Content uploaded by Alexander Ruban
Author content
All content in this area was uploaded by Alexander Ruban on Jan 06, 2023
Content may be subject to copyright.
Absence of the Lhcb1 and Lhcb2 proteins of the
light-harvesting complex of photosystem II ± effects
on photosynthesis, grana stacking and ®tness
Jenny Andersson
1,
, Mark Wentworth
2
, Robin G. Walters
2,y
, Caroline A. Howard
2
, Alexander V. Ruban
2
, Peter Horton
2
and Stefan Jansson
1
1
Umea
ÊPlant Science Centre, Department of Plant Physiology, Umea
ÊUniversity, S-901 87 Umea
Ê, Sweden, and
2
Robert Hill Institute, Department of Molecular Biology and Biotechnology, University of Shef®eld, Western Bank,
Shef®eld S10 2TN, UK
Received 30 October 2002; revised 24 March 2003; accepted 7 May 2003.
For correspondence: (fax 46 90 7866676; e-mail jenny.andersson@plantphys.umu.se).
y
Present address: Department of Plant Sciences, University of Oxford, South Parks Road, Oxford OX1 3RB, UK.
Summary
We have constructed Arabidopsis thaliana plants that are virtually devoid of the major light-harvesting
complex, LHC II. This was accomplished by introducing the Lhcb2.1 coding region in the antisense orienta-
tion into the genome by Agrobacterium-mediated transformation. Lhcb1 and Lhcb2 were absent, while
Lhcb3, a protein present in LHC II associated with photosystem (PS) II, was retained. Plants had a pale green
appearance and showed reduced chlorophyll content and an elevated chlorophyll a/bratio. The content of
PS II reaction centres was unchanged on a leaf area basis, but there was evidence for increases in the relative
levels of other light harvesting proteins, notably CP26, associated with PS II, and Lhca4, associated with PS I.
Electron microscopy showed the presence of grana. Photosynthetic rates at saturating irradiance were the
same in wild-type and antisense plants, but there was a 10±15% reduction in quantum yield that re¯ected
the decrease in light absorption by the leaf. The antisense plants were not able to perform state transitions,
and their capacity for non-photochemical quenching was reduced. There was no difference in growth
between wild-type and antisense plants under controlled climate conditions, but the antisense plants per-
formed worse compared to the wild type in the ®eld, with decreases in seed production of up to 70%.
Keywords: antisense, Arabidopsis, ®tness, LHC II, state transitions, photosynthesis.
Introduction
Over 60% of all the chlorophyll in plants is bound to light-
harvesting-antenna complexes, a group of related pig-
ment-binding proteins encoded by the Lhc gene family
(Jansson, 1999). These antenna complexes absorb sun-
light and transfer the excitation energy to the core com-
plexes of photosystem (PS) II and PS I in order to drive
photosynthetic electron transport. Six different proteins
comprise the light-harvesting antenna of PS II ± three
minor proteins CP29, CP26 and CP24 encoded by the
Lhcb4,Lhcb5 and Lhcb6 genes, and the major complex
LHC II, which consists of three proteins encoded by the
Lhcb1,Lhcb2 and Lhcb3 genes ± of which the Lhcb1 and
Lhcb2 proteins are by far the most abundant. These com-
plexes bind chlorophyll a, chlorophyll band xanthophylls,
but in different amounts. LHC II seems to bind seven
chlorophyll a, ®ve chlorophyll b, two luteins, one neox-
anthin and a variable content of xanthophyll cycle caro-
tenoids (Ruban et al., 1999). Whilst the minor complexes
are monomers, LHC II is trimeric (Ku
Èhlbrandt et al., 1994).
The core complex of PS II is dimeric and binds two copies
per dimer of each of the minor light-harvesting com-
plexes, two strongly bound LHC II trimers (S-LHC II) and
one to two moderately bound LHC II trimers (M-LHC II) to
form the PS II supercomplex (Boekema et al.,2000,
Yakushevska et al., 2001). Although they transfer energy
to PS II, further LHC II units consisting only of Lhcb1 and
Lhcb2 exist that are not present in the PS II supercomplex.
The amount of this population of peripheral LHC II is
highly variable, with plants grown in high light containing
less of them than do plants grown in low light (for exam-
ple, Anderson, 1986; Bailey et al., 2001; Ma
Èenpa
Èa
Èand
Andersson, 1989).
The Plant Journal (2003) 35, 350±361
350 ß2003 Blackwell Publishing Ltd
doi: 10.1046/j.1365-313X.2003.01811.x
LHC II has been implicated in the short-term regulation of
light harvesting. First, the fraction of LHC II bound to PS II is
regulated by state transitions via phosphorylation of LHC II
by a speci®c kinase (reviewed in Allen and Forsberg, 2001).
The LHC II kinase is controlled by the redox state of the
inter-photosystem electron transfer chain, notably the Q
o
site of the cytochrome b
6
fcomplex (Vener et al., 1997), and
therefore state transitions may function to maintain an
excitation balance of the photosystems in limiting light.
In State I, peripheral LHC II delivers energy to PS II, but in
State II, it dissociates from PS II and instead delivers energy
to PS I (Lunde et al., 2000). LHC II has been proposed to be
involved in the stacking of thylakoid membranes to form
the grana (reviewed in Allen and Forsberg, 2001), a process
that has been suggested to play a crucial role in state
transitions. Second, a direct role of LHC II in the regulation
of the dissipation of excess excitation energy from the
antenna of PS II, a process known as feedback de-excitation
(Ku
Èlheim et al., 2002) formerly named the qE type of non-
photochemical quenching (NPQ), has also been suggested
(Horton et al., 1991).
There is a great deal of uncertainty concerning all of
these roles of LHC II; even the physiological importance of
LHC II has not been proven. For example, chlorophyll b-
less mutants de®cient in LHC II appear to grow as well as
wild-type plants do (Habash et al.,1994),asdomutants
de®cient in the PS I subunit H (Lunde et al., 2000), which
leads to the loss of state transitions. The requirement of
LHC II for grana stacking has also been questioned as there
is evidence that the chlorophyll b-less mutants can form
grana (Bassi et al., 1985; Murray and Kohorn, 1991).
Similarly, the chlorophyll b-less mutants still show sig-
ni®cant levels of feedback de-excitation (Gilmore et al.,
1996; Ha
Èrtel and Lokstein, 1995), indicating that LHC II
cannot have an obligatory role in the process. Conversely,
a recently discovered mutant of Chlamydomonas with
reduced feedback de-excitation has been shown to have
a mutation in a gene encoding one of the main LHC II
proteins (Elrad et al., 2002).
Genetic methods are powerful tools in physiology
research because comparison of mutant and wild-type
plants can de®ne the function of the missing gene/protein.
Forward genetics has not so far led to the isolation of a
mutant lacking a speci®c LHC protein, despite screening of
many pigment-de®cient mutants of barley (e. g. Bossmann
et al., 1997) and Arabidopsis (M. Rosenqvist and S. Jans-
son, unpublished). Instead, many experiments have uti-
lised plants grown in conditions known to reduce the
amount of LHC II (for example, intermittent light) or
mutants lacking chlorophyll b, where the LHC II proteins
fail to accumulate. However, such plants lack not only LHC II
but also most of the other LHC proteins (Kro
Âlet al., 1995).
Hence, these studies rather show the function of the chlor-
ophyll b-binding antenna as a whole, not speci®cally LHC II.
In part, this lack of speci®city has also contributed to the
uncertainties concerning the role of LHC II.
Our approach to explore the function of the light-harvest-
ing proteins makes use of reverse genetics. In this paper, we
describe an antisense construct that inhibited successfully
not only the Lhcb2 genes, but also the related Lhcb1 genes,
producing Arabidopsis plants lacking practically all the
protein subunits that form the LHC II complexes. We show
that, although plants devoid of Lhcb1 and Lhcb2 are largely
unaffected in some aspects of photosynthesis, they show a
marked reduction in ®tness under ®eld conditions.
