Content uploaded by Thomas Pfannschmidt
Author content
All content in this area was uploaded by Thomas Pfannschmidt on Apr 09, 2018
Content may be subject to copyright.
Available via license: CC BY 4.0
Content may be subject to copyright.
Retrograde Plastid Redox Signals in the Expression of Nuclear
Genes for Chloroplast Proteins of Arabidopsis thaliana*□
S
Received for publication, June 8, 2004, and in revised form, November 11, 2004
Published, JBC Papers in Press, November 23, 2004, DOI 10.1074/jbc.M406358200
Vidal Fey‡§, Raik Wagner‡§, Katharina Bra¨ utigam‡, Markus Wirtz¶,Ru¨ diger Hell¶,
Angela Dietzmann储, Dario Leister储, Ralf Oelmu¨ ller‡, and Thomas Pfannschmidt‡**
From the ‡Department for Plant Physiology, Friedrich-Schiller-University Jena, Dornburger Strasse 159, 07743 Jena, the
¶Heidelberg Institute of Plant Sciences, University of Heidelberg, Im Neuenheimer Feld 360, 69120 Heidelberg, and the
储Max-Planck-Institute for Plant Breeding Research, Carl-von-Linne´-Weg 10, 50829 Ko¨ln, Germany
Excitation imbalances between photosystem I and II
generate redox signals in the thylakoid membrane of
higher plants which induce acclimatory changes in the
structure of the photosynthetic apparatus. They affect
the accumulation of reaction center and light-harvest-
ing proteins as well as chlorophylls aand b. In Arabi-
dopsis thaliana the re-adjustment of photosystem stoi-
chiometry is mainly mediated by changes in the number
of photosystem I complexes, which are accompanied by
corresponding changes in transcripts for plastid reac-
tion center genes. Because chloroplast protein com-
plexes contain also many nuclear encoded components
we analyzed the impact of such photosynthetic redox
signals on nuclear genes. Light shift experiments com-
bined with application of the electron transport inhibi-
tor 3-(3ⴕ,4ⴕ-dichlorophenyl)-1,1ⴕ-dimethyl urea have been
performed to induce defined redox signals in the thyla-
koid membrane. Using DNA macroarrays we assessed
the impact of such redox signals on the expression of
nuclear genes for chloroplast proteins. In addition, stud-
ies on mutants with lesions in cytosolic photoreceptors
or in chloroplast-to-nucleus communication indicate
that the defective components in the mutants are not
essential for the perception and/or transduction of light-
induced redox signals. A stable redox state of glutathi-
one suggest that neither glutathione itself nor reactive
oxygen species are involved in the observed regulation
events pointing to the thylakoid membrane as the main
origin of the regulatory pathways. Our data indicate a
distinct role of photosynthetic redox signals in the cel-
lular network regulating plant gene expression. These
redox signals appear to act independently and/or above
of cytosolic photoreceptor or known chloroplast-to-nu-
cleus communication avenues.
The light environment of plants is highly variable. This is of
particular importance for photosynthesis, because changes in
incident light intensity or quality can reduce the efficiency of
photosynthetic electron transport and therefore the net energy
fixation. Plants have developed many acclimatory mechanisms
at the molecular level that enable them to cope with such
changes. Most prominent responses are dynamic changes in
the structure and composition of the photosynthetic apparatus
(1–3).
Light quality and quantity gradients that occur e.g. in dense
plant populations induce an imbalance in excitation energy
distribution between the two photosystems (which work elec-
trochemically in series) and therefore reduce photosynthetic
efficiency. To counteract such imbalances plants re-distribute
light energy in a short term by state transitions (4, 5) and in a
long term by a re-adjustment of photosystem stoichiometry.
This results in a supply of more light quanta to the less active
side of the electron transport chain (6 –8). Both processes are
regulated by light-induced changes in the redox state of pho-
tosynthetic components (9 –11). While the short term response
acts via post-translational phosphorylation of existing antenna
proteins, the long term response (LTR)
1
requires the synthesis
of new components and hence has to affect gene expression.
This implies signaling routes that connect photosynthetic elec-
tron transport/efficiency with the expression machinery. Stud-
ies in the last decade show that such functional connections
exist at multiple levels and in virtually all classes of photosyn-
thetic organisms. In higher plants photosynthetic redox control
has been found at the levels of transcription (12–19), transcript
stability (20 –23), ribosome loading (24–26), translation initia-
tion (27), and protein accumulation (28).
The origin of the respective signal transduction pathways
can be very different. To date three classes of redox signals can
be distinguished: the first one is generated directly within the
electron transport chain, the second is represented by photo-
synthesis-coupled redox-active compounds such as thioredoxin
or glutathione, and the third is constituted by reactive oxygen
species, which are unavoidable by-products of photosynthesis
(29 –31). Such signals operate within the chloroplast, but have
also been shown to affect the expression of some nuclear genes
for plastid proteins. Therefore, they may represent a new class
of the so-called “plastid signals” (32–35). Retrograde signaling
represents an important feedback control that couples the ex-
pression of nuclear encoded plastid proteins to the functional
state of the chloroplast. Underlying signaling mechanisms in
this communication still represent a great field of open ques-
tions in plant cell biology. To date neither the impact of retro-
* This work was supported by grants from the Deutsche Forschungs-
gemeinschaft (DFG) (to T. P.) and the DFG Research Group 387. The
costs of publication of this article were defrayed in part by the payment
of page charges. This article must therefore be hereby marked “adver-
tisement” in accordance with 18 U.S.C. Section 1734 solely to indicate
this fact.
□SThe on-line version of this article (available at http://www.jbc.org)
contains Supplemental Table SI.
§ Both authors contributed equally to this work.
** To whom correspondence should be addressed. Tel.: 3641-949-236;
Fax: 3641-949-232; E-mail: Thomas.Pfannschmidt@uni-jena.de.
1
The abbreviations used are: LTR, long term response; Chl, chloro-
phyll; DCMU, 3-(3⬘,4⬘-dichlorophenyl)-1,1⬘-dimethyl urea; PAM, pulse-
amplitude modulation; psaAB, chloroplast genes for PsaA and PsaB
reaction center proteins of photosystem I; psbA, chloroplast gene for
reaction center D1 protein of photosystem II; PSI, photosystem I; PSII,
photosystem II; PIPES, 1,4-piperazinediethanesulfonic acid; GST, glu-
tathione S-transferase; PEP, plastid-encoded RNA polymerase;
E,
mol photons per m
2
and s.
THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 280, No. 7, Issue of February 18, pp. 5318–5328, 2005
© 2005 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.
This paper is available on line at http://www.jbc.org5318
by guest on April 9, 2018http://www.jbc.org/Downloaded from by guest on April 9, 2018http://www.jbc.org/Downloaded from by guest on April 9, 2018http://www.jbc.org/Downloaded from
grade redox signals on the nuclear transcriptome of chloro-
plasts nor possible interaction with other retrograde signals or
with photoreceptor-mediated light signals are known while an
interaction with sugar signals has been reported (19).
In this study we characterize the role of plastid redox signals
in the regulation of plastid and nuclear genes during photosys-
tem stoichiometry adjustment in Arabidopsis thaliana.Bythe
use of this model organism we take advantage from the mutant
and array resources available for this organism offering exper-
imental strategies, which are not possible with tobacco and
mustard used in earlier studies (9, 14). We describe for the first
time the molecular response to PSI or PSII light in chloroplasts
of A. thaliana. Determinations of glutathione content and re-
dox state were performed to check possible interactions of
different redox signals in this event. Cross-talk of the LTR with
other signaling routes has been tested in mutants lacking
either photoreceptors or components of plastid-to-nucleus sig-
nal pathways. By using a macroarray approach we determined
the impact of plastid redox signals on the nuclear transcrip-
tome of chloroplasts. Our study indicates that chloroplast redox
signals from the thylakoid membrane represent a novel and
separate class of plastid signals.
EXPERIMENTAL PROCEDURES
Plant Growth—Plants were grown in temperature-controlled growth
chambers at 22 °C under continuous light. Arabidopsis seeds (var. Col
0orLandsberg erecta and mutant lines in the respective backgrounds)
were sown either sterile on half-strength Murashige and Skoog (MS)
medium containing 1.35% sucrose or on earth substituted with vermic-
ulite. Density of seeds was adjusted in such a way that 16-day-old
plants did not shadow each other. After 2 days at 4 °C plants were
grown for 10 days under white light provided by 30-watt white stripe
lamps (OSRAM, Mu¨ nchen, Germany) with a photosynthetic-active ra-
diation of ⬃35
E. This white light pre-treatment was found to be
necessary for the plants to develop a normal leaf anatomy and hence a
true acclimatory response. Direct germination and growth under the
PSI or PSII light sources resulted in aberrant leaf anatomy due to the
lack of blue radiation of these light sources. After growth in white light,
plants were acclimated to PSI (photosynthetic active radiation, ⬃20
E)
or PSII (photosynthetic active radiation, ⬃30
E) light for 6 days or
they were first acclimated to one light source for 2 days followed by 4
days under the respective other light source. PSI and PSII light sources
have been described earlier (9, 12); however, the incandescent bulbs of
the PSI light source were replaced by 18-watt fluorescent stripe lamps
“Red” (OSRAM, Mu¨ nchen, Germany) of the same photon flux density to
reduce thermal radiation. The photosynthetic active radiation was de-
termined by using the lightmeter LI-250 (Heinz Walz GmbH, Effeltrich,
Germany). It must be noted that the far-red spectrum of the PSI light
is outside of the detection range of the LI-250. White light control plants
were grown for 16 days under the white light source alone.
