawa, Y., Kobayashi, Y., and Shikanai, T. (2008).
Characterization of factors affecting the activity of photosystem I cyclic
electron transport in chloroplasts. Plant Cell Physiol. 49: 825–834.
Peng, L., Ma, J., Chi, W., Guo, J., Zhu, S., Lu, Q., Lu, C., and Zhang,
L. (2006). Low PSII accumulation1 is involved in the efficient assembly
of photosystem II in Arabidopsis thaliana. Plant Cell 18: 955–969.
Peng, L., Shimizu, H., and Shikanai, T. (2008). The chloroplast NAD(P)
H dehydrogenase complex interacts with photosystem I in Arabidop-
sis. J. Biol. Chem. 283: 34873–34879.
Prommeenate, P., Lennon, A.M., Markert, C., Hippler, M., and Nixon,
P.J. (2004). Subunit composition of NDH-1 complexes of Synecho-
cystis sp. PCC 6803. J. Biol. Chem. 279: 28165–28173.
Rumeau, D., Becuwe-Linka, N., Beyly, A., Louwagie, M., Garin, J.,
and Peltier, G. (2005). New subunits NDH-M, -N, and -O, encoded by
nuclear genes, are essential for plastid Ndh complex functioning in
higher plants. Plant Cell 17: 219–232.
Sazanov, L.A., Burrows, P.A., and Nixon, P.J. (1998). The plastid ndh
genes code for an NADH-specific dehydrogenase: Isolation of a
complex I analogue from pea thylakoid membranes. Proc. Natl. Acad.
Sci. USA 95: 1319–1324.
Shikanai, T. (2007a). Cyclic electron transport around photosystem I:
Genetic approaches. Annu. Rev. Plant Biol. 58: 199–217.
Shikanai, T. (2007b). The NAD(P)H dehydrogenase complex in photo-
synthetic organisms: Subunit composition and physiological function.
Funct. Plant Sci. Biotechnol. 1: 129–137.
Shikanai, T., Endo, T., Hashimoto, T., Yamada, Y., Asada, K., and
Yokota, A. (1998). Directed disruption of the tobacco ndhB gene
impairs cyclic electron flow around photosystem I. Proc. Natl. Acad.
Sci. USA 95: 9705–9709.
Shikanai, T., Munekage, Y., Shimizu, K., Endo, T., and Hashimoto, T.
(1999). Identification and characterization of Arabidopsis mutants with
reduced quenching of chlorophyll fluorescence. Plant Cell Physiol. 40:
1134–1142.
Shimizu, H., Peng, L., Myouga, F., Motohashi, R., Shinozaki, K., and
Shikanai, T. (2008). CRR23/NdhL is a subunit of the chloroplast NAD
(P)H dehydrogenase complex in Arabidopsis. Plant Cell Physiol. 49:
835–842.
Shimizu, H., and Shikanai, T. (2007). Dihydrodipicolinate reductase-
like protein, CRR1, is essential for chloroplast NAD(P)H dehydroge-
nase in Arabidopsis. Plant J. 52: 539–547.
Sirpio
¨, S., Allahverdiyeva, Y., Holmstro
¨m, M., Khrouchtchova, A.,
Haldrup, A., Battchikova, N., and Aro, E.-M. (2009a). Novel nuclear-
encoded subunits of the chloroplast NAD(P)H dehydrogenase
complex. J. Biol. Chem. 284: 905–912.
Sirpio
¨, S., Holmstro
¨m, M., Battchikova, N., and Aro, E.-M. (2009b).
AtCYP20-2 is an auxiliary protein of the chloroplast NAD(P)H dehy-
drogenase complex. FEBS Lett. 583: 2355–2358.
Storf, S., Stauber, E.J., Hippler, M., and Schmid, V.H.R. (2004).
Proteomic analysis of the photosystem I light-harvesting antenna
in tomato (Lycopersicon esculentum). Biochemistry 43: 9214–
9224.
Suorsa, M., Sirpio
¨, S., and Aro, E.-M. (July 21, 2009). Towards
characterization of the chloroplast NAD(P)H dehydrogenase complex.
Mol. Plant http://dx.doi.org/10.1093/mp/ssp052.
Takabayashi, A., Ishikawa, N., Obayashi, T., Ishida, S., Obokata, J.,
Endo, T., and Sato, F. (2009). Three novel subunits of Arabidopsis
chloroplastic NAD(P)H dehydrogenase identified by bioinformatic and
reverse genetic approaches. Plant J. 57: 207–219.
Wesley, S.V., et al. (2001). Construct design for efficient, effective and
high throughput gene silencing in plants. Plant J. 27: 581–590.
Zhang, H., Wang, J., and Goodman, H.M. (1994). Differential expres-
sion in Arabidopsis of Lhca2, a PSI cab gene. Plant Mol. Biol. 25:
551–557.
Zhang, P., Battchikova, N., Jansen, T., Appel, J., Ogawa, T., and
Aro, E.-M. (2004). Expression and functional roles of the two distinct
NDH-1 complexes and the carbon acquisition complex NdhD3/
NdhF3/CupA/Sll1735 in Synechocystis sp PCC 6803. Plant Cell 16:
3326–3340.
Zhang, P., Battchikova, N., Paakkarinen, V., Katoh, H., Iwai, M.,
Ikeuchi, M., Pakrasi, H.B., Ogawa, T., and Aro, E.-M. (2005).
Isolation, subunit composition and interaction of the NDH-1 com-
plexes from Thermosynechococcus elongatus BP-1. Biochem. J. 390:
513–520.
Zolla, L., Rinalducci, S., and Timperio, A.M. (2007). Proteomic anal-
ysis of photosystem I components from different plant species.
Proteomics 7: 1866–1876.
3640 The Plant Cell