Results
Construction of Lhcb2 antisense lines
After vacuum in®ltration of ¯owering Arabidopsis with
Agrobacterium tumefaciens carrying the antisense con-
struct and a kanamycin resistance gene, seeds were col-
lected. Germination on MS plates containing kanamycin
revealed eight putative transformants that were transferred
to soil. Immunoblot analysis of these plants with antibodies
against Lhcb2 showed that some of them had severely
decreased amounts of this protein. Two lines were chosen
for further analysis, based on high antisense effect. These
lines were self-pollinated for two generations, and either
second- or third-generation plants were used in this study.
The degree of antisense effect varied between lines ± in one
line (asLhcb2-12), Lhcb2 was not detectable, whereas in the
other (asLhcb2-1), small amounts of the protein remained
(Figure 1a,b). In addition, Lhcb1 was not detectable in
asLhcb2-12.
Cross-antisensing between Lhcb1 and Lhcb2
We sought to determine if the reduction in Lhcb1 is the
result of an effect on RNA level (`cross-antisensing') based
on the high similarity between Lhcb2 and Lhcb1, or
whether it was caused by decreased stability of Lhcb1
in the absence of Lhcb2. RNA was prepared from wild-type
and antisense leaf material sampled 3 h into the photo-
period, separated on agarose gel and blotted onto Hybond
N membrane. When membranes were hybridised with
speci®cprobesagainstLhcb1 and Lhcb2 genes, it was
evident that the expression of these genes was minimal in
both antisense lines (Figure 1c). Apparently, the Lhcb2.1
construct was not only able to reduce the mRNA level
from all three Lhcb2 genes beyond the level of detec-
tion, but also just as ef®cient in abolishing Lhcb1 gene
expression.
We also investigated potential changes in the Lhcb3
transcript level. High similarity between Lhcb1 and
Lhcb3 transcripts (both in size and sequence) presented
Arabidopsis lacking Lhcb1 and Lhcb2 351
ßBlackwell Publishing Ltd, The Plant Journal, (2003), 35, 350±361
dif®culties in distinguishing between the signals from these
mRNA species using standard gel blot technique ± the
hybridisation signal from the Lhcb3 transcript can (in
wild-type plants) be obscured by cross-hybridisation to
the Lhcb1 transcripts, which can be present in 10-fold
higher levels (Jansson, 1999). In contrast to Lhcb1 and
Lhcb2,Lhcb3 is in Arabidopsis encoded by a single gene
allowing detection with reverse transcription (RT)-PCR.
RNA samples (as above) were reverse-transcribed, the
Lhcb3 and actin (as a control of template quantity) cDNA
fragments were PCR-ampli®ed and the products were sepa-
rated on ethidium-bromide-stained agarose gels. Analysis
of the intensity of the bands revealed no difference between
the wild-type and the antisense lines in the amount of Lhcb3
transcript (Figure 1c).
Both Lhcb1 and Lhcb2 are encoded by gene families in
Arabidopsis. An analysis of Arabidopsis expressed
sequence tag (EST) clones indicated the possible presence
of four Lhcb2 genes, two (Lhcb2.1 and Lhcb2.3) differing by
only one amino acid residue (Jansson, 1999). When the
complete Arabidopsis genome sequence became available
(The Arabidopsis Genome Initiative, 2000), it was apparent
that only three loci were present. We have followed Legen
et al. (2001) and re-named Lhcb2.4 as Lhcb2.3. The
sequence formerly known as Lhcb2.3 is seemingly allelic
to Lhcb2.1.Lhcb1 is encoded by ®ve genes.
Antisense inhibition may function by a mechanism in
which double-stranded RNA (dsRNA) molecules silence
similar genes (for example, Hutvagner and Zamore,
2002). In this process, double-stranded (ds) RNA (in this
case, a transgene antisense construct±endodgenous sense
mRNA duplex) is cleaved into approximately 22-bp frag-
ments (Bernstein et al., 2001). These fragments are incor-
porated into an enzyme complex and used as a template for
the degradation of mRNAs that contain the 22-nt sequence
in question. This implies that the sequence requirement for
ef®cient antisense inhibition of heterologous genes should
be absolute sequence identity in an approximately 22-nt
sequence. We compared the nucleotide and amino acid
sequences of the genes encoding LHC II proteins, obtained
from the genomic database.
The polypeptide sequences are highly similar. The iden-
tity between the deduced amino acid sequence of Lhcb2.1
(that was used for the antisense construct) and the Lhcb1
genes is >81%, Lhcb2 genes >98% and Lhcb3 >73%. The
diversity at the nucleotide level is slightly higher (compared
to Lhcb2.1, the identity is 72±74% of the Lhcb1 genes, 96%
of Lhcb2.2, 85% of Lhcb2.3 and 68% of Lhcb3).
We deduced, for each gene, the number of exact 22-nt
matches with the antisense gene. The three Lhcb2 genes
contain numerous 22-nt perfect matches, the ®ve Lhcb1
genes two each and Lhcb3 none. The two perfect matches
in the Lhcb1 genes were contiguous, making one 23-nt
perfect match. This is fully consistent with the assumption
that one approximately 22-nt match is suf®cient for ef®cient
antisense inhibition.
Decreased chlorophyll content in the antisense plants
The antisense lines were clearly less pigmented than the
wild type, having a pale green colour. For plants of the same
age, the chlorophyll content per unit leaf area decreased by
35% in asLhcb2-12 and 20 % in asLhcb2-1. This resulted in a
decrease in light absorption by the leaf: in asLhcb2-12, an
absorptivity of 82% was measured compared to 91% in the
wild type. There was an increase in the chlorophyll a/bratio,
again with asLhcb2-1 giving a value between the wild type
and asLhcb2-12 (Table 1). The contents of functional reac-
tion centres of PS II were found to be higher when
Figure 1. Protein and RNA analysis.
(a) Wild-type (WT) and asLhcb2-12 (2-12) thylakoid samples were probed
with antibodies speci®c for the proteins indicated to the left of each blot. The
amount of sample in each lane corresponds to 3 mg of chlorophyll.
(b) Wild-type (WT) and asLhcb2-1 (2-1) thylakoid samples probed as in (a).
(c) RNA samples were probed with radiolabelled, gene-speci®c PCR frag-
ments hybridising with Lhcb1 or Lhcb2 as indicated to the left of the lanes.
The most prominent rRNA band visible on the ethidium-bromide-stained
gel is shown as an evidence of the presence of similar amounts of RNA in
each lane. RT-PCR was performed with primers amplifying the Lhcb3
transcript. A control of template quantity was provided by ampli®cation
of the actin2 transcript.
ßBlackwell Publishing Ltd, The Plant Journal, (2003), 35, 350±361
352 Jenny Andersson et al.
expressed on a chlorophyll basis (Table 1). We found an
increase in PS II/chlorophyll of about 25% in asLhcb2-1 and
60% in asLhcb2-12 compared to the wild type (Table 1).
These changes can be completely accounted for by the
reduction in chlorophyll content arising from the loss of
Lhcb1 and Lhcb2, with no changes in the content of PS II per
unit leaf area.
Changes in the composition of the photosynthetic
antenna in the antisense lines
Leaf material from wild-type and antisense lines was ana-
lysed. Thylakoid proteins were prepared from these leaves,
and samples corresponding to 3 mg chlorophyll each were
separated on SDS±PAGE, blotted onto nitrocellulose mem-
branes and hybridised with antibodies against all LHC
proteins as well as PsbS and CP43. CP43 is a part of the
PS II core complex and is present in one copy per PS II
(Barber et al., 1997), so that it could be used as an internal
reference to allow determination of the amounts of each
LHC protein relative to PS II content. The validity of the
quanti®cation of the immunoblots was established from
the linearity of the contents of both Lhcb1 and CP26 mea-
sured from densitometry of immunoblots in which the
amount of chlorophyll applied to the gel was varied from
0.5 to 10 mg (data not shown).