Chlorophyll Fluorescence Measurements—In vivo Chl afluorescence
parameters were determined at room temperature with a pulse ampli-
tude-modulated (PAM) fluorometer (PAM101/103, Heinz Walz). 10 –15
seedlings grown on MS medium were measured simultaneously as
described previously (14). After dark acclimation (8 –10 min) the meas-
uring beam was turned on, and minimal fluorescence (F
o
) was deter-
mined. Then leaves were exposed to a 500-ms flash of saturating white
light (6000
E) to determine maximal fluorescence (F
m
) and the opti-
mum quantum yield F
v
/F
m
value was calculated as F
m
⫺F
o
/F
m
(36).
Subsequently, leaves were illuminated with 100
mol of photons m
⫺2
s
⫺1
of actinic red light of 600 nm (Walz 102-R). Fluorescence was
recorded in the saturation pulse mode by application of saturating
flashes every 30 s to determine maximal fluorescence of illuminated
leaves (F
m
⬘) until a stable fluorescence level (F
t
) was reached. Actinic
light was switched off, and far-red light (Walz 102-FR) was turned on to
oxidize the electron transport chain and to determine minimal fluores-
cence (F
o
⬘) in the light-acclimated state. The steady-state fluorescence
F
s
was then calculated as F
t
⫺F
o
⬘⫽F
s
. The optimum quantum yield
describes the maximal photosynthetic capacity of a plant and was taken
as a measure for photosynthetic efficiency of the mutant lines analyzed
in this study in comparison to wild type. For wild type we found F
v
/F
m
values of 0.8 –0.83, which typically indicate that the plant analyzed has
no decreased photosynthetic efficiency. Only plants with a wild type
like behavior were tested for their response to the two light sources. A
proper acclimation response to PSI or PSII light is characterized by a
significant change in the F
s
/F
m
value as shown earlier (14) and reflects
the structural differences in the photosynthetic apparatus of these
plants. The difference of 10
E in photosynthetic active radiation be-
tween PSI and PSII light has no detectable impact on this acclimation,
because in control experiments PSI plants showed the same decrease in
F
s
/F
m
after acclimation to either 20 or 30
E PSII light (data not
shown). One-way analysis of variance was used to reveal significant
differences in F
s
/F
m
values of plants grown under the defined condi-
tions. Light treatment was used as a factor, and the F
s
/F
m
value as a
dependent variable. If a significant influence of light treatment was
determined, post-hoc tests (pairwise multiple comparison test for lowest
significance difference) was performed to find out which groups differ
from each other. p⬍0.5 determines significant differences between
various samples (see Supplementary Table SI). All tests were per-
formed using SPSS 11.5.
Chlorophyll Content Determination—Total chlorophyll was deter-
mined spectroscopically after grinding of leaves in liquid nitrogen and
extracting chlorophylls with 80% (v/v) buffered acetone. Concentrations
of chlorophylls aand bwere calculated by using the extinction coeffi-
cients from previous studies (37).
Western Analyses of Chloroplast Proteins—20 g of leaf material of
plants grown on soil were harvested under the respective light source
and directly homogenized in ice-cold buffer containing 0.05 MHepes/
KOH, pH 8.0, 0.33 Msorbitol, 0.001 MMgCl
2
, and 0.002 MEDTA. The
material was filtered through four layers of muslin and one layer of
Miracloth, followed by a centrifugation (10 s at 6000 rpm). The pellet
was washed twice in homogenization buffer and resuspended in 1 ml of
the same buffer. Concentrations of chloroplasts were determined mi-
croscopically by counting diluted aliquots in a Fuchs-Rosenthal cham-
ber. 2 ⫻10
5
plastids of each preparation were lysed and denatured in
5⫻SDS sample buffer (final concentrations: 0.4% SDS, 0.1%

-mercap-
toethanol, 2% glycerol, 0.02% bromphenol blue) by incubation for 5 min
at 95 °C. Insoluble particles were removed by centrifugation, and sam-
ples were loaded on denaturing 10% SDS-polyacrylamide gels (38) and
separated overnight at 45 V. Proteins were transferred to a nylon
membrane (Roti-Nylon Plus, Roth, Karlsruhe, Germany) at 400 mA for
1 h using a semi-dry blotting apparatus, and the membrane was satu-
rated in Tris-buffered saline containing 2% fat-free milk powder. Incu-
bation with polyclonal antisera followed standard protocols (39). Anti-
bodies for D1, Lhca3, and Lhcb1 were purchased from AgriSera
(Vannas, Sweden). Detection of the first antibody was performed with a
goat-anti-rabbit-IgG-peroxidase conjugate (Sigma, Mu¨ nchen, Germany)
and the enhanced chemiluminescent (ECL) detection system. For visu-
alization of marker proteins and to prove blotting efficiency, mem-
branes were stained with Amido Black (39) after the ECL reaction.
RNA Preparations—RNA for primer extension analyses was isolated
from plants grown on MS medium. RNA for array analyses was isolated
from wild-type (Col-0) plants grown on soil. Leaf material was har-
vested and frozen in liquid N
2
under the respective light source. Total
RNA was isolated using the TRIzol reagent (Invitrogen) following a
protocol described earlier (40). Concentration and purity of RNA sam-
ples were determined spectroscopically in a Biophotometer (Eppendorf,
Hamburg, Germany). Intactness was proven by ethidium bromide
staining of rRNA species after electrophoretic separation of aliquots on
denaturing 1.2% agarose gels containing formaldehyde (39). Isolated
RNA was stored at ⫺80 °C until further use.
Primer Extension Analyses—Primer extension analyses (41) were
carried out according to a protocol from Li-Cor (Bad Homburg, Ger-
many). 5
g of total RNA was resolved in 20
l of hybridization mixture
containing 1.25
Minfrared dye 700-labeled psaA-specific and infrared
dye 800-labeled psbA-specific primers and 18
l hybridization buffer
(50% formamide, 1 mMEDTA, 400 mMNaCl, 40 mMPIPES, pH 6.4).
After denaturation at 80 °C for 15 min RNA/primer hybrids were al-
lowed to form at room temperature for 1 h. Hybrids were precipitated
with 2.5 volumes of 96% EtOH at ⫺80 °C for 30 min and washed with
100
l of 70% EtOH. Precipitates were dried and resolved in 2
lof5⫻
buffer for Moloney murine leukemia virus reverse transcriptase, 4
lof
5mMdNTPs, 3
lofH
2
O, and 1
l of Moloney murine leukemia virus
reverse transcriptase (MBI Fermentas, 200 units/
l)), and incubated
1hat42°C.1
l of the samples was mixed 1:1 with formamide loading
dye (Amersham Biosciences), applied onto a sequencing gel (4% acryl-
amide, 1⫻TBE (0.089 MTris, 0.089 Mboric acid, 0.002 MEDTA, ph 8.0),
7Murea, 66 cm ⫻0.25 mm) and separated according to Li-Cor (Bad
Homburg, Germany) recommendations. Gene-specific primer se-
quences: psaA,5⬘infrared dye 700, 5⬘-CCC ATT CCT CGA AAG-3⬘
(sequence position ⫹65 to ⫹79 relative to ATG); psbA,5⬘infrared dye
800, 5⬘-AGA CGG TTT TCA GTG-3⬘(sequence position ⫹69 to ⫹83
Photosynthetic Redox Control of Nuclear Gene Expression 5319
by guest on April 9, 2018http://www.jbc.org/Downloaded from
relative to ATG). The same primers were used to sequence the respec-
tive region of Arabidopsis chloroplast DNA using a cycle sequencing kit
(MBI Fermentas, St. Leon-Roth, Germany).
Determination of Thiol Group Content and Redox State of Glutathi-
one—For isolation of total glutathione and cysteine 25 mg of leaf ma-
terial was ground in liquid N
2
and extracted with 0.5 ml of buffer E (100
mMphosphate, pH 7.1, 50% methanol, 5 mMdithiothreitol) for 10 min
at 60 °C while shaking. Homogenates were centrifuged twice at
15,400 ⫻gfor 5 min at room temperature, and supernatants were used
for further analysis. Determination of oxidized glutathione was based
on the same extraction, but dithiothreitol in buffer E was replaced by 5
mMN-ethylmaleimide to block reduced glutathione (42). Reduction of
oxidized thiols in the extracts (0.02 ml) was carried out at room tem-
perature for 60 min in a total volume of 0.27 ml containing 134 mMTris,
pH 8.3, 1 mMdithiothreitol. Then thiols were derivatized for 15 min by
adding 0.03 ml of monobromobimane (Calbiochem, La Jolla) to a final
concentration of 3 mM(2.5-fold excess above total thiol concentration).
Resulting monobromobimane derivatives were stabilized by addition of
0.7 ml of 5% acetic acid and detected fluorometrically (Fluorometer RF
551, Shimadzu) at 480 nm by excitation at 380 nm after separation by
reverse-phase HPLC using a Waters HPLC-system (Waters 600E Mul-
tisolvent Delivery system, Autosampler 717plus) connected to a Nova-
Pak C18 4.6 ⫻250-mm column (pore size, 4
m). Glutathione and other
thiols were separated by applying an isocratic flow (1.3 ml/min) of
buffer A (100 mMpotassium acetate, pH 5.5, 9% methanol) for 12.5 min.