Large differences were found between the wild type and
the asLhcb2-12 plants (Figure 1a). The CP43 blot shows that
the number of PS II reaction centres per unit chlorophyll
increased by about 50% in asLhcb2-12 in line with the
content of functional reaction centres (Table 1). As shown
in Figure 1(a), the asLhcb2-12 lacked both Lhcb1 and Lhcb2,
the two proteins that constitute the major fraction of LHC II.
The third component of LHC II, Lhcb3, occurred at similar
levels in the asLhcb2-12 as in the wild type on a chlorophyll
basis, and consequently asLhcb2-12 appears to have
slightly less Lhcb3 per PS II (approximately 75% of the
wild-type level). The level of PsbS was also unchanged
on a chlorophyll basis, indicating that PS II of asLhcb2-12
contained approximately half the amount of PsbS as PS II of
the wild type.
The asLhcb2-12 plants had greatly increased amounts of
the PS II minor antenna protein CP26, and to a lesser extent
CP29 and CP24. Densitometric scanning of the blots indi-
cated that the CP26/chl ratio increased at least sixfold, and
the CP26/CP43 ratio at least fourfold. The PS I light-harvest-
ing proteins were all more abundant (per chlorophyll) in the
antisense plants than in the wild type, with the Lhca4
protein accumulating more than the other proteins. The
increase of Lhca1, Lhca2 and Lhca3 proteins appeared
similar to that shown by CP29 and CP24.
Photosynthetic function in antisense plants
The photosynthetic capacity (P
max
) of the antisense plants
expressed on a leaf area basis was the same as in wild-type
plants, again expected if the decrease in chlorophyll con-
tent arose only from a loss of LHC II (Table 1). The rate of
CO
2
uptake was also recorded in attached leaves in
350 p.p.m. CO
2
as a function of incident irradiance
(Figure 2). The irradiance curves were almost identical
for the antisense plants and the wild type. However, exam-
ination of the light-limited region of the curve revealed a
10±15% decrease in apparent quantum yield. Nevertheless,
the quantum yield based upon light absorbed was almost
the same.
Antisense plants have a reduced capacity for
feedback de-excitation
Illumination of both wild-type and antisense plants caused
NPQ of chlorophyll ¯uorescence (Figure 3a). In the ®rst
illumination period, this was biphasic, and in the second,
after a brief intervening dark relaxation period, the slower
phase was eliminated. This is a typical behaviour for Ara-
bidopsis leaves, the second phase being attributed to the
de-epoxidation of the xanthophyll cycle pool (Andersson
et al., 2001). Non-photochemical quenching up to a value of
around 2.5 was attained in wild-type plants. Most of the
NPQ relaxed rapidly in the dark, re¯ecting the DpH-depen-
dent feedback de-excitation of excess energy in the PS II
antenna (qE type of NPQ; Horton et al., 1996). Very similar
Table 1 Chlorophyll content (chl), chlorophyll a/bratio (chl a/b), PS II content (PS II/Chl), photosynthetic capacity (P
max
), F
v
/F
m
, feedback
de-excitation (qE) and state transitions of the wild type (WT), asLhcb2-1 (2-1) and asLhcb2-12 (2-12)
Line Chl (mgcm
2
) Chl a/bPS II/Chl P
max
F
v
/F
m
qE State transition
WT 29.06 1.19 3.31 0.02 3.27 0.21 20.34 1.43 0.825 0.007 1.87 0.10 0.94 0.02
2-1 23.44 0.84 3.67 0.03 4.10 0.43 18.17 0.72 0.819 0.003 1.61 0.06 0.56 0.11
2-12 19.34 0.60 4.29 0.04 5.16 0.21 19.80 1.84 0.806 0.003 1.40 0.12 0.02 0.04
Measurements were made on fully expanded leaves of plants grown at 100 mmol photons m
2
sec
1
. PS II was measured by oxygen flash
yield and expressed as mmol mol Chl
1
. Photosynthetic capacity (P
max
) was measured as the light-saturated rate of O
2
evolution in
saturating CO
2
and is expressed as mmol m
2
sec
1
. The reversible component (qE) of NPQ was measured as in Figure 4. State transitions
were measured as in Figure 5. Data are means SE (n7).
ßBlackwell Publishing Ltd, The Plant Journal, (2003), 35, 350±361
Arabidopsis lacking Lhcb1 and Lhcb2 353
behaviour was shown by the antisense plants, although the
maximum extent of NPQ was reduced to around 1.8 in
asLhcb2-12. NPQ was saturated at around the same light
intensity in wild-type and antisense plants (Figure 3b). The
reduction in NPQ was mostly because of a reduction in
amplitude of the qE component (feedback de-excitation),
from a value of around 1.9 in wild-type plants to 1.6 in
asLhcb2-1 and 1.4 in asLhcb2-12 (Table 1).
Both antisense lines had reduced values for the ¯uores-
cence parameter F
v
/F
m
, a measure of the intrinsic ef®ciency
of PS II (Table 1). This provided a possible explanation for
the reduced NPQ, as calculation of NPQ depends on the
value of F
m
(the maximum ¯uorescence yield) ± the
changes in F
v
/F
m
corresponded to reductions in F
m
of
3±4% for asLhcb2-1 and 10±13% for asLhcb2-12. However,
this only partially accounts for the reduced NPQ, and it was
therefore concluded that the absence of Lhcb1/Lhcb2 leads
to a partial inhibition of the formation of feedback de-
excitation.
This could arise if there was an alteration in the activity of
the xanthophyll cycle. Some changes in carotenoid com-
position of the antisense plants were detected. There was
no difference in the ratio of total carotenoid to chlorophyll,
but there was a reduction in lutein and particularly neox-
anthin in asLhcb2-12. These are the carotenoids that are
preferentially found in LHC II (Bassi et al., 1993; Ruban
et al., 1994). PS II core complexes are enriched in b-caro-
tene, explaining the increase in proportion of this carote-
noid. However, there was no signi®cant alteration in the
content of the xanthophyll cycle carotenoids (Table 2),
consistent with their distribution throughout the light-har-
vesting complexes of PS II (Ruban et al., 1999). Further-
more, in illuminated leaves, the de-epoxidation state of the
xanthophyll cycle pool was the same in wild-type and
antisense plants. This suggests that the change in NPQ
arises from an intrinsic decrease in feedback de-excitation
capacity rather an alteration in xanthophyll cycle activity.
Antisense plants could not perform state transitions
We have also investigated state transitions in wild-type
and antisense lines. This was done by measuring PS II
¯uorescence, exposing the plants to low-intensity red
light absorbed preferentially by PS II (light 2), supplemen-
ted with far-red light that is only absorbed by PS I (light 1).
For plants in State I (reached by prolonged illumination
with light 1 plus light 2), when the light 1 turned off, there
is an initial increase in PS II ¯uorescence (indicative of
reduction of the acceptor side of PS II). On the time scale of
seconds to minutes, the ¯uorescence then decreases
because of oxidation of PS II, which occurs when abso-
rbed light is re-distributed as a result of LHC II phosphor-
ylation and its migration away from PS II and association
with PS I; the system is now in State II. If light 1 is
Figure 3. Non-photochemical quenching (NPQ).
(a) Kinetics of the formation and relaxation of NPQ. Fluorescence was
monitored on leaf discs (wild type, black symbols; asLhcb2-12, white sym-
bols) during two successive periods of illumination with strong light
(1000 mmol photons m
2
sec
1
), indicated by the white bars, with a
17-min period of darkness in between, indicated by the black bars.
(b) Extent of feedback de-excitation (labelled qE) determined at a series of
irradiances as the difference between the values for total NPQ during
illumination and following 15 min dark relaxation. Data are means SE
(n3).
Figure 2. Photosynthetic CO
2
uptake.
The rate of CO
2
uptake in wild-type (black symbols) and antisense plants
(white symbols) was determined as a function of incident irradiance (PPFD).
Each data point is the mean SE of equivalent leaf disks from four different
plants. Inset: data for low PPFD with linear regression to show apparent
quantum yield.
ßBlackwell Publishing Ltd, The Plant Journal, (2003), 35, 350±361
354 Jenny Andersson et al.
re-applied, there is a rapid fall in ¯uorescence caused by
oxidation of PS II, followed by a slow rise in ¯uorescence
as State I is restored. In the antisense plants, the removal
of light 1 induced negligible reduction of PS II, and no
¯uorescence decrease was observed after the initial
increase (Figure 4).
State transitions can be quanti®ed by measuring the ratio
of the ¯uorescence values ((dc)(ab))/(dc) (see
Figure 4), which expresses the differences in the effect of
removing light 1 on the redox state of PS II in State I and
State II. A theoretical maximum State Transition has a value
of 1, and no State Transition gives a value of 0. Wild-type
plants gave values of around 0.9, decreasing to 0.5 in
asLhcb2-1 and 0.02 in asLhcb2-12 (Table 1). Thus, a reduc-
tion in the levels of Lhcb1/Lhcb2 leads to a reduction in the
extent of state transitions. The reduced effect on ¯uores-
cence of removing light 1 (Figure 4) also indicates that PS II
is not overexcited by light 2 in the asLhcb2-12 plants
(compare the value for dcin wild type and asLhcb2-12
plants) ± the low amount of Lhcb1/Lhcb2 means that there
is no imbalance between PS II and PS I under these
conditions.
Thylakoid stacking is retained in antisense plants
The effects of an LHC II de®ciency on thylakoid ultrastruc-
ture was studied by electron microscopy of transverse
sections of leaves from wild-type and asLhcb2-12 plants.
It was found that asLhcb2-12 exhibited granal stacking of
thylakoid membranes to an extent similar to that of the wild
type, and no marked difference was identi®ed in the dimen-
sions of individual stacks (Figure 5).
Antisense plants have reduced fitness under
natural conditions
Under controlled, medium-light conditions in the labora-
tory, the growth rates of wild-type plants and the two
antisense lines were similar, in terms of both leaf area
(Figure 6a) and fresh weight (Figure 6b). We wished to
investigate if the antisense plants suffered from the lack of
Lhcb1/Lhcb2 under the highly variable conditions encoun-
tered in the ®eld. We have recently developed a method to
directly determine plant ®tness under ®eld conditions by
measuring seed production in Arabidopsis (Ku
Èlheim
et al., 2002). Under two consecutive years (2001 and
2002), plantlets of the wild type and asLhcb2-12 were
transferred to the ®eld just after germination, and were
left without watering and fertilisation. The experimental
sites differed between the years but consisted each year of
two plots, one more shaded by other vegetation than the
other. The asLhcb2-12 plants grew more slowly than the
wild type (data not shown) and also showed decreased
seed production in both years and in both plots (Figure 7).
The number of seeds/silique was fairly constant in all
experiments (Figure 7c), and seed weight did not differ
between either genotypes or growth conditions (data not
shown).
Table 2 Carotenoid composition of the wild type (WT) and asLhcb2-12
Lines Car/chl Lut Neo b-car XC DES
WT 0.268 0.004 46.17 0.95 11.95 0.20 30.69 0.59 11.20 0.47 32.63 1.10
asLhcb2-12 0.276 0.005 41.24 0.88 7.68 0.22 37.47 0.80 13.61 0.73 31.75 1.90
Measurements were made on fully expanded leaves of plants grown at 100 mmol photons m
2
sec
1
. Data were obtained both from dark-
adapted and illuminated leaves. Contents of lutein (Lut), Neoxanthin (Neo), b-carotene (b-car) and the xanthophyll cycle (XC, the sum of
violaxanthin, antheraxanthin and zeaxanthin) are expressed as a percentage of total carotenoid. The de-epoxidation state of the
xanthophyll cycle pool, (zeaxanthin 1/2antheraxanthin)/XC100, was measured for leaves exposed to 1000 mmol quanta m
2
sec
1
illuminated for 23 min as in Figure 3. Data are means SE (n6).
Figure 4. State transitions.
State transitions were measured for wild-type and asLhcb2-12 plants by
assaying chlorophyll ¯uorescence during illumination with light 2 in the
presence and absence of light 1 (L1). Shown are two successive inductions
of a State 1±State 2 transition, with an intervening period in L1 to induce a
State 2±State 1 transition.
ßBlackwell Publishing Ltd, The Plant Journal, (2003), 35, 350±361
Arabidopsis lacking Lhcb1 and Lhcb2 355
The average number of seeds per plant varied, not sur-
prisingly, greatly for the wild type between the experiments
(Figure 7a). This is because of differences in weather
between the seasons (warmer and dryer in 2002; all weather
parameters for the experimental site were logged and are
available upon request), the date for the start of the experi-
ments (later start in 2001 than in 2002) and herbivory
(mainly by the diamond-backed moth Plutella xylostella
that was most abundant in the shaded sites). Large differ-
ences in seed production in each genotype between years
was also observed by Ku
Èlheim et al. (2002); hence, the
number of seeds should not be compared between the
four experiments. Compared to the wild type, the number
of seeds per plant was always decreased for asLhcb2-12 in
both conditions and in both years, showing that the ®tness
of the antisense plants was reduced, although to various
extents, in all studied conditions. This was mainly caused
by a reduction in the number of siliques/plants. Apparently,
Lhcb1/Lhcb2 affects ®tness of Arabidopsis plants both
under shaded and unshaded natural conditions, with or
without signi®cant grazing.
Figure 5. Chloroplast ultrastructure.
Electron micrographs were obtained from wild-type (a) and asLhcb2-12 (b)
plants.
Figure 7. Fitness of wild-type and asLhcb2-12 plants in the ®eld.
Wild-type (WT) and asLhcb2-12 plants were grown in the ®eld at two sites,
one in open sun and the other occasionally shaded during two years (2001
and 2002). Data are means SE.
(a) The number of seeds per plant, calculated from the number of siliques
per plant and seeds per silique, n24.
(b) The number of siliques per plant, n24.
(c) The number of seeds per silique, n32.
Figure 6. Growth and morphology of wild-type and antisense plants.
(a) Plants after growth in 8-h photoperiod (150 mmol photons m
2
sec
1
)for
8 weeks.
(b) The FW of the biomass above soil for wild-type (WT) plants, asLhcb2-1
(2-1) and asLhcb2-12 (2-12) grown as in (a). Mean SE, n10.
ßBlackwell Publishing Ltd, The Plant Journal, (2003), 35, 350±361
356 Jenny Andersson et al.
Discussion
Antisense plants lacking the major LHC II complexes
Plants lacking LHC II have previously been generated by
mutations abolishing chlorophyll bsynthesis or by growing
plants in intermittent light regimes (for example, Andrews
et al., 1995; Bassi et al., 1985; Bossmann et al., 1997;
Habash et al., 1994; Ha
Èrtel and Lokstein, 1995; Murray
and Kohorn, 1991). These have provided indications as to
the importance or otherwise of Lhcb1 and Lhcb2 complexes
in state transitions, granal stacking, protective energy dis-
sipation and plant growth. However, the interpretation of
such studies has been complicated by parallel loss of other
LHC components, including the minor LHC complexes and
LHC I, and/or changes in their pigment composition. For a
more clear-cut determination of the roles of the major LHC II
complexes, it is desirable to analyse plants with a speci®c
de®ciency in these abundant proteins. In this work, we have
been successful in obtaining plants lacking detectable
levels of Lhcb1 and Lhcb2.