The column matrix was washed with 100% methanol for 3 min and
re-equilibrated for 8.5 min in buffer A. Data acquisition and processing
was performed with Millenium
32
software (Waters). Reduced glutathi-
one concentrations were calculated from the difference between total
and oxidized glutathione. Recovery rates were higher than 95% for
reduced and oxidized glutathione and higher than 90% for cysteine,
respectively, as determined by spiking of samples with internal stand-
ards. Samples were analyzed in quadruplicate.
Expression Profiling—The 3292-GST nylon array, including 2661
nuclear chloroplast genes and 631 genes coding for non-chloroplast
proteins, has been described previously (43). Experiments were per-
formed with plant material corresponding to pools of at least 250 –500
individuals. To obtain larger amounts of tissue of healthy and un-
stressed plants, seedlings were initially grown 22 days under white
light (short day periods, 8-h light/16-h dark) on soil. Plants were then
acclimated to: (i) PSI light (5 days), (ii) PSI light (3 days) followed by
PSII light (2 days), (iii) PSI light (3 days) followed by PSII light plus 5
MDCMU (2 days), or (iv) PSI light plus 5
MDCMU. DCMU (Sigma)
has been applied to plants directly before performing the light shifts
using a fine sprayer as described before (14). DCMU stock solution was
10 mMin 50% ethanol, and the applied concentration was prepared by
dilution in sterile water directly prior use. The drug was found to be
completely stable during the 2-day period of experiment as determined
by the effect on chlorophyll afluorescence using a PAM101 fluorometer.
Effects of DCMU on photosynthetic electron flow have been proven by
determination of ⌽PSII (44) at the end of the treatments (PSI: 0.72 ⫾
0.02; PSI–II: 0.8 ⫾0.02; PSI–II plus DCMU: 0.49 ⫾0.05; PSI plus
DCMU: 0.53 ⫾0.05). Three independent experiments with different
filters and independent cDNA probes were performed thus minimizing
variation between individual plants, filters, or probes. cDNA synthesis
was primed by using a mixture of oligonucleotides matching the 3292
genes in antisense orientation and hybridized to the GST array as
described (43, 45). Images were read using a Storm PhosphorImager
(Amersham Biosciences). Hybridization images were imported into the
ArrayVision program (version 6, Imaging Research Inc., Ontario, Can-
ada), where artifacts were removed, background correction was per-
formed, and resulting values were normalized with reference to inten-
sity of all spots on the array (45). In the next step, those data were
imported into the ArrayStat program (version 1.0 Rev. 2.0, Imaging
Research Inc.), and a z-test (nominal
␣
set to 0.05) was performed
employing false discovery rate (46) correction to identify statistically
significant differential expression values. Only differential expression
values fulfilling the criteria of this statistical procedure were used for
the expression profiling.
RESULTS
Changes in Photosystem Structure of Arabidopsis during Ac-
climation to PSI and PSII light and Transcriptional Regulation
of Plastid Reaction Center Genes psaA and psbA—Imbalances
in excitation energy distribution between the photosystems can
be induced by illumination with light sources that differentially
excite PSII or PSI (PSII or PSI light, respectively) resulting
either in a more reduced or more oxidized state of the electron
transport components (data not shown). To study how plants
deal with and acclimate to such imbalances, Arabidopsis seed-
lings were grown first under white light until the four- to
six-leaf stage before they were subjected to PSI or PSII light
(PSI or PSII plants). Responses of such plants were compared
with responses of plants acclimated to PSI or PSII light fol-
lowed by an additional acclimation to the respective other light
source (PSI–II plants or PSII–I plants). The analysis of PSI and
PSII plants show the acclimation to the two light sources in
general, whereas the analysis of the plants shifted between the
light sources proves the reversibility of the observed responses
(an indicator for true acclimatory effects). To test photosystem
stoichiometry adjustment in response to light quality in Ara-
bidopsis we analyzed photosystem protein abundance and chlo-
rophyll contents. The overall protein pattern of whole tissue
protein extracts did not reveal any major differences between
the four growth conditions when analyzed by SDS gel electro-
phoresis. In Western analyses with antisera raised against the
D1 protein and the P700 apoproteins (representing the core
proteins of PSII and PSI) (Fig. 1A) the D1 protein exhibited
more or less constant amounts under all conditions, whereas
the amounts of P700 apoproteins increased in PSI–II plants in
FIG.1. Changes in photosynthesis protein and chlorophyll
amounts during long term acclimation. A, Western immunological
analysis of photosystem core and antenna protein content. Chloroplasts
of the differentially grown plants were isolated, and proteins of ⬃2⫻
10
5
organelles were separated by SDS-PAGE per lane and transferred
to nylon membranes. Respective growth conditions are given at the
bottom. D1 protein, P700 apoproteins, Lhcb1, and Lhca3 were detected
with polyclonal antisera and a peroxidase-coupled secondary antibody
using enhanced chemiluminescence. Representative results from three
independent experiments are shown. A Coomassie Blue-stained SDS
gel is shown below as loading control. Marker proteins (lane M) range
from 116 to 14 kDa. Growth conditions are given at the bottom.B, Chl
a/bratio. Chlorophylls of acclimated plants were extracted, spectropho-
tometrically determined, and calculated as described under “Experi-
mental Procedures.” Growth conditions are given at the bottom. All
experiments were repeated three times.
Photosynthetic Redox Control of Nuclear Gene Expression5320
by guest on April 9, 2018http://www.jbc.org/Downloaded from
comparison to PSI plants and decreased in PSII–I plants in
comparison to PSII plants. Taking the amount of the reaction
center proteins as an indicator for the relative number of the
photosystems, the PSII/PSI ratio is high under PSI light and
decreases after a shift to PSII light, whereas the opposite can
be observed under PSII light and a shift to PSI light. Further-
more, we tested the abundance of antenna proteins Lhcb1 and
Lhca3, two important components of the PSII and PSI anten-
nae, respectively. Lhca3 showed a similar accumulation under
the light sources as the P700 apoproteins, whereas the opposite
effect was observed for the Lhcb1 protein suggesting a concom-
itant increase in antenna size of the respective rate-limiting
photosystem. Such changes in the antennae are also indicated
by characteristic changes in the Chl a/bratio. After acclimation
to PSI light the Chl a/bratio is low and increases significantly
after a shift to PSII light (Fig. 1B, PSI–II). Under PSII light the
Chl a/bratio is high and decreases after a shift to PSI light.
Because Chl bis mainly associated with the PSII antenna,
these observations are consistent with the observed changes in
the amounts of antenna proteins.
In mustard the adjustment of photosystem stoichiometry is
controlled by changes in the transcription of the reaction center
genes psaA and psbA (9, 12). To test if this is also true for
Arabidopsis we performed primer extension analyses (Fig. 2)
for these genes that allowed us to check for changes in tran-
script initiation sites and amounts of the respective RNAs in
the same experiment. Both psaA and psbA transcripts exhib-
ited the same 5⬘-ends under all conditions investigated, al-
though in varying amounts. For psaA we found two prominent
5⬘-ends in a distance of 197 and 111 bases upstream of the
translation initiation codon. The first (more prominent) end
corresponds to the transcription start sites for psaA in mustard
(47), the second one has not been reported in any other orga-
nism and might represent a species-specific start or processing
site. The regulation of psaA transcript accumulation is compa-
rable to the situation observed for the respective proteins with
an increase in transcripts after a PSI–II light shift (in compar-
ison to PSI light) and a decrease after a PSII–I light shift (in
comparison to PSII light). For psbA we found a single promi-
nent 5⬘-end 78 bases in front of the translation start site,
consistent with earlier reports (27, 47, 48). The accumulation of
this transcript showed only a slight decrease after a PSI–II
light shift and a slight increase after a PSII–I light shift. The
observed changes in RNA amounts are in agreement with the
observations at protein level suggesting that redox-regulated
transcription plays an important role also in Arabidopsis.
Light Quality Acclimation in Photoreceptor and Chloroplast-
to-Nucleus Signaling Mutants—Adjustment of photosystem
stoichiometry in higher plants requires coordinated changes in
the expression of plastid- and nuclear-encoded photosynthesis
genes. To test whether cytosolic photoreceptors or components
of plastid retrograde signaling pathways are involved in the
detection and/or transduction of PSI or PSII light-induced re-
dox signals, we analyzed the LTR in various Arabidopsis mu-
tants (Fig. 3). We used the Chl fluorescence parameter F
s
/F
m
,
which in wild type typically increases after acclimation to PSI
light and decreases after acclimation to PSII light (13, 14)
2
and,
therefore, can be used as a non-invasive indicator for a LTR. In
the photosynthesis mutant hcf109 (49), which exhibits partial
impairment of PSII and PSI activities, no significant changes
in the F
s
/F
m
values could be observed (data not shown) indi-
cating that perturbations of photosynthetic electron transport
lead to a loss of the LTR and/or its detectability. Therefore,
before assessing the F
s
/F
m
value each mutant line was tested
for its Chl fluorescence parameter F
v
/F
m
as indicator for the
general photosynthetic function. All mutants revealed wild
type-like F
v
/F
m
values of ⬎0.8 (data not shown) indicating that
they can perform normal photosynthesis. We then tested the
LTR in mutants lacking functional phytochrome A (phyA),
phytochrome B (phyB), or both (phyA/phyB) (50) as well as for
a transgenic line overexpressing phytochrome B (phyB oe) (51).