As the Lhcb1 and Lhcb2 proteins are encoded by a multi-
copy gene family, we have adopted an antisense approach
so as to facilitate the use of a single transgene in eliminating
the expression of multiple genes. The introduction of an
antisense copy of Lhcb2.1 into Arabidopsis eliminated
expression from the ®ve Lhcb1 and three Lhcb2 genes ±
there was no effect on the Lhcb3 mRNA, which lacked the
22-bp stretches of sequence identity to the antisense trans-
gene believed to be necessary for RNAi suppression of
gene expression (Hutvagner and Zamore, 2002). Previous
attempts to abolish expression of LHC II in tobacco were
also successful in markedly reducing mRNA levels for both
Lhcb1 and Lhcb2, but with little effect on protein levels
(Flachmann and Ku
Èhlbrandt, 1995), leading to the sugges-
tion that there is post-transcriptional regulation of LHC II
levels. It is interesting to note that we also identi®ed an
antisense line, which retained detectable quantities of
Lhcb1 and Lhcb2 proteins despite having negligible levels
of mRNA (data not shown). It thus appears that, in Arabi-
dopsis, as in tobacco, signi®cant levels of Lhcb1 and Lhcb2
protein synthesis can be supported by very low message
levels, and that transcription and/or mRNA levels do not
directly regulate LHC II levels.
The antisense Arabidopsis plants produced in this work
have speci®cally lost the two protein subunits constituting
basically all the major LHC II trimers, with only small
parallel reductions in the relative levels of Lhcb3 and PsbS.
Plants of the same age showed no compensating change in
the levels of PS II per unit leaf area, instead compensating to
some degree for the loss of Lhcb1 and Lhcb2 by increasing
the amounts of just two other antenna proteins ± CP26 and
the PSI antenna protein Lhca4. These increases were repro-
ducibly found for multiple sets of plants grown in a variety
of conditions. It seems likely that this adjustment resulted
from increased synthesis of the proteins, although we
cannot rule out the possibility of slower protein turnover
caused by the decreased amount of LHC II. In either case,
the increase is consistent with the low-light acclimation
response of wild-type plants ± CP26 and Lhca4 levels are
signi®cantly increased by the growth of wild-type Arabi-
dopsis in low light (Bailey et al., 2001) ± which might be
activated in the antisense plants via detection of the de®cit
in antenna size by a regulatory signal (e.g. reduced PS II
excitation pressure: Huner et al., 1998). As PsbS levels
decreased in LL-grown wild-type plants, it is tempting to
speculate that the increase in CP26 and Lhca4 and the
reduction in PsbS might also re¯ect perception by the plant
of such a `low light' signal; this is now the subject of further
detailed studies. We have recently found that the additional
CP26 in the antisense line form trimers, probably together
with Lhcb3, that substitute for LHC II trimers in the PS II
supercomplex (Ruban et al., 2003).
LHC II is not essential for granal stacking
Thylakoid ultrastructure (granal stacking) was preserved
despite the absence of the Lhcb1 and Lhcb2 proteins that
constitute the major fraction of LHC II. Although there have
been reports that chlorophyll b-less plants that lack most
outer LHC sometimes could form membrane stacks (Bassi
et al., 1985, Murray and Kohorn, 1991), interpretation of the
ultrastructural work performed with different mutants and
plants grown under different light regimes (resulting in
different LHC polypeptide composition) is far from easy,
and many researchers have questioned these results and
assert that attractive forces between LHC II trimers cause
membrane adhesion (reviewed by Allen and Forsberg,
2001). Prior to this work, plants speci®cally lacking LHC II
have not been available for analysis. We believe that our
®ndings conclusively show that Lhcb1 and Lhcb2 are not
essential for granal stacking. However, this does not neces-
sarily mean that LHC II does not play a role in membrane
appression in wild-type plants. Rather, stacking could be a
property of the different surface charges on a number of
thylakoid proteins (including LHC II) that determines lateral
segregation and stacking (Barber, 1980). In this respect, it is
possible that the compensatory increase in other PS II-
associated LHC proteins in the antisense plants could con-
tribute to the stabilisation of the granal stacks.
Photosynthesis in antisense plants
The loss of LHC II in the antisense lines resulted in a
reduction in leaf chlorophyll by up to 35%, and a resulting
10% decrease in light absorption. Accordingly, in the anti-
sense lines, there was a small but signi®cant reduction in
the quantum yield of photosynthesis (expressed on an
ßBlackwell Publishing Ltd, The Plant Journal, (2003), 35, 350±361
Arabidopsis lacking Lhcb1 and Lhcb2 357
incident light basis), to an extent similar to that for the
reductionin light absorption, which cantherefore be ascribed
solely to the reduction in the proportion of light that was
captured. Conversely, the maximum rate of photosynthesis
per leaf area, determined as both O
2
evolution and CO
2
uptake, revealed no measurable difference between anti-
sense lines and wild-type plants. This is readily explained
because light capture does not limit photosynthesis under
saturating light, and strongly indicates that changes in the
antisense plants were restricted to the light-harvesting sys-
tem with no signi®cant effects on electron transport or car-
bon ®xation. Indeed, although antisense plants had a higher
P
max
when expressed per unit chlorophyll or per unit
absorbedlight (as a result of the reductions in leafchlorophyll
and light absorption), it was unaltered if expressed per PS II,
indicating that maximum electron transport rates were iden-
tical in wild-type and antisense plants.
Regulation of light harvesting is impaired in the
antisense plants
Here, we have unequivocally demonstrated that LHC II
(Lhcb1/Lhcb2) is essential for state transitions, and that
the other LHC proteins that accumulated as a response to
their depletion cannot substitute for this function of Lhcb1
and Lhcb2. Previous work has shown that overall depletion
of LHC proteins as a result of a de®ciency of chlorophyll b
synthesis results in an inhibition of state transitions
(Andrews et al., 1993; Chow et al., 1981). Only Lhcb1 and
Lhcb2, and not other Lhcb or Lhca proteins, are known to be
substrates for the LHC II protein kinase, which provides a
clear explanation for the State Transition defect. Although
the absence of state transitions might partially be explained
by the experimental dif®culty in producing a state of excess
excitation of PS II over PS I, because of altered PS II antenna
characteristics in the antisense plants, an altered PS II redox
state was, nevertheless, achieved, to which the antisense
plants failed to respond.
The antisense plants also had a reduced capacity for
feedback de-excitation (qE). Similar reductions in NPQ have
been found in chlorophyll b-less mutants (Ha
Èrtel and Lok-
stein, 1995). As there was no impairment of the maximum
photosynthetic rate (and therefore in the generation of a
DpH) and xanthophyll cycle activity was unchanged, the
decrease in maximum capacity of qE may result from a
decrease in the number of quenching sites, or a decrease in
their quenching ef®ciency. The latter may arise simply from
a decrease in connectivity within the PS II antenna in the
absence of Lhcb1 and Lhcb2, while the former could be
taken as evidence of a direct involvement of LHC II in
feedback de-excitation. It was found that the level of PsbS
was also decreased in the antisense plants, and we believe
that this explains the lower qE because PsbS has an essen-
tial role in this mechanism (Li et al., 2000).
LHC II is important under natural conditions
The ®nding that there were only minor effects on photo-
synthesis in antisense plants raises important questions
concerning the role of the LHC II complexes in light harvest-
ing. Despite a dramatic reduction in PS II antenna size, there
was only a small reduction in quantum yield and no impair-
ment of maximum photosynthesis, and under controlled
growth conditions, the antisense plants grew as well as
wild-type plants. Previous work on pigment-de®cient plants
has also shown that minimal differences in growth and
photosynthesis were caused by loss of LHC proteins
(Andrews et al., 1995; Habash et al., 1994).