In addition, we tested mutants lacking cryptochrome 1 (hy4)or
2(cry2-1) (52, 53). A significant decrease or increase of F
s
/F
m
after the respective light switch was observed for all photore-
ceptor mutants indicating their ability to perform an appropri-
ate LTR (Fig. 3A). Only the phyA/phyB double mutant revealed
no significant decrease of the F
s
/F
m
value after a shift from PSI
to PSII light, whereas the cry2–1 mutant exhibited a signifi-
cant LTR, however, with a less strong increase in F
s
/F
m
than
usually observed after a shift from PSII to PSI light (compare
Supplementary Table SI).
We also analyzed the response of genome-uncoupled (gun)
(54) and cab underexpressed (cue) (55) mutants (Fig. 3B). Both
types of mutants exhibit defects in chloroplast signaling routes
2
R. Wagner and T. Pfannschmidt, unpublished observations.
.
FIG.2. Primer extension analysis of position and accumula-
tion of 5ⴕ- ends of psaA and psbA transcripts. Plants were grown
under the respective light sources, and total RNA was isolated. Fluo-
rescence dye-labeled primers were designed to anneal within the first
50 bp of the coding region of the psaA and psbA genes and were used
both in a reverse transcription reaction with isolated total RNA and a
sequencing reaction of chloroplast DNA fragments covering the psaA
and psbA 5⬘-gene and promoter regions. Products were separated in
parallel on a denaturing 4% acrylamide gel containing 7 Murea and
detected by laser excitation in a Licor 4200 sequencer. A, sequencer
images of the primer extension analyses. The DNA sequences within
the psaA (left part) and psbA (right part) promoters are shown each on
the left, primer extension products on the right. Detected 5⬘-ends are
marked by dots (black for transcript start; white for unknown end), and
respective transcription start nucleotides are given in bold letters.
Growth conditions are given at the bottom.B, structure of the Arabi-
dopsis psaA and psbA promoter regions. Positions of 5⬘-ends are marked
by the same dots as in Fig. 5A. Transcription start sites are indicated by ⫹1,
and all other positions are given relative to it. Pairs of white boxes indicate
⫺10/⫺35 regions; a black box indicates a TATA-like cis-element.
Photosynthetic Redox Control of Nuclear Gene Expression 5321
by guest on April 9, 2018http://www.jbc.org/Downloaded from
toward the nucleus. gun1-1 has still unknown defects, gun5
encodes the H subunit of the magnesium chelatase in the
chloroplast envelope, and gun4 encodes a product that binds
the substrate of the magnesium chelatase (56, 57). The cue1
gene encodes the phosphoenol pyruvate/phosphate translocator
of the chloroplast envelope (58). In our test system all gun
mutants exhibited a wild type-like behavior with significant
LTRs. The cue1-6 mutant, however, showed no significant de-
crease of F
s
/F
m
after a shift from PSI to PSII light, whereas the
expected increase after a shift from PSII to PSI light is present
to a full extent (compare Supplementary Table I).
These data indicate that the defective components in the
photoreceptor and retrograde signaling mutants are not essen-
tial for the LTR, otherwise we would have observed a complete
loss of it. Thus, chloroplast redox signals represent a unique
class of retrograde signals. The less pronounced effects in
phyA/phyB and cue1-6 mutants might be caused by general
developmental effects (see “Discussion”) suggesting that redox
signals are an integral component of the intracellular signaling
network.
Photosynthetic Control of the Nuclear Transcriptome of the
Chloroplast—To analyze the global effects of light quality and
redox signals on the expression of genes for chloroplast pro-
teins, we performed a macroarray analysis using a GST array
with probes covering respective nuclear genes (45). This pre-
selection of genes guarantees that a high proportion of light-
regulated genes are investigated. Light regulation is a pre-
requisite for the study of redox regulation under our conditions.
Furthermore, this array has been shown in earlier studies to
produce statistically reliable and reproducible expression pro-
files (43, 59). To assess the impact of redox signals we followed
a three-step strategy. 1) First we compared gene expression
profiles of PSI and PSI–II plants (Fig. 4, comparison 1). This
showed the overall impact of a reduction signal induced by the
shift from PSI to PSII light. Non-light-regulated genes could be
identified and omitted from further analysis. 2) Next we com-
pared gene expression profiles of PSI plants with PSI–II plants
pre-treated with DCMU (Fig. 4, comparison 2). Genes with the
same expression under both conditions represent either non-
light-regulated genes or light-regulated genes whose expres-
sion change is abolished by the electron transport inhibitor.
FIG.3. LTR in Arabidopsis mutants. Mutant lines were accli-
mated to PSI or PSII light, and F
s
/F
m
values were determined using a
PAM fluorometer. All values were determined in at least three inde-
pendent experiments with 15–20 plants each, and the statistical signif-
icance of differences was proven using the SPSS statistic program (for
details see Supplementary Table SI). The indication of the respective
lines is given in the upper left corner of each graph (for designation see
text). A, photoreceptor mutants; B, chloroplast-to-nucleus signaling
mutants.
FIG.4. Macroarray strategy to define redox-regulated genes
encoding chloroplast proteins. White light-grown plants were accli-
mated to PSI light for 6 days (PSI) or 2 days followed by an acclimation
to PSII light for additional 4 days (PSI–II). Parallel samples were
treated with 5
MDCMU after 2 days in PSI light and then shifted to
PSII light or left under the PSI light. Large circles represent the
respective expression profiles (test condition is given inside). Profiles of
these conditions (large circles) were compared (for details see text).
Intersections represent genes that do not differ significantly in their
expression under the conditions compared. Small circles represent gene
groups resulting from these comparisons (number is given inside), and
their respective origin is indicated by arrows.Gray boxes list category
and number of up- or down-regulated genes in the small circles that
originate from comparisons between expression profiles represented
by large circles (for details, see text). PEF, genes that are regulated by
photosynthetic electron transport. PEF ⫹?, genes that are regulated
by photosynthetic electron transport and an unknown additional redox
signal from the thylakoid membrane.
Photosynthetic Redox Control of Nuclear Gene Expression5322
by guest on April 9, 2018http://www.jbc.org/Downloaded from
The latter are defined as redox-regulated genes and could be
identified by comparing this group of non-regulated genes with
those responsive to the light signal from step 1 (Fig. 4, com-
parison 3). 3) Finally, we compared gene expression profiles of
PSI plants and PSI plants treated with the same amounts of
DCMU as in step 2 (Fig. 4, comparison 4). Redox-regulated
genes whose expression change is completely abolished by the
DCMU treatment were controlled by the photosynthetic elec-
tron flow when the same DCMU treatment as in step 2 had no
effect in step 3 indicating that the DCMU treatment has only
neutralized the PSII light effect. Such “ideal” redox-regulated
genes are defined by a comparison of non-regulated genes from
comparison 4 with the group of redox-regulated genes of com-
parison 3 (Fig. 4, comparison 5).
Comparison 1 indicates that a light quality shift has a mas-
sive impact on the expression of genes encoding chloroplast
proteins. A set of 2133 genes significantly responded to the
shift from PSI to PSII light; 1121 genes were up-regulated
while 1012 were down-regulated. Among these we found genes
for all major functional classes of proteins (Fig. 5), including
genes for photosynthesis, gene expression, metabolism, and
transport. We found no gene class exhibiting unidirectional
expression changes. As a general tendency it emerged that all
gene classes responded in a balanced way with around 50% up-
and 50% down-regulated genes. Out of the 2133 light-regulated
genes, we identified 286 that are directly regulated by redox
signals from the photosynthetic electron transport chain. 86
genes were up-regulated by a reduction signal while 200 genes
were down-regulated by it. From these 286 redox-regulated
genes 54 matched the theoretical constraints for an “ideal”
expression profile to be expected for a gene regulated by redox
signals from photosynthetic electron flow. The remaining 232
genes still represented redox-regulated genes but seemed to be
regulated by more than one redox parameter (see “Discussion”).
Only 76 of the 286 redox-regulated genes encode products
with known functions, including all major gene groups such as
photosynthesis, gene expression, metabolism, or signal trans-
duction (Table I). The great majority of genes, however, codes
for putative, hypothetical, or even unknown proteins (not
shown). Nevertheless, several groups of functionally related
genes can be identified that exhibit similar expression patterns
pointing to concerted regulatory events. The largest groups
among the down-regulated genes include: (a) a large group of
metabolic genes mainly encoding enzymes (or enzyme sub-
units) involved in amino acid or nucleotide metabolism; (b)
several chaperones and signal recognition particle components
partially involved in photosystem assembly; (c) genes for tran-
scription and its regulation in the nucleus and chloroplast; and
(d) genes for components involved in sulfur and glutathione
metabolism. The largest groups among the up-regulated genes
include (a) metabolic genes for amino acid and nucleotide me-
tabolism as well as energy metabolism and (b) photosynthesis
genes. Beside these major groups many individual genes en-
coding products with functions necessary for the establishment
of the LTR are identified (see “Discussion”). This result dem-
onstrates that redox signals from the thylakoid membrane
have an extensive influence on the expression of nuclear genes
reflecting the multiple functional involvements of chloroplasts
within the metabolic pathways of the cell.
Effects of PSI and PSII Light on Thiol Group Content and
Glutathione Redox State—Glutathione is an important cellular
redox buffer and functions also as a potent regulator of gene
expression especially in chloroplasts (16, 17, 22, 60). The array
analysis exhibited several regulated genes involved in gluta-
thione metabolism. To test if changes in glutathione redox state
are involved in the LTR, we determined the content of cysteine
and glutathione as well as the redox state of glutathione in
plants (Table II) grown under the four different conditions. We
found comparable thiol contents under all growth conditions,
and only PSII plants exhibited slightly increased amounts in
glutathione and cysteine. In addition, glutathione appeared to
be mainly reduced (around 90%) under each light regime, as it
is described for Arabidopsis grown under standard white light
sources. Therefore it is unlikely that changes in the glutathione
redox state are responsible for the observed changes in plastid
gene expression. Aside from this, the highly reduced state
indicates that the plants do not suffer from strong reactive
oxygen species-mediated stresses, which is typically indicated
by an increase in oxidized glutathione concentrations. Thus,
superimpositions from reactive oxygen species-induced redox
signaling cascades under the different light qualities are un-
likely, and we conclude that in Arabidopsis light quality
changes are reported mainly via redox signals from intersys-
tem electron transport components.
DISCUSSION
Light Quality Effects on Photosystem Stoichiometry—We
found significant acclimatory changes in the structure of the
photosynthetic apparatus in an extent comparable to those
reported for other higher plants (9, 14, 61, 62). Our Western
analyses, however, suggest that photosystem stoichiometry ad-
justment in Arabidopsis is mainly regulated by changes in PSI
complexes and PSII antenna size. This differs from observa-
tions in pea and mustard were antiparallel changes in both PSI
and PSII were observed (12, 61), whereas it is in accordance
with observations in spinach and cyanobacteria for which
mainly changes in PSI were reported (7, 62). Changes in PSII
content in Arabidopsis have been reported to occur only under
higher light intensities (63). Spectroscopic analyses might help
to determine more precisely the absolute changes in photosys-
tems in Arabidopsis under our conditions. The immunologically
detected changes in D1 and P700 apoprotein levels are accom-
panied by corresponding changes in respective transcript pool
sizes as observed earlier (9, 12). At both promoters redox reg-
ulation occurs at the major transcription start site (Fig. 2),
which is located directly behind typical promoter elements for
FIG.5. Relative distribution of light quality-induced expres-
sion changes of genes sorted by function of gene product. Only
genes with significant expression changes were included. Genes were
grouped according to the known or predicted function of the encoded
product (given at the bottom). Numbers of genes with increased (gray
part of bar) or decreased (white part of bar) expression are given in
percentages.
Photosynthetic Redox Control of Nuclear Gene Expression 5323
by guest on April 9, 2018http://www.jbc.org/Downloaded from
TABLE I
Redox-regulated genes encoding known products
ATG
a
Ratio
b
PSII/PSI
Ratio
b
PSII_DCMU/PSI
Ratio
b
PSI_DCMU/PSI Description
c
Down-regulated
Metabolism
At5g38530 0.63 0.81 0.49 Tryptophan synthase,

chain
At4g16700 0.64 0.85 0.72 Decarboxylase-like protein
At2g43090 0.69 0.82 0.71 3-Isopropylmalate dehydratase, small subunit
At5g13280 0.69 0.88 0.45 Aspartate kinase
At4g27070 0.70 0.96 0.54 Tryptophan synthase,

-subunit (TSB2)
At4g16800 0.71 0.85 0.50 Enoyl-CoA hydratase
At4g19710 0.72 0.81 0.51 Aspartate kinase-homoserine dehydrogenase-like protein
At5g03650 0.72 0.88 0.48 1,4-
␣
-glucan branching enzyme isoform SBE2.2
At3g10050 0.73 0.81 0.57 Threonine dehydratase/deaminase (OMR1)
At4g18440 0.74 0.82 0.57 Adenylosuccinate lyase-like protein
At4g09740 0.74 1.03 0.51 Cellulase-like protein
At5g08300 0.76 0.81 0.65 Succinyl-CoA-ligase,
␣
subunit
At4g11010 0.76 0.82 0.47 Nucleoside diphosphate kinase 3 (ndpk3)
At2g03220 0.79 0.85 0.50 Xyloglucan fucosyltransferase AtFT1
At4g31180 0.76 0.89 0.40 Aspartate-tRNA ligase-like protein
Other
At4g35770 0.68 1.15 2.01 Senescence-associated protein sen1
At5g18810 0.77 1.05 0.99 Serine/arginine-rich protein-like
At5g25380 0.77 0.92 0.62 Cyclin 3a
At5g24020 0.80 0.95 0.76 Septum site-determining MinD
Photosynthesis
At4g15530 0.69 0.80 0.54 Pyruvate, orthophosphate dikinase
At1g76450 0.78 0.86 0.27 Unknown thylakoid lumen protein, PsbP domain
Protein modification
and fate
At2g39990 0.59 0.82 0.50 26 S proteasome regulatory subunit
At5g15450 0.73 0.81 0.70 ClpB heat shock protein-like
At4g36040 0.75 0.90 1.25 DnaJ-like protein
At2g28800 0.76 0.86 0.98 Chloroplast membrane protein ALBINO3 (ALB3)
At5g03940 0.79 0.96 0.51 Signal recognition particle 54CP (SRP54) protein
At4g37910 0.79 0.85 0.51 Hsp70.3
Protein phosphorylation
At4g23650 0.64 0.83 1.19 Calcium-dependent protein kinase (CDPK6)
Stress response
At4g29890 0.62 1.16 0.89 Choline monooxygenase-like protein
At3g45140 0.76 0.82 0.81 Lipoxygenase AtLOX2
Transcription
At1g59940 0.75 0.87 0.90 Response regulator ARR 12
At3g57040 0.79 0.96 0.54 Response regulator ARR 9
RpoB 0.62 1.05 1.05 Plastid gene; RNA polymerase catalytic chain
At5g24120 0.63 0.81 0.67 Sigma-like factor (emb CAA77213.1)
At3g56710 0.73 0.88 0.38 SigA-binding protein
At3g60490 0.73 0.92 0.39 Transcription factor-like protein
At1g68990 0.75 0.99 0.51 DNA-directed RNA polymerase (mitochondrial)
At1g03970 0.79 0.81 0.69 G-box binding factor, GBF4
Transport
At4g33650 0.68 1.03 0.35 Arabidopsis dynamin-like protein ADL2
At4g18290 0.68 0.94 0.51 Potassium channel protein KAT2
At4g36580 0.75 1.04 0.92 ATPase-like protein
At5g59030 0.76 0.90 0.53 Copper transport protein
Unclassified
At2g13870 0.79 1.06 0.65 En/Spm-like transposon protein
S-metabolism
At4g02520 0.57 0.85 0.91 Atpm24.1 glutathione S-transferase
At5g56760 0.66 0.81 0.47 Serine O-acetyltransferase (EC 2.3.1.30) Sat-52
At4g39940 0.67 1.23 0.95 Adenosine-5-phosphosulfate kinase
At5g43780 0.75 0.97 0.72 ATP sulfurylase precursor (gb AAD26634.1)
Up-regulated
Metabolism
At1g29900 1.26 1.18 0.60 Carbamoyl phosphate synthetase, large chain (carB)
At4g24620 1.27 1.17 0.70 Glucose-6-phosphate isomerase
At3g11670 1.28 0.98 0.67 Digalactosyldiacylglycerol synthase
At2g43100 1.34 0.96 0.81 3-Isopropylmalate dehydratase, small subunit
At1g01090 1.41 1.08 0.77 Pyruvate dehydrogenase E1, alpha subunit
At1g24280 1.53 1.03 0.75 Glucose-6-phosphate 1-dehydrogenase
At5g16290 1.74 1.15 0.72 Acetolactate synthase-like protein
At1g76490 1.81 0.86 1.77 Hydroxymethylglutaryl-CoA reductase (AA 1–592)
Photosynthesis
At3g16140 1.26 0.92 0.67 PsaH1
At5g66190 1.27 1.25 0.62 PetH2; FNR; ferredoxin-NADP⫹reductase
At5g66570 1.30 1.17 1.60 PsbO1
At3g08940 1.49 0.92 0.84 Lhcb4.2 (CP29)
At1g79040 1.53 0.91 1.66 PsbR
At1g15820 1.61 0.80 0.70 Lhcb6 (CP24)
Photosynthetic Redox Control of Nuclear Gene Expression5324
by guest on April 9, 2018http://www.jbc.org/Downloaded from
the plastid-encoded RNA polymerase (PEP) (64). This suggests
the existence of specific regulatory protein factors that might
mediate the redox signal to the RNA polymerase. It is inter-
esting to note that components for the PEP complex are found
in the group of redox-regulated genes (see below). Furthermore,
our primer extension studies identified a not yet described
psaA 5⬘-end; however, this does not provide hints on putative
redox-responsive cis-elements because the technique does not
distinguish between transcript initiation and processing. In
vitro DNA-protein interaction studies in spinach suggest that
the psaA promoter may contain additional important regula-
tory elements, i.e. a so-called region D (65). Transcript initia-
tion at this promoter therefore may play a key role during light
quality acclimation in Arabidopsis. Experiments are in prog-
ress to characterize this regulation in more detail.