In laboratory experiments, plants are routinely given
near-optimal conditions in terms of temperature, nutrient
and water availability, light intensity and day-length, and
also in terms of the stability of these conditions. Many
experiments depend on the homogeneity of the plant
material that is given by these growth methods, but an
important aspect of plant development is missing, namely
the remarkable ¯exibility induced by the high variability of
the natural environment that can occur over timescales
from seconds to season. In order to assess the importance
of a gene product for overall ®tness, the plant must be
challenged with the unpredictable ¯uctuations in growth
conditions that are found in nature. When asLhcb2-12
plants were grown in ®eld conditions, an effect on ®tness
was evident. It should be noted that wild-type plants under
®eld conditions have a very small amount of LHC II trimers;
the chl a/bratio was 4.1±4.3 (depending on plant age), in
other words as high as that of the antisense plants in
climate chamber (which lacks Lhcb1/Lhcb2) conditions.
Apparently, even the small complement of LHC II trimers
present in wild-type plants under high light conditions is
highly important; the increase in CP26 and Lhca4 could not
compensate for the loss in the antisense plants. It is thus
clear that an intact antenna system provides a marked
adaptive advantage to the plant in a variable environment.
There are a number of possible explanations for the ®tness
reduction inthe ®eld. First, plant productivity may be affected
by a reduction in light harvesting in situations where the
incident light intensity is low (e.g. twilight, cloudiness). Sec-
ond, the reduced capacity for feedback de-excitation may
also be responsible ± npq4 mutants lacking feedback de-
excitation suffer a signi®cant loss of ®tness (Ku
Èlheim et al.,
2002) ± and indeed, this may account for the reduced ®tness
of antisense plants in an open-sun environment. Third, a
reduction in antenna size under low-light conditions may
make PS II vulnerable to low-light photoinhibition caused by
an increased probability of charge recombination, and
®nally, the reduction in ®tness may result from the plants'
inability to perform state transitions. As green leaves trans-
mit preferentially light exciting PS I, occasional shading
results in ¯uctuations not only in light quantity but also in
ßBlackwell Publishing Ltd, The Plant Journal, (2003), 35, 350±361
358 Jenny Andersson et al.
light quality. The effects of partial and ¯uctuating shade are
not present in the growth chamber, where there is no wind
to move the leaves, or movement of the light source to mimic
the sun's moves across the sky. Thus, state transitions
would be most needed under partially shaded conditions
in the ®eld, but not at all under constant light quality and
quantity (as in the growth chamber). We can, at present, not
distinguish between these possibilities but hope to address
them in future experiments.
To conclude, we have demonstrated that grana stacking,
light-saturated photosynthesis and growth in optimal con-
ditions were not affected by the loss of Lhcb1 and Lhcb2.
However, the reduction of LHC II had a detrimental effect on
state transitions, reduced light-limited photosynthesis and
the capacity for feedback de-excitation, and imposed a
decrease in ®tness in the ®eld. Taken together, this demon-
strates that in natural conditions Lhcb1/Lhcb2 is necessary
for maximum performance.
Experimental procedures
Plant material and growth conditions
Arabidopsis thaliana, ecotype Columbia (Col-0) and antisense lines
derived from it were grown in climate chambers in an 8-h photo-
period with an irradiance of 150 mmol photons m
2
sec
1
and day/
night temperature of 23/188C.
Antisense constructs and Arabidopsis transformation
Antisense construct was made in the plant expression vector
pSJ10 (Ganeteg et al., 2001). Arabidopsis expressed sequence
tag (EST) clone 31F8T7, harbouring a full-length Lhcb2.1 cDNA,
was ampli®ed by PCR with primers 5
0
-TAATACGACTCACTA-
TAGGGA-3
0
and 5
0
-TCGCGAATTCGCGTACGTAAGCTTGGATCC-
3
0
, digested with EcoRI, and inserted into EcoRI-digested pSJ10.
The correct orientation of the inserts was co n®rmed by sequencing
of the vector/insert junction. A. thaliana cv. Columbia was used as
the background for transformation. The in planta vacuum in®ltra-
tion method was performed according to Ganeteg et al. (2001).
Seeds from transformed plants were selected on MS plates con-
taining 50 mgml
1
kanamycin. Kanamycin-resistant plants were
collected, screened by immunoblotting to verify the ef®ciency of
the antisense inhibition, and allowed to self-pollinate. Kanamycin-
resistant individuals from the T
2
and T
3
generations were used for
the experiments described below.
Seed stocks of asLhcb2-1 and asLhcb2-12 are deposited at the
Arabidopsis Biological Resource Centre (ABRC), accession num-
bers CS6363 (asLhcb2-12) and CS6362 (asLhcb2-1).
Clones and sequence analysis
Genomic sequences were identi®ed in the TAIR databa se using the
BLAST
tool with EST sequence information. The EST sequences
that were used were the same as in Jansson (1999). Accession
numbers (TAIR): Lhcb1.1 (At1g29920); Lhcb1.2 (At1g29910);
Lhcb1.3 (At1g29930); Lhcb1.4 (At2g34430); Lhcb1.5 (At2g34420);
Lhcb2.1 (At2g05100); Lhcb2.2 (At2g05070); Lhcb2.3 (At3g27690);
Lhcb3 (At5g54270).
Immunoblot analysis
Immunoblotting was performed essentially according to Jansson
et al. (1997). Fully expanded leaves were used for thylakoid pre-
paration, and sample corresponding to 3 mg chlorophyll was run in
each lane. Antibodies were the same as those described by
Andersson et al. (2001) and Ganeteg et al. (2001). Antibodies
speci®c for PsbS and CP43 were kindly provided by Kris Niyogi
(University of California, Berkeley) and Roberto Barbato (Univer-
sity of Padua, Italy), respectively.
RNA analysis
RNA preparations and RNA gel blotting were performed as
described by Ganeteg et al. (2001), using leaf material (fully
expanded leaves) sampled 3 h into the light period. Blots were
hybridised with
32
P-labelled PCR fragments of EST clones
138O13T7 (Lhcb1.3) or 31F8T7 (Lhcb2.1) as described by Andersson
et al. (2001). For RT-PCR, RNA samples were DNase-treated for
10 min and diluted to 1.25 ng ml
-1
. In the process of optimising
the protocol, several dilutions were used to ensure that the
chosen amount of template RNA generated products in the linear
range of the method. RT-PCR was performed with Clonetech Tita-
nium One-Step RT-PCR Kit, according to the manufacturers instruc-
tions, except that the reaction volume was scaled down to 50%,
using 5
0
-CAGCAAAACAAGAAGAACAACAACACTAA-3
0
in combina-
tion with 5
0
-GAGGAACAAGAAAGGTTATGTTCATATCA-3
0
(ampli®-
cation of Lhcb3 fragment) or 5
0
-GGAAGGATCTGTACGGTAAC-3
0
in
combination with 5
0
-TGTGAACGATTCCTGGACCT-3
0
(ampli®cation
of actin2 fragment) as primers amplifying 1.25 ng of temp- late RNA.
The temperature cycles for Lhcb3 ampli®cation were: 508C for 1 h;
948C for 5 min; 26 cycles of 948C for 30 sec, 608C for 30 sec, 688Cfor
90 sec; and ®nally 688C for 2 min. The temperature cycles for actin2
ampli®cation were: 508Cfor1 h;948C for 5 min; 26 cycles of 948Cfor
30 sec,558C for 30 sec and 688C for30 sec; and ®nally 688Cfor2 min.
The products were separated on 1 % agarose gels in TAE, supple-
mented with ethidium bromide, visualised with UV light and quan-
ti®ed with
CHEMIIMAGER 4000
software.