LTR in Photoreceptor and Chloroplast-to-Nucleus Signaling
Mutants—Our PSI light source contains wavelengths over 700
nm, whereas the PSII light does not, resulting in different
red/far red ratios that might affect the intracellular ratio of the
phytochrome P
r
and P
fr
forms. However, because the LTR is
present in all photoreceptor mutants tested, we conclude that
the acclimatory response operates either independently from or
above the photoreceptor signaling network. The observation
that the LTR is only partially functional in the phyA/phyB
mutant is most probably caused by pleiotropic side effects,
because the double mutant exhibits severe developmental ef-
fects that may interfere with the LTR even if the general
photosynthetic performance does not seem to be disturbed. The
reversibility of the LTR within the single mutants provides a
strong argument that the LTR is regulated without the signal-
ing avenues of phyA or phyB. Both the PSI and the PSII light
do not contain blue or UV-light, which is consistent with the
observation that the LTR is not mediated by cryptochromes.
The observed weaker response in the cry2-1 mutant after a
PSII–I light shift (Fig. 3A) must therefore be caused by a
developmental side effect in this mutant. These data do not
exclude interactions between redox and photoreceptor signal-
ing networks, especially because many more genes are light-
than redox-regulated, however, for the LTR, this appears to be
meaningless. Arabidopsis photoreceptor mutants have also
been used to test the involvement of photoreceptors in photo-
synthetic acclimation responses to high light (66, 67). In these
studies the photoreceptor mutants acclimated to shifts in light
intensity in a wild type-like manner. Although acclimations to
light quality or light quantity involve different responses (1,
63), they all function in the absence of photoreceptors under-
lining the importance of photosynthetic acclimation in the re-
sponse to environmental changes.
In the chloroplast-to-nucleus signaling mutants we also de-
tected clear responses to the PSI and PSII light, indicating that
the LTR operates independently of the lesions in these mu-
tants. Only cue1-6 lacks a significant LTR after a PSI–II light
shift (as phyA/phyB). The lack of the phosphoenol pyruvate
carrier in cue1-6, however, has a strong impact on the energy
metabolism of the mutant, and adult plants exhibit a reticular
phenotype (58). Similar to the phyA/phyB double mutant, these
developmental lesions might affect the LTR. None of the mu-
tant lines investigated here lack the LTR completely except
hcf109, which is the only mutant with defects in photosynthe-
sis. The observation that in phyA/phyB,cue1-6, and cry2-1 only
one response is affected while the other is not could be a hint
that reduction and oxidation signals can be separated and may
operate via different pathways. It is interesting to note that in
cue mutants a connection between phytochrome and plastid
regulation of nuclear gene expression has been observed (68),
although a connection between photosynthetic redox signals
and other plastid retrograde signals or photoreceptors was not
found here.
TABLE II
Thiol group content and redox state of glutathione in differentially
acclimated Arabidopsis seedlings
Growth light
regime Cysteine
a
Glutathione Reduced
glutathione
pmol/mg %
b
PSI 10.7 ⫾0.7 319.5 ⫾33.9 90.0 ⫾2.4
PSI–II 10.8 ⫾2.1 344.6 ⫾65.0 87.3 ⫾3.6
PSII 12.4 ⫾1.2 390.2 ⫾18.8 92.0 ⫾1.5
PSII–I 10.0 ⫾2.2 322.2 ⫾73.1 88.7 ⫾1.7
a
Each value represents the average of four independent samples
based on fresh weight, and S.D. is given.
b
“%” refers to the proportion of reduced glutathione of total glutathi-
one content.
TABLE I—continued
ATG
a
Ratio
b
PSII/PSI
Ratio
b
PSII_DCMU/PSI
Ratio
b
PSI_DCMU/PSI Description
c
At1g31330 1.71 0.91 1.17 PsaF
At4g29670 1.89 0.82 1.58 Thioredoxin-like protein
Protein modification
and fate
At4g20740 1.30 1.11 0.78 Similarity to CRP1
At5g42390 1.31 1.14 0.81 SPP/CPE
Protein phosphorylation
At5g25930 1.47 1.24 1.78 Receptor-like protein kinase-like
Protein synthesis
At4g17300 1.28 1.14 0.83 Asparagine-tRNA ligase
Secondary metabolism
At4g20230 1.46 1.18 1.36 Terpene cyclase-like protein
At5g38120 1.49 1.22 1.49 4-Coumarate-CoA ligase-like protein
At4g32540 1.56 1.22 0.93 Dimethylaniline monooxygenase-like protein
Stress response
At4g11230 1.36 1.12 0.89 Respiratory burst oxidase homolog F-like protein
Transport
At4g36520 1.30 1.15 0.77 Trichohyalin-like protein
At1g80830 1.38 1.23 0.93 Metal ion transporter
S-metabolism
At5g27380 1.26 1.20 0.96 Glutathione synthetase gsh2
a
Accession number.
b
Expression data under the respective test condition relative to the expression data under PSI-light.
c
Those genes among the classified 286 genes that have a clear functional assignment have been listed according to their down- or up-regulation
under PSII light in comparison to PSI light. Genes are grouped into functional categories and listed according to their degree of regulation. Genes
matching conditions of “ideal” redox regulation are given in bold letters.
Photosynthetic Redox Control of Nuclear Gene Expression 5325
by guest on April 9, 2018http://www.jbc.org/Downloaded from
Impact of Light Quality on the Nuclear Chloroplast Tran-
scriptome—The major goal of our array study was to determine
the global impact of light quality and photosynthetic redox
signals on the expression of nuclear genes for chloroplast pro-
teins to assess the importance of such signals for higher plants.
Light quality affects over 2000 genes encoding not only photo-
synthesis but also many other structural and functional com-
ponents. Around 15% of these genes appear to be regulated by
redox signals suggesting that many genes among the 2000 may
be secondary or tertiary targets that are affected through the
long term impact of redox signals on the overall cellular sig-
naling network and/or the action of other light perceiving sys-
tems. Many genes exhibit relatively small changes in their
expression. This can be best explained by the fact that the
expression profiles were determined at the end of the acclima-
tory response when a new expression equilibrium has been
established. Genes transiently affected only for a short time
after a light switch or an inhibitor application might be not
detected by this approach and will be identified by further,
more detailed, studies.
The expression profile after acclimation to a reduction signal
exhibits similar numbers of up- and down-regulated genes (Fig.
5). In a hierarchical cluster analysis of expression profiles in 35
different physiological situations or mutants with this mac-
roarray, our profile was found to be the most prominent repre-
sentative of the so-called class 2 profiles, which are character-
ized by balanced expression changes (43). Class 1 profiles
showed mainly up-regulated and class 3 profiles mainly down-
regulated genes. Among the latter two classes the profiles of
the gun (class 1) and cue mutants (class 3) were found. The
different profile clustering is an independent confirmation that
in these mutants gene expression regulation appears to be
totally different from that observed under our conditions. This
again argues for the independence of light quality-induced
redox signals from the plastid signaling pathways, which are
defective in the gun and cue mutants.
It is difficult to discuss complex results such as transcript
profiles on the level of individual genes, however, the study
uncovered many interesting genes responding to redox signals.
Some of them that are of special interest for the LTR and its
regulation are highlighted in the following. We found several
groups of redox-regulated genes encoding products with related
functions, including those for photosynthesis (Table I, up-reg-
ulated). All affected genes encoding components of the photo-
synthetic machinery were found to be up-regulated by a reduc-
tion signal. A prominent representative is the PsaF gene,
which exhibits essentially the same expression profile as ob-
tained earlier with transgenic tobacco lines containing a
PsaF-promoter::uidA construct (14), demonstrating the repro-
ducibility of the expression data. We also found a thioredoxin-
like protein that is of interest because thioredoxins regulate
many processes in chloroplasts such as light induction of Cal-
vin cycle enzymes or translation initiation of psbA (69). In
general, up-regulation occurred for both PSII and PSI genes
suggesting that the stoichiometric adjustment of the nuclear
encoded components includes additional regulatory steps at
other levels of expression and/or complex assembly (see below).
This might also be the reason why we did not find all nuclear
photosynthesis genes to be regulated in this array.
Metabolic genes represent the most prominent group among
the redox-regulated genes identified here. Most encode compo-
nents involved in amino acid and nucleotide metabolism and
are regulated in the opposite way to photosynthesis genes.
Amino acids and nucleotides are central molecules in many
biosynthetic pathways demonstrating that the acclimation re-
sponse is not restricted to photosynthesis but has also a deep
impact on the metabolism of a plant. A metabolic gene of
special interest here is the succinyl-CoA-ligase, which produces
the precursor molecule for aminolevulinic acid, the entry sub-
stance for chlorophyll biosynthesis, a process that is clearly
affected during the LTR (Fig. 1C). In addition, we found the
digalactosyldiacylglycerol synthase, which produces the major
lipid of thylakoid membranes (70). The LTR involves major
re-arrangements of the thylakoid membrane system in chloro-
plasts.
2
Because of these results we have started further stud-
ies to investigate the LTR effects on plant metabolism in more
detail.
A further striking observation is the regulatory impact on
components of the chloroplast PEP enzyme (rpoB, sigma-like
factor, SigA binding factor; Table I, transcription, down-regu-
lated), which is responsible for the redox regulation at the psbA
and psaAB promoters (Fig. 2). The rpoB gene is plastid-local-
ized, encodes the catalytic

-subunit of PEP, and is transcribed
by the nuclear encoded RNA polymerase (64). This suggests a
redox regulation of nuclear encoded RNA polymerase activity.
Interestingly the paralogue nuclear encoded RNA polymerase
gene, which encodes the mitochondrial nuclear encoded RNA
polymerase, appears to be redox-regulated in its expression
like ARR9, ARR12, and GBF4, transcriptional regulators of
nuclear transcription, as well as the sigma-like-factor and SigA
binding factor, transcriptional regulators of chloroplast tran-
scription. This suggests a complex signaling network control-
ling in parallel the expression of the different components of
the plastid gene transcription machinery in the nucleus and in
the organelle. Furthermore, regulation of the PEP enzyme has
been shown in vitro to be under phosphorylation control via the
plastid transcription kinase, which itself is under control of
glutathione redox state (16). Our results do not indicate major
differences in the glutathione redox state under the various
light conditions thus supporting the idea of several different
redox control pathways in chloroplast transcription (60, 71),
depending on environmental conditions as to be expected for
different acclimation responses under low or high light (see
above).
The photosystems are multiprotein complexes, which have to
be assembled in a highly coordinated manner. Several chaper-
ones and assembly proteins were identified as being redox-
regulated (Table I, up-regulated, protein modification, and
fate). Important in this context are ALB3 and SRP54, two
proteins of the SRP complex in thylakoid membranes that are
responsible for the import of light-harvesting proteins into the
thylakoid membrane (72, 73).
Of special interest is the observation that several genes for
enzymes involved in sulfur and glutathione metabolism (Table
I, S-metabolism) together with genes for products involved in
various stress responses (choline monooxygenase, lypoxygen-
ase, and respiratory burst oxygenase) are found to be redox-
regulated. It is possible that these changes in glutathione genes
are responsible for the relatively stable glutathione redox state
found here. It is also possible that these genes are regulated
together with stress genes (see above) in a kind of overlap
reaction between photosynthetic redox signals and other envi-
ronmentally induced stress signals such as cold (indicated by
choline monooxygenase (74)) or pathogen attack (indicated by
lypoxygenase, respiratory burst oxygenase (75)), which are also
mediated by redox signals. It is well known that interactions
between photosynthesis, temperature, or pathogen attack exist
and that redox signals of various origin play a central role in
this scenario. Antioxidant molecules such as glutathione are
involved in all these processes indicating that multiple connec-
tions between the responses to the different environmental
stress situations exist (76). The dominant regulatory signals
Photosynthetic Redox Control of Nuclear Gene Expression5326
by guest on April 9, 2018http://www.jbc.org/Downloaded from
controlling the LTR appear to come from the electron transport
chain, because the glutathione redox state remained relatively
stable under all conditions. Because the redox state of ascor-
bate is tightly coupled to that of glutathione (31), we expect
that the antioxidant network remains in homeostasis during
the LTR, which makes it very unlikely that in our light quality
system reactive oxygen species play a significant regulatory
role.
To our knowledge this is the first report describing the ef-
fects of DCMU on gene expression in a higher plant using an
array approach. A similar study has been performed so far only
with a whole genome array of Synechocystis (77). By the use of
DCMU and 2,5-dibromo-3-methyl-6-isopropyl-p-benzoquinone,
140 genes have been reported to be affected by redox signals
from the electron transport chain, which is in the same order of
magnitude as in our experiment. However, Hihara et al. (65)
concluded that the redox regulation of photosynthesis genes in
Synechocystis might be totally different from that in algae and
plants. A gene-by-gene comparison between both studies does
not provide much useful information, even if we consider that
in our array the eukaryotic complement of the cyanobacterial
genome is present, because the physiological conditions used in
both studies are very different.
DCMU also affected photosynthetic electron flow in plants
grown continuously under PSI light indicating that these
plants perform linear electron transport. The expression profile
of these plants, however, is different from that of PSI–II plants
treated with DCMU suggesting that possibly more redox-reg-
ulated genes exist than described here. The combined action of
DCMU and PSI light on photosynthetic electron flow is difficult
to understand to date and requires further detailed analyses;
therefore, we described only those genes as redox-regulated
that allow us to conclude unambiguously on such a regulation.
Data from different studies suggest the existence of several yet
unknown redox signals originating from the electron transport
chain, including PSII (30, 78, 79). Furthermore, any change in
linear electron flow will affect the redox state of components
downstream of PSI such as thioredoxin, which in turn will
affect the efficiency of the Calvin cycle (69). Whether such
signals influence gene expression events in our experimental
system is currently under investigation.
Our study indicates that photosynthetic redox signals play
an important role in the intracellular signaling network. The
photosynthetic redox signals contribute essential information
about the light environment in addition to cytosolic photore-
ceptors thus significantly expanding the ability of plants to
sense environmental cues. It appears that this information is
transferred from the organelle to the nucleus by mechanisms
that differ from other chloroplast-to-nucleus signaling avenues
and without the help of photoreceptor-mediated signaling.
Acknowledgments—We thank Meta Brost for skillful technical as-
sistance and Karen Ko¨hler for help with analysis of variance. Arabi-
dopsis seeds were obtained from the Nottingham Arabidopsis Stock
Center or were kindly provided by U.-I. Flu¨ gge and B. Grimm. The P700
antiserum was a generous gift from M. Hippler.