Electron microscopy
Leaf fragments, sampled during the day, were ®xed in 2% parafor-
maldehyde and 2.5% glutaraldehyde in 0.1
M
phosphate buffer for
3 h at 48C. They were then washed three times at 30-min intervals at
48Cin0.1
M
phosphate buffer containing 10% sucrose. Secondary
®xation was carried out in 2% aqueous osmium tetroxide for 1 h at
room temperature. Dehydration was performed through a graded
series of ethanol solutions at 208C. Samples were then washed two
times in propylene oxide. In®ltration was accomplished by placing
the specimens in a 50/50 mixture of propylene oxide and Spurrresin
overnight at room temperature, following which samples were
placed in full-strength Spurr resin for 6±8 h and embedded into
fresh resin for 8 h at 708C. Ultrathin 70±90-nm thick sections were
cut on a Reichert Ultracut E ultramicrotome (Reichert-Jung, Vienna,
Austria) and stained for 15 min with 3% uranyl acetate in 50%
ethanol followed by staining with Reinold's lead citrate for 2 min.
The sections were examined using a Philips CM10 transmission
electron microscope at an accelerating voltage of 80 kV.
Pigment analysis
The composition of carotenoids was determined as described
by Ruban et al. (1994), with the modi®cations described by
ßBlackwell Publishing Ltd, The Plant Journal, (2003), 35, 350±361
Arabidopsis lacking Lhcb1 and Lhcb2 359
Andersson et al. (2001). Chlorophyll content was assayed in 80%
acetone according to Porra et al. (1989).
Photosynthesis measurements
The kinetics of formation and relaxation of NPQ were obtained by
measurements of chlorophyll ¯uorescence using leaf discs
exactly as described previously by Andersson et al. (2001). State
transitions were measured on attached leaves in air using a
multibranch ®bre optic, which provided connection to the emitter
detector unit of the Walz PAM ¯uorometer (Walz, Effeltrich,
Germany), and to the actinic lights: Light 1 was de®ned by a
Schott RG 710 glass ®lter (Schott, Mainz, Germany) and a 707-nm
interference ®lter, the intensity being adjusted empirically to that
suf®cient to cause maximum oxidation of PS II and PS I; light 2
was de®ned by a Schott RG610 glass ®lter and a 660-nm inter-
ference ®lter, the intensity again being empirically set to the
maximum that would still allow complete PS II oxidation by
the addition of light 1, and this was found to be approximately
10±20 mmol photons m
2
sec
1
. The protocol was to give dark-
adapted leaves 10 min in light 1 plus light 2 to activate photo-
synthesis, followed by 20-min illumination with light 2 only to
induce State II. Light 1 was then switched on, and illumination
withlights1and2continuedforafurther20mintoinducea
return to State I. Finally, light 1 was turned off again to follow the
transition back to State II. This protocol ensures that all redox
changes arise from the state transition and are not complicated
by photosynthetic induction phenomena that can arise upon
switching between light 1 only and light 2 only. Oxygen evolution
under saturating CO
2
was measured using leaf disks as described
by Andersson et al. (2001). CO
2
assimilation (350 p.p.m. CO
2
)was
measured using a PP Systems CIRRAS IRGA (PP systems,
Hitchin, UK) with attached leaves as described by Walters et al.
(1999). The content of functional PS II was determined by the O
2
¯ash yield method (Chow et al.,1991).
Field experiment
The ®eld experiment was conducted essentially as described by
Ku
Èlheim et al. (2002). Two sites were chosen for the experiment.
One located in open sun, the other nearby but partially shaded.
The maximum light intensity was about 2000 mmol photo-
ns m
2
sec
1
in both sites, but on the shaded site, trees and
other vegetation occasionally (several hours a day) shaded the
experimental plants. In brief, ®ve seeds of wild-type and anti-
sense line were germinated in separate pots, all but one seedling
per pot was removed, and pots were randomised in trays (using a
block design) and grown outdoors in July±August 2001 and 2002,
Umea
Ê, Sweden without watering, fertilising or pest control, to
resemble natural conditions as closely as possible. Temperature,
light, wind, humidity and precipitation varied largely over the
experiment, but were logged at a site close to the experimental
site. All meteorological data is available on request. At the time of
bolting, the growth area was sealed into a large tent made of
insect net, in order to prevent spread of transgenic pollen.
Siliques were collected and counted and to determine the num-
ber of seeds per silique, three siliques from each individual
(fewer for those individuals that produced less than three sili-
ques) were opened, and the seed content was counted. There
was a signi®cant mortality among the plants (mainly because
of grazing), but mortality did not differ between the geno-
types, similar results were produced regardless of whether
plants not producing any seeds were included in the analysis
or not.
Acknowledgements
We are grateful to Jenny Jonsson-Lindvall for participation in the
initial part of this work, Lars Ericsson for provision of growth space
for the ®eld experiment and Carsten Ku
Èlheim, Daniel Eriksson and
Martin Frenkel for participation in the ®eld experiments. This work
was supported by grants to S.J. from the Swedish Research
Council for Environment, Agricultural Sciences and Spatial Plan-
ning and the Foundation for Strategic Research and to P.H. from
the UK Biotechnology and Biological Sciences Research Council,
the European Union, Natural Environment Research Council and
the Joint Infrastructure Fund.
References
Allen, J.F. and Forsberg, J. (2001) Molecular recognition in thyla-
koid structure and function. Trends Plant Sci. 6, 317±326.
Anderson, J.M. (1986) Photoregulation of the composition, func-
tion, and structure of thylakoid membranes. Annu. Rev. Plant
Physiol. Plant Mol. Biol. 37, 93±136.
Andersson, J., Walters, R.G., Horton, P. and Jansson, S. (2001)
Antisense inhibition of the photosynthetic antenna proteins
CP29 and CP26: implications for the mechanism of protective
energy dissipation. Plant Cell,13, 1193±1204.
Andrews, J.R., Bredenkamp, G.J. and Baker, N.R. (1993) Evalua-
tion of the role of state transitions in determining the ef®ciency
of light utilization for CO
2
assimilation in leaves. Photosynth.
Res. 38, 15±26.
Andrews, J.R., Fryer, M.J. and Baker, N.R. (1995) Consequences of
LHCII de®ciency for photosynthetic regulation in Chlorina
mutants of barley. Photosynth. Res. 44, 81±91.
The Arabidopsis Genome Initiative (2000) Analysis of the genome
sequence of the ¯owering plant Arabidopsis thaliana.Nature,
408, 796±815.
Bailey, S., Walters, R.G., Jansson, S. and Horton, P. (2001) Accli-
mation of Arabidopsis thaliana to the light environment: the
existence of separate low light and high light responses. Planta,
213, 794±801.
Barber, J. (1980) Membrane surface charges and potentials
in relation to photosynthesis. Biochim. Biophys. Acta,594,
253±308.
Barber, J., Morris, N.E.P., Zheleva, D. and Hankamer, B. (1997) The
structure, function and dynamics of photosystem two. Physiol.
Plant. 100, 817±827.
Bassi, R., Hinz, U. and Barbato, R. (1985) The role of the light-
harvesting complex and photosystem II in thylakoid stacking
in the chlorina-f2 barley mutant. Carlsberg Res. Commun. 50,
347±367.
Bassi, R., Pineau, B., Dainese, P. and Marquardt, J. (1993) Carote-
noid binding proteins of photosystem II. Eur. J. Biochem. 212,
297±303.
Bernstein, E., Caudy, A.A., Hammond, S.M. and Hannon, G.J.
(2001) Role for a bidentate ribonuclease in the initiation step
of RNA interference. Nature,409, 363±366.
Boekema, E.J., van Breemen, J.F.L., van Roon, H. and
Dekker, J.P. (2000) Arrangement of photosystem II supercom-
plexes in crystalline macrodomains within the thylakoid
membrane of green plant chloroplasts. J. Mol. Biol. 301,
1123±1133.
Bossmann, B., Knoetzel, J. and Jansson, S. (1997) Screening
of chlorina mutants of barley (Hordeum vulgare L.) with anti-
bodies against light-harvesting proteins of PS I and PS II:
absence of speci®c antenna proteins. Photosynth. Res. 52,
127±136.