REFERENCES
1. Walters, R. G., and Horton, P. (1994) Planta 195, 248 –256
2. Anderson, J. M., Chow, W. S., and Park, Y. I. (1995) Photosynth. Res. 46,
129 –139
3. Aro, E. M., and Andersson, B. (2001) Regulation of Photosynthesis: Advances in
Photosynthesis and Respiration, Kluwer Academic Publishers, Dordrecht
4. Haldrup, A., Jensen, P. E., Lunde, C., and Scheller, H. V. (2001) Trends Plant
Sci. 6, 301–305
5. Allen, J. F., and Forsberg, J. (2001) Trends Plant Sci. 6, 317–326
6. Allen, J. F., and Pfannschmidt, T. (2000) Philos. Trans. R Soc. Lond. B Biol.
Sci. 355, 1351–1359
7. Fujita, Y. (1997) Photosynth. Res. 53, 83–93
8. Melis, A. (1991) Biochim. Biophys. Acta 1058, 87–106
9. Pfannschmidt, T., Nilsson, A., and Allen, J. F. (1999) Nature 397, 625– 628
10. Allen, J. F., Bennett, J., Steinback, K. E., and Arntzen, C. J. (1981) Nature 291,
25–29
11. Rintamaki, E., Martinsuo, P., Pursiheimo, S., and Aro, E. M. (2000) Proc. Natl.
Acad. Sci. U. S. A. 97, 11644–11649
12. Pfannschmidt, T., Nilsson, A., Tullberg, A., Link, G., and Allen, J. F. (1999)
IUBMB Life 48, 271–276
13. Sherameti, I., Sopory, S. K., Trebicka, A., Pfannschmidt, T., and Oelmuller, R.
(2002) J. Biol. Chem. 277, 46594 – 46600
14. Pfannschmidt, T., Schutze, K., Brost, M., and Oelmuller, R. (2001) J. Biol.
Chem. 276, 36125–36130
15. Tullberg, A., Alexciev, K., Pfannschmidt, T., and Allen, J. F. (2000) Plant Cell
Physiol. 41, 1045–1054
16. Baginsky, S., Tiller, K., Pfannschmidt, T., and Link, G. (1999) Plant Mol. Biol.
39, 1013–1023
17. Karpinski, S., Escobar, C., Karpinska, B., Creissen, G., and Mullineaux, P. M.
(1997) Plant Cell 9, 627– 640
18. Karpinski, S., Reynolds, H., Karpinska, B., Wingsle, G., Creissen, G., and
Mullineaux, P. (1999) Science 284, 654 – 657
19. Oswald, O., Martin, T., Dominy, P. J., and Graham, I. A. (2001) Proc. Natl.
Acad. Sci. U. S. A. 98, 2047–2052
20. Petracek, M. E., Dickey, L. F., Nguyen, T. T., Gatz, C., Sowinski, D. A., Allen,
G. C., and Thompson, W. F. (1998) Proc. Natl. Acad. Sci. U. S. A. 95,
9009 –9013
21. Eguchi, S., Takano, H., Ono, K., and Takio, S. (2002) Plant Cell Physiol. 43,
573–577
22. Liere, K., and Link, G. (1997) Nucleic Acids Res. 25, 2403–2408
23. Alexciev, K., and Tullberg, A. (1997) Physiol. Plant 99, 477– 485
24. Petracek, M. E., Dickey, L. F., Huber, S. C., and Thompson, W. F. (1997) Plant
Cell 9, 2291–2300
25. Sherameti, I., Nakamura, M., Yamamoto, Y. Y., Pfannschmidt, T., Obokata, J.,
and Oelmuller, R. (2002) Plant Journal 32, 631– 639
26. Tang, L., Bhat, S., and Petracek, M. E. (2003) Plant Physiol. 133, 1979 –1990
27. Shen, Y. X., Danon, A., and Christopher, D. A. (2001) Plant Cell Physiol. 42,
1071–1078
28. Yang, D. H., Andersson, B., Aro, E. M., and Ohad, I. (2001) Photosynth. Res. 68,
163–174
29. Pfannschmidt, T., Allen, J. F., and Oelmuller, R. (2001) Physiol. Plant 112, 1–9
30. Dietz, K.-J., Link, G., Pistorius, E. K., and Scheibe, R. (2002) Prog. Bot. 63,
207–245
31. Baier, M., and Dietz, K. J. (1998) Prog. Bot. 60, 283–314
32. Pfannschmidt, T., Schutze, K., Fey, V., Sherameti, I., and Oelmuller, R. (2003)
Antioxid. Redox. Signal 5, 95–101
33. Rodermel, S. (2001) Trends Plant Sci. 6, 471– 478
34. Surpin, M., Larkin, R. M., and Chory, J. (2002) Plant Cell 14, S327–S338
35. Gray, J. C., Sullivan, J. A., Wang, J. H., Jerome, C. A., and MacLean, D. (2003)
Philos. Trans. R Soc. Lond. B Biol. Sci. 358, 135–144
36. van Kooten, O., and Snel, J. F. H. (1990) Photosynth. Res. 25, 147–150
37. Porra, R. J., Thompson, W. A., and Kriedemann, P. E. (1989) Biochim. Biophys.
Acta 975, 384 –394
38. Laemmli, U. K. (1970) Nature 277, 680 – 685
39. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A
Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring
Harbor, NY
40. Chomczynski, P., and Sacchi, N. (1987) Anal. Biochem. 162, 156 –159
41. Ghosh, P. K., Reddy, V. B., Swinscoe, J., Choudary, P. V., Lebowitz, P., and
Weissman, S. M. (1978) J. Biol. Chem. 253, 3643–3647
42. Fahey, R. C., and Newton, G. L. (1987) Methods Enzymol. 143, 85–96
43. Richly, E., Dietzmann, A., Biehl, A., Kurth, J., Laloi, C., Apel, K., Salamini, F.,
and Leister, D. (2003) EMBO Rep. 4, 491– 498
44. Genty, B., Briantais, J.-M., and Baker, N. R. (1989) Biochim. Biophys. Acta
990, 87–92
45. Kurth, J., Varotto, C., Pesaresi, P., Biehl, A., Richly, E., Salamini, F., and
Leister, D. (2002) Planta 215, 101–109
46. Benjamini, Y., and Hochberg, Y. (1995) J. Roy. Stat. Soc. Series B Method 57,
289 –300
47. Summer, H., Pfannschmidt, T., and Link, G. (2000) Curr. Genet. 37, 45–52
48. Liere, K., Kestermann, M., Muller, U., and Link, G. (1995) Curr. Genet. 28,
128 –130
49. Meurer, J., Berger, A., and Westhoff, P. (1996) Plant Cell 8, 1193–1207
50. Reed, J. W., Nagatani, A., Elich, T. D., Fangan, M., and Chory, J. (1994) Plant
Physiol. 104, 1139 –1149
51. Aukerman, M. J., Hirschfeld, M., Wester, L., Weaver, M., Clack, T., Amasino,
R. M., and Sharrock, R. A. (1997) Plant Cell 9, 1317–1326
52. Ahmad, M., and Cashmore, A. R. (1993) Nature 366, 162–166
53. Guo, H., Yang, H., Mockler, T. C., and Lin, C. (1998) Science 279, 1360 –1363
54. Susek, R. E., Ausubel, F. M., and Chory, J. (1993) Cell 74, 787–799
55. Li, H.-M., Culligan, K., Dixon, R. A., and Chory, J. (1995) Plant Cell 7,
1599 –1610
56. Mochizuki, N., Brusslan, J. A., Larkin, R., Nagatani, A., and Chory, J. (2001)
Proc. Natl. Acad. Sci. U. S. A. 98, 2053–2058
57. Larkin, R. M., Alonso, J. M., Ecker, J. R., and Chory, J. (2003) Science 299,
902–906
58. Streatfield, S. J., Weber, A., Kinsman, E. A., Hausler, R. E., Li, J., Post-
Beittenmiller, D., Kaiser, W. M., Pyke, K. A., Flugge, U. I., and Chory, J.
(1999) Plant Cell 11, 1609 –1622
59. Kubis, S., Baldwin, A., Patel, R., Razzaq, A., Dupree, P., Lilley, K., Kurth, J.,
Leister, D., and Jarvis, P. (2003) Plant Cell 15, 1859 –1871
60. Link, G. (2003) Antioxid. Redox. Signal 5, 79–87
61. Glick, R. E., McCauley, S. W., Gruissem, W., and Melis, A. (1986) Proc. Natl.
Acad. Sci. U. S. A. 83, 4287–4291
62. Deng, X. W., Tonkyn, J. C., Peter, G. F., Thornber, J. P., and Gruissem, W.
(1989) Plant Cell 1, 645– 654
63. Bailey, S., Walters, R. G., Jansson, S., and Horton, P. (2001) Planta 213,
794 – 801
Photosynthetic Redox Control of Nuclear Gene Expression 5327
by guest on April 9, 2018http://www.jbc.org/Downloaded from
64. Liere, K., and Maliga, P. (2001) in Regulation of Photosynthesis (Aro, E. M.,
and Andersson, B., eds) Vol. 11, pp. 29 –49, Kluwer Academic Publishers,
Dordrecht
65. Cheng, M. C., Wu, S. P., Chen, L. F., and Chen, S. C. (1997) Planta 203,
373–380
66. Walters, R. G., Rogers, J. J. M., Shephard, F., and Horton, P. (1999) Planta
209, 517–527
67. Weston, E., Thorogood, K., Vinti, G., and Lopez-Juez, E. (2000) Planta 211,
807– 815
68. Lopez-Juez, E., Jarvis, R. P., Takeuchi, A., Page, A. M., and Chory, J. (1998)
Plant Physiol. 118, 803– 815
69. Schurmann, P., and Jacquot, J. P. (2000) Ann. Rev. Plant Physiol. Plant Mol.
Biol. 51, 371– 400
70. Kelly, A. A., and Dormann, P. (2004) Curr. Opin. Plant Biol. 7, 262–269
71. Pfannschmidt, T., and Liere, K. (2004) Antioxid. Redox. Signal, in press
72. Woolhead, C. A., Thompson, S. J., Moore, M., Tissier, C., Mant, A., Rodger, A.,
Henry, R., and Robinson, C. (2001) J. Biol. Chem. 276, 40841– 40846
73. Schunemann, D. (2004) Curr. Genet. 44, 295–304
74. Iba, K. (2002) Annu. Rev. Plant Biol. 53, 225–245
75. Schaller, F. (2001) J. Exp. Bot. 52, 11–23
76. Foyer, C. H., LopezDelgado, H., Dat, J. F., and Scott, I. M. (1997) Physiol. Plant
100, 241–254
77. Hihara, Y., Sonoike, K., Kanehisa, M., and Ikeuchi, M. (2003) J. Bacteriol. 185,
1719 –1725
78. Krieger-Liszkay, A., and Rutherford, A. W. (1998) Biochemistry 37,
17339 –17344
79. Sane, P. V., Ivanov, A. G., Hurry, V., Huner, N. P., and Oquist, G. (2003) Plant
Physiol. 132, 2144 –2151
Photosynthetic Redox Control of Nuclear Gene Expression5328
by guest on April 9, 2018http://www.jbc.org/Downloaded from
Dietzmann, Dario Leister, Ralf Oelmüller and Thomas Pfannschmidt
Vidal Fey, Raik Wagner, Katharina Braütigam, Markus Wirtz, Rüdiger Hell, Angela
Arabidopsis thalianaChloroplast Proteins of
Retrograde Plastid Redox Signals in the Expression of Nuclear Genes for
doi: 10.1074/jbc.M406358200 originally published online November 23, 2004
2005, 280:5318-5328.J. Biol. Chem.
10.1074/jbc.M406358200Access the most updated version of this article at doi:
Alerts:
When a correction for this article is posted• When this article is cited•
to choose from all of JBC's e-mail alertsClick here
Supplemental material:
http://www.jbc.org/content/suppl/2004/12/10/M406358200.DC1
http://www.jbc.org/content/280/7/5318.full.html#ref-list-1
This article cites 76 references, 29 of which can be accessed free at
by guest on April 9, 2018http://www.jbc.org/Downloaded from
Additions and Corrections
Vol. 280 (2005) 5318 –5328
Retrograde plastid redox signals in the expression of nuclear genes for chloroplast proteins of Arabidopsis
thaliana.
Vidal Fey, Raik Wagner, Katharina Bra¨ utigam, Markus Wirtz, Ru¨ diger Hell, Angela Dietzmann, Dario Leister, Ralf Oelmu¨ller,
and Thomas Pfannschmidt
Page 5321, Fig. 2: The positions of the psaA transcripts given in Fig. 2 are incorrect. The correct figure is shown below. As a
result, a sentence appearing in the left column, lines 29 –31 of the same page, should read as follows: “For psaA we found two
prominent 5⬘-ends in a distance of 189 and 76 bases upstream of the translation initiation codon.”
FIG.2
THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 280, No. 17, Issue of April 29, p. 17572, 2005
© 2005 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.
We suggest that subscribers photocopy these corrections and insert the photocopies at the appropriate places where the article to be
corrected originally appeared. Authors are urged to introduce these corrections into any reprints they distribute. Secondary (abstract)
services are urged to carry notice of these corrections as prominently as they carried the original abstracts.
17572