ßBlackwell Publishing Ltd, The Plant Journal, (2003), 35, 350±361
360 Jenny Andersson et al.
Chow, W.S., Hope, A.B. and Anderson, J.M. (1991) Further studies
on quantifying photosystem II in vivo by ¯ash-induced oxygen
yield from leaf-disks. Aust. J. Plant Physiol. 18, 397±410.
Chow, W.S., Telfer, A., Chapman, D.J. and Barber, J. (1981) State
1±State 2 transition in leaves and its association with ATP-
induced chlorophyll ¯uorescence quenching. Biochim. Biophys.
Acta,638, 60±68.
Elrad, D., Niyogi, K.K. and Grossman, A.R. (2002) A major light-
harvesting polypeptide of photosystem II functions in thermal
dissipation. Plant Cell,14, 1801±1816.
Flachmann, R. and Ku
Èhlbrandt, W. (1995) Accumulation of plant
antenna complexes is regulated by posttranscriptional mechan-
isms in tobacco. Plant Cell,7, 149±160.
Ganeteg, U., Strand, A., Gustafsson, P. and Jansson, S. (2001) The
properties of the chlorophyll a/b-binding proteins Lhca2 and
Lhca3 studied in vivo using antisense inhibition. Plant Physiol.
127, 150±158.
Gilmore, A.M., Hazlett, T.L., Debrunner, P.G. and Govindjee (1996)
Photosystem II chlorophyll a ¯uorescence lifetimes and in tensity
are independent of the antenna size differences between barley
wild-type and chlorina mutants: photochemical quenching and
xanthophyll cycle-dependent nonphotochemical quenching of
¯uorescence. Photosynth. Res. 48, 171±187.
Habash, D.Z., Genty, B. and Baker, N.R. (1994) The consequences
of chlorophyll de®ciency for photosynthetic light use ef®ciency
in a single nuclear gene mutation of Cowpea. Photosynth. Res.
42, 17±25.
HaÈ rtel, H. and Lokstein, H. (1995) Relationship between quenching
of maximum and dark-level chlorophyll ¯uorescence in vivo ±
dependence on photosystem II antenna size. Biochim. Biophys.
Acta,1128, 91±94.
Horton, P. Ruban, A.V. Rees, D. Pascal, A.A. Noctor, G. and Young,
A.J. (1991) Control of the light-harvesting function of chloroplast
membranes by aggregation of the LHC II chlorophyll protein
complex. FEBS Lett.,292, 1±4.
Horton, P., Ruban, A.V. and Walters, R.G. (1996) Regulation of light
harvesting in green plants. Annu. Rev. Plant Physiol. Plant Mol.
Biol.,47, 655±684.
Huner, N.P.A., O
Èquist, G. and Sarhan, F. (1998) Energy balance and
acclimation to light and cold. Trends Plant Sci. 3, 224±230.
Hutvagner, G. and Zamore, P.D. (2002) RNAi: nature abhors a
double-strand. Curr. Opin. Genet. Dev. 12, 225±232.
Jansson, S. (1999) A guide to the Lhc genes and their relatives in
Arabidopsis.Trends Plant Sci. 4, 236±240.
Jansson, S., Stefansson, H., NystroÈ m, U., Gustafsson, P. and
Albertsson, P.A. (1997) Antenna protein composition of PS I
and PS II in thylakoid sub-domains. Biochim. Biophys. Acta,
1320, 297±309.
Kro
Âl, M., Spangfort, M.D., Huner, N.P.A., O
Èquist, G., Gustafsson, P.
and Jansson, S. (1995) Chlorophyll a/b-binding proteins, pig-
ment conversions, and early light-induced proteins in a chlor-
ophyll b-less barley mutant. Plant Physiol. 107, 873±883.
Ku
Èhlbrandt, W., Wang, D.N. and Fujiyoshi, Y. (1994) Atomic model
of plant light-harvesting complex by electron crystallography.
Nature,367, 614±621.
Ku
Èlheim, C. A
Êgren, J. and Jansson, S. (2002) Rapid regulation
of light harvesting and plant ®tness in the ®eld. Science,297,
91±93.
Legen, J., Misera, S., Herrmann, R.G. and Meurer, J. (2001) Map
positions of 69 Arabidopsis thaliana genes of all known nuclear
encoded constituent polypeptides and various regulatory fac-
tors of the photosynthetic membrane: a case study. DNA Res. 8,
53±60.
Li, X.P., Bjorkman, O., Shih, C., Grossman, A.R., Rosenquist, M.,
Jansson, S. and Niyogi, K.K. (2000) A pigment-binding protein
essential for regulation of photosynthetic light harvesting. Nat-
ure,403, 391±395.
Lunde, C., Jensen, P.E., Haldrup, A., Knoetzel, J. and Scheller, H.V.
(2000) The PSI-H subunit of photosystem I is essential for state
transitions in plant photosynthesis. Nature,408, 613±615.
MaÈ enpaÈaÈ , P. and Andersson, B. (1989) Photosystem II heteroge-
neity and long-term acclimation of light-harvesting. Z. Natur-
forsch. 44C, 403±406.
Murray, D.L. and Kohorn, B.D. (1991) Chloroplasts of Arabidopsis
thaliana homozygous for the Ch-1 locus lack chlorophyll b, lack
stable LHCPII and have stacked thylakoids. Plant Mol. Biol. 16,
71±79.
Porra, R.J., Thompson, W.A. and Kriedemann, P.E. (1989) Deter-
mination of accurate extinction coef®cients and simultaneous
equations for assaying chlorophyll aand chlorophyll bextracted
with 4 different solvents ± veri®cation of the concentration of
chlorophyll standards by atomic-absorption spectroscopy. Bio-
chim. Biophys. Acta,975, 384±394.
Ruban, A.V., Young, A.J., Pascal, A.A. and Horton, P. (1994) The
effects of illumination on the xanthophyll composition of the
photosystem II light-harvesting complexes of Spinach thylakoid
membranes. Plant Physiol. 104, 227±234.
Ruban, A.V., Lee, P.J., Wentworth, M., Young, A.J. and Horton, P.
(1999) Determination of the stoichiometry and strength of bind-
ing of xanthophylls to the photosystem II light harvesting com-
plexes. J. Biol. Chem. 274, 10458±10465.
Ruban, A.V., Wentworth, M., Yakushevska, A.E., Andersson, J.,
Lee, P.J., Keegstra, W., Dekker, J.P., Boekema, E.J., Jansson, S.
and Horton, P. (2003) Plants lacking the main light-harvesting
complex retain photosystem II macro-organization. Nature,421,
648±652.
Vener, A.V., VanKan, P.J.M., Rich, P.R., Ohad, I. and Andersson, B.
(1997) Plastoquinol at the quinol oxidation site of reduced
cytochrome bf mediates signal transduction between light
and protein phosphorylation: thylakoid protein kinase deactiva-
tion by a single-turnover ¯ash. Proc. Natl. Acad. Sci. USA,94,
1585±1590.
Walters, R.G., Rogers, J.J.M., Shephard, F. and Horton, P. (1999)
Acclimation of Arabidopsis thaliana to the light environment:
the role of photoreceptors. Planta,209, 517±527.
Yakushevska, A., Jensen, P., Keegstra, W., van Roon, H., Scheller,
V., Boekema, E. and Dekker, J. (2001) Supermolecular organiza-
tion of photosystem II and its associated light-harvesting
antenna in Arabidopsis thaliana.Eur. J. Biochem. 268,
6020±6028.
Accession numbers: asLhcb2-12 (seed stock) CS6363; asLhcb2-1 (seed stock) CS6362.
ßBlackwell Publishing Ltd, The Plant Journal, (2003), 35, 350±361
Arabidopsis lacking Lhcb1 and Lhcb2 361
